Report No. 52037-PL Poland Mazowieckie Public Expenditure Review Local Responses to the Global Economic Crisis April 29, 2010 Poverty Reduction and Economic Management Unit Central Europe and Baltic Countries Unit Europe and Central Asia Region Document of the World Bank ÿþCURRENCY EQUIVALENTS Currency Unit PLN 1=US$0.34 WEIGHTS AND MEASURES The Metric System is used throughout the report. FISCAL YEAR January I to December 31 ABBREVIATIONS AND ACRONYMS BGK State Bank of Poland CDD City Development Department (Warsaw) CIT Corporate income tax DFW Debt Forecast of the City of Warsaw EC European Commission ERDF European Regional Development Fund ESF European Social Fund EU European Union GDDKiA General Directorate of Road and Motorways GPS Global Positioning System GRSP Global Road Safety Partnership GRP Gross Regional Product ISPA Instrument for Structural Policies for Pre-Accession KM Regional passenger rail company (Voivodship) MDF Mazowieckie Development Fund MJWPU Mazowieckie EU Funds Implementation Unit MTBF Medium Term Budget Framework NFZ National health insurance fund OECD Organization for Economic Cooperation and Development PIT Personal income tax PLK Polish Railway Lines PPP Public Private Partnership ROP Regional Operating Program ZTM Warsaw transport management agency Vice President : Philippe H. Le Houerou Country Director Peter C. Harrold Sector Director Luca Barbone Sector Leader Bernard Funck Task Team Leader William Dillinger/Emilia Skrok FOREWORD This report is based on the findings of a series of missions to Poland from December 2008 through June 2009 and subsequent data obtained through December 2009. The report was prepared by William Dillinger (ECSP4) with the assistance of Emilia Skrok (initial Task Team Leader, ECSP2), Ewa Korczyc (ECSP2), Radoslaw Czapski (ECSS5), Jukka-Pekka Strand (ECSSD), Jung Eun Oh (AFTTR), Marc Schiffbauer (ECSP2), Lili Liu (PRMED), Ngoc-Bich Tran (LCSPE), Ying Li (PRMED), and Ken Gwilliam (consultant). The peer reviewers were Andreas Kopp, ETWTR and Ahmad Ehtisham, London School of Economics. The report was prepared in conjunction with a transport policy note and a policy note on property asset management in Warsaw. Both are issued separately. The report was produced with the participation of a wide range of Polish counterparts. These included officials and staff of the Voivodship of Mazowieckie, the City of Warsaw, and the cities of Plock, Radom, and Siedlce.  Table of Contents Executive Summary ...................................................... i Introduction.. .......1 Regional Context ..................................................... 1 Economic Context.. ......6 Institutional Context 7............7 Warsaw ............................................................... 10 Expenditures ........................................................ 10 Revenues .......11 Fiscal Prospects........................................................ 13 Rationalizing the School Network .................................... ..... 22 Cutting Transport Subsidies..................................... ......... 23 Targeting Capital Investments ........................................... 26 The Voivodship ....................................................... 28 Expenditures ........................................................ 28 Revenues .......................................................... 31 Fiscal Prospects..................................................... 31 Restraining Spending ......................................... ......... 34 Targeting Capital Investments ........................................... 35 Siedlce, Plock, and Radom .......................................... 39 Revenues .......................................................... 39 Expenditures ........................................................ 39 Fiscal Prospects..................................................... 41 Long Term Reforms .................................................... 42  Executive Summary 1. Regional Context Mazowieckie is the outstanding example and greatest beneficiary of Poland's successful economic transition. Much of the prosperity of the Mazowieckie region, in turn, is concentrated in Warsaw. Economic disparities among the subregions of the Voivodship have sparked concerns in some quarters. The regional development strategy for Mazowieckie notes the gap between economic conditions in different parts of the voivodship and calls for a strategy aimed, inter alia, at improving conditions in peripheral areas. But it wisely makes no attempt to equalize levels of economic activity among the subregions of the Voivodship. 2. Subregional variations in household income-which are the proper concern of regional policy-do not vary as widely within Mazowieckie as per capita GRP. Moreover, there is evidence that low levels of household income in outlying subregions largely reflect the characteristics of the people who live there, rather than the inherent characteristics of the subregions themselves. Under these circumstances, efforts to bring higher-wage employment to poor regions may have little benefit for the resident population. This does not mean consigning peripheral areas to perpetual decline. There are spatially neutral measures that can lead to economic growth. In addressing economic conditions in peripheral areas, the regional development strategy puts particular emphasis on improving transportation facilities to facilitate access to the Warsaw market. In addition to its economic benefits to the City, this would permit peripheral areas to take advantage of the large consumer market and wide range of suppliers of inputs and business services available in the metropolitan area. 3. Fiscal Outlook While Poland is weathering the current global economic downturn fairly well, the major subnational governments of Mazowieckiel--the Voivodship and the City of Warsaw--face considerable fiscal uncertainties. Several years of steady growth in revenues have come to an end. The outlook for an immediate recovery in revenues is uncertain. The two jurisdictions' immediate response has been to slash capital spending and engage in large scale borrowing. But this is not sustainable. Key capital investments cannot be postponed indefinitely. And excessive borrowing could ultimately lead to insolvency. 4. Neither the City nor the Voivodship is in a position to respond by increasing revenues. Both derive the majority of their revenues from fixed shares of the centrally administered personal income (PIT) and corporate income (CIT) taxes. Locally-administered tax sources are limited and subject to central government caps. Because the PIT and CIT are sensitive to economic fluctuations, both jurisdictions are vulnerable to slower-than-expected rates of recovery in the economy. 5. This report therefore focuses on finding solutions on the expenditure side. It focuses on three issues in particular. The first is subsidies to public transport. At present, subsidies are equal to nearly half the operating costs of the two jurisdictions' transport companies. While some Mazowieckie is one of the sixteen voivodships into which Poland is divided. Its territory includes the City of Warsaw, Poland's economic core and the seat of the national government. i degree of subsidy can be justified on environmental and equity grounds, the present level appears to be excessive. To reduce the need for subsidies, the two jurisdictions should consider selective tariff increases. They should also introduce competition into public transit services in order to drive down costs. 6. The second concerns the process of investment planning, evaluation, and funding. Improvement in this area would allow the City and the Voivodship to reduce capital spending while still providing funds to high priority projects. Both jurisdictions now have ambitious investment plans. But increasingly scarce resources do not appear to be reaching the highest priority projects. Both jurisdictions ostensibly base their capital programs on sectoral and (in the case of the Voivodship) regional spatial plans. Both ostensibly employ systems of medium term capital budget planning to ensure reliable funding for individual projects. In practice, the connection between the strategies and actual capital investment funding decisions appears to be tenuous, at best. And although annual funding decisions are consistent with medium term capital investment plans, this is only because the medium term plans are adjusted retroactively to reflect annual changes in priorities. There are several measures the City (as well as the Voivodship) could take to address these shortcomings. Project identification could be improved by strengthening the data bases on which investment decisions are made. Project evaluation methodologies could also be improved, for example, by adopting a _lig%t' version of the framework used to evaluate major EU-funded projects. The medium term capital budget should continue to serve as a guide, rather than a straight jacket, on annual funding decisions. But this flexibility should be used to allow for the vagaries of project execution, not to accommodate short term changes in political priorities. 7. The third area--spending on education-applies particularly to the City. Warsaw spends nearly ten percent of its budget on subsidies to education-over and above the financing provided by the central government education subvention. This is in part because the City pays substantially higher wages than the education subvention is designed to finance. But it also reflects the City's failure to respond to declining enrollment rates. While Warsaw's student population has decreased by ten percent over the last five years, the number of teachers has not declined proportionately. The result is very low pupil-teacher ratios in some districts of Warsaw. To address this problem, the City should conduct an inventory of pupil-teacher ratios by class in each school and then devise a scheme for consolidating classrooms, reassigning pupils, and reducing teaching staff. 8. The Voivodship, for its part, would be wise, particularly in the current fiscal climate, to exploit EU structural funds. The Voivodship has repeatedly budgeted for large inflows from the EU, but then failed to see them disburse. Part of explanation for the slow disbursement of EU funds is a human and institutional one. Local governments lack the capacity to prepare proposals and payment applications. This problem could be addressed through increased training and field visits by Voivodship staff. In the past, project identification and evaluation have also been a problem. The process used to allocate funds under the 2004-07 program was subject to too little technical evaluation and too much political influence. The Voivodship has now reformed the process, using a roster of technical experts to undertake technical evaluations. Whether this reform will improve the quality of evaluations without slowing down the pace of disbursements will soon be seen. ii 9. Three smaller cities--Siedlce, Plock, Radom-were briefly analyzed for this report. All appear to be subject to the same revenue-side vulnerabilities as Warsaw and the Voivodship. Like Warsaw, they are largely dependent on fixed shares of the centrally administered PIT and have little ability to raise revenues independently. Moreover, they have less flexibility than Warsaw on the expenditure side, due to the relatively small proportion of spending they devote to capital works and the corresponding large share they devote to salaries. Under these circumstances, it behooves the three cities to monitor changes in PIT revenues closely and be prepared to cut the expenses that they can control at the first sign of a downturn. 10. Over the long term, the Government of Poland might consider more fundamental structural solutions to the fiscal vulnerabilities of its local governments, either by increasing local autonomy over revenue generation or by removing centrally imposed impediments to the management of personnel costs. 111  Introduction 1. This report examines the fiscal prospects of major units of subnational government in the voivodship (province) of Mazowieckie and proposes a series of reforms in recurrent spending and capital investment planning. Mazowieckie, the seat of Warsaw, is the heart of the Polish economy. While Poland is weathering the current global economic downturn fairly well, the subnational governments of Mazowieckie face considerable fiscal uncertainties. A long period of rising revenues has come to an end. At the same time, both the province and the City of Warsaw are embarked upon major programs of capital investment. While EU funds could help finance some of these investments, there is evidence that the money from Brussels is not reaching the highest priority projects. The objective of this report is to examine how the major units of subnational government can respond - how they can restrain the growth of deficits and allocate funds more wisely. The report is organized in four sections. The first describes the regional, macroeconomic, and institutional context of the subnational governments. The following three sections examine the fiscal situation of Warsaw; the Voivodship, and three _case study' cities. Regional Context 2. Mazowieckie-and more specifically greater Warsaw-is the outstanding example and greatest beneficiary of Poland's successful economic transition. Poland's GDP grew at an average annual rate of 5.1 percent between 2003 and 2008 and achieved rates of 6.2 percent and Figure 1. Variations in GDP per Capita among 6.8 percent in 2006 and 2007 Voivodships respectively. This was largely on the strength of growth in services GDP per inhabitant, in PPS, byNTS2egions,average20M (particularly business services) and IpreofEU27=1 labor intensive manufacturing. Much S<= so of this growth has been concentrated owiec Ms0-=75 in the voivodship of Mazowieckie. 75 - <= 100 M v d a Il 100-<= 125 The Mazowieckie voivodship as a m > 125 whole has enjoyed a wide and Data notavailable sustained lead over Poland's other voivodships as measured by GDP per 2 capita. In 2002, Mazowieckie's per capita GDP was roughly 75 percent above the (unweighted) average of all Source: Eurostat other voivodships. By 2006, that margin had increased to 83 percent. See Figure 1. 3. Much of the prosperity of the Mazowieckie region, in turn, is concentrated in Warsaw. The statistical bulletin of Mazowieckie Voivodship, along with Eurostat and OECD, divides the voivodship into five subregions: the city of Warsaw (Miesto Warsaw), the surrounding Warsaw region, and the three outlying subregions of Ciechanowsko-plocki (to the west), Radomski (south) and Ostrolecko/Siedleckie (east). As shown in Figure 2, the per capita GDP of the city of 2 Adjusted for purchasing power. Purchasing-power adjusted GDP per capita is the key variable for determining the eligibility of NUTS 2 regions in the framework of the European Union's structural policies. 1 Warsaw outranks that of the other four regions by a wide margin. Its per capita GDP-roughly PLN 21,600 in 2006-is over three times that of the (unweighted) average of the other four regions. 4. The Voivodship's regional development strategy is outlined in the Development Strategy for the Mazowieckie Voivodship for the Year 2020 (updated) and its companion Spatial Development Plan. Both were issued Figure 2. Trends in GDP/capita among Subregions in 2006. The Development Strategy within Mazowieckie Voivodship begins with a brief overview of the 40,000 economic challenges facing the 35,000 -4- PL121:Ciechanowsko- voivodship as a whole and the plocki weaknesses of various infrastructure 30,000. --- PL122: Ostrolecko- services. It goes on to note that there 25,000 siedlecki are wide regional disparities within 20,000 the voivodship, that peripheral 15,000 -- PL126: Warszawski subregions do not attract foreign 10,000 direct investment because of poor 5,000 - - PL127: Miasto Warszawa transport access and the absence of a E________,_ local markets but notes the role of 2000 2001 2002 2003 2004 2005 Warsaw as growth pole within region. The Strategy then proceeds to Source: Eurostat diagnose the causes of these disparities. For this purpose, it divides the voivodship into seven subregions, with Warsaw metropolitan area at the core, a surrounding area of -positive influence of the agglomeration', and five outlying -ploblem areas.' 3 As shown in Figure 3 the metropolitan areas and the zone of positive influence of the Warsaw (shown in the two lighter shades of pink) extend as far as 75 km from the city center. 5. According to the Development Plan, the difficulties of each problem area vary. The Radom subregion is said to suffer from _high structural unemployment, low entrepreneurship, and poor road infrastructure'; Ostrolecki, from poor soils (as an impediment to agriculture) and weak transport links to Warsaw. The Nadbuzanski subregion has good soils. Agriculture is, however, impeded by the dominance uneconomically-small farms, run by elderly people whose children have moved off to Warsaw in search of better prospects. The structure of agriculture landownership in the Mlawsko-zurominski subregion, in contrast, is said to be favorable, particularly for cattle-raising. The impediments to growth in this subregion are said to be the low quality of agricultural production, manufacturing, high unemployment and low levels of the private services sector'. The Plock subregion benefits from the presence of good colleges and a thriving petrochemicals industry, but suffers from poor transport links to Warsaw. 6. The remainder of the Strategy largely consists of an abstract discussion of objectives and actions, beginning with a _vision' (Maowieckie as a competitive region in the European and global perspective'), a mission statement, and three primary strategic objectives: (1) shaping the information society and improving the quality of life of the inhabitants of the voivodship; (2) 3 The plan's seven subregions roughly correspond to Eurostat's five-subregion division of Mazowieckie, with the Ciechanowsko-Plocki subregion divided into the Plocki and Mlawa/Zurominski subregions, and the Ostroleka/Sied1ce divided into Ostroleckie and Nadbuzanski. 2 strengthening the competitiveness of the region in the international dimension; and (3) improving social economic and spatial cohesion within the framework of sustainable development. (The last of these includes _modernization and expansion of the communication networks as well as improvement in the living conditions of the inhabitants of the extra-metropolitan areas'). This is followed by a list of five secondary objectives: (1) developing social capital; (2) increasing innovation and competitiveness (3) stimulating the development of Warsaw's metropolitan functions (4) activating and modernizing the extra-metropolitan areas, and (5) developing civil society and shaping the image of the region. Each secondary objective is divided into several ations'. _I1veloping social capital', for example, is divided into seven actions: (1.1) developing the information society and ICT infrastructure; (1.2) increasing the level of education and quality of the labor force; (1.3) combating unemployment; (1.4) improving standards of health care; (1.5) striving for improved housing conditions; (1.6) solving social problems; and (1.7) improving public security. The strategy concludes with a discussion of funding sources. Figure 3. Mazowieckie Regional 7. As such, the Strategy is quite Development Spatial Strategy abstract. It provides little guidance as to the location or nature of spending priorities. (On the contrary, it seems to be aimed largely at meeting the documentation requirements for EU regional funding.) The Strategy is, nevertheless, successful in avoiding the major pitfalls of spatially-oriented development strategies. It does not call for a freeze on the growth of Warsaw or special incentives to subsidize economic activity in outlying areas. Instead, it recognizes the role of Warsaw as the economic engine of the entire voivodship.4 To the extent there is any specific support for peripheral areas, it is in the area of transport investment, where the strategy calls for construction of ring roads and axial lines to better connect cities in the peripheral areas with Warsaw and with each other. 8. Most importantly, it makes no attempt to equalize levels of economic Source: Mazowieckie Region Development Strategy: activity among the subregions of the Voivodship. While the regional development strategy for Mazowieckie notes the gap between economic conditions in different parts of the voivodship and calls for a strategy aimed, inter alia, at improving conditions in peripheral areas, it wisely makes no attempt to equalize levels of economic activity among the subregions. 4 This would be consistent with recent geographic patterns of growth in Western Europe, where economic growth, as measured at the national level, has been accompanied by growing economic concentration, despite the efforts of EU regional funds. According to EU data, this has produced growing geographical economic disparities within member states, even while economic disparities among member states have declined. 3 9. There are several reasons not to do so. First, it is not obvious that household income- which is the proper concern of regional policy-varies as widely as per capita GRP. While there is little doubt that per capita output levels in Warsaw are considerably higher than in other parts of the voivodship, it is not so clear that living standards vary as widely. Regional variations in output per capita are not necessarily correlated with regional variations in living standards. Box 1 explains why. Box 1 Why Variations in GRP Per Capita Do Not Always Reflect Variations in Living Standards By definition, GRP, as calculated on an output basis, should be equal to GRP calculated on income basis. This might suggest that the two should be correlated-that income generated by regional production units would be reflected in the income of households within the region. But this relationship applies only to the proportion of GRP that constitutes returns to labor-compensation to employees and the incomes of self-employed persons. GRP also includes the gross operating surpluses of firms. Although this may be paid out in stock dividends or interest payments (or held as retained earnings) it does not necessarily accrue to households in the region. GRP also includes taxes, which can consume a significant share of the gross operating surpluses generated by production units. As the majority of taxes are collected and spent by central governments, they are unlikely to contribute directly to the income of households in the region. Household income, on the other hand, includes not only returns to labor but also income from transfer payments. These include social protection payments such as pensions and unemployment benefits-both of which are important sources of income in Poland. Regional variations in prices may also undermine the correlation. Regional variations in the cost of living are not reflected in GRP calculations but are reflected in the data sets used to measure household expenditure and poverty rates. Commuting could also distort the relationship between GRP per capita and poverty. In calculating GRPs, production is recorded at the place of work. Household income statistics, on the other hand, are based on the place of residence. Thus regions with high inflows of productive commuters might have high GRPs but relatively low levels of per capita household income. 10. Detailed household income data for the subregions of Mazowieckie would allow this assertion to be empirically tested. In the absence of such data, there is some preliminary statistical evidence, for Poland as a whole, that wide geographical discrepancies in GDP per capita are not reflected in equally wide discrepancies in living conditions. The first is the weak correlation between per capita GDP and household income or poverty rates at a more aggregate level--the 16 voivodships. As shown in Figure 4, poverty rates at the voivodship level have very little correlation (r=-.42) with average rates of poverty. 11. It is also suggested by data on variations in salaries within Mazowieckie, shown in Table 1. Although these reflect only the income of wage earners (as opposed to the self employed) in the formal economy, they suggest that people who manage to obtain formal wage paying jobs do not experience the scale of geographical income discrepancies that the GDP data would suggest. Site visits to the three major outlying cities in the voivodship tend to bear this out. While all three cities (Siedlce, Radom, and Plock) have faced job losses due to the closure or privatization of state enterprises, they all claim to be recovering-in some cases on the strength of a revival of The likelihood of obtaining such a job is indicated in the figures for unemployment, which show considerably higher rates in the outlying regions than in the city of Warsaw. 4 large scale manufacturing but more often through the growth of small scale firms in the manufacturing and services sectors. 12. There are three other arguments against attempting to equalize levels of economic activity among the subregions. First, as a strategy for improving living standards, a focus on geographical equalization may miss the target. Figue 4odrratonertwong DAlthough geographical concentrations of poverty andiModerae Pmay exist in Mazowieckie, poor people Voivodships presumably also live in Warsaw. Again, data on 36% variations in household income within 34% Mazowieckie would permit this calculation to be 32% made on a sub-regional level. In the absence of 30% such data, it is worth noting that Poland's poorest 28% voivodship (Lubelski) accounts for only three 26% Mazowieckie percent of Poland's poor--less than a third of the 26% nme nube of poor residing in its richest, 24% Mazowieckie. Focusing poverty reduction efforts 22% on the poorest region would therefore appear to 20% bypass a large proportion of the target. 2000 4000 6000 8000 10000 13. Second, efforts directed at regional Source: Eurostat, World Bank staff calculations: disparities in economic activity may misdiagnose the sources of the problem. The available data suggest that poverty is, at best, only partly related to where people are. It is also strongly related to who people are. While it can be argued that regional concentrations of poverty reflect the particular disadvantages of certain regions-remoteness, absence of infrastructure, limited local consumer demand-it can also be argued that regional poverty is a reflection of the characteristics of the people who live there: low levels of education or age profiles that render them too young or too old to participate in the labor force. Under the latter circumstances, efforts to bring higher-wage employment to poor regions may have little benefit for the resident population. If people characteristics, rather than place characteristics, are the overriding explanation for regional concentrations of poverty, it suggests that efforts to improve living standards should focus on improving the educational level of the labor force and on providing income support to households with large numbers of unemployable dependents, rather than on attracting new sources of high wage employment for which the resident population may not be qualified. 14. Third, relying on regional economic development to address regional concentrations of low living standards ignores the potential role of migration. In areas with few economic prospects, the out-migration of labor-rather than the in-migration of capital-is likely to be a more effective means of improving the living standards of individual members of the labor force. A regional policy that allows for a decline in the population of some lagging areas is therefore entirely consistent with the objective of improving living standards. 15. Finally, and most importantly, there is some risk that an attempt to equalize levels of economic output raises the risk of killing the goose that lays the golden egg. In a sense, the most important economic priority for Mazowieckie is to make sure that Warsaw continues to thrive- that it is competitive and remains competitive, continuing to provide good jobs both for existing 5 residents and for potential migrants from the rest of the voivodship. Which among other things means making sure that Warsaw works, from an infrastructure standpoint within its current territorial limits, and that infrastructure expands to accommodate the growth of the metropolitan area. 16. This does not mean consigning peripheral areas to perpetual decline. There are spatially neutral measures that can lead to economic growth. These include measures to improve local business climates, such as one-stop business registration. They include improvements in transportation facilities to facilitate access to the Warsaw market, permitting peripheral areas to take advantage of the large consumer market and suppliers of inputs and business services in the metropolitan area. But it has to be recognized that good regional policies may imply a decline in some sectors that now account for a significant share of employment in peripheral areas (e.g., agriculture). And they may imply the out-migration of labor. Improving the quality of education and training, for example, may make peripheral areas more attractive to potential investors, but it may also equip young people in problem areas with the skills required to obtain better jobs in Warsaw. Table 1. Characteristics of Subregions with Mazowieckie total % of total % A PL121: Ciechanowsko-plocki 626,807 12% -1% 7500 17.6 2247 PL122: Ostrolecko-Siedleckie 750,445 15% -1% 5200 24.2 2257 PL124: Radomski 730,955 14% -1% 5300 11.9 2684 PL126: Warszawski 1,356,642 26% 5% 7050 4.6 3790 PL127: Miasto Warszawa 1,699,860 33% 0% 21600 19.8 2552 Total 5,166,716 100% -1% 11400 17.6 2247 *in PLN, 2006 ** 2006 gross zl/month, Source: Population from OECD; GDP per capita from Eurostat, registered unemployed and average salary (both 2006) from Mazowieckie Statistical Bulletin. GDP data for the Warszawski subregion is the average of the W-wschodni and W- zachodni into which Eurostat divides the subregion. Economic Context 17. After five years of strong economic performance, the global financial crisis has dimmed Poland's macroeconomic and fiscal outlook, although Poland is faring better than other countries in the region. Relatively strong economic fundamentals and a strong financial system as well as the Government's timely response to the crisis have helped Poland avoid a deep recession. However, since late 2008 Poland has been adversely affected by reduced demand for its exports by high income countries and by reduced capital inflows. Growth in the first half of 2009 fell to one percent and was expected to remain in the range of 1.0 - 1.3 percent for the year as a whole. 18. This slower growth has resulted in a deterioration of the Government's fiscal position. According to preliminary data, the general government deficit in 2009 will be around 7.2 percent of GDP, compared to the 2.5 deficit percent originally planned for 2009. According to the new Convergence Program , the deficit will stay close to 7 percent of GDP in 2010, reflecting the impact of rising capital investment and still high social transfers. Higher deficits will lead to a rising public debt to GDP ratio, approaching the EU threshold. Overall, the fiscal envelope for 6 public spending is tightening. Under Poland's highly decentralized structure of government, adjustments will be required at both the central and subnational levels. Institutional Context 19. Poland has a three-tier system of subnational government. The national territory is divided into 16 provinces (voivodships). These are divided into 314 districts (powiats) which are in turn divided into 2,478 municipalities (gminas). The 65 more-urbanized gminas have the concurrent status of a gmina and a powiat. Each of the three tiers has an elected executive and council. 20. Gminas. The basic unit of local government-and the only one cited directly in the Constitution--is the gmina. According to the Constitution, the gminas are the foundation of the self-government system. As a result, gminas are responsible for most of public services performed by subnational governments. These include: * education (kindergartens, elementary schools and lower-level high schools), * local public transportation, * maintenance and lighting of local roads, * maintenance of urban green areas, burial grounds, libraries and community centers; * provision of water supply, sewerage, refuse collection and disposal, and distribution of electric power and gas. In addition, gminas play a more limited role in the provision of primary health care, social assistance (each gmina is required to operate a Gmina Center for Social Assistance) and social housing. 21. Powiats. The range of powiats' responsibilities is less clear-cut. While they are explicitly responsible for certain tasks (constructing and maintaining powiat roads, operating upper-level high schools and vocational schools and certain general hospitals and cultural institutions) they also perform tasks on behalf of the central government, mainly in the area of social assistance. The heads of the powiats (starosts) are also in charge of managing State property. 22. Voivodships. In contrast to gminas and powiats, the principal task of voivodships is not the direct provision of services to citizens. Instead it is to plan and coordinate the development of its territory. This entails drafting development strategies and managing EU regional development funds. Voivodships are also charged with coordinating the activities of other self-governments in such areas as environmental protection, transportation, spatial planning and rural development. Nevertheless, voivodships are directly responsible for the maintenance of regional roads, financing regional railroads, running certain cultural institutions, and protecting historical monuments. Voivodships also organize some health service facilities, including some hospitals. 23. Expenditures. As shown in Figure 5, gminas (including gminas with powiat status) accounted for nearly eighty percent of total subnational expenditure in 2008. (Of this total, 35 percent was spent by gminas with powiat status and 43 percent by other gminas.) Powiats accounted for about 12.5 percent of total subnational spending and voivodships the remaining nine percent. 7 24. Revenues. Subnational governments in Poland derive the vast majority of their revenues from the central government, either in the form of shared taxes or as transfers aimed at particular functions or at revenue equalization. While all of these are formula based, some formulas are more stable than others. Figure 5. Distribution of Subnational Spending, by The largest source consists of shares of Tier, 2005-2008, percent the centrally administered personal and 0 Cities with powiat status corporate income taxes. These shares 0 Powiats are distributed on the basis of origin; m Gminas excl. cities with powiat status a Voivodships i.e., according to where they were 100% -collected. The percentage of each 90% 80% shared tax that is assigned to each tier 70% of subnational government is adjusted 60% --1. 30 1. 25slightly each year (and more 40% - substantially when a specific function is 30% -decentralized). As shown in Table 2, as 20% of 2009, gminas are entitled to nearly 10%l 40 percent of the PIT collected within their jurisdictions; powiats about ten percent; and voivodships only 1.6 Source: Central Statistical Office, World Bank staff calculations. percent. 25. The second most important source is a set of grants under the general rubric of the =general grant'. It has four components. The Tabe 2. Disribtio G of hre'axsAmn largest is the educational subvention. This is ffintended to compensate subnational giia 'v governments for the costs of providing the Share of PIT 39.34% 10.25% 1.60% tlevel of education assigned to their respective Share of CIT 6.71% 1.40% 14.00% tiers. The level of the education subvention is Source: Ministry ofFinance, Law on Revenues of Local not fixed by formula but is instead set Government Units, annually in the budget law. The annual budget law in turn, does not lay down a specific process for allocating the subvention among individual jurisdictions, but instead delegates this role to an ordinance issued by the Ministry of Education. Although the actual distribution algorithm varies from year to year, the bulk of the education subvention is allocated on the basis of enrollment, with higher weights for pupils in rural areas and small towns, pupils with disabilities, schools in which classes are taught in minority languages, and vocational schools. 26. The second component of the general grant is the so called equalization subvention. This is allocated to all subnational governments whose per capita revenues from shared taxes (i.e. the PIT and CIT) and local taxes (in the case of gminas) is lower than the national average. The formula does not attempt complete equalization. If a gmina's per capita revenues are between 40 percent and 75 percent of the national average, the grant (per capita) is only equal to 75 percent of the difference between the gmina's per capita revenues and 83 percent of the national average. If the gap is bigger, the proportion of it that is filled by the grant is proportionately larger. The 6 In calculating revenues from own taxes--where gminas have some control over the rate--the subvention formula assumes that the maximum rate has been applied. 8 formula for higher tiers of subnational government is simpler: all powiats with per capita revenues less than the national average receive a transfer equal to 90 percent of the difference. All voivodships with per capita revenues less than the national average receive a transfer equal to 72 percent of the difference. 27. The centrally financed equalization subvention is supplemented by a fraternal equalization mechanism funded from mandatory contributions from richer jurisdictions. Gminas with per capita revenues (again, defined as local taxes and shared taxes) over 150 percent of the national average are required to contribute 30 percent of the excess to the equalization mechanism. As shown in Table 3, richer powiats and voivodships are required to make substantially greater contributions. Powiats with per capita revenues above 123 percent of the national average must contribute 98 percent of the excess to the fraternal equalization mechanism. The criteria used to distribute these grants are defined in the law on the incomes of local government. Table 3. Mandatory Contributions to Fraternal Equalization Mechanism Tier of Jurisdictions pays Of the amount over X percent Government of the national per capita average, where X= gmmnas 20 % 150% 25% 200% 30% 300% powiats 80% 110% 95% 120% 98% 123% voivodships 80% 110% 95% 170% Source: Ministry of Finance, Law on Revenues of Local Government Units. 28. Subnational government units also derive a relatively small amount of funding from so- called dotacje. These are central government grants that are earmarked for specific purposes. They include transfers to finance centrally-mandated social assistance programs, such as the child allowance and unemployment benefits. 29. Locally administered taxes are reserved exclusively for the gmina tier. The property tax is the largest such source. It is imposed on buildings, land not subject to taxes on agriculture or forests, and _non-building assets'. The taxes on buildings and land are assessed on the basis of a fixed amount (in zlotys) per square meter. As such, both the base and rate are subject to a centrally-determined ceiling, which is adjusted annually for inflation. The current ceilings are extremely low, particularly for residential property.7 The residential part of the property tax therefore generates very little revenue. 2Non-building assets', for purposes of the property tax, include industrial plants and equipment. These are assessed on the basis of market value and therefore constitute a considerable source of revenue in gminas with large industrial facilities, including, of course, Warsaw. Like the residential property tax, however, industrial and commercial property is subject to centrally dictated caps. For 2010, for example, the maximum 7 In Warsaw, the rate is PLN 0.62 per square meter (equivalent to three US cents per square foot) for residential property. 9 ax on commercial property cannot exceed PLN 20.51 per square meter. The property transfer tax is a third significant source of revenue. This tax is imposed at the behest of the central government at centrally determined rates. Its yield, moreover, is sensitive to economic conditions. As an ad valorum tax, it is susceptible to declines in prices as well as to changes in the volume of transactions. 30. Subnational governments' reliance on centrally administered taxes and transfers has several virtues. It allows them access to broad based, buoyant, and relatively easily administered taxes, while preserving the central governments' control over the instruments of fiscal policy. It also permits the central government to redistribute tax money from poorer jurisdictions to richer ones. But it also has disadvantages. It renders subnational governments vulnerable to the discretionary decisions of the central government. Warsaw, for example, is now suffering the effects of the Government's decisions to dramatically reduce the rate of the personal income tax and introduce a child tax credit. The level of education subventions, similarly, is subject to annual adjustments.8 The fiscal condition of subnational governments largely depends on their ability to adjust to these changes in revenues. Warsaw 31. The city of Warsaw has the status of both a gmina and a powiat in Poland's three-tier system of subnational government. Accordingly, the City has a wide range of responsibilities including the operation of kindergartens, primary, and secondary schools (grades 1-12), the provision of public transit-including subways, trams, and bus companies--the construction and maintenance of gmina and powiat streets and boulevards, the provision of social assistance and cultural activities, the operation of water, sewer, and other utilities and the management of a considerable real estate portfolio. Expenditures 32. As shown in Figure 6 the composition of Warsaw's expenditure is correspondingly quite diverse.9 The largest single item of city expenditure is education, accounting for 20 percent of the total in 2008. Public transportation accounted for 18 percent of the total. Another ten percent of expenditure was allocated to roads. Property management (gospodarka mieszkaniowa) consumed a roughly equivalent share. (This figure includes expenditures arising from the city's ownership of industrial, commercial and residential real estate, much of which is revenue generating). Mandatory contributions to the fraternal equalization mechanism consumed seven percent. The remaining one-third of expenditure is comprised of a wide array of activities, including social assistance (9 percent), public administration (9 percent) and culture (5 percent). 8 Although the Local Revenue Act prohibits any reduction in the nominal level of the education subvention from one year to the next, it does not require that increases match the rate of inflation. 9 The graph reflects the accounting conventions used by the city administration. Expenditures on public transport and property management are therefore reported gross, rather than net, of revenues from ticket sales and rental income, respectively. Spending on education, health and social assistance includes spending financed from central government subventions. Transfers to other jurisdictions largely represent legally mandated contributions to the supplementary equalization mechanism. 10 Revenues 33. The City's largest source of revenue is its fixed share of the centrally administered PIT. This accounted for roughly one third of the city's receipts in 2008. Revenues from the city's share of the corporate income tax contributed about eight percent of the total. The two principal locally-administered taxes-the property tax and the property transfer tax-contributed 7 percent Figure 6. Warsaw: Composition of Expenditures by subvntioscntr ied cent; Fucton208,pecetearmarked grants for centrally- delegated functions; 3 percent. 1% a health care Revenues from fees-including the 2% 2% 2% a debt service ticket revenues of the public transit 4% menvironment companies and rents on municipally a s owned property--contributed 9 a culture percent of the total. So-called capital publicvenuesproceeds from the sale of 0 pubil admsn city property and EU capital grants-- Ssocial asstcontributed only 2 percent and 3 Etransfers to other juris. percent of the city's 2008 revenues, anproperty management respectively. 10 % roads 34. The City is just coming off a s public transport period of spectacular revenue growth. 0 education Figure 7 illustrates the growth of the eother major components of the city's revenues, in real terms since 2003. As Source: Central Statistical Office, World Bank staff calculations, shown, total revenues increased by about 85 percent in real terms over the five year period 2003-2007. This was led by growth in the personal income tax, which more than doubled in real terms over the five year period. Revenues from the corporate income tax nearly quadrupled (although the CIT's impact on the City's total revenues was relatively modest due to its relatively minor contribution to total revenues). Revenues from charges and fees increased by 50 percent. Expenditures, on the other hand, grew more slowly, increasing by 30 percent in real terms over the period. As shown in Figure 8, this permitted the city to run surpluses in 2006 and 2007. 35. This era is now over. In 2008, expenditures grew by 1I percent in real terms, while revenues stagnated. There is some evidence that this anemic growth in revenues is likely to persist. There are two reasons. First, the central government has decreased the rate of the PIT. Effective January 1, 2009, the Government converted three rates of 19 percent, 30 percent, and 40 percent to two rates of 18 percent and 32 percent. As the City does not have the authority to counteract this measure, the one-time rate reduction will have a lasting impact. Second, the base of the city's principal revenue source--the PIT--is sensitive to downturns in the economy. The result is reflected in the projected slowing of growth in revenues in 2009, shown in Figure 8. O These graphs reflect the city's accounting practices. _Revenue' as defined by the city includes receipts from the sale of property and other assets, and therefore overstates the level of the city's recurrent income. SI 36. In the short term, the City has responded in two ways. First, it has cut spending, particularly on capital Figure 7. Warsaw: Composition of Revenues, 2003 - 2009, investment. The City's original million PLN constant prices 2009 budget for 2009 included an 12,000 ambitious capital spending program, totaling PLN 15.3 10,000 m other billion over the period 2009- income from property 2013. Spending on capital 8,000 investment in 2009 alone was a subventions to be PLN 2.9 billion. The 6,000 1 corporation tax revised budget for 2009 cuts a personal income tax that figure by about twenty 4,000 m revenue from services percent. The City's second 2 transactions tax response has been to increase 2,000 a vehicle tax borrowing. In April 2009, the 0E real estate tax City issued PLN 883.5 million in Eurobonds. In July it issued 0 8 8 0 8 8 8 another PLN 617 million on W__ _ _ _ _ Ln _ _)_*_ __0_ _UD_ domestic market. Source: Central Statistical Office, Warsaw's Budget draft for 2010, World Bank staff calculations. 37. This strategy is not sustainable. A permanent cutback in investment spending could significantly dampen the City's long-run growth prospects, as infrastructure deteriorates and the city's productivity declines. Reliance on borrowing may also be unsustainable. Until 2009, the City had very little debt. The ratio of debt to revenue was about 23 percent at the end of 2008. The recent borrowing activity Figure 8. Warsaw: Projected Trends in Revenues, has increased that ratio to 43 Expenditure and Debt, 2004-2020 percent. While some level of debt is to be expected, - Current revenues, LHS - Total expenditures, LHS particularly in a city whose Debt:revenue ratio, RHS functions require investment in 14000 0.9 long-lived capital works, 12000 - - -0.8 continued reliance on /...0 - 0.7 borrowing to close the budget - 0.6 gap may jeopardize the City's 8000 ------------------0.5 fiscal future. The City could 6000 -- 0.4 find itself forced to borrow 4000 0.3 merely to pay the interest on its - 0.2 existing debt, resulting in an 0.1 ever-increasing stock of debt 0 ------------------------0 and reducing the resources LA 0 0 0 O 00 0 (N 0f 0 0 0 O 0 available to finance other 200 - - ~ -~ -~ --0. 0 0 (N 00 00 municipal responsibilities. The Source Data for 2007 and 2008 from Warsaw City budget execution reports; for City's fiscal prospects therefore 2009-2020, from Debt Management Plan. bear close investigation. 12 Fiscal Prospects 38. The City's medium term budget framework (MTBF) and Debt Forecast for the City of Warsaw (DFW) provide an indication of the City's own view of its fiscal prospects and some indication of its fiscal strategy. The MTBF covers the period 2010-2014 and provides a detailed breakdown of projected revenues and expenditures through 2012. The DFW covers a longer period--2010-35--but provides only aggregate data. On the receipts side, it projects current revenue, so-called capital revenue", other proceeds from the sale of assets (e.g. past privatization receipts), and proceeds from borrowing. On the expenditure side, it projects total recurrent expenditure (separately distinguishing interest payments) and capital expenditure (excluding repayments, which are reported as financing items below the line. For the period 2010-2012 the aggregate figures in the DFW are consistent with the detailed figures in the MTBF. 39. The first ten years of the plan are illustrated in Figure 8. In the aggregate, the MTBF foresees a quick recovery in current revenues beginning in 2010. This would be accompanied by slowing expenditure growth in 2010 and absolute declines in expenditures in 2011 and 2012. (All fiscal results are discussed in terms of constant PLN of 2009.) The result, in aggregate, would be a rapid shrinking of the City's cash flow deficit, from 32 percent of revenues in 2010 to four percent of revenues in 2014.12 Over this period, the City expects to generate considerable sums from so called capital revenues (EU grants and sales of assets) and from privatization. Its borrowing requirements would therefore be limited. As shown in Figure 8, the ratio of debt to revenue would rise to about 55 percent in 2010, but then decline steadily thereafter. 40. Receipts In many respects, these projections are quite optimistic. As noted above, the City derives the largest portion of its revenues from shares of the PIT and, to a smaller extent, the CIT. As the City cannot increase the rates of the PIT or CIT, any growth in revenue from these sources will depend upon growth in its underlying tax base-the Warsovian economy. In making its revenue projections, the City normally relies on the central government's forecasts of growth in the Polish economy. These foresee a slow recovery, with 1.2 percent growth in 2010, followed by growth rates of 2.8 percent in 2011, 3.0 percent in 2012 and 3.4 percent in the two subsequent years. For 2010, however, the MTBF projects substantially higher rates of growth in the two taxes. Revenues from the PIT and CIT would increase by 4.3 percent in real terms, more than three times the rate of economic growth. It is conceivable that revenues from these sources could grow substantially faster than the projected rate of growth in the Polish economy. Warsaw's own economy may grow faster than the national average. The Government may increase the rate of the PIT or CIT or increase the share of them that the City is allowed to retain. Nevertheless, it appears unlikely. 41. The MTBF's projection for growth in _other current revenues' is less ambitious. This category incorporates a wide range of sources, including gross revenues from ticket sales on public transportation, gross property income, and subventions from the central government. It accounts for about forty percent of current revenues. The MTBF projects that revenues from " Revenues, for purposes of this analysis exclude EU grants and proceeds from the sale of assets, including real estate. If these items were included in the definition of revenues (as is Warsovian practice) the deficit would decline from 22 percent of revenues in 2010 to three percent of revenues in 2014. 12 References to debt:revenue ratios in this report refer to current revenues, rather than total revenues', a term which in Warsovian parlance, includes so called capital revenues. 13 these sources will grow by 2.4 percent in real terms in 2010 and 2.0 percent in 2011. While this would require a combination of real increases in rents on municipal property, tariff increases in public transport, and increases in other fees and charges, it is not outside the realm of possibility. In principle, Warsaw could be vulnerable to Government decisions affecting the funding of education subventions and, to a smaller extent, the social assistance dotacje. As noted earlier, the level of the education subvention is not directly tied to changes in the cost of providing education. The Government, as well as Parliament (through its control over the annual budget law) can change the level of the subvention as they see fit, although nominal reductions in the education subvention are legally prohibited, as noted earlier. Historically, the Government has increased the level of the education subvention to match any increase resulting from central collective bargaining agreements with the teachers' unions. There is a risk that in fiscally stressful times, it will cease to do. But the risk is considered small. 42. The City's projections for its other major revenue sources are also relatively modest. Revenues from property taxes are expected to increase by about three percent in 2010 but remain stable thereafter. But because the PIT (and to a smaller extent, the CIT) account for so large a share of City's revenues, overestimates of revenues from these two sources may overstate the level of growth in total current revenues in the short term. 43. Expenditures With respect to expenditures, the MTBF projections also appear somewhat rosy. Total expenditures are projected to increase by only 2.8 percent in real terms in 2010, and then decline, in real terms, by 3.3 percent and 4.2 percent in the following two years, respectively. Current expenditures Figure 9. Warsaw: Composition of Expenditures would remain constant in real terms in by Economic Category 2008, percent 2010, and would fall by 3.5 percent and one percent, respectively, in 2011 and Composition of Expenditure by Economic 2012. Capital expenditures (excluding Category principal repayments) would increase by 2% m Salaries and related ten percent in real terms in 2010, and .Contribto then decline by 2.4 percent in 2011 and equalization fifteen percent in 2012. * Transport subsidies 44. The plausibility of these m Other current exp expenditure projections depends in part on the economic composition of U Capital works Warsaw's expenditures. Some categories of expenditure are more 1 Debt service readily cut than others. The MTBF provides only limited detail on projected expenditures, distinguishing only between current and capital expenditure, Source: Central Statistical Office, World Bank staff calculations. and within those categories, interest and principal repayments, respectively. The composition of spending in 2008, nevertheless, provides some perspective on the flexibility of the City's expenditures. Figure 9 illustrates the economic composition of Warsaw's expenditures in 2008. As shown, spending is dominated by five major items: salaries, contributions to the equalization mechanism, transfers to the transport companies, purchases of other goods and services, and capital works. 14 45. Salaries account for about one-quarter of total spending. In most situations, they are considered inflexible. Governments find it politically difficult to reduce salaries in nominal terms or to dismiss large numbers of staff simultaneously-particularly in the midst of an economic downturn. More subtle approaches-freezes on nominal wages, freezes on the creation of new positions (or more drastically, freezes on all hiring) tend to be more acceptable. The immediate impact of a nominal wage freeze in Warsaw would be fairly limited in the short term. With inflation projected to be only one percent in 2010, a nominal freeze in wages would reduce total spending only 0.25 percent. A salary freeze in 2011 would reduce them by another 0.63 percent.13 Larger reductions would require either more drastic measures, such as large scale personnel cuts or reductions in nominal salaries. As discussed in the section on education spending, there do appear to be opportunities to cut staffing at the primary and secondary level. But it is unlikely that they could be carried out quickly enough and on a sufficient scale to achieve the DMP's projected reductions in current expenditures. 46. The City's contribution to the equalization mechanism also appears unlikely to decline. Warsaw's contributions to the national equalization mechanism consumed about 15 percent of its expenditures in 2008. As described earlier, this obligation adjusts automatically to changes in the city's fiscal situation, as it is based on a percent of Warsaw's revenues (more specifically, on the difference between Warsaw's per capita revenues and the average per capita revenues of all gminas and powiats, multiplied by the City's revenues). If Warsaw's revenues were to fall, its mandatory contributions to the equalization mechanism would also fall. If they rise (as the DMP projects) they would increase (provided Warsaw's increase was equal to or greater than the increase of all other powiats and gminas in Poland). 47. Payments to the transport companies account for about 14 percent of total expenditures. Raising tariffs sufficiently to permit a nominal freeze on payments to the transport companies would, again, have a relatively modest impact in 2010, but a more significant one in 2011. As discussed in the section on transport subsidies, such opportunities appear to exist. 48. Spending on capital works accounted for nearly 20 percent of total expenditure in 2008. This is one area where substantial cuts would be possible, by postponing new starts and stretching out the implementation of ongoing projects. A sharp cut in capital spending in 2010 could therefore go a long way toward fulfilling the DMP's fiscal targets for that year. As the DMP already projects a decline in capital spending after 2010, actual cuts would have to be even deeper. 49. Financing Plan Despite the optimism of some of its projections, the DFW foresees a cash flow deficit running from 2010 to 2014. The DFW, however, anticipates that most of the resulting financing gap will be filled using resources from so-called capital revenues and privatization. The MTBF anticipates a modest inflow of so-called capital revenues in the form of EU grants and proceeds from the sale of assets. EU grants are expected to generate PLN 554 million in 2010; PLN 755 mn in 2011 and PLN 864 mn in 2012. For 2010, the MTBF expects 1 In the Polish context, a locally-imposed wage freeze would be technically feasible but somewhat complicated. Changes in the base wages for all public employees are determined through national collective bargaining agreements between the Government and the various unions. Subnational governments are, however, free to 'top up' the wages of their employees with their own resources. Warsaw does so for teachers, for example. Warsaw could, in principle, impose a local wage freeze by scaling back its top-ups to compensate for any centrally-imposed increases in the basic wage. Note that the estimated impact of a local wage freeze described in the main text also assumes that the wage freeze would be accompanied by a freeze on the creation of new positions. 15 that PLN 167 million will be generated from the sale of assets. In 2011, the total is projected to fall to PLN 128 million, and then rise to PLN 308 million in 2012. In addition, the City projects to realize substantial revenues from privatizations. These are reported as _other receipts' in the DFW and are projected to generate PLN 727 million in 2010, PLN 1.15 billion in 2011 and PLN 283 in 2012. (All figures are in PLN of 2009). Based on these estimates, these sources are expected to finance roughly one third of the cash flow deficit in 2010, 85 percent of it in 2011 and roughly half of it in 2012. 50. Although these numbers are large, they are, in principle, plausible. According to City's debt forecast, proceeds from privatization alone generated PLN 1.56 billion in 2008 and PLN 1.23 billion in 2009.14 There is reportedly much more to sell. According to a forthcoming report on the management of the City's assets 15 the value of City-owned fixed assets totaled PLN 88.5 billion of June 30, 2008; of which PLN 73.7 billion consisted of land (which would presumably be readily marketable). But the City should avoid selling off its property assets haphazardly, particular in the midst of a recession. As described in the asset management paper, the City should develop a carefully considered strategy for the disposition and use of its real estate, as it represents not only a valuable financial resource but also a key instrument for managing urban growth and public services. 51. The remainder of the deficit would be financed from borrowing. As borrowing would be limited, City's debt stock would increase modestly from 45 percent of total revenue in 2009 to 59 percent in 2010 and then decline. As shown in Figure 8, the level would skirt the 60 percent ceiling but never entirely rise above it. (See Box 2.) Box 2. Current and Proposed Controls on Debt The Public Finance Law of 1998 (reaffirmed in the Public Finance Law of 2005) limits the stock of a local government's debt to 60 percent of its current revenues and its debt service obligations to 15 percent of its current revenue. (In the event that local government debt exceeds 55 percent of total revenue, the ratio of debt service to total revenue has to be reduced to no more than 12 percent.) New legislation approved by Parliament would replace these ceilings with a single, somewhat riskier, rule limiting debt service to the average of a local government's current account surplus (with revenues defined to include receipts from the sale of assets) over the preceding three years. Source: Public Finance Act 52. Alternative Scenarios Like all fiscal projections, the MTBF and DFW are subject to a variety of uncertainties. These include changes in the economic environment that lie beyond the City's control, as well as political constraints that might prevent the City from carrying out the spending cuts that would appear to be needed to achieve the MTBF and DFW fiscal targets-- even if economic conditions developed exactly as foreseen. To test the vulnerability of the City's fiscal projections to such uncertainties, a model of the City's fiscal prospects was devised. The model uses the debt: revenue ratio as an indicator of the City's fiscal condition. It assumes that 14 Figures are in constant PLN of 2009 and exclude so called capital revenues. 1 Olga Kaganova , Public Land and Property Asset Management in Warsaw: Strategic Opportunities Discussion Draft January 2010. Background Paper for Warsaw Urban Policy Note (forthcoming) 16 any shortfalls in revenues or increases in expenditure will have to be met by increased borrowing.16 Rising debt: revenue ratios therefore indicate declining fiscal health. 53. Overall, this analysis yields two conclusions. First, the principal external threat confronting the City is continued recession. The effect of a double-dip recession in the U.S. and major EU countries on the Warsovian economy would undermine the projected growth in revenues from the PIT and CIT and lead to a significant increase in the City's debt burden. Rising interest rates (within the range evaluated by the model) and falling exchange rates (ditto) would not have this effect. But the other threat to the City's finances comes from within: the inability of the City to control recurrent expenditures. If the City is unable to restrain growth in salaries, subsidies, and spending on supplies, it faces the prospect of unsustainable debt 54. Basic framework. The analysis is based on the following equation: d, =+± +C,a d(I , - b, (1+ g,)(1+T) where: dt represents the City's outstanding stock of debt in period t; bt. represents the City's primary balance (the difference between total receipts (other than from borrowing) and the sum of current and capital expenditures, net of interest and principal repayments) as a share of GDP in the period t gt is the real annual growth rate nt is the inflation rate, 1 =h f is a weighted average of the domestic interest rate I and the foreign interest rate i a is the fraction of foreign-denominated debt, a _ the fraction of domestic-denominated debt, et = Ae,/et_1 the rate of depreciation of the local currency. 55. As such the model incorporates factors that are (to varying degrees) under the city's control: its revenues and its (non-debt) expenditures. But it also incorporates factors that are not 16 As such, the model is not constrained by the existing legal ceilings on local borrowing. If these ceilings are enforced, Warsaw will not be able to finance growing deficits indefinitely and would have to either increase revenues or reduce expenditures. 17 under its control: the growth rate in the City's economy, the rate of inflation, the prevailing interest rate on City's debt, and changes in the nominal (PLN/EUR) exchange rate. 56. Four scenarios were tested. The first tests the impact of alternative rates of GDP growth and inflation; the second, rising interest rates; the third, depreciation in the exchange rate; and the fourth, a failure to control spending. All compare outcomes against a baseline scenario. 57. The baseline scenario is based on the City's debt forecast (DFW). All budget variables- current revenues, current and capital expenditures (excluding debt service) along with macroeconomic variables--the growth rate of real GDP, the inflation rate and the exchange rate - are taken from the DFW. The interest rates and maturities of existing and projected debt are computed from the debt service data in the plan.17 The baseline scenario (and all other scenarios) also accept the City's projections of receipts from so called capital revenues and proceeds from privatization. (In keeping with international accounting practices, capital revenues are classified as capital receipts and treated as a financing item in these projections.) Table 4: Baseline scenario based on government projection Baseline scenario (government's projection) nominal nominal nomina nominal growth of growth of nomi- nomi- implied year 1 growth of primary primary GDP- infl. EXR nal nal share growth current total current growth interest interest of euro of total revenues expenditure expenditure rates rates debt revenu dom. euro es 2010 7.04 4.38 2.49 0.15 1.2 1.0 4 9.43 7.26 19.81% 2011 7.49 4.44 -1.32 -1.66 2.8 2.5 4 7.84 5.91 18.82% 2012 5.78 5.14 -2.22 0.98 3.0 2.5 4 8.87 5.30 17.94% 2013 5.56 5.56 3.88 5.40 3.4 2.5 4 8.99 5.13 17.08% 2014 -0.24 5.96 -4.43 6.05 3.4 2.5 4 8.95 5.32 18.57% 2015 0.56 3.50 0.34 1.54 3.0 2.5 4 9.00 5.57 17.59% 2016 3.41 3.50 3.75 3.77 3.0 2.5 4 9.00 5.57 16.65% 2017 4.99 3.50 3.72 3.74 3.0 2.5 4 9.00 5.57 15.61% 2018 3.37 3.50 3.71 3.73 3.0 2.5 4 9.00 5.57 14.54% 2019 3.37 3.50 3.48 3.48 3.0 2.5 4 9.00 5.57 16.53% 2020 3.38 3.50 4.21 4.27 3.0 2.5 4 9.00 5.57 18.79% 58. Scenario 1 considers two different GDP and inflation forecasts. The first assumes that the economy will recover more rapidly than is projected under the baseline scenario. Based on the 17 As the plan does not provide information on average maturities and the share of foreign debt after 2014, these were estimated. It was assumed that 20 percent of the government's newly issued debt would be denominated in euro. The maturity profiles for newly issued debt were calibrated to match the government's information on annual debt service payments. It was assumed that 50 percent of the City's recent bond issue will mature after 5 years while the other 50 percent after 10 years. Other recent borrowing is assumed to have a maturity of ten years. It is assumed that half of any new borrowing will take the form of bonds, 30 percent as domestic credits and loans and 20 percent as euro debt, with ten year maturities on bonds and 15 year maturities on other debt.. The resulting debt service estimates were checked against the City's debt service projections and found to match closely. 18 World Bank's most recent macroeconomic projections, it assumes that economic growth will be one percentage point higher than the baseline in 2010, 0.6 percentage points higher in 2011 and 0.5 percentage points higher in 2011. As shown in Table 5, inflation is expected to be slightly higher than the baseline in 2010, but lower in the following year and roughly equal in the third. 59. The second assumes that growth will be slower than the baseline. This could occur, for example, if the U.S. and major EU economies fall back into a recession in 2010 as the one-time effects of counter-cyclical public spending programs (including the Polish one) expire while overall consumption and investment levels still remain weak. This event, which is often referred to as a double-dip recession, would reduce the external demand for Polish goods and services as well as domestic demand due to lower capital inflows into the Polish economy. Both effects would reduce economic growth in Poland and substantially delay its economic recovery. As shown in Table 5, projected growth rates under this scenario would lag the baseline until 2015, while inflation would parallel it (except in 2011).18 60. These differing economic outlooks would directly affect the City's revenues, which Table 5 Alternative scenarios: GDP-growth and need Th onequecs re inflation ised in Figu e Te Baseline scenario Sensitivity Sensitivity (government) Analysis I Analysis II middle (red) line shows the (fast recovery) (double-dip) projected trend in the debt revenue year GDP- GDP- GDP- infl. ratio under the baseline scenario. growth infl. growth infl. growth As discussed earlier, the ratio 2010 1.2 1.0 2.24 2.3 0.5 1.0 would peak in 2010 and then 2011 2.8 2.5 3.40 2.1 1.2 1.0 decline slowly thereafter. Not 2012 3.0 2.5 3.50 2.5 2.8 2.5 2013 3.4 2.5 3.36 2.5 3.0 2.5 2014 3.4 2.5 3.26 2.5 3.4 2.5 recovery has a salutary effect on 2015 3.0 2.5 3.25 2.5 3.4 2.5 City's fiscal situation, relative to 2016 3.0 2.5 3.23 2.5 3.0 2.5 the base scenario. As shown in the >2016 3.0 2.5 ,3.04 2.5 3.0 2.5 lower (blue) line, the debt-to- revenue ratio would be peak at around 56 percent in 2010 and decline rapidly thereafter. The City's debt would be completely paid off by 2020. In contrast, the double-dip recession scenario would result in a steadily rising debt: revenue ratio. Even though the assumed rates of growth and inflation would converge with those of the base scenario after 2013, the difference in the initial years would be enough to send the city into possibly unsustainable levels of indebtedness. Warsaw would exceed the debt-to-revenue limit of 60 percent in 2010 (if current, rather than total, revenues are used as the denominator) and remain above it throughout the projection period. This underscores the City's vulnerability to economic conditions. 18 All four alternative scenarios also employ a more plausible estimate of current revenue growth in 2010. This is based on an analysis of the relationship between GDP and current revenues implicit in the DFW. The implied elasticity of current revenue with respect to GDP ranges from .83 to 1.0 between 2011 and 2014 and is .64 thereafter. In estimating current revenue in 2010, the model assumes an elasticity of 1.64. 19 Figure 10: Impact of Alternative Economic Growth Scenarios 80%. e 70%6 2 0% 20 40% 0% 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Time GDP-inflation scenarios - Debt model (government baseline) - Alternative two (double dips) - Alternative one (WB-IBS) 61. The second scenario examines the impact of higher interest rates--specifically, an increase of 200 basis points in the rates paid by the City on its domestic and foreign debt. As shown, the impact in the initial years would be relatively small. As in the baseline scenario, the debt revenue ratio would rise to about 60 percent of revenues in 2010, rise slowly through 2012, and then decline. Higher interest rates would, however, slow the rate of decline. In 2020, Warsaw's debt stock would still equal about 45 percent of revenue, as opposed to 30 percent under the baseline scenario.. Figure 11: Impact of Rising Interest Rates 65%- e 60% S45% O S0% 40% M 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Time Interest rate scenarios - Debt model (government baseline) - Higher interest rate (2010-2020) 62. Figure 12 shows the impact of a decline in the exchange rate. (Specifically, it assumes that the zloty falls to its ten year low of PLN 4.9: 1 Euro.) Due to the small proportion of Euro- denominated debt in the City's portfolio--about twenty percent--the impact would be negligible. Higher levels of exposure would of course increase this risk. 20 Figure 12: Impact of Exchange Rate Depreciation 65% - * 60% - 0% 2 40%- 0% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Time Exchange rate scenarios - Debt model (government baseline) - Exchange rate depreciation 63. The fourth scenario illustrates the impact of uncontrolled spending. As discussed earlier, the base scenario, reflecting the MTBF and DFW projections, assumes that current expenditures will remain constant in real terms in 2010, and will fall by 3.5 percent and one percent, respectively, in 2011 and 2012. Such reductions may be difficult to achieve. The fourth scenario therefore assumes that the City will be unable to control these costs; more specifically that current expenditures (other than interest) will grow at the same rate as nominal GDP over the period of the forecast horizon. Spending on capital works is assumed to remain unchanged. The result, illustrated in Figure 13, would be an explosive increase in the debt-revenue ratio. By 2014, debt levels would equal 104 percent of revenues. By 2020, they would exceed revenues by a factor of 2.5. Figure 13: Impact of Uncontrolled Current Spending 260% 240% *220% *200% S180% 160% 140% S120% S100% 80% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Time Expenditure scenarios - Debt model (government baseline) - Alternative expenditure scenario 64. The remainder of this section will focus on two areas of potential savings in recurrent spending: (i) reductions in subsidies to education, through school network rationalization; (ii) 21 reductions in subsidies to the city's transport companies. It will also explore one area of indirect savings: improved targeting of infrastructure investments. This would allow the City to reduce capital spending while still providing funds to high priority projects. Rationalizing the School Network 65. The city of Warsaw spent in 2008 some PLN 2.1 billion on education, equal to about 21 percent of its overall expenditures. Only 45 percent of these educational expenditures were funded by the subvention from the state budget, with the remainder funded from the City's own general revenues. Figure 14. Trends in educational subvention, 66. As described earlier, the expenditures and number of pupils, Warsaw, 2003-2010 education subvention is intended - Educational subvention, million PLN, constant prices to compensate subrational Current education expenditures, million PLN, constant prices - No of pupils, thous (RHS) governments for the costs of 1,800 185 1,800- 185 providing the level of education 1,600 180 assigned to their respective tiers. 1,400The level of the education 1,40 175 subvention is not fixed by 1,200 170 formula but is instead set annually 1,000 165 in the budget law. Although the actual distribution algorithm 800 160 fo 6 varies year to year, the bulk 600 155 of the education subvention is 2003 2004 2005 2006 2007 2008 2009 2010 allocated to each unit of Notes: 2009 and 2010 is based on the Budget draft for 2010, No of pupils subn ationalment wt hhe includes students from primary, gymnasia and secondary schools (excl. wig ts f or lss fn e students ofvocational schools and adults) ua Source: Budget execution reports 2003 - 2008, Budget draft for 2010, Central and small towns, pupils with Statistical Office and World Bank staff calculations. disabilities, schools in which classes are taught in minority languages, and vocational schools. 19 67. In principle, this allocation formula should prompt subniational governments to consolidate under-enrolled classrooms and reduce teaching staff. Falling enrollment should result in falling levels of subventions, forcing subnational governments to make such adjustments. To date, however, the City has failed to respond to the efficiency incentives embedded in the education subvention formula. While Warsaw's student population in primary and secondary education decreased by 10 percent between 2003 and 2008 (see Figure 14) the number of teachers has not declined proportionately. The result is very low pupil-teacher ratios in some districts of Warsaw. As a result, the City is forgoing an opportunity to achieve substantial savings. 19 The algorithm has been the subject of constant political bargaining but is widely accepted and has played an important role in increasing the transparency of education funding. It has undergone significant changes between 2000 and 2008. What began as a fairly straightforward per-student based algorithm now includes both students (their number and characteristics) and teachers, although in the latter case, it is teacher qualifications rather than number of teachers, that is taken into account. 22 68. The City's failure to respond is due to two factors. First, the central government continues to increase the amount of the education subvention in the country as a whole, despite nation-wide declines in school enrollment. As a result there is little pressure from the central government funding system to consolidate classrooms or dismiss teachers. At the same time, there is strong resistance, particularly from the _mayors' of the 18 districts into which the city is divided, to any such move. Presumably there is also resistant from teachers, since network rationalization would imply job losses. As a result, spending on education in Warsaw continues to grow--financed not only by the subvention but also by ever-increasing allocations from the City budget. As shown in Figure 14, the City's current expenditures on education increased by over 50 percent in real terms between 2003 and 2008, while the educational subvention increased by 16 percent over that period. 69. From the central government's perspective, there is a case for reducing the amount of the subvention to reflect falling enrollment levels in Poland as a whole (while adjusting them upward to reflect centrally-dictated increases in salaries). From the City's perspective, there is strong case for rationalizing the network.20 This process could being with an inventory of pupil-teacher ratios by class in each school, to determine where pupil teacher ratios are excessively low. The City could then devise a scheme for consolidating classrooms, reassigning pupils, and reducing teaching staff. If successful, this would enable the City to substantially reduce the amount it allocates to education from its own revenues, without reducing the salaries paid to individual teachers or reducing education quality. The City should also look into the possibility of reducing administrative staff employed in the education sector. Warsaw has twice as many administrative staff as teachers. Cutting Transport Subsidies 70. There is also substantial room for savings in transport subsidies. Seventy-five percent of public transport services in Warsaw are provided by three City- owned companies: a metro, a tram company and a bus company. (See Box 3) Although the transport companies are organized as separate enterprises, their ticket revenues are treated as general revenues in the city budget. The City separately compensates each company on the basis of service contracts. 71. Subsidies to the transport companies are substantial. According to data for 2008, total payments to the transport companies for recurrent expenditures totaled about PLN 1.2 billion. Receipts from ticket sales totaled about PLN 500 million, yielding a de facto subsidy of about PLN 700 million. This is equal to about sixty percent of the total payment to the transport companies, and nearly six percent of the City's total expenditures. 72. Subsidies are large for two reasons. First, tariffs are low. Relative to operating costs, Warsaw's tariffs are the lowest in Poland. Although the fares were adjusted in late 2008 (the first increase in seven years) the aim of the adjustment was to cover only 50 percent of total costs. In the event, the proposal submitted to council would have covered only 40 percent of costs and was subsequently amended downward. 73. There is, of course, a case for subsidizing public transport in Warsaw. Cheap public transport can help divert travelers from private cars, thus reducing traffic congestion (with its 20 The 2009 budget includes PLN 30 million for school restructuring, but this refers to modernization not facilities rationalization. 23 accompanying costs in terms of lost productivity) and adverse environmental impacts. Subsidies may also have positive distributional impacts, to the extent the poor rely disproportionately on public, as opposed to private, transportation. It is not clear, however, that the present level of subsidies is justified. Box 3. Warsaw's Municipal Transport Companies Metro company (MW) operates the existing single metro line (north-south). Annual service contracts have been concluded, but a long-term contract is being negotiated to allow MW to invest in rolling stock. Infrastructure was built by the City, but all the equipment was installed and is owned, operated and maintained by MW. In addition, MW is currently a contractual proxy investor on behalf of the City for the second line (east-west) that was recently tendered and is expected to be constructed by 2014. Tram company (TW), owned by the City of Warsaw, is required to provide services in terms of train-km as set by the ZTM. The City and the company recently signed a 20-year contract for operating the service and maintaining and upgrading the infrastructure and rolling stock. The amount of public subsidy was negotiated based on the operation and capital expenses. The company's capital investment plan for 2009-2013 has been prepared and approved by the City. The financial plan for the investment is being prepared including potential financing sources. Municipal bus company (MZA) is wholly owned by the City of Warsaw. It provides bus services within Warsaw and connects the City with almost all neighboring gminas based on bilateral agreements. The MZA owns buses and all infrastructure related to bus services (bases, garages, workshops). All its revenues come from Warsaw City. MZA recently negotiated a long-term contract with the City to be able to purchase new buses. Fast urban railways (SKM) is a passenger railway operator created a few years ago and wholly owned by the City of Warsaw. It currently provides railway services along one diagonal' railway line form the west to the south-east of Warsaw via the city center. SKM plans to purchase more rolling stock and operate new lines within Warsaw using significant financial support from the City. Source: World Bank staff information 74. Any major increase in tariffs should be preceded by an assessment of the impact of higher tariffs on ridership (and the diversion of passengers to private cars) as well as on lower- income groups. There are, nevertheless, several proposals that bear immediate consideration.21 * Raise tariffs on time-based tickets. Presently, the City offers a standard ticket good for one day on any of the city-owned transport systems in Warsaw's central zone. But passengers also have the option of paying a lower fare for a ticket that is good for only 20 minutes. Although this may generate additional trips, it is not clear that it does so. Instead, the city appears to be sacrificing revenues, as people who use time based tickets would otherwise pay full fare. The city should examine the merits of time-based tickets and consider eliminating those that result in net losses in revenue. 21 For further details, see Gwilliam, Kenneth, Warsaw Public Transport Financial Sustainability Study November, 2009 24 * Raise tariffs on long-distance commuters. Presently passengers pay a higher tariff for trips from the center city to the immediate suburbs of Warsaw, but the fare, in terms of zlotys per kilometer, drops off rapidly with increasing distance. There is evidence that passengers are willing to pay higher fares for these trips. (The private, unsubsidized, bus companies serving some of these routes charge substantially more, for example.) Fares on these long distance commutes should be raised. The City is now considering extending bus services into a further ring of outer suburbs. Before doing so, the City should evaluate the fare it intends to charge for this service in order to minimize the fiscal burden it imposes. * Reduce the number of groups eligible for concessionary fares. While the central government mandates concessionary fares for certain groups (mostly students) 22 the City provides concessionary fares to other groups at its own expense and discretion. These include the elderly, the disabled, city employees and certain other categories. The city should reduce the number of discretionary groups eligible for concessionary fares, or at least limit to concessionary fares to off peak hours. 75. The City should also exploit the potential for real estate development around transport stations. In this regard, the City is already renting space above some central metro stations for commercial parking. It has not, however, exploited the potential for major developing the -air space" above the metro stations for a broader range of commercial or even residential development. 76. The second reason subsidies are high is because expenses are bloated. In addition to raising tariffs, the City can encourage the transport companies to improve efficiency. At present, the City's transport management agency, ZTM, remunerates the individual transport companies on the basis of vehicle kilometrage. The amount paid to the public bus company, for example, is based on the number of vehicle-kilometers specified in the company's service contract, multiplied by the average operating cost per kilometer for vehicles of various sizes. The operating cost per kilometer figures are based on the company's own estimates as approved by ZTM. Public sector tram and metro services are procured on a similar basis. The companies' cost figures are subject to financial audits. These check the ex ante legitimacy of the items included in the calculation as well as the ex post veracity of their accounting. In addition, the audit reviews whether the companies have met agreed service quality standards such as frequency and reliability. Non-compliance is penalized in financial terms. But the auditing focuses solely on financial and service quality. It fails to assess whether the companies are providing services as efficiently as possible. 77. There is evidence that they are not. International experience suggests that procurement arrangements based on vehicle kilometrage fail to generate the competitive pressures that could lead to greater operating efficiency. An alternative approach would be to invite competitive bidding for the City's bus services. Competitive bidding has been instrumental in driving down costs in other European countries. In the EU, the cost reductions associated with this step have ranged from 20 percent to 40 percent. While these are initial effects which may be partly lost in subsequent rounds of tendering, the evidence suggests that they do produce lasting cost 22 While similar nationally-mandated concessions offered on services provided by the voivodship receive direct compensation from central government, those applying in Warsaw do not. 25 reductions.23 The City could therefore consider converting the negotiated contracts with the city companies' to competitively tendered contracts. Because Warsaw has already engaged in the competitive tendering of some routes, much of the preparatory work has already been done in this respect.24 Targeting Capital Investments 78. The City should also improve its system of capital investment planning. While the recent cuts in capital investment are a reasonable response to the City's financial reverses, not all capital investments should be delayed indefinitely. To ensure that increasingly scarce resources reach the highest priority projects, the City should improve its process for identifying, evaluating, and funding capital investments. 79. Capital spending accounts for a surprisingly large share of the city's expenditures. In 2007 and 2008, it accounted for twenty percent of total expenditure.25 In the 2009 revised budget, it was expected to account for 20 percent. 80. In principle, Warsaw's investment program is guided by a series of strategic plans. The City Development Plan until the Year 2010 (1999) provided a sector-by-sector development framework for the City. The City Spatial Development Plan (Polityka Przestrzenna Miasta Stolecznego Warszawy, 2007) lays out, in map form, a plan for investment in each of the infrastructure sectors-public transport, roads, water supply, sewerage, etc. The City also has sector specific strategic plans, such as the Strategy for the Sustainable Development of the Warsaw Transport System by 2015. This was formally adopted by the City Council in July 2009 and provides a comprehensive transport development plan, with well defined and well-evaluated investment priorities. 81. All of these plans, however, appear to have only a loose relationship with actual project funding decisions. Although the earlier (1995) transport strategy was adopted by the City Council, for example, it was never fully reflected in the City's capital investment program. The new strategy has also yet to be reflected in the budget. 82. Project evaluation is also deficient. A standardized feasibility study following EU guidelines is required only for the EU-financed projects, not for those financed from the budget. Projects therefore appear to be evaluated on more subjective criteria, without sufficient supporting technical analysis. For instance, the decision to undertake the metro extension appears to have been made without sufficient consideration of likely levels of ridership or possible alternatives (e.g., bus rapid transit). The cost per kilometer of constructing the second line is more than four times that of the first line (PLN 724 million vs. 169 million per km26). Given its short length (5.8km), its benefit to transport users is questionable. 23 While the difference between the costs of the public and private operators in Warsaw appears to be smaller than the outcome differences elsewhere this may well be because of the very limited commercial freedom presently accorded to operational management of the companies. 24 The City's bus companies currently provide 75 percent of the City's bus capacity. Only the remaining 25 percent are procured through competitive tendering from the private sector. 25 Note that this figure overstates the budgetary impact of cutting capital works, as larger projects are partially financed through EU grants and international loans. Cutting or postponing such projects, would, however, reduce the city's counterpart funding obligations. 26 Department of Roads and Public Transport of the City of Warsaw 26 83. The City does have a medium term investment plan. The current plan covers the period 2009-2013 and includes over 300 separate projects, showing the amount to be spent on each project in each year over the six year period.27 A check of the 2009 annual budget shows that the six year plan's allocations for 2009 match exactly: the amount allocated to each project in the 2009 annual budget is precisely the same as the amount specified for that project in 2009 in the six year plan. But this does not necessarily reflect a commitment to medium term capital investment planning. It is reported that allocations in the six year plan are routinely adjusted retroactively to reflect spending priorities in the current year. 84. The process for allocating funds to individual projects is instead considerably more flexible. Responsibility for this task rests, in the first instance, with the City Development Department (CDD). The CDD is charged with preparing the City's annual investment budget. In carrying out this role, its decisions are largely based on readiness and availability of financing. Funding is reduced for projects that are expected to have implementation delays or where funding is uncertain. The CDD's recommendations are then forwarded to the mayors' (President's) coordination commission, and finally to the city council for final approval. 85. Some degree of flexibility in the annual capital budget process is desirable. One would not want to allocate money to projects where it will not be spent or to start projects with no prospect of funding. But flexibility can be taken too far. Political pressures, for example, may prompt funding to be shifted from ongoing projects to new starts. This can result in costly implementation delays. 86. Improve data International experience suggests several solutions to this problem. Project identification can be improved by strengthening the data base on which investment decisions are made. In the road sector, for example, a segment-by-segment inventory of the city's road network, showing road surface conditions, traffic volumes and congestion levels, can provide the basis for prioritizing maintenance, rehabilitation and new investments expenditures. Asset management software can make this process more objective and mechanical. 87. Improve project evaluation Project evaluation methodologies could also be improved. One approach would be to apply the EU's project evaluation framework to projects funded from the budget. The EU methodology, however, may be too costly and data-intensive for small projects. The City could consider modifying the EU framework to create a data-light, yet standardized, approach that applies the same criteria to all projects under evaluation. The practices used in other EU member states provide good reference points, including-in the transport sector-- the Transport Users Benefit Appraisal program and Evaluation Guidance for Major Local Transport Scheme developed by UK Department for Transport (http://www.dft.gov.uk/pgr/evaluation/evaluationguidance/). 88. Funding continuity could be improved by allowing projects to be funded on a multi-year basis. The Irish example (see Box 4) is a good example of this approach. 27 A revised plan covering the period 2010-2014 has since been adopted. 27 The Voivodship Expenditures 89. As noted earlier, the fundamental task of a voivodship is to plan and coordinate the development its territory; a task that entails drafting development strategies and managing EU Box 4. Multi-Year Budgeting - Experience of Ireland Ireland's system of multi-year budgeting for infrastructure investments has provided predictability and continuity in project funding, while ensuring sufficient flexibility to respond to changes in rates of implementation. Under the Irish system, funding for a project-a funding envelope-is first authorized on a multi-annual basis in its entirety once a contract is awarded. Then an annual budget envelope is created taking into account budget availability and the expected rate of implementation. Up to ten percent of the unspent funds from any given year can be carried forward into the following year. Funding can also be transferred between projects when this will accelerate implementation or when progress is slower than anticipated. Such funds are then transferred back to delayed projects once their implementation is resumed. Source: World Bank staff information and other regional development funds. The voivodship's expenditure responsibilities, insofar as specific public services are concerned, are limited. As shown in Figure 15 one of the largest Figure 15. Mazowieckie: Trends in Expenditures in consists of compulsory 2006-2009, million PLN, constant prices 2008 fatenal' equalization transfers that the voivodship is required to 4,500 make to other voivodships with 4,000 - - other 4,000 _______other__________ lower levels of per capita tax 3,500 sports 28 a communal 3,000 -- awefr 2,0* welfare 90. The remaining 2,500 U health expenditures take two forms. 1 education One consists of payments to its 0 equalization own direct and indirect budget 1,000 - debt entities, e.g., the voivodship road 500 - admin authority. The other is the 0 2 2 2 2 m transport distribution of grants to other 2006 2007 2008 2009revised levels of subnational budget I government, NGOs, training Source Central Statistical Office, Budget for Mazowieckie Voivodship for 2009, institutions, and similar World Bank staff calculations. organizations. The available voivodship expenditure data 28 As described in the introduction to this report, the voivodship is required to contribute eighty percent of its excess receipts from the CIT and PIT (defined as receipts over 110 percent of national per capita average) to an equalization fund, which is distributed to poorer jurisdictions. In 2008, the voivodship's per capita revenues were 148 percent of the national average. If they had exceeded 170 percent of the national average, the required contribution would have been higher. 28 does not permit the two to be separated. As a result, it is not clear how much the voivodship spends in its capacity as a direct service provider and how much it spends as a distributor of grants. 91. The largest functional area of aggregate voivodship expenditure is transport. Transport spending is normally dominated by roads.29 The Voivodship has steadily increased its spending on roads in recent years. As shown in Figure 16, spending on roads increased (in nominal terms) from PLN 104 million in 2004 to PLN 357 million in 2008. The original budget for 2009 called for an increase to PLN 415 million.30 92. The Mazowieckie Voivodship also spends considerable amount on subsidies to two passenger railway operators. As described in Box 5, the voivodship is the majority owner of two regional railroad companies-the regional passenger railway (KM) and the Warsaw commuter railways (WKD). As in Warsaw, ticket revenues fall far short of operating costs-covering 47 percent of KM's operating costs and 72 percent of WKD's. As a result, both of the lines are heavily subsidized. KM's net operating loss totaled PLN 202 million in 2008 (although direct Figure 16. Mazowieckie: Expenditure on transport by Mazowieckie Voivodship 32 Expenditures Expenditure on Voivodship Roads Transport Expenditures and Share by Mode 450 50% 1,400 400 4 46% -.9 ./.45% 1,200 350 40% 1,000 35%/ 300 30% 800 2 250 600 25% 0 200 20% 150 2%400 15% 200 50 5% 2008 2009 (plan) 0 0% 2004 2005 2006 2007 2008 2009 2 Voivodship roads (plan) E Roads for powiats, cities with a powiat status and gminas m Capital expenditure U National passenger railway services mmCurrentexpenditure a Domestic passenger bus Voivodship roads spending/transport spending U Other Source: Central statistical office, Budget for Mazowieckie Voivodship for 2008 and 2009, World Bank staff calculations subsidies from Voivodship were somewhat smaller). WKD's net operating loss totaled PLN was 7.7 million. Over the 2005-2007 period, KM's operating costs rose more than twice as fast as its ticket revenues. (See Figure 17.) 29 In 2008, however, the voivodship purchased a majority interest in KM and spent a considerable amount on the purchase of the company and on new rolling stock. With these capital expenditures largely out of the way, spending on rail service will drop in 2009. 30 The revised IIIQ budget increases the amount to PLN 454 million. 31 The figure shows operating costs less ticket revenues, with ticket revenues in WKD including central government subsidies to targeted discounts and transfers for the Warsaw transport company to reimburse multimodal tickets and additional income from selling trip permits. 32 Mazowieckie Voivodship Road Administration and Public Information Bulletin of the Mazowieckie Voivodship, available at http://www.bip.mazovia.pl 29 Box 5. Mazowieckie's Transport Companies Regional passenger railway (KM) is a passenger railway operator practically owned by Mazowieckie Voivodship with PKP (Polish State Railways) having a minority stake. It operates all the regional lines connecting Warsaw with all other cities in the Region. KM uses state-owned railway lines managed by PLK (Polish Railway Lines) and pays to PLK track access charges as any other railway operator. It also partly owns and partly leases rolling stock from the Voivodship. The Voivodship is planning further investments in rolling stock. A long-term service agreement (15 years) was recently signed between KM and the Voivodship. The other is Warsaw commuter railways (WKD) is an integrated railway operator owning one dedicated line that connects both the western neighboring gminas and western suburbs of Warsaw with the city center. The company is controlled by Mazowieckie Voivodship with minority stakes owned by gminas located along the line. All owners pay for the services with most costs covered by Mazowieckie Voivodship. WKD is also in charge of all investments in infrastructure and rolling stock. Source: World Bank staff. 93. Spending on health care consumed about eight percent of total expenditure in 2008. Mazowieckie Voivodship owns 36 health care units, employing close to 18 thousand employees. Among them are large specialized hospitals like the St. Anne's Voivodship Injury Hospital or Specialized Voivodship Hospital in Ciechanow, and smaller health entities providing out-patient Figure 17. Mazowieckie: Trends in Operating services. The voivodship is the owner of Losses of Regional Rail Services, million PLN certain hospitals which were distributed among the various levels of local -KolejeMazowieckie -Warsaw Commuter Train government in the late 1990s. The 250 distribution did not reflect any particular 200 attribute of the hospitals. The distribution 150was not made on the basis of specialized, 150 as opposed to general, hospitals, for 100 example. In principle, the voivodship's 50 financial obligation is limited to building 50 and maintaining hospital buildings and 0 purchasing equipment. Operating costs 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009are supposed to be financed through Source: Financial statements of KM and WKD, World Bank staff direct transfers from the national health calculations, fund. In practice, the hospitals tend to run deficits, which are financed by running up arrears to suppliers and the social security administration, among others, and borrowing from commercial banks. The stock of outstanding hospital debt in Mazowieckie voivodship, including arrears, stood at PLN 400 million at the end of the second quarter of 2009. This figure includes the debts of hospitals owned by the voivodship and by powiats and gminas within Mazowieckie region. The debt of health care units owned by the Voivodship stood at PLN 86 million in 2009. The debt is highly concentrated - six largest hospitals are responsible for nearly 80 percent of total debt outstanding. The voivodship is legally prohibited from financing these deficits or paying off these arrears. The Polish hospital system as a whole is, nevertheless, subject to periodic bailouts by the central government, to which the voivodship is expected to contribute. 30 94. Education constituted five percent of total expenditures in 2008. The voivodship's responsibility in this field is limited. As noted earlier, gminas provide grades 1-6; powiats, grades 7-9 and 10-12. The voivodship is responsible only for optional post-high school vocational schools. Recurrent costs of these schools are partly reimbursed by the Government education grant, noted above. Revenues 95. Mazowieckie's revenues are largely derived from fixed shares of centrally administered taxes. As shown in Figure 18, about nearly three quarters of revenues in 2008 were derived from statutory shares of the corporate and personal income taxes. Another ten percent was derived from other central government transfers, including education subventions and subsidies to cover the costs of bus services to certain socially vulnerable groups. According to the 2009 budget, the voivodship also expected to derive a considerable volume of resources from EU structural funds, particularly the regional operating program and the EU social fund. As shown in Figure 18, these made only a small contribution to the voivodship's receipts in 2008 (PLN 75 million) but were expected to generate PLN 918 million under the revised 2009 budget. Figure 18. Mazowieckie: Trends in Revenues 2006-2009, million PLN, constant prices 2008 Fiscal Prospects 96. Over the last three Source: Central Statistical Office, Budget for Mazowiec years (2006-2008) the 4,000 voivodship's expenditures 3,500 other revenue _have grown at a 3,000other revenue considerably faster pace N welfare than its revenues, resulting 2,500 health in a widening deficit. 2,000 -EU, other grants (Figure 19). In 2008, the 1,500 - education subv. deficit-the difference 1,000 * CIT, PIT between revenues 500 * property mgmt (including grants) and 0 2 r transport expenditures (including 2006 2interest payments) was budget equal to PLN 514 million, kie about twenty percent of Voivodship for 2009, World Bank staff calculations, revenues. According to voivodship budget execution data, this deficit was largely financed through debt. Proceeds from long term borrowing (kredyt dbugoterminowe) generated PLN 580 million in receipts in 2008. As a result, according to the voivodship's long term budget projections, the stock of debt rose from PLN 352 million at the end of 2007 to PLN 924 million at the end of 2008. e In addition, the voivodship reported receipts of PLN 24.5 million from credits for co-financing EU investments, PLN 21.3 million in debt repayments (from loans extended to other units of government) and PLN 108 million in other sources of financing. As a result, the Voivodship finished the year with a cash flow surplus of PLN 61.1 million. 31 Figure 19. Mazowieckie: Recent and Projected Trends in Revenue, Expenditure and Debt - Total revenues, LHS - Total expenditure, LHS Debt:revenue ratio, RHS 4,500 60% 000 -50% 3,500 3,000 40% 2,500 -30% 2,000 1,500 20% 1,000 -10% 500 0 0% oD r 0 0 C3 M~ -ZT Ln ,- D r, oo a) C, 0 0 0 0 0, 0 0 0 0 0, 0 0 0 0 0, Source Central Statistical Office, Budget for Mazowieckie Voivodship for 2009, World Bank staff calculations. 97. Mazowieckie's revised 2009 budget34 foresees a substantial increase in spending-a nominal increase of 35 percent over the level of budget execution in 2008. Much of the increase would consist of grant programs, largely financed from EU funds. Spending on _enterprise development' would increase from PLN 29 million in 2008 to PLN 295 million in 2009; an 80 percent increase.35 Nearly ninety percent of this would consist of special programs financed by the EU social development fund and channeled through Mazowieckie EU Funds Implementation Unit (Mazowiecka Jednostka Wdrazania Programow Uninych (MJWPU). Another PLN 73.9 million is allocated to the MJWPU under the title _public administration'. Spending on social services would also increase very substantially. Spending on education would double. Spending on social services would quadruple. These increases would be offset by a drop in spending on passenger rail service. (As noted earlier, spending on passenger rail were unusually high in 2008, due to costs associated with the purchase of KM and new rolling stock for the rail line. In 2009, they will revert to more normal levels.) The revised budget also foresees a substantial increase in interest payments service: from PLN 23 million in 2008 to PLN 121.5 million in 2009.36 98. On the revenue side, the budget projects a large increase in the disbursement of grants from the EU. Relative to the amount actually received in 2008, EU grants under the regional operating program are budgeted to increase from PLN 12 million to PLN 352 million. Receipts from the EU social fund are budgeted to increase by a full order of magnitude: from PLN 53 million to PLN 566 million. Revenues from the CIT-the voivodship's principal revenue source, 34 As of the third quarter of 2009. 35 Note that the original 2008 budget allocation for enterprise development was PLN 216 million. Only PLN 29 million was actually spent, an execution rate of 16 percent. 36 One third of the increase is due to growth in spending on the budget item other settlements' (dzial 758). Of this increase, 45 percent is, in turn, due to budgeted increases in fraternal equalization payments. The remainder consists of PLN 57.8 million allocated to various financial settlement', including PLN 39.5 million allocated to refunds of grants from the central government used contrary to designation or collected in excessive amounts' and PLN 145 million allocated to general and earmarked reserves of which PLN 78 million is allocated to provisions for investment projects and investment purchases. 32 are projected to increase at a more modest rate of 13 percent. If the 2009 budget were executed as planned, the increase in revenues would more than offset the budgeted increase in expenditures. As a result, the deficit would decrease to 472 million in 2009; 13 percent of revenue. 99. The voivodship proposes to finance this deficit by borrowing. According to the revised budget's financing plan, the voivodship would borrow a total of PLN 845 million in 2009; PLN 485 million in the form of credits and loans and PLN 360 million, the form of bonds.37 The financing budget also includes of allocation of PLN 467.4 million for the repayment of existing debt. The net proceeds from financing would nevertheless be sufficient to finance the deficit and in fact generate a small cash flow surplus. 100. As a short term fiscal strategy, the revised 2009 budget has proven to be optimistic in its particulars but apparently successful in its overall result. As of the end of the third quarter of 2009, the voivodship had met only 52 percent over its overall revenue target. But it had managed to constrain its expenditures to only 46 percent of the budgeted amount. As a result, the budget was roughly balanced: expenditures exceeded revenues by only 0.6 percent. Proceeds from borrowing (PLN 495 million) were already sufficient to repay the Voivodship's principal repayment obligations through the entire year (although only 72 percent of the total had been repaid by the end of the third quarter). 101. Given the limitation of the available data, it is unclear how this result was achieved. Most importantly, it is unclear how expenditures were constrained to match the shortfall in revenues. It may be that actual spending obligations were reduced in a reasonably rational way: salaries may have been reduced or frozen, low priority capital works may have been cancelled. Or it may be that payment obligation were merely been postponed, allowing arrears to accumulate which will eventually have to be paid. 102. What is the prognosis for longer term fiscal sustainability? The Voivodship's current stock of debt is still within the 60 percent ceiling. The total stock of debt at the end of 2008 was equal to 36 percent of the Voivodship's revenues (or 50 percent of its revenues net of compulsory transfers to the equalization mechanism; a more accurate measure of the voivodship's discretionary income). As of the end of the third quarter of 2009, the stock of outstanding debt had risen to 42 percent of total revenues38 103. The Voivodship has prepared a financial projection running through 2034.39 This projects a somewhat erratic trend in both revenues and expenditures. Revenues would decline by eight percent in 2010; rise by 12 percent in 2011, decline by 15 percent in 2012, and then stabilize (except for a one-time drop in 2016) throughout the rest of the decade. Expenditures would decline 12 percent in 2010, rise by seven percent in 2011, drop by 17 percent in 2012, and then roughly stabilize (against, except for the 2016 drop) through the rest of the decade. This projection yields a small overall surplus by in 2013. Although this would be followed by a deficit (equal to one percent of revenues) in 2014, it would be followed by modest surpluses throughout 37 An additional PLN 111 million would be generated from debt repayments (on loans made by the voivodship to other levels of government) and other sources. 3 Assuming that the level of revenue collection in the first three quarters of the year persists through the last quarter. 39 As the projection was prepared in early 2009, it does not reflect later budget revisions or subsequent fiscal developments. Figures for 2009 in the Figure 15 are based on the revised 2009 budget and the stock of debt as of the end of the third quarter of 2009. Figures beginning in 2010 are based on the Voivodship's projection. 33 the rest of the projection period. The debt stock-measured as a percentage of revenues-would peak in 2016 and would gradually decline thereafter. 104. In the absence of any details on the assumptions underlying these projections, their plausibility is difficult to evaluate. Ultimately, from a fiscal sustainability standpoint, what will matter is not so much the absolute level of revenues and expenditures in any given year but the difference between them. Maintaining a positive gap between revenues and expenditures will require the Voivodship to continue to adjust its expenditures to match the somewhat erratic and largely exogenously determined changes in its revenues. If it can manage this, its fiscal prognosis is good. 105. Regardless of the short term result, the longer term outlook for the voivodship, in any case, will require careful management. The voivodship is vulnerable to any slowing in the rate of recovery in the Polish economy. Its principal sources of revenue-fixed shares of the corporate income tax (and to a smaller extent the personal income tax) are sensitive to fluctuations in economic activity. As the voivodship has no control over the rates of the two taxes, it cannot increase rates to compensate for declines in their bases. 106. Fortunately, the voivodship's largest spending obligation, fraternal equalization, can be expected to fluctuate roughly in parallel with its revenues from the CIT and PIT. Under the current formula, if Mazowieckie's economy outpaces that of Poland as a whole (or more specifically, if the bases of the corporate and personal income taxes grow faster in Mazowieckie than in the nation as a whole) the voivodship's obligations to the fraternal equalization mechanism will increase in absolute terms but its revenues from the two shared taxes will increase even faster.40 Restraining Spending 107. The voivodship, moreover, enjoy considerable discretion over its remaining expenditures. Spending on capital works constituted just over half (52 percent) of Voivodship expenditures in 2008 (net of expenditures on fraternal equalization). Although this figure is distorted by the high level of capital spending on passenger rail in 2008, a large proportion of net spending in future years will presumably consist of discrete capital works, which in principle, could be postponed or rescheduled if CIT and CIT revenues fall short. By the same token, expenditures on EU funded projects can be scaled back if EU funding fails to materialize. (As discussed below, EU projects are no longer financed on a reimbursement basis. The EU will now fund start-up costs in advance, and fund ongoing works on a pay-as-you-go basis.) 108. Spending on salaries-a considerably less flexible category of expenditure-constituted only 13 percent of net spending in 2008. Part of the salary bill, moreover, is financed from earmarked dotacje, which in principle, should increase in response to any increase in wages. As in Warsaw, part of the recurrent costs of education in the Voivodship is financed from earmarked grants from the central government. Despite the vagaries of the allocation process at the central level (noted earlier), education subventions have generally increased in tandem with centrally- negotiated increases in teachers' salaries. 40 This is subject to an important caveat. Since the voivodship's obligations to the fraternal equalization mechanism are based on tax revenues from two years earlier, a sharp fall in tax revenues can result in the Voivodship having to increase its contributions even as its revenues from shared taxes are falling. 34 109. There are some specific sectoral areas where cuts seem feasible. Financial vulnerabilities could, for example, be reduced by cutting railway subsidies. As in Warsaw, the source of the railway companies operating losses lies partly in the low level of tariffs and partly in the inefficiency of the companies themselves. Mazowieckie, like Warsaw, pays the operating companies on the basis of a service contract, which is expressed in vehicle-kilometers. Any shortfall between ticket revenues and operating costs must be made up by the Voivodship, undermining any incentive the operators might have to increase revenues or reduce operating costs on their own. The voivodship should also reexamine its transport subsidy policy from a multimodal perspective. Rather than focusing exclusively on railway subsidies, it should review the entire range of public transport options to find the most cost-effective means of meeting its social objectives. For instance, if regional railway is not economically viable for some low- traffic corridors, less expensive options such as inter-city bus services could be introduced instead. 110. In principle, solutions to the looming debts of the Voivodship's hospitals also lie partly within the Voivodship's domain. In April 2009, the Council of Ministers adopted a program aimed at providing financial assistance to indebted hospitals. The program, -Support for Regional and Local Self-Government Units with Regard to Activities that Stabilize the Health Care System", allows subnational governments to convert their hospitals into corporate entities operating under the Commercial Code. Hospitals that do so are eligible for debt forgiveness on their outstanding obligations to State entities. To qualify for assistance, each hospital must obtain the approval of the national health insurance fund (NFZ) and submit a satisfactory business plan to the state bank (BGK). In principle, this program would allow the Voivodship to offload part of its potential hospital debt liabilities. But the program only finances debt to state entities, not those to private suppliers. Moreover, it does not preclude the emergence of new liabilities in the future. This would require more fundamental reforms in the hospital sector. According to a recent study by the Gdansk Institute41 the persistent financial problems of Polish health institutions are due to a combination of inadequate reimbursement levels from the NFZ and high levels of inefficiency in the institutions themselves. Reforms at both levels would be required to remove this threat to the Voivodship's finances. Targeting Capital Investments 111. The voivodship can also improve the quality of capital spending. Like Warsaw, the Voivodship has ambitious investment plans. The Regional Operating Program for Mazowieckie 2007-201342 projects a total of C 3,032 billion in capital spending over the six year period (PLN 12.35 billion at the 2006 exchange rate) of which sixty percent would be financed by EU funds. The budget for 2009 alone allocates a total of PLN 1.288 billion for capital investment in 2009, of which PLN 329 would be financed by the EU.43 Slowing growth in corporate income tax revenues, combined with slower than expected disbursement of EU funds, threaten these plans and increase the urgency for reforms in the system of capital investment planning, evaluation, and funding. 41 Centralized financing of the health care and education in Poland" - Regional Analysis, Warsaw 2008 42 Szczogolowy Opis Priorytetow Regionalnego Programu Operacyjnego Wojewodztwa Mazowieckiego 2007-2013 (2008) 43 This is a subset of the PLN 407 million in ROP funds anticipated in the 2009 budget. 35 112. In principle, capital investment plans in Mazowieckie derive from the Voivodship's regional development strategy (Development Strategy for the Mazowieckie Voivodship for the Year 2020 (updated) and its companion Spatial Development Plan. Both were issued in 2006. As discussed in the Annex, the economic principles of the Development Strategy makes sense. While noting that economic disparities exist among the subregions of the voivodship, the strategy and plan call for the reinforcement of Warsaw's role as the economic engine of the region. 113. The Strategy and Plan do not, however, provide much detail on project-specific investment priorities or timing. While they do specify certain macro-transportation priorities, including the Greater Mazovia Ring Road, the three pan-European transit corridors running through region ( I Helsinki-Warsaw, II Berlin-Nizhniy Novgorod, and VI Gdansk-Ostrava) and their complementary radials (Warsaw-Radom and Warsaw-Plock) they call in only very general terms for the investments in the modernization of rail lines, the reconstruction of high voltage transmission lines; new connections to natural gas pipelines; and investments in bulk water production, sewerage and sewage treatment in Greater Warsaw. 114. In principle, the Regional Operating Program for Mazowieckie provides a more specific, time-bound list of investment priorities. The ROP is the product of negotiations between the voivodship, the central government, and the EU. As per EU regulations, the ROP sets out seven priority areas (plus an eighth, which provides the funding for program administration). These are in turn divided into 27 activities. Table 6 lists the 27 activities, ranked by funding level. As shown, the largest is road infrastructure, followed by development of entrepreneurship', urban renewal, and improvenent of information." Within each activity, funds are allocated in two modes. First, a set of so-called _hrge' and _key' projects are identified. These are specifically listed in the ROP. In principle, these are distinguished not by size but by their essential contribution to the development of the region'. Projects on the list have a prior claim on ROP funds. Although they are still required to go through detailed appraisal, they are not subject to competitions (see below). Of the 27 large and key projects listed in the ROP, most (although not all) fall under the purview of the voivodship. 115. The other form of allocation consists of a series of competitions, held at periodic intervals for specific activities.44 The procedures governing competitions for the 2004-2006 plan period were subject to considerable criticism. Thousands of individual projects were received. Each was evaluated in isolation. The process of technical evaluation was cursory. The level of political evaluation, on the other hand, was reportedly thorough. The procedure also entailed high costs for potential participants. Candidates were required to present fully prepared projects at the initial stage of competition. As a result projects that had no reasonable prospect of funding nevertheless had to undergo substantial project preparation costs. 116. Procedures in Mazowieckie have been substantially revised for the 2007-2013 period. Under the new approach, candidates are required to submit only a short (6 page) project concept at the initial stage. This is then subject to multi-stage evaluation. The first, so-called _frmal' The impact of the procedures governing the EU funds goes beyond the EU funds themselves. The Voivodship also operates its own Mazowieckie Development Fund (MDF), which is financed from the voivodship's own revenues (i.e., the CIT). The MDF is available only to gminas and powiats that are eligible for equalization subventions (i.e., that have per capita revenues less than 92% of the national average) but are otherwise allocated on the same basis as the regional fund. 36 evaluation, determines if all the required forms had been submitted correctly. (This is apparently a major hurdle for some candidates. Candidates have one additional chance to get their forms right.) This is to be followed by a so-called strategic evaluation, to determine if the proposal is consistent with the regional strategy, and a technical evaluation of the project's merits. A maximum of 35 points can be rewarded for consistency with the regional strategy; sixty points for technical merit. 117. The voivodship has taken Table 6. ROP Allocations by Activity (in order of size) steps to improve the rigor of the C __ __Aqil____._______ evaluation process. A 40-person 3.1 road infrastructure 522.2 X bureau of regional planning has 1.5 development of entrepreneurship 473.8 X been established to undertake 5.2 urban renewal 220.0 X evaluation of major regional 2.1 improvement of information 193.1 X projects. The voivodship also 6.2 Tourism 171.6 X 1.4 business environment 164.0 X eplys appoxmae 500 s 4.3 air and energy 146.8 a te t roets 7.2 education infrastructure 127.1 X 4.1 sewerage and SWM 110.0 X (although their evaluations are 6.1 culture 97.7 x largely limited to the so-called 3.2 regional public transport 96.4 frmal' evaluations). Technical 7.1 health infrastructure 90.0 x evaluations are performed by 1.3 infrastructure for new activities 90.0 X consultants, drawn from a 4.2 land protection 77.4 X national list of experts maintained 2.2 development of services 65.3 x 1.7 economic promotion 57.8 1.1 research and development 57.0 Development. 3.3 airports 50.4 118. At the conclusion of this 8.1 project administration 48.4 4.4 environmental protection 27.0 X poce all proet al re 1.6 regional cooperative networks 26.3 X to r e that receiv 1.8 business technology 23.4 X 2.3 IT for SMEs 23.2 X maximum number of points 1.2 economic/scientific cooperation 22.5 X (generally 100) are to be 5.1 urban transport 17.6 submitted to the marshal's office. 7.3 social assistance infrastructure 17.3 X The marshal has the authority to 8.2 publicity 16.3 add five points to the ranking, but TOTAL 3032.5 otherwise must approve projects X=competition completed in the order in which are x=competition process started the Source: Mazowiecka Jednostka Wdrazania Programow Uniinych (MJWPU). tentatively approved, in order of rank, until the money allocated to the particular competition is exhausted. Projects that earn at least 60 percent of the maximum points but for which funding is not sufficient will be assigned to the so-called freezer'. b The bureau's evaluation methodology remains somewhat opaque, however. The regional planning bureau uses a so-called multi-criterion decision making matrix' to evaluate each project. The criteria it maximizes against are: competitiveness, cohesion and integration' 37 119. At this point, detailed feasibility studies are to be undertaken for projects that have been tentatively approved for funding. The feasibility studies are to be evaluated by the technical experts. The criteria applied at this stage include economic rates of return, although the methodology to be used in these calculations is not standardized, as yet. Proposals that pass muster at this stage are to be approved by the marshal, who will then sign contracts with the beneficiaries. Projects that do not pass muster at this stage would be sent to the freezer. 120. On paper at least, this new approach appears to solve several of the problems associated with the 2004-2006 allocation process. The new two-stage evaluation process will reduce the initial cost of participation. The use of rosters of technical experts may improve the quality of technical evaluations. But it is not clear that these problems will be solved in practice. As shown in Table 5, sixteen competitions have been completed and another three have been launched. Their results have not yet been evaluated. 121. In identifying, evaluating, and funding projects under its direct control, the Voivodship could also benefit from some of the recommendations applied earlier to Warsaw. Project identification could be improved by strengthening the data base on which investment decisions are made. Project evaluation methodologies could be standardized and made less susceptible to ad hoc political influence. 122. Steps could be taken to improve funding continuity. In this respect, the Voivodship should pay particular attention to the problems it has had in absorbing EU funds. In the period 2004-2006, Mazowieckie had the lowest absorption rate in Poland. At the end of the third quarter of 2008, only 89 percent of the 2004-2006 allocation had been used. The record does not appear to have improved. As of the end of the third quarter of 2009, the voivodship had absorbed only 33 percent of the amount allocated to the ROP. 123. There are several possible explanations for the low rate of EU funds absorption. According to a recent OECD study, initial absorption problems were a result of _public scrvants' inexperience'- a problem the OECD reports is now largely resolved. The report blames current disbursement problems on a _lak of collaboration between private and public actors', and recommends reducing administrative barriers to private-public collaboration'. It also recommends strengthening the political role of voivodships (perhaps through the direct election of marshals) to increase their ability to play a _strAegic and arbiter role in regional development'. 124. At a more prosaic level, several of the case study cities studied for this report have claimed that they cannot undertake EU-financed projects because of the EU's pre-financing requirements. It is true that under the 2004-2006 program, EU funds could only be disbursed retroactively. Local governments were, in principle, required to provide construction financing. Even during that period, however, EU rules permitted disbursements in tranches (i.e., against completion of specific stages in construction, rather than completion of the work in its entirety). In addition, the Government bank, BGK, provided construction financing for transport projects. Since the adoption of the EU's anti-crisis package, financing arrangements have become even more liberal. As of February 12, 2009, EU funds can be paid in advance for up to 95 percent of the project's value. Funding for large projects can be released even before the EU has issued its formal approval. Pre-financing requirements are therefore no longer a constraint on the absorption of EU funds, if they ever were. 125. Inconsistencies between Polish environmental laws and those of the EU guidelines have also reportedly delayed project approvals. Proposals for road project were held up, until recently, 38 because although they were consistent with Polish law they failed to meet EU requirements. As the Polish law has now been brought into EU compliance, this is reportedly no longer a constraint. 126. The final and most plausible explanation for the slow absorption of EU funds is a human and institutional one. According to the Mazowieckie EU project implementation unit director, local governments lack the capacity to prepare proposals and payment applications. At the same time, the low reimbursement rate for technical evaluations (PLN 200 per evaluation) makes it difficult to attract qualified experts. These problems have solutions. The incapacity of local government to prepare applications and disbursement requirements could be addressed through increase training and field visits by Voivodship staff. Reimbursement levels for technical evaluators could be increased. Siedce, Plock, and Radom 127. Like Warsaw, the three case study cities (in their roles as both gminas and powiats) are responsible for a wide range of service responsibilities, including primary schools and secondary education, various social services, and the construction and maintenance of gmina and powiat roads. Like Warsaw, they operate public utilities, including those responsible for water, sewerage, street cleaning, solid waste management, and district central heating. (As in Warsaw, these are off-budget enterprises.) Revenues 128. The three cities derive virtually all their budgetary revenue from central government transfers. Figure 20 illustrates the major sources of revenues of each of the cities in per capita terms (based on averages of 2006, 2007 and 2008). As shown, shares of the PIT and CIT made the largest contribution to revenues, generating about PLN 1200 per capita in Siedlce and Radom, and about twice that figure in Plock (presumably due to the presence of an oil refinery and related industries). Receipts from the education subvention are the second largest source, generating about PLN 900 per capita in all three jurisdictions. Subventions to cover the costs of social assistance comprised much of the remainder. In total, shared taxes and transfers accounted for about 85 percent of each of the three cities' revenues over the 2006-2008 period. Expenditures 129. As shown in Figure 21, education is the largest single item of the each of the cities' expenditures, although it accounts for widely varying proportions of the total (42 percent in Siedlce, 36 percent in Radom and 28 percent in Plock). As in Warsaw, revenues from the education subvention fall short of the city's actual education expenditures, although the shortfall is considerably smaller in the three case study cities than in their larger neighbor. Over the 2006- 2008 period, education subventions to Siedlce were, on average, four percent higher than the city's recurrent expenditures on kindergartens, primary schools, secondary schools, pupil transportation and teacher training, and roughly 75 percent of the city's total education expenditures. Similarly, the education subvention was more than sufficient to cover the recurrent costs of kindergarten, primary, and secondary education and related costs in Radom, although not so in Plock. (See Figure 22A.) 39 Figure 20. Sources of Revenues in Siedkce, Figure 21. Expenditures in Siedkce, Plock and Plock and Radom in per capita terms, average Radom in per capita terms, average 2006-2008 2006-2008 5,000 5,000 4,500 Other 4,500 m other 4,000 N Transport 4,000 0 public transport 3,500 I Public transport 3,500 3,000 0 Housing 3,000 m communal services 2,500 U Social asst 2,500 m public admin 2,000 *trsbetos2,000 rod 1,500 1,500 . roads M Equalization subvention 1,000 1,000 U social asst 0 Education subvention 500 500 idi PIT,CIT U* education Siedlce Plock Radom Siedlce Plock Radom Source: Central Statistical Office, Ministry of Finance data, World Bank staff calculations. 130. Spending on social assistance-chiefly the centrally mandated child allowance- accounts for ten to twenty percent of expenditures in the three cities. As shown in Figure 22B, 60-75 percent of spending on social assistance (and a presumably higher percentage of spending on the specific centrally-mandated social assistance programs) is covered by the corresponding central government earmarked transfer. Figure 22. Spending on Education, Social Assistance and Transport in Siedkce, Plock and Radom A: Percent of Education Costs Covered by Subvention B: Percent of Social Assistance Costs Covered by Subvention m ed grant/total education M ed grant/social education 80% 140% 70% 120% 60% 100% 50% 80% 40% 60% 30% 40% 20% 20% 10% 0% 0% Siedlce Plock Radom SiedIce Plock Radom 40 C: Percent of Public Transport Costs Covered by Tariff Revenue . revenue/expenditures 0 as %expenditures 120% 100% 80% 60% 40% 20% 0% N w Siedlce Plock Radom Source: World Bank staff calculations based on Central Statistical Office and Ministry of Finance data. 131. Spending on roads consumed an average of slightly less than ten percent of expenditures in Siedlee and Radom and roughly twice that proportion in Plock, over the 2006-2008 period. As it has no corresponding dedicated revenue source, recurrent spending on this function is financed from general revenues. Subsidies to public transport constitute only a minor proportion of the expenditures of the three cities-unlike the case in Warsaw. As shown in Figure 22C, the proportion of transport costs covered by tariff revenues varies widely among the three cities. Siedlce does not subsidize its public transport services at all. While Plock and Radom do, such subsidies constitute less than two percent of either city's expenditures. Fiscal Prospects 132. Like Warsaw, the three case study cities would appear to be vulnerable to economic downturns. Their principal sources of revenue-fixed shares of the personal and corporate income taxes-are sensitive to fluctuations in economic activity. Because education subventions (and to a smaller extent, the social assistance dotacje) loom so large in their revenue budgets, they also vulnerable central government decisions over the funding of these Figure 23. Distribution of Expenditure by transfers. Economic Category, average 2006-2008 133. The three cities do not have as 100% 90% much flexibility on the expenditure side as 0% their larger neighbor. Unlike Warsaw, the 70% .debtservice three cities' budget have a relatively high 60% N capital works proportion of fixed expenditures. As shown 50% U goods and service in Figure 23, salaries and related social 40% N wages and benefits contributions accounted for nearly forty 30% * social asst transfers of ota epenitues20% N transfersto enterprises percent of total expenditures in Siedlce and 10 Radom, and nearly thirty percent of 0% expenditures in Plock. The proportion Siedlce Plock accounted for by goods and services is Source: Central Statistical Office, World Bank staff calculations. relatively small3slightly less than one quarter in all three cases. And unlike Warsaw, the capital budget of the Siedlce and Radom is 41 relatively small. Capital investment accounted for only 15 percent of total spending in Siedlce and Radom over the 2006-2008 period. (In Plock, on the other hand, the proportion was 28 percent.) As a result, Siedlce and Radom would appear to be particularly vulnerable to changes in the fiscal environment. 134. The two charts in Figure 24 illustrate the scale of their vulnerability. Figure 24A illustrates the impact of a five percent decline in revenues from shared taxes and education subventions in 2009 and 2010 on each city' current account surpluses.46 This illustrates the level of revenues that would be left to finance capital expenditure under the assumed scenario. As shown, the current surplus would fall dramatically in all three cases, falling to 15 percent of revenues in Plock and to only seven percent of revenues in Siedlce and Radom. Figure 24B shows the impact of the assumed drop in revenues on the overall balance of the three jurisdictions. As shown, all three were currently running overall deficits in 2008. Assuming capital expenditures remain at their present levels, the deficits would grow to 12-15 percent of revenues over the next three years. This would clearly not be sustainable. Figure 24 Impact of 5 percent revenue drop on Fiscal balances of Radom, Plock and Siedlce A: Recent and Projected Trends in Current Balance, Case Study B: Recent and Projected Trends in Overall Balance, Case Study Cities Cities 30% 10% 25% 5% 20% -4-radom 2006 007 0 2009 2010 2011 -4-radom 15% ---plock -5% ---plock 10% siedlce siedIce 5% -10% 0% 2006 2007 2008 2009 2010 2011 -20% Source: Central Statistical Office, World Bank staff calculations. Long Term Reforms 135. All five of the jurisdictions reviewed in this report operate, to varying degrees, within a fiscal straitjacket. Neither the voivodship, the city of Warsaw, nor the three smaller cities have significant control over the level of revenues they receive. They have only limited control over largest single item of their expenditures: salaries. As a result, they are in a difficult position. They bear the risk of any shortfall in revenues. They also bear the risk of any increase in salary levels arising from centrally negotiated collective bargain agreements. And because they have virtually no control over their revenues and limited control over their expenditures, their ability to manage these risks is constrained. To date, their response to the fiscal crisis has been to slash spending in categories they do control-principally capital works-and to borrow. As noted earlier, this is not sustainable. While all five jurisdictions still have considerable scope to cut 46 Defined as current revenues less current expenditures. These projections treat all existing debt service as a recurrent expenditure. Due to the low level of debt in all three jurisdictions, this does not affect the result. 42 expenditures, over the long term, the Government of Poland might consider more fundamental structural reforms that would loosen the straitjacket itself. These could take two forms. 136. Increasing Revenue Autonomy The first would be to loosen the reins on revenue. The most attractive candidate for liberalization would be the personal income tax. It has a number of advantages, not the least of which is the fact that it is already assigned to subnational governments. As described in Box 6, it also functions well in its economic role as a benefit tax. In Poland, liberalization would merely require that local governments be allowed to adjust the rate of the tax; specifically the rate on that proportion of the tax retained in their jurisdictions. This is already the practice in countries in Western Europe. Local governments in Scandinavia have long had the power to adjust the rate of local personal income taxes. This practice has more recently spread to Spanish regions, Italian regions and municipalities, and counties and municipalities in Croatia. In principle, local autonomy over PIT rates could complicate the task of revenue equalization. But this problem is readily solved. For purposes of calculating equalization transfers, PIT revenues could be assumed to be imposed at a nationally standardized rate. An individual local government could then be given the authority to impose a surcharge on the gross PIT receipts collected within its jurisdiction, with the revenues retained in the municipality rather than being subject to equalization. Box 6 Choosing Among Local Taxes Local taxes, in principle are an appropriate means of financing services whose benefits cannot be confined to individual consumers, but nevertheless do not extend beyond the municipal boundaries. In a sense, local taxes are the collective analogue of user charges. W here user charges are the means by which individuals can express their demand for services whose benefits are largely private, local taxes are the means by which taxpayers within a community can express their demand for services that are consumed collectively. Benefit taxes also provide a degree of geographical equity to the financing of municipal services, imposing the costs of municipal services on the people who benefit from them. Property Taxes. The property tax is the most common local tax. As a local incidence tax, it performs well: the incidence of the tax (at least its residential component) is geographically confined and its coverage is broad enough to reach the majority of beneficiaries. The property tax is, however, difficult to administer successfully. In part, the reasons are technical: the number of taxpaying units is large. Assessments are often controversial because values must be imputed, rather than observed from actual transactions, and the base must be revalued annually to maintain its buoyancy. (Conditions in transitional economies make the property tax particularly difficult to administer, since the basic market data on which the property assessments are based is inaccessible or unreliable.) But there are also particular political costs associated with the property tax. As a direct tax, its burden is readily visible. Unlike the personal income tax, it cannot be withheld at source. And as a tax on wealth held in the form of real estate, its burden tends to fall disproportionately on groups with great political influence. Income and Payroll Taxes. Local income and payroll taxes are an attractive alternative to the property tax. Personal income taxes have the geographically defined incidence sought in a benefit tax, and where imposed in conjunction with a national income tax, are cheap to administer. The constraint on income taxes in some countries is their narrow coverage. Many countries impose the income tax only on formal sector employment, collecting revenues through payroll deductions from firms rather than directly from 47 K. Davey, et.al. The Impact of the Economic Downturn on Local Governments, What is Happening and What Can Be Done About It? Report by the Open Society Institute to the Council of Europe 2010. 43 individuals. In small towns with little formal employment, such payroll taxes fall on too small a proportion of the population to serve effectively as benefit taxes. Utility Taxes. Local taxes imposed in connection with utility bills--particularly with electricity bills--are also not uncommon. In principle, such taxes have localized incidence and--given the ubiquity of electric service--reasonably broad coverage. Costs of administration are low, as the tax can be imposed jointly with electricity bills. The extent to which an electricity surcharge can be relied exclusively is limited, however, by the basis of assessment: if the tax is imposed ad valorem, it distorts the price of power. If it is imposed as a flat fee it fails to capture variations in ability to pay, and is therefore limited by the tax paying ability of the poorest electricity consumers. Taxes on Automobiles. Local automobile taxes--both recurrent taxes on automobile ownership, and one- time taxes on purchase--are also significant auxiliary sources of revenue. Except in multi-jurisdictional metropolitan areas, the incidence of such taxes is easy to confine; such taxes are relatively easy to administer, and enforce. Potential yields, however, are limited. 137. The Government could also relax controls on the assessment of property taxes on residential properties. While the current assessment methodology (based on a fixed amount per square meter, adjusted annually for inflation) is administratively straightforward, the current rate is derisory. The Government could raise the rate, as well as add additional variables to the assessment formula in order to better reflect differences in the market values of different properties. 138. Expenditure Autonomy There may also be a case for relaxing central regulations on expenditures-particularly on personnel. This could include facilitating dismissal of redundant staff and allowing individual municipalities to adjust wage levels according to local economic conditions and spending priorities, rather than conforming to the terms of nationally-negotiated collective bargaining agreements. Such measures are likely to be controversial. Public sector unions are likely to forcefully defend their benefits; particularly protections against dismissal. Moreover, public sector unions typically prefer to negotiate wage agreements with national governments, where the full force of the national union can be brought to bear, rather than negotiating individual agreements with each municipality where more parochial interests might prevail. Reform is nevertheless possible. One approach , for example, would be to reduce job protection only for newly hired staff. 44