99400 Report No. 99400-PE PERU Building on Success Boosting Productivity for Faster Growth GMFDR Document of the World Bank © 2015 International Bank for Reconstruction and Development / International Development Association or The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@ worldbank.org. Cover and design credits: Florencia Micheltorena and Alejandro Espinosa (cover), The World Bank. Contents Background and acknowledgments 1 Executive summary 3 Introduction 13 Part I. Understanding Peru’s growth story 17 1. Peru’s growth success 19 2. Growth in Peru—drivers and challenges 24 Part II. Looking closely at firm productivity in Peru 31 3. Firm dynamics and productivity 33 4. Issues with factor allocation 40 Part III. Correcting misallocation of resources to raise productivity growth 45 5. Addressing regulatory and competition issues 47 6. Streamlining the labor market and building skills 53 7. Improving access to credit for enterprises 59 Part IV. Other opportunities to reduce misallocation and further boost productivity growth 65 8. Using spillovers from international trade 67 9. Unleashing innovation 77 Part V. Conclusions 85 Annex. Background papers prepared for this report 88 References 90 Abbreviations BCRP Banco Central de Reserva del Perú INEI Instituto Nacional de Estadística CAF Corporación Andina de Fomento e Informática CEBB Comisión de Eliminación IPR Intellectual Property Rights de Barreras Burocráticas LAC Latin America and the Caribbean Region CEDLAS Centro de Estudios Distributivos, LACPOV Latin America and the Caribbean Region Laborales y Sociales WB poverty database CENEC Censo Nacional Económico LHS Left-hand Side CEPAL Comisión Económica para América LPI Logistics Performance Index Latina y el Caribe MEF Ministerio de Economía y Finanzas CITE Centros de Innovación Tecnológica MFN Most Favored Nation Empresarial MIC Middle-Income Country CONCYTEC Consejo Nacional de Ciencia, NBFI Non-bank Financial Institution Tecnología e Innovación Tecnológica NTM Non-Tariff Measures EEA Encuesta Económica Anual OECD Organization for Economic Cooperation ENAHO Encuesta Nacional de Hogares and Development EVA Exports Value Added PISA Programme for International Student FIDEICOM Fondo de Investigación y Desarrollo Assessment para la Competitividad PMR Product Market Regulation FINCyT Fondo para la Innovación Ciencia PWT Penn World Table y Tecnología R&D Research and Development FOMITEC Fondo Marco para la Innovación, REDES Red de Ecología Social Ciencia y Tecnología RHS Right-hand Side GCI Global Competitiveness Index RIA Results Impact Assessment GDP Gross Domestic Product RICYT Red de Indicadores de Ciencia GFDD Global Financial Development y Teconología Database SEDLAC Socio-Economic Database for Latin GIPC Global Intellectual Property Center America and the Caribbean GMM Gaussian Mixture Models STRI Services Trade Restrictiveness Index GVC Global Value Chain SUNAT Superintendencia Nacional de IDB Inter-American Development Bank Administración Tributaria IMF International Monetary Fund TFP Total Factor Productivity INDECOPI Instituto Nacional de Defensa de la WDI World Development Indicators Competencia y de la Protección de la WEF World Economic Forum Propiedad Intelectual WITS World Integrated Trade Solution Background and acknowledgments This report has been prepared at the request of the Government of Peru, as part of the Road to Lima process leading to the Annual Meetings of the World Bank Group and the International Monetary Fund (IMF) in Lima, Peru, in October 2015. Other activities of the Road to Lima have been part of its preparation, as have intense technical consultations. The report, by design, is focused on diagnosing selected microeconomic constraints to accelerating productivity growth, while acknowledging that the country may also have other development challenges that are not covered. The report also tries to provide policy directions with the same focused approach. The report was prepared by a World Bank team led by Ekaterina Vostroknutova (GMFDR) under the guidance of Alberto Rodriguez (Country Director, LCC6C), Pablo Saavedra (Practice Manager, GMFDR), and John Panzer (Director, GMFDR). It is based on 25 background papers written by multi-sectoral and multi-Global Practice (GP) and Cross-Cutting Solutions Area (CCSA) teams (a full list of contributors is in the Annex). The following team members led work on the specific topics and sections of the report. Leonardo Iacovone (T&C GP) led a team working on firm-level and productivity analysis that included Trang Thu Tran, Roberto Fattal, Reyes Aterido, and Siddharth Sharma. Thomas Farole (T&C JP and Jobs CCSA) coordinated the team working on trade-related issues, which included Sebastian Saez, Denisse Pierola, Ana Fernandes, Guillermo Arenas, Bruce Fitzgerald, and Erik Van Der Marel. David Robalino (Jobs CCSA) led the team working on labor market issues that included Elizabeth Ruppert Bulmer, Reyes Aterido, Dino Merotto, Mathilde Perinet, Angela Elzir, Tamara Arnold, and Adrian Garlati; the human capital section was based on inputs from Ines Kudo and Miguel Székely; 1 Jennifer Keller was responsible for the corresponding section. Martha Licetti (T&C GP) and Donato de Rosa (MFM GP) led the work on competition and related regulatory policy issues; Alvaro Quijandria led the IFC team working on business regulation; Rong Qian was the main author of the corresponding section; this team included Lucía Villarán, Tanja Goodwin, Karina Rodriguez, Rachel Li Jiang, Congyan Tan, Ernesto Franco Temple, and Jessica Michelle Victor. The innovation section was produced by Rong Qian based on inputs from Ha Nguyen and Pluvia Zuniga; the team also included Patricio Jaramillo and Ekaterina Vostroknutova. The financial sector section was prepared by Daniel Barco (MFM GP) under the guidance and inputs of Steen Byskov (F&M GP). The macroeconomic and growth story was crafted by Ekaterina Vostroknutova and Faruk Khan drawing on inputs from Cristina Savescu, Daniel Barco, and Harry Moroz. Research assistance was provided by Melanie Laloum, Miguel Saldarriaga, and Teresa Peterburs. Administrative assistance was provided by Silvia Gulino and Patricia Chacon Holt. Writing and editing support was provided by Bruce Ross-Larson, Bruce Fitzgerald, and John Burgess. The report has benefited greatly from comments, advice, guidance, and technical discussions with Norman Loayza, Augusto de la Torre, Daniel Lederman, Carlos Silva Jauregui, Jorge Araujo, Marcelo Selowsky, Ivailo Izvorski, Vinaya Swaroop, Edgardo Favaro, and many others. The report owes a lot to the government agencies that supported the team with provision of data, comments, guidance, and advice; the list includes but is not limited to the Ministry of Economy and Finance, the National Institute of Statistics and Informatics (INEI), the Central Bank (BCRP), Ministry of Labor, Ministry of Production, and Ministry of Trade. Vice President: Jorge Familiar Country Director: Alberto Rodriguez Senior Practice Director: Marcelo Giugale Practice Director: John Panzer Practice Manager: Pablo Saavedra Task Team Leader: Ekaterina Vostroknutova Executive summary Executive summary Growth with shared prosperity improved labor incomes rather than redistribution 3 P policies—the latter explain only 15 percent of the eru has emerged as a new growth star in the poverty reduction. Latin America and Caribbean (LAC) region. The economy surged at an average of 6.4 These successes have been driven by strong percent per year during the last decade, the second- macroeconomic and structural reforms over the last fastest growth performance in the region. Over the 20 years, and supported by highly favorable external same period, Peru doubled its per capita income, conditions over the last decade. Beginning in the early well ahead of the region as a whole, which increased 1990s, Peru adopted an ambitious mandate of reform. this figure by only half. This has been an impressive Macroeconomic stabilization included a more flexible performance by global standards as well (Box ES1). exchange rate regime, inflation-targeting, fiscal discipline, and continued debt reduction. Structural The nation’s economic growth has been widely reforms covered areas such as financial liberalization, shared among its 30 million people. Since trade, and product and factor market regulations (Box 2000, almost a quarter of Peru’s population has ES1). As a commodity exporter, Peru also benefited escaped poverty. Across the country, lower-income significantly from the commodity boom, particularly households—the bottom 40 percent—have gained between 2004 and 2013. more from growth than the national average: their per capita income rose by about 6.8 percent per year Peru’s convergence with higher income levels has over the last decade, against a 4.4 percent average for accelerated. From the 1960s to the mid-1990s, the whole population. As with national growth, these convergence was slow, even negative at times. gains were high by regional (and global) standards. During that period other countries —including East Moreover, inequality plummeted during the period at Asian middle income countries (MICs)— continued one of LAC’s fastest rates: 12.6 percent, against a 5.3 converging, leaving Peru behind. But in the last decade percent regional average. Importantly, growth was the Peru’s income per capita has been catching up rapidly main driver of reduced poverty and inequality through with those of high-income countries and—albeit from Peru Building on Success: Boosting Productivity for Faster Growth Box Reforms and prosperity in Peru In the early 1990s, Peru was caught in an the central bank public; and expansion of economic and social crisis. Output had plunged international trade, through agreements with by a quarter, public debt was ballooning, and major trading partners. ordinary citizens were struggling to survive amid hyperinflation that peaked at 7,650 The Third Round began in the mid-2000s with percent. Aside from political and security measures focused on strengthening human measures, to confront these forces, Peru capital and fiscal reforms, through such embarked on a deep and wide reform plan steps as better career rewards for teachers, to target multiple institutions, systems, and tax reform, and more accumulation options practices—public and private. Indeed, few in the pension system; and closing gaps in countries have attempted a shake-up so infrastructure via, for example, multiyear thorough. targets and a framework for public–private partnerships (PPPs). Major accomplishments of the First Round of reforms, which promoted macroeconomic In 2014-15, and in the context of a strategy stabilization and efficiency of markets, to cope with the slowdown in output, this included the creation of an independent central round of reforms continued with measures bank, with limits on how much it could lend to that encompass lower business costs of entry, the public sector; a liberalized trade regime operation, and exit, including reducing red and a more flexible exchange rate regime; tape and simplifying taxes; initial measures restructuring of external debt; lower food and to making the labor market more flexible; and 4 utility subsidies; and the introduction of a a simplification of insolvency proceedings. framework for competition law. The pension This round also tackles some initial aspects system was also overhauled. of trade facilitation, including reducing tariffs for some intermediate goods and supporting The beginning of the 2000s brought the firms to accredit and certify products in order Second Round, with the following basic goals: to reach new export markets. It has also promotion of transparency and accountability initiated regulations to accelerate a prudent in the macro-fiscal framework, creation of a implementation of PPPs. fiscal stabilization fund and fiscal rules, and elimination of full-salary pensions for certain There is no question that further reforms will civil servants; control of inflation, by setting be needed, but Peru can feel proud of its solid an inflation target and making decisions of reform track record. ES Fi ur 1. R forms l d to st ll r p rform nc in ES Fi ur 2. Growth cut pov rt incom p r c pit , r l tiv to th r ion nd th world nd in qu lit sh rpl (real average annual growth, percent) (percent) 3.5% 25% 3.0 Peru LAC average 3.0% 20% 2.5% 15% 2.2 2.0 10% 2.0% 1.7 5% 1.5% 1.3 0% 1.0% -5% 0.5% 0.2 -10% 0.0% -15% Peru World LAC World LAC Peru Percent of population lifted Percent change in 1961-1990 1990-2013 out of poverty 2000-2013 Gini coe cient 2004-2013 Sources: World Development Indicators (WDI) and Background paper prepared for this Report by Céspedes and Lengua- Lafosse (2015), Central Bank and Ministry of Economy and Finance of Peru. Executive summary a low base—is matching the convergence speed of and apparel) accounted for 91 percent of merchandise other fast-growing MICs like Malaysia and Thailand. exports in 2014, virtually unchanged from 1994, with But as with other MICs, Peru now faces the hardest minerals having a significant share. In the context of part—moving into high-income status. a seemingly non-temporary adjustment in commodity prices, growth in high value-added manufacturing and Increasing productivity is the fastest services would help to support sustained growth. way to boost growth, particularly without external tailwinds Structural changes contributed to productivity growth in the last decade. Growth since the 1990s In recent years, high labor participation and capital has been broad across economic sectors, with industry accumulation helped to fuel growth. The country and services leading in the 2000s. Breaking down achieved substantial labor accumulation, supported aggregate labor productivity (measured as value added by a strong demographic dividend and high labor- per worker) into productivity growth within economic force participation that reached 73 percent by 2013, sectors and that due to structural shifts in employment higher than that in Malaysia, and Chile, and above shows that starting in the 2000s structural shifts the Organization for Economic Cooperation and contributed about 20 percent of overall productivity Development (OECD) average. Capital accumulation growth. Labor moving from agriculture to services was was at the level expected for Peru’s income per capita. the principal source of labor productivity gains from There could also be significant gains by improving the structural shifts, although firm-level data indicate quality of those inputs, as it is discussed later in this that the services sector has not been as productive as Executive Summary. it could have been. Peru saved the windfall of the commodity boom. The large informal sector is a drag on labor Savings increased from just above 10 percent of productivity growth. While only 21 percent of GDP Gross Domestic Product (GDP) in the early 1990s to is produced informally by official estimates, by some 5 22 percent in 2014, driven by both private and public measures 70 percent of workers are informal.1 Labor sectors. These helped finance a sharp increase in productivity in informal sector is lower (in manufacturing investment rates, well above regional levels. In that, it is only a quarter of its formal counterpart’s), dragging Peru resembles most of the successful fast-growing down the whole economy’s potential. High informality MICs in East Asia, such as Korea in the 1990s. But it may also be a reflection of constrained productivity differs from regional comparators which consumed growth, which in turn means that job opportunities are most of the commodity windfall and financed not created in the formal sector at an adequate pace, investment through foreign savings. Moreover, external and resulting in informality. savings that helped finance growth were mostly FDI, since Peru reduced its debt. These outcomes have Recent growth has been increasingly driven by made Peru more resilient and provided strong buffers productivity gains, comparable to rates in other to negative external shocks. fast-growing MICs globally. Before the 1990s, growth was driven mainly by labor and capital Despite sharp improvements, infrastructure accumulation, but over the last 15 years total factor gaps across sectors remain wide, and export productivity (TFP) contributed about a third of growth, diversification is narrow. For example, transport and similar to that in East Asian MICs, such as Thailand. logistics infrastructure—the backbone of domestic Yet despite recent gains, Peru still suffers from large commerce and international trade—is less developed income and productivity gaps with high-income compared with that of peers and competitors. Road countries. Its output per worker is only 25 percent of density coverage and percentage of paved roads are that of the United States, and is somewhat lower than relatively low, offering limited connectivity between that of Chile and Mexico. While potential gains from the most important agricultural, consumption, and increasing the capital stock, human capital, and labor export areas, and little capacity to link production to are significant, a TFP increase would have the highest ports and airports. These obstacles impede the growth payoff for income per capita. of exports, which remain relatively undiversified: five 1  There are many different forms of informal employment, sectors (minerals, metals, vegetables, food, and textiles including in the formal sector. Peru Building on Success: Boosting Productivity for Faster Growth Against the backdrop of a less favorable external global productivity frontier (defined as the top 25 environment, growth will have to rely much more percent of firms in the United States)—resembling on productivity. Enhancing productivity growth is the situation in Colombia (5.5 percent) and Mexico the fastest way to close the still-large income gap (8.5 percent). Like other successful countries in the between Peru and high-income countries. For Peru, region, Peru is closing the gap (by 11 percent between increasing TFP renders a larger growth dividend than 2007 and 2011, for example). for the average country in LAC: the region would only double its GDP per worker if it had U.S. levels of There is a large dispersion in productivity growth efficiency, while Peru would almost triple it. among firms. Very different levels in productivity among firms are usually a first indication that Moreover, avoiding the “middle-income trap”, is a markets do not channel factors of production to firms challenge that Peru can overcome by deepening efficiently.2 While firm level productivity heterogeneity reforms to spur productivity. The concept of MIC is common, Peru is characterized by a high level of trap refers to the fact that transitions from low- to productivity dispersion vis-a-vis other countries in middle- income have been more frequent than those LAC and especially the United States. In Peru, firms from middle- to high-income: of 101 MICs in 1960, only in the 90th percentile of the productivity distribution 13 had become high income by 2008. As in countries are 500 percent more productive than those in the that avoided this trap, the policies that propelled Peru 10th percentile; this contrasts with around 200 into upper middle-income status will not suffice to percent in the U.S. Such disparities mean that some continue or accelerate the pace of convergence. The firms are able to produce much more given the same new engine of growth will be greater productivity, inputs, within the same industry. This disparity assisted by a suite of structural reforms. The past could be attributable to variations in technologies, decade’s performance has laid the groundwork. processes, human capital, and managerial skills. But, more importantly, this is a sign of strong disparities in 6 The rest of this Executive Summary examines Peru’s the allocation of factors of production, and a reason challenges and opportunities to boost productivity to look further into this issue. growth from a microeconomic perspective. Growth and convergence toward high income status require a There are also significant differences in firm number of important ingredients. The value-added of productivity across geographic areas of the country. this report is in approaching the issue of growth and Firms in Lima have been, on average, more productive raising productivity from the firm-level perspective, than those in the Sierra, Selva, and Costa regions. But using microeconomic data that has not been firms in the Costa region have started to converge systematically analyzed in Peru and providing selected toward Lima’s productivity level more recently. This policy directions. This analysis uses primarily data of shows some potential for productivity catchup to raise firms in the formal sector (accounting for roughly 80 aggregate productivity. percent of national output), while that of the labor market uses household and labor force survey which Not all subsectors contribute positively to contain data on workers in the informal sector. Thus productivity. The number of industries with the findings have implications for the dynamics of the negative TFP growth rates is higher in the services formal sector and the overall economy. sector. In manufacturing, some subsectors, such as food manufacturing, have significant roles in Peruvian firms’ productivity growth raising aggregate productivity, while others, such as and the challenges ahead chemicals or metallic products in manufacturing and wholesale trade in services, have reduced it. These Peru’s firm-level productivity is comparable to patterns contrast with those in other fast-growing that of other MICs and has been growing fast, but countries, such as China, where almost all subsectors it remains distant from the global productivity contribute positively to productivity growth. frontier. The country’s productivity at company level is on a par with that of similar MICs, such as Malaysia and South Africa. But the average firm in 2  World Bank. 2014. “Latin American Entrepreneurs: Many Firms but Little Innovation,” World Bank, Washington, DC. https:// the country is only 5 percent as productive as the openknowledge.worldbank.org/handle/10986/16457. Executive summary Unlike in other countries, larger firms in Peru firms. This misallocation effect is far stronger than are not more productive than smaller ones. In an in Mexico, Colombia, Slovenia or Hungary where environment where firms have access to appropriate similar analysis was carried out. A set of the “usual inputs and output markets are efficient, productive suspects” contribute to explaining this problem: labor firms can expand while less productive firms are likely market rigidities, lack of competition pressures and to stay small or ultimately exit. Thus, the expectation regulatory biases, issues with capital allocation, skill would be to find older and larger firms that are mismatches/deficiencies, among others.3 more productive (as they have survived and grown) than their younger and smaller counterparts. This This problem of misallocation is concentrated in is the case in the U.S. and in dynamic MICs, such as the services sector. Firm-level productivity grew in Indonesia or Vietnam. Peru has only a small share of both manufacturing and services over 2007–12, but young firms: only 10 percent of firms are between 1 was markedly slower in services. In manufacturing, and 5 years of age, while two-thirds are between 6 the allocation of factors between firms advanced and 19 years. The young and smaller firms, however, from contributing negatively to positively, a critical are more productive. This finding holds across sectors trend that was likely driven by the country’s increased and geographic regions. trade openness and competition. But in services, the slower growth in productivity is explained by the high Firms grow slowly in Peru, and that may also be negative contribution from misallocation of labor encouraging informality. In countries where input between firms, as labor moved from more productive and output markets work more efficiently, firms that to less productive firms. Unlike manufacturing, the survive tend to enjoy rapid growth: in the United States, services sector has not seen any improvement in this for example, old firms (40 years or more) are about trend over the last 15 years. eight times larger than young startups. Peru, however, shows evidence of stunted growth: old companies are, Misallocation in services is also hindering the on average, only about twice the size of younger ones. growth of other sectors that rely on service inputs. 7 This finding indicates that, beyond high costs of entry, Services—especially “other business services”— the costs of operation are large and hamper firms’ are an important input into other sectors, such as growth. Stunted growth in formal firms may also be manufacturing. If services are overpriced or low encouraging the growth of the informal sector, as job quality due to poor allocation of resources, other opportunities are not created at the pace needed in sectors would tend to rely less on these external the formal sector. Experience in other countries shows inputs and sub-optimally produce them internally in that reforms can substantially mitigate stunted the firms (e.g., transport or professional services). This growth. India, for example, shows a striking difference in turn hampers the productivity growth that comes in firm growth before and after reforms that included from specialization, as firms would produce services privatization, improving licensing, and increasing outside their core competencies. The Peruvian services efficiency of the trade system (which also boosted sector contributes only 4.8 percent of domestic competition). These reforms are directed at improving manufacturing value added, much lower than that in factor allocation between firms. Malaysia (20 percent), Thailand (17 percent), or South Africa (27 percent). These findings point to structural and market anomalies that tend to channel labor and other Overcoming the problem of misallocation through factors of production into less-productive public policy is feasible and would render a large firms. This phenomenon is referred to as productivity dividend. Policy actions that remove “factor misallocation.” Productivity may grow by distortions in factor, product, and intermediate goods increasing workers’ output within a firm (within-firm and services markets can spur productivity, pulling productivity), or by employees (and capital) moving up lagging industries and companies to national from less efficient to more efficient firms (between prevailing levels of productivity. In Peru, this could firms productivity). In Peru, while overall productivity 3  Taxes can also be an issue. For more on taxation in Peru, grew at a good pace, it was driven by increases in see World Bank. 2015a. “Peru: Selected Issues in Fiscal Policy within-firm productivity but was dragged down by — Taxation and Equity.” World Bank. Washington, DC. https:// hubs.worldbank.org/docs/imagebank/pages/docprofile. inefficient allocation of factors of production between aspx?nodeid=24311031 Peru Building on Success: Boosting Productivity for Faster Growth potentially increase overall productivity by up to 130 competition pressures. For example in 2013, 10 out percent and double output per worker. of 21 of the country’s larger municipalities had no procedures in place to authorize the construction While acknowledging that productivity growth has of telecommunications systems—limiting firms’ many ingredients, the report focuses on selected ability to compete on network expansion in a opportunity areas to reduce the problem of crucial infrastructure for modernizing the economy misallocation and boost firms’ productivity growth as a whole. In the same year, Instituto Nacional de in Peru. It examines issues and opportunities in the Defensa de la Competencia y de la Protección de la areas of business regulation, labor-market policies, Propiedad Intelectual (INDECOPI), the government’s firms’ access to low-cost financing, skills, trade, and agency tasked with fostering competition, found innovation. In all these areas the focus is chiefly on that 76 percent of complaints filed by businesses constraints that lead to misallocation of resources or were attributable to local problems. The country has are strongly affected by it. implemented some initial and positive measures in this direction recently. Enhancing competition pressures through regulatory reforms Loosening rigidities in the labor market and strengthening skills Competition drives efficient resource allocation, but may be restricted by regulations that unduly Inflexible and segmented labor markets can cause obstruct firms’ entry, operations, and growth. The significant labor misallocations. For an economy to small share of young firms in the market and stunted expand, it is vital that employers have the freedom firm growth in Peru point to barriers to entry and exit, to hire workers efficiently, rotate their duties, and high costs of operation for existing firms, and overall slim down or expand their staffs to reflect market potential barriers to competition pressures (e.g., conditions. Workers need freedom and the tools to 8 growing firms placing pressure on larger/older ones). move from low- to high-wage jobs. Firms’ growth and efficiency in the economy are compromised if labor Reforms in this area have helped in preceding rules restrict this movement and freeze companies decades, but progress needs to be reinforced and and workers at subpar productivity and wages. reforms deepened. The country now ranks as one of the LAC region’s best reformers in global indexes Restrictive dismissal regulations discourage that measure the business environment. Despite this employers from taking on new workers, progress, government bureaucracy and restrictive unintentionally encouraging informality and labor regulations remain as major constraints to doing hampering productivity. By some measures, Peru’s business. Young, small, and medium firms are hit informal sector now includes as many as 70 percent of particularly hard. In a recent survey, nearly 30 percent all non-agricultural workers in the country, well above of such companies identified business licensing and Mexico’s 60 percent and Colombia’s 54 percent. Peru permits as a major constraint on their ability to expand ranked 12th among 15 countries of the LAC region in and innovate. In 2014-15, the government has started a recent Doing Business survey on the ease of hiring to target these areas with a new wave of reforms that and firing. The fact that most workers are employed will need to be consolidated and continued. in the informal sector—outside labor regulations and unable to enjoy their benefits—suggests that Most barriers to competition stem from the overburdening labor policies are reducing their implementation of the legal framework by sub- protection of workers by cutting the coverage of national governments and decentralized bodies. formal policies, contrary to the intended purpose Weak implementation of existing regulations directly of the established rules. Labor productivity levels limits competition in sectors that matter most for in informal manufacturing, for example, are just a local market development. Central and decentralized quarter of its formal counterpart’s. Low productivity agencies and local governments, particularly at also translates into low pay for employees. Excessive the municipal level, sometimes add rules outside labor regulations affect most sectors of the economy, the legal framework. They often fail to implement but are especially detrimental in labor-intensive adequately national laws that would increase industries, such as services. Executive summary The country would do well to strike a better balance tend to perpetuate a cash economy with its myriad between worker protection and job creation. Labor inefficiencies. Political commitment to improving regulations come up as a major reason for allocative access to finance has been demonstrated by the distortions in nearly all areas analyzed: regulation, recent adoption of a national financial inclusion labor markets, trade, and innovation. In terms of strategy. adding flexibility and easing job mobility, hiring and firing regulations could be made less costly and The total volume of bank credit is low and the loan- employers given more discretion to approve lay-offs, to-deposit interest-rate spread is high. Although whether collective or individual. Reducing the non- many firms report using credit, total credit volume is wage costs of labor—which in Peru are very high by well below what the income level of the country and international standards—is a key reform. A better stable economic conditions would predict. Banking balance would also encourage formal employment and concentration and return on equity are high in Peru thus expand coverage of workers’ protection. relative to the rest of region, although competition has stiffened in recent years (the number of banks Skilled workers are a fundamental requirement for increased from 11 in 2006 to 17 today, and return raising productivity and incomes. Although skilled on equity decreased from 30 percent in the late first workers are in growing demand in Peru, they remain in decade of the 2000s to just above 20 percent in 2014). short supply. This is highlighted by the increasing labor These factors, together with the still-high spread income returns for people with 10–11 years of education between loans and deposit rates, suggest room to compared to those with 7–9 years and those with only increase competition further (but more analysis is primary education. Further evidence shows that the needed to confirm this view). quality of schooling is low—Peru has one of the lowest scores in math on the Programme for International The cost of credit for small and medium firms Student Assessment (PISA). Public vocational training is relatively high, and access to finance shows programs, which account for 30 percent of continuing geographic variation. Peru ranks well in access to 9 training, underinvest in equipment and often fail to give credit in most cross-country rankings (e.g., Doing graduates the skills that the labor market is seeking. Business). However, the cost of credit is thought to be Improving these outcomes would require higher public high, especially for small and medium firms. For micro spending on education and training, which remains low firms with access to the financial sector, the annual despite recent improvements as well as involve better cost of credit was above 30 percent of the loan amount matching of vocational training to the needs of the in 2014. Although measurement is difficult, there is private sector. evidence that effective borrowing rates are high for informal firms outside the formal financial system. Improving access to and quality of finance For the financial system as a whole, the average for small and medium enterprises lending rate was 16 percent in 2013, which is high compared with other economies at similar levels of Improving access to efficiently priced finance development. Given that large corporates can obtain can facilitate firms’ expansion and technology credit at very low and competitive rates, smaller adoption, contributing to productivity growth. An and medium firms’ costs of borrowing must be high efficient financial sector is central to productivity to lift the average rate to that level. Geographically, growth because it connects savers and borrowers 85 percent of firms in Lima have bank loans, but in for credit access and capital accumulation, and it the city of Arequipa—where the penetration of the can contribute to better resource allocation in the financial system is similar to that in Lima—the figure economy. Although the country’s savings rate is is only 47 percent. relatively high, many Peruvians do not save in formal financial institutions. An estimated 29 percent of The last two sections of the Executive Summary the population have savings accounts, lower than explore further opportunities to reduce misallocation in Bolivia (36 percent), Colombia (35 percent), and of factors in areas that are typically examined in the Ecuador (51 percent). Instead, people save their cash context of within-firm productivity growth. Indeed, at home or through such traditional methods such performance in the areas of international trade and as purchasing property or livestock. These patterns innovation is strongly affected by misallocation, and Peru Building on Success: Boosting Productivity for Faster Growth therefore large productivity gains could come from as labor regulations and behind-the-border barriers improvements in these areas. (e.g., improving customs and trade facilitation) would encourage domestic and foreign companies to start Expanding foreign trade export projects. Improvements in trade and logistics infrastructure could reduce the large sunk costs International trade can help reduce the misallocation that firms need to undertake to be able to export, of resources and foster productivity. Exposure to encouraging smaller firms to innovate, trade and grow. export markets has provided strong competition Building more tailored skills that can be used in higher pressures, as shown by the progress in reducing the value-added manufacturing and services industries misallocation of resources in Peruvian manufacturing is important as well. A more robust innovation over 2007–12. environment, as discussed below, would also facilitate diffusion of GVC-acquired knowledge and improve Peru should be commended for having one of the downstream links for exporters. world’s most liberal trade regimes, yet it trades very little. Peru’s total exports are low (22.4 percent Helping companies innovate of GDP in 2014) relative to other countries at similar income levels. Despite some impressive successes in Shortcomings in the innovation system perpetuate exporting fruits, vegetables, and apparel, Peru has inefficient factor allocation. For example, in Peru weak ties to global value chains, the international small and young firms are more productive, but if they networks of companies that team up for sequential have less access to inputs such as finance, or if their production in different countries to bring high-value innovations are not well protected, overall productivity products to world markets.4 Peru’s role in these growth is likely to suffer. Thus, beyond assuring chains is mainly as a supplier of low-value primary general preconditions for innovation—high-quality inputs (e.g., minerals). education, investment in research and development,— 10 the government may need to help firms tackle more Higher value-added traded goods and services are a specific obstacles. large opportunity for Peru. In addition to small trade volumes overall, so far, Peru has done little to adapt In Peru, firms invest too little in innovation. Peruvian to the more sophisticated services that are expanding firms invest on average 2.5 percent of their sales their share of global exports. Services also provided in innovation, whereas their peers in Chile invest 3.5 little support to other exports, compared to other MICs. percent. Moreover, technology adoption (acquired The value added of services in total exports in Peru is through licenses of new technology and imported only about 20 percent, compared with 30-40 percent capital goods), which is a straightforward way to in Thailand, Malaysia and South Africa. To support move closer to the innovation frontier, is scarce. Only exports in higher value added products requires 7 percent of Peruvian firms have licensed technology improving infrastructure for trade and logistics, and from abroad, half the LAC region average and 2.5 behind-the-border issues. High costs of trade logistics times less than in OECD countries. amount to about 32 percent of product value in Peru, one of the highest levels in Latin America—above Peruvian firms that do invest in innovation are more Colombia’s (23 percent) and Chile’s (18 percent) for likely to introduce new products, but low returns example—and far higher than the OECD average of 9 to innovation (sales) may be contributing to low percent. Some key elements that should support this investment. If a Peruvian firm spends on innovation- agenda are in fact going the wrong way: in customs, related activities, it is more likely to introduce a new Peru’s Logistics Performance Index ranking has product or new process than firms in other LAC countries. declined markedly, from 49th to 96th. But Peruvian firms that introduce new products or processes have an average of only 38 percent higher These findings suggest that reforms can have a sales per employee, compared to about 100 percent in substantial impact. Easing domestic constraints such five other surveyed countries of the region. 4  World Bank. 2015. “Latin America and the Rising South: Small and young firms seem to have less access to Changing World, Changing Priorities.” https://openknowledge. worldbank.org/handle/10986/21869. innovation support than larger ones, reinforcing Executive summary poor allocation of labor and capital. The average age of firms that innovate is 22 years, five years older than the average age of firms that do not. The average number of employees of innovating firms is 75 percent higher than in other firms. Public policy could help improve the innovation system. Enabling better use of knowledge entails improving governance and legal enforcement environment for institutions and companies to engage in innovative ventures. The expansion of agencies that diffuse existing technology from abroad to local firms, particularly small and medium firms, could help boost productivity. Opening the research and university system to various forms of knowledge transfer and to collaboration with the private sector would also be helpful. Conclusion A third of Peru’s economic growth has been driven by improvements in productivity, but in the next stage of convergence, and under the new external conditions, a larger contribution to economic growth will have to come from productivity. The country emerged as 11 one of strongest growth performers in the region and accelerated its convergence to higher income levels. Although the country’s productivity growth at the company level is at par with similar middle-income countries that aspire to avoid the middle-income trap, the gap to the global productivity frontier remains large. This report shows that enhancing productivity growth remains the fastest way to close it. There is strong firm-level evidence that Peruvian markets tend to misallocate labor and capital into less productive workplaces. This signals that some aspects of product, factor, and intermediate goods and services markets do not function properly. The country’s services sector is a poor performer in this regard. Eliminating these distortions could increase overall productivity by up to 130 percent and double Peru’s output per worker. But that in turn requires deepening reforms in selected areas. 13 Introduction Introduction P eru’s recent robust growth has raised its economic prospects and has been broadly shared. Peru grew slowly from the 1960s through the end of the 1990s. But the defeat of terrorism and a series of far-sighted reforms raised environment. Rising above middle-income status is a policy challenge for all MICs. Strategic shifts are usually needed to move the economy from growth driven by accumulation of factors to productivity-based growth. As factor accumulation slows, an economy that its growth rate above regional and world averages adapts by increasing productivity is more resilient. in the 2000s. During the last decade, the economy Productivity is an important determinant of differences grew at an average rate of 6.4 percent per year. The in incomes across countries, and productivity growth growth translated into higher incomes, and lower is the ultimate long-run driver of economic growth. 15 poverty and inequality. Peru has been consistently But if productivity growth is moderate and growth is among top regional poverty and inequality reducers: driven principally by an accumulation of capital and poverty was more than halved over the last decade, labor, economic growth is likely to taper as returns to while inequality fell by 12 percent between 2004 and such factor accumulation diminish. 2013. As a result, the country reached upper–middle- income status in 2008 and is now seeking to join This report looks at Peru’s principal challenges the Organization for Economic Co-operation and and opportunities to maintain strong and inclusive Development (OECD). growth through higher productivity. Success would require it to boost productivity by further improving The next step in development—high-income conditions for firms’ growth through lower costs of status—is challenging for all middle income entry and operation, higher completion and, through countries and requires sustained growth. Peru would that, an improved factor allocation to productive need to maintain an average 6.5 percent growth rate firms. It would also require a reduction of inefficiencies over the next 15 years to reach high-income status. by reaping more benefits from international trade Many countries develop rapidly to reach middle- and an improved innovation framework. The report income status but few take the next step to high- has five parts. Part I analyzes the sources of Peru’s income status: of 101 middle-income countries (MICs) recent strong growth. Part II examines more closely in 1960, only 13 had become high income by 2008. the dynamics and challenges to Peruvian firms’ Most of these successful countries adopted policies productivity growth. Part III analyzes elements that that spurred productivity. The policies that propelled would increase the efficiency of Peru’s labor and capital Peru into upper–middle-income status are not those resources—regulation, labor markets, and human that will take it to the next level. capital. Part IV is focused on further opportunities to reduce misallocation while at the same time speeding Raising productivity growth is key to sustaining up within-firm productivity through expanding Peru’s Peru’s high growth rates of income per capita, technological possibilities, particularly through including in the context of a less favorable external innovation and the spillovers from international trade. 17 Part I Understanding Peru’s growth story 1. Peru’s growth success 1. Peru’s growth success The country showed a strong drive to comparators (Figure 1). Peru grew at an average of reach a middle-income status, with 4.5 percent per year during 1990–2013 (compared to prosperity broadly shared regional and global growth of around 3 percent), and O at an even faster average rate of 6.4 percent during ver the last 12 years, Peru has doubled the last decade. As a result, Peru doubled its GDP and 19 its real per capita income. Peru grew per capita income in the last 12 years, while the Latin more slowly than the world and regional America and the Caribbean (LAC) region increased averages between the 1960s and the 1990s. But them only by half. after defeating terrorism and adopting a series of macroeconomic stabilization and structural reforms Growth has been widely shared, placing Peru among in the 1990s and 2000s, it saw growth rates of GDP the fastest poverty and inequality reducers in the and income per capita accelerate, exceeding those of region. Low-income households have gained more Fi ur 1. P ru h s s n f st p r c pit nd r t Fi ur 2. Growth w s bro dl sh r d GDP rowth ov r th p st qu rt r c ntur , r l tiv to th r ion nd th world (real average annual growth, %) (average annual per capita income growth, %) 5 10 4.4 4.5 4.1 Overall Population Bottom 40% 4 8 3.1 3.0 6 3 2.9 2.7 2.2 2.0 4 2 1.7 1.3 2 1 0 0.2 0 -2 Peru World LAC World LAC Peru Argentina-urb. Uruguay-urb. Dominican R. Brazil El Salvador Costa Rica Guatemala Nicaragua Honduras Peru Ecuador Paraguay Chile LAC Panama Colombia Mexico Bolivia 1961-1990 1990-2013 GDP per capita GDP Source: WDI. Source: Socio-Economic Database for Latin America and the Caribbean (SEDLAC) and WDI. Note: Annual percentage growth rate of GDP at market prices and GDP per capita Note: 2003–13, except for Bolivia (2002–13), Argentina, Peru, and El Salvador based on constant local currency. GDP per capita is gross domestic product divided (2004–13), Costa Rica (2003–08), Nicaragua (2005–09), Panama (2008–13), Mexico by midyear population. Aggregates are based on constant 2005 U.S. dollars. (2002–12), and Guatemala (2000–11). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 3. Pov rt r duction w s b tt r th n Fi ur 4. P ru w s mon th f st st in oth r LAC’s infl tion-t r tin countri s in qu lit -r duc rs in th r ion (% of population lifted out of poverty) (% change in Gini coe cient, 2004–13) 25 0 -2 20 -4 15 -6 10 -8 -10 5 -12 0 -14 Peru Colombia Brazil LAC average Chile Mexico El Salvador Brazil Honduras Peru Ecuador Uruguay Dom. Rep. Paraguay LAC Colombia Mexico Bolivia 2000-13 2002-13 2001-13 2000-13 2000-13 2000-12 Source: Authors’ calculations based on data from INEI, Comisión Económica para Source: LAC Equity Lab tabulations of SEDLAC (Centro de Estudios Distributivos, Laborales América Latina y el Caribe, (CEPAL). y Sociales, CEDLAS and the World Bank) and World Development Indicators (WDI). Fi ur 5. H vin l d b hind oth r MICs sinc Fi ur 6. Structur l nd st bili tion v ri bl s th 1970s, P ru is now firml on th p th h v h d stron r imp ct on rowth in P ru of conv r nc to hi h incom p r c pit (GDP per capita relative to the United States, %, (contributions to real GDP per capita growth in Peru, 40% constant 2005 US$ per capita, 1950–2014) 5 2000–13, percentage points) 35% 4 30% 25% 3 20 20% 2 15% 1 10% 5% 0 Growth between Growth between Growth between 0% -1 2006-10 2001-05 2006-10 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2014 and 1996-2000 and 1996-2000 and 2001-05 Chile China Malaysia Peru Thailand Persistence Structural Stabilization External Source: Authors’ calculations based on Feenstra et al. (2015): Background paper Source: Authors’ calculations: Background paper prepared for this report by Savescu prepared for this report by Moroz and Vostroknutova (2015). and Vostroknutova (2015), based on Araujo et al. (2015a) and the National Accounts Note: This gure uses Penn World Table (PWT) 8.1 and WDI. data by INEI, BCRP, Ministerio de Economía y Finanzas (MEF), and WDI. from growth than the national average. Between making Peru one of the most successful countries in 2004 and 2013, real income per capita of the bottom the region, although some geographic and urban–rural 40 percent grew at an average 6.8 percent, above the disparities remain (Figure 4). Importantly, growth 4.4 percent national average. These gains were high was the main driver of lower poverty and inequality by regional standards (Figure 2). They helped Peru through improved labor incomes: 85 percent of poverty to reduce poverty from 63.4 to 29.3 percent of the reduction between 2004 and 2010 is explained by population.5 In 2013 alone, almost 500,000 people this factor. Redistribution policies only account for 15 were lifted out of poverty. Peru reduced poverty percent of that reduction.6 faster than other countries with similar income levels and macroeconomic policy stances (Figure 3). Peru’s convergence with higher income levels has Income inequality, measured by the Gini coefficient, is been accelerating over the last decade. From the still high but fell from 0.49 in 2004 to 0.44 in 2013, 1960s to the mid-1990s, convergence was slow (or even negative). During that period several comparator 5  The national poverty line is the cost of a minimum basket of countries continued their convergence, leaving Peru goods (including food) and services (NS292 per month per person in 2013), while the extreme poverty line is the cost of a minimum food basket (NS155 per month per person in 2013). 6  Inchauste et al. (2012). 1. Peru’s growth success Fi ur 7. Pr dict d contributions to rowth from st bili tion, structur l ch n , xt rn l conditions, nd p rsist nc : Ch n s in p r c pit GDP b tw n l t 1990s nd l t 2000s (%, countries ordered by contribution of external factors) 7 External Conditions 6 Persistence 5 Structural Change Stabilization 4 3 2 1 0 -1 St. Kitts & Nevis El Salvador Honduras LAC average Venezuela Brazil Argentina Guatemala Chile Mexico Costa Rica Nicaragua Guyana Ecuador Peru Belize Colombia Jamaica Panama Dominica Uruguay St. Lucia Grenadines Paraguay Grenada Haiti Bolivia Dominican R. -2 St.Vincent & Source: Araujo et al. (2015a). Note: “Late 2000s” stands for 5-year average, 2006–10, and “late 1990s” stands for 5-year average, 1996–2000. behind (Figure 5). Yet, in the last decade Peru’s income Much of the rest of this report diagnoses the selected per capita has been converging rapidly with that of the challenges faced to achieve sustained growth high-income countries and—albeit from a low base—is through higher productivity, and provides options matching the convergence speed of other fast-growing for overcoming them and enabling Peru to continue MICs like Chile, Malaysia, and Thailand. But as with its path to high-income status. The fundamental other MICs, Peru now faces the hardest part—moving premise is that Peru will have to increase productivity— into high-income status. obtaining greater output from its resources—and 21 better align its economic incentives with international Strong growth over the last two decades was best practices. This involves increasing the efficiency the result of macroeconomic stability, structural with which the economy employs its resources. reforms, and favorable external conditions. As in any commodity-exporting country, growth in the late 2000s relative to the late 1990s was significantly supported by favorable external conditions (Figure 6). However, it also benefited from a wave of structural reforms initiated in the 1990s (Box 1). Structural and external factors contributed an estimated 0.8 and 0.7 percentage points, respectively, to growth annually, thus explaining 25 and 21 percent of the 3.5 percent annual growth over this period. Growth in the early 2000s shows high contribution from “persistence”— the likely lingering positive effect of past macro- structural reforms. Growth from 2005 onward shows a larger contribution of structural factors and the highly favorable external environment (Figure 7).7 7  The graph depicts System Gaussian Mixture Models (GMM) estimations of a differences equation (Δlnyct= θ(Δlnyct-1)+ ΓΔln(X) ct + Δbt + Δect) of contribution to growth of a set of variables grouped in four categories: structural variables combine schooling, private credit/GDP, trade openness, and infrastructure (telephone lines), government consumption, and institution quality; stabilization variables include inflation, a proxy for exchange rate misalignment, and the banking crisis dummy; external conditions captures the effects of past shocks and interventions, which is are captured by terms-of-trade growth, growth in (country- introduced by the lagged dependent variable. In this formulation, specific) commodity prices, and time dummies (capturing global the coefficient for the persistence term likely incorporates the effects such as liquidity); and a growth-persistence parameter effects from past stabilizations and structural reforms. Peru Building on Success: Boosting Productivity for Faster Growth Box 1 Reforms and prosperity: A quarter century of reforms in Peru At the beginning of the 1990s, Peru was in state monopolies in food imports. Financial turmoil. Between 1987 and 1990, output fell liberalization entailed eliminating controls by 25 percent, the average public deficit for on interest rates, creating a credit bureau, the period was more than 8 percent of GDP, implementing the Deposit Insurance Fund, and hyperinflation reached 7,650 percent. strengthening banking supervision and Public debt was in arrears and the stock of regulation, and restructuring and privatizing international reserves barely reached 2 percent most state financial institutions. Labor of GDP. Price controls had introduced large reforms saw reduced dismissal costs distortions into the structure of relative prices, through replacing absolute job security with and the parallel-market exchange rate reached protecting against arbitrary dismissal, and five times the official rate. relaxing wage determination, rules on working hours, and types of employment contracts. First Round of reforms Deep pension-system reform and creation of regulatory agencies in telecommunications, Against this backdrop, an ambitious policy energy, and transport completed the package. package was implemented in the early 1990s to promote macroeconomic stabilization These policies boosted confidence, helping lift and support the operation and efficiency of GDP growth to around 4 percent between 1991 markets. The main policies included central and 1999. The public deficit was reduced to bank independence and limits on the financing it below 1 percent of GDP between 1997 and 1998 could provide to the public sector; liberalization and inflation fell to single digits toward the end 22 of the multiple exchange rates system and of the decade. adoption of a managed float regime; removal of food subsidies and reforming the pricing Second Round of reforms of public utilities; simplification of the tax system by reducing the number of taxes and During the 2000s, reforms were mainly exemptions; and restructuring of the external oriented to consolidate and deepen the gains debt owed to official and private creditors. in macroeconomic stability and efficiency in the functioning of the public sector. They were Reforms that fostered the functioning and based on three axes: efficiency of markets included opening up to international trade, via unilateral reduction • Developing a macro-fiscal framework, with laws of tariffs and other barriers to imports that promote transparency and accountability and exports, as well as the elimination of in managing public resources. These included Box Fi ur 1. Fisc l r sults Box Fi ur 2. Infl tion-t r tin r im s tool to nchor infl tion xp ct tions (% of GDP) (%) 6.0 12 4.5 3.0 10 1.5 8 0.0 -1.5 In ation 6 target -3.0 range -4.5 4 -6.0 -7.5 2 -9.0 0 -10.5 -12.0 -2 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Primary Balance Overall Balance In ation (eop) Expectations (eop - one year in advance) 1. Peru’s growth success Box 1 Reforms and prosperity: A quarter century of reforms in Peru (cont.) incorporating quantitative fiscal rules resources in the public sector, and introduced to limit public spending; creating a fiscal clear definition of roles, career development, stabilization fund, which allowed public and pay scales for public servants. resources to be saved in years of high growth • Tax and pension reforms: These were geared for use during recessions through temporary to start bringing the tax system closer expansion of public spending; switching to OECD standards; and pension-system to results-based budgeting, which linked reform included more accumulation options. budget planning with measurable objectives • Closing gaps in infrastructure: This included to boost efficiency of public spending; and defining multiyear targets at the sectoral abolishing the cédula viva regime, in which a level; and developing a framework for group of former civil servants received a full- public–private partnerships (PPPs) and salary pension on retirement. other arrangements to fund and develop • Adopting an inflation-targeting regime, to infrastructure. anchor inflation expectations via an explicit inflation target. In addition, decisions of the In 2014–15, and in the context of a strategy to central bank were made public. cope with the slowdown in output, a new wave • Deepening international trade links, through of reforms has been initiated. The reforms are trade agreements with main trading the first steps on a medium-term agenda, partners. which is highlighted as critical by this report. The measures focus on: These reforms had a positive impact on economic outcomes. In 2000–09, the average • Lowering costs of entry, operation, and public deficit was reduced to around 0.5 exit of firms, with measures to reduce 23 percent of GDP and public debt fell to 27 burdensome regulation, simplification of percent of GDP. Inflation targeting anchored tax procedures, a first step to make the expectations within the target range. labor market more flexible, and insolvency proceedings simplification. Third Round of reforms • Initial aspects of trade facilitation, including the reduction of tariffs for some intermediate In the late 2000s and the beginning of this goods, and support for firms in terms of decade, the reforms focused on: product accreditation and certification to reach new export markets. • Strengthening human-capital development and • In addition, regulations to accelerate public sector. Healthcare reform emphasized a prudent implementation of the PPPs new remunerative and management framework have been initiated. systems; education reform increased budgetary allocations for improvements There is no question that further reforms will in teaching careers, learning systems, and be needed to hold to the needed direction, but infrastructure; civil service reform created Peru can feel proud of its remarkable reform the SERVIR program to manage human achievements. Source: Background paper prepared for this report by Céspedes and Lengua-Lafosse (2015), Central Bank, and Ministry of Economy and Finance of Peru. Peru Building on Success: Boosting Productivity for Faster Growth 2. Growth in Peru - drivers and challenges Labor and capital accumulation were Peru increased domestic savings sharply and they some of the early drivers of growth helped to finance capital accumulation, unlike in W other LAC countries. In the 2000s, growth fueled by ith relatively high labor and capital favorable external conditions led to an almost one- 24 accumulation, Peru has had enough to-one increase in consumption in LAC58 countries; input accumulation to fuel growth. It as a consequence, increases in savings were only a achieved substantial labor accumulation, which was small fraction of GDP, 0.3 percentage points (Figure supported by a strong demographic dividend and high 10). In Peru, conversely, 1.8 percentage points of the labor force participation (Figure 8). Capital accumulation 5.7 percent growth went into savings, and domestic was at the level expected for Peru’s income per capita savings became the main driver of investment growth (Figure 9). But there could be significant gains by as domestic financing reached 86 percent (Figure 11). improving the quality of those inputs, as it is discussed In almost every LAC5, country investment increases later in this report. (see Sections 8 and 9). were financed mostly by external savings (apart from Fi ur 8. L bor forc p rticip tion is hi h r Fi ur 9. C pit l ccumul tion is s pr dict d th n in comp r tor countri s for P ru’s incom p r c pit (average labor-force participation, 1991–2013, %) (Output per worker, constant 2005 US$, 2011) 80 160000 70 140000 60 120000 50 100000 40 80000 30 60000 20 40000 10 20000 Peru 0 0 Peru OECD LAC Chile Malaysia Uruguay 0 100000 200000 300000 400000 Capital per worker (2005 US$) Source: WDI. Source: WDI. 8  LAC5 are Argentina, Brazil, Chile, Colombia, and Mexico. 2. Growth in Peru—drivers and challenges Fi ur 10. P ru incr s d s vin s sh rpl durin Fi ur 11. Dom stic s vin s w r th r pid rowth p riod, unlik LAC5 countri s th m insourc of inv stm nt (contribution to GDP growth, percentage points (change in investment and sources of nancing and annual change, average 2001–2013) between 2000 and 2013, percentage points of GDP) 6 12 10 5 8 4 6 4 3 2 2 0 -2 1 -4 0 -6 LAC5 Peru LAC5 Peru Consumption Savings Average annual growth Domestic savings External savings Investment Source: Sta calculations based on UN and BCRP data. Source: Sta calculations based on UN and BCRP data. Fi ur 12. Both priv t nd public s ctors Fi ur 13. P ru is simil r to oth r succ ssful MICs durin contribut d to boost s vin s th ir rowth spurts in ccumul tin dom stic s vin s (% of GDP) (% of GDP) 25 50 40 20 30 15 20 25 10 10 0 -10 5 -20 0 -30 1971 1990 1971 1990 1987 2006 1994 2013 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Korea Malaysia Chile Peru Public saving Private saving Domestic savings External savings Investment (% of GDP) Source: Sta calculations based on Central Bank data. Source: Sta calculations based on WDI and Central Bank data. Colombia, where external savings financed 17 percent account surpluses and only a small current account of the increase). deficit for most of the last decade. In Peru, the external savings that were used to finance growth were mostly Peru saved domestically and used external sources FDI, not debt. These outcomes have made Peru more to finance its growth, without running fiscal or resilient to positive and negative external shocks. current account deficits. Savings increased from just above 10 percent of GDP in the early 1990s to Despite significant improvements, infrastructure 22 percent in 2014 (Figure 12), driven mainly by the gaps across sectors remain large while export private sector, with a stronger push from the public diversification is limited. Public capital spending sector in the mid-2000s. Peru resembles most of the has increased over the last years, from 3.8 percent successful, fast-growing MICs during their boom years, of GDP in 2000 to 6.1 percent of GDP in 2013—a and differs from regional comparators in this area highly positive development. But infrastructure gaps (Figure 13). Prudent macroeconomic and stabilization across sectors are large. For example, the transport policies and the high saving rates allowed Peru to and logistics infrastructure, which is the backbone reduce debt to 20 percent of GDP, accumulate ample of domestic commerce and international trade, international reserves and fiscal buffers, run fiscal stands as less developed when compared to relevant Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 14. P ru’s infr structur is low r qu lit th n in p r MICs, sp ci ll in ro d n twork nd ports (World Economic Forum, WEF indicators 1–7; higher indicates higher quality) 6 Roads 6 Ports 5 5 4 4 3 3 2 2 1 1 0 0 Malaysia Chile Thailand Mexico Uruguay LAC Peru Colombia Malaysia Chile Uruguay Thailand Mexico Peru Colombia LAC 6 Railroads 6 Air Transport 5 5 4 4 3 3 2 2 1 1 0 0 Malaysia Mexico Chile Thailand Peru Colombia Uruguay MalaysiaThailand Chile Mexico Colombia Peru Uruguay LAC 26 Source: World Economic Forum (2015). peers and competitors (see Figure 14). Road density today is similar to that of two decades ago, with coverage and percentage of paved roads is relatively services accounting for 58 percent, industry 36 low, and with limited connectivity between the most percent, and agriculture 6 percent of GDP. The sectorial important agricultural, consumption, and export contributions to growth therefore largely reflect areas, and limited capacity to link production to ports differences in sector size rather than different growth and airports for export. These constrains also impede rates. Of the 4.9 percent average annual growth the growth of exports, which remain undiversified. during 1995–2013, services accounted for 2.9 percent, Overall, five sectors (minerals, metals, vegetables, industry 1.8 percent, and agriculture 0.3 percent. The foods, and textiles and apparel) accounted for 91 three sectors grew at similar average rates over the percent of merchandise exports in 2014, virtually whole of the two decades, but in the 1990s growth of unchanged since 1994, and with minerals taking a agriculture outpaced that of industry and services, significant share. In the context of a non-temporary and in the 2000s industry and services grew faster. adjustment in commodity prices, growth in high value- added manufacturing and services will help to support Positive structural change contributed to sustained growth. productivity growth. At the aggregate level, average labor productivity (measured as value added per worker) Peru also gained from structural can increase for two reasons: higher productivity within shifts, though a large informal sector one or more sectors of the economy; or structural shifts remains a drag on growth in employment from lower- to higher-productivity sectors.9 Measures of value added per worker and Growth since the 1990s has been broad across economic sectors, with industry and services 9  In the analysis behind this paragraph, labor productivity is measured as value added (GDP) per worker, based on sectorial leading in the 2000s. Peru’s economic structure data from the Groningen Growth & Development Centre 2. Growth in Peru—drivers and challenges Fi ur 15. Althou h l bor productivit rowth h s b n Fi ur 16. A structur l shift tow rd s rvic s driv n m inl b improv m nts within s ctors, structur l h s support d ov r ll rowth in VA/work r, shifts b tw n s ctors h v lso h lp d r c ntl but mor productiv m nuf cturin h s sh d l bor (value added per worker growth contributions, percentage points) (change in average employment share, and deviation from average value added per worker, ratios) 5 1.0 4 0.8 Mining Utilities Log (sector prod / average prod) 3 0.6 Financial Serv. 2 0.4 0.2 Manufacturing 1 Construction 0.0 Transport 0 -0.2 -1 -0.4 Gov. Serv. Commerce -2 -0.6 Agriculture -3 -0.8 -4 -1.0 1961-1975 1976-1992 1993-2001 2002-2011 -0.10 -0.05 0.00 0.05 Within sectors Between sectors VA/worker Change in employment share (1990-2011) Source: Authors’ calculations based on Groningen Growth & Development Centre Source: Authors’ calculations based on Groningen Growth & Development Centre data: Background paper prepared for this report by Barco and Vostroknutova (2015). data: Background paper prepared for this report by Barco and Vostroknutova (2015). employment patterns across 15 economic sectors were Notably, however, manufacturing—which has been used to estimate the relative contributions of these two more productive—also lost some labor to services. sources, and show that overall productivity growth was Manufacturing’s share fell from 12 to 9 percent largely the result of improvements within sectors, though between 1990 and 2011, reflecting both natural structural shifts across sectors also played a role. The structural change and recent favorable terms of 27 between-sector shift subtracted from aggregate value trade, which tended to expand labor in non-tradable added per worker growth over 1993–2001 (although it sectors. Manufacturing’s job loss is linked to the post- was outweighed by the within-sector shift), but both crisis shift in the speed of manufacturing growth: it contributed positively during 2002–11: out of overall tended to lead overall GDP growth before 2008 and labor productivity growth of almost 4 percent per year, trail it after. After 2008 the more complex and less shifts in employment from agriculture to construction labor-intensive manufacturing grew fastest (at 7.7 and services accounted for about 1 percent, while percent) while basic labor-intensive manufacturing within-sector gains accounted for the rest (Figure 15). grew more slowly than GDP (at 2.6 percent). Section 8 sheds further light on the dynamics of wages relative Labor relocation from agriculture to services was to labor productivity. an important source of labor productivity gains from structural shifts. Between 1990 and 2011, the Peru also has a large informal sector that drags rural–urban transformation continued, reducing the growth, since most of its workers have far lower agriculture’s share of employment from 31 percent productivity than in the rest of the economy. Peru’s to 23 percent; that of services and construction informality is high even by LAC standards: it was at increased from 55 percent to 67 percent. Labor 70 percent outside of agriculture, above Mexico’s productivity grew in all sectors. While labor moving 60 percent and Colombia’s 54 percent in 2009.10 As from agriculture to services was an important source discussed in Section 6, workers in the informal sector of labor productivity gains from structural shifts, it is are on average only one-third as productive as formal also important to highlight that the services sector workers. By official estimates, however, the share of has not been as productive as it could have been, as informal sector in GDP was at only about 21 percent the report finds at firm-level (Sections 3 and 4). database. Following the standard critique of this approach, we 10  ILO (2012). Peruvian authorities calculate the share of also constructed marginal labor productivity measures by sector informal employment at 77.2 percent in 2009 (74.3 percent in and conducted similar analysis (using the World Bank I2D2 2012), referring to the total number of informal jobs in the formal database for wages). The ranking using marginal measures of sector (without social contributions by employers), informal labor productivity was the same. companies, or the household sector; see INEI (2014). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 17. Th contribution from TFP to rowth turn d Fi ur 18. P ru comp r s w ll to oth r f st- rowin positiv ft r th r forms countri s on productivit ’s contribution to rowth (real GDP growth and contributions, percentage points) (contributions to growth, %, 1990–2013) 8 6 45 40 6 5 35 4 4 30 25 2 3 20 0 2 15 10 -2 1 5 -4 0 0 1960-1970 1970-1990 1990-2000 2000-2013 1990-2013 1990-2012 1990-2012 1990-2012 Peru Chile Thailand Malaysia Capital Stock Labor Human Capital per Labor Capital Stock Labor Total Factor Productivity Real GDP Human Capital per Labor Total Factor Productivity Source: Authors’ calculations: Background paper prepared for this report by Source: Authors’ calculations: Background paper prepared for this report by Savescu and Vostroknutova (2015). Savescu and Vostroknutova (2015). in 2010.11 This report focuses attention on the informal (Figure 17). In the 2000s, TFP contributed around sector mostly in the context of the labor market, a a third of Peru’s growth.14 This kind of productivity- decision driven both by the importance of informality driven growth is similar to that in other fast-growing for workers, as well as by the absence of firm-level data countries (Figure 18). on informal firms (see Section 6).12 Indeed, the analysis 28 in this report uses mostly data of firms in the formal Despite its recent gains, Peru still has large income sector (which represents roughly 80 percent of the and productivity gaps with high-income countries. country’s output), yet the findings have implications Its output per worker15 is only 25 percent that of for the dynamics of the informal sector and the overall the United States, and is low compared to Chile and economy. Mexico (Figure 19 and Figure 20). While potential gains from increasing the capital stock, human capital, and In the last decade, one-third of growth labor are significant, a TFP increase would have the is explained by productivity growth, highest payoff for income per capita.16 but the productivity gap with high- income countries is still large To avoid being bogged down in the middle-income trap and in the context Strong TFP growth has played a central role in of a less benign external environment, Peru’s growth in the 2000s. Before the 1990s, growth Peru would need to lift further was driven by factor accumulation, with a negative productivity growth contribution from TFP.13 Macroeconomic stabilization and structural reforms during the last two decades After successfully growing from low-income to have improved efficiency in allocating resources middle-income, many countries find themselves and have led to more productivity-driven growth 14  Growth accounting estimates the contribution to growth of labor and of capital (investment). When the sum of their 11  INEI (2014). contributions is not equal to the actual growth, the residual is 12  See a more detailed discussion of this in Sections 3 and 6. attributed to total factor productivity (TFP)—an interactive effect 13  The analytical tool at the core of development accounting is that is not directly observable and is a measure of an economy’s the aggregate production function that relates inputs to outputs: dynamic technological change. TFP is often seen as the real driver the amount of physical and human capital the amount produced. of growth. It may account for up to 60 percent of growth within If increasing the amount of inputs leads to an equivalent increase economies. See Easterly and Levine (2001), among others. in outputs (constant returns to scale), the aggregate production 15  This part of the analysis uses Caselli (2014) results and function can be written in per-worker terms. Thus, we use the applies them to a different measure of output per worker. For augmented Cobb-Douglas aggregate production function, in output per worker estimates, this report uses PWT8.2 GDP in which output per worker is a product of human capital per worker constant 2005 US$, projected to 2011–2013 by growth rates of (adjusted for quality), physical capital per worker, and a term similar variable in WDI. called total factor productivity (TFP) or “efficiency.” 16  Caselli (2005). 2. Growth in Peru—drivers and challenges Fi ur 19. P ru could tripl its incom p r c pit Fi ur 20. P ru would b n fit from incr sin if its TFP w r s hi h s th t in th Unit d St t s TFP mor th n th v r LAC countr (GDP per worker as % of that in the United States) (GDP per worker as % of that in the United States) 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Relative With US With US With US LAC Peru Mexico Chile Income Human Capital Capital TFP Actual Counterfactual Source: Caselli (2015) and authors’ calculations. Source: Caselli (2015) and authors’ calculations. Note: The counterfactual refers to output per worker that countries would have had if their TFP were equal to that of the United States. in a “middle-income trap.” The concept of MIC efficiently as the United States can, it could increase trap refers to the fact that transitions from low to its output per worker by a factor of 2.6, from 25 to 65 middle income have been more frequent than those percent of that in the United States. from middle to high income: of 101 MICs in 1960, only 13 had become high income by 2008 (see Box Given the importance of firms to productivity 2). After reaching Peru’s current level of income per growth, the rest of the report focuses exclusively 29 capita, some Asian countries were able to transition on the challenges and policy opportunities to boost to high income in less than 10 years. It took Chile firms’ productivity in Peru. In Part I, the report around 15 years, yet other countries in LAC have not dealt with some of the macro aspects of growth been able to transition in over 30 years. The limited and productivity, and highlighted some challenges transitions to high income status in the region and (e.g., infrastructure and diversification). While globally is related to the fact that the ingredients for acknowledging that convergence toward high income growth—policy and otherwise—change as countries status requires a number of important ingredients, become wealthier. Thus, the policies that propelled the rest of the report focuses on the underpinnings Peru into upper–middle–income status would not be, of firms’ productivity growth in Peru and the policy on their own, sufficient to continue or accelerate the opportunities available to foster firm productivity in convergence pace achieved over the last decade. selected areas. Maintaining a clear focus and depth of analysis on this area is a core value added of this Moving from middle- to high-income status is report. achievable for Peru, but in the context of a less favorable external environment, growth will have to rely much more on productivity. To double its GDP per capita in the next 12 years, Peru would need to grow at 6 percent annually. At that rate, it could catch up with the lower bound of OECD countries’ current income per capita in 12 years. Increasing productivity is more important for Peru than for an average country in LAC (Figure 20). On average, if LAC had productivity levels equal to those of the United States, it could almost double its output per worker, from 24 percent of that in the United States to 47 percent. If Peru, however, could use its inputs as Peru Building on Success: Boosting Productivity for Faster Growth Box 2 The middle-income trap: a policy issue In the last few decades, a significant number In essence, avoiding or escaping from the of countries transitioned from low- to middle- middle-income trap requires the evolution income status: the percentage of low-income of policymaking to recognize new economic countries in the World Bank classification fell realities and three transitions related to this from 30 to 16 percent between 1987 and 2013. evolution: diversification to specialization; This suggests that even over a relatively short investment- to productivity-led growth; time frame progress can be made in raising per and centralized to decentralized economic capita incomes. management. The first requires policymakers to promote competition and creative Yet countries in the middle-income group seem destruction. The second involves investment to languish in this category for many years in advanced education and systems to or experience decelerations or stagnations promote innovation. The third demands in growth (see Box Figure). The notion of a policymaking close to where economic “middle-income trap” was first formulated activities are occurring. to describe the situation faced by countries having difficulty navigating this transitional The East Asian countries that were able stage. Many middle-income countries (MICs) to avoid the middle-income trap managed seem to be caught between two models of these transitions. They were characterized growth, at a stage of development in which by advanced infrastructure networks and the gains to be made from labor-intensive, low a sound intellectual property rights regime value-added activities have been realized but that permitted evolution from technological 30 the systems of innovation, specialization, and imitation to technological innovation. They technological advancement necessary for high invested in skills upgrading and supported value-added, high-productivity production are Research and Development (R&D). And not yet in place. Unable to shift strategies, some they encouraged flexible labor markets and MICs fail to make a timely transition from economic openness. In general, these countries resource-driven growth, with low-cost labor adopted policies that spurred productivity. and capital, to productivity-driven growth. Several policy outcomes seem to be correlated Box Fi ur . Most LAC countri s r t kin b tw n 15-40 rs to b com n HIC ft r r chin with successful transitions to high-income P ru’s incom p r c pit l l v l status: higher levels of secondary and tertiary (GDP per capita since reaching Peru’s current level of income) education help prevent growth slowdowns; countries that avoided the middle-income trap $35000 Taiwan had more diversified, sophisticated, and non- $30000 Korea standard exports baskets, lower agricultural $25000 share of GDP or a higher industrial share, a higher trade share of GDP, and lower inequality. $20000 Other factors are also found to be important $15000 Chile Mexico to the growth of MICs: the size of government, Malaysia $10000 Argentina regulation, and infrastructure (roads and $5000 Costa Rica telephone lines); advanced infrastructure can help countries avoid the middle-income trap $0 5 10 20 25 30 35 40 by promoting a shift in production from labor- number of years since reaching Peru's current income intensive to skill-intensive activities and by Note: Horizontal line is MIC-HIC threshold. increasing their pace of innovation. Source: sta calculations on PWT 7.1. Source: Background paper prepared for this report by Moroz and Vostroknutova (2015), drawing on Gill and Kharas (2007), Kharas and Kohli (2011), Eichengreen et al. (2013), Felipe et al. (2012), Bulman et al. (2014), Aiyar et al. (2013), and Agénor et al. (2012). 31 Part II Looking closely at firm productivity in Peru Peru Building on Success: Boosting Productivity for Faster Growth 32 3. Firm dynamics and productivity 3. Firm dynamics and productivity T o understand the challenges Peru faces to increase productivity, we need to look at the microeconomic drivers of growth at the firm level. Aggregate productivity growth ultimately at the global productivity frontier (proxied by the top 25 percent most productive firms in the United States) in the same industry. This is, however, not too different from Colombia (6 percent) or Mexico (8 percent; Figure depends on firm productivity, which is driven by factors 21). Peruvian firms have also shown a relatively fast 33 within and outside the firm. Within-firm productivity convergence with the global productivity frontier: depends on the speed of technology absorption and between 2007 and 2011, Peru’s median firm reduced improvements in processes and products (innovation), its distance to the global productivity frontier by 11 which may come from the firm’s own efforts, spillovers percent (Figure 22). from other firms, or its international trading partners. Outside the firm, what matters is the efficiency of Not all industries and subsectors contribute product and factor markets that drive the allocation positively to the overall TFP growth. There is wide of resources, so that more productive firms have variance within both manufacturing and services, easier access to the necessary factors of production, where high TFP growth in some subsectors is offset can expand more easily, and place more competitive by negative growth in many other subsectors. The pressures on larger incumbent firms. number of industries with negative TFP growth rates is higher in services than in manufacturing (Figure 23). Peru shows relatively high and Wholesale and gambling services have been the most growing productivity, but with many productivity-reducing, while telecommunications, lagging industries software, and civil engineering have added the most to overall sector productivity (see Box 3). In manufacturing, Although distant from the global frontier, Peru’s some sectors (such as food manufacturing) have firm-level productivity growth is comparable to that significant roles in raising aggregate TFP both due to of other MICs and has been accelerating. A median formal17 firm in Peru is 5 percent as productive as firms these peculiarities into the analysis. Further, as shown in the recent report analyzing entrepreneurship in Latin America (World Bank, 2014), it is important to take into account general 17  It is important to note that this report’s firm-level analysis equilibrium effects as the size and growth of formal firms is likely is based on data covering formal firms only. While formal sector to influence the behavior and size of informal ones as discussed in accounts for 80 percent of GDP, its employment share is only this Section. However, throughout the analysis in this report, it is about 20 percent, according to INEI (2014). Accordingly, the important to keep in mind the caveat that— although every effort inferences we make about the overall economy are especially has been made to account for the informal sector in drawing relevant for production and value added given the coverage of conclusions—they often refer exclusively to the formal-sector formal firms. In Section 6, we make a special effort to incorporate companies. Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 21. P ruvi n firms’ productivit is t p r Fi ur 22. Firm productivit h s b n conv r in with comp r tors with th Unit d St t s (average value added per worker and TFP) (distance in value added in manufacturing per worker between median rm in Peru and United States, top 25% most productive rms, times; and change in distance, %) VA/worker, thousand 2005 USD TFP (RHS) 16 0.8 25 14 0.7 20 12 0.6 15 10 10 0.5 5 8 0.4 0 6 0.3 -5 4 0.2 -10 2 0.1 -15 0 0.0 -20 Mexico Colombia Peru Malaysia Chile SouthAfrica 2010 Colombia 2010 Czech 2009 2007 Mexico 2010 Peru 2010 2007 Thailand 2006 2007 2011 Change, percent Source: Authors’ calculations based on Enterprise Surveys. Source: Authors’ calculations: background paper prepared for this report by Iacovone and Tran (2015), based on Annual Economic Survey (Encuesta Económica Anual, EEA) by INEI. high growth and large shares of output, while others Fi ur 23. Productivit d n mics diff r (such as apparel, chemicals, or metallic products) have b s ctor nd industr reduced it. These patterns stand in contrast with those (cumulative TFP growth, 2007–12) in, for example, China, where almost all manufacturing 0.25 100% subsectors have contributed to TFP growth positively. 90% 0.20 Rubber and plastics products 34 80% Textiles There is high dispersion in productivity 70% 0.15 60% between firms and also differences by 50% 0.10 Beverages geographic region 40% Food Civil engineering 0.05 products Programming 30% Telecommunications The differences in productivity between firms are 20% 0 Wearing Wholesale trade larger in Peru than in other countries, suggesting 10% apparel Gambling and -0.05 betting activities that there might be significant anomalies in 0% -10% factor allocation. In adopting a micro-lens to -0.10 China Peru Peru assess productivity, the first crucial finding is the manufacturing manufacturing services sector sector sector degree of firm-level heterogeneity. While micro-level productivity heterogeneity is a common stylized fact Source: Background paper prepared for this report by Iacovone and Tran (2015); around the world,18 Peru is characterized by a high level and Brandt et al. (2012) for China. Note: The graph depicts contribution to TFP growth from industries (sub-sectors) of productivity dispersion vis-a-vis other countries in within manufacturing and services sectors. The industries below zero have negative TFP growth rates, and therefore slow overall TFP growth. China’s results LAC and especially the United States. In Peru, firms in are a percentage of total TFP growth. the 90th percentile of the productivity distribution are 500 percent more productive than those in the 10th to look further into this issue, because the presence percentile; this contrasts with around 200 percent of large productivity dispersion is a condition that in the U.S. (Figure 24). Such disparities mean that increases the importance of resources allocation. some firms are able to produce much more given the same inputs, within the same industry. This disparity There are also notable differences in productivity could be attributable to variations in technologies, and employment growth among firms in different processes, human capital, and managerial skills. But, regions of Peru. Firms in Lima have been on average more important, this is a sign of strong disparities in more productive than those in the Sierra and Selva the allocation of factors of production, and a reason regions, while firms in the Costa have moderately 18  Syverson (2011). 3. Firm dynamics and productivity Box 3 Telecommunications: a booming industry in services sector The value added in the telecommunications the demand for labor in the sector, which now sector has been growing faster than GDP employs more than 40,000 workers. since 2003, and accelerated even more since 2006—to 14.1 percent a year. Similarly, the These positive developments have occurred sector’s contribution to GDP growth increased even though the expansion of physical dramatically between 2006 and 2013. infrastructure has not kept pace with the growth in demand. Currently, the quality of As a result of this growth, the provision of telecommunications services can be slow, partly telecommunication services expanded rapidly due to a deficit of about 14 thousand antennas in Peru. Fixed lines increased from 4 per in Peru, according to official estimates. This 100 people in 1995 to 9.9 in 2014. Similarly, has been associated with regulatory burdens mobile lines increased from 0.3 per 100 imposed by some local governments that have people in 1995 to 103 in 2014. Internet use blocked the installation of infrastructure (see also grew, from almost none in 1994 to 40.2 Section 5 of this report). As a result, the average people per 100 in 2015. While impressive, this mobile data consumption in Peru is less than a success has yet to bring Peru to the level of quarter of that recorded in Chile, less than half telecommunication services experienced by of that in Colombia and Venezuela and is also its peer regional MICs. lower than those in Argentina and Ecuador. A new regulation was enacted in April 2015 that The rise of telecommunications has enabled seeks to unify the processes in municipalities the launch of call centers, which have had a and creates a mechanism for the automatic remarkable evolution in Peru, thanks to still approval of applications for installation of 35 competitive costs of labor and real estate. In infrastructure. The implementation of the this context, the size of the call center industry Fiber Optic National Network to be completed more than doubled between 2005 and 2011. by mid-2016 that connects 90 percent of This was largely sustained by exports of call the provincial capitals of the country would center services, which increased from 1 percent also provide much-needed infrastructure of the total turnover of the industry in 2005, to that is expected to increase penetration of more than 40 percent in 2011. This increased information technology. Box Fi ur 1. Fix d t l phon Box Fi ur 1. Mobil t l phon Box Fi ur 1. Int rn t us rs subscriptions p r 100 p opl subscriptions p r 100 p opl p r 100 p opl 30 150 80 60 20 100 40 10 50 20 0 0 0 Malaysia Peru South Chile Colombia LAC Malaysia Peru South Chile Colombia LAC Malaysia Peru South Chile Colombia LAC Africa Africa Africa 1995-1999 2000-2004 1995-1999 2000-2004 1995-1999 2000-2004 2005-2009 2010-2014 2005-2009 2010-2014 2005-2009 2010-2014 Sources: official data, INEI, WDI. Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 24. Th r r l r diff r nc s in productivit Fi ur 25. Firms in P ru row slowl , sp ci ll in s ctors l v ls cross firms in P ru oth r th n m nuf cturin (percentage di erence in productivity between the 90th (average employment by age, relative to employment of young rms) and 10th percentile of the productivity distribution) 8 7 United States 1997 6 China 2005 Argentina 2002 5 Ecuador 2005 4 Chile 2006 Bolivia 2001 3 Uruguay 2005 2 El Salvador 2005 Mexico 2004 1 Peru 2012 0 Venezuela 2001 <5 5-9 10-14 15-19 20-24 25-29 30-34 35-39 >=40 Colombia 1998 Age 0 1 2 3 4 5 6 7 8 Peru Peru manufacturing India Manuf before reforms India Manufacturing Mexico Manufracturing US Manufacturing Source: Peru data are from a background paper prepared for this report by Source: Hsieh and Klenow (2014); and background paper prepared for this report by Iacovone and Tran (2015); data for other countries are from Inter-American Aterido et al. (2015), based on Annual Economic Survey (EEA) by INEI. Development Bank (2010). Note: India data are for 1989–94 (before reforms) and 1994–2011 (after reforms). started to converge toward Lima more recently.19 formal firms may also be encouraging the growth in the There have also been notable differences between informal sector, as job opportunities are not created at firms in Lima and the Costa region versus those in the the pace that is needed in the formal sector. That is, Sierra and Selva regions in employment growth: firms in the absence of better employment prospects, many outside the coast did not grow at all (Box 4). people may end up working for themselves, fueling a 36 vicious cycle of small-size and few good jobs for future Firms grow slowly, indicating rigidities job seekers.22 Experience in other countries shows in factor allocation that reforms can have a strong impact on mitigating stunted growth: India, for example, shows a striking Firm grow slowly in Peru, and that may also be difference in firm growth before and after reforms encouraging informality. In countries where input directed at improvements in licensing, privatizations, and output markets work well, firms that survive and increasing efficiency of the trade system that also experience rapid growth, a process that is best led to increased competition and productivity growth described as creative destruction.20 This sort of (Figure 25).23 These reforms are directed at improving pattern is observed in the United States, for example, factor allocation between firms, as discussed in the where old firms (i.e., those 40 years old or more) are next sections. about eight times larger than young startups. In Peru, however, there is evidence of stunted growth: old Peru, unlike other countries, seems to show a companies are, on average, only about twice the size of negative relationship between firm size and younger ones (Figure 25).21 This finding indicates that, productivity. In an environment where firms have beyond high costs of entry, the costs of operation are access to appropriate inputs and output markets large and hamper firms’ growth. Stunted growth in are efficient, productive firms can expand while less productive firms are likely to stay small, or ultimately 19  These geographical regions translate as Costa (coastal), exit. The expectation would be to find older and larger Sierra (mountainous), and Selva (jungle) and cover 24 regions firms that are more productive (as they have survived (“departments”) and one constitutional province. See Box 4. 20  The term “creative destruction” was coined by Schumpeter and grown) and young firms that are, on average, (1942). The growing modern literature on the life cycle of firms less so.24 There are only a few young Peruvian firms: and the relationship with productivity is reviewed in, for example, Akcigit et al. (2015). 21  Hsieh and Klenow (2014) construct a synthetic panel with 2 22  See World Bank (2014) for a framework in the regional years of manufacturing data. They assume that the same cohort context. experiences similar rates of exit and growth. They compute 23  See, for example, Krishna and Mitra (1998). growth of the synthetic cohort weighted by sector. We regress age 24  This pattern is consistent with a large class of Input-Output categories on size (log) and control for sector, location, and year. models of industrial dynamics as in Jovanovic (1982). 3. Firm dynamics and productivity Box 4 Geographic dimensions of productivity in Peru Peru is geographically diverse with three main commerce, and services. Almost half (47 areas: Costa (coastal), Sierra (mountainous), percent) of all firms are in Lima, including and Selva (jungle), which cover 24 regions 60 percent of manufacturing, 38 percent (“departments”) and one constitutional of construction, and 65 percent of services province. Economic activity is concentrated firms. Following Lima, the main regions are in the coastal area, which contributes almost Arequipa (5.2 percent of GDP), La Libertad 80 percent of national output. Also there, the (4.2 percent), and Piura (3.6 percent). These regions of Lima and Callao make up the Lima have diverse economic profiles: Piura and Metropolitana geographic area, which accounts La Libertad have a large contribution from for 34 percent of the country’s population and primary activities, and Arequipa has a strong 48 percent of national output. industrial sector. Peru’s main ports are also in these regions: Callao (Lima), Matarani The coastal region benefits from proximity to (Arequipa), Salaverry (La Libertad), and Paita the Pacific, greater population density, and (Piura); activity of the river ports in the Selva better-developed infrastructure. Secondary is small. (manufacturing) and tertiary (services) activities predominate there. Tertiary activities, Lima Metropolitana is the most productive closely linked to domestic demand, are also region (Box Figure 1). Other areas—Costa relatively sophisticated in the coastal area, (excluding Lima Metropolitana), Sierra, but underdeveloped in the Sierra and Selva and Selva—were about 30 percent less areas. Primary activities are more widespread productive than Lima Metropolitana in 2012. geographically. Although the valleys of the But Costa and Sierra have been catching up. 37 north coast lead in agribusiness, the Sierra Productivity is more homogeneous across is the main producer of crops for domestic regions in the non-tradable sectors than consumption as well as the main metals in manufacturing (Box Figure 2). There are mining region. Oil extraction is concentrated in also notable differences between firms in the Selva. Lima and the Costa region versus those in the Sierra and Selva regions in employment Lima Metropolitana is the largest city, and growth: firms away from the coast did not the focus of manufacturing, construction, grow at all (Box Figure 3). Box Box ur ur Fi Fi 1. Productivit 1. Productivit is hi is hi h rh in in Lim r Lim , , Box Box ur ur Fi Fi 2. Productivit 2. Productivit cross cross ions r rions is mor is mor butbut n nr ll r ll pickin pickin in oth upup in othr r r rions ions homon n homo ous ous in non-tr in non-tr d d s ctors bl bl s ctors In (TFP) In (TFP) by region by region TFPTFP productivity productivity levels, 2007-2012 levels, average 2007-2012 average 2.4 2.4 2.5 2.5 2.2 2.2 2.0 2.0 2.0 2.0 1.8 1.8 1.5 1.5 1.6 1.6 1.0 1.0 1.4 1.4 0.5 0.5 1.2 1.2 1.0 1.0 0.0 0.0 2007 2008 2009 2007 2008 2009 2010 2010 2011 2011 2012 2012 Manufacturing Manufacturing Construction, Construction, trade trade andand services services Total Total Costa Costa Lima Lima metropolitana Selva metropolitana Selva Sierra Sierra Costa Lima Costa Metropolitana Lima Metropolitana Selva Selva Sierra Sierra Peru Building on Success: Boosting Productivity for Faster Growth Box 4 Geographic dimensions of productivity in Peru (cont.) Overall—and based on National Accounts data productivity. This helps explain the relative that include the informal sector—there has performance of geographic areas in overall been some gradual convergence on output productivity dynamics (Costa is a leading per worker between regions, but the large manufacturing producer while Sierra has gaps remain vis-à-vis Lima. Departments with most of the mining—Box Figures 1 and 4). With lower labor productivity in 2001, on average, large regional differences in labor productivity, experienced faster growth. The convergence a process of sustained productivity catchup in labor productivity in manufacturing and by lagging regions has much potential to raise mining has been sizable enough to lead to aggregate productivity and incomes, and to convergence in aggregate departmental labor reduce regional inequalities. Box Fi ur 3. Onl firms in th co st l r row Box Fi ur 4. R ions r conv r in on m nuf cturin v lu dd d p r work r (employment di erences between young and old rms, log of employment) (manufacturing productivity annualized growth rate, %) 2.5 8 CUS ICA 7 6 TUM CAJ AMA 5 AYA 2.0 4 MAR LOR 3 PIU ARE PUN LAM 2 LIB LIM HUA 1 TAC ANC 1.5 0 HUC 38 -1 9 10 11 12 MAD JUN -2 UCA -3 1.0 -4 MOQ young age old Costa Selva Sierra Metropolitan Log of initial VA per worker Sources: Background papers prepared for this report by Iacovone et al. (2015) and Iacovone and Tran (2015), based on data provided by EEA/INEI and national accounts. 10 percent of firms are between 1 and 5 years of age, firms. This problem is sometimes called “factor while two-thirds are between 6 and 19 years old. These misallocation.” Productivity may grow by increasing young firms, however, are the most productive—a workers’ output within a firm (within-firm), or by somewhat counter-intuitive finding given international employees (and capital) moving from less efficient experience. If productive firms grow, there should be to more efficient firms (between firms). Section 4 a positive relationship between size and productivity. decomposes productivity growth into that driven by But Peru seems to have a negative correlation between within-firm productivity growth and that driven by the size and productivity (Figure 26). TFP by firm size does efficiency in allocation of factors of production between not differ significantly on the different dimensions firms. It suggests policy avenues that could reduce (geographic area or economic sector) and similar misallocation, and estimates the potential impacts of relationships hold when productivity is measured as reducing misallocation on productivity and economic value added per worker. growth. The findings above point to structural and market anomalies that tend to channel labor and other factors of production into less-productive 3. Firm dynamics and productivity Fi ur 26. Th r is n tiv r l tionship b tw n firm si nd productivit in P ru (relationship between productivity (TFP or VA/worker) and employment) All formal rms TFP by geographic location TFP by economic sector 2.25 5.5 2.25 2.0 2.0 2.0 1.8 1.75 1.75 VA worker (log) 1.6 TFP TFP TFP 1.50 4.5 1.50 1.4 1.25 1.25 1.0 1.0 1.2 0.75 2.5 0.75 1.0 0 2 4 6 8 10 0 2 4 6 8 0 2 4 6 8 Employment (log) Employment (log) Employment (log) 95% CI Total factor productivity Costa Selva Manufacturing 95% CI Value added per worker (log) Sierra Metropolitan Services Source: Authors’ calculations: Background paper prepared for this report by Aterido and Iacovone (2015), based on EEA data by INEI. Note: Weighted local Epanechnikov kernel polynomial smoothing is used. The sample of large rms is small and results at that end of the distribution should be interpreted with care. Standard errors of large rms in Costa, Selva, and Sierra are larger than 0.5. 39 Peru Building on Success: Boosting Productivity for Faster Growth 4. Issues with factor allocation M isallocation of factors is usually to blame for the stunted growth and other inefficient firm dynamics described in Section 3. Main causes of misallocation are inefficiencies in product and factor markets, which are Inefficient allocation of factors dragged down aggregate productivity growth, which was driven by within-firm productivity improvements Peru’s overall productivity growth is driven by 40 induced by poor competition, inefficient regulations, within-firm productivity but dragged down by or other market imperfections—such as those in labor inefficient allocation of labor between firms (or and credit markets. A specific type of misallocation “misallocation”). As discussed earlier, productivity responsible for stunted growth is that which may grow by increasing workers’ output within a discourages firms from investing, growing, and raising firm or by employees (and capital) moving from less productivity.25 A possible set of causes driving this efficient to more efficient firms. Figure 27 shows the misallocation include contractual frictions in hiring results of the productivity decomposition for Peru, nonfamily workers,26 financial market imperfections which breaks the drivers of productivity growth limiting access to capital to better-connected (and into those originating from within-firms and those not necessarily more productive) entrepreneurs,27 coming from various allocation issues.33 While overall difficulty in recruiting skilled managers,28 higher tax enforcement on larger firms,29 and difficulty in 33  Productivity growth and its sources cannot be observed access to land and its misallocation.30 Other research directly but must be inferred using statistical techniques. The purpose of this analysis is to estimate productivity of firms suggests that managerial quality is an important and sectors and the factors that lead to rising productivity. The factor in empowering transformative entrepreneurs Annual Economic Survey (EEA) has data for 2007–12 on firms’ inputs, outputs, and productivity. These can be grouped to who strive to grow, and thus has a strong impact on estimate productivity by sector. Using statistical techniques from allocation.31 In India, for example, an intervention to Restuccia and Rogerson (2008) and Olley and Pakes (1996), it is possible to split productivity into an unweighted average of firm- improve management quality led to a 12 percent TFP level productivity in a sector and a term that measures whether growth after a year.32 labor is flowing to the most efficient firms (covariance between firm’s productivity and its market (or input) share). The covariance term measures whether resources are efficiently allocated 25  Hsieh and Klenow (2014). between firms: a positive value indicates that more productive 26  Recent research provides some evidence on this; see, for firms have higher market share, while a negative value indicates example, Caria and Falco (2014) for the case of Ghana. that less productive firms do. An increase in the covariance term 27  Virgiliu and Xu (2014); Buera and Shin (2010). would therefore mean improvements in labor allocation, that is, 28  Bloom et al. (2013). that labor flows increasingly to more productive firms. Depending 29  Caprettini (2014), Leoal-Ordónez (2013). on the magnitude of the relationships, the statistics can also 30  Duranton et al. (2015). impute increases in aggregate productivity that would result 31  Akcigit et al. (2015). from moving workers from less productive to more productive 32  Bloom et al. (2013). firms or sectors. 4. Issues with factor allocation Fi ur 27. Within-firm productivit rowth driv s ov r ll Fi ur 28. P ru’s n tiv imp ct of f ctor lloc tion productivit rowth, but is dr d down b f ctor mis lloc tion si n ls s rious m rk t distortions (contributions to aggregate value added per worker growth, %) (contributions to aggregate value added per worker growth, %) 200 27 145 Between rms factor allocation Within rms productivity 150 25 95 100 23 50 21 45 0 19 -5 -50 17 -100 15 -55 2007 2008 2009 2010 2011 2012 United Kingdom United States Netherlands 1993-2001 Colombia 2012 1993-2001 Germany 1993-2001 1993-2001 Mexico 2012 Hungary 1993-2001 1993-2001 Slovenia 1993-2001 Romania 1993-2001 Peru 2007-2012 France Factor allocation contribution, % Within- rm productivity contribution, % Labor productivity (RHS, thousands USD) Source: Authors’ calculations: Background paper prepared for this report by Iacovone Source: Authors’ calculations: Background paper prepared for this report by Iacovone and and Tran (2015), based on EEA by INEI. Tran (2015), based on EEA by INEI; Brown et al. (2015); and Bartelsman et al. (2009). productivity grew, it was driven by increases in within- services might have a less competitive environment firm productivity, while it has been dragged down by or find domestic regulations more binding than inefficient allocation of factors of production. This firms in manufacturing, which are subject to foreign misallocation effect is far stronger in Peru than in competition. Mexico, Columbia, and other countries (Figure 28). Misallocation in the services sector is also reducing 41 Misallocation in the services sector productivity growth indirectly by constraining other harms aggregate productivity directly sectors’ growth that rely on it for inputs. Services— and through intermediate product especially “other business services”—are an important markets input into other sectors, including manufacturing. If services are overpriced or of low quality due to Inefficient allocation of factors among firms has poor allocation of resources, other sectors, such as limited aggregate productivity growth, especially manufacturing, would tend to rely less on these external in the services sector. Firm-level productivity grew in inputs and sub-optimally produce them internally (e.g., both manufacturing and services sectors in 2007–12, transport services). This hampers the productivity but growth was slower in services. In manufacturing, growth that comes from specialization, because firms the allocation of factors between firms went from would produce goods and services outside their core contributing negatively to positively (Figure 29), a competencies and inside their production possibility critical trend in this sector. But in services, the slower frontiers. The Peruvian services sector contributes only growth in productivity is fully accounted for by the 4.8 percent of domestic manufacturing value added high negative contribution from allocation of labor and only 8.4 percent in exported manufacturing value between firms (Figure 30), as labor moved from more added, which is much lower than in other countries to less productive firms. (Figure 31). In exports—where services contribute only 19 percent of exported value added—it performs There are several possible explanations for the poorly versus comparators: Thailand (30 percent), Chile differences in factor allocation. Because rigidities (32 percent), Colombia (26 percent), South Africa (42 in product and factor markets usually cause percent), and even China (26 percent). Peru’s impact misallocation, these in turn could be due to sector- of allocative inefficiencies in services is especially large specific regulations. Common regulations could relative to other countries in LAC. Reducing these also become more binding in the services sector distortions would increase TFP in Peru by an estimated due to its higher labor intensity (such as hiring and 20 percent, far more than in Mexico and Colombia firing constraints); and as a non-tradable sector, (below 5 percent) and Chile (10 percent) (Figure 32). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 29. Alloc tion of f ctors in m nuf cturin Fi ur 30. S rvic s h v s n no improv m nt in h s improv d, boostin productivit rowth lloc tion, dr in down r t productivit rowth (manufacturing) (services) 28 30 30 0 28 26 20 26 24 -50 22 000 USD/employee 24 10 20 18 -100 22 0 16 14 20 -10 12 -150 10 8 18 -20 6 -200 4 16 -30 2 2007 2008 2009 2010 2011 2012 0 -250 2007 2008 2009 2010 2011 2012 Factor allocation contribution, % Aggregate productivity Factor allocation contribution, % Aggregate productivity Source: Authors’ calculations: Background paper prepared for this report by Source: Authors’ calculations: Background paper prepared for this report by Iacovone and Tran (2015), based on Annual Economic Survey (Encuesta Iacovone and Tran (2015), based on Annual Economic Survey (Encuesta Económica Anual, EEA) by INEI. Económica Anual, EEA) by INEI. Fi ur 31. S rvic s in P ru do not provid suffici nt Fi ur 32. P ru is mon th countri s th t would b n fit int rm di t inputs to support th m nuf cturin s ctor most from limin tin distortions in s rvic s (total services value-added into manufacturing, (TFP gains from removing distortions, %) domestic and exports, 2011) 30 0.35 0.30 25 42 0.25 20 0.20 15 0.15 0.10 10 0.05 5 0.00 0 -0.05 PER COL MYS THA CHL ZAF CHN Brazil Mexico Argentina Colombia Chile Ecuador Peru Bolivia Linkage domestic Linkage in exports Source: Authors’ calculations: Background paper prepared for this report by Saez Source: Authors’ calculations: Background paper prepared for this report by Fattal et al. (2015), based on Exports Value Added (EVA) Database. (2015), based on Censo Nacional Económico (CENEC) 2008 by INEI. Removing distortions in product and Chile on output per worker (Figure 34). To understand factor markets would boost Peru’s some broad dimensions of the impact of productivity productivity improvements, consider the following: Significant gains in productivity can be achieved • If labor were reallocated as efficiently as in the United by improving the allocation of factors between States, productivity (value added per worker) could firms. When resources are misallocated, the economy increase by 130 percent (Figure 33).34 Output per operates inside its production possibility frontier. TFP worker would more than double, making Peru close will be lower, and the economy will produce less output to the hypothetical level if it had U.S. levels of TFP with its resources than it is capable of. Removing and well above that of Chile (compare Figure 20 and rigidities would yield additional productivity gains. In Figure 34). Peru, the benefits from eliminating distortions are large, • Further, if the distortions in intermediate product estimated at 25–130 percent of current productivity. markets were eliminated, TFP could increase by Productivity improvements of this magnitude could 34  This case is likely to include market distortions as well as place Peru well above the LAC average and at par with capture other inefficiencies. See Iacovone and Tran (2015). 4. Issues with factor allocation Fi ur 33. P ru could si nific ntl incr s productivit Fi ur 34. Such n incr s in productivit could mor b r movin lloc tiv distortions… th n doubl P ru’s output p r work r (value added per worker, real 2005 US$; TFP, units; increase in productivity, %) (GDP per worker relative to the United States, %) 250 70 60 200 50 150 40 132 100 30 62 20 50 54 10 25 0 0 Total VA/worker / Manufacturing TFP / Overall TFP / U.S. allocation Chile Actual With U.S.'s TFP of labor Random allocation of labor 25 % TFP increase 62% manufacturing TFP increase LAC Mexico labor allocation HK framework intermediate product markets only Actual aggregate labor productivity (average 2007-2012) If eployment is allocated as in the US (50% contribution of labor reallocation) Productivity gain, percent of current Productivity gain from random allocation Source: Background papers prepared for this report by Fattal (2015) Actual relative and Iacovone and Tran (2015). Peru income for Note: TFP estimates are based on the CENEC 2008 by INEI. Value-added estimates are comparable countries based on EEA average over 2007–12 and assume that labor allocation contributes Source: Authors’ calculations using Caselli (2015). half to aggregate productivity, as in the United States. 25 percent. This would correspond to an inter- taxes, labor market segmentation, other labor market industry production function the same as that of rigidities, poor management, incomplete information, 43 the United States.35 The existing distortions lead to unions, lack of competition, and regulatory biases, prices for intermediate goods that are high, resulting among others.36 in relatively low services inputs into the exports of other sectors, as well as relatively low direct services Part III examines these “usual suspects” further and exports. The 25 percent increase in TFP would mean suggests ways to overcome the obstacles imposed by a similar percentage increase in Peru’s GDP per them: worker (Figure 34). Eliminating these distortions could be additional to that in manufacturing, leading • Section 5 looks at competition and regulation. to a combined increase of around 35 percent in Barriers to competition in upstream or downstream output per worker (see below). markets can distort prices of intermediate and • Finally, if the distortions in manufacturing could final goods and lower returns to investment or be eliminated, that sector’s productivity could to factors of production. Excessive or inefficient increase by 62 percent. These distortions generally regulations can institutionalize rigidities and include financial and labor market frictions, or inflexibility in factor and product markets, unfair advantages gained by certain types of firms create sectorial or geographical discretion or sectors, due to discretion that may effectively and distortions, or create unfair advantages amount to subsidies. This increase in manufacturing for select firms, entrepreneurs, or activities. productivity would mean an 8 percent gain in • Section 6 takes a deeper look at the labor market economy-wide productivity. and the quality of education. Inflexible or segmented labor markets could cause significant labor Behind the rigidities in product and factor markets misallocations. Low-quality human capital or weak are a set of “usual suspects” that prompt an education and training could reduce the supply of inefficient allocation of factors between firms. These labor for higher value-added jobs. market frictions can arise for a variety of reasons: • Section 7 highlights some issues that might restrict 35  This case implies assumptions about Input–Output structure of the economy. See Fattal (2015). 36  Jones (2015). Peru Building on Success: Boosting Productivity for Faster Growth access to credit, because unequal, expensive, or insufficient access to credit could prevent firms from innovating and growing. The last two sections explore further opportunities to reduce misallocation of factors to boost productivity in areas that are typically examined only in the context of within-firm productivity growth. • Section 8 looks at patterns of international trade, since international trade is an important driver of productivity growth at firm level, through the spillovers offered by import of technologies from abroad. Trade also is a good instrument to correct misallocation of factors of production, as shown by the Peruvian manufacturing sector. • Section 9 analyzes constraints to innovation since innovation is one of the strongest drivers of firm-level productivity growth in other countries. Innovation can also illustrate the benefits of removing distortions—especially if this means more equal access to inputs necessary for innovation. 44 Part III Correcting misallocation of resources to raise productivity growth Peru Building on Success: Boosting Productivity for Faster Growth 46 5. Addressing regulatory and competition issues 5. Addressing regulatory and competition issues P roductivity has been growing, but with more efficient allocation of factors of production between firms, productivity growth would have been higher. Inefficient allocation of resources is reflected in the slow growth of firms and in allocation Strong overall progress on competition policy and pro-competition regulation The reforms undertaken in the last 20 years laid the foundations for Peru’s solid competition policy. 47 of labor to less- productive firms. In this part, we look These structural reforms have introduced a best- at factors that may well have reduced aggregate practice competition legal framework, embedded productivity growth, under three areas: regulation and competition principles in network sectors regulation, competition issues, labor market rigidities, and access reined in state economic activity, and put in place a to credit. unique ex-post control mechanism for regulations Competition drives efficient resource allocation, but may be restricted by regulations that unduly Fi ur 35. P ru scor s hi h r th n th LAC v r obstruct firms’ entry, operations, and growth. on lmost ll pill rs of th GCI Competition in the marketplace contributes to (absolute values from 0 to 7, higher values re ect more competitive policies and institutions) productivity growth: it shifts market shares toward more efficient producers (between-firm), and it induces 1st pillar: Institutions 7 2nd pillar: Infrastructure firms to become more efficient so as to survive (within- 12th pillar: Innovation 6 firm).37 Market competition can be enabled but also 11th pillar: 5 3rd pillar: Business 4 Macroeconomic restricted by regulations. Regulations are key to 3 sophistication environment addressing market failures and achieving other valid 2 policy objectives. However, some regulatory options 1 4th pillar: Health 10th pillar: and primary unnecessarily limit the number of competitors, Market size education facilitate anti-competitive practices, and lock in unfair 5th pillar: Higher 9th pillar: advantages for particular firms, entrepreneurs, or Technological education and readiness training activities. As a result, upstream markets—such as 6th pillar: services—would not develop fully and downstream 8th pillar: Financial Goods market e ciency market development 7th pillar: firms might be less competitive than their foreign Labor market e ciency rivals and less likely to compete globally. LAC South Africa Peru Mexico Colombia Brazil Source: WEF, Global Competitiveness Report 2014–2015. 37  Kitzmuller and Licetti (2013). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 36. P ru p rforms b tt r th n th LAC v r Fi ur 37. But it l s b hind in som s ctors on th ov r ll product m rk t comp tition nd (OECD Product Market Regulation (PMR) Indicator; (OECD Product Market Regulation (PMR) Indicator; absolute values from 0 to 6, higher values are associated absolute values from 0 to 6, higher values are associated 3 with regulations more restrictive to competition) with regulations more restrictive to competition) Electricty 4 2 0.55 3 0.66 2 Road 1 Gas 0.13 0.87 1 0.70 0.43 0.58 0.58 0.82 0 OCDE TOP 5 a/ Peru LAC Average b/ Airlines Telecom State Control Barriers to Barriers to trade Peru OECD Top 5 a/ LAC average b/ entry and rivalry and investment Source: World Bank/OECD PMR data 2013–2014; Background paper prepared for this Source: World Bank/OECD PMR data 2013–2014; Background paper prepared for this report by Licetti et al. (2015). report by Licetti et al. (2015). Notes: a/ Top 5 OECD countries are the Netherlands, United Kingdom, Austria, Notes: a/ Top 5 OECD countries are the Netherlands, United Kingdom, Austria, Denmark, and New Zealand. b/ LAC countries include Argentina, Brazil, Chile, Denmark, and New Zealand. b/ LAC countries include Argentina, Brazil, Chile, Colombia, Costa Rica, El Salvador, Honduras, Jamaica, Mexico, Nicaragua, and Peru. Colombia, Costa Rica, El Salvador, Honduras, Jamaica, Mexico, Nicaragua, and Peru. that unduly burden private initiative. The Commission to trade (Figure 36). Peru’s extent of state control of Elimination of Bureaucratic Barriers of INDECOPI is comparable to the OECD’s, consistent with the (Instituto Nacional de Defensa de la Competencia government’s policy of encouraging growth through 48 y de la Protección de la Propiedad Intelectual) can enabling private initiative. However, as discussed declare an existing public norm or administrative act just below, barriers to competition are hindering well- as non-compliant with legal framework (illegal) or functioning markets in important service sectors. unreasonable, and sanction the responsible institution Although product market regulations in Peru are or public servant. generally conducive to competition, for example in the network sector (electricity, airlines, and Peru performs well on competitiveness indexes. telecommunications, Figure 37), these lessons learned On the Global Competitiveness Index (GCI), Peru’s will need to inform regulation of emerging network performance in the macroeconomic environment is sectors, such as gas and payment systems. Despite strong even relative to the OECD average (Figure 35). having liberalized economy for trade, some barriers to Financial market development in Peru outperforms FDI and trade facilitation remain. that of major LAC countries such as Brazil, Colombia, Mexico, as well as the LAC average. Peru has very good But some barriers still hamper product performance in improving its business regulations, market competition as shown by the overall ranking in Doing Business, as it has been steadily improving its overall distance Some services sectors remain shielded from to frontier score from 67.2 in 2010 to 72.11 in 2015 competition. Barriers to entry and rivalry are still (Figure 40). According to Doing Business 2015, Peru restrictive to competition in certain services sectors ranked 35th out of 189 economies and 2nd in LAC on (Figure 38). Unlike in Chile, Mexico, and Colombia, ease of doing business, showing impressive progress professional service providers in Peru self-regulate the from 71st in the world in 2006. entry conditions for new service providers. Professional services (such as lawyers, accountants, architects, Peru has made great strides toward OECD and engineers) are licensed by professional bodies and standards in competition policy. On product market it is a mandatory requirement to be a member of the regulation (PMR) indicators, Peru compares well in association. Also, there are more tasks than in the areas of pro-competition network sector regulation, OECD on which these professionals have a monopoly or limited state economic activity, and general openness shared monopoly right. By some estimates, removing 5. Addressing regulatory and competition issues Fi ur 38. Compl xit of r ul tor proc dur s, r ul tor prot ction of incumb nts, nd b rri rs in s rvic s top th list of b rri rs to ntr nd comp tition (OECD PMR Indicator decomposition for barriers to entry and rivalry; absolute values from 0 to 6; higher values are associated with regulations more restrictive to competition) 3.0 3.0 3.0 2.5 2.5 2.5 Communication and simpli cation 2.0 2.0 2.0 of rules and procedures Barriers in services sectors 1.5 1.5 1.5 Legal barriers 1.0 Antitrust exemptions 1.0 Administrative burdens 1.0 Licence and for sole proprietor rms permits system 0.5 Barriers in network 0.5 0.5 sectors Administrative burdens for corporations 0.0 0.0 0.0 Regulatory protection of incumbents Administrative burdens on startups Complexity of regulatory procedures Source: World Bank/OECD PMR data 2013–2014; Background paper prepared for this report by Licetti et al. (2015). Note: The gure depicts decompositions of PMR indicator for barriers to entry and competition. The larger the share of a barrier is, the more restrictive this barrier is to competition. For example, in complexity of regulatory procedures, the license and permits system is a stronger restriction on competition than are communication and simpli cation of rules. Fi ur 39. B rri rs to FDI nd tr d f cilit tion issu s h v h mp r d comp tition in th tr d bl s ctors (OECD PMR Indicator decomposition through barriers to trade and investment; absolute values from 0 to 6; higher values are associated with regulations more restrictive to competition) 3.0 3.0 2.5 2.5 49 Tari barriers 2.0 2.0 1.5 1.5 Di erential treatment of foreign suppliers 1.0 Barriers to FDI 1.0 Barriers to trade 0.5 0.5 facilitation 0.0 0.0 Explicit barriers to trade and investment Other barriers to trade and investment Source: World Bank/OECD PMR data 2013–14; Background paper by Licetti et al. (2015). Note: The gure depicts decompositions of PMR indicator for trade and investment. The larger the share of a barrier is, the more restrictive this barrier is to competition. For example, in explicit barriers to trade and investment, barriers to FDI are a stronger restriction on competition than tari s. these restrictive regulations would increase value Some barriers that limit trade and foreign entry, added in 14 Peruvian sectors that use professional hamper competition in the tradable sectors. Barriers services intensively by 0.8 percent, equivalent to 0.2 to trade and investment are concentrated in FDI percent of GDP.38,39 regulations and trade facilitation (Figure 39). Peru has eliminated tariff barriers and ensures foreign suppliers’ equal treatment. But it lags behind in trade 38  Barone and Cingano (2011). 39  “Intensive use” is defined as an above-average technical facilitation (see also Section 6): it has not agreed with coefficient with respect to professional services in the input- other countries on mutual standard recognition nor output matrix. The 14 sectors are chemical and chemical products; electric machinery and apparatus; information service equivalence of regulatory measures in key services activities; other transport equipment; electricity, gas, and water sectors such as manufacturing energy, distribution, supply; transport and storage; post and telecommunications; finance and insurance; renting of machinery and equipment; maritime and air transport, telecommunications health and social work; public administration; other business (fixed and mobile), insurance, banking, and hotels activities; other community social and personal services; and professional, scientific, and technical activities. and restaurants. Removing these barriers would Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 40. P ru h s shown stron pro r ss on ov r ll Fi ur 41. … but r ul tion r m ins k probl m Doin Busin ss indic tors … for busin ss (distance to frontier score, 2010–15, 0 to 100, 100 as best performer) (top barriers to doing business, % of responses) 73 60 72.11 72 50 71 14.9 14.4 12.2 40 11.8 13.6 % of responses 70 30 69 15.3 16.8 16.1 15.7 15.1 20 68 67.2 67 10 19.4 16 17.5 19.6 21.8 66 0 2010-2011 2011-2012 2012-2013 2013-2014 2014-2015 DB2010 DB2011 DB2012 DB2013 DB2014 DB2015 Restrictive Labor Regulations Tax Regulations Ine cient Government bureaucracy Corruption Source: Doing Business (World Bank). Source: Executive Opinion Survey 2014 (World Economic Forum). improve competition in the tradable sectors and help bureaucracy. Peru’s performance on structural increase imports and exports—key elements for Peru’s issues, business sophistication, and innovation is productivity growth (see Section 7). weak. Moreover, inefficient government bureaucracy and corruption have been consistently ranked as top Government bureaucracy, constraints for firms (Figure 41).41,42 Importantly for 50 labor market regulations, and productivity, labor regulations have been consistently implementation and enforcement among the top three barriers to doing business of the legal framework remain (Section 8). problematic Bureaucratic barrier —often imposed Barriers to entry, and high costs of operation by subnational governments—stifle and expansion hold down the growth of small firm entry, growth and competition and medium firms. The share of firms reporting business licensing and permits as a major constraint Illegal or unreasonable implementation of the contributes to overall complexity of regulatory law does not only obstruct investments and raise procedures and is considerably higher in Peru, at costs for businesses but also harms competition. 20.4 percent, than the OECD average of 5 percent.40 Estimations by INDECOPI show that bureaucratic Medium firms are most affected in their ability to barriers generate an estimated cost of 0.1 percent grow, expand, and innovate. Nearly 30 percent of such of GDP for citizens and companies either directly or firms identified business licensing and permits as a indirectly affected.43 The majority of these cases major constraint, against 17 percent of small firms restrict competition to a significant degree by (i) and 15 percent of large firms. The most recurrent limiting the number of firms in the market or inhibiting obstacle had to do with procedures “for a technical private initiative, (ii) imposing rules on modes of excuse,” meaning obstacles caused by a public agency service provision that increase the business risk or that blocks a specific authorization or report in the facilitate anti-competitive practices, or (iii) instituting procedures. Administrative burdens on startups are rules that discriminate against certain companies. reflected mainly in barriers in the services sector. 41  World Bank Enterprise Surveys 2010 and World Economic Forum Indicators 2015. More progress can be achieved on key regulatory 42  WEF 2014–2015. 43  INDECOPI Observatorio de Mercados (2014), areas, such as the labor market and government Midiendo el costo económico potencial de las barreras burocráticas en el Perú, available at http://www.indecopi. g o b . p e/ re p o s i t o r i o a p s/ 0/ 0/ j e r/ p u b l i c a c i o n e s q s/ 40  The average in LAC is 18.4 percent (Enterprise Surveys 2010). ObservatorioCostoBarrerasBurocr%C3%A1ticas2013.pdf 5. Addressing regulatory and competition issues Fi ur 42. Most compl ints bout bur ucr tic b rri rs Fi ur 43. Most of th bur ucr tic b rri rs t th r inst bodi s oth r th n th c ntr l ov rnm nt subn tion l l v l r not in lin with th l l fr m work (% out of 613 processed complaints, 2013) (out of 613 processed complaints, 2013) 1.3% 0.2% 24% 76% 98.5% Not in line with legal framework Irrational Local governments and Central government Not in line with legal framework and irrational decentralized bodies Source: INDECOPI 2013. Source: INDECOPI 2013. Note: processed complaints are rms’ claims about a bureaucratic procedure that has been declared illegal or irrational by INDECOPI. For example, between 2013 and 2014, the Transport top reasons for the complaints were restrictions to Ministry effectively closed the market for further operation of new transport businesses, suspension entry on 50 different interstate bus routes. of procedures after initial commencement of the registration process, and illegal requirements for Many bureaucratic barriers were imposed by operating licenses. The number of complaints subnational government (especially at the increased dramatically in 2014: through July, more 51 municipal level) or other non-central bodies. than 1,200 bureaucratic barriers were declared Important aspects of market access and investment non-compliant with the legal framework and/or are the responsibility of subnational governments, unreasonable. most often municipalities. Procedures relating to company startup, licensing or permits, inspections, Whether imposed by national or subnational bodies, and access to land, for example, commonly require the majority of cases of undue implementation of procedures at national and subnational levels, even legal frameworks affect sectors that are critical when subnational technical implementation capacity for the development of local markets. In 2013, 70 may be weak. INDECOPI processed 613 complaints percent of the cases of undue secondary norms, on bureaucratic barriers in 2013; of these, 76 percent temporary dispositions, or general administrative were on barriers imposed by local governments and acts affected key sectors for the development of local decentralized bodies. Most of these barriers were markets such as transport, telecommunications, classified by INDECOPI as non-compliant with the retail, construction/real estate, or tourism, limiting legal framework (Figure 42 and Figure 43).44 The the provision of adequate (higher-quality and lower-cost) services, and efficiency-enhancing 44  The database includes claims identified at the national labor allocation.45 For example, the second mobile level—the headquarters Commission to Eliminate Bureaucratic operator is constrained in its network expansion Barriers (Comisión de Eliminación de Barreras Burocráticas, CEBB)—and 12 subnational CEBBs. It is classified by sector as US$80 million in investments were stalled when (14), illegal or unreasonable, and type of barrier. Barriers deemed municipalities denied it permits to install antennas. As illegal include suspending authorizations to provide passenger interstate road transport in the national road network; requiring of November 2013, 10 out of 21 larger municipalities proof of a minimum capital of $1.2 million to offer public had no procedure in place to authorize, for example, transport for passengers; and requesting presentation of a marketing, financial, and administrative feasibility study for any telecommunication infrastructure. And almost one- route starting or ending in Lima. Barriers deemed unreasonable third of all municipalities did not comply with the legal include requiring transport companies to have authorized land terminals and stations at both ends of the route and commercial framework on operational permits for commercial stops to access and remain in the sector; requiring vehicles to be no more than three years old to access the market for public transportation of goods; and restricting opening hours. 45  INDECOPI (2014). Peru Building on Success: Boosting Productivity for Faster Growth establishments in 2014.46 These local barriers raise the to local development affect investment, entry, and cost of doing business and hamper competition. competition. There is currently no legal mechanism that can effectively address this issue with an ex The mechanisms in place to enforce the removal ante approach. Therefore, based on best international of bureaucratic barriers have been limited. The practice, a mechanism to ensure that the potential scope of INDECOPI’s decision as a result of a regulatory impact on competition is taken into account particular complaint is limited: INDECOPI can only could be a regulatory option to check the quality of order the government agency to revoke the specific regulations issued by the executive. Covering a range administrative action declared a bureaucratic barrier. of regulatory goals, this system should assess (as part As other parties would need to initiate a complaint of the potential impacts of regulations) the specific themselves and wait for their case to be resolved, the effects that the regulation could have on competition same administrative action has been declared illegal and markets. Once the general RIA regime is enacted, or unreasonable more than 140 times.47 Further, the operating guidelines should be promulgated to Free Competition Commission at INDECOPI plays no provide policymakers with criteria for conducting the real role in revising new regulations on their impact competition assessment within RIA. on competition nor in advocacy for pro-competition regulation in general. Finally, a formal coordination mechanism between the CEBB and this Commission is lacking, despite the obvious links among the issues covered by both. The first steps to improve competition: dealing with discretion at the subnational level and assessing 52 competition more generally Enhancing INDECOPI’s power will be critical to more effectively remove barriers to market access and competition. Actions should include empowering INDECOPI to achieve the objective of effectively removing illegal and/or unreasonable bureaucratic barriers; granting INDECOPI the legal power to analyze bureaucratic barriers that affect competition; setting a legal limit on public entities for challenging INDECOPI’s declaration of an illegal and/ or unreasonable bureaucratic barrier before the judicial courts; and giving INDECOPI, as the competition authority, the power to issue ex ante opinions on the potential impact on competition of regulations (issued by the central government) in key economic sectors as well as on draft concession contracts and PPPs. Peru would also benefit from introducing a tailor- made and well-implemented regulatory impact assessment (RIA) system on competition. Barriers 46  For more details on the decentralization process and challenges, see Rendon et al. (2015), a background paper prepared for this report. 47  This is the number of times a bureaucratic barrier (the Ministry of Transport has kept suspended passenger interstate road transport on the national road network) has been declared illegal or unreasonable in 2013 and 2014 (up to August). INDECOPI (2014), p. 22. 6. Streamlining the labor market and building skills 6. Streamlining the labor market and building skills A well-functioning labor market is important for supporting growth and productivity. There is evidence that restrictive regulations impede the efficient allocation of labor to most productive firms. In Peru, these relate mostly to firing Wages kept up with productivity growth, supported by structural shifts Growth over the last decade has served workers well. In 2003–13, despite higher labor force growth 53 restrictions, as well as non-wage labor costs. Making the than most regional and upper–middle–income labor market more flexible and reducing excessive costs comparators (averaging 2.4 percent a year), job would help increase formality and overall productivity growth kept pace. Supported by rapid productivity of the economy. Reforms in this direction would also growth averaging 4.3 percent since 2013, Peru has improve the well-being of Peruvian workers, who would seen average real wage earnings increase by 3.7 be able to better benefit from their employment and percent a year. Over the period for which there is receive higher returns on their education. comparable data (2006–11), Peru’s real wage growth Fi ur 44. P ru’s work rs h v s n r lw rowth Fi ur 45. Productivit nd w dv nc s support d bov th LAC v r b shiftin sourc s of mplo m nt (real wage growth 2006–11, %) (contributions to real output per worker and real wage growth, percentage points, 2003–13) 16% 5 4 Sectoral Sectoral 12% shift in Spatial shift in shift in employment employment employment 3 8% Within Within Services Within 2 Services +Constrn urban +Constrn 4% 1 Within Within industry Within agriculture rural 0% 0 2006 2007 2008 2009 2010 2011 Real output Real wage Real wage per worker growth, growth, Peru wage growth LAC wage growth growth, by sector by sector by rural/urban Source: Authors’ estimates from WDI; Encuenta Nacional de Hogares (ENAHO), Source: Authors’ estimates from WDI; ENAHO, based on background paper based on background paper prepared for this report by Ruppert Bulmer et al. prepared for this report by Ruppert Bulmer et al. (2015). (2015). Note: Industry excludes construction. Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 46. L bor r ul tions r mor r strictiv Fi ur 47. L bor r ul tions lso fi ur mor in P ru th n in oth r countri s in busin ss compl ints (1=heavily impeded by regulations; 7=extremely exible) (top business complaints, 2010) 6 100% 5 80% 4 60% 3 40% 2 1 20% 0 0% Peru Peru Colombia Mexico Costa Rica Panama Bolivia Brazil Practices of informal sector Inadequately educated workforce Crime Labor regulations Other Source: Authors’ calculations: Background paper by Quijandria et al. (2015), based on The Global Competitiveness report 2014–15. Source: Authors’ estimates from Enterprise Surveys. of more than 6 percent a year far exceeded the LAC informal employment. Peru’s informal sector is large regional average of 2.1 percent a year (Figure 44). relative to comparators’, and productivity and wages among informal workers are a fraction of that among Productivity and wage growth has been aided by those having formal contracts. Transitions of workers large structural shifts in the sources of employment. from informal to formal employment are impeded Chief among these shifts has been the decline in on the supply side by still-large educational divides 54 agriculture and growth in the non-tradables sector. among workers, and on the demand side by costly Over 2003–13, about 600,000 jobs were shed in and restrictive labor market regulations, holding down agriculture (about 14 percent of agricultural jobs), overall employment growth. More generally, worker and in their place about 3.4 million net nonfarm jobs productivity and wage growth are also impeded by were created, the majority in the non-tradables sector quality of skills, the result of an educational system (services and construction).48 Large construction that has both underinvested and lacked appropriate projects (predominantly nonresidential) almost mechanisms to ensure that skills are pertinent to doubled construction jobs over the period, while the private sector needs. generalized rise in incomes stimulated the economy’s demand for services, with the largest job gains coming Labor demand is stifled by a costly, from commerce, government, and transport and restrictive labor code communications. In total, the agriculture sector’s share of employment fell from 35 percent to 25 percent, Labor regulations are more restrictive in Peru than in while that of services and construction increased from most of LAC. The ease of hiring and firing index ranks 56 percent to 64 percent. With the average wage in the Peru as 12th of 15 LAC countries on simplicity of hiring services and construction sector more than three times new or firing current employees. The only countries that that of agriculture, this shift has been strongly wage trail Peru are Brazil, Argentina, and Venezuela (Figure enhancing (Figure 45).49 46). Underlying this figure is a highly prohibitive system of restrictions to firing, which severely limits dismissals Nevertheless, earnings have been dampened by the for economic reasons and in which dismissal outside share of workers engaged in lower-productivity of economic reasons must be through negotiated compensation. Because these controls do not apply to 48  The industry of employment is determined by the worker’s declared “primary occupation.” small enterprises, they are disincentives for businesses 49  The average annual wage growth by sector during 2003- to grow beyond a certain threshold. In addition, at over 2013 was 8.2 percent in agriculture, 0.4 percent in industry, and 2.1 percent in services and construction. The average share of 35 percent, Peru’s tax wedge (the difference between employment has been 16 percent for agriculture, 13 percent for the labor costs to the employer and the take home pay industry excluding construction, and 71 percent for services and construction. of the employee) is among the highest in the region, 6. Streamlining the labor market and building skills Fi ur 48. Form l job cr tion l d b th priv t s ctor Fi ur 49. Th bulk of form l priv t s ctor job rowth c m und r fix d-t rm contr cts (employment, millions persons, 2004-13) (formal private sector job growth by contract type, thousands persons) 6 400 Fixed Open ended 350 5 300 250 4 200 150 3 100 2 50 0 Restaurants and hotels services Other services Mining Electricity Fishing Agriculture, hunting and forestry Transportation and communications Government Construction Manufacturing Commerce and water 1 -50 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Formal wage public Formal wage private Informal wage Own account Source: Authors’ estimates on ENAHO, based on background paper prepared for Source: Authors’ estimates from ENAHO panel 2007–11, based on background this report by Ruppert Bulmer et al. (2015). paper prepared for this report by Ruppert Bulmer et al. (2015). further reducing incentives to hire. More generally, labor Restrictive labor regulations regulations are among the most difficult obstacles for encourage the growth of the large businesses in Peru. While they feature in all competing informal sector, which is much less Latin American countries, they figure more prominently productive than the formal in Peru (Figure 47). A large portion of workers are in the informal sector, Job growth in the formal sector is shackled by where productivity is one-sixth that of the formal 55 restrictive dismissal regulations. While regulations sector. Restrictive labor code, especially where it for hiring are almost non-existent, restrictions to firing concerns firing restrictions and non-wage labor costs, are highly prohibitive. The labor code allows employers contributes to high levels of informality. Comparable to dismiss workers for economic or technical reasons estimates of informality outside of agriculture suggest but they require authorization from third parties that Peru has one of the highest levels of informality (Ministry of Labor or Judges), which effectively hamper in the region.51 Informality is generally high in Latin necessary adjustments of the labor force. Section 4 America but Peru’s level of informality outside of discussed the problem of stunted employment growth agriculture, at 70 percent, is above Mexico’s and at the firm level, which is linked to misallocation of labor Colombia’s (Figure 50). Both output per worker and and rigidities in the labor market—especially the firing wages are far lower in the informal sector. Estimates restrictions. Over 2004–13, the formal sector increased put the wage differential in 2007 at three to one, from around 1.6 million jobs to about 3.2 million jobs while the productivity differential for the same year is (from around 11 percent of overall employment to 20 estimated at six to one (Figure 51). percent). More jobs were created in the formal sector than the informal, and the vast majority of those formal jobs were created by the private sector (Figure operational definition for the “economic motives” that could 48). However, more than three-quarters of that job trigger collective dismissals (a definition previously missing in the growth came from fixed-term employment (Figure 49). Legislative Decree N 728). Another draft law (4008-2014, still to be approved by Congress) proposes a reduction in threshold for Fixed-term hiring has facilitated more efficient labor collective dismissals from 10 to 5 percent of employees. allocation, but cannot fully compensate for a rigid 51  From ILO 2012 (http://laborsta.ilo.org/applv8/data/ INFORMAL_ECONOMY/2012-06-Statistical%20update%20-%20 labor code. Thus, greater flexibility in hiring procedures v2.pdf). Informal employment is defined as the sum of persons will continue to be needed to allow businesses to whose main job was in the informal sector (non-registered businesses) or was informal (a job lacking basic social or legal better adapt their labor forces to the needs of their protection, whether in the informal sector or otherwise). Peruvian businesses. Recent reforms are a step in the right authorities have a similar estimate of informal employment: they calculate the share of informal employment at 77.2 percent in 2009 direction, although they do not resolve this problem.50 (74.3 percent in 2012), referring to the total number of informal jobs in the formal sector (without social contributions by employers), 50  The recent Presidential Decree 013-2014-EF provides an informal companies, or the household sector, see INEI (2014). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 50. P ru h s hi h l v l of inform lit Fi ur 51. Th work r productivit diff r nti l b tw n r l tiv to LAC form l nd inform l s ctors is 6 to 1 (share of non-agricultural employment, 2009) (average output, in N. Soles per worker, 2007) 80 Agriculture/ shing 70 Restaurants/hotels 60 W&R trade 50 Govt. services Transport/communications 40 Construction 30 Other services 20 Manufacturing 225 316 10 Mining Total economy 0 Peru Colombia Mexico 0 50 100 Informal employment: Informal sector Total Formal Informal Informal employment: Outside informal sector Source: ILO (2012). Source: INEI (2014). Fi ur 52. Tr nsitionin out of inform lit is difficult Fi ur 53. Educ tion improv s th odds of form l mplo m nt (probability of transitioning to alternate labor market (marginal improvement in probability of transitioning category or remaining in same category, 2007–12) to formal sector work, odds ratio) 100% Elder members Youth 80% Children HH size Post-secondary 60% Secondary Primary 56 40% Male Age Years in job 20% Self-employed 0% Farmer Disengaged Informal Formal -4% -2% 0% 2% 4% 6% 8% Original labor market state, 2007 Became/stayed disengaged Became/stayed informal Became/stayed formal Source: Authors’ estimates based on ENAHO: background paper by Ruppert Bulmer Source: Authors’ estimates ENAHO: background paper by Ruppert Bulmer et al. et al. (2015). (2015). Informal employees have little chance of moving post-secondary, which remains almost exclusive to to formal employment. Over 2007–12, less than 4 those working in the formal sector (Figure 53). But percent of informal workers made the shift to formal while education remains paramount for improving employment (Figure 52), versus 13 percent exiting the formal-sector employability of workers, it does not labor force and 3 percent becoming unemployed in the fully explain the limited transitions from informality, space of 12 months. These transition rates for informal nor the differences in productivity between informal workers to inactivity or unemployment are double the and informal workers. Other things being equal, a rates in Brazil or Mexico. worker with higher education has a probability only 6 percentage points higher of transiting into a formal Part of the difficulty with transitioning relates to job than a worker without education. Other factors sharp differences in educational achievement— such as age, gender, and the size of the household also but wage differences are significantly driven influence transitions into formal jobs, but their effects by the nature of informal work itself. Informal are even more muted. Moreover, regression analysis of workers average 4 years less of education than the wage gains from transition to formality suggests formal workers—9.2 years versus 13.4. One of the that a relatively large portion of the observed wage determinants important for transitioning from differential (and thus, potentially, the productivity informal to formal work is education, specifically differential) cannot be explained by differences in worker 6. Streamlining the labor market and building skills characteristics. Rather, it reflects a range of features Fi ur 54. P ru h s s n r l tiv r turns to duc tion of informal work that lower the productivity and incr s , unlik th r st of th r ion wages of those engaged there. For example, informal (relative returns to education, 1999-2011) sector workers who transitioned to the formal sector 2.6 over 2007–11 (and controlling for changes in sector/ 2.4 region of employment), experienced a 54 percent 2.2 increase in their wage, not associated with any change 2.0 in the worker’s underlying characteristics. That wage gap cannot be attributed to productivity differentials 1.8 related to the individual but to the nature of informal 1.6 work itself.52 Precarious transitions between formal 1.4 and informal employment highlight the importance of 1.2 making sure young people have access to the formal 1.0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 sector jobs. Otherwise, they might become caught by Return in LA completed 10 to 12 relative to 7-9 the trap of low productivity jobs, with little chance of Return in LA completed 7-9 relative to primary moving to high-productivity jobs in the formal sector. Return in Peru completed 10 to 11 relative to 7-9 However, the evidence described here also points to Return in Peru completed 7-9 relative to primary the limits of supply-side interventions in driving more Source: Authors’ calculations and background paper prepared for this report by Kudó and Székely (2015). rapid transitions from informal to formal jobs, and the importance of addressing demand-side constraints. Low educational quality inhibits productivity Advancing access to and quality of growth. Various sources paint a picture of low quality post-primary education is becoming of education relative to comparator countries.53 increasingly important A cross-country comparison of results from the Program for International Student Assessment (PISA) 57 While most countries in LAC have experienced suggests that Peru’s quality of education is deficient declining returns to education, returns to secondary across the board. Only a third of Peruvian 15 year olds and post-secondary skills in Peru have continued exhibit proficiency in reading comprehension, while to rise. There has been a substantial rise in the only a quarter demonstrate adequate mathematical educational attainment of the Peruvian labor force over skills (Figure 55). Peru’s government has moved to the past two decades. School attendance rates to age address many of the problems associated with basic 12 are above LAC regional averages, Peru has achieved education,54 including devoting greater resources to near- universal primary education, and secondary education (Figure 56). Despite this, a comparison of and tertiary education completion rates also surpass Peru with other LAC countries and other regions shows LAC comparators. Despite this increase in supply, the that the level of both public and private investment in demand for secondary and tertiary educated workers higher education was below the level corresponding continues to outpace supply. As shown in Figure to Peru’s income, considered in aggregate and in per 54, relative returns to education in Peru have been capita terms.55 increasing, unlike in the rest of the LAC region. Although the good news is that the demand for higher skills Technical training programs do not seem to train remains strong, the flip side is that higher skills may workers for private-sector needs. Many public constrain stronger growth and employment creation vocational training programs, which account for 30 (particularly in the services sector). Ensuring that percent of continuing training, have underinvested in businesses can continue to acquire needed technical skills to expand will require both continued advances in 53  Castro and Yamada (2012). post-primary education and increased quality. 54  Reforms center on increasing public resources for education, making greater use of public–private partnerships for improving education infrastructure, instituting meritocratic career 52  The increase in wage earnings associated with transition to advancement for teachers based on continuous evaluation, and formality in 2007–11, controlling for worker characteristics, varied revamping regulations for better management of the system (see from 22 percent in utilities to 168 percent in mining, and averaged the background paper prepared for this report by Yamada et al., 52 percent economy-wide (according to authors’ estimates on (2015)). ENAHO 2007–11 as reflected in the background paper prepared 55  From background paper prepared for this report by Botero for this report by Ruppert Bulmer et al., 2015). (2015), citing World Bank (2007). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 55. P ru’s PISA r sults point to low chi v m nt Fi ur 56. P ru’s public sp ndin on duc tion is low cross th bo rd (PISA math score, 2012) (relationship between public spending on education and PISA math scores in 2012) 600 600 550 500 500 450 Peru 400 400 2012 Peru 350 Peru 2013 300 300 8 9 10 11 12 0 5000 10000 15000 Ln (GDP per capita, 2012 PPP) Government expenditure per primary and secondary student (in PPPUS$, 2011) Source: PISA, 2012. Source: Authors’ estimates from ENAHO panel 2007–11. equipment, and they suffer from problems of quality providers can compete on the basis of appropriate and relevance to labor-market needs. Much of the contracting and payment systems; removing the vocational training for mining workers, for example, constraints that affect investment in training by firms; does not meet international industry standards. As a and diversifying sources of funding.57 result of the poor coordination between educational training and private-sector requirements, the average On the demand side, an appropriate, functioning internal rate of return to professional degrees in public balance needs to be struck between worker 58 and private technical institutes are low. Firm-level protection and job creation. The fact that most training has a better record for addressing specific workers are employed in the informal sector—outside firm needs, but the incidence is low, in part because labor regulations and unable to enjoy the benefits— firms find few incentives for investing in training that suggests that overburdening labor policies are in fact can be carried away to new jobs.56 reducing protection to workers by reducing coverage of formal policies, contrary to the intended aim. It Removing distortions will help workers seems appropriate to reform the regime to offer a acquire more relevant skills and consensually defined level of security to employees allocate them to the most productive while making the regime attractive enough to include jobs with the highest earning potential the majority of workers. On the supply side, improving access to quality basic This increased flexibility and reduced costs of labor and vocational education is important for reducing regulations should be accompanied by redundancy the obstacles to transitioning to higher-wage work, arrangements. In terms of adding flexibility and easing whether in the formal sector or even within the job mobility, hiring and firing regulations could become informal sector. This involves higher public spending less costly and require less administrative discretion on education and training, and greater monitoring in firing approvals, including collective and individual and evaluation of all education programs, including dismissals. Making temporary contracts more preparing students for higher education exams, and mainstream and incorporating benefits would also better matching technical training supply to private help. Finally, reducing the non-wage costs of labor— sector needs. Based on experience in other countries, which are very high by international standards—would in addition to policies that provide an adequate system also encourage formal employment. of labor regulations, active policies should include creating a contestable market for training and active labor-market programs where public and private 57  See the background papers prepared for this report by 56  Background paper prepared for this report by Yamada et Ruppert Bulmer et al. (2015) and Yamada et al. (2015) on al. (2015). continuous training. 7. Improving access to credit for enterprises 7. Improving access to credit for enterprises An efficient financial sector is central is allocating and intermediating resources between to productivity growth savers and borrowers: since only intermediated E savings can be used as credit, the extent to which fficiently allocated credit allows firms to savers use the financial system is important. As pursue economical investments. Capital illustrated in Figure 60, domestic savings in the 59 investment is not only a direct factor financial system are lower that would have been of production, but it also increases total factor predicted by the level of savings alone. productivity (Figure 57). Higher returns to capital that are associated with increases in total factor Use of the financial system by savers is low. This is productivity, could also reward higher investment.58 An reflected in lower deposits than expected for the savings efficient, deep financial sector intermediates between rate (Figure 60): only 29 percent of the population have savers and borrowers to mobilize and pool savings and savings accounts, lower than in Bolivia (42 percent), channel them to productive borrowers.59 It also helps Colombia (39 percent), Ecuador (46 percent), and Chile to reduce the costs and risks associated with savings (63 percent).60 This was not due to lack of information, and investments. It fosters capital accumulation, as 72 percent of respondents to a recent survey had contributes to better resource allocation in the knowledge about savings accounts.61 More than half economy, realizes economies of scale, and overcomes the people surveyed did not use or own any financial investment indivisibilities. product at all, while only 22 percent kept savings in the financial system (Figure 61). But Peru’s credit to the private sector seems low given the high savings rate and income level. Peru Access to credit varies by firm size and has a relatively high savings rate, comparable to that geographically of other successful MICs during their high growth periods (as discussed in Section 1) and relative to its Although formal businesses do not report access to income (Figure 58). But the volume of credit to the finance as a major constraint, cost of credit is high private sector is lower than predicted by the country’s for SMEs. Peru ranks 12th (out of 188 countries) in level of development (Figure 59). This suggests that 60  Global Findex Database: http://datatopics.worldbank.org/ there may be inefficiencies in how the financial sector financialinclusion/. 61  Survey measuring financial capacities in the Andean countries. Corporación Andina de Fomento Comisión de 58  Caselli (2005). Eliminación (CAF) (2015). Encuesta de medición de las capacidades 59  Acemoglu and Zilibotti (1997). financieras en los Países Andinos. Informe comparativo 2014. Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 57. C pit l nd productivit show positiv Fi ur 58. P ru’s dom stic s vin s r hi h for its r l tionship incom l v l TFP (proportion, US TFP= 1) (gross domestic savings,% of GDP, 2013, 130 countries) 1.4 50 1.2 40 1.0 30 0.8 Peru 20 0.6 Peru 10 0.4 0.2 0 0 10000 20000 30000 40000 50000 60000 0.0 -10 0 100000 200000 300000 400000 Capital per worker (2005 US$) GDP per capita (current USD) Source: PWT8.2. Source: WDI–Global Financial Development Data (GFDD), World Bank. Fi ur 59. Low cr dit to priv t s ctor, Fi ur 60. Low d posits in th fin nci l s st m, comp r d to incom p r c pit comp r d to s vin s (dom. credit to private sector, % of GDP, 2013, 173 countries) ( nancial system deposits, % of GDP, average 2007–11, 173 countries) 250 160 140 200 120 150 100 60 100 80 60 50 40 0 Peru 0 10000 20000 30000 40000 50000 60000 20 Peru -50 0 0 20 40 60 80 GNI per capita PPP (international currency) Domestic savings (% of GDP) Source: WDI. Source: WDI and GFDD. the Doing Business indicator of ease of getting credit; the financial statistics may underestimate issues of the percentage of firms with a bank loan or credit access and the cost of credit. Although measurement line is higher in Peru than in comparator countries is difficult, there is evidence that effective borrowing (60 percent versus the LAC average of 45 percent); rates are high for informal firms outside of the formal only a fraction of firms report access to finance as financial system.63 For the financial system as a a major constraint (8.5 percent, against a regional whole, despite falling steeply from the early 1990s, and global average of around 30 percent). However, the real lending rate was 16 percent in 2013—high the cost of credit is thought to be high, especially for compared with other economies at similar levels small and medium firms. For those micro firms who of development (Figure 62).64 The fact that large had access to the banking sector, the annual cost of corporates can obtain credit at low competitive credit was above 30 percent of the loan amount in 2014 and for those without access it could be much 63  A survey by the association of banks (ASBANC) found that the annual effective rate in such loans could be as high as 400 higher.62 Moreover, given the high rate of informality, percent. Informal loans also can involve pledges, whose value can go from 1.3 to 5 times the value of the loan. This segment is not small, as informal activity accounts for about 20 percent of GDP 62  Choy et al (2015) measure the average cost of credit by and units of production categorized as informal were 87 percent estimating a rate that includes all the payments associated to of all units of production in 2012, according to INEI (2014). loans obtained by firms, mainly interest-rate payments and other 64  Corresponds to the average lending interest rate, adjusted commissions, as a percentage of the loan amount. by the GDP deflator, as reported by WDI Data, from World Bank. 7. Improving access to credit for enterprises Fi ur 61. Most p opl do not us Fi ur 62. Hi h r l int r st r t s, th fin nci l s st m for s vin s comp r d to incom p r c pit (saving instrument used during last year, % of surveyed individuals) (lending rate, percent, 2013, 106 countries) 25 SAVED IN THE FINANCIAL SYSTEM 22 By a saving account 16 20 By other nancial investment 6 15 Peru SAVED OUTSIDE THE FINANCIAL SYSTEM 42 10 Saved at home 26 5 By collective saving modality (juntas) 9 By other modality (buy properties 7 0 0 10000 20000 30000 40000 50000 60000 70000 80000 or livestock, family loan) -5 NO SAVINGS 45 -10 0 5 10 15 20 25 30 35 40 45 GNI per capita, PPP Source: Survey measuring nancial capacities in the Andean countries, CAF (2014). Source: WDI; interest rates are based on IMF IFS. Note: 1,210 individuals were surveyed in 2013. Fi ur 63. Acc ss to cr dit is hi h r in Lim Fi ur 64. Sm ll firms borrow l ss from th fin nci l s st m ( rms with a bank loan in 2010, %) ( rms with a bank loan or credit line in 2010, %) 100 100 85.3 80 80 59.4 60 55.6 60 46.5 61 40 40 20 20 0 0 Lima Trujillo Chiclayo Arequipa Small Medium Large Source: Enterprise Surveys, World Bank. Source: Enterprise Surveys, World Bank. rates, also indicates that smaller and medium firms’ credit line in 2010, compared to around 90 percent costs of borrowing must be high to lift the average to for medium and large firms (Figure 64). Large that extent. corporations and firms obtain finances mainly with banks, but small firms and micro-enterprises face Access to credit also varies geographically. different circumstances (Figure 65). When they Lima accounts for 44 percent of GDP, but receives have access, there are big differences in the cost of 72 percent of the credit. In Lima, 85 percent of credit faced by firms of different sizes (Figure 66). firms have a bank loan, compared to 46 percent in The annual cost of credit for micro enterprises is 33 Arequipa (Figure 63). This difference is partly related percent, when they use the formal financial system. to variations in capital intensity, productivity, and the Among SMEs surveyed that obtained credit, almost size of firms. However, differences in access to credit 20 percent were not satisfied with the terms. might also have played a role in the productivity gap between Lima and other geographical areas, as Peru has a low-volume, high-cost financial discussed in Section 3. system. Figure 67 shows a positive correlation between income per capita and the depth of the Small firms have both higher costs and less access financial sector, showing Peru in the high-cost and to credit. Only half of small firms had a bank loan or low-volume situation compared to other countries. Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 65. Sourc of cr dit b t p of firm Fi ur 66. Cost of cr dit nd v r outst ndin lo ns in dom stic curr nc (% of total credit for nancing working capital in 2010) (by type of rm, %; and US$, rst half of 2014) 100 35 40 30 35 80 25 30 60 25 20 Millions 32.9 20 40 15 15 10 19.7 10 20 5 9.6 5 6.3 4.6 0 0 0 Small (5-19) Medium (20-99) Large (100+) Micro- rms Small rms Med. rms Large rms Corp. Banks Supplier credit Other nancing sources Interest rate, % Av. Loan (USD, RHS) Source: Enterprise Surveys, World Bank. Source: Choy et al. (2015), Central Bank of Peru. Fi ur 67. Hi h int r st-r t spr d, Fi ur 68. Hi h b nkin conc ntr tion nd profit bilit iv n cr dit to priv t s ctor (Banking spread, %, 2013, 117 countries) (ROE , %, 2011) 20 30 18 Peru Peru 25 16 14 20 COL 12 ECU MLY 62 10 15 BRZ 8 CHL KOR 6 10 THA SPA 4 UK 5 MEX 2 USA JAP FRA ITA GER 0 0 0 50 100 150 200 250 40 50 60 70 80 90 100 Domestic credit to private sector (% of GDP) 5-bank asset concentration Source: WDI. Source: Global Financial Development Data, World Bank. Note: ROE (return on equity) is calculated as the average for 2007–11. 5-bank asset concentration shows the assets of the ve largest banks as a share of total assets in commercial banking. Figure 68 shows that the interest rate spread is high Deepening the financial sector and in absolute and relative terms—the third highest in equalizing access would be the key the sample and the highest for countries with similar directions of reform credit depth (30–50 percent of GDP). On the demand side, the operating and administrative cost of small- Macro-financial conditions, inertia, and financial scale loans and the risk of operating with limited industry competition may be some of the supply information in an environment of informality raise factors explaining the low financial depth in Peru. the costs of financing. On the supply side there may Dollarization remains substantial in spite of low be insufficient competition (Figure 68). inflation and robust fiscal and external positions, suggesting that the inertia of past financial instability continues to influence the level of development and sophistication of the financial system. Banking concentration and return on equity are high in Peru relative to elsewhere, although competition has 7. Improving access to credit for enterprises increased in recent years. The number of banks increased from 11 in 2006 to 17 today, and return on equity decreased from 30 percent in the late 2000s to just above 20 percent now. Also, indexes of competitive behavior have increased and financial margins have fallen.65 Moreover, micro-finance and other non-bank financial institutions (NBFI) providing banking services have developed well in Peru, a fact not measured by banking concentration. The still-high interest rate spread, however, suggests that there is still room to increase competition. Horizontal policies that promote formality and growth of firms would naturally increase access by enterprises as well as improve competition in the banking system. More efficient financial intermediation will lower the wedge between lending and deposit rates and expand access. These horizontal reforms would eventually help equalize access to credit for small firms and increase productivity in the long term. However, more targeted interventions might be needed in the short to medium term to improve access for small firms. Given the complexity of the issues involved, further research and analysis of the Peruvian financial sector is needed, and it is beyond the scope of this 63 report. Deeper analysis would help to shed further light on these issues and arrive at more detailed policy recommendations and conclusions. 65  Cespedes and Orrego (2014) estimated the Panzar and Rose index for competition in the banking industry and found that it has been increasing since 2008. The Panzar and Rose index is measured as the sum of elasticities of financial to factor prices (expenditure on interest for loanable funds, personnel wages, and capital depreciation expenditure). Moron et al. (2010) estimated the residual elasticity of demand for seven financial products and found that in five of them it increased since 2006. Residual demand is the demand that a bank faces after the rest of the banks have served the market. 65 Part I V Other opportunities to reduce misallocation and further boost productivity growth Peru Building on Success: Boosting Productivity for Faster Growth 66 8. Using spillovers from international trade 8. Using spillovers from international trade W hat factors can both help reduce misallocation and offer an additional impetus to Peru’s already strong productivity growth within firms? As in other countries, Peru’s trade has a statistically significant and enhanced management capabilities. Most importantly, trade exposes firms to foreign market competition, thus inducing more efficient allocation of resources domestically. This can happen directly in the firms that export and import, or indirectly 67 relationship to within-firm productivity growth. through the pressure exerted on suppliers via supply Imports of inputs for production are often an chains. All of these features of trading internationally instrument to introduce new technology. But export can be viewed as spillovers. markets also provide strong competition pressures, as shown by progress in reducing misallocation in the Peru has one of the world’s most Peruvian manufacturing sector. At the same time, liberal trade regimes, but trades very innovation is a key driver for productivity globally, little and in a small number of goods but firms do not innovate as much as they should in Peru. Moreover, there are differences in access to Peru has one of the world’s most liberal trade policy innovation inputs between firms of different sizes regimes. This resulted from deep and comprehensive and ages. Because small and young firms are more reforms that began in 1990 as it sharply cut tariffs, productive in Peru—unlike in other countries—this eliminated most non-tariff barriers, and liberalized feeds the misallocation of factors. The next two services markets (Figure 69). In 2000 it had an average sections analyze the innovation environment and Most Favored Nation (MFN) tariff66 of 12.8 percent, international trade and discuss policy opportunities compared with Brazil’s 12.7 percent and Malaysia’s 4.5 in these areas to spur productivity. percent. By 2013, Peru had lowered its average tariff to 1.9 percent (against Brazil’s 10.1 percent and Malaysia’s International trade provides an impetus to 3.6 percent). Only around 15 percent of products are increase productivity. International trade introduces subject to nontariff measures (NTMs) versus around competition by integrating countries within larger 50 percent in Brazil and Chile, and nearly 80 percent world markets while offering the benefits of greater in China. Furthermore, Peru was one of the founders of specialization and economies of scale. Over the the Pacific Alliance, a Latin American trade bloc with long term, it boosts domestic productivity through features of further integration. learning by exporting and importing. It spurs rapid adoption of cutting-edge technology, modern work 66  Weighted average (by trade volume) most-favored-nation processes, best-practice production standards, tariff. Peru Building on Success: Boosting Productivity for Faster Growth Yet it trades much less than countries at similar from 747 products and 102 markets to 1,328 products income. Peru leveraged the boom in global commodity and 118 markets. This apparent success is, however, markets for rapid growth in the 2000s, but did not modest when compared that of other countries in the take advantage of the productivity-enhancing region (Figure 71). Peruvian exports consist mainly of potential of deeper trade integration. And despite minerals and metals (copper, gold, zinc, silver) and its large minerals trade, its exports are low relative their products. A far smaller share consists of some to other countries with similar income levels (Figure vegetables and food products (fishmeal, asparagus, 70), at 22.4 percent of GDP in 2014. Peru also has avocados, grapes, and coffee); petroleum and low services trade and is poorly integrated into global derivatives; and apparel and textiles. Exports of goods value chains (GVCs). grew rapidly in the last decade, by an average of 17 percent annually in nominal US$ terms, but minerals Peru’s exports basket is not diversified and and metals significantly increased their shares in dominated by commodities. Export diversification can exports: from 54 percent in 1994 to 72 percent in 2013 be measured by the number of products the country (Figure 73). Overall, five sectors (minerals, metals, exports and the number of markets the country vegetables, foods, and textiles and apparel) accounted reaches. Between 2000 and 20013, the number of for 91 percent of exports in 2013, virtually unchanged goods and destinations in Peru’s exports basket grew from the 92 percent share in 1994. Fi ur 69. P ru h s v r lib r li d tr d r im … Fi ur 70. … but tr d s r l tiv l littl (simple average tari rates) (trade as share of GDP, 2013) 20 200 160 MYS 15 THA 68 120 10 80 ZAF CHL 5 PER 40 BRA 0 0 Thailand South Colombia Chile Malaysia Peru 6 8 10 12 Africa 2000 2013 ln GDP per capita (PPP) Source: World Integrated Trade Solution (WITS). Source: WITS and WDI. Fi ur 71. P ru’s xport div rsific tion succ ss f d s in comp rison to th t of th r ion (number of products) 2000 2013 5000 5000 4000 4000 ZAF ZAF THA THA MYS BRA 3000 3000 MYS BRA 2000 2000 COL ARG COL CHL CHL PER 1000 URY 1000 URY PER 0 0 0 50 100 150 0 50 100 150 Number of countries Number of countries Source: Authors’ calculations based on UN Comtrade via WITS: background baper prepared for this report by Farole and Gutierrez (2015). 8. Using spillovers from international trade Fi ur 72. Export rowth com s m inl Fi ur 73. Exports r incr sin l conc ntr t d from xp ndin xistin m rk ts (decomposition of export growth by product and market, 2007–13) (% of total exports, total export as % of GDP) Increase of existing products 100 40 in established markets 133.55 Decrease in existing products -36.02 in established markets 75 30 Extinction of exports of products -2.52 in established markets Introduction of new 0 50 20 products in new markets Introduction of new products 0 in established markets Introduction of existing 0.14 25 10 products in new markets Product diversi cation 4.84 in established markets 0 0 -40 20 80 140 1994 2003 2013 Minerals & Metals Vegetable Foodstu s Textiles & Apparel Rest Total Exports as % GDP (RHS) Source: Authors’ calculations based on WITS. Source: WDI. Fi ur 74. Tr dition l xport rowth Fi ur 75. Nontr dition l xport rowth w s driv n m inl b pric s w s driv n b qu ntit (decomposition of export growth by quantity and price, 2004–11) (decomposition of export growth by quantity and price, 2004–11) 2004 2004 2005 2005 2006 2006 2007 2007 69 2008 2008 2009 2009 2010 2010 2011 2011 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 Price Quantity Price Quantity Source: INEI, CEPAL. Source: Authors’ calculations based on Latin America and the Caribbean Region World Bank poverty database (LACPOV) and WDI. Export of new products to new markets is negligible. Peru has seen a commendable growth in its Between 2007 and 2013, almost all of export growth nontraditional exports, but has not significantly in Peru occurred at the intensive margin (95.1 changed its export profile. There has been very limited percent of export growth is explained by existing shift into nontraditional or upgraded products. Peru’s trade relationships). Furthermore, 133.6 percent of exports are mostly in raw or semi-processed form export growth is explained by increases of exports rather than products with significant value added. of old products in old markets, which was offset by Only around 5 percent of copper and zinc exports are a reduction of 36 percent in old products in existing value-added products. Growth of commodity exports markets as well as a reduction of 2.5 percent due to has been driven mainly by higher prices than volume extinction of exports of existing exports in existing (Figure 74). For nontraditional products, more than 90 markets. The extensive margin overall contributed percent of export growth between 2004 and 2011 was 4.9 percent to export growth in Peru in 2007–2013. due to quantity expansion (Figure 75). Nontraditional Product diversification in old markets explain 4.8 exports grew strongly (albeit from a low base), percent of total export growth, increases of old especially textiles and agriculture. Nontraditional products in new markets 0.14 percent, and both new agricultural exports grew by more than 18 percent a products in old markets and new products in new year (see Box 4), and some firms became international markets zero (Figure 72). “export superstars.” Growth of textile exports was Peru Building on Success: Boosting Productivity for Faster Growth Box 5 GVCs: an overview A GVC is a group of firms in multiple countries There is untapped potential to expand that operate at sequential stages of a nontraditional exports and reap productivity production process. One firm organizes and gains by integrating into GVCs. This is manages the GVC with the goals of minimizing especially true for integration in the middle costs and increasing productivity. GVC- of the chains, where value added is higher based globalization is driven by firms’ global than from providing natural resource inputs strategies rather than by traditional country- downstream. Falling transport costs, greater based comparative advantages. Developing global openness, and cooperative trade policies countries join GVCs to become competitive, have given rise to GVCs. Each step is carried and they industrialize by deepening their out where skills and materials are available at participation. (The Box Figure illustrates the competitive cost and quality. New empirical channels through which GVCs benefit an evidence suggests that being in the middle economy and raise productivity.) Government of GVCs plays a role in the way international policies on trade, investment, employment, trade fosters economic growth. GVCs also and infrastructure need to be formulated to include trade in services because they add take advantage of emerging opportunities value and are necessary to link activities to join GVCs. For Peru, good policies would across countries. By integrating into new maximize the value added generated by its GVCs, Peru could diversify its exports and find exports and increase it over time. The level of nontraditional sources of export growth. GVCs that value added is tied to the breadth, variety, link to dynamic, leading global firms to expose and sophistication of tasks and activities that local firms to sources of productivity growth can be located in Peru. through leading-edge technology, business 70 processes, and demanding standards and Traditional trade involves goods made almost technical regulations. entirely in one country and sold in another, but the essence of GVC trade is “importing Box Fi ur . GVCs b n fit conomi s nd r is to export.” GVC measures domestic value productivit throu h multipl ch nn ls added by subtracting the value of imported inputs from the value of exports. One country Backward / forward • Demand effect linkages • Assistance effect (for example, Peru) exports parts that are incorporated in the exports of another country • Diffusion effect (for example, China). Flows of intermediate • Availability and Impact of participation at home quality effect Technology goods provide two measures of supply chain spillovers integration and a country’s role in GVCs. On the • Demostration effect sales side, it indicates that exporters are selling into GVCs—forward integration or indirect Market • Pro-competition effect restructuring • Demostration effect value added. On the sourcing side, it indicates that a country is buying from a GVC. Patterns • Amplification of Minimum scale on the buying side—backward integration achievements pro-competitioneffect or foreign value added—provide information • Sustainability effect on the source of technology transfer and the Labor • Demand effect types of GVCs a country is likely to join. markets • Training effect • Labor turnover effect Source: Taglioni and Winkler (2014). 8. Using spillovers from international trade Fi ur 76. P ru h s limit d links to GVCs, s r fl ct d b th low Fi ur 77. P ru’s xports r mostl low-t ch int r t d v lu dd d nd hi h sh r of downstr m inputs (upstream and downstream intermediate (sectoral contribution to foreign value added in exports, %) value added, % of gross exports, 2011) 0.7 100 0.6 80 0.5 0.4 60 0.3 40 0.2 20 0.1 0.0 0 Colombia Peru Mexico Chile China Thailand South Malaysia Peru Chile Argentina Thailand South Africa Africa High tech manufacturing Low tech manufacturing Upstream Downstream Services Primary Other Source: World Bank calculations based on data from Eora dataset. Source: World Bank calculations based on data from Eora; sectoral technology classi cations based on OECD classi cation of technology content. Fi ur 78. W k forw rd link s of s rvic s 50 (share of gross, direct, and forward linkages of the services sector, 2011, %) 40 30 71 20 10 0 Gr Di Tot Gr Di Tot Gr Di Tot Gr Di Tot Gr Di Tot Gr Di Tot Peru Thailand Chile Colombia S Africa China Transport Finance Communications Water & utilities Construction Distribution and trade Insurance Other business Other commercial Source: World Bank Trade in Value Added Database. Note: Total value added is split up into direct and forward linkages, i.e. the di erence between gross value added and direct value added illustrates forward linkages. primarily to the U.S. market, with significant new performance in information and communications firm entry and product experimentation, supported technology. Services exports are still concentrated in by trade agreements and liberalization. The success traditional services such as transport and travel. of the non-traditional exports is often linked to the participation in the global value chains (GVCs) as Peru’s low participation in GVCs described in the following section. reflects its behind-the-border structure, especially low intermediate Peru has the opportunity to become a significant inputs from the services sector exporter of modern services, but it would need to exploit it. Technological advances are making Peru’s participation in GVC is limited, but there services tradable, such as communications, financial are opportunities to develop it. (Box 5 presents an and business services, and call centers. Services overview of GVCs.) Peru’s role in GVCs is mainly to now constitute around 20 percent of world trade, supply primary inputs via downstream linkages (see but just 5.2 percent of Peru’s exports, a low rate just below), although notable successes have also given the country’s human capital endowment and taken place in the non-tradable sectors (Box 6). Peru Peru Building on Success: Boosting Productivity for Faster Growth Box 6 Successes and opportunities of GVCs for upgrading and productivity spillovers There have been some notable success stories capabilities, and opening global markets. in trade integration. In agribusiness, Peru It exports grapes to over 70 countries, has earned a strong position in global retail including sophisticated markets in Europe, chains: mainly fruits and vegetables and and has generated positive spillovers for more recently quinoa. Horticultural exports smaller producers to export and for the future expanded from asparagus to paprika, avocado export of other crops. There was extensive and citrus, then grapes and, most recently, knowledge transfer to Peru, especially from blueberries (Box Figure). In apparel —Peru’s Chile. Developing backward linkages to input largest manufacturing export sector— the providers and forward linkages into grape juice number of exporters in knitted apparel (HS 61) or raisins would strengthen the sector further. increased from around 200 during 1997–2001 to over 350 during 2002–2006; the number In mining equipment, the success of early of products exported jumped from 50 during firms was largely the result of technology 1997–2001 to almost 70 during 2002–2006. transfers from foreign firms. New products However, “export superstars” did not emerge have been developed, and Peruvian firms have on the same scale as in agriculture, and the built capabilities in design and development sector has struggled since the crisis peak. by working with foreign firms in the domestic Plastics comprises a promising complex market and in Chile. There has been of manufacturing sectors, which is among upgrading with a greater focus on quality Peru’s fastest-growing exports. Interestingly, and safety standards. Opportunities exist to virtually all of Peru’s “export superstars” in strengthen connections between firms and these sectors make significant use of imports, other stakeholders to generate synergies or 72 reinforcing the importance of external spillovers for the overall economy. So far, most spillovers in productivity growth. spillovers have come from labor rotation. Peru’s successes in nontraditional exports point to future opportunities. High-quality Box Fi ur . Som not bl xport succ ss s w r h lp d b ntr into GVCs cotton apparel, for example, has had high value capture within Peru, from agriculture (market share of U.S. imports from Peru in total U.S. imports for that product, %) to finished apparel exports, and it upgraded 100 the unit value of products by 50 percent 80 between 1998 and 2014. It has diversified its export markets within the region and 60 generated demand for cotton inputs from small producers. The sector has opportunities 40 to upgrade further as a fast-fashion supplier 20 in high value brands or by design and branding activities. 0 -20 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Table grapes are an exemplary sector that has had the most success recently. Their Asparagus, fresh or chilled Asparagus, preserved Avocados, fresh or dried Bananas, including plantains, fresh success builds on earlier agricultural exports Fresh grapes Guavas, mangoes and mangosteens such as avocados and asparagus, diversifying Mandarins, clementines, wilkings horticulture, deepening its marketing Source: Authors’ calculations on Superintendencia Nacional de Administración Tributaria (SUNAT) data. Source: Authors and Background papers prepared for this report by Pierola et al. (2015), Fernandez-Stark et al. (2015a), Fernandez-Stark et al. (2015b), Fernandez-Stark, et al. (2015c). 8. Using spillovers from international trade Fi ur 79. R turns on importin r si nific nt… Fi ur 80. … t most xport rs do not import (average percentage di erence between (total number of exporters) exporter-importers and exporters-only) 80% 1.6 70% 1.4 Average percentage di erence 60% 1.2 50% 1.0 Thousands 40% 0.8 30% 0.6 20% 0.4 10% 0.2 0% 0.0 Number of Export value Average relative Export growth 2000 2006 2012 2000 2006 2012 2000 2006 2012 destinations unit values Agri Business Plastics Apparel of export Exporter-Importers Exporters only Source: Authors’ calculations based on customs data: background paper prepared Source: Authors’ calculations based on customs data: background paper prepared for this report by Pierola et al. (2015). for this report by Pierola et al. (2015). participates little in sectors typically associated with among Peru’s fastest-growing exports. Interestingly, dynamic GVCs like motor vehicles, electronics, and virtually all of Peru’s “export superstars” in these services offshoring. Critically, Peru has extremely sectors make significant use of imports, reinforcing limited upstream linkages in GVCs, that is, it makes the importance of external spillovers in productivity very little use of imported inputs (and their embedded growth. These are discussed in more detail in Box 6. technology) in its exports (Figure 76). And just 2 percent of Peru’s GVC imports are embedded with high The services sector is lowering competitiveness 73 technology—far lower than, for example, Thailand’s through weak forward linkages to exporters. The (50 percent) and Argentina’s (30 percent, Figure 77). quality, cost, and reliability of services are important for competitiveness. Services are critical in GVCs and • Downstream linkages. Peru’s share is one of the can be value-added, exportable elements embedded highest among comparator countries. Most Peruvian within the chains. They can also be major indirect exports are commodities (metals/minerals) that are exports, accounting for a share of the value added easily incorporated into other countries’ exports, in goods-producing sectors. In Peru, however, the and almost a third of domestic value added that is forward linkages of services exports are low. Services’ exported ends up in other countries’ exports. value added in total exports is 19.6 percent, and only • Upstream linkages. The share of foreign value 8.4 percent of manufactured goods exports (see also added in Peruvian exports is low versus similar Section 4). In most comparator countries, the ratio countries and is only one-third of the countries like of services in value added of exports reaches 30–40 Mexico or Malaysia that are highly integrated into percent and about 20 percent of manufactured GVCs. This is because most Peruvian exports are exports (Figure 78). On value added, business services natural resource–based and most of the value added and finance are the services most linked to goods is from extracting and processing minerals and other exports; traditional services, like distribution, trade, natural resources with scant foreign inputs. and transport, have limited forward linkages. Peru’s past and recent success in boosting non- Peru’s use of imported inputs is low, even though traditional exports points to future opportunities the firms that use them have measurably better to join global value chains. In agribusiness, Peru outcomes. Notwithstanding some notable successes has earned a strong position in global retail chains. in integration (see Box 6), low imports are a reflection In apparel—Peru’s largest manufacturing export of low participation in GVCs and thus an indicator sector—the number of exporters and products of poor use of spillovers. Importing to export affects increased dramatically over the last 15 years. Plastics multiple dimensions of firms’ performance. Exporters is a promising complex-manufacturing sector, which is that import mainly from high-income countries (this Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 81. P ru r nks low on lo istics p rform nc … Fi ur 82. … sp ci ll customs nd infr structur (LPI rank, higher rank means lower performance) (LPI rank, higher rank means lower performance) 100 120 80 100 80 60 LPI rank 60 40 40 20 20 0 0 2007 2010 2012 2014 2007 2014 2007 2014 Customs Infrastructure Peru China Colombia Colombia Peru Chile Malaysia Chile South Africa Thailand Thailand South Africa China Malaysia Source: Logistics Performance Index (LPI, World Bank). Source: LPI (World Bank). can be indicative of high-technology imports) show as well as Brazil’s (26 percent) and Argentina’s (27 stronger productivity, export value, diversification, percent). The OECD average is only 9 percent of and export quality. And any exporter that also imports product value. Peru ranked 71st in the world in 2014, enjoys better export performance—more diverse having fallen from 59th in 2007.68 This is substantially destinations and higher export value, export quality, worse than comparator countries such as Malaysia and growth (Figure 79). Despite the benefits from (25th), Thailand (35th), and Chile (42nd; Figure 81). 74 importing, in 2012, 60 percent of Peru’s agricultural Customs and infrastructure are areas where Peru’s exporters and 80 percent of apparel exporters did not trade facilitation and logistics ranking has declined import (Figure 80).67 While all “exports superstars” markedly, the former from 49th to 96th and the were also strong importers, this may suggest that latter from 57th to 67th (Figure 82). Bureaucratic small firms face heavy barriers to importing. and infrastructure obstacles weigh on some of Peru’s more dynamic exporters (Box 7). Infrastructure for trade, including logistics—and other behind-the-border The narrow contribution of services to exports issues—remain a constraint may reflect shortcomings in domestic competition and regulatory regimes. As discussed in Section 4, Improving infrastructure for trade and logistics, domestic constraints affecting firms in the services and dealing with other behind-the-border issues sector misallocate factors of production and introduce present a significant opportunity to boost trade. significant distortions in the intermediate goods Peru’s logistics costs—about 32 percent of product markets, which are hurting exporters in all sectors of value—are among the highest in Latin America, well the economy. Peru’s services value added and services above Colombia’s (23 percent) and Chile’s (18 percent), exports are low despite its relatively low services trade barriers (Figure 83). This suggests that other 67  The data used for this analysis allow the identification of supply-side constraints such as skills availability and inputs and other goods imported directly by exporters. There may be other inputs and goods imported indirectly through third electronic and physical infrastructure play a bigger parties (e.g., distributors and traders); however, these transactions role in services trade performance. cannot be identified with the data available for the analysis. Importing indirectly may be an efficient strategy for small firms and for the import of non-core inputs, as it can reduce the fixed The government has opened services to foreign costs of establishing relationships with international buyers and may confer some scale-related cost benefits. However, for competition, but there are impediments in the the purpose of analyzing the degree of integration of exporters into GVCs, transactions covering direct imports are the most 68  The World Bank’s Logistics Performance Index (LPI) rates critical—firms that are integrated into GVCs would normally the transport and logistics environment in 160 countries. LPI source directly, as these sourcing relationships are critical from a rates countries on six factors: customs clearance, infrastructure, quality and technology spillover perspective. See the background international shipping, logistics services, tracking and tracing, paper prepared for this report by Pierola et al. (2015). and timeliness and reliability. 8. Using spillovers from international trade Fi ur 83. S rvic s tr d costs r hi h Fi ur 84. Nonw l bor costs r hi h (Services Trade Restrictiveness Index, STRI and development (LHS in % of gross remuneration, RHS times the remuneration, 2011) & value added in services as part of GDP, 2013) 80 80 1.6 70 70 1.4 60 60 1.2 50 THA MYS 50 1.0 40 CHN 40 0.8 ZAP 30 CHL 30 0.6 20 20 0.4 10 PER COL 10 0.2 0 0 0.0 20 40 60 80 100 General regime Agrarian regime Micro entreprises regime Services VA/GDP Familly allowance Vacations Life insurance Pension Supplementary payments (grati cations) Severance compensation (CTS) Training service Social health insurance Complementary labor risk indemnity for dismissal (times the remuneration) Source: World Bank STRI; WDI. Source: Sta calculations based on ENAHO, MEF, and Central Bank data. domestic regulatory regime, which limits entry and High non-wage labor costs reduce the competition (Section 5). Regulatory structures and competitiveness of Peru’s exports. During the institutions lag behind those in comparable countries 2000s, labor earnings grew faster than productivity, affecting services trade performance.69 Firm-level particularly in manufacturing, driven by nonwage analysis concludes that domestic reforms aiming at costs which, under the General Labor Regime, account increasing foreign firm participation and enhancing in for 68.2 percent of the basic wage, by far the highest services markets have an economy-wide downstream in the region. Nonwage labor costs in Chile and Mexico 75 effect for Peruvian firms. The analysis shows that are below 30 percent and in Brazil around 56 percent. both foreign services firm participation and increased In practice, special regimes—including one specifically competition (measured through concentration for nontraditional exporters—and a high degree of indicators) in Peru’s services markets increase informality mean that few firms actually pay all the downstream productivity. In other words, downstream nonwage costs (Figure 84). But reliance on special industries that are more reliant on services as part regimes raises information and administration costs, of their input use benefit more from foreign firm which tend to hit smaller exporters hardest. participation and enhanced competition in upstream services. Finally, the analysis shows that both foreign Sector-specific constraints restrict nontraditional presence and increased competition in services also export sectors. In agriculture, for example, increasing have a significant and positive effect on downstream constraints on access to water are a barrier to productivity growth in Peru; also, next to the level of growth. In the textile value chain, limited access to productivity, the services reforms in Peru also have land prevents efficiently sized cotton suppliers from a significant effect on the growth of productivity. emerging, undermining the competitiveness of the To increase the services trade role in the Peruvian high value-added Pima cotton value chain. Major economy, reducing access barriers, improving market skill gaps are emerging in most sectors, contributing competition, identifying skills shortage relevant for to weak productivity growth and lower export services activities, and improving both electronic and competitiveness. These checks tend to hit smaller physical infrastructure would be crucial. exporters harder. 69  Reforms that were carried out in the 1990s (see Box 1) Several policies could improve Peru’s trade included the deregulation of some key service markets, such as the financial, electricity, and telecommunications sectors, performance, raising productivity and exposing them to increased foreign competition. Moreover, in international competitiveness. Peru’s experience the 2000s, Peru signed several trade agreements with its major trading partners, deepening the conditions of competition in has shown that trade liberalization is a necessary— service sectors. However, important restrictions to competition but not sufficient—condition for increasing in services markets, such as professional services, still prevail, as described in Section 5 of this report. competitiveness; it is now turn for domestic Peru Building on Success: Boosting Productivity for Faster Growth Box 7 Bureaucratic and infrastructure obstacles lower export competitiveness While the three successful nontraditional private investments, more than 200,000 export sectors have largely overcome new hectares are expected to be added constraints, they still face obstacles, which to agriculture by 2020. But the two fruit- other less successful firms also grapple with. exporting ports are not fully prepared to receive the export volume, especially during In high-quality cotton apparel, Peru’s irrigation the high season, when delays are common. The infrastructure offers favorable conditions road infrastructure has improved recently but for cotton production. However, it has been road quality is still not good and delays imperil uncompetitive in attracting foreign direct the quality of produce. The sector is burdened investment (FDI) to the sector because of by costly delays and bureaucratic procedures relatively high labor costs, absence of tax related to government services necessary incentives, economic processing zones, and to open facilities, clear customs, or handle bureaucratic and infrastructure constraints. phytosanitary issues. Inadequate port and highway infrastructure increased costs of trade, eroding the “fast In mining equipment, Peru is at the geographic fashion” advantages of proximity to key center of Latin America’s mining region. This markets and the ability to produce clothing is a competitive advantage for shipping large, rapidly. Transportation costs can be up to heavy items, but infrastructure weaknesses twice that of shipping from Chile. erode it. Road, rail, and port infrastructure undermine the industry, despite investments In table grapes, the coastal growing to expand the Port of Callao, and regulatory region benefits from abundant irrigation challenges delay land zoning and the 76 infrastructure, and with planned public– establishment of industrial operations. Source: Background papers prepared for this report by Fernandez-Stark et al. (2015a), Fernandez-Stark et al. (2015b), and Fernandez-Stark et al. (2015c). reforms. These include improving the behind-the- border environment, including streamlining customs clearance regimes; increasing the capacity and quality of transport infrastructure, logistics, and storage and distribution services; improving incentives for consolidating freight services; consolidated reforms aimed at adoption and consolidation of the sanitary and phyto-sanitary standards is also needed. A more robust innovation environment would also facilitate diffusion of GVC-acquired knowledge and improve downstream links for exporters. Addressing domestic regulations in services would facilitate their role of supplier of intermediate inputs for exports of manufacturing and more complex services. Finally, improving labor market flexibility and reducing nonwage costs would also contribute to higher trade volumes, as would building human capital and more complex skills that can be used in higher value-added manufacturing and services (these are analyzed in Section 6). 9. Unleashing innovation 9. Unleashing innovation I nnovation is a driver of within-firm productivity, but if innovation inputs are not equally accessible it can also feed misallocation of factors. Innovation in developing countries relates more to the adoption and diffusion of existing knowledge and respectively (Figure 85). Research and Development (R&D) is low and mostly public: 29 percent of firms invested in it in 2004 (the latest year with data), while in Chile and Colombia 40 percent did that year, but today 45 percent do so. 77 technologies than to the creation of new-frontier technologies (Box 8). But any kind of innovation is Low rates of innovation may be reflected in Peru’s at the heart of raising productivity. However, if not industrial composition and its undiversified exports. all firms have equal access to innovation inputs, it A more diversified export basket with diverse levels of might also feed factor misallocation. For example, in value added can make an economy more competitive Peru small and young firms are more productive (see and resilient in the face of external shocks. Peru’s Section 3), and if they have less access to inputs for manufacturing exports, however, are concentrated innovation such as finance, this shortcoming in the in low-technology industries and primary products. innovation system would contribute to reducing overall productivity, while still increasing within-firm Fi ur 85. P ru l s b hind th r ion productivity for large firms. This section looks into the in t chnolo doption specific features of the Peruvian innovation system (% of rms, latest year available) and considers options for improvements that would be 100 most effective in raising overall productivity. 80 Peruvian firms innovate little 60 Peru lags behind other comparable countries on 40 adopting existing technologies. According to Peru’s latest Enterprise Survey, 14.2 percent of firms have 20 an internationally recognized quality certification, slightly lower than the LAC average of 16.2 percent 0 Percent of rms Percent of rms Percent of rms but far lower than the OECD average of 32.8 percent. using technology using e-mail to with an internationally Only 7.7 percent of firms in Peru adopt technology licensed from interact with -recognized foreign companies clients/suppliers quality certi cation licensed from foreign companies while the LAC and High income: OECD Latin America & Caribbean Peru OECD averages are 14.2 percent and 18.2 percent, Source: Enterprise Surveys, World Bank. Peru Building on Success: Boosting Productivity for Faster Growth Box 8 What is meant by innovation? Innovation refers to activities (creation of for example) or a grassroots entrepreneur new products and improvement of existing (who starts using a phone for financial knowledge and production processes) that transactions). The most productive and either expand the technological frontier inclusive kind of innovation seems to be in or absorb and adapt existing technology. the middle when firms can be “piggyfrogging” Innovating at the frontier requires high- through technological change: leapfrogging to quality education, continuous investments in wide use of new technologies by piggybacking research and development, and well-defined on the existing knowledge and patent base. and enforceable property rights. These create Importantly, this kind of innovation requires incentives for innovation as well as provide the sufficient absorptive capacities in the economy, required inputs for “innovators” (entrepreneurs). which are akin to those required for frontier innovation, although less sophisticated. But most innovative activities are not at the frontier. This category is based on absorbing In this report we use the term “innovation” to knowledge—typically from abroad—through identify activities that are new to the firm or international transfers and spillovers. Even a country (and thus can be based on existing simple applications of existing knowledge technologies adapted to local context), and can be innovative from the perspective of a “frontier innovation” to refer to the creation of company (that adopts a new product line, completely new knowledge. Sources: Freeman (1987), Lundvall (1992), Nelson and Rosenberg (1993), Metcalfe (1995). 78 Only 0.4 percent of exports are considered high tech, Firms that do invest in innovation in Peru are more against 2.2 percent in Argentina, 3.3 percent in Brazil, likely to introduce new products, but low returns (in 0.9 percent in Colombia, and 6.0 percent in Chile. sales) to innovation may be contributing to the low level of investment observed. Innovation investment Low investment in innovation might be correlates strongly with firms’ sales, but firms’ returns related to low returns to innovating are lower in Peru than in other countries. If a Peruvian firm spends on innovation-related activities, it is more Firms invest too little in innovation, indicating that likely to introduce a new product or new process than the innovation system might not be functioning firms in other LAC countries except Chile (Figure 87).71 properly. In Mexico, Colombia, and Peru within- Firms that introduce new products or processes have firm productivity growth has been the main driver an average of 38 percent higher sales per employee. of productivity convergence with the international While a substantial return, this is significantly lower and domestic frontiers. In Mexico and Colombia, than in five other countries (where the increase is on innovation has been the main driver of firm-level average 100 percent, Figure 88). A 1 percent increase productivity convergence, but this is not the case in in innovation expenditures leads to a 0.22 percent Peru.70 Peruvian firms invest on average 2.5 percent increase in sales per worker in Peru while in other LAC of their sales in innovation whereas peers in Chile countries—Costa Rica aside—the elasticity ranges invest 3.5 percent. Firms from advanced economies from 0.2 in Chile to 0.69 in Panama. invest even more: up to 5.6 percent of sales in the European Union (Figure 86). 71  Innovation expenditures include technology transfer, computing hardware and software purchases, design and industrial engineering, marketing activities for innovation, other internal R&D, and other external R&D. Source: Encuesta Nacional 70  Brown et al. (2015) and Iacovone and Tran (2015). de Innovacion 2013 by INEI. 9. Unleashing innovation Fi ur 86. M nuf cturin firms in P ru inv st littl in innov tion (% of rms’ sales, countries grouped roughly by their income per capita level) 6 R&D intensity Innovation Expenditure Intensity 5 4 3 2 1 0 zil en ge Arg ca s Ge k e a ay Co a bia Ne rage um d K ria m ay Co u De e Lu any urg in nd ar tin nc il m r a do Ri Bra rw gu Ch Pe ed era t lom Sp nm na lgi bo Un Aus Fra rla rm en e ing sta Uru Sw No Av Av Pa Be xem the ite Source: Innovation Surveys based on data from OECD (2009), IDB (2011), and National Innovation Survey. Note: Data refer to manufacturing industries only. Fi ur 87. Innov tion inv stm nt Fi ur 88. ...but r turns to innov tion s ms to work... nd to inv stm nts in innov tion r low in P ru (marginal increase in probability to have new process (sales per worker increase due to (sales per worker increase due to or product due to innovation expenditure, %) new product or process, %) innovation expenditure, %) 1.2 2.5 0.8 0.7 1.0 2.0 0.6 0.8 1.5 0.5 79 0.6 0.4 1.0 0.3 0.4 0.2 0.2 0.5 0.1 0.0 0.0 0.0 a Co a a a ile ru Co y bia Arg a a ia Uru a ile Arg ru Co a Uru ia Arg ay ru Co ile a Co ay a tin Ric tin tin m m am Ric Ric mb b gu gu gu Ch Ch Ch Pe Pe Pe lom lom na na en en en n sta sta sta Uru lo Pa Pa Pa Co Source: Authors’ calculations based on Encuesta Nacional de Innovación Source: Authors’ calculations based on Encuesta Nacional de Innovación 2012: Background paper prepared for this 2012: Background paper prepared for this report by Nguyen et al. (2015). report by Nguyen et al. (2015). Note: Coe cients reported are marginal e ects, i.e. they predict the Note: Percentage increase in sales from a 1% spending increase on innovation, or from an introduction of new likelihood of introducing product or process innovation. Controls include size product of process. Controls include rm size, age, nontechnological innovation, sector, and region dummies. of the rm, whether it exports, and whether it is foreign owned. Young and small firms—which are the most financing and market domination by other firms as productive in Peru—appear to be especially major constraints (see Figure 89). constrained in their innovation efforts. When innovation happens, it is more likely among larger Outside-of-firm constraints, especially in product firms. During 2009–11, about 30 percent of firms had and intermediate goods and services markets could some technological innovation. The average age of the be causing low returns and reinforce biased access firms that innovate is 22 years, 5 years older than the to inputs. The fact that investment in innovation average age of firms that do not. The average number gives high returns to innovation outcomes, which then of employees of innovating firms is 75 percent higher give low returns on sales, points to possible product than in other firms—350 versus 200. Small and young market competition and regulation issues. As pointed firms therefore seem to be facing specific constraints out by more than 25 percent of firms surveyed in the to innovating. Indeed, while old firms mostly complain Encuesta Nacional de Innovación (National Innovation about quality of personnel, young firms also perceive Survey 2012), market dominance by established Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 89. Firms’ constr ints to innov tion diff r b Fi ur 90. Mor firms innov t in th co st l r (% of rms declaring it very important) ( rms that innovate by region, %) 40 Lack of quali ed personnel 35 High cost 30 Easy to imitate 25 20 Lack of internal nancing 15 Market dominated by other rms 10 Lack of external nancing 5 0 0 5 10 15 20 25 30 35 40 Costa Lima Sierra Metropolitan and Selva 40+ 10-39 <10 Source: Encuesta Nacional de Innovación 2012. Source: Authors’ calculations: background paper by Nguyen et al. (2015), Note: only constraints declared very important by 25% or more rms are reported. based on National Innovation Survey 2012 (NIS). Note: Costa excludes Lima Metropolitana. firms is one of the top five constraints to innovation important part of the innovation system—have also (Figure 89). The fact that firms are facing unequal been shown to be important in firms’ decisions to access to innovation inputs could also reinforce invest in innovation. In LAC in general, improvements factor misallocation issues—especially when the in property rights are shown to have strong impacts most productive young and small firms are the ones on firms’ decision to innovate, unlike in other regions.73 that suffer. Similarly, it is possible that sales returns 80 are low not only because of product market issues, The adoption and creation of knowledge are but because of the intermediate goods and services restricted by having too few scientific researchers markets: if transportation services or financial or university graduates in science and technology. services for expansion are not available, expensive, Peru ranks 113 out of 144 countries (Figure 91) in or have uneven access points, not all firms can terms of availability of scientists and engineers, and receive equal access to inputs necessary to expand is the lowest among peer countries. Industries point production after having innovated. to shortages of engineers and technicians as an important constraint on their performance, innovation, The underdeveloped innovation and growth. Several industries highlight the need for system constrains invention, experienced engineers whom they cannot find in Peru. adoption, and diffusion (The number of engineering graduates is growing, however, and around 25 percent of university students Low innovation effort reflects Peru’s innovation are taking science and engineering programs.) system shortcomings. The national innovation system (Box 9) is still at an embryonic stage. The elements are The diffusion of new knowledge is limited, indicating there but the capacity of actors to fully undertake transmission problems in the innovation system. activities formally remains limited: the system is The linkage between industry and science for uncoordinated and does not seem to have a common strategic vision, such that it lags behind the systems 73  Nguyen and Jaramillo (2014). Regarding the intellectual in other emerging economies.72 These deficiencies are property rights, given the low levels of frontier innovation and firms’ knowledge about intellectual property rights (IPR), they are reflected in inputs and investment in innovation and unlikely to be the main cause of Peru’s problems at this point in knowledge (human capital and education, science the development process. Peru does have a less developed IPR environment however, compared to Colombia, Mexico, and Chile. and technology, technology adoption, software The 2015 Global Intellectual Property Center (GIPC) Index is a investment, and other intangibles), and in knowledge composite indicator that measures the national IPR environment based on the state of legal frameworks and their enforcement. and innovation outputs. Property rights—another The Index maps the IPR environment of 30 economies, accounting for nearly 80 percent of global GDP. Economies’ GIPC Index scores 72  Several innovation reviews have been conducted including are evaluated on the basis of 30 indicators indicative of a robust by OECD (2009) and UNCTAD (2011). IPR system. 9. Unleashing innovation Box 9 Peru’s national innovation system Building a policy framework for innovation participants of the innovation process, as well requires coherent actions in several policy as agglomeration effects in cities, facilitate areas: education, trade, investment, finance, knowledge spillovers. and decentralization. Innovation can be likened to gardening: prepare fertile ground Peru has innovation and competitiveness (education), nurture the soil (R&D, information funds and public and private instruments for transmission, and connectivity), remove technology transfer, such as the National weeds (competition and regulation policy), and Fund for Science, Technology and Innovation irrigate and fertilize (finance, and other support Fondo para la Innovación Ciencia y Tecnología for innovators). (FINCyT), the R&D Fund for Competitiveness Fondo de Investigación y Desarrollo para The national innovation system comprises la Competitividad (FIDEICOM), and the the firms, research institutes, universities, Framework Fund for Innovation, Science and financial institutions, and R&D industry Technology Fondo Marco para la Innovación, that jointly contribute to innovation. The Ciencia y Tecnología (FOMITEC). The system’s success depends on linkages among government has introduced fiscal incentives these agents, and on the environment and (tax deductions) for R&D investment and incentives for collaboration: the education acquisition of equipment. system develops absorption capacity for learning, technologies, and ideas; finance is Peru lags behind other Latin American necessary to commercialize ideas and develop countries in offering broader support for new projects; the links between academia, science and technology. The Box Table shows R&D, and firms allow for creation; absorption the range of science and technology programs 81 and diffusion of knowledge among all the for LAC countries. Box Table: Peru implements fewer supply-side instruments than other LAC countries Instrument/ URY ARG MEX PAN CHL BRA COL CRI PER PRY DOM GTM SLV Country Science a nd x x x x x x x x x x x x technology funds Support to x x x x x x x centers of excellence Scholarships: x x x x x x x x x undergrad., grad., or postgrad. Support for x x x x x x x x x national postgrads. Salary x x x x incentives to research Affiliation x x x x x with national researchers abroad Sources: Background paper prepared for this report by Zuniga (2015), World Bank (2010) Innovation Policy, Freeman (1987), Lundvall (1992), Nelson and Rosenberg (1993), Metcalfe (1995); Table updated from Políticas e Instrumentos en Ciencia, Tecnología e Innovación en América Latina y el Caribe 2009: Inter-American Development Bank (IDB), Red de Ecología Social (REDES), and Red de Indicadores de Ciencia y Teconología (RICYT). Peru Building on Success: Boosting Productivity for Faster Growth Fi ur 91. P ru’s low innov tion b s Fi ur 92. Poor univ rsit -industr coll bor tion in R&D slows diffusion (1=among the worst in the world, 7=among the best in the world) (1=among the worst in the world, 7=among the best in the world) 6 6 5 5 4 4 3 3 2 2 1 1 0 0 sia sia ile o ica o CD nd So bia C u CD ic a ile d bia C u LA LA xic xic an r r Ch Ch Pe Pe ila OE OE lay lay Afr Afr lom lom ail Me Me a Ma Ma Th uth Th uth Co Co So Availability of scientists and engineers Quality of scienti c research institutions Source: WEF Executive Opinion Survey. 2013–14 weighted average. Source: WEF Executive Opinion Survey. 2013–14 weighted average. knowledge sharing in Peru is the weakest among peer (see also Section 8). Firms in the coastal area, where countries (Figure 92). In Peru, most of firms innovate in density is higher, are more likely to innovate (Figure isolated fashion. Firms’ production of innovation of any 90)—an encouraging sign. kind—product, process, or marketing- or organization- based—are carried out with their own funds and Public policy can help improve 82 without any collaboration with other entities. University innovation outcomes and industry collaboration in R&D is the weakest in Peru among peer countries (Peru ranks 109 out of 144 Public investment in innovation is low but also countries in the corresponding WEF indicator). This lacks effectiveness. Peru’s public expenditures on suggests the need for exploiting spillovers, increasing science, technology, and innovation as a share of interactions with research institutions and public GDP are low compared with other LAC countries. The programs to increase efficiency in innovation, raising Consejo Nacional de Ciencia, Tecnologia e Innovación incentives to innovate, and lowering innovation costs Tecnológica (CONCYTEC) is a principal public-funding for firms.74 Diffusion of technologies from the frontier agency for science and technology, yet its 2013 firms to other firms in the same industry is low in Peru, budget was a mere 0.009 percent of GDP—innovation while such spillovers have been important factors of productivity convergence in Mexico and Colombia.75 Vertical spillovers from firms to suppliers — that Fi ur 93. Low public inv stm nt in innov tion improve quality of inputs for other firms in the same (% of GDP) industry — have been shown to be most effective 0.25% 74  Tello (2013). 75  Brown et al. (2015) and the background paper by Iacovone 0.20% and Tran (2015) regress firm-level labor productivity growth (change in value added per employee) on a range of variables 0.15% including “spillover” terms that are the change in the local and global frontiers of labor productivity. Including these terms controls for technological shifts, domestically and globally. If 0.10% there are spillovers from the best firms to other firms (such as through learning) so that productivity in the average firm improves with productivity of the frontier firms, these terms will 0.05% be positive. Contrary to results for Mexico and Colombia, they find the terms are not significant, so there is no evidence that firms, 0.00% even those in Lima and Callao, were able to grow faster when Chile Mexico Colombia Peru there were positive technological shifts at the frontier. This could be related to data quality and Peru sample size and should be Source: Informe de Rendición de Cuentas de la Administración Publica 2011 interpreted with care, however. (Mexico); CONICYT Institutional Brochure 2013 (Chile); Informe de Gestión 2012 National Survey of Innovation in Manufacturing Industry 2012. (Colombia); Plan Operativo 2012 (Peru). 9. Unleashing innovation agencies in regional countries had budgets that were 6 Fi ur 94. Sh r of firms r c ivin public support to 20 times higher (Figure 93). But public support also to fin nc innov tion reaches very few firms—less than 1 percent (Figure (%) 94)—among the smallest coverage in LAC countries Peru 0.7 for which data are available. In Brazil and Chile, Uruguay 1 Panama 3 innovation policy instruments are being used by 6 and Costa Rica 3.4 8 percent of firms, respectively. Thus, the amount of Argentina 5 Colombia 5 public spending alone is not an indicator of success or Brazil 6 Chile 8 policy direction, if effectiveness is not improved. Switzerland 5.7 Australia 7 The improvement of agencies that diffuse existing Japan 7.8 Denmark 12.6 technology from abroad to local firms, particularly United Kingdom 12.8 France 13.1 small and medium-size enterprises, could help Germany 13.2 boost productivity. More specific actions could Belgium 16.9 Netherlands 20.5 include improving the capacity of the CITEs (Centros Normway 22.7 Luxembourg 22.8 de Innovación Tecnológica Empresarial) by introducing Finland 23 Austria 24.5 a performance-oriented approach and revising their Korea 32 goals, operation, governance, and management Canada 40.4 0 10 20 30 40 50 capacity. The centers that serve the “Pisco” industry in Peru have shown a potential path for these type of Source: Ministry of Production, Peru. agencies. Horizontal policies are also important. By ensuring equal access to innovation avenues and financing for small firms, such policies would improve factor 83 allocation toward more productive firms. Policies aimed at correcting misallocation, such as easing regulatory burden on start-ups (Section 5), elimination of firing constraints (Section 6), improving access to credit (Section 7), or increasing imports that provide technology spillovers (Section 8), would also positively impact incentives to innovate and innovation outcomes. Opening the research and university system to different forms of knowledge transfers and collaboration with the private sector would also be helpful. Moreover, enabling better use of knowledge entails improving the governance and the legal rights enforcement for institutions and companies to engage in collaboration and innovative ventures. Part V Conclusions Conclusions P eru has emerged as a new growth star in the Latin America and Caribbean region, with prosperity widely shared among its 30 million people. Peru’s economy was the second- fastest growth performer in the region over the last The country’s services sector is a chronically poor performer in this regard. This weakness has hurt overall productivity and output directly and also indirectly by crippling an intermediate goods and services market that could spur exports and decade, with per capita income doubling— well ahead production in other sectors such as manufacturing. of the region as a whole. Since 2000 almost a quarter Eliminating these distortions could increase overall of Peru’s population has broken free of poverty, and productivity by up to 130 percent and double Peru’s inequality plummeted at one of LAC’s fastest rates. output per worker. Across the country, lower-income households, the bottom 40 percent, have seen their income grow more Reducing those frictions will spur productivity than 2 percentage points faster than that of the average growth, but that in turn requires deepening key income. While the country struggled to converge in the reform areas. Macroeconomic and structural reforms 1970s and 1980s, the speed of convergence recovered over the last 20 years have played a central role in 87 in the mid-1990s, and accelerated like never before over Peru’s growth. But the path toward higher income the last decade. Throughout this period of growth and status, including in the context of a less-benign good external conditions, and unlike many countries, external environment, hinges on reducing market Peru saved the windfall, leaving itself with significant frictions that lead to inefficient allocation of resources savings to continue to afford needed investments in and enabling the environment for firms to grow. Such infrastructure and with strong macroeconomic buffers growth would require a suite of microeconomic reforms to face more challenging times. that are challenging, but are doable and within reach of public policy in Peru.  Specific policy directions include A portion of Peru’s growth has been driven by (i) reducing costs of entry and operation that hamper improvements in productivity, but in the next SMEs’ growth and reduce competition pressures stage of convergence, and under the new external in the market; (ii) reducing the rigidities imposed conditions, a larger contribution to economic by labor laws, allowing employers more flexibility, growth will have to come from higher productivity. which could also help reduce the informal sector that Although the country’s productivity growth at the hampers overall productivity; (iii) improving quality of company level is at par with similar middle-income education, particularly higher secondary education countries that aspire to avoid the middle-income trap, and training, to improve the skills match between the gap to the global productivity frontier of higher- labor supply and demand, which also has an effect income countries remains large. This report showed on reducing informality; (iv) deepening the financial that enhancing productivity growth remains the sector, particularly to reduce borrowing costs for fastest way to close it. SMEs so they can invest in their growth; (v) improving infrastructure, with a focus on infrastructure for trade, There is firm-level evidence that Peruvian markets logistics, and facilitation as well as lifting other behind- tend to misallocate labor and capital into less- the-border constraints to trade; and (vi) improving productive workplaces. This signals that some the innovation framework to help firms innovate and aspects of product, factor, and intermediate goods adopt new technologies, enabling them to compete in and services markets do not function properly. different markets. Peru Building on Success: Boosting Productivity for Faster Growth Annex Background papers prepared for this report Background paper title Authors GPs 1 Reforms and growth Nikita Céspedes and Patricia Peru MEF and BCRP Lengua-Lafosse 2 Economic growth in Peru Cristina Savescu and Ekaterina GMFDR Vostroknutova 3 The middle-income trap and Peru Harry Moroz and Ekaterina GTIDR, GMFDR Vostroknutova 4 Disentangling (more than) two decades of Daniel Barco and Ekaterina GMFDR fast growth in Peru Vostroknutova 5 Growth, poverty & shared prosperity in Elizaveta Perova, Maria Eugenia GPVDR Peru Genoni, Anna Bonfert, and 88 Santiago Garriga 6 Regional productivity convergence in Peru Leonardo Iacovone, Luis Fernando GTCDR, DECWD Sanchez Bayardo and Siddharth Sharma 7 Firm-level convergence of productivity in Leonardo Iacovone and Trang Thu GTCDR, DECWD Peru Tran 8 The role of imports for exporter Martha D. Pierola Castro, Ana M. DECTI, GCJDR performance in Peru Fernandes, and Thomas Farole 9 Policy actions for improved Thomas Farole and Juan Julio GCJDR competitiveness Gutierrez 10 Innovation and firms’ productivity in Peru Ha Nguyen, Patricio Jaramillo and DECMG, GMFDR Ekaterina Vostroknutova 11 Innovation system in development –the Pluvia Zuniga GTCDR case of Peru 12 Investment climate and competitiveness Alvaro Quijandria, Ernesto Franco- GTCDR in Peru Temple, Jessica Michelle Victor, and Peter Kusek 13 Peru—Tackling regulatory barriers Martha Licetti, Donato de Rosa, GTCDR, GMFDR to competition and local economy Tanja Goodwin, Congyan Tan, Lucia development Villarán, and Rachel Li Jiang Annex Background paper title Authors GPs 14 Services trade performance and Sebastian Saez and Erik L. Van GTCDR productivity in Peru: A competitiveness der Marel analysis 15 Measuring Peru’s integration into global Thomas Farole, Daria Taglioni, GCJDR, GTCDR value chains Guillermo Arenas 16 Peru in the table grape global value Karina Fernandez-Stark, Penny Duke Center on chain: opportunities for upgrading Bamber, and Gary Gereffi Globalization, Governance & Competitiveness 17 Peru in the high quality cotton textile Karina Fernandez-Stark, Penny Duke Center on and apparel global value chain: Bamber, and Gary Gereffi Globalization, Governance opportunities for upgrading & Competitiveness 89 18 Peru in the mining equipment global Karina Fernandez-Stark, Penny Duke Center on value chain: opportunities for upgrading Bamber, and Gary Gereffi Globalization, Governance & Competitiveness 19 Jobs in Peru: dynamics, constraints and Elizabeth Ruppert Bulmer, GCJDR policy implications Mathilde Perinet, Angela Elzir, David Robalino 20 Firms dynamics in Peru: analysis of jobs Reyes Aterido and Leonardo GCJDR, GTCDR and productivity Iacovone 21 Giving Peru a productivity boost: Gustavo Yamada, Jamele LCC6C towards a system of continuous Rigolini, and Pablo Lavado education and training 22 Tertiary education in Peru Javier Botero GEDDR 23 XXI century challenges in education in Inés Kudó and Miguel Székely GEDDR, DFGPE Peru 24 Considerations of the decentralization Carolina Rendon, Adrienne GGODR process in Peru Hathaway, and Katherine Grau 25 TFP gains from removing input–output Roberto Fattal DECMG distortions in Peru Peru Building on Success: Boosting Productivity for Faster Growth References Acemoglu, D., and F. Zilibotti. 1997. “Was Prometheus Barone, G., and F. Cingano. 2011. “Service Regulation Unbound by Chance? Risk, Diversification, and Growth.” and Growth: Evidence from OECD Countries.” Economic Journal of Political Economy 105.4: 709–51. http:// Journal 121.555: 931–57. http://www.oecd.org/eco/ www.econ.nyu.edu/user/debraj/Courses/Readings/ growth/46329580.pdf. AcemogluZilibotti.pdf. Bartelsman, E., J. Haltiwanger, and S. Scarpetta. 2009. Agenor, Pierre-Richard, Otaviano Canuto, and Michael “Cross-Country Differences in Productivity: The Role Jelenic. 2012b. “Avoiding Middle-Income Growth of Allocation and Selection.” Working Paper 15490, Traps,” Economic Premise Number 98, World Bank, National Bureau of Economic Research, Cambridge, Washington, DC. MA. http://www.nber.org/papers/w15490.pdf. Aiyar, Shekhar, Romain Duval, Damien Puy, Yiqun Wu, Brandt, L., J. Biesebroeck, and Y. Zhang. 2012. and Longmei Zhang. 2013. “Growth Slowdowns and “Creative Accounting or Creative Destruction? Firm- the Middle-Income Trap.” Working Paper, International Level Productivity Growth in Chinese Manufacturing.” Monetary Fund, Washington, DC Journal of Development Economics 97.2: 339–51. http://groups.chass.utoronto.ca/brandt/wp-content/ Akcigit, U., H. Alp, and M. Peters. 2015. “Lack of uploads/JDE_2012.pdf. Selection and Limits to Delegation: Firm Dynamics in Developing Countries.” Working Paper, University Bloom, N., E. Eifert, M. Aprajit, D. McKenzie, and J. 90 of Pennsylvania, Philadelphia, PA. https://economics. Roberts. 2013. “Does Management Matter? Evidence wustl.edu/files/economics/imce/akcigit_paper.pdf. from India.” Quarterly Journal of Economics 128: 1–51. https://web.stanford.edu/~nbloom/DMM.pdf. Araujo, J., M. Brueckner, M. Clavijo, E. Vostroknutova, and K. Wacker. 2015a. “Beyond Commodities: The Botero, J. 2015. “Tertiary Education in Peru.” Growth Challenge of Latin America and the Caribbean.” Background paper prepared for this report. World Bank, Washington, DC. http://documents. worldbank.org/curated/en/2014/12/24349586/ Brown, J. D., G. A. Crespi, L. Iacovone, and L. Marcolin. benchmarking-determinants-economic-growth-latin- 2015. Productivity Convergence at Firm Level: New america-caribbean. Evidence from Americas. In J. Araujo, E. Vostroknutova, K. Wacker, and M. Clavijo, eds. “Understanding Latin Araujo, J., E. Vostroknutova, K. Wacker, and M. America and the Caribbean’s Income Gap,” World Clavijo, eds. 2015b. “Understanding Latin America Bank, Washington, DC. http://documents.worldbank. and the Caribbean’s Income Gap” World Bank. LCR org/curated/en/2015/07/24838845/understanding- MFM. http://documents.worldbank.org/curated/ latin-america-caribbean%E2%80%99s-income-gap. en/2015/07/24838845/understanding-latin- america-caribbean%E2%80%99s-income-gap. Buera, F., and Y. Shin. 2010. “Financial Frictions and the Persistence of History: A Quantitative Exploration.” Aterido, R., and L. Iacovone. 2015. “Firms Dynamics in Working Paper 16400, National Bureau of Economic Peru: Analysis of Jobs and Productivity.” Background Research, Cambridge, MA. http://www.nber.org/ paper prepared for this report. papers/w16400.pdf. Barco, D., and E. Vostroknutova. 2015. “Disentangling Bulman, David, Maya Eden, and Ha Nguyen. 2014. (More Than) Two Decades of Fast Growth in Peru.” “Transition from Low-Income Growth to High-Income Background paper prepared for this report. Growth: Is There a Middle Income Trap?” Policy Research Working Paper 7104, World Bank, Washington, References DC.Caprettini, B. 2014. “Taxes, Misallocations and Corperación Andina de Fomento. 2013/2014. Productivity.” http://pedl.cepr.org/sites/default/ “Encuesta de medición de capacidad financiera en files/Taxes%2C%20Misallocations%2C%20and%20 Perú.” Informe. http://scioteca.caf.com/bitstream/ Productivity.pdf. handle/123456789/740/ENCUESTA%20DE%20 MEDICION%20DE%20LAS%20CAPACIDADES%20 Caselli, F. 2005. “Accounting for Cross-Country F I N A N C I E R A S % 2 0 E N % 2 0 LO S % 2 0 PA Í E S E % 2 0 Income Differences.” In P. Aghion and S. N. Durlauf, ANDINOS%20-%20COMPARATIVO. eds., Handbook of Economic Growth. Amsterdam: North pdf?sequence=5&isAllowed=y. Holland, vol. 1A, p. 1060; vol. 1B, p. 146. Correa, P. 2014. “Public Expenditure Reviews in Caselli, F. 2015. “The Latin American Efficiency Gap.” In Science, Technology, and Innovation: A Guidance J. Araujo, E. Vostroknutova, K. Wacker, and M. Clavijo, Note.” World Bank, Washington, DC. https://www. eds., “Understanding Latin America and the Caribbean’s innovationpolicyplatform.org/sites/default/files/rdf_ Income Gap” World Bank. LCR MFM. http://documents. imported_documents/public%20expenditure%202014. worldbank.org/curated/en/2015/07/24838845/ pdf. understanding-latin-america-caribbean%E2%80%99s- income-gap Duranton, G., S. Ghani,A. Goswami, andW. Kerr. 2015. “The Misallocation of Land and Other Factors of Production Castro, J., and G. Yamada. 2012. “‘Convexification’ and in India.” World Bank Policy Research Working Paper ‘Deconvexification’ of the Peruvian Wage Profile: A Tale 7221. World Bank, Washington, DC. http://www-wds. of Declining Education Quality.” Discussion Document worldbank.org/servlet/WDSContentServer/WDSP/ DD/12/02. University of the Pacific, Stockton, CA. IB/2015/03/23/090224b082c6a137/1_0/Rendered/ http://www.up.edu.pe/ciup/SiteAssets/Lists/JER_ PDF/The0misallocat00production0in0India.pdf. Jerarquia/EditForm/On_convexity_of_wage_profile_ v10.pdf. Easterly, W., and R. Levine. 2001. “It’s Not Factor 91 Accumulation: Stylized Facts and Growth Models.” Céspedes-Reynaga, N., and F. Orrego. 2014. World Bank Economic Review 15.2: 177–219. https:// “Competencia de intermediarios financieros en williameasterly.files.wordpress.com/2010/08/33_ Perú.” Working Paper 2014-010. Central Reserve easterly_levine_itsnotfactoraccumulation_prp.pdf Bank of Peru, Lima, Peru. http://www.bcrp.gob.pe/ docs/Publicaciones/Documentos-de-Trabajo/2014/ Eichengreen, Barry, Donghyun Park, and Kwanho Shin. documento-de-trabajo-10-2014.pdf. 2013. “Growth Slowdowns Redux: New Evidence on the Middle-Income Trap,” Working Paper 18673, National Céspedes, N., and P. Lengua-Lafosse. 2015. “Reforms Bureau of Economic Research, Cambridge, MA. and Growth.” Background paper prepared for this report. Executive Opinion Survey. 2014. “Global Risks of Highest Concern for Doing Business.” In Global Risks 2015. World Choy, M., E. Costa, and E. Churata. 2015. “Radiografía Economic Forum. London School of Economics and del costo del crédito en el Perú.” Working Paper 2015- Political Science, London, UK. http://www3.weforum. 001. Central Reserve Bank of Peru, Lima, Peru. http:// org/docs/GCR2013-14/GCR_Chapter1.3_2013-14.pdf. www.bcrp.gob.pe/docs/Publicaciones/Documentos- de-Trabajo/2015/documento-de-trabajo-01-2015.pdf. Falco, P., and S.A. Caria. 2014. “Do employers trust workers too little? An experimental study of trust Coronado, J., et al. 2014. “Midiendo el in the labour market.” Working Paper WPS/2014- costo económico potencial de las barreras 07. Center for the Studies of African Economies. burocráticas en el Perú.” Observatorio de University of Oxford, Oxford, United Kingdom. http:// Mercados 8.16: 1–393. http://www.indecopi.gob. www.economics.ox.ac.uk/materials/papers/13752/ p e / re p o s i t o r i o a p s/ 0/ 0/ j e r/ p u b l i c a c i o n e s q s/ csae-wps-2014-07.pdf. ObservatorioCostoBarrerasBurocr%C3%A1ticas2013. pdf. Peru Building on Success: Boosting Productivity for Faster Growth Farole, T., and J. J. Gutierrez. 2015. “Policy Actions Hsieh, C-T and P. J. Klenow. 2014. “The Life Cycle for Improved Competitiveness.” Background paper of Plants in India and Mexico.” Quarterly Journal of prepared for this report. Economics (129) 3: 1035–84. Farole, T., D. Taglioni, and G. Arenas. 2015. “Measuring Iacovone, L., and T. Thu Tran. 2015. “Firm-level Peru’s Integration into Global Value Chains.” Convergence of Productivity in Peru.” Background Background paper prepared for this report. paper prepared for this report.Iacovone, L., L.F. Sanchez Bayardo, S. Sharma. 2015. “Regional Productivity Fattal, R. 2015. “TFP Gains from Removing Input– Convergence in Peru.” Background paper prepared for Output Distortions in Peru.” Background paper this report. prepared for this report. ILO. 2012. Statistical update on employment in the Feenstra, Robert C., Robert Inklaar, and Marcel P. informal economy http://laborsta.ilo.org/applv8/ Timmer. 2015. “The Next Generation of the Penn World data/INFORMAL_ECONOMY/2012-06-Statistical%20 Table.” American Economic Review 5.15: 1–41. http:// update%20-%20v2.pdf www.rug.nl/research/ggdc/data/pwt/v81/the_next_ generation_of_the_penn_world_table.pdf. Inchauste, G., Sergio Olivieri, Jamie Saavedra, and Hernan Winkler. 2012. “What Is Behind the Decline in Felipe, Jesus, Arnelyn Abdon, and Utsav Kumar. 2012. Poverty since 2000? Evidence from Peru, Bangladesh, “Tracking the Middle-Income Trap: What Is It, Who Is and Thailand.” Policy Research Working Paper in It, and Why?” Working Paper 715, Levy Economics 6199, World Bank, Washington, DC. http://www. Institute, Annandale-on-Hudson, NY. researchgate.net/publication/256034569_What_ is_Behind_the_Decline_in_Poverty_Since_2000_ Fernandes, A. M., M. D. Pierola Castro, and T. Farole. Evidence_from_Bangladesh_Peru_and_Thailand. 92 2015. “The Role of Imports for Exporter Performance in Peru.” Background paper prepared for this report. INDECOPI. 2014. “Propuesta para Mejorar la Competitividad del País.” Press Conference, November Fernandez-Stark, K., P. Bamber, and G. Gereffi. 2015a. 25, Lima, Peru. http://agenciaorbita.org/indecopi- “Peru in the High Quality Cotton Textile and Apparel presenta-conjunto-de-medidas-para-mejorar-la- Global Value Chain: Opportunities for Upgrading.” competitividad-del-pais/. Background paper prepared for this report. INEI. 2014. Producción y Empleo Informal en el Perú: Fernandez-Stark, K., P. Bamber, and G. Gereffi. 2015b. Cuenta Satélite de le Economía Informal 2007–2012. “Peru in the Mining Equipment Global Value Chain: http://www.inei.gob.pe/media/MenuRecursivo/ Opportunities for Upgrading.” Background paper publicaciones_digitales/Est/Lib1154/index.html. prepared for this report. Informe de Rendición de Cuentas de la Administración Fernandez-Stark, K., P. Bamber, and G. Gereffi. Pública Federal. 2006–2012. SHCP, Instituto Federal de 2015c. “Peru in the Table Grape Global Value Chain: Acceso a la Información y Protección de Datos. Mexico Opportunities for Upgrading.” Background paper City, Mexico. prepared for this report. Inter-American Development Bank. 2010. The Age Freeman, C. 1987. Technology Policy and Economic of Productivity. Transforming Economies from the Performance: Lessons from Japan. London: Pinter Bottom Up. Development in the Americas. New York: Publishers. Palgrave McMillan. Gill, Indermit, and Homi Kharas. 2007. “An East Asian Jones, Charles I. 2015. “The Facts of Economic Growth.” Renaissance: Ideas for Economic Growth.” World Bank, Preliminary Version 0.4. April 6. http://www.hoover. Washington D.C. org/sites/default/files/jones-facts040.pdf. References Jovanovic, B.1982. “Selection and the Evolution of Morón, E., J. Tejada, and A. Villacorta. 2010. Industry.” Econometrica: Journal of the Econometric “Competencia y concentración en el sistema financiero Society 50.3: 649–70. http://www.andrew.cmu.edu/ en el Perú.” Revista de la Competencia y la Propiedad course/88-737/dynamic/selection.pdf. Intelectual 6.11: 41–85. http://www.slideshare.net/ NamerThomasRoger/competencia-y-concentracin- Kharas, Homi, and Harinder Kohli. 2011. “What Is the en-el-sistema-financiero-en-el-per. Middle-Income Trap, Why Do Countries Fall into It, and How Can It Be Avoided?” Global Journal of Emerging Moroz, H., and E. Vostroknutova. 2015. “The Middle- Market Economies 3.3: 281–89. Income Trap and Peru.” Background paper prepared for this report. Kitzmuller, M. and M. Martinez Licetti. 2012. Competition Policy: Encouraging Thriving Markets National Commission for Scientific and Technological for Development. Financial and Private Sector Research (CONICYT). 2013. Science and Technology for Development. Viewpoint note 331, World Bank, Chile. Institutional brochure. http://www.conicyt.cl/wp- Washington, DC. content/uploads/2012/07/Conicyt_Brochure_2013_ en.pdf. Krishna, P., and D. Mitra. 1998. Trade Liberalization, Market Discipline and Productivity Growth: New Nelson, R. R., and N. Rosenberg. 1993. “Technical Evidence from India. Journal of Development Economics Innovation and National Systems.” In R. R. Nelson, ed., 56: 447-62. National Innovation Systems—A Comparative Analysis, 3–21. New York: Oxford University Press. Kudó, I., and M. Székely. 2015. “XXI Century Challenges in Education in Peru.” Background paper prepared for Nguyen, H., and P. Jaramillo. 2014. “Institutions and this report. Firms’ Return to Innovation: Evidence from the World Bank Enterprise Survey.” Working Paper 6918, World 93 Leal-Ordónez, J. C. 2013. “Tax Collection, the Informal Bank, Washington, DC. http://www-wds.worldbank. Sector, and Productivity.” Review of Economic Dynamics org/servlet/WDSContentServer/WDSP/IB/2014/0 17.2: 262–86. 6/11/000158349_20140611115622/Rendered/PDF/ WPS6918.pdf. Licetti, M., D. de Rosa, T. Goodwin, C. Tan, and R. Li Jiang. 2015. “Peru—Tackling Regulatory Barriers Nguyen, H., P. Jaramillo, and E. Vostroknutova. to Competition and Local Economy Development.” 2015. “Innovation and Firms’ Productivity in Peru.” Background paper prepared for this report. Background paper prepared for this report. Lundvall, B. A. 1992. “User-Producer Relationships, Olley, S., and A. Pakes. 1996. “Dynamic Behavioral National Systems of Innovation and Responses in Longitudinal Data Sets: Productivity in Internationalization.” B. A. Lundvall, ed., National Telecommunications Equipment Industry.” University Systems of Innovation: Towards a Theory of Innovation of Pennsylvania, Philadelphia, PA. http://econweb.umd. and Interactive learning, 45–67. London, UK: Pinter edu/~haltiwan/olley_pakes.pdf. Publishers. Perova, E., M. E. Genoni, A. Bonfert, and S. Garriga. Metcalfe, S. 1995. “The Economic Foundations of 2015. “Growth, Poverty & Shared Prosperity in Peru.” Technology Policy: Equilibrium and Evolutionary Background paper prepared for this report. Perspectives.” In P. Stoneman, ed., Handbook of the Economics of Innovation and Technological Change, 409– Pierola Castro, M. D., A. M. Fernandes, and T. Farole. 512. Oxford, UK: Blackwell. 2015. “The Role of Imports for Exporter Performance in Peru.” Background paper prepared for this report. Midrigan, V., and D. Xu, 2014. “Finance and Misallocation: Evidence from Plant-Level Data.” American Economic Quijandria, A., E. Franco-Temple, J.M. Victor, and P. Review 104.2: 422–58. Kusek. 2015. “Investment Climate and Competitiveness in Peru.” Background paper prepared for this report. Peru Building on Success: Boosting Productivity for Faster Growth Rendon, C., A. Hathaway, K. Grau, and Z. Navarro. Bank, Washington, DC. http://www-wds.worldbank. 2015. “Considerations of the Decentralization Process org/external/default/WDSContentServer/WDSP/IB/2 in Peru.” Background paper prepared for this report. 006/09/13/000112742_20060913111024/Rendered/ PDF/359990WDR0complete.pdf. Restuccia, D., and R. Rogerson. 2008. “Policy Distortions and Aggregate Productivity with World Bank. 2014. “Latin American Entrepreneurs: Heterogeneous Establishments.” Review of Economic Many Firms but Little Innovation,” World Bank, Dynamics 11.4: 707–20. http://www.uh.edu/~bsorense/ Washington, DC. https://openknowledge.worldbank. RestucciaRogersonRED2008.pdf. org/handle/10986/16457. Ruppert Bulmer, E., M. Perinet, A. Elzir, and D. Robalino. World Bank. 2015. “Latin America and the Rising 2015. “Jobs in Peru: Dynamics, Constraints and Policy South: Changing World, Changing Priorities.” https:// Implications.” Background paper prepared for this openknowledge.worldbank.org/handle/10986/21869. report. World Bank. 2015a. “Peru: Selected Issues in Fiscal Saez, S., E.L. van der Marel. 2015. “Services Policy — Taxation and Equity.” Report No. ACS12766. Trade Performance and Productivity in Peru: A https://hubs.worldbank.org/docs/imagebank/pages/ Competitiveness Analysis.” Background paper docprofile.aspx?nodeid=24311031. prepared for this report. World Bank Enterprise Surveys. 2010. http:// Savescu, C., and E. Vostroknutova. 2015. “Economic microdata.worldbank.org/index.php/catalog/658. Growth in Peru.” Background paper prepared for this report. World Economic Forum. 2015. “Global Competitiveness Report 2014–2015.” http://reports.weforum.org/ 94 Schumpeter, J.A. 1942. “Capitalism, Socialism and global-competitiveness-report-2014-2015. Democracy”. New York: Harper & Row. 81-106. Yamada, G., J. Rigolini, and P. Lavado. 2015. “Giving Peru Syverson, C. 2011. “What determines productivity?” a Productivity Boost: Towards a System of Continuous Journal of Economic Literature 49.2: 326–65 http:// Education and Training.” Background paper prepared home.uchicago.edu/syverson/productivitysurvey.pdf for this report. Taglioni, D., and D. Winkler 2014. “Making Global Value Zuniga, P. 2015. “Innovation System in Development– Chains Work for Development.” Working Paper 143, The Case of Peru.” Background paper prepared for this World Bank, Washington, DC. https://openknowledge. report. worldbank.org/bitstream/handle/10986/18421/ 8 8 0 5 3 0 B R I 0 E P 1 4 0 d 0VC 0 k n o w l e d g e 0 n o t e s . pdf?sequence=1. Tello, M. 2013. “Quantitative Analysis Component: On the Determinants of Innovation in Services and Its Linkages with Productivity.” Working Paper 2013 (SS- IP)-05, Pontificia Universidad Católica del Perú, Lima, Peru. http://www.clacso.org.ar/libreria_cm/archivos/ pdf_617.pdf. Virgiliu, M., and D. Xu. 2014. “Finance and Misallocation: Evidence from Plant-Level Data.” American Economic Review 104.2: 422–58. World Bank. 2007. “Development and the Next Generation.” World Development Report 2007, World 95 The World Bank 1818 H Street, NW, Washington, DC 20433, USA. www.worldbank.org