DEPENDENCE | DISTANCE | DISPERSION OPTIONS FOR UPGRADING KAZAKHSTAN’S ECONOMY KAZAKHSTAN COUNTRY ECONOMIC MEMORANDUM May 2022 World Bank Kazakhstan Country Economic Memorandum i Contents Acknowledgments ............................................................................................................................... iv Executive summary .............................................................................................................................. v Motivation ............................................................................................................................................................................... v Reversing dependence, distance, and dispersion .................................................................................................. vii Three priorities: Productivity, mobility, and connectivity ...................................................................................viii From oil to cities: Structural and spatial transformation ..................................................................................... xii 1 Restarting Kazakhstan’s stalled structural and spatial transformations...................................... 2 Productivity drives growth and transforms economic structures ...................................................................... 3 A limping private sector ..................................................................................................................................................... 9 Agglomeration for productivity ................................................................................................................................... 14 Conclusion ............................................................................................................................................................................ 17 References ............................................................................................................................................................................ 18 2 Migration and domestic mobility .................................................................................................. 19 Migration and structural transformation.................................................................................................................. 19 Past migration trends and path dependency ......................................................................................................... 22 Recent migration patterns ............................................................................................................................................. 25 Housing and housing costs ........................................................................................................................................... 28 Modern residency registration ..................................................................................................................................... 31 Housing policy reform ..................................................................................................................................................... 38 References ............................................................................................................................................................................ 41 Annex 2A Migration statistics ....................................................................................................................................... 44 Annex 2B Registration procedure ............................................................................................................................... 45 Annex 2C The legal basis of residency registration ............................................................................................. 46 3 Spatial dimensions of intraregional and interregional trade in Kazakhstan ............................. 47 The dynamics and cross-sectional variation of intranational trade ............................................................... 47 Intranational trade and regional development ...................................................................................................... 51 Conclusion ............................................................................................................................................................................ 60 References ............................................................................................................................................................................ 61 Annex 3A Calculating the market access index ..................................................................................................... 64 Annex 3B Distance as a source of friction for intraregional and interregional trade.............................. 65 Annex 3D Testing for the impact of connectivity and firm level performance ......................................... 67 Kazakhstan Country Economic Memorandum ii Annex 3E Testing for market integration and its spatial dimension.............................................................. 69 Annex 3F Short-run movement of regional prices ............................................................................................... 70 4 Assessing place-based investments ............................................................................................... 72 Principles for assessing place-based policies and projects ............................................................................... 72 References ............................................................................................................................................................................ 80 5 Agglomeration patterns ................................................................................................................. 81 Reshaping Kazakhstan’s economic geography...................................................................................................... 82 Breaking with the territorial development vision of the past........................................................................... 90 Building the right foundations for territorial development with structural reforms, including greater labor mobility ................................................................................................................................................................... 94 Empowering subnational governments is crucial for multiple aspects of territorial development .. 96 Moving beyond a top-down approach and supporting integrated, multisectoral local economic development programs implemented by subnational authorities ............................................................. 98 Areas and recommendations for improving territorial development policy............................................ 102 References .......................................................................................................................................................................... 103 Annex 5A Territorial development in Canada ...................................................................................................... 106 Annex 5B The Greater Manchester Combined Authority: Coordinated metropolitan management ............................................................................................................................................................................................. 108 Annex 5C Reforming the transport system in Moscow and Paris ................................................................ 110 6 Special economic zones, industrial zones, and territorial clusters ...........................................112 The zones and clusters .................................................................................................................................................. 113 Kazakh stakeholders’ views on challenges in existing zone and cluster policies ................................... 122 Learning from international experience: The importance of tailoring zones to local context .......... 125 Conclusion and policy recommendations.............................................................................................................. 126 References .......................................................................................................................................................................... 128 Annex 6A Pilot territorial clusters .............................................................................................................................. 130 Annex 6B Special economic zones ........................................................................................................................... 133 Annex 6C Industrial zones............................................................................................................................................ 135 Annex 6D History of cluster policy in Kazakhstan .............................................................................................. 137 Kazakhstan Country Economic Memorandum iii Acknowledgments This country economic memorandum was prepared by a team led by Somik Lall (lead economist EFI chief economist office and former World Bank global lead on territorial development) and Ivailo Izvorski (practice manager for the Global Macro, Growth, and Debt Unit of the World Bank and former lead economist on Kazakhstan). The team included Paulo Correa (program leader), Will Seitz (senior poverty economist), Sjamsu Rahardja (senior economist on Kazakhstan), Mariana Lootty (senior private sector specialist), Dmitry Sivaev (urban specialist, Europe and Central Asia), Elwyn Davies (economist) and Olena Bogdan (economist). The team received substantial contributions from Alena Sakhonchik (private sector specialist), Kenan Karakulah, and Tleugazy Bespaliev. The team was also capably assisted by Artem Gebelev, Asset Bizhan, Azamat Agaidarov (economist), Douglas Zhihua Zeng (senior economist), Kassymkhan Kasparov, Mari Chichagova, Omoniyi (Niyi) Alimi, Shawn Tan (senior economist), and Zarina Adilkhanova. The work has benefited from guidance and insightful comments from Indermit Gill (vice president EFI), Sandeep Mahajan (practice manager Europe and Central Asia), and Shigeo Katsu (president, Nazarbayev University). The team acknowledges the excellent cooperation with the Ministry of National Economy during the preparation of this work, led by Alibek Kuantyrov (minister), Bauyrzhan Omarbekov (vice minister), Zhaslan Madiev (vice minister of telecommunications and former vice minister of economy when the project was conceived), and Alisher Abdykadyrov (deputy akim of Almaty and former vice minister of national economy). Gaziz Selikhanov (deputy director of the Economic Research Institute at the Ministry of National Economy) was instrumental in helping the team process the data on which the analysis of firm-level productivity in Chapter 1 was based. The team is also grateful to representatives of the following stakeholders for valuable data and insight on the work of special economic zones, industrial zones, and territorial clusters: JSC QazIndustry (under the Ministry of Investment and Infrastructural Development), the managing companies of SEZ Ontustik, SEZ Pavlodar, SEZ Chemical Park Taraz, and the Almaty Industrial Zone, as well as the Association of Construction Companies as part of the construction materials cluster in Qaragandy Oblast, and participants in some of the SEZs. Valuable comments were received during presentations to the Ministry of National Economy, the Nazarbayev University, and the Academy of Public Administration. We are grateful to Bruce Ross-Larson and his team for the superb editorial service, and to Sarah Babirye and Gulmira Akshatyrova for their support to the team. Cover image courtesy of Kazangapov and Ospan Ali. Kazakhstan Country Economic Memorandum iv Executive summary Motivation In principle, helping fewer than 20 million people get to high income and sustain it should be easy for a nation with Kazakhstan’s natural resources. It is the ninth largest country in land mass, with 2.7 million square kilometers; 14th largest in arable land, with more than 22 million hectares; 12th largest in oil wealth, with 30 million barrels of proven reserves; and it has the 15th largest gas reserves, with more than 2,400 billion cubic meters. It is rich in minerals both common, such as coal, and rare, such as uranium: it has the 10th largest coal deposits, which add up to more than 25 billion tons; and the second largest uranium reserves, with deposits of nearly a million tons. But despite generally favorable global conditions during the past three decades, becoming an advanced economy has proved to be difficult for middle-income countries. Middle-income economies are not all the same, of course, and Kazakhstan is different from Malaysia, Mexico, and Romania. It is different also from the world’s high-income economies. What sets Kazakhstan apart from the advanced economies in the Organisation for Economic Co-operation and Development is the coincidence of resource dependence, distant location, and dispersed population. Because of these special circumstances, Kazakhstan will have to devise its own strategy to succeed. Kazakhstan’s economy depends heavily on extractives, with many other industries linked to them. Oil and gas contribute more than 20 percent to GDP, more than half of it from extraction and more than a third from the services linked to the sector. Mining contributes another 15 percent of GDP. Commodity revenues account for nearly half of overall government revenues. Commodity exports, meanwhile, account for more than 80 percent of total shipments abroad. Manufacturing’s contribution to GDP has been static since 2010, accounting for around 11 percent of K azakhstan’s GDP in 2018. Between 2015 and 2018, the fastest annual growth was in resource-heavy manufacturing (metals, materials, petroleum, and pharmaceuticals) followed by the textiles and apparel industry. Perhaps the most telling indicator of a stalled structural transformation is the stagnant spatial organization of the economy spread on such a huge territory. One of the most durable findings in economic development is a tight three-way relationship between a country’s per capita income, the sectoral composition of its output, and the share of urban settlements in its population (Grover, Lall, and Maloney 2022). It appears that during the past decade and a half, Kazakhstan’s spatial transformation has stalled. The urban share of the population—measured using Kazakhstan’s own definitions—has essentially remained unchanged (Figure 1). Indeed, it has not changed much since 2000. Kazakhstan Country Economic Memorandum v Figure 1 A boring chart—urban share of population in Kazakhstan, 2006–20 60.00 50.00 40.00 30.00 20.00 10.00 0.00 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 And because urbanization has stalled, Kazakhstan suffers from having the bulk of its population dispersed in distant locations. On average, population density is six people per square kilometer across this large country—making it challenging to provide basic services and infrastructure. Low densities also stymie economies of agglomeration and specialization. Mobility toward urban areas is dampened by high housing costs, a severely limited rental market, and the residency registration system.1 In the highest demand urban areas, housing is extremely unaffordable when compared with local incomes. Homeowners in urban areas would typically not be able to afford their housing if they did not already own it. At the same time, housing subsidies and legacy policies contribute to one of the highest home ownership rates in the world (Figure 2). Especially for rural areas where market demand for housing is low, these features dampen regional mobility, ultimately leading to labor market dysfunction and lower productivity. Finally, the residency registration system deters many people—particularly low-income people receiving financial support from the state—from relocating within the country, as benefits are tied to where a person is registered. For those who are unregistered, the system can lead to exclusion from many social services and benefits. 1 The high influx of temporary migrants from Russia after the war in Ukraine has further increased rental prices of apartments and housing in Kazakhstan’s major cities. Kazakhstan Country Economic Memorandum vi Figure 2 Home ownership rate, by country, 2015 100 90 80 70 60 50 40 30 Austria France Iceland Latvia Serbia Denmark Finland Italy Norway Romania Netherlands Ireland Cyprus Greece Poland Hungary Russian Fed. Germany Luxembourg Slovenia Spain Bulgaria Croatia Belgium Malta Estonia Slovakia Lithuania N. Macedonia Switzerland Sweden European Union Portugal Kazakhstan Czechia United Kingdom Sources: EuroStat, RusStat, and the Household Budget Survey of Kazakhstan. Reversing dependence, distance, and dispersion Kazakhstan’s challenges can be traced to three interrelated economic attributes: the dependence on oil and gas revenues instead of taxes and contributions from individuals and enterprises, large economic distances across regions within the country, and its dispersal of people and economic activity. Productivity growth was strong in the 1990s and the early 2000s but has since fallen dramatically. One reason for the decline is the sluggishness of structural transformation. On the structural side, the economy needs to complete the shift from plan to market and accelerate the transformation from oil dependence to a more diversified economy, and permit for much more economic freedom for companies and individuals. On the spatial side, existing structure reflects the vestiges of the planned economy: a lot of production is still in places where it would not be under a market system. It would be near consumers, other firms, or near the borders. And since the 1990s, the spatial transformation, though advancing, has been slow. If unhindered, following markets, new and existing firms will choose to locate in a particular place if it has a good business environment, which requires good local governance, amenities, and infrastructure and place-based government policies. Low transport and trade costs facilitated by good infrastructure will help firms source inputs and produce their outputs more cost effectively. Free internal movement of workers will ensure that firms can locate in a region of their choice and still have access to workers from other parts of the country. So, the spatial structure of the economy is rapidly becoming more important for Kazakhstan’s future development. The Kazakhstan’s national development plan offers a new model of economic growth featuring diversification, reduced dependence on natural resources, and a transition toward productivity growth driven by technology, innovation, and human capital accumulation. At the same time, Kazakhstan wants to maintain territorial cohesion and ensure that growth benefits all regions. The government has improved its territorial development policy but needs to further adjust its goals. The creation of new regions (Abai, Ulytau, and Zhetisu) may help improve linkages to the region’s centers and public service delivery in the districts However, territorial cohesion requires complementing spatially uneven economic growth with reforms and investments to ensure access to Kazakhstan Country Economic Memorandum vii opportunity, no matter where businesses and people are, and targeted measures to address market failures that keep places from reaching their full economic potential. Three priorities: Productivity, mobility, and connectivity Productivity: Stalled Kazakhstan’s ability to improve its standard of living over time depends almost entirely on its ability to raise its productivity.1 It experienced strong productivity growth in the early 2000s, with a contribution to annual GDP growth of 6 percentage points on average. This period of strong productivity growth was followed by productivity declines in 2008–09, a productivity growth resumption (at a much lower level) in 2010–13, followed by declining productivity in 2015–16 with the main driver of GDP growth being the capital base. Together with a need to increase investment in human capital, a concerted effort to invest in productivity growth is crucial to foster economic growth and create better job opportunities. Productivity growth can be the result of three processes, usually concurrent: 1. Improving firm capabilities: Firms can adopt new technologies and better managerial practices (within-firm productivity growth). 2. Improving factor allocation: Productivity can also improve if the factors of production, labor, and capital move from less efficient to more efficient firms (between-firm growth). The failure of more productive firms to grow can be a sign that resources are being misallocated —and that there are barriers to the growth of more productive firms. 3. Productive entry and exit: Factor allocation is improved when new firms enter that are more productive than the average firm and when less productive firms leave the market (dynamic productivity growth). Productivity growth in Kazakhstan’s manufacturing sector, albeit low, in recent years has been driven by firm upgrading (“withinâ€? growth). But the between-firm growth was a drag on aggregate productivity, indicating a worsening in allocations of production factors between firms in 2015–18. The exit of firms made a small but positive contribution: firms exiting the markets were less productive than those that continued to exist. Kazakhstan has room to leverage all three sources of productivity growth to catch up with higher income peers. Policies should incentivize firms to innovate, allow productive firms to grow, and remove constraints on entry. Firms can improve their productivity by innovating, adopting better technologies, and implementing better managerial practices (“within-firmâ€? productivity growth). Productivity can also improve if production factors move from less-efficient to more-efficient firms (“between-firmâ€? productivity growth, associated with improved allocations of resources), either through firm growth and decline or through entry and exit (“dynamicâ€? productivity growth). The muted contribution of within-firm growth in Kazakhstan, particularly in the services sector, suggests that there is space to improve firms’ capabilities. Policies that target worker and manager skills, the innovative capacity of firms, and the quality of management can be expected to enhance productivity. Removing regulatory restrictions to competition is fundamental for efficient resource allocation and thus productivity growth. Competitive pressures between firms drive efficiency gains by allowing more productive firms, which usually produce higher-quality products and services at lower cost, to outcompete less efficient firms. Competition can also pressure firms to improve their production processes through innovation and more efficient operations. Kazakhstan Country Economic Memorandum viii Kazakhstan needs to step up efforts to reform state-owned enterprises (SOEs) and reduce the regulatory restrictiveness captured caused by the level of public ownership. Kazakhstan is an outlier on the degree of public ownership among comparator economies when the development level is considered. Strong presence of state in the economy undermines the overall efficiency of resource allocation and competitive playing field between SOEs and private firms, with the former granted better access to resources, markets, credit, and licenses. Deregulation of infrastructure services also inhibits competition. Except for railways, where many comparable countries show equal or higher levels of restrictiveness, most network services in Kazakhstan present regulations that are more restrictive than its peers, including in air transport, and fixed and mobile communications. Existing evidence suggests that regulatory barriers in services, including entry barriers, have diverse effects on downstream manufacturing performance, depending on the type of regulatory measure in question. In addition, anti-competitive policies (conduct regulations) on the operations of the firm in the services sector appear to play an essential role in explaining downstream performance across services and goods firms. Mobility: Constrained Skilled migrants predominate in the domestic population flows to the largest agglomerations in Kazakhstan, confirming that, under market conditions, skilled people seek places where other skilled people are abundant. By contrast, between 2014 and 2019, the mobility of the unskilled population declined in Kazakhstan (as did the share of unskilled people in the population). But the mobility of highly skilled people kept growing, so that by 2019, 9 percent of the country’s highly skilled population had moved to a different district.2 Roughly two-thirds of highly educated migrants in 2019 moved to one of the four largest cities. And as the share of highly skilled people among migrants grew, so did the role of those four cities as migration destinations. Internal migration in Kazakhstan closely follows spatial patterns of income and consumption inequality. Officially measured national income inequality has been largely stable and moderate by international standards. But at the local level, monetary wellbeing is much higher in the most prominent migrant- receiving areas than in migrant-sending areas. Those spatial gaps in monetary welfare can be decomposed into two sources.3 The first reflects differences in the composition of people living in rural versus urban areas (for instance, differences in the average age, endowments, and education level of the resident populations). The second reflects the portion of inequality that comes from simply living in a particular place and cannot be accounted for by differences in people’s observ able characteristics—a returns-to-place effect. The decomposition of spatial inequality into those two sources highlights the importance of access to the more dynamic labor markets in urban agglomerations, especially for the highly educated. The portion of the income gap between rural and urban areas accounted for by place rather than by the composition of the resident population is positive—about 39 percent at the mean—and rises with income. The returns-to-place effect between rural and urban areas explains about 15 percent of the gap at the 5th income percentile but about half the gap at the 95th percentile. As would be predicted from the direction of migration flows, the returns-to-place effect is much larger for Almaty and Astana, even when compared with the average rural–urban gap. In Almaty, it accounts for 55 percent of the gap on average, again rising with income. In Astana, it accounts for the entire gap on average. Also, in Astana and in contrast with the rest of the country, the resident composition effect contributes to closing the overall income gap. Kazakhstan Country Economic Memorandum ix Several factors restrict internal migration. One is the rural–urban difference in the cost of living. Living in Almaty is 190 percent more expensive than the country average.4 With that difference, many households simply cannot afford to move. Housing affordability is the main contributor, reflecting the rapid growth and volatility of housing prices over the past 20 years. Insufficient supply of housing in the most desirable places, particularly big cities, is due to some extent to local policy. But other matters can be tackled at a national level through supply-side and demand-side interventions, regulatory actions, and incentives to develop the severely limited rental housing market. The registration system also deters many people—particularly those receiving financial support from the state—from relocating within the country. For those who are unregistered, the system can lead to exclusion from many social services and benefits. Although new e-governance systems have simplified registration process, further deregulation and streamlining could substantially reduce administrative burden and the risk of exclusion. Because barriers to the internal mobility of labor can render the best-designed place-based policy irrelevant, enabling labor mobility should be fundamental to territorial development efforts in Kazakhstan. No matter how great an opportunity a place offers, if people cannot move there, for financial or bureaucratic reasons, it will not succeed. Addressing labor mobility constraints should thus be a top priority if Kazakhstan wants to see its cities emerge as engines of growth and prosperity. Internal trade: Concentrated Almaty and Astana are the main contributors to Kazakhstan’s increasing domestic trade. Between 2015 and 2019, they and their surrounding regions accounted for about 75 percent of intranational trade. The size of their imports is not surprising, because they are vital international hubs and have higher per capita income from nonmining industries than other regions, except Atyrau. High economic concentration in the two cities and surrounding regions also has a positive spillover. Almaty city, Almaty region, Astana city, and Akmola region contributed about 60 percent to growth in sourcing goods from intraregional and interregional trade. Most regions source more goods from Almaty city and region than from adjacent areas. The high level of goods purchased from Almaty also reflects the indirect sourcing of imports because 67 percent of Kazakhstan’s imports come from Almaty and its region. The share of interregional trade with them is more than 20 percent in nine of 14 regions, including Astana. But the share of goods sourced from adjacent regions is generally lower than that from Almaty, except in Zhambyl, Karagandy, and East Kazakhstan. Firm-level data suggest increasing concentrations of manufacturing activities across regions (Figure 3). The industry-level concentration index, developed by Ellison and Glaeser (1997), suggests a higher concentration of manufacturing activities in 2018 than in 2014, especially in regions near the borders, such as Akmola, Kostanay, North Kazakhstan, East Kazakhstan, and South Kazakhstan. Concentration in manufacturing also increased in two regions with access to large metropolises: Almaty region with Almaty city, and Karagandy region with Astana. Kazakhstan Country Economic Memorandum x Figure 3 Concentration index of manufacturing activities, 2014 and 2018, and import growth, by region and major city 2014 (left axis) 2018 (left axis) Import growth (right axis) 0.060 50 40 0.050 30 0.040 Concentration index 20 0.030 Percent 10 0.020 0 -10 0.010 -20 0.000 -30 -0.010 -40 Sources: Adapted from Tan and Tusha (2020) for the index and World Bank calculations for import growth. Note: Aktobe region is excluded because of an anomaly in trade data in the observed period. Areas with higher concentrations of manufacturing activities, especially those with access to border crossings, had higher imports.5 Between 2014 and 2018, imports surged in the northern regions of Akmola, Kostanay, and North Kazakhstan and the southeast of Almaty, where manufacturing concentration picked up. The positive association between imports and manufacturing concentration could indicate businesses’ preference to locate where they can better access labor, intermediate inputs, and imported materials. For example, the automotive and machinery industries are expanding within the Kostanay region, benefiting from rail and road links to international markets. Almaty region is experiencing growing activity in the food, machinery, and pharmaceutical industries. Regional markets tend to integrate with nearby regions, and markets in higher-density regions adjust more quickly to shocks. Specifically, trade in food products suggests that regional markets react faster to changes in nearby regions, so that prices that have deviated return more rapidly to parity. Similarly, markets in regions with higher population density (suggesting more economic concentration) also adjust with greater alacrity to shocks. Therefore, fostering the integration of local markets with nearby markets through local infrastructure and improving the links with international markets will strengthen overall internal market integration. Regional markets tend to integrate with nearby regions, and markets in higher-density regions adjust more quickly to shocks. Specifically, trade in food products suggests that regional markets react faster to changes in nearby regions, so that prices that have deviated return more rapidly to parity. Similarly, markets in regions with higher population density (suggesting more economic concentration) also adjust with greater alacrity to shocks. Therefore, fostering the integration of local markets with nearby markets through local infrastructure and improving the links with international markets will strengthen overall internal market integration. Economies of scale in production and transport are expected to reinforce and perpetuate these trends. Distance remains a crucial part of intranational trade costs in Kazakhstan. Improved transport links Kazakhstan Country Economic Memorandum xi would strengthen the incentives for businesses to locate closer to markets and labor pools, which would continue to benefit from agglomeration. The rapid market and population expansion in Almaty and Astana are crowding in businesses to their surrounding regions. Regions with access to border crossings are also experiencing greater concentration of manufacturing around specific locations. These agglomeration forces can further increase intraregional trade, so unless economically distant places have a strong natural advantage, they will likely find it hard to join production chains competitively. From oil to cities: Structural and spatial transformation The success of the Kazakhstan economy relies on the success of cities, which requires good infrastructure, good planning and land management, and a good regulatory ecosystem that enables the power of the markets. And territorial cohesion requires complementing spatially uneven economic growth with reforms and investments to ensure people’s access to opportunity, no matter where they live, and with targeted measures to address market failures that keep places from reaching their full economic potential. The government has a chance to address the shortcomings of its territorial development policy and make it a development force. It has been 15 years since the country’s first regional development strategy was established and nine years since the initiation of the State Program for Regional Development (SPRD)—the flagship program for territorial development. In that time, the territorial development effort has demonstrated some results, but the approach to addressing its challenges continues to evolve. Its economic geography has been defined by the legacy of central planning and the dependence on extractive industries. New forces are shaping the future, and in the past two decades, the largest cities have driven economic growth and attracted skilled migrants, signaling that agglomeration forces are boosting productivity. And as Kazakhstan starts to benefit from increasing returns to agglomeration, its territorial development vision and priorities are catching up with global best practices. The key Forecast Scheme for Territorial Development identifies four primary and 14 secondary functional urban areas and rightly identifies large cities as critical to improving economic development and people’s wellbeing. That recognition confirms the country’s departure from the spatially blind development doctrines of the past. And by introducing service access standards and continuing to implement sectoral state programs, Kazakhstan is maintaining a focus on access to services and infrastructure across all territories and settlements. This combination suggests that Kazakhstan has the right foundations to build on. Kazakhstan has begun, if tentatively, to enjoy increasing returns to the agglomeration of economic activity, which—with the migration of skilled workers and integration of regional markets— can fundamentally reshape its economic geography. Higher regional market integration also confirms the predominance of market mechanisms as drivers of territorial development today. The correlation in prices in different regions of Kazakhstan varies inversely with the distance between the regions. The speed of one region’s reaction to a price shock in another also appears to be determined by distance, the transportation network, and population density. These features indicate that regional markets are well integrated and confirm that market forces have emerged as the key determinants of Kazakhstan’s economic geography. The responsiveness of regional markets to price shocks indicates an economy where market forces incentivize the reallocation of factors of production toward places where they are used most efficiently. For labor, that usually means moving to big cities—exactly what is seen in the migration of skilled workers to Kazakhstan’s largest urban agglomerations. Kazakhstan Country Economic Memorandum xii There are disturbing signs, however, that returns to agglomeration are starting to diminish. The four largest cities saw far less impressive productivity growth in 2015–18 than the national leaders, Mangystau and Atyrau, which are extractive regions. Cities have also experienced contracting GDP per capita, explainable by the initially lower productivity of newcomers, but still a negative trend. The diminishing returns to agglomeration would be less worrisome if they indicated a redistribution of capital incentivized by high agglomeration costs in the largest cities following the growth of regional disparities due to earlier high returns to agglomeration. But that does not seem to be happening Kazakhstan’s three largest urban areas account for 23 percent of the country’s population, while in developed countries that share averages 49 percent. So, the decline in urban productivity at this early stage of urbanization and concentration of the urban system is worrisome. Breaking with the past vision of territorial development The most recent territorial development policy, at its core, is defined by two policy documents: the Forecast Scheme for Territorial Development and the SPRD.6 The Forecast Scheme sets the main priorities, and the SPRD—a level down from the Forecast Scheme—outlines the implementation mechanisms. The two documents set an ambitious agenda centered on enabling the productive growth of agglomerations. Kazakhstan is also finalizing the Law on Agglomeration and amendments to the Budget Code, which will grant more autonomy to Cities of Republican Significance and their surrounding districts to coordinate and jointly promote agglomeration-based development in their respective areas, Such coordination is needed for cities and districts to coordinate reforms and fund joint initiatives such as developing clean energy, housing, green urban transit, and waste management. The Forecast Scheme positions the territorial development policies within the context of national plans: the Kazakhstan 2050 Development Strategy and the Kazakhstan 2025 Strategic Plan. It prioritizes the development of growth poles as the drivers of the national economy and focuses on density as an enabling condition for the benefits of economic agglomeration in major cities. It also sets a goal of access to basic infrastructure and services across the country, thus meeting global best practices in territorial development. And it introduces the concept of functional urban areas, thus dissociating urban development policy from city administrative boundaries, which are often restrictive.7 Finally, it identifies four primary and 14 secondary functional urban areas as priorities and acknowledges that monotowns and small towns, as well as remote small villages, have limited development potential and so should not be prioritized for investment—a big shift from the development principles of the recent past. The SPRD describes the main challenges of territorial development and offers policy tools to address them. It identifies the key issues of urban development in Kazakhstan—under-urbanization, or the prevalence of small towns and the relatively small size of the leading cities; the low density of urban settlements; low internal mobility; and the high cost of living in metro areas, attributable partly to a virtually nonexistent rental housing market. Although it rarely articulates the policy tools clearly, the program sets the overall implementation framework. Positioning the SPRD against other policy efforts is the challenge. Linking targets to implementation mechanisms—a struggle Kazakhstan’s territorial development policy efforts have yet to produce the desired outcomes. The official results of the previous round of the SPRD in 2016–20 suggest that it was one of the drivers of further agglomeration and concentration of population, but its contribution was unclear. Almost all those indicators focus on infrastructure, few prioritize location, and none focuses on agglomerations. Kazakhstan Country Economic Memorandum xiii The government may want to use the Law on Agglomeration and changes in the Budget Code to promote bottom-up initiatives by cities and district governments to address problems undermining agglomeration. The funding allocation under the SPRD confirms that spending on spatial targeting was limited. Only 18 percent of SPRD spending in 2015–19 was on components that implied spatial targeting. That included 7 percent on agglomerations, 3 percent on monotowns, 2 percent on key rural settlements, and 4 percent on housing subsidies for people moving to villages. The rest of the program’s funding went to activities that were not explicitly spatially targeted. It aimed predominantly to provide a basic level of services across the country and to meet the service and infrastructure provision goals in which the only spatial target is to distinguish between urban and rural areas. The SPRD operates much more as a vehicle for grants for infrastructure and services than as a vehicle for place-based policy. Place- based policies have a clear geographic scope and a goal of advancing local economic development. But most interventions implemented under the SPRD were not motivated by economic development goals, according to interviews with selected local government officials, who struggled to explain how investments connected to the priorities of economic development. All this supports the hypothesis that SPRD has followed the structures and mechanisms that it inherited from the sectoral programs it absorbed in 2014. The sector-specific interventions should not be seen as wrong or bad because they elevate people’s quality of life and build human capital, a fundamental pillar of development. But they hardly meet the stated goal of supporting the economic development of growth centers. Other state programs, though they implement initiatives much closer to the definition of a place-based policy, lack clear ties to the priorities of territorial development stated in the Forecast Scheme. A good example is the development of infrastructure for the industrial zones in Shymkent, supported by the “roadmap for businessâ€? of the State Program for Business Development. Such development is based on seeing a highly populated area surrounded by an agricultural region close to the large Tashkent agglomeration across the border in Uzbekistan as having potential for developing a food processing industry, currently held back by the lack of serviced land for manufacturing—a market failure. Although the full analysis of the intervention’s effects will require assessing possible relocation and displacement effects and its success will require complementary policies in transportation infrastructure, workforce training, housing, and so on, this policy comes much closer in design and intent than most SPRD investments to meeting the criteria of a place-based policy. Building the right foundations Improving territorial development requires not just adjustments to the SPRD but a comprehensive rethinking of the territorial development policy ecosystem—with place-neutral interventions as well as place-based ones. Without favorable structural, macroeconomic, and fiscal conditions, alongside a favorable business environment, even well-designed support for a specific sector in a specific region is unlikely to contribute to overall improvement. National structural policies may not intuitively seem the starting point for addressing lagging regions or underperforming cities. But addressing structural distortions is a necessary, if not sufficient, condition to unlock the potential of regions and cities. National conditions influence territorial development by shaping the overall business environment and enabling the optimal spatial allocation of factors of production. They shape the potential of cities and regions because they are critical to establishing an environment conducive to investment, productivity growth, and active participation in the labor force. With growing climate challenges and aging and coal-based energy infrastructure, structural reforms are important for cities and regions to mobilize resources to transition toward low-carbon development.8 Further, structural and regulatory factors Kazakhstan Country Economic Memorandum xiv determine whether market forces can drive the spatial allocation of factors of production to their most effective uses—the foundation for making the most of the economic potential of different locations and thus maximizing results for the national economy. The allocation of factors of production will be suboptimal if labor and capital mobility are heavily restricted within the country; if foreign capital and labor face substantial barriers to entry; or if access to domestic and foreign markets is restricted due to regulatory burdens, poor transportation infrastructure, and services. Some places will fail to reach their potential, and labor and capital will be used inefficiently. Empowering subnational governments Strong subnational governments are critical for successful territorial development—a strength that Kazakhstan lacks. Most subnational authorities do not have the capacity and resources to design and implement strategic programs for local economic development. The system curbs the potential contribution of territorial development to economic growth in two ways. First, it leads to inefficient prioritization of investments as local governments follow the guidelines and incentives of state programs—the resulting decisions are often disconnected from local realities and uninformed by any central understanding of local needs. Second, the failure to build subnational governments’ capacity limits the effectiveness of policies aiming to develop specific territories—whether large cities and agglomerations or monotowns. The framework for territorial development today is almost entirely top-down. The regions identify projects, apply for funding, and implement investments, but their actions are almost entirely predefined by the restrictive parameters and complex procedures of state programs. And while their own sources of revenue are limited for state programs, they remain the main source of funding for local capital investments and development initiatives. Most state programs limit funding to narrowly defined, ring-fenced projects pursuing sector-specific objectives that local governments must address with their project proposals. As the government prepares a new rural development program, it should consider cementing strong reforms to promote markets, an effective public sector, and greater empowerment to subnational governments. The national government is taking the first steps toward decentralizing powers, among the national priorities of Strategy 2020. But elements of the decentralization of authority in Kazakhstan’s current territorial development policy need to go further. The current SPRD mandates the governments of large cities to develop strategies and define key priorities for economic development but offers only brief guidelines for developing these strategies. Most large cities appear to have adopted strategies following the requirements of the program, a step in the right direction. But the quality varies. In practice, the implementation of plans is often derailed by incentives dictated by the state programs. This highlights the weakness of local strategies that often are not linked to budget planning or, thus, to implementation. The proposed changes to the Budget Code to allow subnational governments to coordinate and fund project initiatives is a positive step in decentralizing autonomy. The government can build on this initiative by providing support for subnational government to have more resources and develop capacity in budget planning and implementation. The way forward should further decentralize the authority to plan local development and entrust resources to the financial capacity of subnational governments, either by providing non-ring-fenced development grants or by delegating more sources of budget revenue. Such decentralization should go hand in hand with building the capacity of subnational governments, requiring a combination of technical support and incentives. Kazakhstan Country Economic Memorandum xv Kazakhstan should broaden the scope of state programs targeting territorial development. For agglomerations to thrive, institutional coordination is critical The Forecast Scheme for Territorial Development, recognizing that fragmentation deters economic development in metropolitan areas, mandated cross-jurisdictional spatial development plans for large agglomerations—a step in the right direction. In 2018, such plans were adopted for Shymkent and Almaty. But it is unclear whether the coordinated system of governance needed to implement those plans exists in the metropolitan areas, and the state program does not offer specific steps to resolve the issue. The only proven way to coordinate management of an agglomeration is to expand the administrative boundaries of the city— most recently done in Shymkent. But expanding boundaries also requires integrating the new territories. Developing monotowns, a legacy of Soviet planning, is a declared priority of Kazakhstan development policy. The policies proposed for them in the SPRD 2025 focus predominantly on infrastructure and service investments in monotowns, while retaining the goal of diversifying the economy. In addition to the SPRD 2025, complementary policies to develop single-industry towns are included in the program for developing productive employment and mass entrepreneurship for 2017–20, known as Enbek, and the state program for supporting business development, Business Roadmap 2025. A comprehensive program is also needed to help monotowns and their population to cope with the transition away from fossil fuel. The national policy should formally drop economic development targets from monotown development programs and reserve business support in them to local small businesses. Most of the targets that the current SPRD sets relate to infrastructure for monotowns, which is consistent with the focus on livability. Ensuring decent access to health care and education, which are predominantly covered by other state programs, is also important. But current policy still includes targets for attracting private investments to large monotowns, which seems unrealistic, given the experience presented here. The natural evolution of monotown development policy is toward directly acknowledging the focus on quality of life and access to services. While some targeted investments could be directed to unlock new economic opportunities in monotowns, those are exceptions, not the rule, and development programs should not be structured around such expectations. The government should also consider offering relocation grants and other forms of support for people aiming to move from monotowns to locations that offer better economic opportunities. Improving territorial development policy The government’s territorial development policy has evolved and improved but needs to further adjust its goals. Kazakhstan is struggling to create a policy framework where favorable structural conditions empower subnational governments to implement targeted multisectoral programs that address local economic development challenges. Three key directions are needed, and we consider them in turn. First, focus on structural reforms that enable production factors to move where they can be most efficiently used—initially, removing barriers to internal migration. If market forces are not allocating labor and capital to their most efficient applications, place-based economic development is doomed to fail. While Kazakhstan’s business environment has improved in recent years, territorial development would benefit from further easing of constraints on foreign investment (including industry-specific restrictions) and on immigration—most critically internal migration. The persistent internal migration barriers stem from unaffordable housing and the restrictive household registration system. Fully abolishing such registration, which has no equivalent in the developed world, is the first step. Housing affordability should be addressed through supply- and demand-side policies to shift the housing Kazakhstan Country Economic Memorandum xvi market toward a lower price equilibrium. And policy should create the environment for a rental housing market and should expand social housing programs. Relocation grants should be offered to people who want to move to a place with greater opportunities but cannot afford to relocate. Second, adopt a systematic strategy of empowering and building the capacity of subnational governments. Augmenting the capacity of local governments and their resources is common to all the best practices of territorial and local development. Empowering local governments is integral to moving from a top-down system and to improving approaches to the specific challenges of local development in agglomerations and monotowns. Kazakhstan is taking steps in the right direction by decentralizing some revenue streams, increasing the accountability of local leaders to residents, and requesting that local strategies be developed in major cities. But more is needed to make local governments into capable leaders of economic development agendas. Capacity building for local authorities, critical to any subnational policy, should be mainstreamed and made integral to territorial development. Current strategies for developing priority agglomerations vary in quality. A more systematic approach to building the integrated planning proficiency of local governments could create detailed guidance and develop interactive tools to give subnational governments analytical foundations for strategic planning. Kazakhstan could establish a national institution devoted to enhancing local planning capacity by building such tools, providing technical support, and ensuring the high quality of local plans and strategies.9 It should also consider tools such as asymmetric decentralization and project competitions as incentives for subnational capacity building. Third, reposition the SPRD as a regional and urban development fund supporting integrated, multisectoral local economic development programs to be implemented by subnational authorities. Today the SPRD, though envisaged as a local integrated development tool, functions as a mix of sector- specific subprograms that did not fit into other state programs, so that its role and impact are diminished. Perhaps the SPRD can become a single program that supports complex multisectoral projects in opening new local economic development opportunities. Such complex projects are probably expensive, however, and not many could be supported, so the SPRD would need a transparent competitive process to allocate funding. The competition could include criteria for spatial targeting and for prioritizing selected industrial sectors. The guidelines for project assessment should be made clear, and we recommend a framework that can guide the development and evaluation of place-based policies and project proposals. The projects selected should clear the high bar of proving that proposed investments and incentives address market failures and will not simply relocate some existing economic activity (Box 1). Box 1 Principles for assessing place-based policies and projects Given the competing claims for assistance and resources the large, upfront costs of many public investments, and the often-long-lasting nature of the assets, the choices of places getting the intervention and the type of interventions should be informed by a realistic, objective, and systematic appraisal of projected policies and projects. An economic appraisal of a proposed place-based policy or project should rest on the following principles: • Provide a clear narrative. Policymakers need to develop a narrative that clearly lays out the market failures or distortions which, if addressed, would foster the region’s progress, and drive the design of the intervention. Kazakhstan Country Economic Memorandum xvii • Fully describe the direct and indirect quantity changes. A complete appraisal must take a dispassionate look beyond the expected direct effects to the indirect effects, which are harder to measure. Such indirect effects are often invoked as the critical tipping consideration in defending a policy. But they usually are not well documented, and the arguments usually are not supported with empirics. • Consider complementary conditions and policies. Market failures often come in multiples, implying that cost-benefit analysis of any one intervention will be misleading because of the complementary effects of resolving several at once. • Consider general equilibrium effects and displacement effects. The former are the quantity changes that occur in response to changes brought about by the policy, and the latter, the changes in one place occurring at the expense of another. • Place values on the quantity changes. After the direct and indirect quantity factors have been identified and assessed, the next step is to place a value on quantity changes (social valuation). • Be candid and explicit about government capabilities. Many place-based policies require interventions with multiple dimensions, large budgets over long periods of time, and well- developed government capabilities for diagnosis, design, and implementation. Limited capabilities may mean, even if a program is appraised well and appears to yield good benefits, that the returns may be low in practice. Policy design should thus look for ways to limit the burden on government. Ideally, even the simplest road project would have a full appraisal that would allow a solid ranking of projects by their social value added. This would offer some disciplining of the often-formidable pressures to “do somethingâ€? to either reverse the declining fortunes of an area or kickstart a long - standing laggard. However, while the direct effects can often be quantified, doing the same for the indirect effects is expensive, time consuming, and beyond the capabilities of even advanced economy governments. Often simpler rules of thumb are employed, sometimes based more on the symptoms than a careful diagnosis of the underlying disease. Given the challenges facing even well established and competent bureaucracies, it is probably better to view the framework here not so much as a mechanical valuation device and more as a heuristic tool that informs the dimensions that should be considered, that focuses debate, and that follows some policy guidelines. Kazakhstan Country Economic Memorandum xviii Part 1 Forces driving growth and transforming economic structures Kazakhstan Country Economic Memorandum 1 1 Restarting Kazakhstan’s stalled structural and spatial transformations Kazakhstan’s economy depends heavily on extractives, with many other industries linked to them. Oil and gas contribute more than 20 percent to GDP, more than half of it from extraction and more than a third from the services linked to the sector. Mining contributes another 15 percent of GDP. Commodity revenues account nearly half of overall government revenues. Commodity exports, meanwhile, account for more than 80 percent of total shipments abroad. Manufacturing’s contribution to GDP has been static since 2010, accounting for around 11 percent of Kazakhstan’s GDP in 2018, with a substantial increase since 2010.10 Between 2015 and 2018, the fastest annual growth was in resource-heavy manufacturing (metals, materials, petro, pharma) followed by the textiles and apparel industry. More than 60 percent of the total manufacturing value comes from the central region of Karagandy and eastern regions of East Kazakhstan, Pavlodar, and Almaty. Karagandy, East Kazakhstan, and Pavlodar are traditionally known for their metallurgy. Karagandy is also the key manufacturing hub, with large industrial plants inherited from the Soviet times specializing in iron, steel, and copper. Manufacturing is mostly concentrated in resource- and capital-heavy activities, with petroleum-based products and basic metals and minerals accounting for more than three-fourths of manufacturing value added. But food and beverage production remains important, particularly in the Almaty region. Services contribute more than 50 percent to Kazakhstan’s GDP,11 and its annual growth rate has been positive since 2010.12 More than 75 percent of services value added comes from the largest agglomeration centers of Almaty and Astana, and from the Atyrau region, with the expansion of commerce and other services. Between 2015 and 2018, commerce and transport and storage have seen the biggest annual growth in value added in services, followed by construction and ICT sectors. Commerce accounts for nearly half of services’ value in the country, followed by transport and other services. Knowledge-intensive services, such as ICT, are largely the cities of Almaty and Astana, which together account for more than 90 percent of ICT’s value. Agriculture’s contribution to GDP has remained static at around 5 percent in the 2010s.13 Traditionally, around 60 percent of agricultural value added comes from the north, including Aqmola, North Kazakhstan, and Kostanay, regions that specialize in crop farming. All three regions have also experienced a decline in their contribution to agricultural value added since 2010. Kazakhstan’s economic prospects since 1990 were shaped by its move from plan to market, its huge surge in oil exports to global markets, and its role as a hub for Central Asia. The first delivered substantial dividends, but the pace of reforms has plateaued, and Kazakhstan’s growth rate has slowed after every downturn. After almost reaching high-income status by 2014, the country’s per capita income has since dropped, and it is now stuck in middle income. The second helped integrate the country with the world and generated large resources for development, but the economic model of dependency on oil is on its last legs. And the third, as an economic hub for Central Asia, is challenged by the emergence of Uzbekistan. Productivity growth was strong in the 1990s and the early 2000s but has since fallen dramatically. One reason for the decline is the sluggishness of structural transformation. The economy remains dependent on oil and international transit, with substantial misallocations of resources across firms and industries. So, the spatial structure of the economy is rapidly becoming more important for Kazakhstan Country Economic Memorandum 2 Kazakhstan’s future development. The existing structure reflects the vestiges from the planned economy: a lot of production is still in places where it would not be under a market system—near consumers, other firms, or near the borders. And since the 1990s, the spatial transformation, though advancing, has been slow. Now, the challenge for the authorities is to harness the power of both the structural and spatial transformation. On the structural side, the economy needs to complete the shift from plan to market and accelerate the transformation from oil dependence to a more diversified economy. On the spatial side, it needs to move toward spatial structures driven by market principles. Productivity drives growth and transforms economic structures Between 2000 and 2010, Kazakhstan’s economy grew by an annual average of 8 percent, placing it among the world’s 20 fastest growers.14 Strong growth in this period—boosted by rising oil and gas prices, rapid domestic demand, and some economy-wide reforms in the early 2000s—drove significant welfare gains, as real wages surged and inequality and poverty declined. Kazakhstan’s GDP per capita also converged toward that of the United States, but it slowed after the global economic downturn and Kazakhstan’s economic crisis of 2014 with the fall in oil prices and a major devaluation weakening domestic demand.15 Only recently has it been on an uptick. Productivity measured by total factor productivity (TFP) has been an important driver of economic growth in Kazakhstan. A country’s ability to improve its standard of living over time depends almost entirely on its ability to raise its productivity.16 Productivity growth is a major determinant of long-term economic growth, which explains up to half of the income differences around the world (Figure 1.1). Kazakhstan experienced strong productivity growth in the early 2000s, with its contribution to annual GDP growth of 6 percentage points on average.17 The period of strong productivity growth was followed by productivity declines in 2008–09, a productivity resumption (at a much lower level) in 2010–13, followed by declining productivity in 2015–16 with the main driver to GDP growth being the capital base (Figure 1.2). Together with a need to increase investment in the capital base and in human capital, a concerted effort to invest in productivity growth is crucial to foster economic growth and create better job opportunities. Kazakhstan Country Economic Memorandum 3 Figure 1.1 Across countries, productivity is a Figure 1.2 TFP’s contribution to GDP long-term determinant of GDP per capita growth has declined GDP per worker and total factor productivity, 2019 Contribution to annual growth, 2002–16 Source: World Bank 2020a. Source: World Bank 2018. At the macro level, TFP has been the main driver of labor productivity and economic growth in Kazakhstan in the past two decades. GDP per capita can grow thanks to higher labor productivity or to higher labor participation. In the early 2000s, both were growing relative to the United States, but most of the convergence was explained by labor productivity (Figure 1.3). 18 In the past decade, labor participation has been declining, particularly among the female population, and labor productivity remained stagnant.19 As of 2017, the Kazakhstan’s labor productivity was nearly 40 percent of the United States’s level, nearly the same as in 2010. Labor productivity can also be explained using capital and by how efficiently labor and capital are used in production (total factor productivity). The capital- labor ratio and TFP growth—the key components of labor productivity expansion—have been experiencing different patterns relative to the United States. The capital-labor ratio has remained like the United States, suggesting a small role of capital deepening in labor productivity. Instead, TFP has been the key driver of labor productivity changes and thus of economic growth in general (Figure 1.4). Figure 1.3 Labor participation declined and Figure 1.4 With capital deepening only labor productivity stagnated relative to the slowly, most labor productivity trends are United States, limiting convergence explained by TFP Labor productivity, participation, and GDP per capita Labor productivity, capital-labor ratio, and TFP (US = (US = 100%) 100%) 100% 100% 80% 80% 60% 60% 40% 40% 31.49% 20% 20% 10.99% 0% 0% Labor productivity Labor participation GDP/capita TFP Capital-labor ratio Labor productivity Kazakhstan Country Economic Memorandum 4 Source: World Bank 2020a. Source: World Bank 2020a. Firm-level data confirm muted TFP growth, with moderate growth in the most recent years. The Strategic Development Plan to 2025 lists productivity growth as one of the key drivers of economic transformation to lift Kazakhstan into the top 30 most developed countries by 2050. This chapter analyzes productivity trends in Kazakhstan using administrative firm-level data between 2009 and 2018 to better understand the drivers of firm-level growth at country and regional levels. With information for more than 70,000 formal firms, covering 1.6 million jobs in the country, the firm-level data confirm a decline in TFP between 2010 and 2015 in both manufacturing and services. In more recent years, TFP growth has been positive, but modest. Between 2015 and 2018, TFP grew by 1.1 percent in the manufacturing sector and by 0.4 percent in the services sector.20 The roles of firm upgrading, factor allocation, entry, and exit Understanding the drivers of productivity growth helps understand why Kazakhstan has seen little growth. Aggregate productivity growth can be the result of three processes, which are usually concurrent: 1. Improving firm capabilities: Firms can adopt new technologies and better managerial practices (within-firm productivity growth). 2. Improving factor allocation: Productivity can also improve if the factors of production, labor, and capital move from less efficient to more efficient firms (between-firm growth). The failure of more productive firms to grow can be a sign that resources are being misallocated —and that there are barriers to the growth of more productive firms. 3. Productive entry and exit: Factor allocation is improved when new firms enter that are more productive than the average firm and when less productive firms leave the market (dynamic productivity growth). Productivity growth in Kazakhstan’s manufacturing sector in recent years has been driven by firm upgrading (“withinâ€? growth), albeit low. This indicates that the average Kazakh manufacturing firm was able to increase their productivity. However, the total between-component was a drag on aggregate productivity, indicating a worsening in allocation of production factors between firms in 2015 –18. Further decomposition of the between-component shows improved allocation of resources within manufacturing industries (intra-industry): more productive manufacturing firms increased their market share at the expense of the less productive. Yet, the deteriorated allocation of resources between manufacturing industries (inter-industry) is what contributed negatively to the overall between-firm growth. The exit of firms nevertheless represented small, but positive contribution: firms exiting the markets were less productive than those who continued to exist. Several manufacturing sectors—such as wood, computers, textiles and clothing, paper and printing, and machinery and electronics—saw deteriorations in allocative efficiency. In these sectors, it was not always the most productive firms that increased their market share, as indicated by the negative intra- industry between-component (Figure 1.5). In the metals and materials sector, the manufacturing sector with the highest employment share, both the within and between components were positive. In food and beverage manufacturing, the manufacturing sector with the second highest employment shares, most components are close to zero, muting overall growth in this sector. Kazakhstan Country Economic Memorandum 5 Figure 1.5 Reallocation reduced TFP growth in several manufacturing sectors Contributors to TFP growth in manufacturing sectors, 2015 –18 10% Within Between (intra-industry) Between (inter-industry) Entry Exit 8% 6% 4% 2% 0% -2% -4% -6% -8% -10% Food, bev., Textiles, apparel Wood, paper and Metals, materials, Computers, Motor vehicles Other tobacco and leather printing petro, pharma machinery, manufacturing electrical Source: World Bank 2020a. Note: TFP growth is averaged over three years. Aggregate TFP is weighted by value added. In services, productivity growth in recent years was driven by positive reallocation of factors of production between firms in a sector (intra-industry “betweenâ€? growth) and to a more limited extent by firm upgrading (“withinâ€? growth). The positive contribution of the intra-industry “betweenâ€? component, suggesting that in a particular services sector more productive firms were able to expand their market share, is almost fully canceled by a negative inter-industry “betweenâ€? component, suggesting that less productive sectors expanded their market share at the cost of more productive industries. The “withinâ€? contribution was low, though positive, at 0.3 percent, suggesting that firm upgrading played only a small role in overall productivity growth in services between 2015 and 2018. Construction, commerce, and transport saw positive effects of resource reallocation (mostly intra- industry effects) between firms to TFP growth—an indication that more productive firms increased their market shares in these sectors. These are also the sectors that have experienced considerable employment boost since 2009. In commerce, the within-firm growth was negative at 1.24 percent. In accommodation and food services, information and communications technology (ICT), and other services, the negative contribution of reallocation to TFP growth has been the largest. Kazakhstan Country Economic Memorandum 6 Figure 1.6 Reallocation increased TFP growth in construction, commerce, and transportation services Contributors’ TFP growth, 2015–18 15% Within Between (intra-industry) Between (inter-industry) Entry Exit 10% 5% 0% -5% -10% -15% ICT Commerce Transport Mining Utilities Other services Agriculture Manufacturing Construction Accommodation / food service Source: World Bank 2020a. Note: TFP growth is averaged over three years. Aggregate TFP is weighted by value added. Compared with other countries, the contribution of within-firm growth, reallocation of resources, and business dynamism has been low and at times muted in Kazakhstan. When compared with Poland, Slovenia, and Serbia, productivity growth in Kazakhstan has been low with low contribution of resource reallocation. For example, in Serbia, most of productivity gains up until 2014 were achieved through better allocation of resources to more productive firms, allowing them to grow. However, this pattern reversed between 2014 and 2017, as economic transformation gave way to workaday market functioning. Similarly, in Poland in early years after its transition to a market economy, 15 percent of productivity growth was attributed to allocative efficiency between firms. Kazakhstan has room to leverage all three sources of improvements in productivity growth to catch up with higher-income peers. This could come from further increasing the contribution of firm capabilities to fostering more efficient production. In the same way, removing barriers to firm growth by improving the business environment would allow for production factors to flow to the most productive firms, both incumbent as well as new entrants. Kazakhstan Country Economic Memorandum 7 Figure 1.7 Experiences from other countries show that firm upgrading (within), better allocation of production factors (between), and entry and exit played a large role in achieving productivity growth Cross-country comparison of contributions to TFP growth in manufacturing 25% 21% 20% Within Between Entry/Exit 15% 15% 10% 5% 5% 4.5% 5% 4% 5% 3% 2% 3% 3% 2% 1% 0.3% 0.16% 0.41% 0.5% 0.0% 0.0% 0.1% 0.1% 0% -0.4% -1% -0.7% -1% -0.6% -0.5% -5% -2% -4% -0.1% -5% -10% -8% -10% -15% 1997-1999 2003-2005 2007-2009 1995-2000 2007-2009 2009-2011 2012-2014 2015-2017 2009-2012 2012-2015 2015-2018 Poland Slovenia Serbia Kazakhstan Source: World Bank 2020a. Note: TFP is weighted by value added. To raise productivity, policies should incentivize firms to innovate, allow productive firms to grow, and remove constraints on entry. Firms can improve their productivity by innovating, adopting better technologies and implementing better managerial practices (“within-firmâ€? productivity growth). Productivity can also improve if production factors move from less efficient to more efficient firms (“between-firmâ€? productivity growth, associated with improved allocation of resources), either through firm growth and decline or through entry and exit (“dynamicâ€? productivity growth). Table 1 summarizes these three components and their links with policy. Most policies are not exclusively associated with only component: for example, reforming business regulations can both facilitate entry of new firms (“dynamicâ€?) as well as facilitate the growth of firms that were previously restricted (“betweenâ€?). Table 1.1 Sources of productivity growth and facilitating policies Within-firm Between-firm Dynamic (Entry and Exit) Firms increasing their capabilities Allocating resources to more Entry of productive and exit of productive firms unproductive firms Capabilities to be targeted are Misallocation of resources indicate Entry of highly productive, fast- human capital skills, management barriers that prevent the movement growing firms (gazelles) and exit and organizational practices, using of capital, labor, and other factors of of less productive firms that are and adopting technology, and production to the most productive not growing (laggards) innovation by the firm firms in the economy Policy levers Policy levers Policy levers Improving education and technical Removing distortions in product Same as within-firm and between- skills; encouraging markets (competition firm productivity levers, with entrepreneurship, technology policy); addressing frictions in land, emphasis on limiting barriers to adoption and innovation; reducing capital and labor markets; opening entry and exit of domestic and administrative burdens markets to trade and investment foreign firms (e.g., licensing, asset recovery) Kazakhstan Country Economic Memorandum 8 The muted contribution of within-firm growth in Kazakhstan, particularly, in the services sector, suggests that there is space to improve firms’ capabilities. Policies that target improvement of worker and manager skills, the innovative capacity of firms, and the quality of management can be expected to enhance productivity. According to the World Bank’s Enterprise Surveys (2019), firms report the absence of skilled workers or a poorly educated workforce as a major constraint. The share of such firms is also considerably larger that in peer economies of Europe and Central Asia (ECA) and at upper- middle income. At the same time, the willingness to train workers remains low. The lack of growth of more productive firms points to the importance of removing distortions in product and input markets. Policies should address current distortions that do not allow factors of production to flow freely to the most productive firms, restrain entry, or reduce firms’ own efforts to become more efficient. Although in Kazakhstan’s manufacturing, productivity decompositions provide some evidence for factors flowing to more productive firms (even though limited), productivity growth could have been faster if regulatory distortions were eliminated. In contrast, in services, factors have been flowing to the least productive firms, a sign of sector-specific distortions and barriers to competition. In some sectors, such as telecommunications and professional and technical services, the most productive firms grew more slowly than the least productive. This is a sign that production factors are not always flowing to the firms that can use them most efficiently. Policies that remove distortions in input and output markets can therefore be expected to lift productivity. Removing regulatory restrictions to competition is fundamental for efficient resource allocation and thus productivity growth in services. Competitive pressures between firms are a driver of efficiency gains by allowing more productive firms, which usually produce higher-quality products and services at lower cost, to outcompete less efficient firms. Competition can also pressure firms to improve their production processes through innovation and more efficient operations. According to the Organisation for Economic Co-operation and Development (OECD) indicator that measures the restrictiveness of product market regulations (OECD PMR), Kazakhstan’s regulatory restrictiveness to competition in 2018 (as it appears on the books) is higher than the OECD average. When compared with ECA peers, Kazakhstan fares worse than most peers, except Türkiye. A limping private sector Comparative and mostly descriptive analysis shows that, on firm performance, Kazakhstan underperforms its peer groups and countries. Firm growth profiles look rather grim, with firms entering the market with a small workforce, and adding few workers, if any, during the first 10 years of operation. But the business environment appears friendly to private sector across many aspects, such as regulatory burden, corruption, informality, and quality of infrastructure (electricity and roads), having seen substantial improvements over the last decade. There is still a room for further improvements, especially across regions within the country. Among the aspects of the business environment, access to finance appears most difficult for firms, with the measures of bank dealings showing specific difficulties, especially for small and medium enterprises (SMEs). And firms in Kazakhstan have a large gap to close in their capabilities, including their potential for international trade. Kazakh firms lag considerably behind comparator groups and countries in firm performance. Firm growth profiles suggest that they enter the market smaller than in Chile, though like Russia. During the first 10 years of their lives, they add fewer workers than in Chile, and many shrink, especially SMEs. On many enterprise survey measures of business environment, Kazakhstan is comparable to, or even more favorable than, ECA and UMIC averages. An exception is access to finance, with a higher share Kazakhstan Country Economic Memorandum 9 of firms in Kazakhstan than in ECA and UMICs facing harder credit constraints. Kazakhstan also has a gap to close in the speed of clearing goods through customs while exporting and importing. Kazakh firms are more constrained in their capabilities than firms in peer groups and countries. Per WBES 2019, only 1.4 percent of firms in Kazakhstan are foreign owned (at least 10 percent)—a measure that is often used as a proxy for economy’s potential to attract foreign direct investments—which is close to Russia, but a fifth the ECA average and a sixth the Chile average. On having internationally recognized quality certification, technology licensed from foreign companies, or using material inputs of foreign origin in their production, Kazakh firms underperform comparator groups and countries, and SMEs are considerably more disadvantaged than large firms, likely limiting their growth potential. There is much room and value for further improvements of the business environment, especially in expanding firms’ access to finance, speeding clearing through customs, and closing regional variations in many measures of business environment. The government appears to be focusing on some of these improvements already. Kazakh firms also show clear weaknesses in international trade that policy could alleviate through opening new markets—as through negotiating deeper and more comprehensive agreements with the existing partners or pursuing trade agreements with more countries. Box 1.1 Gauging the performance of Kazakh firms To gauge relative performance of Kazakhstan’s private sector, the comparators are ECA countries,UMICs, the Russian Federation, and Chile. The Russian Federation is a peer country because it is among the biggest trading partners of Kazakhstan and shares economic similarities. Chile shares a richness in hydrocarbons and mining, yet its growth model has been quite successful in reducing the reliance on mineral rents, developing autonomous engines outside mining to buffer the economy against commodity price fluctuations, and lifting the country to high income in 2013. Lifting Kazakhstan into the top 30 most developed economies is one of the key objectives of economic transformation set by the government of Kazakhstan in its 2025 Strategic Development Plan. This would also require diversification of the economic base, for which Chile’s experience could be helpful. Firm performance Kazakhstan’s private sector performance—in terms of labor productivity (measured by sales per worker)—is below the ECA and UMIC averages, as well as Chile. Sales per worker in Kazakhstan is considerably and statistically significantly below the comparator groups and countries. The median is around $15,000 in 2019 (down from the levels of 2009 and 2013), while ECA and UMIC medians are around $40,000 and $31,000, respectively.21 Kazakhstan’s private sector also underperforms Russia ($29,000) and Chile ($51,000 in 2010). Sales per worker for SMEs (five to 99 employees) is considerably lower than for large firms (100 or more employees). The gap in median sales per worker is considerable between SMEs ($14,000) and large firms ($25,000) in 2019—a 78 percent difference. In comparison, the gap is 27 percent in Russia, and 49 percent in Chile. Kazakhstan Country Economic Memorandum 10 Figure 1.2 Median sales per worker by firm size (in USD 2009) in the formal private sector: Kazakhstan versus selected peers Source: World Bank 2020b. Moreover, evidence also shows that formal private firms do not grow as they age in Kazakhstan, reflecting a stagnant private sector. A crucial driver for economic development is the speed at which the average business grows over its lifecycle (Hsieh and Klenow 2014; Eslava, Haltiwanger, and Pinzón 2019). Analysis based on 2019 WBES shows that private firms in Kazakhstan appear to enter the market with a lower number of workers and face stronger pressures during the first 10 years of their life. Kazakh formal private firms up to 10 years old employ on average only 14 workers, compared with 63 in Chilean peers. This gap grows wider (to almost 80 employees) for firms 20 or more years old. The gap between Russian and Kazakh formal firms also widens over time, though on a smaller scale: it jumps from a gap of two employees among firms up to 10 years old to 10 for firms 20 or older. These results suggest that Kazakh private firms struggle to expand as they age. Figure 1.3 Average workforce by firm age group in the formal private sector: Kazakhstan versus selected peers Source: World Bank 2020b. Note: The top decile (of firm size times World Bank Enterprise Survey (WBES) sampling weights) was removed for each country to minimize volatility. Exporting and innovation On various measures of exporting and innovation, firms in Kazakhstan considerably underperform peer groups as well as Chile. Only 3.9 percent of firms in Kazakhstan export directly at least 10 percent of Kazakhstan Country Economic Memorandum 11 their sales, compared with 15.6 percent for ECA and 12.1 percent for UMICs. Chile’s average of 8.2 percent is more than twice the level of Kazakhstan. Among measures of innovation, prevalence of spending on research and development (R&D) shows the most underperformance, including in comparison with Russia. Only 2.1 percent of firms in Kazakhstan spend on R&D around a fourth that in ECA (and Russia), an eighth that in UMICs, and around a 20th that in Chile. Business environment On several important aspects of business environment, Kazakhstan performs similar to or better than comparator groups and countries. On bureaucratic burden of regulations, corruption, and informality, Kazakhstan performs comparatively well, with considerable regional variations offering room for further improvements. While the main infrastructure in Kazakhstan (electricity and roads) appears in better shape than in comparator groups and countries, firms seem to have difficulties finding workers well versed on the tasks at hand. Accessing finance appears harder for Kazakh firms than in ECA and UMICs with a higher share of firms facing full credit constraints, though the share of firms facing full or partial credit constraints is similar. Banks loans are difficult to obtain in Kazakhstan, especially for SMEs. Senior managers of firms in Kazakhstan spend considerably less of their time dealing with regulations than those in ECA, UMICs, and Chile. In this measure of regulatory burden, Kazakhstan is like Russia. Large firms face considerably higher regulatory burdens than SMEs, with senior managers spending 9 percent of their time on dealing with government regulations, more than twice the 4 percent for SMEs. Measures of corruption steadily declined in Kazakhstan after 2009, with the 2019 levels comparable to ECA and UMIC averages, but worse than Chile. On a perception-based measure of corruption, 7.9 percent of firms in Kazakhstan report being expected to give gifts to public officials to get things done, compared with an average of 10 percent in ECA and UMICs. Kazakhstan’s level is better than Russia’s, but considerably worse than Chile’s, where fewer than 1 percent of firms report the same. While informality in Kazakhstan does not appear as severe as elsewhere, the businesses seem to be persistently troubled by it. Sixteen percent of firms in Kazakhstan in 2019 and in 2013 name informal competition as the top obstacle to their operations (this share in 2009 was 13 percent). This is the second most frequently named top obstacle in Kazakhstan in the list of 15 different potential obstacles (after tax rates). Furthermore, 24 percent of firms identify informal competition as their major constraint in 2019, the second highest share of firms identifying a major constraint (after inadequately educated workforce). Although the differences are not statistically significant, SMEs are more likely than large firms to be troubled by informal competition, with higher share naming informality as their top obstacle: 16.4 versus 14.8 percent, and higher share identifying informal competitions as their major constraint: 24.3 versus 15.1 percent. It takes longer to clear exports and imports through customs in Kazakhstan than in ECA and UMICs. The speed with which goods can cross borders is an important element in firms’ ability to be part of the global value chains. On average, it takes nine days to clear exports and 14 days to clear imports in Kazakhstan, considerably and statistically significantly above the ECA (4.2 days and 6 days, respectively) and UMIC (6.8 and 9.8, respectively) averages, though like Chile (10.8 and 11.3, respectively). Kazakhstan outperforms Russia (15.6 and 20.4, respectively) on these measures. As there are few exporters (and importers) in Kazakhstan, these averages may be relatively noisily estimated. Kazakhstan Country Economic Memorandum 12 Firm capabilities Kazakhstan underperforms most comparator groups and countries in almost all the Enterprise Survey measures of firm capabilities. Only 6 percent of firms in Kazakhstan have internationally recognized quality certification (Figure 1.8), considerably and statistically significantly below averages for ECA and UMICs (around 18 percent), and Chile (22 percent). A similar pattern is observed for the share of firms using technology licensed from foreign companies as well as other measures not shown in Figure 1.8, namely, the share of firms that have their annual financial statement reviewed by external auditors and the share of firms with their own website. Furthermore, top managers in Kazakhstan have on average five years’ less experience working in the firms’ sector than the average for ECA and UMICs, and 10 years less than the average for Chile (statistically significant differences). Firms in Kazakhstan are less adept at attracting foreign ownership than in ECA, UMICs, and Chile. While firms’ foreign ownership is likely affected by factors that are outside of firms’ control, it is still an important indicator of firms’ ability to find important partners across the border. In Kazakhstan, only 1.4 percent of firms have at least 10 percent foreign ownership, less than half the levels in 2009 and 2013, five times less than the ECA average of 7.2 percent, six times less than the average for Chile (8.5 percent), and almost seven times less than the UMIC average of 9.7 percent. Figure 1.4 Firm capabilities in Kazakhstan and comparator groups and countries Source: World Bank 2020b. Firms in Kazakhstan underperform ECA and UMIC averages in intensive and extensive margins of the use of inputs of foreign origin in manufacturing. Sixty percent of manufacturing firms in Kazakhstan use imported inputs or supplies, statistically significantly below the averages for ECA, UMICs, and Chile. Among the firms that use foreign inputs, the proportion of imported inputs used in production is 53 percent in Kazakhstan, below the ECA and UMIC averages of 60 percent. On almost all measures of firm capabilities, SMEs underperform large firms. The differences are considerable and statistically significant (except for foreign ownership). Notably, while a lower share of SMEs than large manufacturing firms use imported inputs (59 versus 78 percent), SMEs use a higher proportion of imported inputs in their production (54 versus 41 percent). Policy recommendations that follow from this analysis include further improvements of business environment and efforts to open new markets for firms in Kazakhstan. The government appears to be focusing on these already. The continued efforts would be of great importance, particularly considering the sizeable and tangible success that the reform agenda has already brought about. An important aspect of business environment, namely the difficulty that firms face in clearing goods at customs for exports or imports, would also benefit from additional attention. Simplifying procedures and speeding Kazakhstan Country Economic Memorandum 13 customs would likely help businesses interact with the rest of the world more efficiently. And firm capabilities that possibly help in this process show clear signs of important weaknesses that can potentially be alleviated through opening new markets for firms in Kazakhstan. Agglomeration for productivity Agglomeration economies, derived from firms locating near one another in industrial clusters, can be a source of productivity growth in Kazakhstan. In general, there are four ways for firm productivity to benefit from agglomeration economies. One, as firms in related sectors cluster together, firms can enjoy reductions in production costs, as there will be multiple competing suppliers and/or greater economies of scale among the suppliers. Two, industrial clusters can increase market access as firms enjoy low transport costs to a larger market. Three, industrial clusters also provide a large labor supply, which increases labor matching and lowers search costs, especially for specialized workers. Four, there are knowledge spillovers between firms, which can learn about new production processes and techniques and can benefit from technology adoption from their neighbors. As industrial clusters grow, more firms are attracted to locate in them, and these benefits are amplified. Kazakhstan has a large land mass, but its population and industry do not seem as agglomerated. Kazakhstan is the world’s ninth largest country by territory, but it is the 12th least densely populated, with just seven persons per square kilometer compared with 141.4 for the OECD average.22 Regional- level location quotients show that Mangistau and Karagandy regions were the most industrially specialized in 2018, followed by Atyrau and Pavlodar. Conversely, in comparison with the industrially specialized west, the northern and southern regions (such as North Kazakhstan, Aqmola, and Zhambyl regions) concentrate most agricultural activities relative to the rest of the country. Activities in the services sector are much less spread out and located mostly in the larger agglomeration centers of Almaty and Astana. While it is possible to examine the extent of agglomeration of a sector by observing the location of firms, it is difficult to compare the sectors: since each sector has a different size distribution of firms, they are not directly comparable. Here, we use an agglomeration index proposed by Ellison and Glaeser (1997), which can accurately measure the agglomeration within an industry and more important, comparable across industries. Kazakhstan’s manufacturing sector has generally exhibited increasing concentration over time, but there is an increasing polarization in the agglomeration patterns. The average agglomeration index for the manufacturing industries has been increasing in 2016–18, with the index in 2018 higher than in 2010, indicating more concentration overall. In addition, there has been more industrial agglomeration in the northern regions and less in the southern regions. At the industry level, transport is on average concentrated throughout the study period and became more concentrated over time, while the clothing and apparel industry is on average dispersed and becoming more dispersed over time. In addition, the agglomeration patterns are becoming polarized as more industries become agglomerated while others become more dispersed, with fewer industries in the middle. Similarly, there has been a trend of growing agglomeration in the services sector. Relative to the manufacturing sector, the average agglomeration index for the service sector is higher, indicating higher agglomeration. The patterns in each region have remained stable over time, with strong agglomeration levels in the northwestern regions. Professional services, such as legal and accounting services and head office and management consultancy services, are the most concentrated industries and became more concentrated, while the food and beverage services industry was among the least concentrated and became even more dispersed. Wholesale and retail trade became more concentrated. Kazakhstan Country Economic Memorandum 14 The patterns of agglomeration among the regions changed between 2010 and 2018, especially in the south and north. The agglomeration in 2010 is higher for regions in the south, but this changed in 2018, where there is more industrial agglomeration in the northern regions and less in the southern regions (Figure 1.9). In 2018, high levels of agglomeration remain around the major cities such as Almaty, Astana, and Shymkent. Figure 1.5 Manufacturing agglomeration, increasing in northern regions in 2018 2010 2018 Source: World Bank (2020c) Note: To obtain the regional patterns, a weighted average for the region is calculated from the industry’s agglomeration index using the region’s employment in that industry as weights. The regions are grouped into four bins based on their average indexes, which are colored from light blue (dispersed) to dark blue (concentrated). The bins are: [-1,0), (0, 0.02], (0.02, 0.05], and (0.05, 1]. Agglomeration forces can reinforce themselves, so concentrated industries become more concentrated over time, while dispersed industries become more dispersed over time. The transport (nonmotor vehicle) industry is among the top three concentrated industries in 2010–18 and became more concentrated between 2010 and 2018. The other two most concentrated industries are textiles and wood products, but they did not increase in concentration over time. Instead, electrical equipment and machinery repair and installation experienced the largest increases in their indexes. In contrast, the clothing and apparel industry is less concentrated and became more dispersed over time. It is among the top three least concentrated industries, alongside basic metals and motor vehicles. Clothing and apparel is also among the top three industries that became more dispersed between 2010 and 2018, but rubber and plastics and metal products exhibited a larger decrease in their indexes, indicating more dispersions. The changes in agglomeration patterns in Kazakhstan share some similarities to those in China. The agglomeration patterns in the concentrated industries can be explained by regional employment shares. In the most concentrated industry, the textiles industry, most of the employment is located in the Turkestan region, which has 55 percent of the industry’s average employment between 2010–18.23 Most other regions barely have an employment share of more than 5 percent. Similarly, in the next most concentrated industry—wood products—East Kazakhstan and Almaty city have almost 50 percent of the average industry employment share over 2010–18. While the regional employment shares are not concentrated in just one region in wood products relative to textiles, most other regions also have less than 5 percent of the average employment share in the wood products industry. Overall, there has been an increase in agglomeration across the services sector between 2010 and 2018. The average agglomeration index for the services sector experienced a slight decrease in 2012 Kazakhstan Country Economic Memorandum 15 and a recovery in 2017, followed by a drop again in 2018. The service sector has an average index of 0.046, close to the threshold for a concentrated pattern (0.05). The index passed this threshold in 2017. In relative terms, the services sector is relatively more concentrated than the manufacturing sector. While there is an increasing trend in the average agglomeration, the agglomeration patterns in the regions remain relatively unchanged over time (Figure 1.9), with higher concentration levels in the northwestern regions. It is noteworthy that the concentration of services sectors is not restricted to regions near large cities (such as Almaty city) but spread throughout Kazakhstan. Figure 1.6 Agglomeration patterns of service industries in regions remains relatively unchanged over time 2010 2018 Source: World Bank 2020c. Note: To obtain the regional patterns, a weighted average for the region is calculated from the industry’s agglomeration index using the region’s employment in that industry as weights. The regions are grouped into four bins based on their average indexes, which are colored from light blue (dispersed) to dark blue (concentrated). The bins are: -1,0), (0, 0.02], (0.02, 0.05], and (0.05, 1]. The two most concentrated services industries have most of their employment centered around just a handful of regions. For water transport, the most concentrated services, the Mangistau region had 60 percent of the average employment in the industry in 2010–18. This is natural given that Mangistau is the only region with water access to the Caspian Sea. The Pavlodar region and East Kazakhstan each have an average employment share of 10 percent in water transport, and there are almost no workers of this industry in the other regions. For the postal and courier activities, the employment is concentrated in the two largest cities and transport hubs in Kazakhstan—Almaty city and Astana. The average employment share in 2010–18 averaged 70 percent in Almaty city and about 20 percent in Astana, with minimal employment in other regions. The changes in concentration of services, especially in the professional and administrative services, are driven by shifts in the employment shares between certain regions. In the rental and leasing services industry, there is a large reduction in the employment share in the Mangistau region from just over 50 percent in 2010 to more than 20 percent in 2018. The employment shares increased in Aqtobe region and Atyrau region by more than 10 percentage points over the same period. But the employment shares are still concentrated in these three regions in 2018, which make up almost 60 percent of the industry’s employment share. In the head offices and management consultancy services, the shares of employment have shifted between Astana and Almaty city. In 2010, Astana had about 20 percent of Kazakhstan Country Economic Memorandum 16 the industry’s employment, while Almaty city had 50 percent. This changed sharply in 2018, as Astana increased its employment shares to almost 50 percent, taking some share from Almaty city and other regions. Despite the reductions in employment share from 2010, Almaty city still had a large share of the industry’s employment of 37 percent in 2018. The increasing agglomeration patterns could be a result of policies and programs that encourage firms to locate in particular areas and that reduce the costs of agglomeration. New and existing firms can choose to locate in a particular region if it has a good business environment, which is manifested through good local governance, amenities, and infrastructure and through place-based government policies. In addition, these firms will want to agglomerate in regions as long as they have access to markets and inputs. Low transport and trade costs facilitated by good infrastructure can help firms to source inputs and send their final products more cost effectively. Free internal movement of workers will ensure that firms can locate in a region and still have access to workers from other parts of the country. The World Development Report 2009 provides a useful framework to understand the different policy instruments available to Kazakhstan (Table 1.2). Further analysis and examination are needed to understand whether Kazakhstan has already implemented any such policy instrument and its effectiveness in affecting the agglomeration patterns during the 2010s. Table 1.2 Policy instruments to encourage agglomeration Institutions Infrastructure Interventions Explanation Policies that are spatially Policies and investments that Programs that are blind in their design and facilitate the movement of goods, spatially targeted should be universal in their services, people, and ideas locally coverage Policy Improve land regulations, Improved investments in transport Programs (such as fiscal levers reduce trade barriers for infrastructure and communication incentives or special (examples) imported inputs, systems; reducing institutional economic zones) for strengthen education and barriers to ease the movement of manufacturing firms to labor policies to goods; removing any barriers that locate in particular encourage a good labor prevent internal migration of regions supply workers Source: Adapted from World Bank (2009). Conclusion After years of declines in both labor productivity and total factor productivity, productivity in Kazakhstan increased, but its growth remains muted, limiting economic growth. Changes in total factor productivity have been driving a large part of the economic growth patterns in Kazakhstan over the past decades. In the 2015–18 period, annual total factor productivity growth—despite no longer being negative—remained low, at 1.1 percent in manufacturing and 0.4 percent in the services sector. Even so, significant differences exist across sectors and regions, with the regions with large extractive industries showing higher rates of productivity growth than regions focusing more on manufacturing or services. Both a lack of firm upgrading and barriers to reallocation played a role in this muted productivity growth. In manufacturing, most of the growth between 2015 and 2018 can be attributed to firm upgrading, while the components representing reallocation and business dynamism are low or negative. In services, some intrasectoral reallocation improved between 2015 and 2018, but low degrees of firm upgrading and misallocation between services sectors dragged down productivity. Kazakhstan Country Economic Memorandum 17 Regional comparisons show similarly muted or negative contributions in key economic regions outside those with large extractive industries. Policies that focus on firm upgrading and removing distortions in product and output markets can be expected to increase productivity. The low or negative contributions of the “withinâ€? and “betweenâ€? components compared with countries with high productivity growth suggest that there is significant scope for improving firm capabilities through the upgrading of technology and management practices as well as improving the regulatory environment to remove barriers to growth of productive firms. Agglomeration increased in the 2010s in Kazakhstan on average in the manufacturing and services sector, which can have positive effects on firm productivity. Agglomeration economies can increase firm productivity, as firms have access to specialized inputs and labor and enjoy knowledge spillovers from their neighbors. In addition, agglomeration is a self-reinforcing phenomenon, and agglomerated sectors and regions tend to become more agglomerated over time. But while the manufacturing and services sector exhibited increasing agglomeration on average, the agglomeration patterns among the industries were uneven. Certain manufacturing and services industries became more agglomerated, but others became less agglomerated over the period, which suggests that the productivity benefits from agglomeration may only accrue to firms in those sectors, firms connected through buyer-supplier linkages, or firms located in regions with more agglomeration. References Eslava, Marcela, John C. Haltiwanger, and Alvaro Pinzón. 2019 Job creation in Colombia vs the US: “up or out dynamicsâ€? meets “the life cycle of plants.â€? Working Paper 25550. National Bureau of Economic Research. Hsieh, Chang-Tai, and Peter J. Klenow. 2014. "The life cycle of plants in India and Mexico." The Quarterly Journal of Economics 129 (3): 1035–84. Kuntchev, Veselin, Rita Ramalho, Jorge Rodríguez-Meza, and Judy S. Yang. 2014. "What have we learned from the Enterprise Surveys regarding access to finance by SMEs." Policy Research Working Paper 6670, updated May 2014. World Bank, Washington, DC. Medina, Leandro, and Friedrich Schneider. 2018. Shadow Economies Around the World: What Did We Learn Over the Last 20 Years? Washington, DC: International Monetary Fund. World Bank. 2018. Kazakhstan Country Economic Memorandum: Reversing Productivity Stagnation. Washington, DC: World Bank Group. World Bank. 2020a. Drivers of Productivity Growth in Kazakhstan. Report prepared as part of the Joint Economic Research Program between the Government of Kazakhstan and the World Bank Group. World Bank. 2020b. Insights into the Private Sector of Kazakhstan. Report prepared as part of the Joint Economic Research Program between the Government of Kazakhstan and the World Bank Group. World Bank. 2020c. Agglomeration Patterns in Kazakhstan. Report prepared as part of the Joint Economic Research Program between the Government of Kazakhstan and the World Bank Group. Kazakhstan Country Economic Memorandum 18 2 Migration and domestic mobility Because economic growth proceeds unevenly across space, it is usually accompanied by elevated migration. Workers are drawn to where wages and demand for their services are highest. Over time, employers in growing sectors pay higher wages to attract needed workers. In the modern era, cities have grown as the meeting places of these workers and employers, and all advanced economies have rapidly urbanized as a result (Figure 2.1a). The same trends, though restrained, are present in Kazakhstan today. Since the country’s independence, the agricultural sector has become far more productive, and both industry and services now account for a much higher share of employment. But the rate of domestic migration is still modest, and structural transformation is proceeding slowly. The country must also contend with the distortions of previous decades’ central planning policies that intentionally “flattenedâ€? its economic geography, leaving many communities isolated and limiting their growth prospects. In the next section, this chapter reviews common patterns seen worldwide in the links between migration and structural transformation and between urbanization and rising income. It then analyzes Kazakhstan’s path dependency in migration patterns, before diving deep into two key impediments that the country still faces: the high cost of housing in the most dynamic urban centers and the continued practice of regulating internal migration through the registration system. Migration and structural transformation Migration—a natural consequence of structural transformation and rising prosperity Migration is instrumental in maintaining long-term economic expansion, as an extensive literature attests (see, for example, Hsieh and Klenow 2009; Morten 2019). The key driver is self-interest: employers offer higher wages when the workers they need are scarce, and the higher wages offered in growing sectors incentivize workers to move to where their work is in greatest demand. Although those forces operate at the level of individual workers and firms, cumulatively they drive an economic transformation that coincides with the public interest. When added together, individual migration decisions push toward a balance between where workers live and where job opportunities (and incomes) are rising, which reduces unemployment and raises productivity. The urban population share typically increases rapidly as a country’s income rises The places that tend to pull in workers are not randomly spread out. Workers concentrate in the places and sectors with the fastest economic growth. A rapidly expanding manufacturing or services sector is common in high-growth economies, at times generating an increase in demand for workers that equals or exceeds population growth. Both industry and services disproportionately depend on economies of scale, concentrated infrastructure, and deep labor markets available only in urban areas. This reliance creates a nearly universal pattern of rising urban population concentration as countries become richer (Lucas 2015) (Figure 2.1a). Following this logic, Lall and Lebrand (2020) show that countries in Central Asia where people are not mobile will experience higher spatial wage inequality and, relatedly, lower welfare from the new transport corridors and trade policies. As a result, in Kazakhstan, real wages are likely to grow five times more in and around Almaty than in northern areas close to the Russian border. The development benefits from closer economic relations with China will depend on Kazakhs’ willingness to move to places like Almaty. Kazakhstan Country Economic Memorandum 19 When urbanization proceeds rapidly, the fate of rural areas is a common concern among policymakers and the public. But the movement of workers from low-productivity rural areas to high-productivity urban ones also drives regional economic convergence, powering catch-up growth in lagging areas (Box 2.1).24 High-income countries such as Germany, Japan, the United Kingdom, and the United States have all followed a similar development path: as rising wages and higher productivity attract workers and their families to cities and mid-size towns, workers in rural areas become scarcer (Figure 2.1a). That scarcity can catalyze rural development by incentivizing employers to change in two important ways. First, with fewer potential workers available, employers in rural areas offer higher wages to retain workers and hire new ones. Second, the increasing labor costs stimulate greater investment in capital and technology, in turn boosting productivity. In high-income countries that have completed the transition, total output has dramatically risen on average, with 5 percent or less of the labor force remaining in agriculture. Why does rural agriculture not absorb these workers? Although agricultural output often rises quickly in rapidly growing economies, that growth does not necessarily lead to increasing employment. Around the world, a continuous stream of technological improvements since the early 1900s has swiftly boosted agricultural production. But because rising productivity is responsible for much of the output growth, fewer workers are typically required to produce the same output, so in recent experience, the agricultural sector alone has rarely expanded quickly enough to raise labor demand and increase wages (Figure 2.1b). Rural-to-urban migration also commonly leads to a higher average standard of living, coupled with rising wages and income. Because providing high-quality public services is more cost-effective in urban areas and because urban wages tend to be higher, urbanization is strongly associated with both higher nonmonetary living standards and lower monetary poverty (Acosta et al. 2007; Adams and Page 2005; Ravallion 2002; Ravallion et al. 2007). Figure 2.1a Urban population as a share of Figure 2.1b Relationship between total population, by country, 1820–2015 agriculture’s share of employment and Percent GDP per capita, by country, 1820–2015 100 100 Share of Agricultural Employment 80 80 Percent Urban 60 60 40 40 20 20 0 0 1820 1851 1882 1910 1941 1971 2002 6 7 8 9 10 11 Year Log GDP per Capita Germany India Japan Kazakhstan Argentina Bolivia China Kyrgyz Republic Korea, Rep. of France Japan Netherlands Tajikistan Türkiye United States Source: World Development Indicators database. Kazakhstan Country Economic Memorandum 20 Structural transformation is under way in Kazakhstan The global economic trends hold true in Kazakhstan. Agricultural output rose by more than 26 percent between 1991 and 2019, while agriculture’s share of employment fell from 39 percent to less than 16 percent. In 1992, agriculture value added as a share of Kazakhstan’s GDP stood at more than 23 percent, but by 2010, it had fallen below 5 percent and has remained there ever since. And as agricultural productivity rose (figure 2.2), workers shifted to the services sector, whose share of GDP rose from 25 percent in 1991 to more than 55 percent in 2019. Figure 2.2 Agriculture value added per worker, 1991–2019 8,000 7,000 6,000 USD, constant 2010 5,000 4,000 3,000 2,000 1,000 - 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Kazakhstan Upper-middle-income average Source: World Development Indicators database. Box 2.1 Lessons from regional income convergence in the United States and other advanced economies From the 1880s to the 1990s, the rate of convergence in per capita income between states in the United States consistently averaged 1.8 percent a year. The main driver of this remarkable achievement was internal migration: people flowed from poor to rich states, raising living standards in both. Affordable housing and jobs-driven growth propelled broadly shared prosperity by encouraging people to live where their skills were most in demand. Kazakhstan might learn from that example, but it might also learn from the US failure to maintain that rate. From 1990 to 2010, US convergence in per capita income was less than half the historical norm. Declining mobility, caused by growth-unfriendly urban policies (such as zoning restrictions that reduce available housing in the most sought-after areas), has been the primary culprit (Ganong and Shoag 2017). As a result, interregional income inequality in the United States is at its highest level in a century (Kemeny and Storper 2020). For similar reasons, in Europe, the long-term trend of regional convergence after 1900 was replaced by divergence after 1980 (Rosés and Wolf 2018). Internal mobility is most sustainably achieved through growth in housing that is affordable and accessible to low- and middle-income families. In Kazakhstan, reducing the administrative impediments to movement, such as registration requirements, would improve mobility and, as a consequence, regional convergence. Empowering business models for rental housing and supporting the emergence of deeper rental housing markets would lower barriers to aspiring migrant workers and establish a bridge for a larger share of the population to enter the more dynamic labor markets in urban areas. Kazakhstan Country Economic Memorandum 21 Past migration trends and path dependency Several phases of migration throughout Kazakhstan’s history have led to a distinctive geographic pattern of development that persists to this day. The shift from a sparsely populated and largely nomadic society to one with a growing, settled population and an economy dominated by agricultural production—and later, fossil fuels—was relatively recent. It took place first with the expansion of the Russian Empire into the region and later with the incorporation of Kazakhstan as a republic of the Soviet Union. The transition was accompanied by a large influx of Russian settlers to areas thought to be well suited to agriculture, including in the northern parts of Kazakhstan that previously supported few permanent settlements. That initial wave was followed by state-directed mass migration throughout the Soviet Union from the 1920s through the early 1950s. During that time, many populated areas of today’s Kazakhstan were founded, often in previously remote areas. By the early 1950s, about 260 new towns in the Soviet Union were created in remote locations that before had been totally uninhabited, and other new towns and cities were built by centrally planned expansion of very small outlying settlements, totaling about 800 (Shkvarikov, Haucke, and Smirnova 1964). Kazakhstan’s history of directed migration led to path dependency and spatial inefficiency Because the new settlements’ locations were not governed mainly by natural population growth and economic considerations, their economic fundamentals were often unsound, in many cases intentionally. Under Stalin, from 1922 through 1953, new Soviet settlements were commonly isolated and made desolate on purpose, especially in Kazakhstan, which hosted several of the largest gulag labor camps in the Soviet Union.25 Although some of those settlements have since thrived despite their challenging origins, other places have not prospered economically, following strong path dependency created by large-scale central planning. After the reforms of the 1950s and 1960s, the founding of totally new settlements became much less common. Even so, modern Kazakhstan remains dotted with monotowns—towns whose economy is dominated by a single industry or company. Today, the aging housing stock in these areas often has low resale value, leaving owners with less wealth than residents in high-demand primary cities. In 1953, another wave of state-directed migration to Kazakhstan was launched under the auspices of the Khrushchev-era “virgin landsâ€? campaign to boost agricultural production throughout the Soviet Union. The modern capital of Kazakhstan (renamed Tselinograd at the launch of the campaign) served as the administrative center of the effort. During the first three years, the rapid influx of more than 300,000 agricultural workers in remote areas of Kazakhstan led to food and housing shortages (Durgin 1962). Following an initial boost in agricultural production, gains proved illusory as intensive monoculture extensively depleted soil nutrients. Much of the area cultivated under the virgin lands campaign remains severely degraded to this day, limiting the economic prospects of modern settlements in these areas. The bursts of directed population migration and investment have led the allocation of both physical and human capital to persist over the long term.26 Specialized Soviet prison camps, which accounted for a large minority of new settlements in Kazakhstan during the 1920s–50s, offer a compelling example. Today, areas around settlements that incarcerated a larger share of “ highly educated prisonersâ€? among prisoners are more economically prosperous than otherwise similar places, as measured by wages, education, firm productivity, and nighttime light emissions per capita. Toews and Kazakhstan Country Economic Memorandum 22 Vézina (2020) suggest that this effect is due to the high education levels prevalent among political prisoners during that time. The artificial spatial allocation of people and capital in Kazakhstan proved unsustainable. In the decade following the virgin lands campaign, Soviet demographers reported a persistent flow of people from remote and rural parts of the Soviet Union to urban areas and particularly to primate city areas. The harsh conditions in rural Kazakhstan at that time drove more than half the trained specialists sent to designated virgin lands to leave soon after arriving (Durgin 1962). Partly as a consequence, Kazakhstan’s urban areas were among the fastest growing in the Soviet Union between 1959 and 1970: they grew 61 percent, more than double the 24 percent in rural areas (Harris 1971).27 But by the 1970s, the republic’s net in-migration had reversed, with a massive surge in out-migration of more than 2 million people after independence in 1991 (Figure 2.3). Figure 2.3 Net in-migration, 1962–2017 Thousands 1,000 500 0 -500 -1,000 -1,500 -2,000 Source: World Development Indicators database. Administrators used to rely heavily on the propiska system To actively manage migration and the pace of urban growth, Soviet Union administrators used the propiska (registration) system, akin to domestic passports for citizens to travel within the country’s borders. Registration at a person’s place of residence was used to regulate migration and to administer public services. It was perennially controversial, especially given its Russian Empire roots as a system for managing serfdom (and later for limiting peasant migration). The Bolsheviks strongly opposed it during the revolution and abolished it in 1917. But the propiska was reintroduced under Stalin with the closure of Moscow and (what was then) Leningrad in 1931, followed by full national propiska registration of all Soviet citizens in 1932. Officials’ policy justification was the need to implement central planning and to control the movement of groups that the state viewed as dangerous. In 1939, a second group of cities was closed to new migration, including Kyiv and Sverdlovsk, and in 1956, a third group, including Almaty and Tashkent. From that time, access to urban housing—especially in the closed cities—was restricted and allocated by the state based on employment assignments. Kazakhstan Country Economic Memorandum 23 Privatization after the breakup of the Soviet Union left Kazakhstan with one of the highest home ownership rates in the world Soviet housing policy left clear marks on the economic geography of Kazakhstan. To this day, the country has one of the highest home ownership rates in the world, a legacy shared by many former Soviet republics (Figure 2.4). In the Soviet Union, housing was centrally allocated, with location decisions based largely on place of employment (Struyk 2000). Widespread housing privatization in the countries of what had become the Commonwealth of Independent States and among other formerly centrally managed economies in Central and Eastern Europe led to much higher home ownership rates in those areas than in Western Europe. None of the formerly centrally managed economies has a rental market share that approaches the European Union average.28 And even in ECA, only Romania’s home ownership rate was higher in 2015. Privatizing government-owned housing after the breakup of the Soviet Union provided significant benefits for most recipient households. Subsidies for the purchase price were moderately pro-poor (Struyk 2000). Under privatization, sitting tenants had the right to purchase their units, typically at a substantial discount or, in some cases, for free, except for a nominal processing fee. The rate of take- up of privatization offers was high. And due in part to the appreciation in home values that followed in many countries of the former Soviet Union, a popular view emerged that housing was a more preferable means of storing wealth than other investments or bank saving (Seitz 2021; Struyk 2000). Figure 2.4 Home ownership rate, by country, 2015 100 90 80 70 60 50 40 30 Austria France Iceland Latvia Serbia Denmark Finland Italy Greece Spain Netherlands Norway Hungary Russian Fed. Romania Germany Ireland Cyprus Belgium Luxembourg Slovenia Bulgaria Poland Croatia Malta Estonia Slovakia Lithuania N. Macedonia Switzerland Sweden European Union Portugal Kazakhstan Czechia United Kingdom Sources: EuroStat, RusStat, and the Household Budget Survey of Kazakhstan. Despite the benefits, home ownership also has drawbacks. A high rate of home ownership, achieved by reducing mobility, directly impedes the functioning of the labor market (Oswald 1996, using cross- country comparisons of unemployment). And the high moving costs related to home ownership are a severe impediment to regional mobility, ultimately leading to labor market dysfunction and lower productivity (Dohmen 2004, using a related theoretical model). Rising home ownership rates commonly precede and are strongly associated with higher unemployment, due primarily to lower worker mobility (Blanchflower and Oswald 2013, using global comparisons).29 Kazakhstan Country Economic Memorandum 24 Recent migration patterns Domestic migration is below the international average, depending on the year and source of measurement Different sources report varied but qualitatively similar magnitudes of migration in Kazakhstan (Annex 2A). Survey data for 2014–19 suggest that, by international standards, domestic migration rates in Kazakhstan range from moderately low to average. Rural-to-urban migration accounted for only one-fifth of urban population growth between 2010 and 2015 (OECD 2017).30 About 1.7–2.3 percent of the current population moves between regions within the country each year.31 That share is much lower than in many advanced economies: 14 percent in Canada, 11 percent in the United States, and even the slightly higher 2.6 percent in the Russian Federation, with which Kazakhstan still shares many structural characteristics (Bell et al. 2015). Administrative records, which define migration differently from surveys, also suggest a modest pace, with a domestic rate of about 5 percent of the population over the five years preceding 2019, ahead of Indonesia and Mexico but slightly behind other large countries, including China (6.7 percent), and well behind the most mobile large countries, such as Brazil (10 percent) and Australia (more than 21 percent). United Nations growth projections predict that the urban share of Kazakhstan’s population will be about 65 percent by 2050, well short of the government’s official target of 70 percent (UN 2019). Domestic migration is concentrated in a few urban districts in Kazakhstan where poverty rates are lower and average per capita consumption far higher (Figure 2.5; Aldashev and Dietz 2014.) In the 2014 national Labor Force Survey, about 58 percent of migrants had relocated to the top 10 destination districts, rising to about 64 percent in the 2019 survey. Of the 172 districts ( rayons) with data, 27 recorded no incoming migrants in 2014; by 2019, 60 did. In 2019, the Tselinograd district surrounding Astana was the clear destination of choice for domestic migrants, attracting about 40 percent of them. In 2015, the share of migrants going to Astana overtook the share going to Almaty, which fell from 16 percent in 2014 to 8 percent in 2019. Figure 2.5 Share of migrants over the previous five years, by district they moved to, 2019 Source: 2019 Kazakhstan Labor Force Survey. Domestic migrants are much more highly educated than average, and have become more so Although highly educated people make up about 30 percent of the population, they account for more than half the migrant population. Destinations are even more concentrated among migrants with Kazakhstan Country Economic Memorandum 25 education above the secondary level (figure 2.6) than among other migrants. In 2019, the top 10 districts were the destinations of about 77 percent of all highly educated migrants, compared with 46 percent of less educated migrants. Less educated people have become less mobile over time. In 2014, about 4 percent of less educated people were migrants, falling to about 3 percent in 2019, according to Labor Force Survey data. This is opposite the increasing trend for the highly educated, about 8 percent of whom were migrants in 2014, rising to 9 percent in 2019. Those trends correspond to recent demographic statistics: in Astana, the share of the workforce that had completed a tertiary degree increased from 43 percent in 2006 to 51 percent in 2015. In Almaty city, the share grew from 45 percent to 61 percent over the same period. Figure 2.6 Top five destination districts (rayons) of highly educated migrants, 2014–19 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 2014 2015 2016 2017 2018 2019 Almaty Tselinogradskiy Atyrau Shymkent Aktobe Source: World Bank, based on Labor Force Survey of Kazakhstan. Official data on individual registration offer an alternative measure of domestic migration. These results provide a slightly different picture than estimates based on survey data. They suggest that the number of new registrations in Almaty is greater than the number accounted for as new migrants in survey data.32 The difference may reflect in part an increase in the registration rate due to 2010 policy changes. But the two sources agree on destinations, with migration concentrated in the two largest growth poles—Astana and Almaty city—followed at some distance by Shymkent and Mangystau (Figure 2.7). Migration trends in Kazakhstan closely follow spatial patterns of income and consumption inequality. National income inequality has been largely stable and moderate by international standards, with a Gini coefficient of 0.35 in 2018. But at the local level, monetary wellbeing is much higher in the most prominent migration destination districts than in migrant-sending areas. Those spatial gaps in monetary welfare can be decomposed into two sources.33 The first reflects differences in the composition of people living in rural versus urban areas (for instance, differences in the average age, endowments, and education level of the resident populations). The second reflects the portion of inequality that comes from simply living in a particular place and cannot be accounted for by differences in people’s observable characteristics—this can be thought of as a “returns to placeâ€? effect. Kazakhstan Country Economic Memorandum 26 Figure 2.7 Net new residential registrations, by region and city of republican significance, 2019 Thousands 50 40 30 20 10 0 -10 -20 -30 Net New Registration (1000s - 2019) Source: World Bank, based on Kazakhstan Bureau of National Statistics. The decomposition of spatial inequality into those two sources highlights the importance of access to the more dynamic labor markets in urban agglomerations, especially for the highly educated. The portion of the income gap between rural and urban areas accounted for by place rather than by the composition of the resident population is positive—about 39 percent at the mean—and rises with income. The returns to place effect between rural and urban areas explains about 15 percent of the gap at the 5th income percentile but about half the gap at the 95th percentile (Figure 2.8). As would be predicted from the direction of migration flows, the size of the returns to place effect is much larger for Almaty and Astana than average, even when compared with the average rural–urban gap. In Almaty, the returns to place effect accounts for 55 percent of the gap on average, again rising with income. In Astana, the place effect accounts for the entire gap on average. Also, in Astana and in contrast to the rest of the country, the resident composition effect contributes to closing the overall income gap. Kazakhstan Country Economic Memorandum 27 Figure 2.8 Income gap decomposed into returns to resident composition and returns to place, by income percentile and for the two largest cities, 2019 Oaxaca-Blinder Decomposition of Income Gap 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 -0.05 -0.10 Total 5th 25th 50th 75th 95th Astana Almaty Resident composition Place effect Source: World Bank calculations, based on Labor Force Survey of Kazakhstan. Note: Oaxaca-Blinder decomposition. Housing and housing costs High housing costs constrict rural-to-urban migration. Living in Kazakhstan’s cities is typically much more expensive than living in its rural areas—the country is an outlier in this respect, with one of the largest urban–rural gaps in the cost of living in ECA. The cost of housing consumed (as measured by imputed rents) is 310 percent higher in Almaty city than the national average, and about 460 percent higher in Astana (Seitz 2021). The large differences are due partly to rapid and volatile appreciation of home values in recent years. Housing prices adjusted for inflation—measured mostly in urban areas—rose sixfold between 2001 and 2016, according to official statistics. In Astana, real housing prices were three times higher in 2016 than in 2001, and in Almaty, prices more than quadrupled over the same period.34 Although housing prices in those two cities have been rising more slowly than the national average, they are remarkable because those areas were already much more expensive than other parts of the country. Urban housing price growth has also been extraordinarily unstable. In real terms, prices rose by an average of more than 50 percent a year between 2001 and 2007, only to fall by more than 60 percent over the following five years (Figure 2.9). Kazakhstan Country Economic Memorandum 28 Figure 2.9 Real housing sale prices in Almaty, Astana, and nationally, 2001–16 2001 = 100 1100 1000 900 800 700 600 500 400 300 200 100 Almaty Astana National Source: World Bank, based on data from Kazakhstan Bureau of National Statistics. Note: 2001 = 100. People’s sensitivity to these price changes relative to their changing incomes will determine the urbanization trajectory. An analysis of housing demand responsiveness to changes in income showed Kazakhstan having slightly higher average sensitivity than other countries, with an elasticity to income of about .66 (Seitz 2021), meaning that housing demand would be expected to rise by 34 percent less than any average increase in income. And as housing prices rise, theory dictates that housing demand will fall.35 In Kazakhstan, housing demand is quite responsive to price, with a national elasticity of about −.85 (Seitz 2021). Taken together, the results indicate that price elasticity dominates income elasticity in absolute terms. So, if both incomes and prices rose by 10 percent over the coming years, the price effect would overwhelm the income effect, and net demand for housing would fall (Figure 2.10). Figure 2.10 Propensity to consume additional housing in response to rising welfare Source: World Bank, based on Kazakhstan Household Budget Survey. Kazakhstan Country Economic Memorandum 29 Rapid price appreciation in Almaty and Astana eroded housing affordability. In 2015, imputed rent accounted for 28 percent of urban household consumption on average, and 36 percent of urban households allocated more than 30 percent of their budget to housing (a common affordability threshold in the housing literature). Imputed rent accounted for 36 percent of total average consumption in Astana and 44 percent in Almaty city—the most expensive urban markets (Figure 2.11). In Astana, more than 66 percent of households live in housing they could not afford if they did not already own it, as do more than 77 percent in Almaty city, despite the highest average incomes in the country. Figure 2.11 Actual budget share allocated to rent versus budget share simulated using average urban rent, by region and city of republican significance Percent 60 50 40 30 20 10 0 Actual budget share Assuming urban average Source: World Bank, based on 2015 Kazakhstan Household Budget Survey. In contrast, rural areas of Kazakhstan are almost always affordable. In 2015, housing accounted for less than 30 percent of spending on average in virtually every rural area. Housing accounted for only about 11 percent of consumption in rural Kazakhstan and even less in the most rural parts of the country, including Zhambyl and Kyzlorda. A simulation of a rural household moving to an urban area offers a useful way to gauge the implications of housing costs for migration. On average, rural households that relocate to the median urban area could expect to spend nearly 20 percent more of their current total budget on housing alone, before considering differences in the cost of other goods and services. In about half of Kazakhstan’s regions, the relocation would mean the share of consumption allocated to housing would more than double. In every region but Mangystau and the cities of Almaty and Astana, the average imputed rent would absorb more than 30 percent of consumption, regarded as the affordability threshold. That approach has a weakness, however, in that it does not consider the higher incomes prevalent in urban areas. Households that relocate would expect higher incomes to compensate for a portion of the increase in the cost of living. Even so, although average income was 66 percent higher than the national average in Almaty and 92 percent higher in Astana, the cost of living would still be 190 percent higher in Almaty and 240 percent higher in Astana (Seitz 2021). So, despite the higher income that Kazakhstan Country Economic Memorandum 30 rural households would expect from migrating to those three urban areas, for most households the high cost of living would more than exceed the expected monetary benefit (Figure 2.11). Migration is further complicated by low home values in rural and lagging areas. That challenge is linked to the age of the housing stock, which varies widely by region. Akmola, East Kazakhstan, Zhambyl, Karaganda, Kostanay, and Pavlodar have older and more deteriorated housing, on average, than the largest cities. In Karaganda, more than 60 percent of housing was constructed in 1970 or earlier (Figure 2.12). At the other extreme, Atyrau and Mangystau, with more recent development related to the oil and gas sector, have much newer housing than average (but far fewer total units and lower population density). Figure 2.12 Year of housing stock construction, by region and city of republican significance, 2018 Percent 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1970 or earlier 1971-1975 1976-1980 1981-1985 1986-1990 1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 2016 2017 Source: World Bank, based on Kazakhstan National Bureau of Statistics. According to UN projections, population growth is expected to continue its current downward trajectory, falling below 1 percent a year by 2030 (UN 2019). As population growth slows, Kazakhstan may see increasing pressure on shrinking areas with aging housing, especially in monotowns and other areas already facing limited growth prospects. Areas struggling with rapid, large-scale population declines face self-perpetuating cycles of home values falling and dwellings being abandoned, which can greatly reduce the quality of life for the remaining residents. Modern residency registration Kazakhstan still has a residency registration system, though no longer officially called propiska. Nevertheless, the system hampers internal migration—particularly by the low-income population and those receiving financial support from the state. For those who are unregistered, the system can lead to exclusion from many social services and benefits. Migration is sometimes impeded by residency registration propiska After gaining independence, Kazakhstan reformed the propiska system to shift its purpose from authorizing residence to notifying the state of a person’s local residence. The country retired the official Kazakhstan Country Economic Memorandum 31 use of the name “propiskaâ€? (Radio Azattyk 2012). During the early years following independence, adherence to registration requirements was lax, as the state encouraged internal migration to lift the population out of poverty (Tukmadiyeva 2016). But restrictions gradually tightened beginning in the 2010s. In policymaking, narratives declaring a need to control “spontaneous migrationâ€? became more prevalent (Tengrinews 2012; Zakon.kz 2014). In 2010, Almaty introduced a “sanitary normâ€? of at least 15 square meters of living area per person— a restriction expressly justified to reduce the migration of low-income people (Zakon.kz 2010). In 2016, a package of “antiterrorism lawsâ€? to combat “uncontrolled migration and illegal settlementsâ€? was introduced in Kazakhstan (Sputnik Kazakhstan 2016). The laws cut the time allowed to citizens moving inside the country to report their location from 90 to 10 days and introduced temporary registration,36 in addition to permanent registration37 (Annexes 2B and 2C). They also introduced administrative penalties on property owners for hosting unregistered tenants.38 According to the Ministry of Internal Affairs, these residency registration procedures were aimed at gathering information about population movement, partly for planning infrastructure and social services. Information on the registration of citizens is used by state and local executive bodies (akimats), in particular for planning the socioeconomic development of regions, settlements (the need for hospitals, schools, educational institutions, and others), forecasting the development of urban and transport infrastructure, for kindergartens, schools, and the like, in the calculation of targeted social assistance and in other areas. Galina Sarsenova, deputy chairperson, Migration Committee of the Ministry of Internal Affairs. June 5, 2020 According to statistics provided by the Committee on Legal Statistics of the General Prosecutor's Office, during 2017–19, there were 180,793 documented cases of citizens living without registration or an identity card. In the same period, the police registered only 6,537 cases of penalties for housing owners.39 Police regularly report on investigations and raids to apprehend individuals without residency permits or registrations. The latest reported large-scale government raid was conducted in late 2018, when the police identified 15,300 individuals living without registration and about 1,100 homeowners who allowed people to live without registration (Sputnik News 2018). At the same time, new e-governance systems simplified registration and reduced the bureaucratic burden on citizens. The new systems have eliminated the requirement to withdraw from one’s previous place of registration (vypiska), the registration book of citizens (domovaya kniga), and the address certificate (adresnaya spravka). State agencies are not allowed to request documents from citizens that they can retrieve from the state information systems, including the confirmation of residency registration. Before 2018, people had to pay a state fee to register, but now registration is free (HOLA News 2018). Application for most social benefits has been digitized, and since March 2020, citizens have been able to obtain an approving digital signature (elektronnaya tsifrovaya podpis) online, making it possible for residents to fully register online without physically visiting a public services center (TsON). In 2019, a simplified process of registering citizens through the M-gov mobile application was introduced, which allowed registration using a one-time password without visiting a TsON.40 These measures have improved the quality of state services, increased access to them, and reduced corruption by limiting human interactions. Many continue to live essentially in an illegal status, and their capabilities are limited. [If needed,] people register, but unofficially, there is a corruption component—they ask someone, pay money somewhere. Almaty, May 4, 2020 Kazakhstan Country Economic Memorandum 32 Even so, according to the 2021 Listening to Kazakhstan survey, about 20 percent of people still reported having a friend or relative who does not live where they are formally registered.41 A considerable share of the population of large cities still experiences difficulties registering, which could prevent people from accessing vital social services. Some resort to the black market (Box 2.2). Knowing an unregistered person is strongly correlated with lower socioeconomic status and relying on social assistance. About 36 percent of people believe that registration procedures are a problem for some people, a share that rises to 48 percent among respondents who receive social assistance. In 2017, an estimated 13 percent of Astana’s population were not registered at their actual place of residence (Toqmadi 2020). I, myself, live in a rented apartment. I have a permanent registration elsewhere, but temporary registration at my current residence is out of the question. This causes some inconvenience; for example, we had loud neighbors and were afraid to call the police because we do not have registration. It bothered us a little. The problem is that it is much more complicated in cities than in villages. In villages, people often live in houses for several generations; these houses might lack formal documents, but no one claims them. In cities, no one wants to register residents; social workers are forced to register [people] at their place. For people who rent cheap housing—temporary huts, rooms—registration is a very big problem. Kostanay, April 30, 2020 Box 2.2. Securing registration on the black market When it is not possible to register legally, people find temporary solutions. According to experts interviewed for this study, people in such situations often either register with relatives or turn to the black market, which is vibrant for registrations in Almaty and Astana. The study systematically reviewed online advertisements for black market registrations. Although the size of the black market was smaller in 2020 than in 2014, when it was last reviewed, it remains competitive and open (Tukmadiyeva 2016).1 The price of black market registration varies by type, location, and duration.2 In Almaty, a one-month temporary registration costs 4,000 tenge on average, and the price of a one-month permanent registration (in this context meaning that the individual is not registered elsewhere; “permanentâ€? registration can nonetheless be withdrawn at the discretion of an authorized landlord) started at 5,000 tenge, with the average about 7,000 tenge, depending on city area, with various limitations on maintaining registration beyond one month. About 65 percent of all the online advertisements examined offered registration in the Alatau rayon of Almaty. Registration in Astana is generally less expensive, with temporary registration costing on average 2,800 tenge and permanent registration costing about 5,000 tenge. In Almaty and Astana, a considerable share (30–50 percent) of advertisements were posted by the same accounts, often representing law offices. But the viewership of advertisements for Almaty and Astana differs strikingly. Although the number of advertisements was about twice as large in Almaty as in Astana, advertisements for temporary registration were viewed as much as 10 times more often in Almaty than in Astana, and those for permanent registration, about four times more often. Many black-market advertisements describe the purposes for which the registration can be used. People most often buy a registration for employment and education (schools and kindergartens), and Kazakhstan Country Economic Memorandum 33 somewhat less often to register a car; access mortgage loans, pensions, and social benefits; perform military registration; and obtain or restore identification documents. Although according to the Ministry of Labor and Social Protection, employers are not entitled to demand registration, in online job postings, employers commonly explicitly require “city registration.â€?3 Employer preferences for candidates registered in the city are most prevalent for low-paid jobs, such as salesclerks, plumbers, drivers, nurses, odd-jobbers, and babysitters.4 Notes: 1. See, for example, a website dedicated to registration for a fee: https://propiska- almatyda.kz/#project. 2. The analysis of the black markets of Almaty and Astana was based on content analysis of advertisements posted on market.kz, one of the popular websites for this kind of advertisement. The relevant advertisements were manually identified using the search engine on May 1, 2020. All relevant advertisements were then entered into an Excel table for further analysis. The keywords in Russian and Kazakh languages used for the search were: “propiskaâ€? (“пропиÑ?каâ€?), “registrationâ€? (“тіркеу,â€? “региÑ?трациÑ?â€?), “temporary propiskaâ€? (“временнаÑ? пропиÑ?ка,â€? “уақытша тіркеуâ€?), and “residency registrationâ€? (“региÑ?трациÑ? по меÑ?ту жительÑ?тва,â€? “тұрғылықты жері бойынша тіркеуâ€?). Irrelevant advertisements, repeated advertisements, and advertisements without prices were excluded from the analysis. Since most of the advertisements offered both temporary and permanent registration but indicated only the starting price, this analysis considers the indicated price by default as the temporary registration price, unless the advertisement explicitly stated otherwise. 3. Market.kz, “Rabota v Almatyâ€? (Work in Almaty), https://market.kz/almaty/rabota/?term=пропиÑ?ка. 4. Electronic Government of the Republic of Kazakhstan, answer of the Minister of Labor and Social Protection of the Population of the Republic of Kazakhstan to question no. 586478, December 30, 2019, https://dialog.egov.kz/blogs/all-questions/586478. Social benefits and services, and bureaucratic elements requiring residency registration Many state social services and benefits are based on place of permanent residence registration. Citizens who move to another region must apply at the new place of residence.42 The competent authority generally confirms the applicant’s residency information from state information systems. Social benefits.43 These include state basic pension payments and pension payments by age; benefits related to childbirth and childcare, disability, and loss of the breadwinner’s job; loss of income due to pregnancy and childbirth; loss of income in connection with adopting a newborn baby; loss of income due to caring for a child over one year old;44 social benefits for parents of a disabled child or a disability care allowance; all special medical services and benefits for the disabled carried out in the place of registration;45 targeted social assistance for individuals and families whose monthly average per capita income is below the official poverty threshold; and a one-time state monetary compensation to citizens affected by nuclear tests at the Semipalatinsk nuclear test site.46 Basically, the absence of registration is revealed when a person applies for some kind of social service, and then he is fined. Many people do not even know that they were deregistered. For example, there was a case when a man turned to the TsON for help, and it turned out that he had been deregistered a long time ago and had to pay a fine. Kostanay, April 30, 2020 Public housing. Selected vulnerable populations and public servants are eligible for public housing from the state housing fund or in housing leased by a local executive body. Access to the public housing waiting list is granted on the basis of confirmation of residence in the settlement.47 To be Kazakhstan Country Economic Memorandum 34 considered for public housing in cities of republican (that is, national) significance and the capital, an applicant must be a permanent resident of the respective city for at least three years.48 Applicants must annually confirm their application and leaving for permanent residence in another locality takes the applicant off the local waiting list.49 [My client] is a graduate of an orphanage who has been living in a state rental apartment for 10 years. She cannot register her newborn son because the apartment building has not yet been accepted in operation. As a result, she has already lost three months of the state social benefits for her child and is forced to go to work with the infant. The residents of four more apartment buildings are in the same situation. In short, people find ways somehow, some paid money, some registered at friends’ [housing], but people lost [state] social assistance. Taraz, April 22, 2020 Unemployment benefits. Because unemployment registration and application for state unemployment benefits and assistance are based on location of permanent residence registration, residency requirements could limit access for citizens from other regions. Until recently, most employers required registration when hiring, including for public employment. According to the former Chairman of the Board of the state Government for Citizens corporation, Abylaykhan Ospanov, almost all the five million address certificates the corporation issued annually were requested by employers and banks.50 According to a 2017 survey of internal migrants in Astana, most difficulties respondents faced due to lacking a registration were related to employment (Toqmadi 2020). Currently, labor contracts should contain information about the registration at the employee’s place of residence.51 The Minister of Labor and Social Protection of the Population emphasized that “the address information is filled out on the basis of the oral account of the employee. An employer is not entitled to demand documents not provided for by paragraph 1 of Article 32 of the Code.â€?52 Yet, the requirement remains relevant in practice for low-skilled workers, where employers traditionally see registration as insurance against theft or fraud, as in industries such as sales and construction and in low-paid or “materially responsibleâ€? jobs, such as babysitters, nurses, plumbers, drivers, odd-jobbers, and bookkeepers. A systematic review of online job postings confirms that local registration requirements are routine. Registration is required when applying for a job; not every employer will accept you, especially if this position is materially responsible. People without registration may have difficulty in obtaining a legal job. They agree to work illegally [without contracts or taxes paid], which entails additional risks for them—they are not protected, and it will be difficult to apply for official protection if their rights are violated. Almaty, May 4, 2020 Health care. Although emergency medical care is guaranteed to all citizens regardless of location, planned public health care is provided at the place of permanent or temporary residence.53 Individuals can choose a doctor and a medical organization within the same administrative-territorial unit (city, town, village, or district in cities of republican significance).54 And while the 2014 amendments to the Code On the Health of the People and the Health Care System provided citizens the opportunity to independently decide on the place of medical care, the administrative-territorial unit criterion makes those services unavailable for internal migrants without local registration. Primary health care organizations use the e-government information systems to determine a patient’s residency registration. Hospitals can refuse registration if they establish that a patient’s permanent or temporary residence is outside the local administrative-territorial unit.55 Hospitals give priority to Kazakhstan Country Economic Memorandum 35 patients residing within their service area and can refuse treatment to a patient from outside that area if the number of patients per doctor exceeds the maximum permitted. But a hospital cannot refuse to treat if a patient lives in its service area and that area has no other primary medical organizations.56 Patients can be referred for highly specialized medical care to republican clinics. The referrals, with all necessary documents, are submitted by territorial polyclinics to the health department.57 Education. Public schools and kindergartens enroll children based on their parents’ temporary or permanent registration.58 The school enrollment application can be submitted directly to the school or through the e-government portal.59 Since the abolition of the address certificate, service providers confirm the applicant’s declared residency and identity information using the e-government information systems.60 Schools should ensure admission to grade 1 for all children living in their service area, but because of the limited number of places in each school, a child could be enrolled in any public school within the same administrative-territorial unit.61 In some cases, registration status may result in discrimination against unregistered children. The choice of kindergartens and schools is limited to the administrative-territorial unit where a child’s parents or guardians are permanently or temporarily registered, excluding those residing in a locality other than the one indicated in their registration document. In practice, however, according to an interview with a human rights activist, secondary schools often ignore the absence of local registration and enroll children, giving priority to the children’s right to an education.62 [I]t is true that if the school does not want to accept [a child for some reason], they look at the registration and send the child to another school according to the registration. But people are often forced to live outside the place of registration. Taraz, April 22, 2020 Given the high demand for public kindergartens, local education departments maintain a single electronic waiting list for their administrative-territorial unit. Service providers confirm the applicant’s declared residency and identity information using the e-government state information systems.63 Identification documents. Issuance of birth certificates does not require residency registration.64 The issuance of passports and identity cards is carried out at the place of permanent registration, as well as at the place of his actual stay once initial issuance has been completed. But for those receiving identity documents for the first time, the absence of a residency permit prevents issuance. Reasons for denial of residence Officially, residence registration “is carried out on the basis of documents confirming the ownership of the dwelling … or evidence of its receipt for use, including under a lease (rental), sub-rent agreement, … as well as by written consent of housing owner.â€?65 In practice, these requirements cannot always be met. Law-abiding citizens may not be able to register legally for several reasons.66 Refusal of landlords to register residents. Renters require written consent of a dwelling’s owner to officially register. But landlords commonly refuse to provide the required documents, often due to a Soviet-era understanding of propiska that granted property rights to residents (Hojaqizi 2008). And landlords frequently seek to avoid taxes on rental income—unlike tax-exempt capital gains and use of owner-occupied housing, income from rental properties is taxable in Kazakhstan. Starting on January 1, 2019, Kazakhstan introduced the single aggregate payment, which provides a simplified procedure for registration with the tax authorities when renting some types of property below a certain value. Kazakhstan Country Economic Memorandum 36 Some landlords seek to avoid utility fees, which are based on the number of people either temporarily or permanently registered in a dwelling (Box 2.3). And some simply wish to avoid bureaucracy. Absent documentation, tenants have little recourse in such cases. Box 2.3 Utility tariffs in Almaty depend on residential registration Electricity. Tariffs depend on whether the consumer uses an electric stove and on how many people live in the apartment. For example, in Almaty, the minimum tariff, under which monthly consumption per person is 90 kilowatt hours, is 17.79 tenge per kilowatt. So, if an apartment has a family of four registered, in order to meet the minimum tariff, they must consume 360 kilowatt hours per month, which will result in a tariff of 6,404 tenge (4 people × 90 kilowatt hours × 17.79 tenge). Water and sewage. Tariffs are calculated either as metered or at a per person rate. In Almaty, for apartment buildings without metering devices, the fees per person per month rise with the number of people registered to the dwelling. Gas. Tariffs are also calculated either as metered or at a standard per person rate. As of late 2018, Almaty residents without meters paid for a standard of 10 cubic meters per month in the summer, and 17 cubic meters per person per month in winter. Waste collection. Fees are based on the number of residents. The average fee in Kazakhstan is 174 tenge per person per month. “Sanitaryâ€? standards in Almaty and Astana. Sanitary standards are another barrier to registration for the residents of Almaty. At least 15 square meters of living floor area should be allowed per person for the registration of residents.67 That restriction was adopted in an attempt to get rid of “rubberâ€? apartments—housing where black market dealers register people for a fee. In February 2019, then- president Nursultan Nazarbayev, referring to the rental housing market, said that “houses were overloaded in Astana—the reason for this is rubber apartments, where too many people are registered.â€? He proposed that Astana follow the example of Almaty, where authorities “made the decision to ban registration if an apartment has less than 15 square meters per personâ€? (Butenko 2019). In March 2019, the Astana city local representative body adopted the restriction.68 Ruslan Akhmetov, head of the Department of Construction and Housing Policy of Astana, explained why it was needed: “These rules were developed in order to ensure monitoring and forecasting of the migration situation, addressing the issue of employment, planning the socioeconomic development of the capital for better registration of the populationâ€? (Sputnik News 2019). The restriction affects principally the low - income population and internal migrants, who rent housing with small spaces or shared rooms in cities. Lack of title deeds. House owners lack title deeds for various reasons, including loss of documents, the death of owners, lengthy litigation relating to inheritance, and a lack of funds to register the inheritance. The interior minister, Kalmukhanbet Kasymov, acknowledged the problem: “Indeed, … we encountered such a problem that many developers moved citizens into new houses, but at the same time they did not prepare any documents of title. And they were not registered with the Ministry of Justice. And since there are no legal documents, we cannot register citizens. … They are no t all to blame, so we do not hold anyone accountableâ€? (Bozhko 2018). Although such residents are not fined, without residency registration they face difficulties in accessing social benefits. Long-term absence of homeowners. The long-term absence of homeowners who may have moved abroad, suffered long-term illness, been imprisoned, and so on often prevents them from administering the property, including registering residents. According to the First Deputy Prime Kazakhstan Country Economic Memorandum 37 Minister, Minister of Finance, Alikhan Smailov, updating the Real Estate Register system identified “843,000 real estate objects registered to deceased persons or those who [had] left the countryâ€? (Ogurtsov 2019). Housing policy reform In areas with high housing demand, supply constraints drive high costs. In most countries, supply constraints in areas struggling with affordability and high housing prices stem largely from regulatory barriers limiting the amount and type of housing that can be added to the market (Gyourko and Molloy 2015). The barriers are often local, including zoning and land use restrictions, title and tenure security, building codes, restrictions on minimum unit size, and ability to use land as collateral in development financing. Restrictions on small units (including the sanitary norms used in Almaty and Astana) are high-priority examples of such distortions. In Kazakhstan, identifying and removing such unnecessary barriers at both the national and local levels are the most essential elements of containing the rising cost of housing in areas struggling with affordability. The small size of the rental housing market (Figure 2.4) further limits access to housing. A larger market is needed, especially one focused on ensuring affordable housing for low-income people. Absent massive public investments, in most countries, not everyone can realistically afford to own homes in desirable areas. Countries with housing policies focused primarily on promoting ownership face severe fiscal challenges addressing affordability among poor and vulnerable people (Peppercorn and Taffin 2013). Nor is universal home ownership usually a desirable goal for public finance. Encouraging home ownership through public spending is often financially regressive, and it concentrates household wealth in a single asset whose value can fluctuate widely, as the past decades in Kazakhstan have shown. But in Kazakhstan, almost all financing provided to official housing support primarily encourages home ownership (including the 5-10-20, 7-20-25, Baqytty Otbasy, Nurly Zher, and similar programs), not supporting people needing help to pay rent. International experience suggests that developing a “social rental sectorâ€? is a far more fiscally sustainable option to promote universal housing and better living conditions. Unrestricted housing rental vouchers could be used to remove place-based restrictions on government low-income housing support. Currently, housing law requires that any beneficiary of state housing support must permanently reside in the location where they seek support. This condition limits low- income people’s mobility by forcing them to remain in one place to access benefits. A common alternative form of housing support provides vouchers that can be redeemed for private rental housing anywhere in the relevant jurisdiction—in this case, ideally, anywhere in Kazakhstan. The portability designed into the program—allowing families to rent a unit of their choice in the private market— enables renters to move to safer neighborhoods and to places where their potential incomes are higher (Lens, Ellen, and O’Regan 2011). Recipients could relocate while retaining housing benefits to which they are eligible. Rental real estate should not carry a higher tax burden than other real estate. The use of and capital gain from land and housing—provided they are owned for more than one year—are typically tax exempt in Kazakhstan’s Tax Code (Article 319), putting renters at significant financial disadvantage to homeowners. Rental income is not tax exempt (Article 227 of the Tax Code), and since 2019, formal rental housing has been subject to a single aggregate payment. The monthly amount of the single aggregate payment is 1 minimum calculation index (MCI)—2,917 tenge in 2021—for individuals living in cities of republican and regional significance and 0.5 MCI for individuals living in other settlements. Kazakhstan Country Economic Memorandum 38 Reforming the Tax Code to treat rental and owner-occupied housing similarly for tax purposes is advisable. Countries facing a cycle of localized population decline and deteriorating housing often intervene with “renewalâ€? initiatives to reduce the supply of housing and improve average housing quality in viable communities. In the most common approaches, the government purchases vacant or distressed properties (effectively setting a price floor) and targets them for either renewal or demolition. Examples come from Japan, often in rural areas due to population decline; Germany, particularly in eastern Germany in the 1990s; the United States, for declining cities such as Buffalo, Detroit, and Youngstown; and other high-income countries (Figure 2.13). In addition to direct government action, many countries provide financial incentives to private investors to invest in distressed housing. Such investors typically purchase the housing at a substantial discount from the price offered to consumers. Some such policies have arrested housing value declines and stabilized prices. One study found that a 10 percent increase in the proportion of houses purchased by investors in a US census block led to a 0.20 percent medium-term increase in house prices (Allen et al. 2018). Figure 2.13 Approaches to managing shrinking cities in Europe, Japan, and North America, 2019 Renewal (housing/urban) Demolition 12% 24% Economic diversification 4% 4% Cultural regeneration/tourism 4% Bottom-up initiatives 5% Innovation/entrepreneurship 7% 17% Environment 8% Active aging 15% Land-use control Other Source: Döringer et al. 2020. If the state continues to invest in public housing, it should be built in high-demand areas. Public housing policy is commonly torn between two competing objectives: containing program cost and locating public housing in auspicious labor markets. To contain costs, many public housing programs prioritize construction in low-cost areas far from high-cost agglomerations. In Kazakhstan, public housing has been intentionally distributed across space widely, because recipients must already live where they receive public housing benefits. Although this approach often reduces project costs, it risks further entrenching spatial inefficiency, and in the worst case, it can create poverty traps. Globally, the most successful programs have leaned heavily on building public housing where labor markets are strongest and market demand for housing is greatest. Kazakhstan Country Economic Memorandum 39 Private rental housing vouchers should be used to reduce queuing for public housing. The queue for social housing in Kazakhstan increases by 50,000 applicants a year, while new social housing can accommodate 20,000.69 The Nurly Zher program and other housing assistance programs ration public housing using a grading system to provide citizens with housing depending on their capabilities and status. Alternative programs using housing vouchers in the United States and other countries often reduce crowding, homelessness, and unaffordability. Voucher systems are not limited by the pace of public housing construction, so with sufficient funding, they could eliminate waiting lists for social housing support. The legal framework regulating rental housing relationships needs review and consolidation. If landlords see the legal system as unfair, they will not invest in rental housing or will keep rental housing informal. Legal revision should include clarifying renters’ rights (correcting Soviet-era understanding of registration and tenant rights). Many of them [landlords] live by the old concepts when registration gave preferences to the one registered. They do not know that they can deregister tenants at any time; they believe that this will be a headache for them. Well, the headache begins with the registration process itself —one needs to go [to the registration office], lose his time —and therefore they say [to tenants], “No, get the registration wherever you want, but I will not do it.â€? Taraz, April 22, 2020 A rental contract should include a standard list of provisions for a strong, two-party agreement (Peppercorn and Taffin 2013). The main items are a definition and description of the rental unit, the duration and termination of the contract, rent setting and rent increases, procedures for resolving conflicts and stability, and the adaptability of legal dispositions. The contract should specify a fixed period for the rental—neither too short, to give the tenant stability, nor too long, to give the landlord some flexibility. Reform of residency registration Many advanced economies do not have any form of resident registration. Discontinuing the system entirely is perhaps the simplest approach to resolving the above issues. To serve the purposes now performed by the registration system, governments commonly use census, administrative, and other sources of statistical information. Even if the registration system is preserved, reforms could significantly improve outcomes, especially for low-income people. Most local public services are available only to registered individuals. And a considerable share of the population cannot register, including people who do not have their own housing (or lack documentation), former inmates, homeless people, orphanage alumni, and many internal migrants.70 Absence of registration makes these groups vulnerable to discrimination and increases corruption risks. Some of the reforms that the government might wish to consider are: Removing the distinction between temporary and permanent registration. That would require ensuring no remaining differences in eligibility for rights, privileges, and social benefits by registration status. Setting utility fees through mechanisms other than official registration. Removing tax disincentives for registering tenants in rental housing. Lowering fines for individual nonregistration. Introducing a fine for requesting registration information from job applicants. Kazakhstan Country Economic Memorandum 40 Disallowing registration validation requests by employers—fully enforcing the prohibition of registration requirements in employment contracts and the hiring process.71 Further reducing administrative burdens. For instance, registration could remain in place without needing to be periodically renewed. And the window for registering in a new place of residence could be widened: for instance, the current 10-day period could be lengthened to three months or more—long enough to find a job and housing. Removing the requirement for proof of an agreement with a landlord in order to register. Ending the requirements that people apply for government benefits and request official documents at their place of registration (permanent or otherwise). Removing the requirement that people reapply for public benefits when they move. Benefits would become fully portable from one place of residence to the next. Finally, the registration procedure itself remains burdensome and cannot currently be considered a simple notification. It requires documents confirming ownership rights or evidence of authorized use of a property.72 Solutions flexible enough that each citizen can register at the place of his or her actual residence are needed. They should consider the circumstances of people without permanent residence or a legally recognized address. Those registering could, for example, provide an address at which they can be reached. Or the requirement of a title deed for registration could be removed. References Abilmazhitova, Aigerim. 2020. “Ya pronto hochu domoi. Kak almatintsy probivalis cherez blokposty v pervyi den karantinaâ€? [“I just want to go home: How Almaty residents made their way through checkpoints on the first day of quarantineâ€?]. Tengrinews, March 19. https://tengrinews.kz/fotoarchive/ya-prosto-hochu-domoy-almatintsyi-probivalis-blokpostyi- 1256/. Akhmetov, Alisher. 2017. “V MVD otvetili na voprosy po povodu vremennoi registratsii po mestu zhitelstvaâ€? [Ð’ МВД ответили на вопроÑ?Ñ‹ по поводу временной региÑ?трации по меÑ?ту жительÑ?тва Tengrinews, January 5. https://tengrinews.kz/kazakhstan_news/mvd-otvetili-voprosyi- povodu-vremennoy-registratsii-mestu-309480/. Allen, Marcus T., Jessica Rutherford, Ronald Rutherford, and Abdullah Yavas. 2018. “Impact of Investors in Distressed Housing Markets.â€? The Journal of Real Estate Finance and Economics, 56 (4), 622–52. Bozhko, Vladimir. 2018. “Novaya sistema propiski v Almaty: ‘rezinovyh’ kvartir bolshe ne budetâ€? [New registration system in Almaty: there will be no more ‘rubber’ apartments]. Kursiv.kz, Ap ril 16. https://kursiv.kz/news/obschestvo/2018-04/novaya-sistema-propiski-v-almaty-rezinovykh- kvartir-bolshe-ne-budet. Bryan, Gharad, and Melanie Morten. "The aggregate productivity effects of internal migration: Evidence from Indonesia." Journal of Political Economy 127.5 (2019): 2229-2268. Butenko, Mimi. 2019. “Nazarbayev zayavil, chto v Astane peregruzheny doma iz-za ‘rezinovyhâ€? kvartir’â€? [Nazarbayev said houses in Astana were overloaded because of ‘rubberâ€? apartments’]. Nur.kz, February 27. https://www.nur.kz/1780948-nazarbaev-zaavil-cto-v-astane-peregruzeny-doma-iz- za-rezinovyh-kvartir.html. Döringer, Stefanie, Yuta Uchiyama, Marianne Penker, and Ryo Kohsaka. 2020. “A Meta-analysis of Shrinking Cities in Europe and Japan: Towards an Integrative Research Agenda.â€? European Planning Studies, 28 (9), 1693–1712. Kazakhstan Country Economic Memorandum 41 Durgin Jr., Frank A. 1962. “The Virgin Lands Programme 1954–1960.â€? Europeâ€?Asia Studies, 13 (3), 255– 80. Gyourko, Joseph, and Raven Molloy. 2015. “Regulation and Housing Supply.â€? In Gilles Duranton, J. Vernon Henderson, William C. Strange, eds., Handbook of Regional and Urban Economics. Vol. 5. Amsterdam: Elsevier. https://online.zakon.kz/Document/?doc_id=34199995#pos=1;-161. Harris, Chauncy D. 1971. “Urbanization and Population Growth in the Soviet Union, 1959– 1970.â€? Geographical Review, 1971, 102–24. Hojaqizi, Guliatir. 2008. “Citizenship and Ethnicity: Old Propiska and New Citizenship in Post -Soviet Uzbekistan.â€? Inner Asia, 10 (2), 305–22. Hsieh, Chang-Tai, and Peter J. Klenow. "Misallocation and manufacturing TFP in China and India." The Quarterly journal of economics 124.4 (2009): 1403-1448. HOLA News. 2018. “Sistema online propiski stala bolee uproshennoiâ€? [Online registration system has become more simplified]. April 3, 2018, http://holanews.kz/view/news/14688. Kemeny, Thomas, and Michael Storper. 2020. “Superstar Cities and Left-behind Places: Disruptive Innovation, Labor Demand, and Interregional Inequality.â€? Working Paper 41, International Inequalities Institute, London School of Economics and Political Science, London. Lens, Michael, Ingrid Gould Ellen, and Katherine O’Regan. 2011. “Do Vouchers Help Low-Income Households Live in Safer Neighborhoods? Evidence on the Housing Choice Voucher Program.â€? Cityscape 13 (3), 135–59. Ogurtsov, Ilya. 2019. “843 tysyachi ob’ektov nedvizhimosti chislitsya za umershimi ili uekhavshimi iz Kazakhstana - Smailovâ€? [843 thousand real estate objects listed as dead or who left Kazakhstan - Smailov]. Liter, December 3. https://liter.kz/13440-843-tysyachi-obektov-nedvigimosti-chislitsya- za-umershimi-ili-uehavshimi-iz-kazahstana-smailov/. Peppercorn, Ira Gary, and Claude Taffin. 2013. Rental Housing: Lessons from International Experience and Policies for Emerging Markets. Washington, DC: World Bank. Pokidayev, Dmitry. 2019. “Ospanov: Banki gotovy otkazatsa of adresnyh spravokâ€? [Ospanov: Banks are ready to give up address certificates]. Kursiv.kz, February 11. https://kursiv.kz/news/banki/2019- 02/ospanov-banki-uzhe-gotovy-otkazatsya-ot-adresnykh-spravok. Pokshishevsky, V.V. 1972. “Urbanization in the USSR.â€? Geoforum, 3 (1), 23–32. Radio Azattyk. 2012. “V Kazahstane vozobnovili institut propiskiâ€? [Kazakhstan revives the Institute of Registration]. 6 February. https://rus.azattyq.org/a/propiska_mvd_sainov_ergalieva_nazkhanov_baiseitova/24474434.html. Rosés, Joan R., and Nikolaus Wolf. 2018. “Regional Economic Development in Europe, 1900–2010: A Description of the Patterns.â€? Economic History Working Papers 278/2018, International Inequalities Institute, London School of Economics and Political Science, London. Seitz, William. 2021. “Urbanization in Kazakhstan: Desirable Cities, Unaffordable Housing, and the Missing Rental Market.â€? International Journal of Urban Sciences, 25 (Supplement 1), 135–66. Shkvarikov, V., M. Haucke, and O. Smirnova. 1964 “The Building of New Towns in the USSR.â€? Ekistics, 18 (108), 307–19. Kazakhstan Country Economic Memorandum 42 Sputnik Kazakhstan. 2016. “Zasedanie Soveta Bezopasnosti RK pod presdsedatelstvom Nursultana Nazarbayevaâ€? [Security Council Meeting chaired by President Nursultan Nazarbayev]. [Video]. YouTube. June 10. https://www.youtube.com/watch?v=oev_VRfc0dk. Sputnik News. 2018. “Po 50 chelovek ‘zhivut’ v odnoi kvartire—MVDâ€? [50 people ‘live’ in one apartment in Kazakhstan—Ministry of Internal Affairs]. December 7. https://ru.sputniknews.kz/society/20181207/8397493/zhile-propiska-kazahstan-mvd.html. Sputnik News. 2019. “V Astane ogranichat propisku v odnoi kvartire - kogo kosnetsa normaâ€? [Astana will limit registration in one apartment - who will be affected by the norm]. March 6. https://ru.sputniknews.kz/capital_life/20190306/9520068/Astana-propiska-kvartiry-maslihat.html. Tengrinews. 2012. “V Kazakhstane khotiat ogranichit’ propisku v Astane i Almatyâ€? [Kazakhstan wants to limit registration in Astana and Almaty]. July 12. http://tengrinews.kz/kazakhstan_news/v- kazahstane-hotyat-ogranichit-propisku-v-astane-i-almatyi224794/. Toews, Gerhard, and Pierre-Louis Vézina. 2020. “Enemies of the People.â€? Working Paper 2020-20, Quantitative Political Economy Research Group, Department of Political Economy, King’s College London. Toqmadi, Malika. 2020. “Registratsiya po mestu zhitelstva i bezopasnost cheloveka v Kazakhstane.â€? PaperLab, Nur-Saltan, Kazakhstan. https://drive.google.com/open?id=1N3Rx0M5UnDj4IjbrURZCSWNj2SH4mjDw. Tukmadiyeva, M. 2016. “Propiska as a Tool of Discrimination in Central Asia.â€? Central Asia Fellowship Papers 12, George Washington University Institute for European, Russian, and Eurasian Studies, Washington, DC. Zakon.kz. 2010. “V Almaty zapresheno propisyvat grazhdan menee chen ne 15 kvadratnyh metrah na chelovekaâ€? [It is forbidden to register citizens in less than 15 square meters per person in Almaty]. April 13. https://www.zakon.kz/169133-v-almaty-zapreshheno-propisyvat.html. Zakon.kz. 2014. “Ogranichit propisku v odnom pomeshchenii po primeru Almaty predlojil mazhilismenâ€? [Mazhilis Deputy suggested limiting residence in one room, following the example of Almaty]. January 15. http://www.zakon.kz/4596463-ogranichivat-propisku-v-odnom.html. Kazakhstan Country Economic Memorandum 43 Annex 2A Migration statistics Two primary sources of domestic migration data were available to the authors: survey data from the Kazakhstan Labor Force Survey (LFS), and officially published data or administrative records that use individual registration and other sources of official statistics at the regional level. The chapter analyzes domestic migration in Kazakhstan using data from the LFS. In a spatial approach, it focuses on migration at the district level. It begins by examining the spatial pattern in domestic migration, then examines variation in these patterns by education level. It combines data on migration rates with consumption data from the household budget survey and small area estimates of poverty to examine the relationship among migration, consumption, and poverty. Specifically, the report discusses: Distribution of domestic migration by district: which districts do domestic migrants in Kazakhstan move to? Distribution of domestic migration by district and education level. Raw immigration rates by district. The relationship between district immigration rates and district average per capita consumption from the household budget survey. The relationship between migration and poverty using information on district poverty rates at the US$3.20 per day and US$5.50 per day levels. The analysis uses information from LFSs between 2014 and 2019 and focuses on migrants who arrived in the five-year period preceding each survey year. For the analysis, a domestic migrant is a person living in a different location from the place of birth.73 Data from the following LFS questions were used to identify migrants and their length of stay: “Have you lived in this location since birth?â€? and “Date of arrival in current place of living.â€? The definition of migration and the data available made it impossible to distinguish between intra-district and inter-district migration or to perform origin–destination analysis. But the analysis is informative about broad patterns in domestic migration flows in Kazakhstan. Kazakhstan Country Economic Memorandum 44 Annex 2B Registration procedure The 2016 amendments to the Law on Migration of the Population introduced a temporary registration (registration at the place of temporary stay),74 as opposed to permanent registration (at the place of residence).75 According to the rules of registration of internal migrants, temporary registration lasts from one month to one year.76 The Ministry of Internal Affairs, clarifying the 2016 amendments, stated that “the period of stay is indicated when obtaining temporary registrationâ€? (not the permanent registration). Both the temporary and permanent addresses are stored in the registry together (Akhmetov 2017). The place of registration is the citizen’s legal address at which tax, military, and other types of accounting are conducted.77 Both temporary and permanent registrations are carried out by the Ministry of Internal Affairs through public services centers (TsON), or through the e-government portal by personal written consent of the property owner or by use of the owner’s digital signature.78 The authorized agency then uses the address registration code through the information systems to confirm the owner’s property rights.79 Currently, registration is authorized in the following places: hostels, hotels, rest houses, boarding schools, apartments, nursing homes, boarding houses, summer houses, cooperatives, dispensaries, residential buildings, sanatoriums, medical institutions, horticultural societies, as well as buildings and premises otherwise used for the residence of people.80 According to the rules for providing public services on issues of documentation and registration of the population, the registration procedure is free of charge and takes up to 30 minutes.81 Surveys show that most violations of the registration regime take place because citizens cannot prove the legitimacy of their stay in the residential premises (for instance, they are living in unauthorized or unregistered residential buildings) or because landlords are not willing to register their tenants (Toqmadi 2020). Identification of violators of migration legislation is carried out when citizens apply to TsONs for registration at the place of residence, as well as during operational and preventive measures as part of monitoring compliance with the requirements. Galina Sarsenova, deputy chairperson, Migration Committee of the Ministry of Internal Affairs, June 5, 2020 “Many [people without registration] continue to live essentially in an illegal status, and their capabilities are limited. For example, now with the introduction of an emergency quarantine of Almaty, these people became outlawed. People turned to us—many people who work in Almaty but live in the region could not get into the city. I witnessed that people without a residence permit were not allowed into the city by the police. On the other hand, people who lived and were registered in the region and traveled to the city for work were cut off, their rights were violated, they were unable to work. In essence, this is discrimination against these people.â€? Almaty, May 4, 2020 Kazakhstan Country Economic Memorandum 45 Annex 2C The legal basis of residency registration The Constitution of the Republic of Kazakhstan guarantees “everyone who has the legal right to stay … the right to free movement throughout its territory and free choice of a place of residence, except in cases stipulated by law.â€?82 It also ensures nondiscrimination “for reasons of origin, social, property status, … place of residence, or any other circumstances.â€?83 The Law on Migration of the Population obliges citizens84 to register at the place of residence.85 There is an administrative liability for accommodation without registration for more than 10 days for residents,86 as well as for property owners for hosting unregistered tenants, for registration of individuals not actually living in the premises, and for failure to deregister former residents.87 Besides the Constitution and the Law on Migration, the residency registration system is regulated by the following: Code of the Republic of Kazakhstan, No. 235-V, On Administrative Offenses, dated July 5, 2014 (amended January 16, 2020). Decree of the Government of the Republic of Kazakhstan, No. 1427, On Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, dated December 1, 2011. Decree of the Government of the Republic of Kazakhstan, No. 296, On Approval of Model Rules for Regulating Migration Processes in Regions, Cities of Republican Significance, the Capital, dated May 26, 2017.88 Order of the Minister of Internal Affairs, No. 267, On Approval of the Rules for the Provision of Public Services on Issues of Documentation and Registration of the Population of the Republic of Kazakhstan, dated March 30, 2020.89 In 2017, migration management functions were transferred from the Ministry of National Economy and distributed between the Ministry of Labor and Social Protection and the newly formed Migration Service Committee within the Ministry of Internal Affairs. The Ministry of National Economy retained the function of gathering statistics on population movement. Currently, residence registration of citizens is carried out through the information system Registration Point Documentation and Registration of Population. The system transmits all information about registered citizens in real time to the state database Individuals, access to which is provided to all state bodies. When someone registers, information about their change of address is also transmitted to the e- statistics information system of the Statistics Committee of the Ministry of National Economy, and so statistics on population migration are compiled by this agency. Galina Sarsenova, deputy chairperson, Migration Committee of the Ministry of Internal Affairs, June 5, 2020 On the other hand, the regions and cities of “republicanâ€? (that is, national) significance got the right to develop their own rules of regulating migration processes “in order to ensure the manageability of migration processes and economic growth, strengthen the country’s security and create conditions for the realization of the socioeconomic rights of migrants.â€?90 Kazakhstan Country Economic Memorandum 46 3 Spatial dimensions of intraregional and interregional trade in Kazakhstan Internal trade powerfully shapes a country’s economy.91 Like international trade, intraregional and interregional trade offers growth and job opportunities. One region’s producers benefit from improved access to other regions’ markets. Through a larger pool of input markets, more efficient information flow, and expanded product markets, producers achieve increased productivity and more diversified production. Businesses in regions near large markets with more concentrated production are likely to sustain higher nominal wages.92 In Kazakhstan—a landlocked country with vast territory and low population density—interregional trade is likely to be higher between regions that are close to one another and between regions that have large economies and denser populations. Although trade between any regions is possible, trade flows will be limited by the quantities that markets demand and by operational and logistics costs. And because low-density areas support scant manufacturing and processing, all regions will seek access to suppliers in other regions where input markets (labor and capital) can competitively sustain manufacturing and services. This chapter addresses the following: How do intraregional trade (within a region) and interregional trade (between regions) flow across Kazakhstan, and how do trade patterns correlate with geographic features? Given the country’s huge territory and low population density, how do geographic features relate to market integration? Is intranational trade or interregional trade associated with the level of regional development? The dynamics and cross-sectional variation of intranational trade Kazakhstan has seen a rapid increase in intranational trade, driven by growth in all three of its components: intraregional trade, interregional trade, and distribution of imported goods.93 The value of intranational (inter- and intraregional) trade in 2018–19 was 5.9 trillion tenge, or 4.6 percent of GDP (the middle bars in figure 3.1).94 This is 62 percent higher than the 3 trillion tenge in 2014–15. The distribution of imported goods contributed about 52 percent to the increase in intranational trade, due to economic recovery after the 2014 crisis, exchange rate depreciation, and the formation of the Eurasian Economic Union, while intraregional and interregional trade contributed about 48 percent (Figure 3.2).95 The fact that intraregional and interregional trade made up almost half of the increase in trade suggests the importance of within-country trade flows in moving the economy. Kazakhstan Country Economic Memorandum 47 Figure 3.1 Intranational trade in Kazakhstan, Figure 3.2 Contributions to the change in 2014–15 and 2018–19 intranational trade between 2014–15 and 2018–19 Trillion tenge Percent 12 10 8 6 4 2 0 Total Intra and inter Import regional Intraregional and interregional 2014-2015 2018-2019 Source: World Bank, based on Committee on Statistics data. Note: Values are averages for 2014–15 and for 2018–19. Kazakhstan is the largest country in Central Asia, with an area of about 2.69 million square kilometers— approximately 4.9 times larger than France. The maximum distance for interregional trade is 3,700 kilometers.96 For example, the distance by road between Aktau in Mangystau (in the western part of the country) and Ust-Kamenogorsk in East Kazakhstan is about 3,600 kilometers, depending on the route. Yet, about 60 percent of the value of intraregional and interregional trade in 2014–19 was between cities less than 1,000 kilometers apart or in regions with a radius of less than 1,000 kilometers (Figure 3.3). Almaty and Astana cities and their surrounding regions are the main contributors to Kazakhstan’s increasing intranational trade. Between 2014/15 and 2018/19, they and their surrounding regions accounted for about 75 percent of intranational trade. The size of their imports is predictable, because these cities of national significance are vital international hubs and have higher per capita income from nonmining industries than other regions, except Atyrau. High economic concentration in the two cities and surrounding regions also has a positive spillover. Almaty city, Almaty region, Astana city, and Akmola region contributed about 60 percent to growth in sourcing goods from intraregional and interregional trade (Figure 3.4). Kazakhstan Country Economic Memorandum 48 Figure 3.3 Contribution to intraregional Figure 3.4 Contribution to change in total and interregional trade, by distance intranational trade, by region and city of between trading locations, 2014–19 republican significance, 2014/15–2018/19 Percent Percent Intraregional Interregional Import 60 50 40 30 20 10 0 -10 Source: World Bank estimates, based on data from Source: World Bank, based on Committee on Statistics the Bureau of National Statistics. data. Note: The red line is the estimated trend from a Note: The sharp decline in intraregional trade in Aktobe locally weighted scatterplot (lowess smoothing). region is probably a data anomaly. Cross-sectional data suggest that regions closer to Kazakhstan’s borders have a larger share of imports in their intranational trade. Imports account for about 59 percent of goods sourced by businesses in Almaty city and region, which are approximately 350 kilometers from Khorgos, a major border crossing with China, and less than 250 kilometers from Korday, a border crossing with the Kyrgyz Republic (Table 3.1). The customs union established under the Eurasian Economic Union is also likely to make imports a relatively high proportion of goods sourced in regions near the Russian border. For instance, in the West Kazakhstan region, which borders the Russian Federation, imports accounted for 64 percent of goods sourced by businesses. But in South Kazakhstan (since 2018, Turkistan region and Shymkent city), which borders Uzbekistan, imports accounted for only 20 percent of goods sourced by businesses. South Kazakhstan may be sourcing a high share of goods from other regions, particularly Almaty. Most regions source more goods from Almaty city and region than from adjacent areas. The high level of goods purchased from Almaty also reflects the indirect sourcing of imports because 67 percent of Kazakhstan’s imports come from that city and its region. The share of interregional trade with them is more than 20 percent in 9 of 14 regions, including Astana (Table 3.1, column 3). On the other hand, the share of goods sourced from adjacent regions is generally lower than that from Almaty, except in Zhambyl, Karagandy, and East Kazakhstan (Table 3.1, column 4). Almaty city and region’s large marketplace and their access to international trade make them attractive to other regions for sourcing goods. They can offer attractive terms, such as higher volume and lower cost per volume shipped, for sourcing goods. Access to imports from the Kyrgyz Republic, China (via the East Kazakhstan region), and the Russian Federation (with which Kazakhstan has a rail link as well as a customs union) also feeds into their market. For regions facing a trade imbalance, sourcing goods from Almaty city and region can offer traders an opportunity for backhaul to sell products to those Kazakhstan Country Economic Memorandum 49 larger markets. Also, the range of products that remote regions can source from nearby markets is limited. Large and high-density regions tend to have higher intraregional trade. Almaty region (including the city) accounts for 21 percent of Kazakhstan’s GDP. Its businesses sourced 30 percent of goods intraregionally in recent years (Table 3.1). Atyrau region, which has the second largest GDP after Almaty and is the heart of the oil industry, sourced 73 percent of its goods intraregionally, and Karagandy, 50 percent.97 These relationships suggest that the concentration of economic activities, which fosters increasing returns to scale, helps ramp up output. Urbanization and population density98 can also affect trade turnover. For example, Almaty accounts for 28 percent of the country’s retail trade, followed by Astana (11 percent), Karagandy (7 percent), and Aktobe (6 percent; Kurmanbekov and Temirkhanov 2015). Table 3.1 Percentage of goods sourced in each region, 2018 and 2019 Distance to nearest land Interregional border Adjacent crossing Nearest land Geographic areas Total Almaty regions Intraregional Import (kilometers)a border East Pavlodar region 27.6 12.4 11.7 55.9 16.5 445 Russian Federation East Kazakhstan region 38.7 23.8 35.8 27.9 33.4 251 China Almaty region 11.1 30.3 1.2 30.3 58.6 215 Kyrgyz Republic South Zhambyl region 71.6 26.8 63.7 16.0 12.4 314 Kyrgyz Republic South Kazakhstan region* 42.7 39.0 3.9 37.4 19.8 168 Uzbekistan Kyzylorda region 50.7 30.4 1.6 32.6 16.7 565 Uzbekistan West Aktobe region 63.5 54.6 2.3 20.1 16.4 110 Russian Federation Mangystau region 42.7 30.5 2.9 42.2 15.1 550 Uzbekistan Atyrau region 13.0 9.0 0.4 72.8 14.2 739 Russian Federation West Kazakhstan region 23.5 16.6 2.5 12.5 64.0 488 Russian Federation North Kostanay region 57.9 43.0 7.0 15.8 26.2 172 Russian Federation Akmola region and Astana 51.4 17.8 6.7 12.3 36.3 739 Russian Federation North Kazakhstan region 58.2 33.0 4.4 31.8 10.0 627 Russian Federation Central Karagandy region 19.8 14.3 26.6 50.2 30.0 879 Russian Federation * Covers Shymkent city and Turkestan region. Source: World Bank, based on Committee on Statistics data. Note: a. travel distance from region’s largest city to the border crossing. Kazakhstan Country Economic Memorandum 50 Some regions have high intensities of interregional trade and rapid industrial development. The volume of goods sourced from other regions as a share of GDP is above the national average in Akmola and Astana, Kostanay, and South Kazakhstan, as indicated by a location quotient index greater than 1.99 Akmola, Kostanay, and Shymkent city are experiencing high growth in manufacturing, construction, and wholesale/retail activities, which probably require intermediate inputs from other regions, particularly Almaty (Figure 3.5). Figure 3.5 Intensity of intranational trade relative to the national average, by trade component and region and city of republican significance, 2018–19 3.5 Intra-regional Inter-regional Import 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Source: World Bank, based on Committee on Statistics data. Note: Values are the share of each type of intranational trade in GDP, divided by the respective national average. For example, if a region’s number for intraregional trade is greater than 1, the region has a higher volume of intraregional trade than the national average. For a country with vast territory, distance can create friction for intraregional and interregional trade. Few regions source a large share of their goods from adjacent regions (Table 3.1), but trading with a more distant region that has lower economic activity can be costly because it might not generate enough two-way trade to produce economies of scale. With no tariffs or duties for goods moving between regions, distance itself is a major trade cost in Kazakhstan. To find whether the data are consistent with the trade theory prediction, we used a gravity model to further analyze the correlation between distance and size of economic activities. This approach uses distance and other observable variables to approximate trade costs that can be sources of friction in intraregional and interregional trade. The result suggests that intraregional and interregional trade decline with distance: a 1 percent increase in the area of a region or the distance between two regions is associated with 0.9 percent less regional trade (Table 3B.1, column 2, Annex 3B). Intranational trade and regional development Market access and regional income per capita Evidence from international trade suggests a strong correlation between market access and wages across countries (Redding and Venables 2004). Access to markets can lower firms’ input costs, allowing them to sustain higher wages. For regions, this topic has been analyzed in the context of international trade and regional economic integration—for example, in the European Union. The results suggest Kazakhstan Country Economic Memorandum 51 that access to markets is important in explaining variations across countries in per capita income (Figures 3.6 and 3.7).100 Figure 3.6 Market access to suppliers and Figure 3.7 Market access to buyers and nonmining income per capita, by region, nonmining income per capita, by region, 2014–18 2014–18 16.0 16.0 Log monmining GDP per capita Log nonmining GDP per capita Atyrau Akmola Akmola Atyrau 15.5 Almaty 15.5 Almaty E S (millions of tenge) E Kazakhstan S Kazakhstan (millions of tenge) 15.0 Kazakhstan Kazakhstan 15.0 Aqtobe Aqtobe Karagandy Karagandy 14.5 Mangystau Kostanay 14.5 Kostanay Mangystau Pavlodar N Pavlodar N Kazakhstan Kazakhstan Kyzylorda 14.0 Kyzylorda 14.0 Zhambyl Zhambyl 13.5 W 13.5 W Kazakhstan Kazakhstan 13.0 13.0 1 2 3 4 5 6 1 3 5 7 9 Log market access to suppliers Log market access to buyers Source: World Bank estimates, based on Committee on Statistics data. Note: Akmola includes Astana, and Almaty includes Almaty city. South Kazakhstan is now Turkistan region and Shymkent city. For Kazakhstan, too, measures of market access suggest positive correlations with regional development. We applied an approach originally proposed by Redding and Venables (2004) and discussed by Redding (2020) that used results from the structural gravity model in column 2 of Table 3.2 (see Annex 3A for a description of the approach).101 The result confirms the positive association across regions between market access indexes and average income per capita seen in Figures 3.5 and 3.6.102 The result is in line with predictions that the wages that businesses in a region can pay (proxied by GDP per capita) are linked to market capacities to obtain inputs and sell products.103 Almaty, Akmola (including Astana), South Kazakhstan, and East Kazakhstan have high market access indexes, which are associated with their higher income per capita from nonmining activities. Better access to suppliers can benefit industries in a region and sustain higher income per capita from nonmining activities. Market access to suppliers is defined as the combination of trade costs and fixed effects from bilateral trade. It relates to the cost of intermediate goods, which feeds into the production cost. The correlation between market access to suppliers and average income per capita is positive and significant (Table 3.2), and it remains significant after the share of employment in the services sector (column 2), which often pays better wages than agriculture, is controlled for. The finding does not confirm a causal relationship between market access and regional nominal wages, but it shows that proximity to suppliers is associated with lower costs for intermediate inputs and the ability to sustain higher income per capita from nonmining activities. In the West Kazakhstan region, the correlation between market access to buyers and income per capita from nonmining activities is not significant. The region is an outlier, with high market access to buyers that probably reflects the economic structure. Mining accounts for 43 percent of the region’s output. West Kazakhstan also produces about 37 percent of Kazakhstan’s natural gas, which has secure internal buyers. But nonmining activity, limited to agriculture, has not benefited from proximity to the Russian Kazakhstan Country Economic Memorandum 52 market, unlike Kostanay, nor has it benefited from the rail line through the region. West Kazakhstan handles 20 percent of freight in Kazakhstan, but poor local infrastructure may be a constraint, with only 73 percent of public roads paved, below the 86 percent national average, and only 2.4 percent of public roads in road quality categories I and II, compared with the 8.5 percent national average (UNECE 2019). Table 3.2. Regression between income per capita and indicators of market access Dependent variable: log of nominal nonmining GDP per capita (1) (2) (3) (4) Log market access to suppliers 0.551 0.388 (6.49)** (4.13)** Log market access to buyers 0.198 0.011 (1.8)† (0.06) Share of employment in services 0.045 0.075 (2.84)** (2.54)* Constant 12.80 11.57 13.8 11.66 (36.88)** (18.63)** (33.3)** (18.95)** Number of observations 14 14 14 14 F-statistics 42.1** 16.5** 3.2* 14.9** R2 0.63 0.76 0.14 0.54 Source: World Bank estimates. Note: Numbers in parentheses are t-statistics, which are based on robust standard error. † significant at p < .1; * significant at p < .05; ** significant at p < .01. Concentration of economic activities Trade can concentrate economic activities in particular locations. Traditional international trade theory suggests that trade allows countries or regions to specialize in activities in which they have the highest comparative advantage. More recent theories suggest that population mobility, lower transport costs, increasing returns to scale, and consumer demand for variety create positive spillovers that attract industries to concentrate in certain locations.104 Natural advantages, such as soil, temperature, or the peculiarity of local demand, can also attract specific industries to concentrate in certain locations. Firm-level data suggest increasing concentration of manufacturing activities across regions in Kazakhstan. Using the industry-level concentration index developed by Ellison and Glaeser (1997), we calculated the regional average index as a weighted average across industries. A higher index value implies that business activities in the region have become more concentrated around certain locations. The index suggests a higher concentration in manufacturing activities in 2018 than in 2014 (Figure 3.7), especially in regions near the borders, such as Akmola, Kostanay, North Kazakhstan, East Kazakhstan, and South Kazakhstan. Concentration in manufacturing also increased in two regions with access to large metropolises: Almaty region with Almaty city and Karagandy region with Astana. Areas with higher concentration of manufacturing activities, especially those with access to border crossings, had higher imports.105 Between 2014 and 2018, imports surged in the northern regions of Akmola, Kostanay, and North Kazakhstan and the southeast Almaty region, where manufacturing concentration picked up (Figure 3.8). The positive association between imports and manufacturing concentration (Table 3.3) could indicate businesses’ preference to locate where they can better access labor, intermediate inputs, and imported materials. For example, the automotive and machinery Kazakhstan Country Economic Memorandum 53 industries are expanding within the Kostanay region, benefiting from rail and road links to the Russian Federation. Almaty region is experiencing growing activity in the food, machinery, and pharmaceutical industries. Figure 3.8 Concentration index of manufacturing activities, 2014 and 2018, and import growth, by region and city of republic significance 2014 (left axis) 2018 (left axis) Import growth (right axis) 0.060 50 40 0.050 30 0.040 20 Concentration index 0.030 10 Percent 0.020 0 -10 0.010 -20 0.000 -30 -0.010 -40 Sources: Index adapted from Tan and Tusha (2020); import growth based on World Bank calculations. Note: South Kazakhstan includes Shymkent city and Turkestan region. Aktobe region is excluded because of an anomaly in trade data in the observed period. Table 3.3 Simple correlations between the growth of intranational trade and change in concentration indexes across regions, 2014–18 Average growth (percent) Intraregional Interregional Import Change in manufacturing concentration a −0.36 −0.11 0.57** Source: World Bank estimates using Tan and Tusha (2020) calculations on the Ellison-Glaeser index. a. Ellison-Glaeser concentration index. A higher value implies higher concentration of activities in certain locations. ** significant at p < .01. Connectivity and firms’ performance Kazakhstan has invested massively in transport connectivity. In 2015, the government launched Nurly Zhol, a US$9 billion nationwide project to develop and upgrade roads, ports, and railroads. The program also aims to develop trade opportunities with two large markets—China and Russia— between which Kazakhstan is located. Long-range transport corridors are also being developed for Kazakhstan Country Economic Memorandum 54 Kazakhstan to tap opportunities from transit cargo between countries in Central Asia with China, Europe, and Russia. Figure 3.9 Road and rail networks of Kazakhstan, 2021 Source: World Bank. Despite heavy investment, road conditions vary greatly: in 2018 about 85 percent of national roads were estimated to be in good or fair condition (Figure 3.10), but about 35 percent of municipal roads, in poor condition (Figure 3.11). Wide regional differences exist, too. Yet, despite roads being relatively bad in the western part of the country, as in Aktobe and Atyrau regions, most of the road projects in 2020–22 are concentrated in the southeastern part of the country. Figure 3.10 Condition of national roads, Figure 3.11 Condition of municipal roads, 2018 2017 Source: UNECE 2019. Source: UNECE 2019. Kazakhstan Country Economic Memorandum 55 The extent to which transport and logistics connectivity shape local economic development also depends on the underlying structural factors. If resources are able to move, better connectivity lowers mobility costs and may cause productive capital and labor to move from rural locations to cities over time. This process enforces the concentration of economic activities in the cities, which can potentially increase productivity through agglomeration. As previously shown, regions with higher import growth are also experiencing higher industrial concentration. Manufacturers require imported machinery, components, and intermediate goods. So, access to border checkpoints or trade hubs can be important in shaping industry concentration. In addition, about 60 percent of trade in Kazakhstan occurs between locations less than a thousand kilometers apart. Thereforem interregion connectivity, such as between-cities connectivity, can play an important role in influencing economic activities. This section seeks to explain the impact of those types of connectivity on firm performance. A panel data analysis was derived from a detailed time series on transport connectivity, based on micro-driving data collected from freight shippers by the Asian Development Bank during 2010–19.106 The CAREC Corridor Performance Measurement and Monitoring data cover detailed cost and time information. For instance, the duration of travel includes not only drive time and border crossing time but also time spent for weight inspections, police checks, and loading and unloading. Transportation costs reflect not only shipping fees but also official—and unofficial—payments incurred during the trip, including customs duties, border clearance fees, and emergency repair costs. Three transport connectivity variables were examined: transport costs to the nearest city (CITY), transport costs to the nearest border point (BP), and the Market Access Index, based on transport costs (MAI). The analysis also used a Kazakh firm registry database that covers the service, industry, and agriculture sectors for 2010–18. The database includes some 20,000 formal enterprises, which employ about 100,000 people. For the indirect effects, firms’ inventory (Q) and agglomeration economies (N and M) were considered (see Annex 3D for the approach). The results suggest that a 10 percent improvement in local connectivity could increase firm production by 2.4 percent (elasticity –0.237). For the direct effects, only the connectivity to the nearest city or market (CITY) has a statistically significant impact on firm production. Other types of transport connectivity— BP and MAI—do not have a clear impact on firm production. This suggests that firm production is most sensitive to the construction of local roads. While the overall market accessibility and the efficiency of border crossing points were found to be insignificant, the signs of the coefficients suggest that improvements along these dimensions also positively affect firm productivity. The MAI was found to have a significant impact on firm inventory with elasticity estimated at –0.87. These results are consistent with the literature (see, for example, Datta 2012 and Shirley and Winston 2004). This implies that besides the observed direct impact of transport connectivity on firm production, the indirect effect of transport connectivity reduces firm inventory, which in turn increases firm productivity. Improving overall market accessibility therefore affects both firm inventory directly and firm productivity indirectly. The results also show that market accessibility has a significant positive impact on agglomeration economies (Table 3.4), which is also consistent with the discussion in the previous section. Here, agglomeration is represented by the number of firms registered in a location. Local transport Kazakhstan Country Economic Memorandum 56 connectivity (CITY) or regional connectivity (BP) does not seem to be conducive to agglomeration economies. Table 3.4 Estimation results of connectivity variables and firm level performance Cost function (𝑦) Inventory (𝑄 ) Agglomeration (í µí±?) Coefficient Standard error Coefficient Standard error Coefficient Standard error CITY –0.237 (0.145) * –0.056 (0.051) 0.014 (0.031) BP –0.106 (0.142) –0.012 (0.056) –0.012 (0.063) MAI –0.029 (0.178) –0.865 (0.436) ** 0.168 (0.076) ** Q –0.637 (0.382) * N −0.046 (0.545) M 0.122 (0.463) Source: World Bank estimates. Note: Standard errors in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01. Factors in internal market integration By creating friction for intraregional and interregional trade, distance can influence internal market integration (Annex 3B).107 The result from a gravity model confirms that distance, which contributes to trade costs, influences intraregional and interregional trade in Kazakhstan. The greater the friction for trade between regions, the harder it is for businesses to arbitrage goods. The friction reduces market responsiveness to excess demand or excess supply, so that regional prices take longer to converge.108 Even in advanced economies, distance, trade imbalance, and regional segmentation in business operations affect price transmission across regions.109 In a developing country such as Kazakhstan with a vast territory and a sparse population, prices can be expected to adjust imperfectly across regions. A region is, predictably, more integrated with nearby domestic markets, and a region near the border, with foreign markets. Figure 3.12 Monthly price of rice, 2011–20 250 200 150 100 50 Akmola Aktobe Almaty region North Kazakhstan Zhambyl Karagandy Kazakhstan Country Economic Memorandum 57 Source: World Bank estimates. Note: January 2011 = 100. Higher correlations appear among food prices in regions close to each other and in regions close to foreign markets.110 Rice prices clustered in the adjacent regions of Akmola and Karagandy over 2011– 20 and moved in parallel in other regions (Figure 3.12). Akmola and Karagandy rice prices gradually showed wider differences from prices in other regions. Prices of flour, in which Kazakhstan has a significant production surplus, provide another example: prices moved together in the regions depicted in Figure 3.13, Aktobe, Karagandy, and North Kazakhstan. As for sugar, Kazakhstan met 90 percent of its domestic demand for sugar with imports from Belarus and the Russian Federation, so it is not surprising to see sugar prices in Kazakhstan following price fluctuations in Russia (Figure 3.14).111 Figure 3.13 Monthly price of flour, 2011–20 Figure 3.14 Monthly price of sugar, 2011– 20 250 200 200 150 150 100 100 50 50 Akmola Aktobe Akmola Aktobe Almaty region North Kazakhstan Almaty region North Kazakhstan Zhambyl Karagandy Russia Source: World Bank estimates. Note Source: World Bank estimates. Note: January 2011 = 100. Note: January 2011 = 100. Analysis of 10 food products finds a wide variation in market integration across Kazakhstan’s regions. We used a panel error correction approach to test for the presence of market integration between regions (Annex 3C). Of 2,400 combinations of pairwise regional prices for 10 food products, 59 percent had significant long-term equilibrium relations (significant λ coefficients). So, for those regions and products, the price in region I will adjust to a price shock in region j, and the two will then converge toward their long-term equilibrium. The analysis finds that the λ coefficients (often referred to as the speed of adjustment) varies between −0.51 and −0.03, with an average of −0.14. We calculated a half- life measure to convert the speed of adjustment into time.112 It indicates how long in months it takes for a deviation of prices in two regions to return halfway to their equilibrium relationship. The half-life measures vary between one and 24 months, with an average of 5.7 (Figure 3.15). The food prices that adjust the fastest tend to involve regions or cities that are closer together. The minimum and maximum values of the estimated half-life measures suggest several instances of speedy adjustment (Table 3.5). For instance, it occurs between South Kazakhstan and Zhambyl for cooking oil and between West Kazakhstan and Atyrau for rice. There are also indications of slow price adjustment for farther apart regions—for example, between Mangystau and North Kazakhstan for rice and between Karagandy and West Kazakhstan for dairy products. But some results also reflect specific relationships between consumers and suppliers or transport and logistics services across region pairs— Kazakhstan Country Economic Memorandum 58 for example, the fast convergence, despite greater distances, between Atyrau and Zhambyl for eggs and between Almaty and Akmola for flour. Finally, the results also capture market characteristics and demand for products originating from specific locations or peculiar nontraded activities in a region or city that might influence price adjustments—such as local logistics market, property market, and retail/warehousing services.113 Figure 3.15 Time needed for regional prices that have deviated to return halfway to their long-term equilibrium Source: World Bank estimates. Table 3.5 Minimum and maximum half-life of price adjustments Product Minimum Region i, region j Maximum Region i, region j Beef 2.4 Akmola, Kyzylorda 19.1 South Kazakhstan, East Kazakhstan Eggs 1.1 Atyrau, Zhambyl 16.2 Zhambyl, Aktobe Fruits 1.8 East Kazakhstan, Astana 16.8 Mangystau, Karagandy Poultry 1.3 West Kazakhstan, Aktobe 21.0 Kostanay, Almaty city Rice 2.5 West Kazakhstan, Atyrau 20.5 Mangystau, North Kazakhstan Sugar 1.0 Almaty city, Karagandy 8.4 Atyrau, Pavlodar Vegetables 1.6 Karagandy, North 14.4 Almaty city, Pavlodar Kazakhstan Dairy 2.3 South Kazakhstan, Akmola 23.6 Karagandy, West Kazakhstan products Flour 2.4 Almaty region, Akmola 14.6 East Kazakhstan, Kyzylorda Oil and fats 1.4 South Kazakhstan, 14.3 Astana, North Kazakhstan Zhambyl Source: World Bank estimates. Statistical analysis reveals that distance and population density are correlated with market integration (Table 3.6). Regression results suggest that the greater the distance between two regions, the slower the adjustment (Annex 3E)—that is, the more time needed for prices that have deviated to return to their equilibrium (column 1). This result also supports findings from the structural gravity model above. Kazakhstan Country Economic Memorandum 59 And higher population density increases the speed of adjustment, reducing the time needed for the price in one region to adjust to a price shock in another (column 4). This last result suggests that more population-dense regions might have more enterprises per capita, which can help markets respond to changes in price signals. Higher density areas could facilitate faster information exchange, which reduces the chance of market segmentation. Table 3.6 Spatial dimension in price convergence across regions Coefficients from Tobit estimator (1) (2) Log distance between regions i and j −0.043 (−2.10)* Log density in region i 0.209 (2.65)** Log density region j 0.123 (1.19) θs, ηs, and ωs dummy variables Yes Yes Constant 0.444 −0.463 (3.01)** (−1.71)† Number of observations 2,400 2,400 F-statistics 17.2 17.2 Log-likelihood 1534 1534 Source: World Bank. Note: Numbers in the parentheses are z-statistics. † significant at p < .1; * significant at p < .05; ** significant at p < .01. Conclusion This chapter highlights three emerging findings on the spatial dimension of trade in Kazakhstan. First, economies in large cities and regions with links to international trade are a critical driver of the increase in Kazakhstan’s trade. Much of the intraregional trade (within region), interregional trade (among regions), and imports are conducted by large cities (Almaty and Astana) and the surrounding regions (Almaty region and Akmola). Higher imports in these cities and regions are also associated with higher manufacturing concentration. The analysis also finds that interregional trade is higher between regions with access to international border crossings. Second, most of the trade within Kazakhstan occurs between locations less than a thousand kilometers apart. Economies of scale in production and transport are expected to reinforce and perpetuate the trend. Distance remains a crucial part of intranational trade costs in Kazakhstan. Improved transport links would strengthen the incentives for businesses to locate closer to markets and labor pools, which would in turn continue to benefit from agglomeration. The rapid market and population expansion in Almaty city and Astana are attracting businesses in the surrounding regions of Almaty and Akmola. Regions with access to border crossings are also experiencing a greater concentration of manufacturing around specific locations. These agglomeration forces can further increase intraregional trade, so unless economically distant places have a strong natural advantage (such as natural resource deposits or favorable land characteristics), they will likely find it hard to join production chains competitively. In this context, allowing resources to flow more freely could increase productivity from agglomeration. Kazakhstan Country Economic Memorandum 60 Third, closely located regions’ markets tend to integrate, and markets in higher-density regions adjust more quickly to shocks. Specifically, our analysis of food products suggests that regions’ markets react faster to changes in nearby regions, so that prices that have deviated return more rapidly to parity. Similarly, markets in regions with higher population density (suggesting more economic concentration) also adjust with greater alacrity to shocks. Given the vast territory and dispersed population, improving trade and transport connectivity is critical for Kazakhstan to exploit growth opportunities from trade. Kazakhstan should consider a broad agenda of policy reforms and infrastructure development, such as: • Promoting agglomeration in cities to sustain growth in trade and market integration. • Improving customs and trade facilitation by expanding implementation of single window/e- customs in all major border crossings and integrating customs control with customs a risk management system; and expanding the use of a electronic clearance process for sending trade- related documents to other relevant agencies and ministries. • Facilitating growth in competitive and quality logistics services, such as by leveling the playing field for private sector investments, reducing barriers to access road and railway infrastructures, and create more competition in railway logistics services by unbundling those services from the state railway company. • Improving connectivity between cities and surrounding districts through local roads, and connectivity between cities through transport infrastructure and quality telecommunication services. References ADB (Asian Development Bank). 2020. CAREC Corridor Performance Measurement and Monitoring Annual Report 2019. Manila: Asian Development Bank. Agnosteva, Delina E., James E. Anderson, and Yoto V. Yotov. 2019. “Intra-national Trade Costs: Assaying Regional Frictions.â€? European Economic Review, 112, 32–50. Anderson, James E., and Eric van Wincoop. 2003. “Gravity with Gravitas.â€? American Economic Review, 93, 170–92. Arvis, Jean-Francois, and Ben Shepherd. 2013. “The Poisson Quasi-Maximum Likelihood Estimator: A Solution to the ‘Adding Up’ Problem in Gravity Models.â€? Applied Economic Letters, 20, 515–19. Arellano, Manuel, and Stephen Bond. 1991. “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.â€? Review of Economic Studies, 58: 277–97. Baffes, John. 1991. “Some Further Evidence of the Law of One Price: The Law of One Price Still Holds.â€? American Journal of Agriculture Economics, 73 (4), 1264–73. Breinlich, Hoger. 2006. “The Spatial Income Structure in the European Union: What Role for Economic Geography?â€? Journal of Economic Geography, 6, 593–617. Cechetti, Stephen G., Nelson C. Mark, and Robert J. Sonora, 2002. “Price Index Convergence among United States Cities.â€? International Economic Review, 43, 1081–99. Kazakhstan Country Economic Memorandum 61 Crucini, Mario J., Mototsugu Shintani, and Takayuki Tsuruga. 2014. “Noisy Information, Distance, and Law of One Price Dynamics across US Cities.â€? Working Paper 216, Federal Reserve Bank of Dallas, Dallas, TX. Datta, Saugato. 2012. "The Impact of Improved Highways on Indian Firms." Journal of Development Economics, 99 (1), 46–57. Ellison, Glenn, and Edward L. Glaeser. 1997. “Geographic Concentration in U.S . Manufacturing Industries: A Dartboard Approach.â€? Journal of Political Economy, 105, 889–927. Federico, Giovanni. 2005. “Market Integration and Market Efficiency: The Case of 19th Century Italy.â€? Explorations in Economic History, 44, 293–316. Hanson, Gordon H. 2005. “Market Potential, Increasing Returns, and Geographic Concentration.â€? Journal of International Economics, 67, 1–24. Kitenge, Erick M., and A.K.M Mahbub Morshed. 2019. “Price Convergence among Indian Cities: The Role of Linguistic Differences, Topography, and Aggregation.â€? Journal of Asian Economics, 61, 34– 50. Kurmanbekov, A., and M. Temirkhanov. 2015. “Trade Sector of Kazakhstan.â€? Halyk Finance, Halyk Bank, Almaty. Mussin, Arman. 2017. “Patterns of Informal Trade in Petropavl, Kazakhstan.â€? Thesis submitted to the School of Humanities and Social Sciences of Nazarbayev University. Nerlove, M. 1963. “Returns to Scale in Electricity Supply.â€? In C.F. Christ, ed., Measurements in Economics: Studies in Mathematical Economics and Econometrics in Memory of Yehuda Grunfeld . Stanford: Stanford University Press. Persyn, D., and J. Westerlund. 2008. “Error-Correction–based Cointegration Tests for Panel Data.â€? Stata Journal, 8 (2), 232–41. Pesaran, M. 2007. “A Simple Panel Unit Root Test in the Presence of Cross-section Dependence.â€? Journal of Applied Econometrics, 22 (2), 265–312. Pesaran, M., and Y. Shin. 1999. “An Autoregressive Distributed Lag Modelling Approach to Cointegration.â€? In Steinar Strøm, ed., Econometrics and Economic Theory in the 20th Century. The Ragnar Frisch Centennial Symposium. Cambridge, UK: Cambridge University Press. Pesaran, M., and R. Smith. 1995. “Estimating Long-run Relationships from Dynamic Heterogeneous Panels.â€? Journal of Econometrics, 68 (1), 79–113. Redding, Stephen J. 2020. “Trade and Geography.â€? Working Paper 27821, National Bureau of Economic Research, Cambridge, MA. Redding, Stephen J., and A.J. Venables. 2004. “Economic Geography and International Inequality.â€? Journal of International Economics, 62, 53–82. Shirley, C., and C. Winston. 2004. “Firm Inventory Behavior and the Returns from Highway Infrastructure Investments.â€? Journal of Urban Economics, 55 (2), 398–415. Silva, Santos J.M.C, and Silvana Tenreyro. 2006. “The Log of Gravity.â€? The Review of Economics and Statistics 88 (4): 641–58. Kazakhstan Country Economic Memorandum 62 Tan, Shawn, and Dea Tusha. 2020. “Agglomeration Patterns in Kazakhstan.â€? Background paper produced for the Kazakhstan Joint Economic Research Program (JERP), Washington, DC. UNECE (United Nations Economic Commission for Europe). 2019. Logistics and Transport Competitiveness in Kazakhstan. Geneva: UNECE. Venables, Anthony J. 2020. “Why some places are left behind: urban adjustment to trade and policy shocks.â€? Oxford Review of Economic Policy, 36 (3), 604–20. Westerlund, J. 2007. “Testing for Error Correction in Panel Data.â€? Oxford Bulletin of Economics and Statistics, 69 (6), 709–48. Kazakhstan Country Economic Memorandum 63 Annex 3A Calculating the market access index The market access index is calculated based on a methodology proposed by Redding and Venables (2004). Because the data represent the flow of imports by region i from region j, the market access variables are defined as: 𝛾 𝜃 𝜃2 𝜃3 𝑀𝐴𝑆𝑖 = 𝑒 𝛼𝑖𝑖 𝑇1−𝜎 𝑖𝑖 + ∑ 𝑒 𝛼𝑖𝑗 exp(ln 𝑑𝑖𝑠𝑡𝑖𝑗 ) exp(𝐷𝐴𝑑𝑗𝑖𝑗 ) 1 exp(ln 𝑑𝑖𝑠𝑡𝑖𝑗 𝐷𝐴𝑑𝑗𝑖𝑗 ) exp(í µí°·í µí°¼í µí±›í µí±¡í µí±’í µí±Ÿí µí±?𝑖𝑗 ) 𝑖≠𝑗 𝛾 𝜃1 𝜃2 𝜃3 𝑀𝐴𝐵𝑗 = 𝑒 𝛽𝑗𝑗 𝑇𝑗𝑗 1−𝜎 + ∑ 𝑒 𝛽𝑖𝑗 exp(ln 𝑑𝑖𝑠𝑡𝑖𝑗 ) exp(𝐷𝐴𝑑𝑗𝑖𝑗 ) exp(ln 𝑑𝑖𝑠𝑡𝑖𝑗 𝐷𝐴𝑑𝑗𝑖𝑗 ) exp(í µí°·í µí°¼í µí±›í µí±¡í µí±’í µí±Ÿí µí±?𝑖𝑗 ) 𝑖≠𝑗 where MAS is the market access index for region i to suppliers, MAB is the market access index for region j to buyers,114 distij is the distance between regions, dAdj is a dummy variable for neighboring regions, and dInterb is a dummy variable if region i and region j have access to international border crossings. The α and β parameters are the mean of the estimated time and region fixed effects for each region i and j, respectively, and γ is the estimated coefficient for distance from the structural gravity equation. The θs are estimated coefficients for dummy variables and their interaction, as described in Table 3.2 in the main text. 𝑇1−𝜎 𝑖𝑖 is the intraregional trade cost. But because the structural gravity result does not provide an estimate for 1−σ, we proxy 𝑇1−𝜎 𝑖𝑖 with a natural log of intraregional distance raised to the power of the estimated coefficient for distance (γ). If the region has access to international border crossings, 𝜃 exp(í µí°·í µí°¼í µí±›í µí±¡í µí±’í µí±Ÿí µí±?𝑖𝑖 ) 3 is multiplied by the first term. Kazakhstan Country Economic Memorandum 64 Annex 3B Distance as a source of friction for intraregional and interregional trade It appears that intraregional trade increases with the size of a region’s economy and that interregional trade increases with the size of both trading regions’ economies. Both types of trade tend to increase in regions with larger combined economic activities (Figure 3B.1) but tends to diminish as the area of a region or the distance between regions increases (Figure 3B.2). To find whether the data pattern is consistent with the prediction of trade theory, we used a gravity model to further analyze the relationship between distance and size of economic activities. This approach uses distance and other observable variables to approximate trade costs that can be sources of friction in intraregional and interregional trade (just below). The result of a traditional gravity model suggests that intraregional and interregional trade declines with distance: a 1 percent increase in the area of a region or the distance between two regions is associated with 0.9 percent less regional trade (Table 3B.1, column 2).115 Also, the combined size of two economies is positively correlated with regional trade. The estimated coefficient of the partner region (the supplier) is larger than that of the importing region, which might suggest that sourcing goods from regions with larger economies is preferred because such regions can also be a source of demand, minimizing the imbalance that can occur in trade with regions with smaller economies. Figure 3B.1 Intraregional and interregional Figure 3B.2 Intraregional and trade increases with the size of the regional interregional trade declines as the area of economy or combined regional economies, a region or the distance between regions 2014–19 increases, 2014–19 Source: World Bank, based on Committee on Statistics data. Using a gravity model to approximate trade costs for intraregional and interregional trade We used a gravity model to analyze how intraregional and interregional trade varies with distance and the size of economic activities. The approach has been extensively used to analyze friction in international trade but is also applied to patterns of intraregional and interregional trade (Agnostenova, Anderson, and Yotov 2019). We used two gravity model approaches: the traditional approach, which follows Newton’s gravitation, and the structural approach, as proposed by Anderson and van Wincoop (2003). In both approaches, distance represents trade costs. We used 0.66 * (area of region in km2/Ï€) ^ 0.5 to approximate distance for intraregional trade. For interregional trade, we used the travel distance between regions’ business centers following the shortest route for a car in Google Maps. The dependent variable is the value of imports by the region from 2014 to 2019. Kazakhstan Country Economic Memorandum 65 We believe that the estimated coefficient in a structural gravity model is more appropriate to capture friction in intraregional and interregional trade because the coefficient takes into account changes that might affect bilateral trade between regions i and k, as opposed to the traditional gravity model, in which the coefficient captures only the average trade relationship. In the structural model, the specific bilateral trade relationship between two regions is captured by time and region fixed effects, which control for unobservable variables that correlate with distance. The following specification is used to approximate trade costs for intraregional and interregional trade in the structural gravity model: (1 − 𝜎) ln 𝑇𝑖𝑗 = 𝛾 ln 𝑑𝑖𝑠𝑡𝑖𝑗 + 𝜃1 𝐷𝐴𝑑𝑗𝑖𝑗 + 𝜃 2 (ln 𝑑𝑖𝑠𝑡𝑖𝑗 𝐷𝐴𝑑𝑗𝑖𝑗 ) + 𝜃3 í µí°·í µí°¼í µí±›í µí±¡í µí±’í µí±Ÿí µí±?𝑖𝑗 where (1—σ)lnTij is a proxy for trade costs between regions i and j based on series of observable variables, in this case: distij, the distance in kilometers between regions’ business centers, or 0.66 * (area of region in km2/Ï€) ^ 0.5 in the case of intraregional trade; dAdj, a dummy variable for neighboring regions; and DInterb, a dummy variable if region i and region j have access to international border crossings. Results from a structural gravity model suggest that a 1 percent increase in distance is correlated with 1.1 percent reduction in trade. Regression results also suggest two other findings (Table 3B.1). Trade between adjacent regions can happen seamlessly, though it has a lower value than the average value of interregional trade. The negative coefficient of the dummy variable (θ1) suggests that the value of interregional trade among adjacent regions is lower than the average value of interregional trade. But a statistical test cannot reject the null hypothesis that the sum of estimated coefficients for distance (γ), and the interaction between distance and the dummy variable for adjacent region (θ2), equals zero.116 That is, distance does not appear to affect trade between neighboring regions. Direct access to international border crossings significantly correlates with higher intraregional and interregional trade. When buyers and suppliers have direct access to markets in neighboring countries, intraregional and interregional trade increases by an estimated 107 percent compared with the average.117 This finding probably shows the movement of indirect imports between regions with a direct link to international trade routes. It also shows how regions with access to international trade gateways can offer more products that reduce the risks of trade imbalance that often impede interregional trade. Kazakhstan Country Economic Memorandum 66 Table 3B.1 Gravity model regressions on intraregional and interregional trade (14 regions, 2014–19) (Poisson pseudo-maximum likelihood estimator) Dependent variable: imports by region i from region j Traditional model Structural model (1) (2) (3) (4) Log GDP region j 1.018 0.942 (12.79)** (11.41)** Log GDP region i 0.771 0.674 (10.13)** (8.53)** Log distance_ij (γ) −0.860 −0.923 −1.072 −1.068 (−9.82)** (−10.38)** (−16.40)** (−17.78)** Dummy: adjacent region (θ1) −9.321 −8.868 (−5.02)** (−6.01)** Adjacent region*log distance_ij (θ2) 1.275 1.170 (4.55)** (5.36)** Dummy: region i and j have international border 0.071 0.728 crossings (θ3) (0.72) (3.45)** Time dummy variables Yes Yes No No Time and region i, and time and region j fixed effects No No Yes Yes Constant −11.99 −8.789 12.88 13.42 (−6.83)** (−4.53)** (12.56)** (15.48)** R2 0.829 0.841 0.945 0.963 Number of observations 1,176 1,176 1,176 1,176 Source: World Bank estimation. Note: Numbers in parentheses are z-statistics. * significant at p < .05; ** significant at p < .01. Annex 3D Testing for the impact of connectivity and firm level performance As in the traditional industrial organization literature (for example, Christensen and Greene 1976 and Nerlove 1963), the following cost function is estimated using the Arellano–Bond (1991) dynamic panel data model: ′ í µí±?𝑖𝑡 = 𝑓(𝑊𝑖𝑡 ′ , 𝑈𝑖′ , í µí±?𝑖𝑡 ; 𝑦𝑖𝑡 ). This supposes that firm i at time t minimizes its production cost (c) to produce output y (total production or sales) by choosing inputs X’s. Five inputs are considered: labor, energy, fuel, interest paid, and other inputs. The input prices are given by W. The firm’s productivity is also assumed to depend on a variety of time-invariant firm-specific characteristics, Ui, such as ownership structure and the number of years in operation as a proxy of the firm’s experience in the industry, as well as time - Kazakhstan Country Economic Memorandum 67 variant characteristics around firms (denoted by Zit). All time-invariant firm-specific effects are removed by the first difference transformation of the dynamic panel model. The elasticity of firm inventory is estimated using the following dynamic panel data model ln 𝑄𝑖𝑡 = âˆ‘í µí±? í µí»¼í µí±? ln í µí±„í µí±–í µí±¡âˆ’í µí±? + ∑𝑘 𝛽𝑣𝑘 ln 𝑣𝑘 + 𝛽𝑦 ln 𝑦 + 𝛽𝐼 í µí°¼í µí±?𝑇 + 𝛽𝐿 ln 𝐿 + 𝑋 ′ 𝛽𝑥 + 𝑢𝑖 + 𝑢𝑡 + 𝜀. The equation follows a simple economic order quantity model, which states that optimal inventory increases with demand size (represented by a firm’s total production or sales, y) and decreases with transport connectivity vk and interest rate (INT). To control for size differences across firms, the number of employees (L) is also included with other firm-specific characteristics, X. For transport connectivity (V), the same variables are used: CITY, BP, and MAI. Again, the equation was estimated by the dynamic panel data model. To examine the impact of connectivity on agglomeration, the following equation is estimated by the dynamic panel model: ln í µí±?𝑗𝑡 = âˆ‘í µí±? í µí»¼í µí±? ln í µí±?í µí±—í µí±¡âˆ’í µí±? + ∑𝑘 𝛽𝑣𝑘 ln 𝑣𝑗𝑘𝑡 + 𝑢𝑗 + 𝑢𝑡 + 𝜀 where Njt is the number of firms registered in location or district j at year t. The model not only eliminates any time-invariant location-specific characteristics uj but also addresses the potential endogeneity of transport connectivity vk. Kazakhstan Country Economic Memorandum 68 Annex 3E Testing for market integration and its spatial dimension Finding the extent to which markets in two regions are integrated involves detecting whether prices in those regions have a long-term, cointegrating relationship. To analyze this, we used the monthly prices from 2011 to 2020 of 10 food products across 16 Kazakhstan regions and cities, collected by the Committee on Statistics. Data for South Kazakhstan region were available only up to 2018, when the region was split into Turkistan region and Shymkent city. We use the following error correction approach to test the presence of market integration:118 Î”í µí±?𝑖𝑘𝑡 = 𝜆𝑖𝑗𝑘 (í µí±?𝑖𝑘 𝑡−1 − 𝛽 0𝑘 − 𝛽 𝑘 í µí±?𝑗𝑘 𝑡−1 ) + 𝜎𝑘 Î”í µí±?𝑖𝑘 𝑡−1 + 𝛾𝑘 Î”í µí±?𝑗𝑘 𝑡−1 + 𝜇𝑖𝑘𝑡 where p is the natural logarithm of prices of product k in region i at time t. The bracketed term in the equation is the cointegration term, which captures the equilibrium relationship between prices in region i and region j. Correspondingly, the 𝜆𝑖𝑗𝑘 is the speed of adjustment for prices that have deviated to return to their equilibrium.119 We use a regression to further pin down the effect of a spatial dimension on the speed of price convergence. The following specification estimates how a spatial variable affects the speed of adjustment for prices in two regions to converge: 𝜆𝑖𝑗𝑘 = 𝑎0 + 𝑎1 𝑋ij +𝜂 𝑖𝑘 + 𝜃𝑗𝑘 + 𝜔𝑖𝑗 + 𝜖𝑖𝑗𝑘 where λijk is the speed of adjustment of price in region i from a shock in region j for product k and Xij is a spatial variable, such as distance between region i and j or density in region i and region j. The terms ηik and θjk are product- and region-specific effects (between 10 products and 16 regions) to control for unobserved factors such as the taste of a certain product in region i and product market characteristics in each region. The term ωij controls for specific but unobserved characteristics between regions i and j (16 x 16 regions), such as connectivity and logistics services between the two regions. Finally, εijk is the error term, which is assumed to be independently and identically distributed. Because the dependent variable is truncated, ordinary least square will not deliver consistent coefficient estimates, so we used the Tobit estimator. Kazakhstan Country Economic Memorandum 69 Annex 3F Short-run movement of regional prices We checked whether each price had a unit root. Because individual unit root tests have limited power, we used Fisher-type and Pesaran panel unit root tests. The Pesaran panel unit root test in the presence of cross-section dependence was developed by Pesaran (2007). The test considers the asymptotic behavior of time series and cross-sectional dimensions of the series by considering cross-sectionally dependent, as well as serially correlated, errors. The test results show that most variables were nonstationary in levels and stationary in month-over-month measures. We performed an error correction–based panel cointegration test proposed by Westerlund (2007). That test has three advantages over other residual-based panel cointegration tests: it is normally distributed and allows for a large degree of heterogeneity in the long-run relationship and short-run dynamics (Persyn and Westerlund 2008), it accounts for cross-sectional dependence across products, and robust critical values can be obtained through the bootstrapping procedure. Because the regional price index series are cointegrated, we then estimated the speed of price convergence between regions using the error correction model approach for nonstationary heterogeneous panels (Pesaran and Shin 1999; Pesaran and Smith 1995). The error correction model describes the short-run movement of regional prices in response to a change in other domestic or international prices as they converge back to the long-run equilibrium. The error correction model is speciï¬?ed using the following autoregressive distributive lag dynamic panel speciï¬?cation: Î”í µí±?𝑖𝑘𝑡 = 𝜆𝑖𝑗𝑘 (í µí±?𝑖𝑘 𝑡−1 − 𝛽 0𝑘 − 𝛽 𝑘 í µí±?𝑗𝑘 𝑡−1 ) + 𝜎𝑘 Î”í µí±?𝑖𝑘 𝑡−1 + 𝛾𝑘 Î”í µí±?𝑗𝑘 𝑡−1 + 𝜇𝑖𝑘𝑡 where pi and pj are logarithms of price indexes in different regions. The subscript k = 1, 2 … 10 corresponds to the number of products, and the subscript t = 1, 2 … 111 corresponds to the time periods; λ is the error correction (speed of adjustment) parameter, which represents the estimated pairwise coefficients for each product. The parameter of interest is expected to be negative to show the speed at which regional price indexes return to long-run equilibrium. The coefficient β0k is a nonzero mean of the cointegrating relationship; βk, σk, and γk are coefficient vectors; and μikt is the error term. Kazakhstan Country Economic Memorandum 70 Part 2 Place-based policies reshaping economic geography Kazakhstan Country Economic Memorandum 71 4 Assessing place-based investments The spatial structure of Kazakhstan’s economy is rapidly becoming more important for its future development. The existing structure reflects the vestiges from the planned economy: a lot of production is still in places where it would not be under a market system—which is near consumers, near other firms, or near the borders. And although the spatial transformation has been advancing since the 1990s, the advances have been slow. Kazakhstan now needs to move toward a spatial structure driven much more by market principles (chapter 4). Following markets, new and existing firms would choose to locate in a particular place if it has a good business environment, which requires good local governance, amenities, and infrastructure and place-based government policies. Low transport and trade costs facilitated by good infrastructure would help firms source inputs and send their final products more cost effectively. Free internal movement of workers would ensure that firms can locate in a region and still have access to workers from other parts of the country. Kazakhstan’s national development plan offers a new model of economic growth featuring diversification, reduced dependence on natural resources, and a transition toward productivity growth driven by technology, innovation, and human capital accumulation. At the same time, Kazakhstan wants to maintain territorial cohesion and ensure that growth benefits all regions. The government has improved its territorial development policy but needs to further adjust its goals. Chapter 4’s analysis suggests that it is struggling to create a policy framework where favorable structural conditions empower subnational governments to implement targeted projects that address local economic development challenges. Territorial cohesion requires complementing spatially uneven economic growth with reforms and investments to ensure access to opportunity for people, no matter where they live, and targeted measures to address market failures that keep places from reaching their full economic potential. The guidelines for assessing place-based policies and projects in this chapter are fairly straightforward, and we recommend that the framework presented here be used for the development and evaluation of place-based project proposals. All policies adopted and projects selected should prove that the proposed investments and incentives address market failures and will not simply relocate some existing economic activity. Principles for assessing place-based policies and projects Given the competing claims for assistance and resources the large, upfront costs of many public investments, and the often-long-lasting nature of the assets, the choices of places getting the intervention and the type of interventions should be informed by a realistic, objective, and systematic appraisal of projected policies and projects. The framework prepared by Duranton and Venables (2018) appraises the impact of place-based interventions in a systematic and rigorous manner. The framework presented here is broad enough to be relevant and adaptable to a variety of circumstances, yet rigorous enough to provide a comprehensive measure of the value of a project (Figure 4.1). Even where quantification is difficult, the framework provides a disciplining exercise to help policymakers ask the right questions about likely impacts and insulate policy decisions from pure political calculus. Although the framework requires gathering considerable information—much of it difficult to obtain—the very process of asking questions and attempting to gather information helps discipline thinking about how to evaluate a project. Kazakhstan Country Economic Memorandum 72 Figure 4.1 A framework for appraising place-based policies Source: Elaboration based on Duranton and Venables 2018. Note: GVCs = global value chains. An economic appraisal of a proposed place-based policy or project should rest on the following principles: Provide a clear narrative. Policymakers need to develop a narrative that clearly lays out the market failures or distortions that, if addressed, would foster the region’s progress and drive the design of the intervention. Fully describe the direct and indirect quantity changes. A complete appraisal must take a dispassionate look beyond the expected direct effects to the indirect effects, which are harder to measure. Such indirect effects are often invoked as the critical tipping consideration in defending a policy. But they usually are not well documented, and the arguments usually are not supported with empirics. Consider complementary conditions and policies. Market failures often come in multiples, implying that cost-benefit analysis of any one intervention will be misleading because of the complementary effects of resolving several at once. Consider general equilibrium effects and displacement effects. The former are the quantity changes that occur in response to changes brought about by the policy, and the latter, the changes in one place occurring at the expense of another. Place values on the quantity changes. After the direct and indirect quantity factors have been identified and assessed, the next step is to place a value on quantity changes (social valuation). Be candid and explicit about government capabilities. Many place-based policies require interventions with multiple dimensions, large budgets over long periods of time, and well- developed government capabilities for diagnosis, design, and implementation. Limited capabilities may mean, even if a program is appraised well and appears to yield good benefits, that the returns may be low in practice. Policy design should thus look for ways to limit the burden on government. Provide a clear narrative It is critical to have a clear narrative specify the main problem and the market failures that motivate the place-based policy. Frequently, the primary motive for place-based policies is the creation of new jobs in the targeted place that are expected to result from increased investment and economic activity. Kazakhstan Country Economic Memorandum 73 However, policy is constrained by lumpiness and persistence of geography, and hence places vary in their suitability for growth and, indeed, fundamental viability. It is thus critical to have a well-argued narrative that is specific and transparent about the barriers to growth a region faces and what the policy is supposed to achieve. The causal mechanism to substantiate why the proposed measures are the best way to address the desired outcomes and point to uncertainties in achieving these outcomes should be described in detail. The narrative is intended to address the key market failures that motivate the policy. These could include externalities created by the location—and concentration—of economic activity. Some are positive (clustering agglomeration effects), while others are negative (urban crowding, congestion, and pollution). The failures could also include imperfect labor mobility; many barriers can impede mobility, especially in the short term. Among them are missing information, inadequate resources to move, mismatches in skills, social attachments to place, and restrictions on internal migration. In addition, the failures could include path dependence. Location choices of firms or households are typically major decisions that entail large sunk costs—and, if structures are being built, that create long-lived assets. Expectations of future returns are therefore critical. Agglomeration economies mean that the returns to investing in a place depend on who else is (or is expected to be) there. This in turn creates a coordination problem: no one wants to move to a new place while uncertain about its future development. Also to be considered are displacement effects: other sectors or areas might contract in response to the broader effects of place-based policy intervention. In general, the fundamental determinants of market failure should be diagnosed first and then addressed by targeted policy. For instance, if unclear land rights hamper investment in infrastructure and housing, the first-best policy is to clarify these rights. Place-based policies may not be the first- best policy. Instead of pursuing interventions to influence the exact location of economic activities (place-based interventions), governments can influence access to opportunities for people (people- based interventions), or they can reduce distortions that constrain markets through broad-based national interventions (institutional interventions). Fully describe direct and indirect quantity changes The impact of a place-based policy is often intended to be wide ranging, affecting many aspects of the economy. The overall effect of a policy can be broken down into direct and indirect quantity changes. These direct and indirect quantity effects have a net social value—which in turn arises from the net interaction of quantity changes with market failures and inefficient resource allocation or from equity concerns. The expected quantity changes arising from the project include the changes in real economic activity induced by the policy compared with conditions without the policy—that is, “business as usual.â€? In turn, quantity changes can be broken down into direct effects and indirect effects. Measuring quantity effects—direct and indirect—as well as placing a “social valuationâ€? of these changes is challenging (Figure 4.1). Direct effects are fairly simple to measure, at least in principle. It is much harder to identify and assign value to indirect effects because they are contingent on underlying assumptions about how economic linkages work in a given context. Direct effects are the immediate impacts on users of the project. They include the impact on, for instance, economic activity, assuming that other factors of production and the technology do not change due to the intervention. In a typical road infrastructure project, for example, direct effects would Kazakhstan Country Economic Memorandum 74 include how a road benefits users by saving time and lowering vehicle operating costs. For a project to build or expand an airport, direct effects would include how the airport could facilitate a larger number of flights, benefiting passengers and those shipping cargo that would otherwise have had to be shipped more slowly, less directly, and at greater expense. Place-based policies also often anticipate indirect effects such as job creation and higher wages: that is, induced changes in the location and levels of activity, alongside changes in inputs and the efficiency of combining those inputs. These effects may arise if the policy changes private sector behavior (such as firms deciding to relocate, or workers switching from an agrarian to an industrial job) or triggers a private sector investment response. It is often claimed (or hoped) that place-based policies will trigger the regeneration of a district, or growth in a lagging region. Consider complementary conditions and policies Location and investment decisions inherently are not choices made at the margin: they are either-or choices, rather than fractional adjustments. They are costly, typically incurring sunk costs and long- lasting assets, and are shaped by expectations. The private sector will decide to invest in a place, for instance, only if multiple conditions are met. These conditions include the natural characteristics of a place, the policy environment, and the business ecosystem. Weaknesses in any of these necessary complements may stymie private sector investment and lead to no or few indirect effects. Again, the basic geographical endowments and distance from markets may imply that a region is nonviable and that place-based policies are a nonstarter: there is no road or big push effort that is going to breathe life into some areas. In a less extreme case, good roads for exports will not be enough if a place lacks reliable electricity or the necessary human capital. It is especially difficult to assess multidimensional projects or big push initiatives. The temptation to assess individual elements separately contradicts the notion that there are strong complementarities among them. Quantifying spillovers and scale effects is also difficult. Relying on the aggregate outcomes—such as overall employment or GDP per capita in the treated regions relative to the untreated—is often difficult to disentangle from other aggregate changes in the economy unrelated to the policies at hand. Moreover, the initiative may have displaced labor or industry from “untreatedâ€? regions, thus overstating the nationwide social benefit. All this occurs against a backdrop in which investments are often dispersed across wide geographical areas, the amounts invested are modest, and many of the polices are ongoing, thus making it difficult to conduct a rigorous impact evaluation. Barriers posed by natural geography or remoteness do not constitute a market failure, but they may impose a high underlying cost revealing that it is not worthwhile to devote resources to that region. The general equilibrium effects of a policy or project, particularly the displacement effects in factor and product markets, need to be taken into account to fairly identify the national benefits. Merely moving a plant or jobs from one place to another has no benefit. Place-based policies thus need to be weighed against other policies or projects, such as those facilitating migration or transitional fiscal transfers and service provision, which may provide better value. Complementarities also include the policy environment and the business ecosystem (Figure 4.1). The policy environment covers national variables and those that are specific to a place: Infrastructure including utilities and transport. Place-specific tax and regulation, as in a special economic zone. Policy as it affects labor supply, covering such factors as public services, commuting, and housing. Kazakhstan Country Economic Memorandum 75 Institutions, including the clarity and enforcement of property rights and contracts. The items included in the policy environment cover different areas of government, so the effectiveness of the policy environment is contingent on coordination across space, function, and time. Policies need to be integrated functionally: that is, they must cover planning, land, and building regulations as well as infrastructure and the provision of utilities and public services. Policies also need to take a long view, meaning that policy makers must be able to make credible commitments to future city development. This would require some coordination and consistency between the different levels of government at the local, regional, and national levels. Having competence, financial resources, and credibility to meet these challenges requires an authorizing environment better integrated between different parts of government than is typically found, especially in developing countries (World Bank 2017). The business ecosystem refers to the network of organizations—including suppliers, distributors, competitors, customers, and workers—that contribute to the performance of firms and the value of investment decisions that they undertake in a particular place. This includes: Related firms: the stock of firms and other productive activities, in particular its suppliers and customers. Workers: the supply of workers with appropriate skills at competitive wages, or the potential to attract migrants to the area. The availability of other complementary factors, land, and capital. Market size: the size of markets to which the place is well connected. The items in the business ecosystem largely mirror those concerning agglomeration and clustering. Conditions that determine one investor’s decision depend not only on the business climate, but also on decisions that have been—or will be—taken by other private agents. This extends across a wide range of agents and depends on expectations. The location decisions of firms thus depend on those of workers and other firms. The decisions of workers depend on firms and on housebuilders, who take a long view about employment and population in the place, and so on. All these points indicate different ways for policy to shape private investment and location decisions. Some are dealt with in the policy environment, but others (called soft complements) involve shaping expectations about the business ecosystem, including credible signals about the government’s commitment to a place, effective investments agencies, and a responsive government that will credibly remove future blockages and obstacles. Evaluating the direct effects individually of packages of complementary policies is misleading because, by design, these packages of policies expect important interactions among them. Consider general equilibrium effects and displacement General equilibrium effects are the quantity changes that occur—possibly in quite different places—in response to changes brought about by the policy. While total factor productivity may change, changes in those factors—notably labor and capital—may merely be shuffled from one location to another. As a result, the investment induced by a place-based intervention may occur in one place at the expense of another: that is, investment may merely be displaced (Figure 4.1). Displacement effects can occur through several distinct routes. The first is competition for a particular project, such as a single factory that will operate in only one of multiple possible places. Second, displacement can occur through a product market in which, if demand is inelastic, an increase in supply in one place will be met by a reduction in supply elsewhere. This effect is most pronounced for Kazakhstan Country Economic Memorandum 76 nontradable goods, where demand comes only from a local or national market. Third, displacement can be channeled through factor markets. If the supply of capital is fixed or labor is fully employed, then expansion of one activity is bound to be accommodated by contraction of another. It is not always either necessary or possible to identify general equilibrium effects with precision. However, if capital and labor are simply being reshuffled between uses, then both sides of these quantity changes must be taken into account. The principle of market valuation discussed next requires establishing both the value of a factor in its new use and the opportunity cost of this employment. Place values on the quantity changes After the direct and indirect quantity factors have been identified and assessed, the second stage in policy appraisal is to place a value on quantity changes (social valuation). Valuing the direct effects (the effect of a change in policy, given other factors constant) is usually straightforward and such valuations are generally the main focus of appraisals. Indirect effects are less straightforward to measure because they have net value only if policy works to correct inefficiencies: that is, if it draws resources from a lower-value use to a higher-value use. Otherwise, the policy will simply be drawing resources from one place to another: it will have no social value. So what matters is the incremental change, captured through the notion of marginal value—the change in the value of labor, capital, or other resources when switched from one activity to another. In the face of market distortions, the private marginal cost on which the firms base their decisions is not the same as the social cost. For example, if labor is being drawn from an underemployed stock of labor with a low opportunity cost, then the social cost of moving labor is lower, creating a further channel of benefit. A way to capture value changes is to appraise labor at a “shadow priceâ€?: that is, capture indirect effects by using the social opportunity cost of labor instead of using the market wage in cost calculations. By comparison, in the benchmark case with no distortions or market failures, capital and labor are fully employed, and their marginal products (the change in output attributed to a change in input) are equated across all industries and regions of the economy. Hence, a project that increases employment in a region will merely be displacing labor from elsewhere. Evaluating the indirect benefits thus requires identifying existing distortions or market failures that policy will effectively be mitigating. In fact, there are many market failures. The task of the appraisal is to identify is the most relevant ones and place a value on redressing them. Some examples: Labor market inefficiencies across space. Most often, the hoped-for indirect impact of a place-based policy is the creation of jobs. As chapter 2 shows, migration is not so fluid as to eliminate differences in unemployment rates or wage gaps among regions, so creating new jobs in a depressed region has social value. Computing this social value for an appraisal depends on capturing the inefficiency of local labor markets. Distortions in capital and land markets. In urban areas, land is the ultimate scarce factor, so there is high return to using it efficiently. However, unclear property rights and obstacles to trading land can impede it from being allocated to the most productive use. Building is impeded by failures in capital markets (particularly for residential mortgages) and in some cases inappropriate building and land use regulation. Place-based policies that aim to address these imperfections yield direct benefits if they enable land to switch from a low-value to a higher-value use. Multiple effects are likely. For example, Kazakhstan Country Economic Memorandum 77 a regulatory change that enables efficient use of a piece of urban land might yield social benefits by raising land values not only directly, but also indirectly by creating better jobs, and encouraging the positive economies in the urban cluster. Nonmarket effects (externalities). Many place-based policies seek to remedy outcomes arising precisely because allocations occur outside or external to markets. For example, a light rail project may seek to diminish congestion or pollution—disamenities not internalized by individual actors through the market and hence overproduced. Similarly, knowledge spillovers are not intermediated through the market and thus are undersupplied. Coordination failures. The balance between agglomeration benefits and costs can lead to clusters or cities that are too big because externalities are not appropriately internalized by firms. But an individual firm will not leave to start a new agglomeration unless others accompany it, lest it lose the benefits of being around other firms. Place-based policies whose indirect effect is to resolve this coordination failure thus add value by reducing the social losses arising from congestion or giving life to a new region. Transport costs in developing countries are still four to five times higher than in advanced economies, which makes the need for coordination greater—and makes the cost of coordination failure greater. Be candid and explicit about government capabilities The Kazakhstan government has finite capabilities to appraise and execute policies. This reality must be integrated into laying the best of plans. The number of market failures that a potentially viable region suffers from, and hence the multiplicity of necessary policy interventions, increases with distance from the frontier, while policy capability to evaluate and implement decreases (Cirera and Maloney 2017). So, not only is the national government facing a spatial landscape with vast inertia, with regions that are simply not viable, but its tools to sort out what is feasible and then implement policy are bounded. Identifying “doableâ€? combinations of policies becomes as important as designing policies that, if perfectly implemented, would yield the highest returns. In closing, four suggestions based on a review of World Bank project documents nest well within the framework presented here (World Bank 2012). 1. Understand investor behavior. Projects pursue a certain linear approach from “input to output to outcomeâ€? that assumed that increased agglomerations would ease coordination failures in sectors and locations and thereby induce a critical mass of firms to enter. However, the projects did not consult sufficiently with the private sector in the planning stage to determine the potential “distance from the thresholdâ€? where costs and risks would justify such entry, as well as the uncertainty surrounding private sector responses and the likely discontinuous nature of payoffs around the threshold. This can be seen as a combination of insufficient understanding of the market failures and distortions as well as the intrinsic viability of the chosen locales. For instance, the agglomeration-related component of projects in Ghana and Afghanistan included building industrial parks. In Ghana, restrictions on access only to export-oriented firms led to deficient demand, which raises the question of what was necessary, in terms of complements or location to attract them. In Afghanistan, insufficient attention to ensuring power-generators were too costly and connection to the public grid was not implemented, hobbling the project as initially designed. 2. Streamline projects. Many projects involved multiple locales, interventions, and ministries—and many of these projects had unsatisfactory outcomes. No project seriously considered the need Kazakhstan Country Economic Memorandum 78 to manage complexity. It is necessary to document the number of activities and institutions involved, to identify the need for institution building, and to embed flexibility into the project design. If the nature of the complementarities permits interventions to be sequenced, then the complexity of any one component can be diminished. 3. Strengthen mechanisms to deal with complexity. Projects need to include mechanisms to ensure a common interest in the project and to sustain pressure to move forward through the inevitable snags in implementation that will accompany even well-conceived projects. Projects especially need to ensure buy-in of key players from the start to ensure forward momentum, including understanding the incentives and constraints that they face. Again, ensuring local government capabilities is key: part of the challenge in adapting industrial park design in the Ghana and Afghanistan projects arose from the limited expertise of the local implementing agency. The monitoring and evaluation framework for the project becomes particularly important, as it needs to provide accurate and timely reports that monitor progress and reasons for delays. 4. Take quantity and value effects into account. Both quantity and value effects should be accounted for. For instance, projects often have an impact on the value of land and on land tenure security. This can arise from large purchases of land for industrial parks (Ghana and Afghanistan) or large investments in irrigation (Burkina Faso) or infrastructure that facilitates tourism and other commercial activities (Madagascar). Understanding who controls the land and how well the land market functions is as much an issue of political economy as technical design—and is just as important. ___________________________ Ideally, even the simplest road project would have a full appraisal that would allow a solid ranking of projects by their social value added. This would offer some disciplining of the often-formidable pressures to “do somethingâ€? to either reverse the declining fortunes of an area or kickstart a longstanding laggard. However, while the direct effects can often be quantified, doing the same for the indirect effects is expensive, time consuming, and beyond the capabilities of even advanced- economy governments. Often, simpler rules of thumb are employed, sometimes based more on the symptoms than a careful diagnosis of the underlying disease. For example, to be eligible for the local economic growth initiative in the United Kingdom, a local area had to rank fiftieth or worse against any of six indexes of multiple deprivation in 2000 or 2004. Likewise, the French urban enterprise zones program also selects lagging areas based on an “indexâ€? measuring socioeconomic conditions in the area (Mayer, Mayneris, and Py 2017). But none of these focus on viability or attempt to quantify and value overall effects. Given the challenges facing even well established and competent bureaucracies, it is probably better to view the framework here not so much as a mechanical valuation device and more as a heuristic tool that informs the dimensions that should be taken into account, that focuses debate, and that follows some policy guidelines. Kazakhstan Country Economic Memorandum 79 References Cirera, X., and W.F. Maloney. 2017. The Innovation Paradox: Developing-Country Capabilities and the Unrealized Promise of Technological Catch-Up. Washington, DC: World Bank. Duranton, G., and T. Venables. 2018. “Place-based Policies for Development.â€? Policy Research Working Paper 8410, World Bank, Washington, DC. Lall, S.V. 2009. “Bridging Spatial Divides in Growth and Welfare: Identifying Priorities and Evaluating Trade-offs for Infrastructure Policies.â€? Sustainable Development Network report, World Bank, Washington, DC. Mayer, T., F. Mayneris, and L. Py. 2017. “The Impact of Urban Enterprise Zones on Establishment Location Decisions and Labor Market Outcomes: Evidence from France.â€? Journal of Economic Geography 17 (4): 709–52. World Bank. 2012. Internal Portfolio Review. Africa Financial and Private Sector Development Department, World Bank, Washington, DC. World Bank. 2017. World Development Report 2017: Governance and the Law. Washington, DC: World Bank. Kazakhstan Country Economic Memorandum 80 5 Agglomeration patterns Kazakhstan’s national development plan should use territorial development as a policy tool. The plan offers a new model of economic growth featuring diversification, reduced dependence on natural resources, and a transition toward productivity growth driven by technology, innovation, and human capital accumulation. At the same time, Kazakhstan wants to retain territorial cohesion and to ensure that growth benefits all regions.120 Growth based on innovation and widely spread throughout the country has clear spatial implications. Innovation-driven economic growth is often a feature of well-run cities, which attract smart and talented workers and help firms increase their productivity through knowledge spillovers and other agglomeration economies. So, the success of the national economy relies on the success of cities, which takes good infrastructure, good planning and land management, and a good regulatory ecosystem that enables the power of the markets. But territorial cohesion requires complementing spatially uneven economic growth with reforms and investments to ensure people’s access to opportunity, no matter where they live, and with targeted measures to address market failures that keep places from reaching their full economic potential. This moment offers the government of Kazakhstan a chance to address the shortcomings of its territorial development policy and make it a development force. It has been 15 years since the country’s first regional development strategy was established and nine years since the initiation of the State Program for Regional Development (SPRD)—the flagship program for territorial development. In that time, the territorial development effort has demonstrated some results, but the approach to addressing its challenges continues to evolve. Its economic geography has been defined by the legacy of central planning and dependence on extractive industries. But new forces are shaping the future, and in the past two decades, the largest cities have driven economic growth and attracted skilled migrants, signaling that agglomeration forces are boosting productivity. And as Kazakhstan starts to benefit from increasing returns to agglomeration, its territorial development vision and priorities are catching up with global best practices. The key Forecast Scheme for Territorial Development identifies four primary and 14 secondary functional urban areas and rightly identifies large cities as critical to improving economic development and people’s wellbeing. That recognition confirms the country’s departure from the spatially blind development doctrines of the past. At the same time, by introducing service access standards and continuing to implement sectoral state programs, Kazakhstan is maintaining a focus on access to services and infrastructure across all territories and settlements. This combination suggests that Kazakhstan has laid the right foundations to build on. This chapter reviews recent trends in Kazakhstan’s territorial development and analyzes current policy and ways it can improve, including: Laying the right foundations for territorial development with structural reforms, starting with removing barriers to migration. Underpinning territorial development policy with a strong institutional structure based on empowered subnational governments. Moving past a top-down approach and repositioning the SPRD to support integrated, multisectoral local economic development programs implemented by subnational authorities. Kazakhstan Country Economic Memorandum 81 Reshaping Kazakhstan’s economic geography Kazakhstan has tentatively begun to enjoy increasing returns to the agglomeration of economic activity, which—with migration of skilled workers and integration of regional markets—is fundamentally reshaping its economic geography. When Kazakhstan was part of the Soviet Union, state planning shaped its territorial development. Although urbanization was the main macroterritorial development trend in Kazakhstan between 1960 and 1990, the forces that fueled it were shaped by economic planning rather than by the market. Much urban growth was therefore captured by settlements near industrial enterprises placed where GosPlan or a relevant line ministry found it desirable to place them, without regard to market forces. Combined with highly controlled and restricted internal migration, the development approach resulted in large towns in unfitting locations associated with state-planned functions, such as Baikonur—in the middle of the steppe around the main Soviet spaceship launchpad—and Kurchatov—in a deserted area next to the Polygon nuclear testing facility. Even so, the overall concentration of population remained low. Almaty, the biggest city in the country in 1990, had 1.1 million people, which accounted for less than 7 percent of the national population. This represented (and still represents) a level of urban primacy—the share of population in the country’s largest city—far below the level common for countries of comparable size (Figure 5.1). So, when Kazakhstan gained independence, its economy was highly spatially fragmented. Population and industries were spread across many small monotowns, often with fewer than 50,000 people, and several medium regional centers, all but one with fewer than 1 million people. None of those towns and cities had the scale to generate the agglomeration economies associated with higher productivity and innovation. Figure 5.1 Share of population in the main metro areas of countries comparable to Kazakhstan in total population, 2000 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Niger Hungary Ecuador Zambia Greece Cameroon Cuba Cambodia Netherlands Malawi Kazakhstan Ghana Portugal Chile Source: Oxford Economics Global Cities Database. For about 30 years after 1960, the main trend in Kazakhstan’s territorial development was urbanization (Figure 5.2), stopped dead in its tracks by the socioeconomic turbulence of the 1990s, which also hit most other former Soviet republics. As the national economy stagnated and many people from other parts of the former Soviet Union left Kazakhstan, its city growth stalled, and urbanization declined from 54.8 percent of the population in 1990 to 50.3 percent in 2001. Kazakhstan Country Economic Memorandum 82 Figure 5.2 Kazakhstan’s urbanization rate, 1960–2018 60% 58% 56% 54% 52% 50% 48% 46% 44% 42% 40% 1964 2016 1960 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012 Source: World Development Indicators database. That trend changed dramatically around the turn of the century, when rising incomes from exports of extractives brought economic growth. By then, the country had completed the initial phase of building free market institutions, and market forces were becoming more important in territorial development. New centers of wealth creation emerged in extractive regions. But they rarely formed large cities, since the oil and gas industry is not labor intensive and climatic conditions in those places were unfavorable. Export incomes flowed, however, into large urban areas, fueling consumption and leading to growth in local trade, services, and real estate development. And people flowed into the bigger cities as migration regulations eased. Almaty, Shymkent, and the newly established capital of Astana grew fast (Figure 5.3). Other regions and cities struggled to provide economic opportunities for their populations—particularly monotowns that never fully adjusted after losing their place in Soviet value chains. In the 2000s, these factors contributed to a wave of migration that benefited the three largest metropolitan areas (Figure 5.4). Kazakhstan Country Economic Memorandum 83 Figure 5.3 Population growth in Aktobe, Figure 5.4 Net migration for Almaty, Astana, Almaty, Astana, Shymkent, and five regional Shymkent, and macro regions excluding those centers, 1981–2014 urban agglomerations, 2000–19 12% 150,000 10% 100,000 8% 50,000 6% 0 4% -50,000 2% -100,000 0% 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 -2% -150,000 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Almaty Astana Net migration Shymkent North South West East Astana Almaty Shymkent Aktobe 5 regional centers (population > 300,000) Source: UNDESA 2019. Source: Committee on Statistics. Emergence, and subsequent faltering, of the largest urban areas as growth engines In the past 20 years, the largest urban areas emerged as the engines of economic growth. Nighttime light emissions data, used as a proxy for economic activity, show that the four largest urban agglomerations—Aktobe, Almaty, Astana, and Shymkent—have been the drivers of the national economy since the turn of the century, consistently growing faster than other areas, including monotowns and second-tier agglomerations.121 Only those four largest cities grew faster than the rest of the territory (including small towns and rural areas), perhaps showing the economic strength of the extractive industries not located in cities and the continuing economic underperformance of second- and third-tier cities (Figure 5.5). Despite some progress, economic growth in monotowns and second- level agglomerations lags that in major cities (Figure 5.6). The four largest urban agglomerations have grown since 2000 (Figure 5.7). Kazakhstan Country Economic Memorandum 84 Figure 5.5. Nighttime light emissions, 1992–2018 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 First-tier agglomerations Secod-tier agglomerations Monotowns Rest of Kazakhstan Source: Data collected from Li et al. 2020. Note: Uses a harmonized nighttime lights time series dataset (Li et al. 2020). City agglomeration boundaries are defined by applying the urban extent algorithm from the degree of urbanization methodology of the European Union (Dijkstra et al. 2018) using the GHSL-POP geographic population database produced by the European Union. Figure 5.6 Change in average annual Figure 5.7 Change in average annual nighttime light emissions for monotowns and nighttime light emissions for Aktobe, Almaty, first- and second-level urban agglomerations, Astana, and Shymkent, 1990–2018 1990–2018 6.0% 10.0% 5.0% 8.0% 4.0% 6.0% 3.0% 4.0% 2.0% 2.0% 1.0% 0.0% 0.0% -2.0% 1990-2000 2000-2008 2009-2018 -1.0% 1990-2000 2000-2008 2009-2018 -4.0% -2.0% -6.0% -3.0% -8.0% First level Secnod level Monotowns Aktobe Almaty Astana Shymkent Source: Uses a harmonized nighttime lights time series dataset (Li et al. 2020). City agglomeration boundaries are defined by applying the urban extent algorithm from the degree of urbanization methodology of the European Union (Dijkstra et al. 2018) using the GHSL-POP geographic population database produced by the European Union. Skilled migrants predominated in the population flow to the largest agglomerations, confirming that under market conditions, skilled people seek places where other skilled people are abundant. Between 2014 and 2019, the mobility of the unskilled population declined in Kazakhstan (as did the share of unskilled people in the population). But the mobility of highly skilled people kept growing, so that by 2019, 9 percent of the country’s highly skilled population had moved to a different district.122 That mobility clearly correlates with the growing role of large cities as migration destinations. Approximately two-thirds of highly educated migrants in 2019 moved to one of the four largest cities (Table 5.1). And Kazakhstan Country Economic Memorandum 85 as the share of highly skilled people among migrants grew, so did the role of those four cities as migration destinations. Between 2014 and 2019, the share of migrants moving to one of the four major cities grew from 35 percent to 51 percent. The growth confirms that these cities offer economic opportunities to the most educated in more productive and knowledge-intensive sectors of the economy, which is fully consistent with global trends in the migration of skilled people (Glaeser and Resseger 2009). Table 5.1 Top 10 districts for internal migration by education qualifications of migrants, 2019 Share of the national total Share of the national total Share of the national total Share of migrants in the of internal migrants with of internal migrants with of internal migrants with population (percent) high education other education low education qualification (percent) qualification (percent) qualification (percent) Tselinogradskiy Tselinogradskiy Tselinogradskiy 38.3 50.1 26.6 Kostanay 7.9 (Astana) (Astana) (Astana) Bukhar- Almaty 7.7 Almaty 11.2 Ulanskiy 4.7 6.5 Zhyrauskiy Atyrau 3.3 Atyrau 3.2 Almaty 4.0 Zelenovskiy 6.2 Ulanskiy 2.9 Shymkent 3.2 Aktobe 3.6 Kyzylzharskiy 4.6 Aktobe 2.8 Aktobe 2.1 Atyrau 3.4 Aktobe 4.2 Shemonaikhinskiy 2.5 Shemonaikhinskiy 1.6 Shemonaikhinskiy 3.3 Taranovskiy 3.6 Shymkent 2.3 Burlinskiy 1.6 Zelenovskiy 3.2 Almaty 3.6 Bukhar- Bukhar- 2.1 Ulanskiy 1.3 3.0 Shemonaikhinskiy 3.4 Zhyrauskiy Zhyrauskiy Bukhar- Burlinskiy 2.1 1.2 Kostanay 3.0 Zerendinskiy 2.9 Zhyrauskiy Kostanay 2.1 Kyzylzharskiy 1.1 Burlinskiy 2.8 Taldyqorghan 2.7 Source: 2019 Kazakhstan Labor Force Survey. High regional market integration also confirms the predominance of market mechanisms as drivers of territorial development today. As chapter 3 shows, the correlation in prices in different regions of Kazakhstan varies inversely with the distance between the regions. The speed of one region’s reaction to a price shock in another also appears to be determined by distance, the transportation network, and population density. These features indicate that regional markets are fairly well integrated and confirm that market forces have emerged as the key determinants of Kazakhstan’s economic geography. The responsiveness of regional markets to price shocks indicates an economy where market forces incentivize the reallocation of factors of production toward places where they are used most efficiently. For labor, that usually means moving to big cities—exactly what is seen in the migration of skilled workers to Kazakhstan’s largest urban agglomerations. There are disturbing signs, however, that returns to agglomeration are starting to diminish. The results of firm productivity analyses (see chapter 1, “Restarting Kazakhstan’s stalled structural and spatial transformationsâ€?) show that being located in Almaty or Astana strongly predicts high productivity for a firm but much more weakly predicts productivity growth. The four largest cities saw far less impressive productivity growth in 2015–18 than the national leaders—Mangystau and Atyrau, which are extractive regions. Cities have also experienced contracting GDP per capita, explainable by the initially lower productivity of newcomers, but still a negative trend. The diminishing returns to agglomeration would be less worrisome if they indicated a redistribution of capital incentivized by Kazakhstan Country Economic Memorandum 86 high agglomeration costs in the largest cities following the growth of regional disparities due to earlier high returns to agglomeration (the peak of the “Williamson curveâ€?; Williamson 1965). But that does not seem to be happening: Kazakhstan’s three largest urban areas account for 23 percent of the country’s population, while in developed countries, that share averages 48.6 percent. So, the decline in urban productivity at this early stage of urbanization and concentration of the urban system is worrisome (OECD 2020). The disaggregation of productivity growth in large cities points to the limited competitiveness of the country’s tradable sectors, linked to the national economy’s dependence on natural resource extraction. The breakdown of productivity dynamics by sector and source (available for Almaty and Astana) suggests that large cities, particularly Almaty, see little productivity growth in knowledge- based services (information and communications technology, or ICT, and “Other servicesâ€?) (Figures 5.8 and 5.9). The negative between-firm component of productivity growth (change of average productivity because of growth of some firms and decline of other firms on the market) suggests that competition in these sectors is not driving productivity growth, as more productive firms don’t grow faster than less productive. The failure of competition to drive productivity of service industries may be at least to some extent a result of Dutch disease—in this case, the dominance of extractive industries in the economy could be leading to higher prices of production factors (labor and capital), harming the competitiveness of other sectors and limiting their innovation potential. In such conditions, city economies may become overly dependent on local trade and services while having weak tradable sectors. Although the data presented here do not support that interpretation conclusively, it is a possible reason for the underperformance of tradable service sectors in cities. But the poor productivity growth of these sectors also points in part to limited agglomeration benefits for firms in those cities and negative impacts on productivity from growing congestion costs. Figure 5.8 Decomposition by source of Figure 5.9 Decomposition by source of total total factor productivity growth for factor productivity growth for sectors in sectors in Almaty, 2015–18 Astana, 2015–18 NTL growth rate by period 6.0% 4.0% 2.0% 0.0% 1990-2000 2000-2008 2009-2018 -2.0% -4.0% First Level Secod Level Monotowns Source: World Bank, based on administrative firm- Source: World Bank, based on administrative firm-level level data. data. Note: “Betweenâ€? component: average productivity change resulting from growth of some firms and decline of others. “Withinâ€? component: change in average productivity because growth or decline of productivity of individual firms. “Entryâ€? component: change of average productivity because of firms leaving the market. “Exitâ€? Kazakhstan Country Economic Memorandum 87 component: change of average productivity because of firms entering the market. ICT = information and communications technology. Monotowns—getting better but generally still struggling Monotowns in Kazakhstan have performed better economically over the most recent decade than in the prior two but still face challenges. For most of the 20 years after the Soviet Union’s dissolution, many monotowns were in rapid decline. Nighttime light emissions data show that only around 2008 did they return, on average, to their level of economic activity of the early 1990s. But their economic dynamism remains low, and while some towns continue to do well—largely because of favorable conditions in their industry of specialization—most stagnate. Two types of monotowns prioritized by the SPRD—monotowns that are near cities and included in functional urban areas and monotowns that have more than 50,000 people—have been performing better than others (Figure 5.10).123 This is encouraging for economic development, showing that the prioritized monotowns are the right ones for targeted policy support. But the overall poor track record of monotowns and of economic revitalization policies for them across post-Soviet space (see below) and the weak performance and low economic potential of most of them suggest limiting targeted economic development, prioritizing investments in quality of life and human capital, and ensuring that people have access to opportunities elsewhere. Figure 5.10 Change in nighttime light emissions in monotowns, 1992–2018 140% 120% Relative to 1992 level 100% 80% 60% 40% 20% 0% 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Large self-standing monotowns Monotowns in agglomerations Other monotowns Source: Data collected from Li et al. 2020. Note: Uses a harmonized nighttime lights time series dataset (Li et al. 2020). City agglomeration boundaries are defined by applying the urban extent algorithm from the degree of urbanization methodology of the European Union (Dijkstra et al. 2018) using the GHSL-POP geographic population database produced by the European Union. The weakness of monotowns is part of a broader, persistent story of regional inequality that deserves attention but should not necessarily cause major concern. Disparities are high between the economic development of regions in Kazakhstan. The gaps between the leaders are pronounced, with the extractive regions and two major cities (Almaty and Astana) accounting for 74 percent of national growth in 2013–18. The country’s Gini index of regional inequality is three times that of two large and fairly sparsely populated comparators—Australia and Canada—but close to that of two large and Kazakhstan Country Economic Memorandum 88 growing middle-income countries—China and the Russian Federation (Figures 5.11 and 5.12). Regional disparities, after rising rapidly in the early 2000s, have stabilized. The divergence in the early 2000s was fueled by the oil industry boom that benefited extractive regions. In recent years, the divergence has stalled, probably due to the more moderate performance of the oil sector. Regional disparity is not itself negative, as it has historically been associated with periods of high economic growth in many countries, explaining, for instance, why regional development in GDP per capita is more unequal in China than in Kazakhstan. The problem that regional disparity in development can create is regional gaps in public services. In Kazakhstan, substantial regional inequalities in health care and education services were highlighted in recent studies by the Organisation for Economic Co-operation and Development (OECD 2014). Because those services relate closely to human development, shortages in them can limit opportunities for people in the lagging regions and reduce the development potential of the country as a whole. Figure 5.11 Gini index of regional inequality Figure 5.12 Gini index of regional inequality in selected countries, 2018 of GDP per capita in Kazakhstan, 1998–2018 Source: OECD 2020. Source: OECD 2020. Lagging regions and monotowns present the most challenging aspect of persistent territorial disparities. Many countries historically have confronted that challenge, including Canada and the United States, and it has been at the center of the EU policy agenda for many years (Böhme et al. 2011). By the EU definition (in lagging regions, GDP per capita is less than 70 percent of the national average), seven of Kazakhstan’s 17 regions qualify. Almaty region, Zhambyl region, and Turkistan region are at the bottom, with GDP per capita less than half the national average. Two of these laggards surround dynamic metro areas—Almaty region surrounds Almaty, and Turkistan surrounds Shymkent— suggesting development potential from economic links. The World Bank framework for lagging regions emphasizes balanced, space-blind structural policies and selective targeted interventions to overcome local constraints (World Bank 2017). Although the challenges of lagging regions often appear especially hard to overcome, the overall approach to them should not differ drastically from the general approach this chapter advocates. Even so, cities must have assured access to opportunities outside the lagging regions, perhaps requiring special policy measures, such as relocation grants for people who seek opportunities in a different place. Agglomeration of economic activity, migration of skilled workers, and integration of regional markets are reshaping Kazakhstan’s economic geography. Yet, scope remains to amplify market forces and Kazakhstan Country Economic Memorandum 89 accelerate economic growth and diversification. Kazakhstan’s territorial development trends and patterns discussed signal consistency with what can be expected from a market economy: integrated regional markets and improved connectivity incentivize better allocation of factors of production, and migration fuels growth of larger cities, which over time should become drivers of innovation and productivity growth. While not all barriers limiting productivity benefits have been addressed, territorial development policy appears to be catching up to this reality, which is a welcome sign. And to build development further, Kazakhstan might want to consider how the world’s most advanced countries nurture agglomeration as a driver of innovation and productivity growth (see, for instance, Canada, in Annex 6A). Breaking with the territorial development vision of the past The territorial policies of the past largely shaped Kazakhstan’s current economic geography—Soviet industrial development policies left a lasting impact, particularly because territorial development was an extension of industrial development. Although spatially equitable development was a core stated principle of centralized planning, in practice priority was given to places important for defense or for selected industries. Several towns in Kazakhstan are a product of this approach: Priozersk in Karagandy region is a former Soviet military base; Kurchatov—a monotown in East Kazakhstan—held the main Soviet nuclear testing facility; and Balhash, Karagandy, and Zhekazgan were centers of mining and metallurgy. Those cities’ economies were fully linked to state-planned supply chains of specific industries. As the planned economy withered, the monotowns faced high risk of economic collapse because their competitiveness in their dominant industry was untested without state protection, and they had few assets—or their people, skills—suitable for other industries. The weak performance of monotowns since the 1990s confirms this fact, and though some of them managed to adapt, most still struggle. As for Kazakhstan, its territorial development was shaped by such industry-specific initiatives, leaving its rural areas underdeveloped, unintegrated, and isolated. Besides the scarcity of basic infrastructure in those areas, the quality of life was hampered by restricted population mobility. A household registration system (featuring the propiska—see chapter 2) and highly controlled and restricted access to housing meant that residents of rural areas had almost no chance to seek their fortune in the cities. Policy development since independence The territorial development policy approach gradually evolved through several stages after independence in 1991: The first stage coincided with the post independence crisis, up to 2006. The country lost population and deurbanized while it formed new sovereign institutions. The first attempts to address territorial development were made, and in 1996, the first Concept of Territorial Development was introduced.124 In 2001, the Concept of Regional Development Policy was put forward.125 Those documents focused on regional economic specialization and territorial inequalities but included no substantial policy or investment mechanisms and had little impact. The second stage started in 2006, when policy turned toward embracing the spatial “lumpinessâ€? in economic development. In 2006, the Strategy for Territorial Development until 2015 became the first policy document in Kazakhstan’s history to embrace spatially uneven economic development and to set policy objectives around forming growth poles and supporting the development of economic clusters.126 The strategy laid the foundation for the first SPRD, adopted in 2011. It defined agglomerations as the key driver of development and Kazakhstan Country Economic Memorandum 90 emphasized identifying the constraints to local development and empowering local governments to address them. The program planned to mobilize private co-financing to fulfill the ambitious investment agenda but failed, due partly to the weak development of public – private partnership practices in Kazakhstan. The third stage started in 2014 and continues. The updated SPRD, greatly expanded in scope, introduced large components in housing, water infrastructure, and other utility networks. The program absorbed existing state programs, most of which were sector-specific infrastructure programs (for modernizing utility infrastructure, for improving access to water, and for developing housing). While this gave the SPRD much more financial power, it also shifted the goal from meeting targets for territorial development toward providing access to basic infrastructure and services. The most recent territorial development policy represents continued evolution. The system at its core is defined by two policy documents: the Forecast Scheme for Territorial Development and the SPRD.127 The Forecast Scheme sets the main priorities, and the SPRD—the second-level document to the Forecast Scheme—outlines the implementation mechanisms. The two documents set an ambitious agenda centered on enabling the productive growth of agglomerations. The Forecast Scheme positions the territorial development policies within the context of national plans: the Kazakhstan 2050 Development Strategy and the Kazakhstan 2025 Strategic Plan. The Forecast Scheme commits to prioritizing the development of growth poles as the drivers of the national economy and focuses on density as an enabling condition for the benefits of economic agglomeration in major cities. It also sets a goal of access to basic infrastructure and services across the country, thus meeting global best practices in territorial development identified in the 2009 World Development Report (World Bank 2009). The Forecast Scheme introduces the concept of functional urban areas, thus dissociating urban development policy from city administrative boundaries, which are often restrictive.128 It identifies four primary and 14 secondary functional urban areas as priorities and acknowledges that monotowns and small towns, as well as remote small villages, have limited development potential and so should not be prioritized for investment—a big shift from the development principles of the recent past. The SPRD describes the main challenges of territorial development and offers policy tools to address them. The program identifies the key issues of urban development in Kazakhstan—under urbanization, or the prevalence of small towns and the relatively small size of the leading cities; the low density of urban settlements; low internal mobility; and the high cost of living in metro areas, attributable partly to a virtually nonexistent rental housing market. Although it rarely articulates the policy tools clearly, the program sets the overall implementation framework. Positioning the SPRD against other policy efforts is the challenge. The following discussion covers the shortcomings hindering the ambitious territorial development plan. Struggles with linking targets to implementation mechanisms Kazakhstan’s territorial development policy efforts have yet to produce the desired outcomes. The official results of the previous round of the SPRD in 2016–20 suggest that it was one of the drivers of further agglomeration and concentration of population, but its contribution was unclear. Those results state that the program led to an increase of 1 million in the population of agglomerations and regional centers. That, of course, represents meaningful movement toward the main targets of the Forecast Scheme. Other results include reduction of wear on water and electricity infrastructure in monotowns (from 70 percent of stock needing repair to 61 percent for water, and from 75 percent to 63 percent Kazakhstan Country Economic Memorandum 91 for electricity), improved quality of life for 700,000 residents of key rural settlements, housing support for 17,500 professionals moving to rural areas, completion of 311 projects in water and sanitation (unspecified in scale and location), and construction of 50,800 kilometers of network infrastructure of various types.129 Almost all those indicators focus on infrastructure, few prioritize location, and none focuses on agglomerations. The funding allocation under the SPRD confirms that spending on spatial targeting was limited. Only 18 percent of SPRD spending in 2015–19 was on components that implied spatial targeting (Figure 5.13). That included 7 percent on agglomerations, 3 percent on monotowns, 2 percent on key rural settlements, and 4 percent on housing subsidies for people moving to villages. The rest of the program’s funding went to activities that were not explicitly spatially targeted. It aimed predominantly to provide a basic level of services across the country and to meet the service and infrastructure provision goals in which the only spatial target is to distinguish between urban and rural areas. Figure 5.13 Spending by segment of the State Program for Regional Development, 2015–19 500,000.0 450,000.0 400,000.0 350,000.0 Tenge (millions) 300,000.0 250,000.0 200,000.0 150,000.0 100,000.0 50,000.0 0.0 Agglomeration Monotowns Key rural Housing Other Goal 2: Goal 3: Water Goal 4: development settlements subsidies for Infrastructure and sanitation Housing educated development systems development professionals improvement (2016-17 only) moving to rural areas Goal 1: Territorial development (by subcomponents) Source: Based on data provided by Ministry of National Economy of Kazakhstan. Nor does regional disaggregation of spending reveal any strong geographic prioritization (Figure 5.14). Although developing the four large urban agglomerations is stated as the main priority of the program, in practice, Almaty, Shymkent, their surrounding regions, and Aktobe region received far less per capita than average. The only territories that appear prioritized for funding received are Astana and the surrounding Akmola region. Karagandy and Kyzylorda—the regions with the highest share of population living in monotowns—also received only average support per capita. The infrastructure- focused components of the SPRD, which accounted for 90 percent of funding distributed in 2012–19, were mostly spatially blind (as suggested earlier). Data on spatial allocation by SPRD subcomponents show that even the funding allocated to Goal 1— territorial development—had little spatial targeting except prioritizing the capital (Figure 5.15). Astana and its surrounding regions received more than double the average national per capita funding under Kazakhstan Country Economic Memorandum 92 that component, while the other major cities recognized as priority areas for development—Almaty and Shymkent—received less than the average. Shymkent, the second largest city in the country, with 6 percent of its population, received only 2 percent of the funding allocated though the SPRD. So, spatial targeting under the SPRD has not followed the priorities declared by the Forecast Scheme. These results show that SPRD operates much more as a vehicle for grants for infrastructure and services than as a vehicle for place-based policy. Place-based policies are defined in one useful framework as activity with a clear geographic scope and a goal of advancing local economic development (Duranton and Venables 2018). But most interventions implemented under the SPRD were not motivated by economic development goals, according to interviews with selected local government officials —the examples they shared included gas, water, and sanitation infrastructure projects, all motivated by the deficiency of infrastructure service in a particular neighborhood.130 The officials struggled to explain how those investments connected to the priorities of economic development. All this supports the hypothesis that SPRD has followed the structures and mechanisms that it inherited from the sectoral programs it absorbed in 2014. The sector-specific interventions should not be seen as wrong or bad because they elevate people’s quality of life and build human capital, a fundamental pillar of development. But they hardly meet the stated goal of supporting the economic development of growth centers. Other state programs, though they implement initiatives much closer to the definition of a place-based policy, lack clear ties to the priorities of territorial development stated in the Forecast Scheme. A good example is the development of infrastructure for the industrial zones in Shymkent, supported by the “roadmap for businessâ€? of the State Program for Business Development. Such development is based on seeing a highly populated area surrounded by an agricultural region close to the large Tashkent agglomeration across the border in Uzbekistan as having potential for developing a food processing industry, currently held back by the lack of serviced land for manufacturing—a market failure. Although the full analysis of the intervention’s effects will require assessing possible relocation and displacement effects and its success will require complementary policies in transportation infrastructure, workforce training, housing, and so on, this policy comes much closer in design and intent than most SPRD investments to meeting the criteria of a place-based policy. Equally, other state programs do not explicitly link their activities to the Forecast Scheme for Territorial Development or follow its territorial priorities. In fact, the only spatial priority stated in the State Program for Business Development is for diversifying monotown economies.131 The complex system of state programs in Kazakhstan is partly responsible for the inconsistencies between the vision of territorial development and the poor coordination of policy instruments. Each program, organized largely by sector, is managed by a separate ministry: the SPRD is managed by the Ministry of Economy, for example. The State Program for Industrial and Innovative Development (which includes several space-based policies), managed by the Ministry of Industry and Development, includes the development of six industrial clusters identified in a concept paper, originally prepared in 2013, which has a short section on agglomeration, emphasizing the development of industrial “countermagnetâ€? towns (never mentioned in the Forecast Scheme) around major agglomerations.132, 133 None of these spatially targeted policy measures is explicitly coordinated with the Forecast Scheme.134 Other sectoral programs related to spatial development include the State Program for Housing Development (a part of SPRD until 2017), as well as state programs for education and health, agriculture and agro- processing, tourism development, innovation development, business development and entrepreneurship, and digitization.135 Kazakhstan Country Economic Memorandum 93 The example of the industrial zone in Shymkent is not an outlier, and the complicated system of state programs includes multiple interventions that can qualify as place-based policy but are not directly aligned with the Forecast Scheme or SPRD. The complexity of the system also results in duplicated policy tools and conflicting targets. For example, the SPRG and the state programs for infrastructure development and for housing all include provisions for financing water and heating infrastructure in residential areas.136 And while the SPRG prioritizes developing functional urban areas around major cities and sets targets around gross value- added growth and capital investment growth in those areas, the State Program for Business Development prioritizes businesses in monotowns and rural areas and has no spatially disaggregated results indicators.137 The decision-making mechanism behind the distribution of funding under the state programs reinforces their sectoral focus and impedes designing and implementing multisectoral programs. The allocation of funding under the state programs is highly formal.138 Money is transferred to regions as ring-fenced grants that can be used only for the projects approved. Applications for financing specific projects are prepared by the local governments, then approved by the regional authorities, and then channeled to the national ministry responsible for the specific state program. The ministry selects projects based on sector-specific targets (for instance, the percentage of residences connected to wastewater treatment). Then the interministerial budget committee—the body prioritizing investment across different sectors based on resource limitations—considers applications and approves them for financing. The process is complex and not fully transparent. More important, the role of line ministries in allocating funding results in sectoral and fragmented investments. Building the right foundations for territorial development with structural reforms, including greater labor mobility Better territorial development would require not just adjustments to the SPRD but a comprehensive rethinking of the territorial development policy ecosystem. Place-based policy is not the only critical condition to produce desired outcomes. The lagging regions policy framework developed for the European Union (World Bank 2017) offers an overall approach that applies to spatially targeted policy in general (Figure 5.16). It calls for place-neutral interventions as well as place-based ones. Without favorable structural, macroeconomic, and fiscal conditions, alongside a favorable business environment, even well-designed support for a specific sector in a specific region is unlikely to contribute to overall improvement. National structural policies may not intuitively seem the starting point for addressing lagging regions or underperforming cities. But although the task is not as simple as getting the macro environment right, with some cities and regions carving out success in an unfavorable national environment, addressing structural distortions is a necessary, if not sufficient, condition to unlock the potential of regions and cities. Kazakhstan Country Economic Memorandum 94 Figure 5.16 Lagging regions policy framework Source: World Bank 2017. National structural conditions influence territorial development by shaping the overall business environment and enabling the optimal spatial allocation of factors of production. They shape the potential of cities and regions because they are critical to establishing an environment conducive to investment, productivity growth, and active participation in the labor force. Further, structural and regulatory factors determine whether market forces can drive the spatial allocation of factors of production to their most effective uses—the foundation of making the most of the economic potential of different locations and thus maximizing results for the national economy. If labor and capital mobility are heavily restricted within the country; if foreign capital and labor face substantial barriers to entry; or if access to domestic and foreign markets is restricted due to regulatory burdens, poor transportation infrastructure, and services, the allocation of factors of production will be suboptimal. Some places will fail to reach their potential, and labor and capital will be used inefficiently. Kazakhstan has made impressive progress in improving the business environment. It has risen to 25th globally on the Doing Business rankings and has moved forward in several areas, such as improving investment attractiveness by establishing a one-stop shop to help investors obtain information, get permits and licenses, and start administrative procedures. Still, critical challenges remain. Key structural issues covered in other chapters range from difficulties accessing finance (particularly for small and medium enterprises), to export–import regulations, to access to skilled labor. Each challenge affects the growth potential of certain industries and cities. The clearest examples are continuing state control of a few sectors—for example, through legislative restrictions on foreign ownership of mass media and restrictions on foreign investment in forest and agricultural land. The employment of foreign staff is another area with substantial restrictions, about which foreign investors have complained for most of the past decade. Kazakhstan’s visa policy thus has unnecessarily impeded investment (OECD 2018b). And domestically, rural-to-urban migration is weak (seen in the stalling urbanization level in Figure 5.2 earlier), despite growth in population mobility in recent years. With people living where they are less productive than they could be and cities smaller than optimal, Kazakhstan needs to catch up to optimal population distribution by letting people move. (Even so, cities offer better opportunities, with a poverty rate of 6 percent versus 25 percent in rural areas, better access to services and infrastructure, and a better quality of life generally; Seitz 2008). Several factors restrict internal migration. One is the rural–urban difference in the cost of living. Living in Almaty is 190 percent more expensive than the country average (Seitz 2008). With that difference, many households simply cannot afford to move. Housing affordability is the main contributor, Kazakhstan Country Economic Memorandum 95 reflecting the rapid growth and volatility of housing prices over the past 20 years. Insufficient supply of housing in the most desirable places, particularly big cities, is due to some extent to local policy. But other matters can be tackled at a national level through supply- and demand-side interventions, regulatory actions, and incentives to develop the almost nonexistent rental housing market. Another barrier is the partially retained Soviet residency registration system (discussed in detail in chapter 2). One example: people who move to a new place in Kazakhstan must register at the place of residence to access social benefits and services. Creating favorable structural conditions and, most of all, enabling labor mobility should be recognized as fundamental to territorial development efforts in Kazakhstan. Business environment limitations— such as restricted access to finance and barriers to foreign investment —hurt the economic potential of places. But barriers to the internal mobility of labor can render the best-designed place-based policy irrelevant. No matter how great an opportunity a place offers, if people cannot move there, for financial or bureaucratic reasons, it will not succeed. Addressing labor mobility constraints should be made top priority if Kazakhstan wants to see its cities emerge as engines of growth and prosperity. Empowering subnational governments is crucial for multiple aspects of territorial development Strong subnational governments are critical for successful territorial development—a strength that Kazakhstan does not have. Most subnational authorities lack the capacity and resources to design and implement strategic programs for local economic development. The system curbs the potential contribution of territorial development to economic growth in two ways. First, it leads to inefficient prioritization of investments, as local governments follow the guidelines and incentives of state programs—the resulting decisions are often disconnected from local realities and uninformed by any central understanding of local needs. Second, the failure to build subnational governments’ capacity limits the effectiveness of policies aiming to develop specific territories—whether large cities and agglomerations or monotowns. A common theme in international best practices in territorial development is that although the national government’s role in setting national goals and priorities is important, the success of cities and regions usually requires local governments to have enough powers to act. The decentralization of powers and public finance, seen as a tool to overcome national government failures in implementing territorial development policy, has been a global trend in recent decades (Pike et al. 2012). But it is not guaranteed to benefit local economic development. Harm can arise when fragmented local government territories produce diseconomies of scale in making policy and providing services (Rodríguez-Pose and Bwire 2004). Global evidence suggests that the critical criteria for decentralizing authority and public resources is the quality of institutions and the collaboration between jurisdictions. The best examples of territorial development policies often combine decentralization of power with incentives and capacity building (Toboso and Scorsone 2010). Some countries have achieved progress in local economic development by empowering subnational authorities to develop and implement local strategies and providing them support and guidance. Examples from Australia, Canada, and the European Union demonstrate the critical importance of subnational government leadership in defining local development priorities. The European Union’s cohesion policy delegates responsibility for local development strategies to regions. Australia and Canada both have regional development agencies charged with identifying needs and market failures in each region and designing and carrying out policy responses (World Bank 2020). These countries’ Kazakhstan Country Economic Memorandum 96 experiences make a strong case for rethinking the role of subnational governments in territorial development in Kazakhstan. Decentralizing powers The framework for territorial development in Kazakhstan today is almost entirely top-down. The regions identify projects, apply for funding, and implement investments, but their actions are almost entirely predefined by the restrictive parameters and complex procedures of state programs. And while their own sources of revenue are limited for state programs, they remain the main source of funding for local capital investments and development initiatives. Most state programs limit funding to narrowly defined, ring-fenced projects pursuing sector-specific objectives that local governments must address with their project proposals. The national government is taking the first steps toward decentralizing powers, which was among the national priorities of Strategy 2020. Earlier legislative amendments introduced in 2012–17 expanded the financial independence of lower levels of governments, enhanced the accountability of local leadership by introducing indirect elections of municipal akims (mayors), and granted them more control over local budgets (OECD 2020). More recently, legislation assigned local corporate income tax from small and medium enterprises to local budgets.139 But subnational government powers and capacity remain restricted, partly because national authorities and higher-level akims can reverse decisions of lower-level akims, which encourages top-down governance. Finally, accountability to citizens is low because few local leaders are directly elected (although since 2018 direct elections for akims have been piloted in rural areas). Elements of the decentralization of authority in Kazakhstan’s current territorial development policy need to go further. The current SPRD mandates the governments of large cities to develop strategies and define key priorities for economic development but offers only brief guidelines for developing these strategies. Most large cities appear to have adopted strategies following the requirements of the program, a step in the right direction. But the quality varies. In practice, the implementation of plans is often derailed by incentives dictated by the state programs. This highlights the weakness of local strategies that often are not linked to budget planning or, thus, to implementation. The way forward should further decentralize the authority to plan local development and entrust resources to the financial capacity of subnational governments, either by providing non-ring-fenced development grants or by delegating more sources of budget revenue. Such decentralization should go hand in hand with building the capacity of subnational governments (below), requiring a combination of technical support and incentives. The proposed new Law of Agglomeration and amendments to the Budget Code can offer a new start for subnational governments of cities and surrounding districts to take the initiative to coordinate and solve their development issues. Nevertheless, technical support is likely needed for those subnational governments to design and deliver programs. International examples Providing detailed guidance and handling on support for subnational strategy development through a dedicated national institution could enhance the quality of subnational strategic planning. The US Economic Development Administration (EDA), established in 1965, today provides an e-tool called the Comprehensive Economic Development Strategy to support subnational governments in building strategy-driven plans for regional economic development. It helps regional governments build an evidence-based foundation for their strategic documents. The tool also promotes participatory Kazakhstan Country Economic Memorandum 97 planning processes and the engagement of community leaders and the private sector (World Bank 2020). Replicating the model in Kazakhstan might strengthen subnational development strategies. Nations all over the world have successfully used policy tools linking the delegation of power and financial resources to subnational authorities to various conditions. One possibility is asymmetric decentralization—granting powers only to governments that meet certain criteria. Another is competitions—funding only well-prepared local development projects from the central budget, as seen, for example, in: City Deals in the United Kingdom rewarded cities such as Manchester and municipalities in its agglomeration that were more advanced in building institutions for collaborative metropolitan governance (Annex 6B; OECD 2015). Colombia used a similar approach based on asymmetric decentralization of responsibilities and financial resources. Each territory could apply for the decentralization of a certain task, such as additional oversight of industrial or social infrastructure operation and investment. Applications were evaluated based on the level of development in the relevant area in the region and the characteristics of specific regional institutions related to the task considered in the application. The model has proven popular: in 2017, the government received 11 applications for decentralization, three of which were approved (World Bank 2020). The Russian Federation has introduced competitions for the best project proposals in a selected category. The prize for the best projects is funding for implementation (Box 5.1). Box 5.1 Local government competition experience in the Russian Federation Starting in 2015–17, Russian territorial development management has used open competitions to optimize funding allocation and incentivize capacity building: • The Ministry of Construction of the Russian Federation holds a countrywide competition for the best projects for public space improvements in small towns and historical settlements. The competition calls for integrated concepts to develop urbans spaces. Up to 400 municipalities participate annually, and 80 winners are selected to receive funding to implement improvement projects. • The Ministry of Economic Development of the Moscow Region holds a “Growth Pointsâ€? competition, with all municipalities in the region participating. The purpose is to prepare comprehensive town development programs, with winners receiving funding to develop their concepts into full-fledged strategies. This experience points to two possible lessons: competitions allow the identification of local governments with the highest capacity and potential and with direct funding to support them, and participation in competitions is itself a valuable education for local authorities. Moving beyond a top-down approach and supporting integrated, multisectoral local economic development programs implemented by subnational authorities Successful investment in infrastructure depends on the ability to operate and maintain it properly, as improved facilities will do little without improved capacity of service providers. Business incentives need to be administered and training implemented properly to achieve results. But the current Forecast Scheme and SPRD neither recognize that nor include provisions or tools for capacity building. Reviewing Kazakhstan’s metropolitan areas and monotowns, the following also brings in several Kazakhstan Country Economic Memorandum 98 international examples of best practices in implementing policies like those of the SPRD and highlights the critical role of subnational institutions in these practices’ success. Metropolitan areas The SPRD aims to develop metropolitan areas and boost agglomeration economies to drive economic growth, but most of its policies focus on infrastructure and service improvements, which on their own are probably insufficient. Its policies for developing agglomerations include enhancing connectivity, improving public spaces, incentivizing the private sector to invest in housing, promoting social opportunities and utility infrastructure at the peripheries of large cities, and launching joint education programs with leading international universities. Yet, these wide-ranging ambitions lack the critical component of building the capacity of local institutions. An SPRD priority is to increase connectivity between the agglomeration’s center and its peripheries. According to the Forecast Scheme, transport infrastructure increases inequalities in access to social, employment, and infrastructure opportunities, thus limiting agglomeration economies and lowering the quality of life. Although the SPRD sets ambitious targets for enhancing connectivity within functional urban areas and increasing the share of trips in them using public transport, the specific activities it outlines are limited to transport strategies and infrastructure investment. The experiences of Moscow and Paris (Annex 6C) suggest that infrastructure investment alone is not enough to enhance connectivity in a metropolitan area. In recent decades, Moscow and Paris ambitiously revamped their transportation systems. While both cities invested heavily in enhancing infrastructure, they also built appropriate institutions. This implied both strategically coordinating local governments in the agglomeration and designing an institutional structure that could improve infrastructure and manage transportation and affiliated services. Both transport systems found success by allocating responsibility for managing a complicated transportation system to a dedicated entity, rather than sharing it across multiple government entities. Another component of success appears to be the deliberate focus on the quality of services through clear parameters in the provider contract and a monitoring and enforcement system. In short, international best practices of agglomeration development suggest that institutions and governance are generally more important for economic development of agglomerations than infrastructure investment alone. To incorporate these best practice lessons, Kazakhstan should broaden the scope of state programs targeting territorial development. For agglomerations to thrive, institutional coordination is critical. Global evidence shows that fragmentation deters economic development in metropolitan areas (Bartolini 2015). The Forecast Scheme for Territorial Development, recognizing this, mandated cross- jurisdictional spatial development plans for large agglomerations—a step in the right direction. In 2018, such plans were adopted for Shymkent and Almaty. But it is unclear whether the coordinated system of governance needed to implement those plans exists in the metropolitan areas, and the state program does not offer specific steps to resolve the issue. The only proven way to coordinate management of an agglomeration is to expand the administrative boundaries of the city—most recently done in Shymkent. But expanding boundaries also requires integrating the new territories. Interviews with government officials of cities at the center of agglomerations have revealed that the expansion areas are often seen as burdens and a punishment on the city government. Kazakhstan Country Economic Memorandum 99 Monotown development Monotowns are a legacy of Soviet planning that remain a major issue for former Soviet republics. These towns were formed around a single industrial facility, which made them extremely vulnerable in the shift to a free market by exposing them to industry-specific shocks. But these towns (some with more than 100,000 people) represent large concentrations of capital, infrastructure, and population, so even if their economic potential is limited, a strong incentive remains for the government to invest in them. Developing monotowns is a declared priority of Kazakhstan development policy. The policies proposed for them in SPRD 2025 focus predominantly on infrastructure and service investments in monotowns while retaining the goal of diversifying the economy. In addition to SPRD 2025, complementary policies to develop single-industry towns are included in the program for development of productive employment and mass entrepreneurship for 2017–20, known as Enbek, and the state program for supporting business development, Business Roadmap 2025. The monotown situation in Kazakhstan is like that in the Russian Federation. Both saw monotowns emerge during the industrial development of the Soviet Union, particularly during the second half of the 20th century. Their development was motivated largely by attempts to position manufacturing factories close to major natural resources. So, monotowns were established far from regional centers in areas with unfavorable climatic conditions and often poor transport connectivity. Kazakhstan now has 27 monotowns. Russia has 319, of which 99 fall into the category of “monotowns with the greatest socioeconomic difficulty.â€? The results of the Russian Federation’s longstanding effort to develop monotowns show that their economic potential is limited and that local governments are critical to any form of success. Despite different approaches to developing monotowns and seeing some successes, Russia failed to change the fortunes of most of them (Box 5.2). Box 5.2 Lessons from monotown development programs in the Russian Federation The crisis of monotowns has been long recognized in the Russian Federation and reads like a litany. At the beginning of the 21st century, monotowns generated about 40 percent of Russia’s GDP; today that share is 15–17 percent. At the turn of the century, 25 percent of the country’s population lived in monotowns; today 15 percent do. About a third of monotowns are in economic crisis, due to changes in market conditions and to the exclusion of town-forming enterprises (the enterprises that dominate the economy of the monotown and employ most of its labor force) from supply chains. The quality of life is worse in monotowns than in larger cities. This, combined with a lack of jobs, caused a population exodus in 2008–09: the working-age population in monotowns fell by 365,600 people in just the two years. The economic decline in monotowns was recognized during the 2008–09 global financial crisis, when residents of Pikalevo blocked a federal highway to protest the interruptions of raw material supply, which halted operations at the Pikalevo Aluminum Plant. In 2010, the Russian government responded to the challenges of monotowns by developing a comprehensive investment plan focused on infrastructure and support for the private sector. In 2016, the Priority Program for Integrated Development of Monotowns was launched, and a nongovernmental organization—the Foundation for the Development of Monotowns—was created. The foundation’s program included infrastructure investments and measures to improve the business environment and support private initiatives. The biggest shift was a new focus on building the capacity of the monotown governments. By 2017, teams responsible for integrated development were Kazakhstan Country Economic Memorandum 100 established in all monotowns and received training and capacity-building support through the program. The Monotown Development Foundation reports that its programs have significantly helped the monotowns. In 2019, the fund signed seven agreements for co-financing and constructing engineering infrastructure. This activity created 5,963 permanent or temporary jobs, launched 19 infrastructure facilities in 10 monotowns, attracted more than $450 million in investment, and led to 14 agreements to issue loans to companies currently or potentially located in those monotowns. Even so, major issues have not been overcome. Depopulation continues, and infrastructure investments financed by the state program had little impact on private activity. Overall, the program’s contribution to the metrics of local economic development were insignificant. For example, most of the jobs created were temporary. Only two monotowns emerged from challenging socioeconomic situations during the program, and none was able to diversify its economy to graduate from being a mono-town. Still, the Russian experience of monotown development offers the following hopeful perspectives: • Government capacity-building activities seem critical for monotown development. They produced better strategic policy in several monotowns. • The biggest successes emerged when new local comparative advantages were unlocked. In the Kemerovo region, the program found that some monotowns had a chance to create local ski resorts, as Sheregesh had done, increasing tourist flows in the region to 2 million a year. • Special economic zones are ineffective in monotowns. According to a report from the Accounts Chamber of the Russian Federation (2018), although free economic zones attracted substantial investment and created many jobs in 2014–19, they went mainly to four zones: Alabuga, Lipetsk, St. Petersburg, and the Technopolis in Moscow. All four were prosperous before the zones were established, and most of the growth in them could be attributed to the relocation of firms from other locations that didn’t offer subsidies. But the zones inside monotowns saw little investment or job creation. The overarching conclusion is that few monotowns can become economically vibrant again. Population outflow to large cities is a constant, so policy should focus on human development—providing better health care, education, and quality of life—even if the people who benefit from those services move to a part of the country offering better economic opportunities. But where dominant businesses are healthy, investment in an environment for economic diversification should be made right away. For Kazakhstan, the main lesson is that ambitions to turn monotowns into growth centers should be reserved for exceptional cases of locations where competitive advantages are maintained or new comparative advantages (such as tourism) are discovered. Priority should be given to investing in human capital and quality of life. Another critical lesson is that the rare success stories are usually associated with proactive and well-organized local governments and that incentives (for instance, competitions) tend to contribute to stronger outcomes, even for policies related to enhancing the quality of life. Although adding a capacity-building component to monotown development initiatives did not solve the economic challenges, it resulted in more efficient use of support funding and better social outcomes of support programs. The current SPRD acknowledges the priority of basic infrastructure and service investments in monotowns, but it should pay more attention to capacity building. SPRD 2025 acknowledges the Kazakhstan Country Economic Memorandum 101 limited effectiveness of the previously favored anchor projects—large-scale subsidies to place new industrial enterprises in monotowns. The anchor projects were focused mostly on extractive industries rather than new and innovative sectors, were ineffective, and failed to produce the anticipated economic benefits. Thus SPRD 2025 recommends deprioritizing anchor projects in monotowns, even though it does not drop them altogether. The state program still includes measures for using state- owned enterprise contracts to diversify monotown economies, but our analysis suggests such measures rarely have an effect. In the experience of Temirtau, 90 percent of state program funding received in 2015–19 was for utility and education infrastructure, and the projects supported by the SPRD were for improving the wastewater treatment system. Such investments should not be challenged, however, as they focus on improving the quality of life and building human capital.140 Given the Russian experience, even in scaling down support to monotowns, it is advisable to add capacity-building measures and incentives for local governments to increase their efficiency in providing services and livability interventions. The national policy should formally drop economic development targets from monotown development programs and reserve business support in them for local small businesses. Most of the targets that the current SPRD sets relate to infrastructure for monotowns, which is consistent with the focus on livability. Ensuring decent access to health care and education, which are predominantly covered by other state programs, is also important. But current policy still includes targets for attracting private investments to large monotowns, which seems unrealistic, given the experience presented here. The natural evolution of monotown development policy is toward directly acknowledging the focus on quality of life and access to services. While some targeted investments could be directed to unlock new economic opportunities in monotowns—such as the ski resorts in the Russian Federation’s Kemerovo region—those are exceptions, not the rule, and development programs should not be structured around such expectations. The government should also consider offering relocation grants and other forms of support for people aiming to move from monotowns to locations that offer better economic opportunities. Areas and recommendations for improving territorial development policy The government’s territorial development policy has evolved and improved but needs to further adjust its goals. This chapter’s analysis suggests that Kazakhstan is struggling to create a policy framework where favorable structural conditions empower subnational governments to implement targeted multisectoral programs that address local economic development challenges. Three key directions are identified: • Focus on structural reforms that enable production factors to move where they can be most efficiently used—initially, removing barriers to internal migration. If market forces are not allocating labor and capital to their most efficient applications, place-based economic development is doomed to fail. While Kazakhstan’s business environment has improved in recent years, territorial development would benefit from further easing of constraints on foreign investment (including industry-specific restrictions) and on immigration—most critically, internal migration. The persistent internal migration barriers stem from unaffordable housing and the restrictive household registration system. Fully abolishing such registration, which has no equivalent in the developed world, is the first step. Housing affordability should be addressed through supply- and demand-side policies to shift the housing market toward a lower price equilibrium. And policy should create the environment for a rental housing market Kazakhstan Country Economic Memorandum 102 and should expand social housing programs. Relocation grants should be offered to people who want to move to a place with greater opportunities but cannot afford to relocate. • Adopt a systematic strategy of empowering and building the capacity of subnational governments. Augmenting the capacity of local governments and their resources is common to all the best practices of territorial and local development. Empowering local governments is integral to moving from a top-down system and to improving approaches to the specific challenges of local development in agglomerations and monotowns. Kazakhstan is taking steps in the right direction by decentralizing some revenue streams, increasing the accountability of local leaders to residents, and requesting that local strategies be developed in major cities. But more is needed to make local governments into capable leaders of economic development agendas. Capacity building for local authorities, critical to any subnational policy, should be mainstreamed and made integral to territorial development. Current strategies for developing priority agglomerations vary in quality. A more systematic approach to building the integrated planning proficiency of local governments could create detailed guidance and develop interactive tools to give subnational governments analytical foundations for strategic planning. Kazakhstan could establish a national institution devoted to enhancing local planning capacity by building such tools, providing technical support, and ensuring the high quality of local plans and strategies.141 It should also consider tools such as asymmetric decentralization and project competitions as incentives for subnational capacity building. • Reposition the SPRD as a Regional and Urban Development Fund supporting integrated, multisectoral local economic development programs to be implemented by subnational authorities. Today, though envisaged as a local integrated development tool, the SPRD functions as a mix of sector-specific subprograms that did not fit into other state programs, so that its role and impact are diminished. Perhaps the SPRD can become a single program that supports complex multisectoral projects in opening new local economic development opportunities? Such complex projects are probably expensive, however, and not many could be supported, so the SPRD would need a transparent competitive process to allocate funding. The competition could include criteria for spatial targeting and for prioritizing selected industrial sectors. The guidelines for project assessment should be made clear, and we recommend that the Duranton and Venables (2018) framework be used as guidance for developing and evaluating place-based project proposals. The projects selected should clear the high bar of proving that proposed investments and incentives address market failures and will not simply relocate some existing economic activity. References Accounts Chamber of the Russian Federation. 2018. Key Conclusions of the Report of Results of Joint Control Activity “Audit of results of support of monotowns in Kirov region as a part of the implementation of the priority program Comprehensive Development of Monotowns in 2016- 2017,â€? Moscow (in Russian). https://ach.gov.ru/upload/iblock/3e4/3e48e4c1ab0289d3781d7791f9c8ef0f.pdf. Bartolini, D. 2015. “Municipal Fragmentation and Economic Performance of OECD TL2 Regions.â€? Regional Development Working Papers 2015/02, Organisation for Economic Co-operation and Development, Paris. Kazakhstan Country Economic Memorandum 103 Böhme, Kai, Philippe Doucet, Tomasz Komornicki, Jacek Zaucha, and Dariusz ÅšwiÄ…tek. 2011. “How to Strengthen the Territorial Dimension of ‘Europe 2020’ and the EU Cohesion Policy.â€? Polish Presidency of the Council of the European Union, Warsaw. Breau, Sébastien, and Richard Saillant. 2016. “Regional Income Disparities in Canada: Exploring the Geographical Dimensions of an Old Debate.â€? Regional Studies, Regional Science 3 (1): 463–81. Deas, Iain. 2014. “The Search for Territorial Fixes in Subnational Governance: City -Regions and the Disputed Emergence of Post-Political Consensus in Manchester, England.â€? Urban Studies 51 (11): 2285–314. Dijkstra, Lewis, Aneta Florczyk, Sergio Freire, Thomas Kemper, and Martino Pesaresi. 2018. “Applying the Degree of Urbanization to the Globe: A New Harmonized Definition Reveals a Different Picture of Global Urbanization.â€? Paper prepared for the 16th Conference of the International Association of Official Statistics, September 19-21, Paris. Duranton, Gilles, and Anthony J. Venables. 2018. “Place-Based Policies for Development.â€? Policy Research Working Paper 8410, World Bank Group, Washington, DC. Glaeser, Edward L., and Resseger, Matthew G. 2009. “The Complementarity between Cities and Skills.â€? Working Paper 15103, National Bureau of Economic Research, Cambridge, MA. Greater Manchester Combined Authority. 2020a. “Investment.â€? Manchester, UK. https://www.greatermanchester-ca.gov.uk/what-we-do/investment/. Greater Manchester Combined Authority. 2020b. “Greater Manchester Local Enterprise Partnership: European Structural and Investment Funds: Investment Plan 2014-2020.â€? Manchester, UK. https://www.greatermanchester- ca.gov.uk/media/1334/gm_eu_investment_plan_2014_2020_feb_2016.pdf. Heddebaut, Odile. 2017. “The Evolution of Public Transport Contracts in France.â€? Discussi on Paper 2017-11, International Transport Forum, Paris. International Transport Forum. 2020. Reforming Public Transport Planning and Delivery. Paris: International Transport Forum. https://www.oecd-ilibrary.org/transport/reforming-public- transport-planning-and-delivery_6c2f1869-en. Li, Xuecao, Yuyu Zhou, Min Zhao, and Xia Zhao. 2020. “A Harmonized Global Nighttime Light Dataset 1992–2018.â€? Scientific Data 7: 168. Lupton, Ruth, Ceri Hughes, Alex Macdougall, Hannah Goldwyn-Simpkins, and Annika Hjelmskog. 2019. Inclusive Growth in Greater Manchester 2020 and Beyond. University of Manchester, Manchester, UK. http://documents.manchester.ac.uk/display.aspx?DocID=46396. OECD (Organisation for Economic Co-operation and Development). 2014. Reviews of National Policies for Education: Secondary Education in Kazakhstan. Paris: OECD. https://dx.doi.org/10.1787/9789264205208-en. OECD (Organisation for Economic Co-operation and Development). 2015. Local Economic Leadership. Paris: OECD. Kazakhstan Country Economic Memorandum 104 OECD (Organisation for Economic Co-operation and Development). 2018a. OECD Reviews of Health Systems: Kazakhstan 2018. Paris: OECD. https://dx.doi.org/10.1787/9789264289062-en. OECD (Organisation for Economic Co-operation and Development). 2018b. Reforming Kazakhstan: Progress, Challenges and Opportunities. Paris: OECD. https://www.oecd.org/eurasia/countries/OECD-Eurasia-Reforming-Kazakhstan-EN.pdf. OECD (Organisation for Economic Co-operation and Development). 2018c. Regions and Cities at a Glance 2018. Paris: OECD. OECD (Organisation for Economic Co-operation and Development). 2020. Regional Policies in Support of Diversification and Productivity Growth in Kazakhstan. Paris: OECD. Pike, A., Andrés Rodríguez-Pose, John Tomaney, Gianpiero Torrisi, and Vassilis Tselios. 2012. “In Search of the ‘Economic Dividend’ of Devolution: Spatial Disparities, Spatial Economic Policy, and Decentralisation in the UK.â€? Environment and Planning C: Politics and Space 30 (1): 10–28. Polese, Mario. 2006. “Regional Economics in Canada.â€? The Canadian Encyclopedia. Rodríguez-Pose, Andrés, and Adala Bwire. 2004. “The Economic (In)Efficiency of Devolution.â€? Environment and Planning A: Economy and Space 36 (11): 1907–28. Savoie, Donald J. 2003. “Reviewing Canada’s Regional Development Efforts.â€? Royal Commission on Renewing and Strengthening Our Place in Canada. Seitz, William. 2018. “Urbanization in Kazakhstan: Desirable Cities, Unaffordable Housing, and the Missing Rental Market.â€? Policy Research Working Paper 8530, World Bank, Washington, DC. Toboso, Fernando, and Eric A. Scorsone. 2010. “How Much Power to Tax Do Regional Governments Enjoy in Spain since the 1996 and 2001 Reforms?â€? Regional & Federal Studies 20 (2): 157–174. UNDESA (United Nations Department of Economic and Social Affairs). 2019. World Population Prospects: The 2019 Revision. New York: UNDESA. Ward, Matthew. 2020. “City Deals.â€? Briefing Paper Number 7158, House of Commons Library, London. Williamson, Jeffrey G. 1965. “Regional Inequality and the Process of National Development: A Description of the Patterns.â€? Economic and Cultural Change 13 (4): 1–84. World Bank. 2009. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. World Bank. 2017. Rethinking Lagging Regions in the EU: Evidence-Based Principles for Future Cohesion Policy. Washington, DC: World Bank. World Bank. 2020. Territorial Development in Argentina: Using Differentiated Policies to Reduce Disparities and Spur Economic Growth. Washington, DC: World Bank. Kazakhstan Country Economic Memorandum 105 Annex 5A Territorial development in Canada In Canada—another unevenly developed, natural–resource-dependent country, with large territorial disparities and a historical economic dependence on exports of natural resources—territorial development has largely achieved the federal and provincial governments’ desired goals. But the effort has required multiple course corrections. Since the 1950s, regional disparities have been the country’s main issue of territorial development, which was historically dictated by strong exogenous factors. Tough climatic conditions led to concentration of population along the southern border with the United States. Natural resources— from fish to fur, wood, and minerals, and to oil and gas—and access to trade routes and markets (across the Great Lakes to the industrial regions of the midwestern United States, across the Atlantic to Europe, and across the Pacific to East Asia) favored some regions over others. Until the 1970s and 1980s, Canada used interventionist measures and market regulations to promote regional development. Protective tariffs benefited the manufacturing areas of Ontario and Quebec provinces. The western provinces and the Atlantic region were effectively subsidizing the industrialization of the heartland. And for most of the 20th century, subsidized transportation of (mostly) agricultural goods increased the ability of Canada’s western regions to compete with imports through better access to the Ontario and Quebec markets. Finally, the so-called Borden Line (1961) for a while limited oil imports into provinces west of the Quebec–Ontario border, thus reserving the largest part of the local market for historically more expensive oil from western Canada (Polese 2006). These policies introduced multiple distortions to how market forces shaped territorial development, likely reducing the growth potential of the national economy. The first wave of policies targeting regional disparities focused on fiscal equalization, aiming to achieve balanced public service provision across regions. From 1960, that agenda was broadened to include measures to spruce up the economies of lagging regions, through tax incentives and direct support for agricultural producers. That trend culminated in 1967 in the establishment of the Department of Regional Economic Expansion (DREE), which focused on promoting growth poles in lagging regions through concentrated upgrading of infrastructure and services, and complementary spatially targeted business incentives. But this policy approach and DREE itself were discontinued in 1974 after running into political difficulties.142 Experiments in decentralizing regional development responsibilities defined the 1970s and 1980s. In 1972, the introduction of General Development Agreements with provinces signaled a push toward decentralization. Ten years later, in a policy U-turn, a more centralized approach with top-down spatial targeting was introduced by the new Department for Regional Industrial Expansion in response to a perceived change in regional development dynamics after the discovery of Atlantic oil and a crisis in key industries in Ontario and Quebec. In 1987, another decisive turn established Regional Development Agencies, subnational bodies with broad authority to design and run development programs in their regions. Although Canadians disagree about the effectiveness of 70 years of regional development policy, outside observation can point to their success in reducing regional disparities and supporting the adjustment of regions to exogenous shocks. Provincial governments gradually grew less dependent on federal funding. Regional variation in per capita incomes has narrowed (Figure 5A.1). And cities (including urban areas in traditionally slow-growth regions such as St. John’s, Halifax, and Moncton) emerged as economic drivers, almost matching the average share of economic growth in cities of Organisation for Economic Co-operation and Development countries (Figure 5A.2). It is hard to Kazakhstan Country Economic Memorandum 106 attribute any of these outcomes to specific policies, but they suggest that territorial development efforts may have had a more positive impact than generally inferred (Savoie 2003). Canada’s experience shows achievement without measures that interfere with market mechanisms, such as limits on factor mobility or access to markets. The more effective policy model for Canadian territorial development is based on removing access regulations and allowing market forces to drive the spatial allocation of factors of production, while decentralizing the design and implementation of place-based policies, to such an extent that, in 2016, regional governments accounted for 87.5 percent of government spending in Canada. Figure 5A.1 Canadian regional economic development disparities, 1981–2010 Source: Breau and Saillant 2016. Figure 5A.2 Contributions of cities with more than 500,000 people to economic growth in Canada and the Organisation for Economic Co-operation and Development average Source: OECD 2018c. Kazakhstan Country Economic Memorandum 107 Annex 5B The Greater Manchester Combined Authority: Coordinated metropolitan management The case of the Greater Manchester agglomeration suggests that metropolitan governance is at the core of advancing an agglomeration’s development. The Greater Manchester model shows how proactive leadership, consensus building, and a wide network of both public and private partners can transform a lagging postindustrial region into one of the most successful regions nationwide. From 2009 to 2014, the Greater Manchester economy grew by 15.2 percent, becoming the fourth largest metropolitan economy in the United Kingdom (Lupton et al. 2019). The Greater Manchester collaborative governance model evolved gradually. The first step was the Association of Greater Manchester Authorities between 1986 and 2011. Under that arrangement, representatives of 10 municipalities set the joint goals for agglomeration development on a collaborative, voluntary basis. They gave priority to transport, regeneration, and economic growth. But there was no real policy implementation capacity or mechanism. As early as 1996, representatives of the 10 municipalities met to discuss possible drivers of economic growth in the region and decided to combine their efforts on developing tourism, the service economy, knowledge-led industries, and connectivity enhancements (Deas 2014). In 2011, the Greater Manchester Combined Authority (GMCA) was formed. GMCA is the executive metropolitan authority represented by leaders of 10 municipalities that oversee agglomeration development. Each member of the executive board supervises a particular strategic direction—health, culture, employment, community, security, economic development, sustainable development, and so on—and chairs the respective sectoral committee. In 2017, the first election of a metropolitan area mayor of Greater Manchester was held. The mayor now chairs the executive board of GMCA. The mayor is the 11th member of the board and is directly responsible for transport strategy, police and fire services, economic development, spatial development of the region, and a new £300 million housing provision project. The gradual development of the partnership of municipalities was helped along by national incentives. In 2011, the UK government announced City Deals—a program that gave the largest cities an opportunity to access additional powers and revenue streams in exchange for developing economic development plans and showing a record of strong governance. The criteria for approving city deals emphasized records of cross-jurisdictional collaboration and engagement with the private sector. The Greater Manchester city deal, signed in July 2012, was one of the most ambitious. It included establishing a revolving infrastructure fund allowing Greater Manchester to “earn backâ€? a portion of additional tax revenue resulting from infrastructure investment (Ward 2020). The deal also included housing, apprenticeship, and skill development initiatives. The revolving fund alone allowed for the return and local reinvestment of about £30 million from taxes paid to the national budget in the region (OECD 2015). One of GMCA’s key features is strong private sector engagement via local enterprise partnerships (LEPs). LEPs are voluntary partnerships between local authorities and businesses, set up in 2011 by the Department for Business, Innovation, and Skills to help determine local economic priorities and lead economic growth and job creation within the local area. In the Greater Manchester area, the LEP enables business community representatives to have a stronger voice in discussing strategic long-term goals for the joint economic development of the 10 municipalities. While the Manchester partnership between the business community and the association of local governments predates the formation of Kazakhstan Country Economic Memorandum 108 the city deals and the LEPs, these structures and incentives provided by the national government resulted in closer, more efficient collaboration. Joint work of the 10 municipalities has resulted in major initiatives, many of them focusing on infrastructure development. Overall, the coordinated investment program of municipalities has resulted in more than £1 billion invested in enhancing transport connectivity, in particular the Metrolink light rail system in 2009. This was enabled largely by the creation of the Greater Manchester Investment Fund, an umbrella organization that coordinates several investment funds in the metropolitan area. The redistribution of resources (national investment, EU investment (available at that time), and local public and private investment) is carried out through an investment framework that prioritizes and allocates funding across 10 municipalities, first developed as a part of the Greater Manchester city deal. Allocations are decided collegially, and the money is directed to projects that will have the most impact on employment benefits and gross value added. By 2020, about £123 million had been invested in more than 110 businesses, creating roughly 8,000 jobs and £445 million had been invested to build 6,100 new homes and redevelop 23 hectares of brownfield land, delivering 349,048 square meters of new commercial property. Input from the private sector exceeded £1 billion (Greater Manchester Combined Authority 2020a). Over the past 10 years, the GMCA, with the support of the Greater Manchester LEPs, has launched several landmark projects (Greater Manchester Combined Authority 2020b): £494.5 million in investment to provide 6,250 jobs in 2015–21, supported by the UK government under the Growth Deal. £140 million in investment for the Skills and Economy Project. £354 million for transport projects in Greater Manchester. The establishment of two enterprise zones: Airport City and Corridor Manchester. The Coronavirus Business Interruption Loan Scheme, which directed £3 million in joint financial support to local businesses to help them survive the coronavirus crisis and safeguard jobs. Kazakhstan Country Economic Memorandum 109 Annex 5C Reforming the transport system in Moscow and Paris Moscow The overhaul of Moscow’s transportation system started in 2011 to relieve the traffic issues for which the city had become internationally notorious. The historical circular structure of the city encouraged using center–periphery roads for trips that did not have the center as the origin or the destination— resulting in horrendous traffic jams. The reform started with the establishment of the Moscow Transport Node nongovernmental organization, an agency put in charge of coordinating transport investment and the operation of the transport networks of Moscow city and the Moscow region. The investment program that followed included large-scale infrastructure projects. They included the Moscow Central Circle (54 kilometers of new track, 31 new stations, 23 new metro interchange stations, and 10 rail interchange stations) and the Moscow Central Diameters that now connect multiple satellite towns to Moscow. The Moscow Central Diameters, with 132 kilometers of lines, 57 stations, and 19 subway connections, are fully integrated into the Moscow transport system. The investments in hard infrastructure were accompanied by measures to improve service quality. They included comprehensively reviewing bus routes to ensure their integration into the system and upgrading the bus fleet, enhancing the overall attractiveness of public transport with better facilities and better signage, and introducing a single ticket card, “Troika,â€? for all modes of transport within the agglomeration. The program was also coordinated with urban regeneration and public space improvement efforts.143 Over the nine years of implementation (2010–19), the program achieved impressive results. Moscow saw a 33 percent reduction in the number of trips made by car, a 15 percent increase in the number made by bus and other modes of surface public transportation, and a 20 percent reduction in the driving time from the city ring road to the city center during peak hours.144 Paris In 2000, an ambitious project of revamping the transportation system in the Paris region began. The goals were to increase connectivity within the ÃŽle-de-France region, make using public transport more convenient, and improve service quality. The first step was to establish a new coordinated transport authority for the region. The national provider Parisian Transport Syndicate was transferred to STIF (Syndicat Transport ÃŽle-de-France, renamed ÃŽle-de-France Mobilités in 2017). ÃŽle-de-France Mobilités is responsible for organizing, coordinating, and financing public transport, new transport infrastructure, and school transport services, as well as developing the region’s general transport strategy. It is financed mostly through revenues from the transport tax, advertising, fines, and public subsidies. Although the reform included large investments, it had a strong focus on setting the right organizational framework. Service contracts were concluded with three major operators: RATP, which operates the Paris Metro, trams, and buses; RER, a system of suburban lines partly operated by RATP and partly by SNCF, the national rail company; and OPTILE, a public transport organization created through a merger of several private bus companies serving the suburbs. To improve service, contracts with the operators include a mechanism to monitor and evaluate their performance, which includes a framework to levy penalties on the operators that fail to properly provide the contracted services. For Kazakhstan Country Economic Memorandum 110 the convenience of passengers, all public transport routes are serviced by one payment system (Navigo) and coordinated fare structure, operated by another private operator. The comprehensive remodeling led to measurable improvements. Between 2000 and 2014, travel by public transportation increased by 21 percent, and between 2000 and 2017, the number of passenger- kilometers shot up by 43 percent (International Transport Forum 2020). And while public investment in transport infrastructure increased every year, from €5.4 billion in 2000 to €9.2 billion in 2014, government subsidies to the operators decreased from €445 million in 2000 to €128 million in 2015, indicating greater efficiency of the system (Heddebaut 2017). Kazakhstan Country Economic Memorandum 111 6 Special economic zones, industrial zones, and territorial clusters In the late 1970s, China’s command economy was lagging its regional peers, such as Japan and the Republic of Korea.145 To address the systemic weakness but essentially remain a centrally planned economy, Chinese policymakers experimented with specially designated areas where the legal regime would differ from that in the rest of the country. The first such area, or special economic zone (SEZ), was created in Guangdong province in 1979. China’s first SEZs were all along the South China Sea coast, designed to accelerate maritime trade, attract foreign investment, and build economic links with neighboring Hong Kong SAR, China and Taiwan, China. After initial success there, in the 1990s, China’s political leaders started to create such zones in other provinces and pursue other goals, such as developing manufacturing and high-tech industries. Many countries saw China’s impressive economic growth from the 1980s and adopted its model for their economies. Globally today, there are three general sets of place-based policies to attract investment and advance local economies, broadly: SEZs, industrial zones (IZs), and territorial clusters (TCs)—“zones and clustersâ€? for short—each with its own features. A SEZ is a commercial area where business and trade laws usually differ from those in the rest of the country. It is a place where the government aims to generate investment, exports, jobs, or fiscal revenue (direct benefits) as well as create spillover technology transfers and export diversification (indirect effects) (Zeng 2011). An IZ146 is similar to a SEZ but focuses on industrial development and/or manufacturing. (A SEZ can contain an IZ but does not have to.) Freestanding IZs do not usually operate under a special legal regime. TCs, in contrast, are geographic areas with a group of interconnected companies, specialized suppliers, service providers, firms in related industries, and related institutions (educational and financial, for instance) (Porter 1998). TCs operate within the local legal system without any particular regulatory incentives and harness intersectoral links and spillovers. SEZs have gained new traction in recent years, due partly to the intense competition worldwide for foreign direct investment. The World Development Report 2020 recognizes the possibility of using SEZs to facilitate participation in global value chains (World Bank 2020). According to the latest World Investment Report, the use of SEZs as an economic development tool has grown rapidly in recent years, with at least 5,400 SEZs in nearly 150 economies (compared with 4,000 in 2015), covering almost three- fourths of developing economies and almost all transition economies (UNCTAD 2019). While Kazakhstan started experimenting with SEZ policies in the early 1990s, it established its first SEZ in Astana city in 2002. Astana–New City SEZ was designed to attract investment to the new capital. The following year, the government created other SEZs, first in Almaty city and at the Aktau Sea Port on the Caspian Sea and then Ontustik SEZ in Turkistan region and the National Industrial Petrochemical Technopark in Atyrau region. A second wave of SEZs came in the 2010s when seven more zones were created across the country. Simultaneously, the government pursued industrial development policies and created dozens of IZs, the majority in the south (in Almaty, Kyzylorda, Turkistan, and Zhambyl regions). In 2015, the government introduced the first six pilot TCs (Annex 7A) and planned to publicly finance them from 2020, but because of the COVID-19-induced economic crisis, this funding was frozen. This chapter assesses these policies. It looks first at the overall policy context of these zones and clusters, the challenges they aim to address, and their development potential, as well as their impact on local economies and their possible indirect effects. Second, it presents Kazakh stakeholders’ views Kazakhstan Country Economic Memorandum 112 on existing zone and cluster policies and on the zones themselves. Third, it briefly reviews some other countries’ experiences. Finally, it provides some policy recommendations to help policymakers maximize the benefits of these policy tools and minimize any unintended negative consequences. The zones and clusters Kazakhstan has several major state development programs for economic diversification and territorial development. The State Program for Industrial Innovative Development 2020–2024 (led by the Ministry of Industry and Infrastructure Development, or MIID), the Business Roadmap 2025 (led by the Ministry of National Economy), the Economy of Simple Things program, Digital Kazakhstan, and the Nurly Zhol (“Bright Pathâ€?) Infrastructure Development Program. The first two are umbrella programs for industry- specific development initiatives in, for example, zones and clusters. Yet these programs are fragmented, lack strategic focus, and emphasize input and output indicators over outcomes (Box 6.1). Box 6.1 A fragmented framework Kazakhstan’s entire public policy framework has over 3,200 goals, objectives, and indicators, with all the current strategic documents more than 5,000 pages long and with every industry having its dedicated state program.1 Additionally, because the state budget relies heavily on commodity export revenues, the private sector is in a nascent stage and nonresource foreign direct investment is meager, state funding of these development programs is extremely volatile, and the public sector lacks capacity to implement these strategies. There is also a practice of prioritizing welfare programs over development programs. Further, responsibilities overlap between central and local government agencies in daily zones’ operations. While QazIndustry, within MIID, is a key stakeholder in creating and setting all zone and cluster policies and can participate directly in zone management, local executive bodies (akimats) are also important in their management, funding, and operations. Note: 1. Draft Concept for Development of Public Administration in the Republic of Kazakhstan in 2020–25. The SEZs and IZs are primarily tools for economic diversification and spatial development (Table 6.1). Policies aimed at improving the business climate and at encouraging efficiency-seeking foreign direct investment focus on boosting the country’s ranking in the World Bank’s Doing Business. TCs, in contrast, are policy initiatives introduced in 2003 that aim to develop a range of industries for import substitution and increased export competitiveness. Table 6.1 Main features of Kazakhstan’s policies on its zones and clusters Feature Special economic zones Industrial zones Territorial clusters • Develop modern, highly productive, and competitive • Develop • Encourage import industries entrepreneurship substitution and increase • Provide high-quality services • Provide infrastructure export competitiveness Goals • Attract investment • Boost production • Boost productivity by • Introduce new technologies in efficiency addressing horizontal economic sectors and regions • Increase employment (strategic) issues • Increase employment • Land tax: 0% (versus 48–5,790 • Land tax: As left Tax incentives No incentives tenge per hectarea or 0.48–28.95 • Land rent: As left Kazakhstan Country Economic Memorandum 113 tenge per square meterb on the general market) • Land rent: 0% for the first 10 years of residency (versus at least 100% of land tax amount) • Annual property tax: 0% (versus 0.5% of book value for sole proprietorship, 0.5–1.5% of book value for legal entities) • Corporate income tax: exempt (versus 10–20% on the general market) • Import value-added tax on goods and raw materials: 0% for up to 15 years (versus 12% on Trade the general market) Not applicable No incentives incentives • Import duties on technological equipment, spare parts, and raw materials: 0% (versus 0–15% on the general market) Permit for hiring foreign workers: Land lease up to 49 Other simplified procedures for the years, simplified issuance No incentives incentives period of construction work and 1 of permits year after project launch Transport (rail, road), utilities Infrastructure (heating, electricity, water supply, As left No incentives provision and sewage) at rates similar to or lower than the market Management Public (with minor co-financing company Public Public by private territorial cluster (public/private) members) Business entities in Small and medium enterprises and industry, agriculture, Small and medium enterprises, Eligible large entities selected by decision tourism, transportation/ local universities, regional participants of local or central public agencies logistics, and waste development organizations management • Transport and logistics • Metallurgy • Tourism • Dairy • Metallurgy • Wholesale cross-border trade • Flour • Petroleum refining Permitted • Information and • Tourism • Machinery activities communications technology • Construction materials • Chemicals • Chemicals • Pharmaceuticals • Food processing • Textiles • Furniture manufacturing • Construction • Other manufacturing Links to • Backward/supply chain domestic Backward/supply chain Backward/supply chain • Forward economy • Horizontal • Special economic • Special economic zones law • State Program on Industrial Governing zones law • Presidential decrees and Innovative Development regulations • Regional akimat • Cabinet of Ministers decrees for 2020–25 orders Kazakhstan Country Economic Memorandum 114 • Ministry of Industry and Infrastructure Development rules Sources: QazIndustry, Law of the Republic of Kazakhstan on Special Economic Zones and Industrial Zones, No. 242-VI, 2 April 2019; Entrepreneurship Code, adopted through Law No. 375-V 3PK, dated 29, 2015; State Program on Industrial and Innovative Development for 2020 –24. a. Basic rates for land used in industrial purposes outside inhabited localities depend on the bonitet rate (land quality). b. Basic rates for land used in inhabited localities depend on the city. Special economic zones The concept of SEZs was introduced into the Kazakh legal system in the early 1990s with the Law on Free Economic Zones in Kazakh SSR of November 30, 1990. That law defined a free economic zone as “a specially designated territory with clearly defined administrative boundaries and a specia l legal regime created to attract foreign capital, foreign technology, and managerial experience for accelerated socioeconomic development of the zone.â€? That law was replaced by the Law on Special Economic Zones and Industrial Zones in 1996, and in turn was updated in 2001, 2007, 2011, and 2019. According to the most recent Kazakh law (2019), a SEZ is “a territory of the Republic of Kazakhstan with exactly marked borders, on which a special legal regime shall be organized to carry out priority activities.â€?147 Under that law, the goals of SEZ creation are to accelerate the development of modern, highly productive, and competitive industries; provide high-quality services; attract investment; introduce new technologies in economic sectors and regions; and increase employment.148 Additionally, for SEZs whose limits fully or partly coincide with the Eurasian Economic Union’s customs border, the objective is to develop cross-border trade and the economy of adjacent border territories, as well as transport infrastructure, tourism, and cultural interactions between the border territories.149 A special legal regime applies to firms operating in SEZs. Each SEZ also has a detailed list of permitted activities, whose scope reflects the government’s targeted sectors and priority areas. The special legal regime is applicable only when a SEZ participant engages in the priority activities defined for that particular SEZ (and consistent with the objectives of that SEZ’s creation).150 Ministries, akimats (state and local executive bodies), and private companies can initiate establishment of a SEZ. A proposal, with a SEZ feasibility study (or “conceptâ€?), is submitted to QazIndustry, in MIID. The proposal is examined against formal requirements defined by MIID.151 Next, the proposal is reviewed technically and economically by an expert interagency council comprising representatives from MIID, the Ministry of National Economy, the Ministry of Finance, and the National Chamber of Entrepreneurs (Atameken), for review. If the expert council approves the proposal, MIID drafts and submits to the Prime Minister’s Office a government decree establishing the SEZ. A SEZ can be closed when its “lifetimeâ€? (usually 25 years) expires, as envisaged in the presidential decree (for SEZs established before 2019) or government decree (for SEZs established from 2019). The current law does not define other mechanisms for SEZ closure. When a SEZ is created, a SEZ management company is established, with no more than 26 percent of the voting shares owned by QazIndustry (in effect giving the government veto power). The first meeting of the SEZ founders is held within 30 days of the government’s decision on its participation share. The board of directors of the management company elects two independent directors from a group of candidates recommended by the Single Coordination Center on the Development of SEZ under KAZNEX Invest JSC (Corporation for Export Development and Promotion joint stock company), Kazakhstan Country Economic Memorandum 115 a development institute under the Ministry of Industry and Trade, and the National Chamber of Entrepreneurs. The management company also develops and agrees with MIID on a three-year SEZ development strategy, along with annual target indicators. SEZs are overseen at the national or regional level, depending on their affiliation or status. The law defines the roles and responsibilities of the relevant government agencies. Nationally, MIID (QazIndustry) implements the state’s policies on SEZs, coordinating government agencies and management companies in establishing, operating, and closing SEZs (except for Astana–New City SEZ). Additionally, MIID develops laws and regulations on SEZ activities, keeps a unified register of SEZ participants (based on the data provided by the management company of each SEZ), approves the requirements for a feasibility study on establishing a SEZ, and develops sample contracts for the lease or sublease of private or state-owned land where a SEZ is created. It also sets the regulations for the aforementioned interagency expert council. The Ministry of National Economy develops the government’s budget, tax, and customs policies, while the State Revenue Committee of the Ministry of Finance implements the tax and customs policies, as related to SEZs. QazIndustry coordinates SEZ activities and is responsible for many aspects of SEZ operations. For registering participants, it interacts with public authorities, SEZ management company shareholders, and the participants themselves. It may also participate directly in SEZ management, as a shareholder, consult SEZ management on marketing strategy, or assist in attracting investors in a SEZ (for example, by holding business forums, exhibitions, conferences, and seminars). Additionally, QazIndustry is a key policymaker and develops proposals on improving SEZ legislation. It also maintains a database of all SEZ participants, conducts market and project analysis, and monitors project implementation and target indicators for each zone. The management company undertakes overall development, promotion, and monitoring of the SEZ. It interacts with state bodies and all its SEZ’s participants, provides land parcels for secondary land use, collects information on key SEZ performance indicators, analyzes ways to improve SEZ operations in terms of legal and institutional frameworks, elaborates development plans and marketing strategies for the SEZ, and confirms actual consumption of imported inputs during the activities corresponding to the SEZ’s objectives. Regional akimats implement state policy on SEZ operations and can initiate the establishment of SEZs. They fund and participate in the SEZ management companies, provide land for SEZ use, and facilitate infrastructure development. Like QazIndustry and SEZ management companies, akimats also monitor SEZ participants for compliance with the terms of the agreement and analyze data on SEZ performance. In mid-2020, Kazakhstan had 13 SEZs, 12 of which were functioning (Figure 6.1; Annex 7B). These SEZs occupied a total of 34,900 hectares of land, an average of 2,685 hectares each (Table 6.2). SEZs are usually located close to big urban centers with at least half a million people. SEZs hosted 277 residents in total (residents is used in a legal sense to mean firms registered in a SEZ or IZ), though only 231 were operating as of mid-2020. A typical SEZ resident is small or medium, averaging 55 employees. About one-fourth of SEZ residents are either fully or partly foreign owned, which shows that SEZs are focused on attracting foreign investors. About 60 percent of SEZ residents were in the Park of Innovative Technologies SEZ in Almaty City, which focuses on developing the information and communications technology (ICT) sector. This SEZ has generated 35 percent of all SEZ jobs. Astana–New City SEZ, which concentrates on local development of Astana city, generated two-thirds of all SEZ value added. Kazakhstan Country Economic Memorandum 116 Figure 6.1 Map of special economic zones, industrial zones, and territorial clusters in Kazakhstan, mid-2020 Source: QazIndustry and KazInvest. Table 6.2 Characteristics of special economic zones and industrial zones Special Characteristic economic zones Industrial zones Number of zones 13 23 Number of functioning zones 12 19 Average zone size (hectares) 2,685 155 Average population within 50 kilometers 625,576 211,297 Number of operating firms 231 134 Total number of registered firms 277 140 Share of foreign owned (%) 24 6 Average firm average (number of employees) 55 47 Source: World Bank calculations, based on QazIndustry data. Industrial zones IZs focus mainly on providing infrastructure for entrepreneurship development in regions (Table 6.1). The law defines an IZ as a territory with installed utility systems provided for private entrepreneurs by the state for creating and operating business entities, including those in tourism, industry, agriculture, transportation and logistics, and waste management.152 The lifetime of an IZ is determined by akimats but may not be less than 20 years. IZs are divided into two broad groups: public and private. The law defines three types of public IZs, depending on their funding source: IZ with republican (national) status. Establishment of republican IZs can be initiated either by regional or major city (such as Almaty, Astana, and Shymkent) akimats. An IZ feasibility study Kazakhstan Country Economic Memorandum 117 is reviewed by QazIndustry (within MIID) and an expert council. If the council approves the study, the akimat issues an executive order to establish the IZ and establishes or selects a management company for it. An IZ with republican status is fully or partly financed from the national budget. IZ with regional status. Establishment of regional IZs can be initiated by regional or major city akimats. The IZ feasibility study is reviewed by the public council (that is, the local community board) of the region or the major city. If the public council approves the study, the akimat issues an executive order to establish the IZ and establishes or selects a management company for it. An IZ with regional status is fully or partly financed from the regional/city budget. Small IZ. A small IZ can be established on the territory of an existing IZ with republican or regional status or on the territory of an industrial entity (for example, a workshop of a plant). Its size should not exceed 100 hectares. Establishment of such an IZ can be initiated by public and private legal entities. The feasibility study of a small IZ is reviewed by regional coordination councils, and the final decision is made by akimats. A small IZ is financed from the regional budget when its establishment is initiated by a regional akimat. IZs can be private when established by individuals or nonpublic entities on their own land and using their own infrastructure. In this case, regional or major city akimats check the feasibility studies of these private IZs against environmental requirements and city/region master plans. Upon concurrence of the akimat, a private IZ owner establishes a management company. Infrastructure of republican and regional IZs must be funded from budgetary or other sources not prohibited by law, whether new building or rebuilding. IZ residents can construct infrastructure on land plots they rent (and operate on) upon approval of the IZ management company. IZ infrastructure includes roads, rail (if applicable), water supply and sewage, natural gas supply (if applicable), electric power supply (transmission and distribution network), and last-mile telecommunications. An IZ can also be established at idle enterprises or existing workshops and production facilities, if that is cost efficient against building new infrastructure. With republican and regional IZs, akimats establish or competitively select a management company. IZs are usually managed by regional development institutions that also manage regional public assets. However, legislation also allows establishment of a management company in partnership with a private entity, including foreign entities. With a private IZ, a private entity that initiated establishment of the IZ may establish or select a management company, which is financed either from the state budget, replenishment of equity capital (when a company is fully owned by the state), participation fees and rents collected from IZ residents, or loans. Within two months of being established, the management company develops a three-year IZ development strategy to be approved by MIID (for republican-level IZs) or akimats (for regional-level IZs). The strategy includes annual key performance indicators for the management company. IZ management companies have the following functions: distributing (renting out) land plots and infrastructure facilities among IZ residents; reporting quarterly to MIID and KazInvest on performance of the IZ; concluding, terminating, and monitoring the implementation of participation agreements with IZ residents; providing logistics support and utility services; attracting new IZ residents; attracting investment for infrastructure construction; constructing and maintaining infrastructure, including a venue for a one-stop shop; and assuring consulting, public relations, and marketing support services for IZ residents. Kazakhstan Country Economic Memorandum 118 Kazakhstan has 23 IZs, 19 of which are functioning,153 as of mid-2020 (Table 6.2; Annex 7C.) Of these, seven have republican status and 16, regional status. The four “desolateâ€? IZs are in the suburban areas of Ust-Kamenogorsk (the administrative center of East Kazakhstan region) and in Turkistan region.154 All IZs occupied a total area of 3,253 hectares, of which only 900 were occupied by existing IZ residents and infrastructure. The average IZ, at about 155 hectares, is therefore much smaller than the average SEZ. IZs are also usually farther away from urban areas, with an average of about 200,000 people within 50 kilometers. Moreover, IZs host far fewer firms than SEZs—140 residents in total, with 134 of them operational in mid-2020—even as the average IZ firm size, at 47 employees, is similar to that in SEZs. Of the 23 IZs, 11 are in Turkistan region and on the outskirts of Shymkent city (Figure 6.1),155 six are in Kyzylorda region, and six are in Almaty city and the regions of Aktobe, Kostanay, and East Kazakhstan. All the IZs are multiactivity zones, with most residents engaged in production of construction materials, light industries (food and textiles), and agriculture. The Ontustik SEZ in Shymkent is the oldest (established in 2010) and the largest zone by number of residents (57, or 40 percent of the countrywide total) and jobs created (3,901, or 54 percent of all jobs created in IZs). The largest by allocated land is the Ondiris IZ in Kyzylorda city (760 hectares). The anchor residents of the zone are a brick factory, a ferrous alloy plant, and a planned glass factory. The largest zones by residents’ investments are the Shieli district IZ in Kyzylorda region (64 billion tenge in total, of which 60 billion tenge is a cement plant launched by a Chinese investor), Ontustik SEZ (37.4 billion tenge) and Tassay IZ on the outskirts of Shymkent (24.4 billion tenge), Almaty IZ in Almaty city (26.6 billion tenge), and Badam IZ in Turkistan region (16.1 billion tenge). These five zones account for 86 percent of total investments made by residents of the 21 IZs. Impacts of special economic zones and industrial zones on local economies An analysis of the SEZ and IZ costs and benefits reveal that SEZs primarily attract foreign investment, while IZs are more domestically oriented (Table 6.3). Moreover, IZs received far more public investment in infrastructure than SEZs while generating much less tax revenue. IZs also created far fewer jobs than SEZs but generated much more output, meaning that IZs are more productive than SEZs and revealing the underlying natures of the two zones’ residents: SEZ residents are usually export -oriented specialized small firms (many in services sectors), and IZ residents are usually industrial and manufacturing plants serving the domestic market. Yet, neither type generates enough fiscal revenue to outweigh the cost of their infrastructure. The net cost of both zone types might be even higher if other types of cost, such as environmental damage and social costs, were considered. Kazakhstan Country Economic Memorandum 119 Table 6.3 Costs and benefits of special economic zones and industrial zones Cost or benefit Special economic Industrial zones zones Costs (million tenge, all years) Public investment in infrastructure 656,007 5,264,061 Tax subsidy — — Environmental cost — — Social cost — — Benefits (million tenge, all years) Tax revenue from residents 84,766 18,174 Private investment 982,456 196,983 Production value added 2,518,869 3,855,322a Exports 1,893,928 326 Number of jobs created 15,469 6,637 Public investment to private investment ratio 0.67 26.72 Export as share of production (%) 75.19 0.01 Production to land use (million tenge per hectare) 72.17 1081.44a Jobs created as share of local labor force (%) 0.40 0.29 Public investment for one job created (million tenge) 42.41 793.14 Value added per worker (million tenge) 162.83 580.88a Fiscal revenue per resident (million tenge) 306.01 129.82 Fiscal spending per resident (million tenge) 2,368.26 37,600.44 Net fiscal benefit per resident (million tenge) −2,062.24 −37,470.62 Source: World Bank calculations, based on QazIndustry data. Note: a. Because data on output are missing for many IZ residents, these estimates should be treated as lower bounds. — = data not available. Both SEZs and IZs, however, have failed to generate local jobs, despite that being a stated policy goal. The SEZs and IZs not only have few operating firms, but the firms are also small or medium. Over the years, these zones have employed far less than 1 percent of the estimated local labor force (Table 6.3), even though SEZs are usually located near big urban centers (Figures 6.2 and 6.3). For example, the two largest SEZs by jobs are the Park of Innovative Technologies in Almaty and Astana–New City, both in major metropolitan areas. Many IZs are also located near cities but create even fewer jobs than SEZs. Additionally, industrial plants gradually employ fewer workers over time as machines replace manual labor. Kazakhstan Country Economic Memorandum 120 Figure 6.2 Special economic zone Figure 6.3 Industrial zone employment and employment and local population, mid-2020 local population, mid-2020 2,000,000 2,000,000 Population living within 50km radius Population living within 50km radius 1,500,000 1,500,000 1,000,000 1,000,000 500,000 500,000 - - 2,000 4,000 6,000 - - 2,000 4,000 6,000 Employment of SEZ residents Employment of IZ residents Source: QazIndustry. Territorial clusters Since 2003, Kazakhstan has explored the place-based policy of TCs. While SEZs and IZs have a narrow sectoral view, TC policy acknowledges interindustry interdependencies and focuses on regional ecosystems of related industries and competences.156 At the initial development stage of TC initiatives, the government selected and supported networking activities for TCs, identifying strategic (horizontal) challenges and preparing the ground for TC organizations. Later, it intended to provide financial and nonfinancial support for the six TCs157 by financing TC organizations and pilot projects to address the horizontal challenges of a particular TC, such as research and development, infrastructure, or skills/input supply. (See Annex 7D for a short history of TC policy in Kazakhstan.) Due to the COVID- 19-induced economic recession, however, the expected public financing has not materialized for the TC organizations that were intended to coordinate TC activities. According to the law, a TC is a geographically concentrated group of interconnected and complementary companies and organizations.158 In contrast to SEZs and IZs, TCs do not offer tax, trade, or other incentives for firms. According to the approved feasibility study, TC stakeholders include central and local government, the private sector, academia, and TC organizations as TC centers. MIID, QazIndustry in particular, is a key TC stakeholder because it is an administrator of state support measures, develops TC eligibility criteria, and provides methodological support (such as selection criteria and monitoring and evaluation of TC initiatives) and technical assistance to TC organizations (including diagnostics, policy notes, and study tours). QazIndustry also coordinates TC policy throughout central government. Regional administrations and regional development institutes (“socioentrepreneurship corporationsâ€?) aim to position a particular TC in a broader context of local government policies. Regional authorities are also expected to address infrastructure issues and business environment issues. Additionally, they calibrate other regional policies, such as SEZ and IZ activities, in ways to account for TC activities. A TC organization is expected to be a formal TC center headed by an experienced industry professional and with some administrative staff. It is expected to be independent of corporate interests of a Kazakhstan Country Economic Memorandum 121 particular TC member and to promote facility sharing, knowledge exchange, and collaboration among TC members. TC organizations are also envisaged to be financed from public funds, with minor co- financing from the private sector. Finally, the private sector and academia are envisaged as participating in TC creation and operations. Existing TC policy identifies small and medium enterprises in manufacturing and services, regional business associations, and universities as ultimate beneficiaries of TC activities, underscoring that TCs are bottom-up initiatives aimed at solving horizontal issues and at increasing productivity. Kazakh stakeholders’ views on challenges in existing zone and cluster policies To understand particular challenges faced by zone and cluster stakeholders, we conducted 10 interviews with residents, management organizations, and QazIndustry policymakers. Interviews were conducted by videoconference between November 2020 and January 2021 and used a standardized questionnaire, which varied slightly depending on the interviewee role (for example, resident). During the interviews, there were mixed views on what works in zone and cluster policies and what incentives attract residents to these areas. Some interviewees mentioned tax incentives; others were attracted by the zone’s proximity to urban areas, the availability of land, and transport infrastructure. Several said that local market size was important, as well as links to the local economy. Cluster participants also highlighted the need to cooperate with upstream firms. It is close to Almaty, the transport hub of Kazakhstan. Chemical Park Taraz SEZ There is no single preference that we can figure out. For some investors, tax preferences are more important than infrastructure; others are more interested in waiver of labor permits. Pavlodar SEZ The main advantage of the zone is that it rents out land free of charge. This is a significant incentive since the market price is very high. Almaty IZ The interviews also exposed several important gaps and challenges in existing zone and cluster policies, along three main axes. First, the current SEZ law does not serve as an umbrella framework for place-based policies, because separate presidential or government decrees regulate individual zones and are too rigid. Consequently, there is a need to allow more autonomy and regulatory flexibility in zone and cluster operation. Even though this concern was addressed partly by the amended SEZ law in 2019 for newly created zones, existing zones still face a cumbersome process and long timelines to change anything in their operations. Current performance is falling short. … The main reason is … inflexibility of SEZ focus. We now realize that development should be driven by market demand, rather than focusing on the chemical industry only. Chemical Park Taraz SEZ We cannot spend money on activities outside of our mandate. … It would [also] be good if SEZs had more flexibility/autonomy in granting access to promising investors outside priority areas. Ontustik SEZ Kazakhstan Country Economic Memorandum 122 Additionally, interviewees repeatedly mentioned the imbalances in infrastructure provided such as lack of warehouses, inadequate rental space, insufficient water supply and sewage, and lack of public transportation from zones to major urban centers. In some cases, there was a mismatch between previously planned public investment and current local circumstances, but due to rigid SEZ and IZ policy frameworks, investment plans were rarely changed after the zones were established. The SEZ needs to have workspaces for rent for first-time investors … SEZ management companies can’t build workspaces for rent—they do not have such items in the SEZ master plan and respective budget. Ontustik SEZ Private companies are sometimes scared away by excessive reporting requirements that accompany public support measures. Almaty IZ Zone management companies struggle with financing their operational activities, and TC organizations have not received any of the proposed public financing. The management company does not have any funding under its direct responsibility. There are two ways to fund management company activities: by appointing it as an administrator of the SEZ infrastructure building activities or by providing equity funding for a management company. The first option is not working since according to the Budget Code, administrators of infrastructure building programs are construction departments of akimats, and we are not engaged in that activity (the SEZ management company is under a different department of the regional akimat). The second option is not working either: […] the regional akimat can’t provide equity funding for the SEZ management company due to the absence of such an expenditure item (or budget program) in the regional budget. Pavlodar SEZ Despite a broad legal mandate of zone management companies, these companies lack autonomy and flexibility in zone management and decision making, a point raised repeatedly during the interviews as a serious issue, impeding zone development. Participants specifically suggested that the zone institutional framework should be amended to give management companies more autonomy and ability to raise funds directly from resident firms. Ideally, the management company should be self-sufficient and earn on services rendered to residents. … Yet given the tiny number of residents, the management company is currently funded from the state budget. Chemical Park Taraz SEZ The issue of funding of the SEZ management company needs to be resolved. Ideally, our funding should come from services to residents, but due to low number of residents, this model is not working. Pavlodar SEZ The second axis is that the monitoring and evaluation framework is missing within MIID (and QazIndustry in particular), resulting in sporadic data collection and no monitoring of results. Consequently, interviewees felt that policymakers do not understand what works and what does not work in zone and cluster policies. Management faces challenges given this absence, including no systematic analysis and planning, leading to unrealistic indicators and goals being set when a zone is Kazakhstan Country Economic Memorandum 123 established as well as lack of knowledge about what works and does not. Related to this, some interviewees mentioned the low quality of zones’ feasibility studies. The management company’s funding does not depend on the number and quality of residents attracted. Almaty IZ No polls among residents. … No systematic monitoring. Chemical Park Taraz SEZ Business is about money, so it would be good to have … some sort of performance -based state- support measures. Dairy TC in Akmola region Third, zones and clusters struggle with the overall economic structural issues related to the business environment, labor market and skills, land property rights, and macroeconomic vulnerabilities. Zone residents often mentioned systemic economic issues of the business environment and labor market, such as lack of needed skills and low-quality labor supply (one of the most frequently cited), no protection from counterfeit products, and inability to acquire land. [The] land issue is a major one … investors … see the absence of land title as a risk to operation of the would-be plant. Chemical Park Taraz SEZ The lack of skilled labor and counterfeit goods are systemic problems. Ontustik SEZ [W]e have high turnover among entry-level employees. Ontustik SEZ Our business plan (projected return on investment) was hurt by the recent devaluation. Pavlodar SEZ [T]here are legal collisions, there is a lack of skills, there are collisions in regional development policies. … The systemic issue we see is the low quality of skills at all levels—from top management of construction companies to basic personnel, there is a low level of competency. Construction TC Residents often admitted that they lacked resources and knowledge about the market in general as well as about potential ways to attract investors, citing lack of communication from central authorities on promotion opportunities and their difficulty understanding policies. These points suggest the need for broad labor force policies to upgrade technical and managerial skills. We would like to … present our project to a wider audience of investors. Chemical Park Taraz SEZ There is distrust, especially at the farmers level. Dairy TC in Akmola region [W]e lack intelligence on markets of neighboring countries. Ontustik SEZ [E]xpensive logistics and customs procedures … as well as the lack of knowledge of the local market. Ontustik SEZ Kazakhstan Country Economic Memorandum 124 In summary, the interviewees saw pros and cons in existing zone and cluster policies. On the one hand, participants seemed satisfied with the incentives provided, such as tax exemptions and land rentals. On the other, residents and management companies repeatedly mentioned challenges related to the zones’ rigid policy design as well as systemic issues that inevitably affect every business in Kazakhstan, such as the low quality of local labor, weak enforcement of property rights, and access to financing. Learning from international experience: The importance of tailoring zones to local context Global experiences show that there can never be a one-size-fits-all approach in economic development, including SEZs and IZs. Even if there are several key success features of any zone program, customized approaches should be developed to fit the local socioeconomic context and policy goals. Even so, there are important global lessons, including the need for integrating zone initiatives into the host economy’s national and regional development strategies and ensuring effective policy implementation, choosing the right locations, understanding business demand, clearly defining and leveraging local comparative advantages or latent potential, and making SEZs and IZs truly “specialâ€? by providing high-quality infrastructure and a conducive business environment. Here, Kazakhstan can benefit from the policy lessons of successful experiences of other countries, such as the Republic of Korea, Poland, and the United Arab Emirates. In the Republic of Korea, a policy of export processing zones (EPZs) was introduced in the early 1970s as part of an export-oriented economic development strategy. The government established its first EPZ in Masan in 1970, followed by one in Iksan in 1973. It provided a favorable business environment to foreign investors within the EPZs (renamed free trade zones, FTZs, in 2000). Today, there are 13 such zones. Since 2000, the country has introduced free economic zones to promote industrial transformation by developing high value-added industries, such as logistics, information technology, and finance. The country now has eight such zones. Among the zones in the Republic of Korea, Masan FTZ is one of the most successful. It has contributed significantly to developing the country’s export industry by acquiring foreign technologies, engaging in foreign trade, and creating local jobs. It is regarded worldwide as a successful and representative SEZ. In about 40 years, Masan FTZ attracted investment of US$6.1 billion, and at its peak created 34,883 jobs and exported US$8 billion worth of goods and services. Of 94 investors in the zone, over 55 percent are foreign. As a highly concentrated industrial complex, by the 1990s, Masan FTZ generated up to 15 percent of the country’s foreign exchange earnings (Jeong and Zeng 2016). The key success factors of Masan FTZ include a superior location within a compact harbor, excellent labor supply from surrounding cities, proximity and links to related local industries, highly developed infrastructure, strong government support for export industries, and efficient one-stop services (Jeong and Zeng 2016). In Poland, SEZ policy in the 1990s prioritized narrowing regional disparities. The Kamienna Gora SEZ (KG SEZ), established in 1997, was one of the earliest and most successful SEZs in Poland. Its primary purpose was to reverse the high structural unemployment in the region resulting from post-Soviet economic restructuring, the phasing out of state-owned agricultural holdings, and the bankruptcy of state-owned enterprises. Small and medium enterprises were seen as important contributors to developing the subregional economy and, therefore, became the focus of the KG SEZ. The zone attracted foreign investment, especially from Germany, due primarily to its location (Kozak 2014). By Kazakhstan Country Economic Memorandum 125 2018, it had attracted 45 investors and US$700 million in investment and had created 6,647 local jobs (Velamuri and Zeng 2021). Several favorable factors contributed to the success of Poland’s SEZs. First, Poland’s location made it a perfect investment destination for enterprises targeting both Western and Eastern Europe. Second, its SEZs were not clearly demarcated enclaves restricted to a single location. Third, the structure of SEZs promoted links with the local economy, targeting mainly small and medium enterprises. The new investment law significantly strengthens these links by uniting zones with the rest of the national economy.159 Fourth, SEZs provide a probusiness environment, contributing to a virtuous cycle of reforms spilling over to the rest of the economy (Velamuri and Zeng 2021). In the United Arab Emirates, Dubai’s flagship Jebel Ali Free Zone (JAFZ) is one of the most successful free zones in the world and home to a plethora of global brands. The JAFZ was established in 1985 and is located adjacent to Jebel Ali Port. Dubai’s location, both geographic and strategic, is a natural business gateway between the East and West and has been an essential element of trade routes between Asia, Africa, Europe, and beyond for centuries. The JAFZ is also widely regarded as having some of the best logistics facilities in the world. Since its creation, it has generated 135,000 jobs and over US$80 billion worth of trade, and in recent years, it contributed over 20 percent of Dubai’s GDP (Jeong and Zeng 2016).160 Key sectors in the zone include logistics and warehousing, medical supply/instruments, machine tools, construction materials and equipment, glassware, paper products, food processing, electronics, chemicals, stationery, textile/fashion industries, and automotive (Jeong and Zeng 2016). The zone remains important in diversifying Dubai’s economy and expanding its manufacturing base. The JAFZ is administered by the Jebel Ali Free Zone Authority (JAFZA), which is responsible for the administrative oversight of all the activities in the zone, including real estate and property-related matters; marketing and sales; civil engineering and planning; environment, health, and safety; and customer services (Jeong and Zeng 2016). One of JAFZA’s operational advantages is its delegated authority. While some critical functions, such as immigration, remain in the hands of the federal government, JAFZA has the autonomy to register and license companies, issue building permits, and enforce its own guidelines (for instance, on construction and on environmental health and safety). The three key JAFZ success factors are its strategic location, conductive business environment, and suitable business facilities that help lower entry costs. Dubai’s location offers a distinct geographic advantage to businesses. Within a four-hour time difference, Dubai offers a significant catchment area, serving several dynamic emerging markets throughout the Middle East, Africa, and South Asia. Further, the government has taken steps to streamline business setup and operation in the JAFZ. Key to this is the authority delegated to JAFZA by the government. Finally, a key strategy of the JAFZ has been to offer the right facilities to the right customers, big or small, ranging from large plots of seafront land for constructing oil rigs to a single desk in a business center with shared facilities (Jeong and Zeng 2016). Conclusion and policy recommendations This chapter has attempted to answer the following three questions: What are the zone and cluster policies in Kazakhstan, and how do they compare to those in some other countries? What are the structural challenges impeding zone and cluster success? Kazakhstan Country Economic Memorandum 126 How can the Kazakhstan improve its zone and cluster policy framework to make these zones more effective? Kazakhstan’s policymakers have continually struggled with defining the goals and frameworks of zones and clusters, leading to overlapping goals and contradictory priorities. In contrast to China’s SEZ approach, Kazakhstan’s rigid top-down approach to zone management and operation excludes any local experimentation with these policy instruments. Chinese authorities designed SEZs primarily as experimental areas where continuous monitoring and evaluation of succeeding and underachieving projects influence the policy response. In Kazakhstan, piloting and experimenting with zone and cluster programs are missing. While the SEZ law states that these zones are designed primarily to develop new, modern, competitive industries and sectors, SEZ founding documents specify a rigid list of allowed economic activities, a priori excluding any potentially innovative activities yet to emerge. Additionally, the law specifies increased local employment as a key goal of both SEZs and IZs, but data show these zones fail to generate jobs locally and employ less than 1 percent of the local labor force. The private sector is seldom included in the design, development, and operation of these zones. Lack of regular data tracking and of evaluation of what works and what does not hinders development of successful zones and precludes knowledge transfer between regions. Consequently, similar policy design mistakes keep being repeated. Firms within SEZs and IZs inevitably stumble upon structural economic challenges present in a broader economy. Despite a special regulatory regime for SEZs, businesses in these zones still face numerous systemic issues related to the labor market (hiring and skills gap), land ownership, availability of capital, market access and size, competition, and red tape. These issues are graver when zones are created away from the urban centers where access to a quality labor force is limited, the market is small, and transportation costs are higher. Because zone residents operate in the overall national business environment, systemic reforms and regional development policies are complements, not substitutes. Still, Kazakhstan’s government designed SEZs and IZs to attract investment, spur entrepreneurship and innovation, and create jobs, and they have achieved some of these goals—for instance, attracting foreign direct investment in SEZs. For other priorities such as job creation, however, they have largely failed. Moreover, the public costs—direct financial costs and indirect social and environmental costs— have likely been much higher than the benefits realized. In addition, even as the government seeks to use these instruments to develop lagging regions, most of the more successful zones and clusters are situated either in or near vibrant regions’ urban areas. This outcome reflects global experience with zones: to be successful, zones should be in locations with comparative advantages, good trade connectivity, and links to the local economy. Many other countries have struggled with designing the right place-based policies to overcome internal regional disparities. While zones and clusters are often used for these purposes, authorities in Kazakhstan could consider three key factors. First, the management companies of the zones, as well as regional and local policymakers, need to have autonomy in decision making on policies for local socioeconomic development, including zone and cluster design, management, and operations. Local stakeholders and business communities should play a bigger role in establishing and operating zones. For instance, local governments might not have the resources to provide all the necessary infrastructure, but the private sector could contribute under public–private partnerships. Moreover, firms usually have more market and factor knowledge and Kazakhstan Country Economic Memorandum 127 could help policymakers both identify key issues and find practical solutions. Hence, local governments should have more autonomy to decide on the relevant policy tools rather than constantly mold their actions to central directives. China’s SEZ success lies primarily in the autonomy granted to those special geographic areas. Second, a pilot approach to zone policies is essential. Local authorities should be able to experiment with their zones—adjusting regulatory provisions, infrastructure plans, and service provision— depending on zone performance. For this, a solid monitoring and evaluation system should be established at the same time as documents on zone creation. The system should outline and track economic, social, and environmental indicators and set target values against which performance is to be measured. Mechanisms and processes to handle underachieving zones should also be in place. China’s first SEZs were established explicitly as “experimental zones,â€? and other zones were subsequently created based on lessons from these early experiments. Finally, because zone residents operate in the national business environment, systemic reforms and regional development policies are complements, not substitutes. In the survey, zone and cluster participants repeatedly mentioned challenges related to labor supply and labor quality, financing, land ownership and property rights, access to domestic and foreign markets, and issues of counterfeit goods. Because these reforms are often hard to implement and require a long time horizon, a pilot approach could also be useful here. The first steps could be providing specialized education and preparing a modern workforce locally, with incentives for firms to train or retrain their employees. Another policy initiative could focus on creating, also on a pilot basis, a land market within a particular region. Other pilots could substitute existing fiscal or tax incentives for zone participants with eased access to financing via the local financial sector. Local economies could be a testing ground for difficult structural reforms that afterward, if successful, could be rolled out nationwide. References Jeong, Hyung-Gon, and Douglas Zhihua Zeng. 2016. “Promoting Dynamic and Innovative Growth in Asia: The Cases of Special Economic Zones and Business Hubs.â€? KIEP Research Paper Policy Analysis16-01, Korea Institute for International Economic Policy, Sejong City, Republic of Korea. Kozak, Marek W. 2014. “Subregion Jelenia Gora as an Example of Border Region: Case Study Report.â€? GRINCOH Working Paper Series 6: Spaces, Territories and Regions. Growth–Innovation– Competitiveness: Fostering Cohesion in Central and Eastern Europe, European Commission, Brussels. Porter, Michael E. 1998. Clusters and Competition: New Agendas for Companies, Governments, and Institutions. Cambridge, MA: Harvard Business School Press. UNCTAD (United Nations Conference on Trade and Development). 2019. World Investment Report 2019: Special Economic Zones. Geneva: UNCTAD. Velamuri, M., and Douglas Zhihua Zeng. 2021. “Kamienna Gora SEZ: An Industrial Hub for SMEs in Poland.â€? Working Paper, Finance, Competitiveness and Innovation Global Practice, World Bank, Washington, DC. World Bank. 2020. World Development Report 2020: Trading for Development in the Age of Global Value Chains. Washington, DC: World Bank. Kazakhstan Country Economic Memorandum 128 Zeng, Douglas Zhihua. 2011. How Do Special Economic Zones and Industrial Clusters Drive China's Rapid Development? Washington, DC: World Bank. Kazakhstan Country Economic Memorandum 129 Annex 6A Pilot territorial clusters Cluster members Cluster Stated goals and challenges Private Academia Other Strengthening the leading position in domestic and foreign markets by expanding and conquering new • Aruana–2010 LLP niches for new products • Kostanay Flour Company • Kostanay Engineering • Tobol Wheat flour LLP and Economics Socioentrepreneurship cluster in • Kostanay Mill JSC University Corporation (regional Key challenges: Kostanay region • Mibeko LLP • Kostanay State University development institute) • Railroad infrastructure • Qazaq Uny LLP • Public export policy • Romana–NAN LLP • Certification • Salamat LLP • Export market intelligence • Lack of knowledge of downstream opportunities Enhancing competitiveness of regional pharmaceutical companies by fostering cooperation between public and private medical • Apteka #5 LLP institutions with a focus on education and research • Ecopharm International and development • Association of LLP Pharmaceutical and Key challenges: • HimPharm JSC Pharmaceutical • Center for Development Medical Organizations • Weak supply chain links • KazMedPribor Holding cluster in South of Science, Education • Association of • Underdeveloped distribution network LLP Kazakhstan and Business Pharmaceutical Cluster • Lack of investment • Merey LLP region • South Kazakhstan State of South Kazakhstan • Poor research and development infrastructure • Pharmacy Synthesis LLP Pharmaceutical Academy region • Planta LLP • Regional Chamber of • Zerde NGO Entrepreneurs, Atameken • Zerde-Fito LLP • ZerdePharma LLP Kazakhstan Country Economic Memorandum 130 • Alamak Stroy LLP Enhance competitiveness of Karagandy region’s • Bömer Armatura LLP construction sector in accordance with global • Braer Color LLP professional and technological requirements, • IP Kulyasov (IC) • Construction Business Construction practices, and trends • JBI-Karaganda LLP Development Center LLP materials cluster • Association of • Kraska Alian LLP • Karagandy Mining in Karagandy Developers of the Key challenges: • NORD Prom NS LLP Industrial College region Karagandy Region • Lack of skilled labor • PROLUX Led LLP • Karagandy State • Lack of cooperation among local firms, design • Santechprom LLP Industrial University bureaus, and research and development entities • TAS-2006 LLP • Dominance of cheap, low-quality construction + 14 construction and materials on the market design firms • Aina LLP • Aina LLP Maximize regional dairy sector’s ability to produce • Akkol farm high-quality products by creating a regional • Beke farm community of suppliers, processors, distributors, • CapitalNaturProduct LLP • Katharkol Agricultural academia, and regional institutions • Esil Milk Refinery LLP College • Esil • Gormolzavod LLP • Kokshetau State Socioentrepreneurship Dairy cluster in • Karnashauskene farm University named after Corporation Akmola region • MMM farm Sh. Ualikhanov • Regional Chamber of Key challenges: • Natizhe dairy factory LLP • S. Seifullin Kazakh Agro- Entrepreneurs, Atameken • Focus of public support measures on large farms • Omarov farm Technical University • Seasonality of raw milk supply • SPK Enbek-2017 • Underdeveloped processing and transportation • Zerendinskoe Moloko infrastructure LLP • Zhaksylyk Agro LLP • Al-Farabi Kazakh • Association Alma-Tau National University • Association of Camping Enhance competitiveness of tourist cluster • Alma University of and Caravan Tourism companies through cooperation focused on Management • Association of Travel strategic (horizontal) issues that cannot be solved Tourism cluster 22 tourism companies, • Almaty Technological Agencies at a single firm level of Almaty and hotels, and transport University • Caravan Club Association Almaty region companies • Kazakh Academy of • Charyn National Park Key challenges: Sports and Tourism • Cycling Development • Lack of skilled labor • Kazakh–Swiss Tourism Fund • Lack of quality standards and certification system Institute • Tanbaly National Reserve at the regional and national levels • Narxoz University Museum Kazakhstan Country Economic Memorandum 131 • Insufficient or poor quality of statistical data • Turan University • Tourist Information • Lack of infrastructure at destination Center Almaty • Underdeveloped tourist transportation market • Union of Kazakhstani Craftsmen • Alma-Furniture TPK LLP Increase the domestic market share and enter • Almaty Furniture Factory export markets LLP • Ariba LLP • AWOOD LLP • Almaty Construction and • Association of Furniture Key challenges: • KARA Company LLP Technical College and Woodworking Furniture cluster • Technological upgrading • Master Grad LLP • Kazakh Head Industries of Almaty city • Sales infrastructure • PKF Tornado PLUS LLP Architecture and Civil • Regional Chamber of • Underdeveloped furniture design market • Sanni Master LLP Engineering Academy Entrepreneurs, Atameken • Standards • SP Yakupov AR • Skilled labor • Sylwia TPK LLP • Dependence on imports of input materials + 11 suppliers of input materials Note: JSC = joint stock company; LLP = limited liability partnership. Kazakhstan Country Economic Memorandum 132 Annex 6B Special economic zones Special economic Establishment Size Location Residents Main features zone and closure (hectares) • Support to upstream Name: Ontustik and downstream participants of textile Management Shymkent city 2005–30 200 15 industry company of • Access to cheap labor, Ontustik SEZ energy, and raw material (cotton) • Petrochemical industry, Name: Pavlodar aluminum processing • Access to cheap Management Pavlodar region 2011–36 3,300 10 electricity, proximity to company of one of three largest oil Pavlodar SEZ refineries and aluminum plant Name: PIT Alatau • Information technology Almaty city 2003–28 163 167 Technopark Alatau and high-tech startups LLP Name: Aktau Sea Port Mangystau 2003–28 2,130 18 • Maritime transportation region Aktau Sea port JSC SEZ Name: Qyzyljar North • Not functioning at time Management Kazakhstan 2019–44 192 1 of study company of region Qyzyljar SEZ Name: Saryarka Karagandy • Metallurgy, waste Management 2011–36 595 7 region processing company of Saryarka SEZ • Real estate development Name: Turkistan • Development of historical sites and Management Turkistan region 2018–43 3,014 1 administrative center of company of Turkistan city Turkistan SEZ • Development of tourism Name: Astana– • Real estate New City development • Development of Management Astana city 2002–27 15,472 48 administrative center of company of Astana Astana– • Industrial enterprises Technopolis SEZ Kazakhstan Country Economic Memorandum 133 Name: Astana– Technopolis • Information technology Management Astana city 2017–42 653 2 and high-tech startups company of Astana– Technopolis SEZ Name: National Industrial Petrochemical Technopark Atyrau region 2007–32 3,476 1 • Petrochemical industry Management company of National Industrial Petrochemical Technopark SEZ Name: International Center for Cross- border Cooperation Khorgos • Cross-border wholesale Almaty region 2017–41 609 1 and retail trade Management • Tourism company of the Khorgos International Center for Cross- Border Cooperation Name: Khorgos– Eastern Gates • Multimodal transport Management Almaty region 2011–35 4,591 5 and logistics center company of Khorgos–Eastern Gates SEZ Name: Chemical Park Taraz • Chemical industry Management Zhambyl region 2012–37 505 1 • Production of mineral company of fertilizers Chemical Park Taraz SEZ Note: JSC = joint stock company; LLP = limited liability partnership. Kazakhstan Country Economic Memorandum 134 Annex 6C Industrial zones Size Industrial zone Location Established Residents Key features (hectares) • Cement factory Name: Shieli district Kyzylorda • Soda ash factory 2014 96 4 region • Storage of hazardous Kyzylorda regional akimat chemicals • Agri-industrial zone Name: Serpin Kyzylorda • Rice processing 2012 22 8 region • Textiles Kyzylorda regional akimat • Cattle market Name: Karmakshy district Kyzylorda • Production of concrete 2017 15 1 region structures Kyzylorda regional akimat • Bottled water Name: Kazaly district Kyzylorda • Greenhouse, vegetable and 2015 6.9 5 region fruit warehouse Kyzylorda regional akimat • Concrete structures Name: Aral district Kyzylorda • Food production 2013 10 2 region • Construction materials Kyzylorda regional akimat • Brick factory. Name: Ondiris • Planned projects include glass Kyzylorda 2014 760 1 factory, ferrous alloy plant, region Kyzylorda regional akimat construction materials, and molybdenum production Name: Kazygurt Turkistan • Food and beverages 2013 35 5 Turkistan Management Company region • Construction materials of Industrial Zones LLP Name: Tulkibas Turkistan 2015 55 0 Not functioning at time of study Turkistan Management Company region of Industrial Zones LLP Name: Baidibek Turkistan 2013 57 0 Not functioning at time of study Turkistan Management Company region of Industrial Zones LLP Name: Shardara Turkistan • Cotton processing 2014 35 3 Turkistan Management Company region • Construction materials of Industrial Zones LLP Name: Maktaaral • Asphalt plant Turkistan 2014 28 2 • Cotton fiber production Turkistan Management Company region factory of Industrial Zones LLP Name: Sozak Turkistan • Services to uranium mining 2012 50 5 Turkistan Management Company region company of Industrial Zones LLP Name: Kentau Turkistan • Production of power 2013 25 3 region transformer parts Kazakhstan Country Economic Memorandum 135 Turkistan Management Company of Industrial Zones LLP Name: Badam Turkistan Next to Shymkent city, no 2014 115 4 Turkistan Management Company region specific industrial focus of Industrial Zones LLP Name: Turkistan Turkistan • Construction materials 2012 350 9 Turkistan Management Company region • Furniture production of Industrial Zones LLP • Power transformers Name: Tassay • Beverages Shymkent 2016 89 10 • Construction materials Shymkent Socioentrepreneurship city • Medical products Corporation • Furniture Name: Ontustik • Metallurgy Shymkent • Lubricants 2010 337 57 Shymkent Socioentrepreneurship city • Construction materials Corporation • Chemicals Name: Ondiris • Recycling of medical waste East • Road construction materials Kazakhstan 2010 76.6 5 Yertis Socioentrepreneurship • Construction materials region Corporation • Basic railroad equipment Name: Orken KShT East Kazakhstan 2007 33.5 0 Not functioning at time of study Yertis Socioentrepreneurship region Corporation Name: Mashinostroiteley Street East Kazakhstan 2007 26.6 0 Not functioning at time of study Yertis Socioentrepreneurship region Corporation Name: Kostanay In industrial zone of Kostanay Kostanay 2014 400 2 city, next to machinery plant Tobol Socioentrepreneurship region • Two tractor assembly plants Corporation Name: Aktobe In industrial zone of Aktobe city Aktobe • Transport logistics center 2015 200 7 Aktobe Socioentrepreneurship region • Chemistry Corporation • Polyethylene plastic bags • Car assembly plant Name: Almaty • Greenhouse Almaty city 2013 490 13 • Machine building Almaty Industrial Zone LLP • Construction materials Note: JSC = joint stock company; LLP = limited liability partnership. Kazakhstan Country Economic Memorandum 136 Annex 6D History of cluster policy in Kazakhstan The first territorial clusters (TCs) were introduced in 2003–04 to diversify the economy based on the country’s competitive advantage. The cluster approach was intended to foster links among enterprises, academia, and local authorities, with a longer-term vision and a stronger flow of high-quality investment opportunities. Since 2003, the country has seen three waves of cluster development policy measures. 2003–10: Identification of regional and subregional territorial clusters in nonresource sectors In 2003, the government identified seven potential TCs (metallurgy, transport logistics, textiles, tourism, oil and gas machinery, construction materials, and food production). In 2005, the government added medical and agriculture TCs (including those for grain, meat, and horticulture). The policy envisaged legislative changes to support TCs, a coordination mechanism between central and local governments and the business community, and pilot investment projects. However, the policy did not achieve its stated objectives of forming long-term nonresource sector TCs for several reasons, including a top-down approach in forming TCs, lack of political support, little private sector participation, and the global financial crisis of 2008–09.161 2010–15: Focus on national territorial clusters In 2010–15, the government turned to creating high-tech and innovation TCs in Astana (centered on Nazarbayev University) and in Almaty (the Sovereign Cluster Fund and Technopark Alatau), as well as a petrochemical TC in Mangystau and Atyrau regions (the oil refinery in Atyrau and two seaports in Mangystau region).162 Development of these TCs was incorporated in the State Industrial Innovative Development Program for 2015–19. Specified policy measures included requiring subsoil resource users (that is, mining companies) to channel 1 percent of annual revenue to the Sovereign Cluster Fund for developing startups and other research and development initiatives. Nazarbayev University has been positioned as a TC with focus on breakthrough technologies and science. The petrochemical TC initiative led to the de facto launch of a special economic zone in Atyrau region in 2014 and conclusion of nonbinding memoranda with major oil companies in the region. 2015–present: Launch of pilot territorial clusters In parallel with national TCs, the government selected pilot TCs in six regions (Annex 7A) within the framework of the Enhancing SME Competitiveness project financed by the World Bank in 2016–19. The project included an explicit focus on small and medium enterprises, a bottom-up approach in selecting pilot TCs, and extensive capacity building of the state agency, QazIndustry, which oversees TCs. The first six TCs were launched in 2015, and the process was led by the public industrial development institute (Kazakhstan Center for Industry and Export, or QazIndustry joint stock company, formerly Kazakhstan Industrial Development Institute, or KIDI). The process focused on providing nonfinancial support measures to TCs driven by small and medium enterprises, including building capacity for TC participants, identifying horizontal problems, and developing TC action plans. In 2016–17, the TC development activities progressed in line with the original schedule, but the process stalled in late 2017. The final endorsement of the six selected pilot TCs was delayed and completed only in 2019 with the adoption of the new State Program for Industrial and Innovative Development for 2020–25. In addition, in October 2018, KIDI was reorganized and merged with another development institute, which entailed change of top management and removal of the key personnel that led the TC 137 work. And in 2020, planned program rollout was again delayed by the public financing reprioritization due to the COVID-19 pandemic. 138 1 Productivity is the efficiency of transforming inputs like labor and capital into production. 2 Calculations based on the 2019 Kazakhstan Labor Force Survey. 3 Decomposition of the urban–rural income gap uses the recentered influence function approach . 4 Seitz 2008. 5 Aktobe region is excluded because of an anomaly in its trade growth. 6 Government of Kazakhstan (2019) Forecast Scheme for Territorial Development until 2030. [In Russian.] 7 A functional urban area consists of a city and its commuting zone. It therefore includes a densely inhabited city and a less densely populated commuting zone whose labor market is highly integrated with the city. This definition, originally proposed by the Organisation for Economic Co-operation and Development, has been recognized and adopted by the European Union and the World Bank. See https://ec.europa.eu/eurostat/statistics- explained/index.php/Glossary:Functional_urban_area. 8 World Bank Group. 2022. Kazakhstan Country Climate and Development Report. CCDR Series. World Bank, Washington, DC. © World Bank Group. https://openknowledge.worldbank.org/handle/10986/38215 License: CC BY- NC-ND. 9 Following the example of the US Economic Development Administration. 10 Data from the National Statistical Committee of Kazakhstan (2018). 11 According to national accounts data, services that exclude oil and gas services account for around 48 percent of GDP. 12 Data from the National Statistical Committee of Kazakhstan (2018). 13 Data from the National Statistical Committee of Kazakhstan (2018). 14 World Development Indicators data on GDP per capita growth (annual percent). 15 Penn World Tables. 16 Productivity is the efficiency of transforming inputs like labor and capital into production. 17 CEM 2018. 18 The compound annual growth rate (CAGR) of labor productivity in Kazakhstan is 0.1 compared with 0.02 annual growth rate in labor participation between 2000 and 2011 (elaboration based on Penn World Tables 9.1). 19 Between 2010 and 2021, the labor force participation rate of female population declined from 65.4 percent to 63.3 percent, and for the male population, from 75.9 percent to 75.4 percent. The figures are based on ILO estimates obtained from the World Development Indicators database. 20 TFP growth is averaged over three years. Aggregate TFP is weighted by value added. For 2013 and 2014, TFP was calculated using interpolated employment figures. 21 For sales per worker, medians as opposed to averages are reported because of the skewed distribution that is characteristic of this variable. In regression analysis and tests for statistical significance of differences, logged sales per worker was used. 22 UN Population Prospects 2020. 23 The cotton-producing Turkistan region hosts the Ontustik special economic zone, specializing in cotton processing and textile. 24 For a comprehensive demonstration, see the models of wage rate convergence in Ganong and Shoag (2017), who suggest a tendency toward roughly equalizing real wages across regions. 25 For instance, Karlag comprised most of the present-day Karagandy region. Akmol, the administrative center of Tselinograd district near what is now the capital, Astana, was similarly established to support a remote prison labor camp. 26 See the 1969 Standard Methodology for Determining the Economic Effectiveness of Capital Investments. 27 The increase in urban population for the entire Soviet Union was 36 percent over the same period. 28 See Lex and Sunega (2014) for more details. 29 Secondary issues include greater commuting times and fewer new businesses in areas with high ownership rates. 30 The large majority was due to natural population growth within cities. 31 About 300,000–400,000 people. 139 32 In such cases, it is customary to use survey data as the reference. Administrative (including registration) data sources are used for administering government programs (see the section on recent migration patterns), which can incentivize strategic behavior. Properly implemented survey data are free of such potential bias. 33 Decomposition of the urban–rural income gap uses the recentered influence function approach. 34 In nominal terms, prices increased by factors of about 12 and 15, respectively. 35 Except in rare and extreme cases for specific housing markets. 36 Law on Migration of the Population, No. 477-IV, July 22, 2011 (amended on December 27, 2019), article 1, paragraph 17-1. 37 Code of the Republic of Kazakhstan on Administrative Offenses, No. 235-V, July 5, 2014 (amended and supplemented as of January 16, 2020), article 1, paragraph 17-1, https://online.zakon.kz/document/?doc_id=31577399#pos=7549;-54. 38 Law on Introducing Amendments and Additions to Some Legislative Acts of the Republic of Kazakhstan on Countering Extremism and Terrorism, No. 28-VI, 2016, https://online.zakon.kz/Document/?doc_id=34199995#pos=1;-161. 39 The statistics were provided upon author request by the Committee on Legal Statistics and Special Accounts of the General Prosecutor's Office of the Republic of Kazakhstan, May 26, 2020, ref. no. 2-20-20-04048. 40 Galina Sarsenova, deputy chairperson, Migration Committee of the Ministry of Internal Affairs, in an official letter to the author, June 5, 2020, ref. no. 8-8-5-53/8-7931. 41 The Listening to Kazakhstan survey has been conducted by the World Bank, the United Nations Children’s Fund, and local partners by phone every month since December 2020, with a panel of approximately 1,500 households, representative of major cities, other urban areas, and rural areas. Each month, households are asked core questions covering their welfare, including shocks, incomes, savings, wellbeing, migration, employment, and opinions on public policy, as well as COVID-19-related questions. 42 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of the Rules for the Provision of the State Basic Pension Payment at the Expense of Budgetary Funds, as well as the Appointment and Implementation of Pension Payments by Age, State Social Benefits for Disability, in Case of Loss of the Breadwinner, Special State Benefits, No. 223, April 14, 2015, paragraph 33, https://tengrinews.kz/zakon/pravitelstvo_respubliki_kazahstan_premer_ministr_rk/sotsialnoe_obespechenie/id- V1500011110/. 43 Law of the Republic of Kazakhstan on State Targeted Social Assistance, No. 246, July 17, 2001, article 3, paragraph 1, http://adilet.zan.kz/rus/docs/Z010000246; Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of Standards of Public Services in the Social and Labor Sphere, No. 279, April 28, 2015, public service standard, Assignment of Benefits for Childbirth and Childcare, chapter 2, paragraph 9, http://adilet.zan.kz/rus/docs/V1500011342#z18. 44 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of Standards of Public Services in the Social and Labor Sphere, No. 279, April 28, 2015, public service standard, Appointment of Social Benefits in Cases of Social Risks: Disability; Loss of Breadwinner; Job Losses; Loss of Income Due to Pregnancy and Childbirth; Loss of Income in Connection with Adoption of Newborn Baby(s); Loss of Income Due to Care after the Child Reaches the Age of One Year, chapter 2, paragraph 9, http://adilet.zan.kz/rus/docs/V1500011342#z18. 45 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of Standards of Public Services in the Social and Labor Sphere, No. 279, April 28, 2015, public service standard, Paperwork for People with Disabilities to Provide Them with the Services of an Individual Assistant for People with Disabilities of the First Group Who Have Difficulty in Moving, and a Sign Language Specialist for People with Hearing Disabilities; Providing Disabled People with Typhlo-technical and Mandatory Hygiene Products; Paperwork for Disabled People to Provide Them with Prosthetic and Orthopedic Care, http://adilet.zan.kz/rus/docs/V1500011342#z18. 46 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of Standards of Public Services in the Social and Labor Sphere, No. 279, April 28, 2015, public service standard, Registration of Citizens Affected by Nuclear Tests at the Semipalatinsk Nuclear Test Site, Payment of One-Time State Monetary Compensation, Issue of Certificates, chapter 2, paragraph 9, http://adilet.zan.kz/rus/docs/V1500011342#z18. 140 47 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules for the Provision and Use of Housing from the State Housing Fund or Housing Leased by a Local Executive Body in a Private Housing Fund, No. 1420, December 1, 2011, chapter 2, article 8, paragraph 7, http://adilet.zan.kz/rus/docs/P1100001420#z31. 48 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules for the Provision and Use of Housing from the State Housing Fund or Housing Leased by a Local Executive Body in a Private Housing Fund, No. 1420, December 1, 2011, chapter 2, article 5, http://adilet.zan.kz/rus/docs/P1100001420#z31. 49 Law of the Republic of Kazakhstan, on Housing Relations, No. 94, April 16, 1997, article 73, http://adilet.zan.kz/rus/docs/Z970000094. 50 See, for example, the comment of Abylaykhan Ospanov, the former chairman of the board of the state corporation Government for Citizens: “Until recently, the personnel service [of the corporation] demanded address certificates when hiringâ€? (Pokidayev 2019). 51 Labor Code of the Republic of Kazakhstan, No. 414-V, November 23, 2015 (amended January 1, 2020), article 28. 52 Electronic Government of the Republic of Kazakhstan, Answer of the Minister of Labor and Social Protection of the Population of the Republic of Kazakhstan to Question No. 586478, December 30, 2019, https://dialog.egov.kz/blogs/all- questions/586478. 53 Private health care providers do not differentiate between unregistered and registered patients when providing paid services. But some private medical organizations have agreements with the Social Medical Insurance Fund (Фонд Ñ?оциального медицинÑ?кого Ñ?трахованиÑ?) for providing public medical services under Compulsory Social Health Insurance (обÑ?зательное Ñ?оциальное медицинÑ?кое Ñ?трахование), in which case the same rules apply as for public health care organizations. 54 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of the Rules of Provision of Primary Health Care and the Rules of Attachment to Organizations of Primary Health Care, No. 281, April 28, 2015, chapter 2, article 8, http://adilet.zan.kz/rus/docs/V1500011268. 55 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of the Rules of Provision of Primary Health Care and the Rules of Attachment to Organizations of Primary Health Care, No. 281, April 28, 2015, chapter 2, article 8, http://adilet.zan.kz/rus/docs/V1500011268. 56 Order of the Minister of Health and Social Development of the Republic of Kazakhstan on Approval of the Standards of Public Health Services, No. 272, April 27, 2015, public service standard, Attachment to a Primary Health Care Provider, chapter 2, paragraph 9, http://adilet.zan.kz/rus/docs/V1500011304#z49. 57 First Almaty City Polyclinic, Frequently Asked Questions, http://1gp.kz/gostevaya/chasto-zadavaemye-voprosy. 58 Private schools and private kindergartens typically enroll children regardless of registration status. But private education services are considerably more expensive than those provided by the state. 59 Order of the Minister of Education and Science of the Republic of Kazakhstan on Approval of Standards of Public Services in the Field of Secondary Education Provided by Local Executive Bodies, No. 179, April 8, 2015, article 1, paragraph 3, https://tengrinews.kz/zakon/pravitelstvo_respubliki_kazahstan_premer_ministr_rk/konstitutsionnyiy_stroy_i_osnovyi_g osudarstvennogo_upravleniya/id-V1500011057/. 60 Order of the Minister of Education and Science of the Republic of Kazakhstan on Approval of Standards of Public Services in the Field of Secondary Education Provided by Local Executive Bodies, No. 179, April 8, 2015, article 9, https://tengrinews.kz/zakon/pravitelstvo_respubliki_kazahstan_premer_ministr_rk/konstitutsionnyiy_stroy_i_osnovyi_g osudarstvennogo_upravleniya/id-V1500011057/. 61 Almaty City Development Center, “V perviy raz v pervyi klassâ€? [“For the first time to the first gradeâ€?], https://open- almaty.kz/ru/pervyy-raz-v-pervyy-klass. 62 Rauf Sabitov, personal interview, April 21, 2020. 63 Order of the Minister of Education and Science of the Republic of Kazakhstan on Approval of Standards of Public Services Provided by Local Executive Bodies in the Field of Preschool Education and Training, No. 172, April 7, 2015, article 9, https://tengrinews.kz/zakon/pravitelstvo_respubliki_kazahstan_premer_ministr_rk/obpazovanie/id-V1500010981/. 64 Order of the Minister of Justice of the Republic of Kazakhstan on Approval of Standards of Public Service of Registration of Acts of Civil Status and Apostille, No. 219, public service standard, Registration of the Birth of a Child, 141 Including the Introduction of Changes, Additions, and Corrections to the Records of Acts of Civil Status, April 17, 2015, article 9, http://adilet.zan.kz/rus/docs/V1500011374. 65 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 4, http://adilet.zan.kz/rus/docs/P1100001427#z15. 66 Based on the interview with Rauf Sabitov, Center for Social Adaptation of Orphans and Graduates of Orphanages “Miracleâ€? (Taraz), April 22, 2020. 67 Decision of the XXI session of Maslikhat of Almaty city of the VI Convocation On Approval of the Rules for Regulating Migration Processes in the City of Almaty, No. 152, September 15, 2017, Article 14, https://tengrinews.kz/zakon/gosudarstvennyie_organyi_almatyi/konstitutsionnyiy_stroy_i_osnovyi_gosudarstvennogo_ upravleniya/id-V17R0001410/. 68 Decision of the Maslikhat of Astana city On Approval of the Rules for Regulating Migration Processes in the City of Astana, No. 356/45-VI, March 6, 2019, Article 10, https://tengrinews.kz/zakon/gosudarstvennyie_organyi_goroda_astanyi/konstitutsionnyiy_stroy_i_osnovyi_gosudarstve nnogo_upravleniya/id-V19AAZ01210/. 69 Official Information Source of the Prime Minister of the Republic of Kazakhstan, “Mortgage on Favorable Terms and Construction of Affordable Housing Development of Housing Construction in Kazakhstan,â€? July 28, 2020. https://www.primeminister.kz/en/news/reviews/mortgage-on-favorable-terms-and-construction-of-affordable- housing-development-of-housing-construction-in-kazakhstan. 70 The quarantine measures enforced by the government and municipalities due to the COVID-19 pandemic in March– May 2020 exacerbated the situation of internal migrants without proper registration, cutting off many commuter workers from their sources of income in cities. On the other hand, checks inside cities resulted in additional inconveniences and penalties for people who stayed in quarantine outside their registration area (Abilmazhitova 2020). 71 Official Information Source of the Prime Minister of the Republic of Kazakhstan, “Otmena adressnoi spravki: kak budut predostavlyatsa gosuslugiâ€? [“Cancellation of an address certificate: How will public services be providedâ€?], October 30, 2019, https://primeminister.kz/ru/news/press/otmena-adresnoy-spravki-kak-budut-predostavlyatsya-gosuslugi. 72 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 2, Paragraph 4, http://adilet.zan.kz/rus/docs/P1100001427#z15. 73 These are the only questions on domestic migration flows in the Kazakhstan labor force survey. 74 Law on Migration of the Population, No. 477-IV, 22 July 2011 (amended on December 27, 2019), Article 1, Paragraph 17-1. 75 The Code of the Republic of Kazakhstan on Administrative Offenses, No. 235-V, 5 July 2014 (amended and supplemented as of January 16, 2020), Article 1, Paragraph 17-1, https://online.zakon.kz/document/?doc_id=31577399#pos=7549;-54. 76 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 13, http://adilet.zan.kz/rus/docs/P1100001427#z15. 77 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 3, http://adilet.zan.kz/rus/docs/P1100001427#z15. 78 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 2, Paragraph 4, http://adilet.zan.kz/rus/docs/P1100001427#z15. 79 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 2, Paragraph 4, http://adilet.zan.kz/rus/docs/P1100001427#z15. 80 Decree of the Government of the Republic of Kazakhstan on Approval of the Rules of Registration of Internal Migrants and Amendments to Some Decisions of the Government of the Republic of Kazakhstan, No. 1427, December 1, 2011, Article 2, Paragraph 4, http://adilet.zan.kz/rus/docs/P1100001427#z15. 142 81 Order of the Minister of Internal Affairs on Approval of the Rules for the Provision of Public Services on Issues of Documentation and Registration of the Population of the Republic of Kazakhstan, No. 267, March 30, 2020, http://law.gov.kz/client/#!/doc/140840/rus. 82 The Constitution of the Republic of Kazakhstan, Article 21, Paragraph 1, https://www.akorda.kz/en/official_documents/constitution. 83 The Constitution of the Republic of Kazakhstan, Article 14, Paragraph 2, https://www.akorda.kz/en/official_documents/constitution. 84 This overview covers only the residency registration legislation regarding citizens of Kazakhstan moving within its borders. 85 Law on Migration of the Population, No. 477-IV, July 22, 2011 (amended on December 27, 2019), Article 51, Paragraph 2, Subparagraph 1, https://online.zakon.kz/document/?doc_id=31038298#pos=1094;-46. 86 The Code of the Republic of Kazakhstan on Administrative Offenses, No. 235-V, July 5, 2014 (amended and supplemented as of January 16, 2020), Article 492, https://online.zakon.kz/document/?doc_id=31577399#pos=7549;-54. 87 The Code of the Republic of Kazakhstan on Administrative Offenses, No. 235-V, July 5, 2014 (amended and supplemented as of January 16, 2020), Article 493, https://online.zakon.kz/document/?doc_id=31577399#pos=7549;-54. 88 Decree of the Government of the Republic of Kazakhstan on Approval of Model Rules for Regulating Migration Processes in Regions, Cities of Republican Significance, the Capital, No. 296, May 26, 2017, https://tengrinews.kz/zakon/pravitelstvo_respubliki_kazahstan_premer_ministr_rk/konstitutsionnyiy_stroy_i_osnovyi_g osudarstvennogo_upravleniya/id-P1700000296/. 89 Order of The Minister of Internal Affairs on Approval of the Rules for the Provision of Public Services on Issues of Documentation and Registration of the Population of the Republic of Kazakhstan, No. 267, March 30, 2020, http://law.gov.kz/client/#!/doc/140840/rus. 90 Decree of the Government of the Republic of Kazakhstan on Approval of Model Rules for Regulating Migration Processes in Regions, Cities of Republican Significance, the Capital, No. 296, May 26, 2017, https://tengrinews.kz/zakon/pravitelstvo_respubliki_kazahstan_premer_ministr_rk/konstitutsionnyiy_stroy_i_osnovyi_g osudarstvennogo_upravleniya/id-P1700000296/. 91 By Sjamsu Rahardja, with contributions from Zarina Adilkhanova (NAC Analytica, Nazarbayev University). Zhansaya Kanatova contributed to updating data. Thanks to Shawn Tan for sharing the data on the Ellison-Glaeser index. 92 Redding and Venables (2004) and Hanson (2005) conducted empirical studies that found positive correlations between wages and market access and between wages and geographic concentration of activities. 93 Based on Committee on Statistics data that cover aggregate trade across regions and cities by businesses with at least 50 employees. Outliers were removed by combining data for Almaty city with data for Almaty region, data for Astana city with data for Akmola region, and data for Shymkent city with data for Turkistan region (to form South Kazakhstan). 94 Values are averages for 2014–15 and for 2018–19. 95 Regional trade data were obtained from the Ministry of Trade and Integration and were collected by the Ministry of National Economy’s Committee on Statistics. However, internal trade might include misreporting of indirect imports— imports purchased from local suppliers. The data also do not capture informal trade through bazaars, which is probably large. Mussin (2017) describes informal trade in northern Kazakhstan, where informal traders bypassed state regulations on tax, licenses, and health and safety. 96 Travel distance calculated through Google Maps. 97 The rapid growth of Shymkent city could attract new businesses to cater to greater and more diversified market demand. 98 Population density was obtained as the ratio of the population in a region to the area of the region. Data on the population, area, and length of automobile roads in each region were also gathered from the Committee on Statistics. 𝑡𝑖𝑗 ( ) 𝐺𝐷𝑃𝑗 99 The location quotient is calculated as â?„ 𝑡𝑖 , where tij is component of intranational trade i for region j, GDPj is ( ) 𝐺𝐷𝑃 the nominal output of region j, and ti and GDP are the corresponding aggregate national values. 143 100 See, for example, Redding and Venables (2004) for the context of international trade and Breinlich (2006) for regional economic integration. 101 Unlike analyses in international trade, the dummy variable for the common border is not included because the sample covers only Kazakhstan. 102 Nonmining GDP and log scale for the vertical axis are used to avoid outliers in GDP of several regions, such as Atyrau. 103 Redding and Venables (2004) refer to this as a “wage equationâ€? in which firms break even if their sales equal the sum of demands (market capacities), weighted by the distance to partner countries. 104 See Venables (2020) for a review. 105 Aktobe region is excluded because of an anomaly in its trade growth. 106 See ADB (2020) for more details. The authors express special thanks to the CAREC CPMM team for sharing raw data. 107 The distance between regions was collected from Google Maps. The distance between Kazakh regions was the shortest distance between their capitals. Distances were the quickest road routes provided by Google Maps. 108 Consumer price index microdata from Kazakhstan were collected monthly between January 2011 and March 2020. They consisted of 111 observations for 10 products across regions. The data for Kazakhstan incorporates all 14 regions (oblasts), as well as three cities of national signiï¬?cance—Almaty, Astana, and Shymkent. Because South Kazakhstan region was divided into Turkistan region and Shymkent city in 2018, price indexes are not available for before 2018. The food price index data in Kazakhstan are gathered for only four products and so are limited. Data were obtained from the Taldau information and analytical system of the Committee on Statistics. The regional price indexes are normalized, with January 2011 prices as the base. Seasonal adjustments were carried out using the X-11 method of the X-12-ARIMA family, following the methodology used by the National Bank of Kazakhstan. 109 Microdata studies on US cities suggest a persistent deviation of law of one price, which correlated with distance. See Cecchetti, Mark, and Sonora (2002) and Crucini, Shintani, and Tsuruga (2014). 110 Prices are turned into indexes with January 2011 as the base and then seasonally adjusted using X –12 method. 111 Astana Times, July 29, 2019. 112 The half-life measure is calculated using ln (0.5) , where λ (k) is the adjustment speed of prices of product k in region ij ln [] i from a shock in region j. 113 Cechetti, Mark, and Sonora (2002) found large differences between minimum and maximum inflation rates for nontraded consumer price indexes (CPIs) across US cities despite the high internal mobility of labor and capital. They concluded that nontraded activities probably contribute to slow price convergence (overall CPI) across US cities. 114 The first term in the equations for MAS and MAB captures intraregional market access. 115 There are 287 zeros in the 980 observations of bilateral trade. The Poisson pseudo-maximum likelihood estimator is used to correct potential bias and heteroskedasticity. The level of trade (instead of its natural logarithm) is used as the dependent variable, but coefficient estimates are interpreted as in ordinary least square, which indicate elasticity. Discussion on the approach can be found in Silva and Tenreyro (2006) and Arvis and Shepperd (2013). 116 Null hypothesis of γ + θ = 0. 2 117 Obtained from (𝑒 0.728 − 1) ∗ 100%. 118 An error correction model is commonly used to test market integration; see, for example, Baffes (1991), Federico (2005), and Kitenge and Morshed (2019). 119 The coefficients of λ are negative. The more negative a coefficient, the swifter the correction from a price deviation to the long-term equilibrium. 120 Government of Kazakhstan (2019) Forecast Scheme for Territorial Development until 2030. [In Russian.] 121 Nighttime light emissions data are widely used as a proxy for economic activity, including such metrics as gross domestic product and gross value added in areas where these metrics are not available through conventional methods. Definitions of agglomerations have been borrowed from the SPRD: second-level agglomerations include the city of Samey and all regional centers except Aktobe. 122 Calculations based on the 2019 Kazakhstan Labor Force Survey. 123 Data only for a sample of monotowns was collected using our methodology. 124 Republic of Kazakhstan (1996) The concept of regional вумудщзьуте of the Republic of Kazakhstan. КонцепциÑ? региональной политики РеÑ?публики КазахÑ?тан (поÑ?тановление ПравительÑ?тва от 9 Ñ?ентÑ?брÑ? 1996 г. â„– 1097). 144 125 Republic of Kazakhstan (2001) The concept of regional policy of Republic of Kazakhstan. КонцепциÑ? региональной политики РеÑ?публики КазахÑ?тан ¨[КонцепциÑ? региональной политики РеÑ?публики КазахÑ?тан на 2002-2006 годы (поÑ?тановление ПравительÑ?тва от 7 декабрÑ? 2001 г. â„– 1598)]. 126 Republic of Kazakhstan (2006) Strategy of territorial development of Republic of Kazakhstan till 2015. [СтратегиÑ? территориального развитиÑ? РеÑ?публики КазахÑ?тан до 2015 года (Указ Президента РеÑ?публики КазахÑ?тан от 28 авгуÑ?та 2006 года â„– 167)]. 127 Government of Kazakhstan (2019) Forecast Scheme for Territorial Development until 2030. [In Russian.] 128 A functional urban area consists of a city and its commuting zone. It therefore includes a densely inhabited city and a less densely populated commuting zone whose labor market is highly integrated with the city. This definition, originally proposed by the Organisation for Economic Co-operation and Development, has been recognized and adopted by the European Union and the World Bank. See https://ec.europa.eu/eurostat/statistics- explained/index.php/Glossary:Functional_urban_area. 129 Data provided by Ministry of National Economy. 130 This information is based on interviews with representatives of akimats of Almaty, Astana, Shymkent, and Temirtau. 131 Republic of Kazakhstan (2014) State program for support and development of business “Road map for business— 2020.â€? 132 ПоÑ?тановление ПравительÑ?тва РеÑ?публики КазахÑ?тан от 11 октÑ?брÑ? 2013 года â„– 1092 «Об утверждении Концепции формированиÑ? перÑ?пективных национальных клаÑ?теров РеÑ?публики КазахÑ?тан до 2020 года». 133 Republic of Kazakhstan (2019) State program for Industrial and Innovative Development for 2015–2019. 134 The Forecast Scheme is not referenced in the Program for Industrial and Innovative Development, and neither the Forecast Scheme nor the SPRD mentions the six cluster programs (three of which are in key agglomerations prioritized by the Forecast Scheme and SPRD). 135 See https://primeminister.kz/ru/documents/gosprograms. 136 Republic of Kazakhstan (2019) State program for Territorial Development 2020–2025. Republic of Kazakhstan (2014) State program for Infrastructure Development Nurli-Zhor. 2020-2025. Republic of Kazakhstan (2018); State program for Housing Construction Nurli-Zher 2017–2021. 137 Republic of Kazakhstan (2019) State program for Territorial Development 2020–2025; Republic of Kazakhstan (2014) State program for support and development of business “Road map for business—2020.â€? 138 Republic of Kazakhstan (2015) The rules of consideration and selection of ring fenced transfers for development. [Правила раÑ?Ñ?мотрениÑ? и отбора целевых транÑ?фертов на развитие, Утверждены приказом МиниÑ?тра финанÑ?ов РеÑ?публики КазахÑ?тан от 25 февралÑ? 2015 года â„– 126]. 139 Law of Kazakhstan «About introducing changes to the selected legal acts related to the issues of procedures of budget and tax legislation» December 27, 2019 No. 290-VI. [«О внеÑ?ении изменений и дополнений в некоторые законодательные акты РеÑ?публики КазахÑ?тан по вопроÑ?ам Ñ?овершенÑ?твованиÑ? процедур реабилитации и банкротÑ?тва, бюджетного, налогового законодательÑ?тва, законодательÑ?тва о железнодорожном транÑ?порте» от 27 декабрÑ? 2019 года â„– 290-VI.] 140 Data provided by the akimat of Temirtau. 141 Following the example of the US Economic Development Administration, 142 Montreal was designated a growth pole and could benefit from subsidies and incentives. 143 Ð?наÑ?таÑ?иева, Е., Ð?нциперова, Ðœ., КраÑ?юк, Е., Мальцев, Г., РумÑ?нцев, Ð?, Серажетдинов, Ф., СтолÑ?ров, Б., Чубуков, Ðœ., Швец, Ð’. 2018. Эпоха агломераций. ГородÑ?каÑ? Ñ?кономика, проÑ?транÑ?тво и политика в новом маÑ?штабе. 1st ed. МоÑ?ква: Ð?льпина нон-фикшн, pp. 341–345. 144 See https://www.diafan.ru. 145 This chapter was prepared by a team comprising Olena Bogdan, Douglas Zhihua Zeng, and Asset Bizhan. 146 Often referred to as an industrial park. 147 Law of the Republic of Kazakhstan on Special Economic Zones and Industrial Zones, No. 242-VI, April 2, 2019. 148 Law of the Republic of Kazakhstan on Special Economic Zones and Industrial Zones, No. 242-VI, April 2, 2019. 149 Law of the Republic of Kazakhstan on Special Economic Zones and Industrial Zones, No. 242-VI, April 2, 2019. 150 Law of the Republic of Kazakhstan on Special Economic Zones and Industrial Zones, No. 242-VI, April 2, 2019. 151 Order of the Minister of Industry and Infrastructural Development of the Republic of Kazakhstan, July 30, 2019, No. 571. 145 152 Order of the Minister of Industry and Infrastructural Development of the Republic of Kazakhstan, July 30, 2019, No. 571. 153 Industrial zones in the country can be divided into three categories depending on the stage of development: publicly announced/planned or under development, functioning, and desolate. Several regional IZs are planned, where local authorities have not completed a zone’s feasibility study or suspended infrastructure construction. In addition, subnational authorities in several regions as well as private entities have initiated or publicly announced several private and small IZs (referred to as industrial parks, often designed as part of a large industrial factory), yet closer study revealed that these small-scale IZs are not functioning. Because QazIndustry does not keep statistics on planned IZs, this chapter does not cover them. Thus this section covers the 22 IZs that QazIndustry provided information on and an additional IZ with publicly available data, Almaty IZ. 154 Mashinostroiteley Street and Orken KShT IZs in East Kazakhstan region, and Tulkibas and Baidibek IZs in Turkistan region. 155 Shymkent was the administrative center of South Kazakhstan region before 2018. On June 19, 2018, Shymkent became a city of republican significance—that is, a separate subnational unit—like Almaty city and the capital, Astana. Consequently, the administrative center of the region was relocated to Turkistan city, and the region was renamed Turkistan region. 156 The current concept of territorial clusters was developed under the World Bank-financed project titled Enhancing SME Competitiveness. The cluster development activities under the project were carried out in 2016–19 and resulted in six cluster development plans and capacity building of the QazIndustry national development institute and of cluster members. 157 See Government Decree No. 1050, December 31, 2019, on Approval of the State Program for Industrial and Innovative Development in 2020-2025, Article 5 “Main directions, ways to achieve the set goals of the Program and corresponding measures.â€? The intention is to support the six pilot clusters at the teething stage and gradually limit state-support measures as clusters become self-sufficient. 158 This includes manufacturers of final or intermediate goods and services, suppliers of components and specialized services, manufacturers of industrial and other equipment, suppliers of specialized infrastructure, scientific and research organizations, higher education organizations, and technical and vocational education and other organizations having a specific industry specialization. Entrepreneurship Code, adopted through Law No. 375-V 3PK, October 29, 2015. 159 In 2018, the entire country was transformed into a SEZ through the New Investment Support Act. The new law intends to expand the area, offering tax incentives and other benefits from the current 25,000 hectares, or 0.08 percent of Polish territory, to almost 100 percent of Poland’s investment space. This is meant to level the playing field for small and medium enterprises and other companies that would no longer have to move their operations to specific locations to benefit from incentives (UNCTAD 2019). 160 See the Jebel Ali Free Zone Authority website (http://www.jafza.ae/about-us/). 161 The 2016 policy review section of the Methodology for Implementation of Cluster Policy (available in Russian at https://qazindustry.gov.kz/docs/8957967.pdf), developed by QazIndustry, briefly mentions that an analysis of the seven clusters revealed that the program did not meet its initial expectations, because government support was limited to subsidies, there was no active participation by the private sector, participating companies lacked motivation to cooperate, and supply of and demand for local innovative products, exchange of innovations, and creation of new knowledge were low. 162 National clusters were defined as a new stage of cluster policy, shifting the focus from developing industrial clusters by creating value chains in traditional sectors of the economy to developing innovation clusters based on key competencies, transfer of knowledge and technology, and innovative entrepreneurship. See the Concept for Formation of Perspective National Clusters in Republic of Kazakhstan till 2020, approved by Government Decree No. 1092, October 11, 2013 (available in Russian at http://adilet.zan.kz/rus/docs/P1300001092). 146