46828 Aurora Ferrari and Inderbir Singh Dhingra India's Investment Climate India's Investment Climate Voices of Indian Business Aurora Ferrari and Inderbir Singh Dhingra Washington, D.C. © 2009 The International Bank for Reconstruction and Development /The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved. 1 2 3 4 12 11 10 09 This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank 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 judgement on the part of The World Bank concerning the legal status of any territory or the endorse- ment or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other 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. ISBN-13: 978-0-8213-7757-4 eISBN: 13: 978-0-8213-7808-3 DOI: 10.1596/978-0-8213-7757-4 Cover design by Candace Roberts, Quantum Think, Philadelphia, PA, United States. Library of Congress Cataloging-in-Publication data has been applied for. Contents Foreword xiii Acknowledgments xv Abbreviations xvi 1 Overview 1 The Indian Economy: Robust Growth but Challenges Remain 1 Manufacturing: Investment Climate Obstacles to Productivity Growth and Employment Generation 2 Retail Sector: Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Software and ITES Sector: Productivity Improvements Key to Continued Growth and Competitiveness 6 Reducing Inequalities between States by Improving the Investment Climate 8 Conclusions 9 2 The Indian Economy: Robust Growth, but Challenges Remain 11 Trends in the Economy 11 Key Growth Drivers in the Economy 12 Challenges in the Way Forward 19 Focus of the Report 23 Notes 27 3 Manufacturing: Investment Climate Obstacles to Productivity Growth and Employment Generation 29 The Manufacturing Sector: A Potential Driver of Growth 29 Firms' Perceptions of Investment Climate Constraints: Power, Taxes, and Lack of Skilled Workforce 31 Productivity Analysis of the Investment Climate and Firms' Perceptions of Key Obstacles 35 Impact of Investment Climate Obstacles on Growth Across Groups of States 37 The Investment Climate for Exporting Manufacturing Firms 40 Conclusions 45 Notes 46 Annex 3.A Obstacle Case Study--Power 47 Annex 3.B Obstacle Case Study--Corruption 50 v Contents 4 Unorganized Manufacturing: Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 53 Importance of Unorganized Manufacturing 53 Low Productivity, Low Wages, Weak Skill Levels, and Limited Integration with Large Supply Chains in Unorganized Manufacturing 54 Benefits of Becoming Organized and Why Some Firms Remain Unorganized 56 Access to Finance and Power Supply: Largest Constraints to Higher Productivity and Growth 61 Conclusions 65 Notes 66 Annex 4.A Obstacle Case Study--Access to Finance 67 5 Retail Sector: Facilitating the Emergence of Large Players to Boost Productivity and Employment 69 Two Sharply Contrasting Segments: Small Stores and Large Chains 69 Sector Performance: Low Productivity and Modest Employment Growth 70 Government Reform Efforts to Improve the Investment Climate 73 Effect of the Investment Climate on Modern-Format Stores 74 Obstacles to Increased Productivity and Employment Generation for Modern-Format Stores 77 Conclusions 79 Notes 79 Annex 5.A Obstacle Case Study--Taxes 81 Annex 5.B Obstacle Case Study--Access to Land 83 6 Software and ITES Sector: Productivity Improvements Key to Continued Growth and Competitiveness 87 The Software and ITES Sector: An Indian Success Story 87 Favorable Conditions Underpinning the Software and ITES Success Story 89 The Potential for Greater Productivity 91 The Importance of Productivity Gains as Favorable Conditions Erode 93 Conclusions 96 Notes 97 Annex 6.A Obstacle Case Study--Skill Shortages 98 7 Reducing Inequalities between States by Improving the Investment Climate 101 Importance of Measuring the Investment Climate at the State Level 101 The ICI: Methodology 102 ICI State Rankings 102 Identifying the Key Investment Climate Differences between Low-Growth and High-Growth States 107 Conclusions 109 Notes 109 vi Contents 8 Conclusions 111 Power 111 Access to Land 112 Access to Finance 112 Taxes 112 Skilled Labor 113 Three Challenges 113 Appendix A Investment Climate Surveys 115 Manufacturing ICS 115 Unorganized Manufacturing ICS 117 Retail ICS 119 Software/ITES ICS 121 Notes 123 Appendix B Econometric Analysis of the Effect of Investment Climate on Firm Performance 125 Manufacturing Sector 129 Unorganized Manufacturing Sector 135 Retail Sector 137 Software and Information Technology­Enabled Services Sector 140 Notes 145 Appendix C Investment Climate Index Methodology 147 Note 149 References 151 Index 155 vii Contents Boxes 2.1 Reforms and TFP during the 1990s 18 2.2 International Comparisons of Income Gaps between States 25 4.1 What Constitutes Unorganized Manufacturing in India? 54 4.2 Understanding the Problem of Access to Finance 61 5.1 Effect of Investment Climate on Labor Productivity and Employment Generation for Traditional Retail Stores 74 6.1 Comparing Perceived Obstacles for Software and ITES Exporters and Manufacturing Exporters 89 7.1 ICI Reliability 103 7.2 Comparing the ICI with the World Bank's Doing Business Study and the Government of India's Infrastructure Quality Index 105 C.1 Formulas Used to Construct the ICI 149 Figures 2.1 Real GDP Growth, by Sector, in India, 1990/91­2007/08 12 2.2 Structural Change in India, 1950/51­2005/06 13 2.3 Real GDP Composition, 2007/08 13 2.4 Savings and Investments, 1950/51­2005/06 14 2.5 Savings as a Percentage of GDP, 1965/66­2005/06 15 2.6 Trends in Consolidated Fiscal Deficit and Public Debt Stock, 1991/92­2006/07 16 2.7 Formal and Informal Labor Force, 1990 and 2003 19 2.8 Evolution of the Labor Force Distribution, by Sector, 1983/84­2004/05 20 2.9 Productivity Growth and Labor Reallocation in India, 1983­2002 22 2.10 States' Growth in Per Capita Income, 1980s and 1990s 23 2.11 Decomposition of Differences in Per Capita Income Growth across Groups of States, 1983­2004 24 2.12 Relative Sector Productivities and Potential 26 2.13 Labor Force Distribution, 2004 26 3.1 Proportion of Firms Reporting Each Obstacle as Major or Severe, 2003 and 2006 32 3.2 Firms Perception of Investment Climate Variables as the Most Important Obstacle or as a Major or Very Severe Obstacle, 2006 33 3.3 Impact of Corruption, by Firm Size, 2006 34 3.4 Percentage of Firms Rating Labor Regulations as a Major or Very Severe Obstacle, 2006 35 3.5 Proportional Impact of Investment Climate Variables on Firm TFP, Employment, Real Wages, Exports, and FDI, 2006 36 3.6 Deviation from the National Average of Percentage of Firms Reporting Each Obstacle as Major or Very Severe in 2006, by Income Level of States, 2006 38 3.7 Change in the Percentage of Firms Identifying Factor as a Major or Very Severe Obstacle between 2003 and 2006, by Income Level of States 39 viii Contents 3.8 Trade Competitiveness: Growth of Indian Exports and Worldwide Exports for the Sectors with the Largest Share of Total Indian Exports, 2001/02­2006/07 42 3.9 Importance of Largest Manufacturing Export Sectors in the Indian Export Basket and Growth of Sectors' Market Share in the World Export Market, 2001/02­2006/07 43 4.1 Per Capita Manufacturing Output, Organized and Unorganized Manufacturing, 2004 55 4.2 Management Qualifications and Experience, by Industry, 2006 57 4.3 Full-Time and Part-Time Workforce, by Industry, 2006 58 4.4 Surveyed Firms Reporting Themselves as Registered, 2006 59 4.5 Enterprises Reporting Employment Growth, by Age, 2006 60 4.6 Major Constraints to Normal Functioning Identified by Unorganized Enterprises 63 4.7 Effect of Power Outages on Unorganized Firms 64 5.1 Growth Rate of Selected Services Sectors, 1990/01­2006/07 71 5.2 Labor Productivity in Retail Relative to the United States, 2001 71 5.3 Overall Productivity in Selected Segments of the Indian Services Sector Relative to the U.S. Sector, 2001 72 5.4 Olley and Pakes Decomposition of Aggregate Labor Productivity for the Retail Segments and for the Retail Sector, 2006 72 5.5 Perceived Major or Very Severe Investment Climate Obstacles, 2006 75 5.6 Differing Cost Structures for Traditional and Modern-Format Food Retail Stores 76 5.7 Perceived Most Serious Investment Climate Obstacles for Modern Format Store, 2006 78 5.A.1 Tax Payments in India and OECD Countries: Number of Payments, Time Needed, and Tax Rates, 2006 82 5.B.1 Cost and Rent of Land Relative to GDP per Capita, Various Years 84 5.B.2 Cost and Time of Registering Property across India 85 6.1 Distribution of Software and ITES Firms and Contribution to the Overall Sector Revenue, 2006 88 6.2 Labor Productivity Growth across Services Sectors, 1982­90 and 1992­2002 91 6.3 Overall Labor Productivity in Software Sector Indexed to U.S. Average, 1998 92 6.4 Most Important Perceived Obstacle, 2006 94 6.5 Average Number of Weeks to Fill Position, 2006 95 6.6 Average Annual Percentage Change in Average Wages by Type of Skills, 2003­06 95 7.1 ICI Ranking of Business Climate in 16 Indian States, 2006 106 7.2 Relative Effect of Infrastructure Indicators on the Overall ICI and on Low-Investment and Low-Growth States 108 7.3 Relative Impact of Institution Indicators on Overall ICI and on Low-Investment and Low-Growth States 109 ix Contents Tables 1.1 Main Investment Climate Obstacles, Identified and Ranked by Sector 9 2.1 India Total Factor Productivity Estimates 17 2.2 Labor Productivity and Employment Shares 21 3.1 Contribution of Manufacturing to GDP in Selected Countries and Regions, 2005 30 3.2 Top 15 Manufacturing Exports, 2006/07 41 3.3 Value Added per Unit Labor Cost 44 3.4 Indirect Costs for Exporting and Nonexporting Firms in India and China 44 6.1 Corruption for Exporters and Nonexporters, 2006 96 8.1 Main Investment Climate Obstacles, Identified and Ranked by Sector 112 A.1 Allocation of Manufacturing ICS 2006 Sample: Bihar, Jharkhand, Orissa, and Rajasthan 116 A.2 Sample Allocation and Actual Coverage of Manufacturing ICS 2006, by State 117 A.3 Industry Sectors Identified for the Unorganized ICS 2006 Compared with Industry Sectors Covered by National Sample Survey's 56th Round 118 A.4 Sample Allocation and Actual Coverage of Unorganized Manufacturing ICS 2006, by City and Manufacturing Subsectors 120 A.5 Sample Allocation, Actual Coverage, and Distribution of Retail ICS 2006, by Sector 121 A.6 Actual Coverage and Distribution of Software/ITES ICS 2006, by Subsector 122 A.7 Actual Coverage and Distribution of Software/ITES ICS 2006, by City 122 A.8 Actual Coverage and Distribution of Software/ITES ICS 2006, by State 122 B.1 Definitions of Firm Performance Variables 126 B.2 Manufacturing Investment Climate Elasticities and Semielasticities with Respect to Productivity, Unrestricted Estimation 127 B.3 Manufacturing Two-Stage Least Squares Estimation of Employment Equation 132 B.4 Manufacturing Two-Stage Least Squares Estimation of Real Wages Equation 133 B.5 Manufacturing Two-Stage Least Squares Estimation of Probability of Exporting Equation 134 B.6 Manufacturing Two-Stage Least Squares Estimation of Probability of Receiving FDI 135 B.7 Unorganized Sector Investment Climate Elasticities with Respect to Log of Labor Productivity 137 B.8 Unorganized Sector Investment Climate Elasticities with Respect to Employment Demand 138 B.9 Retail Investment Climate Elasticities and Semielasticities with Respect to Log of Labor Productivity 139 B.10 Software and ITES Investment Climate Unrestricted Elasticities and x Semielasticities with Respect to Productivity 140 Contents B.11 Software and ITES Two-Stage Least Squares Estimation of Employment Equation 141 B.12 Software and ITES Two-Stage Least Squares Estimation of Real Wages Equation 142 B.13 Software and ITES Two-Stage Least Squares Estimation of Probability of Exporting Equation 143 B.14 Software and ITES Two-Stage Least Squares Estimation of Probability of Receiving FDI 144 C.1 Variables Used in the Construction of the Composite ICI 148 xi Foreword India has grown impressively in recent years and is emerging as a major world eco- nomic power. The economy has surged over the last decade, posting an average annual GDP growth of 8.5 percent since 2005, after two relatively slow decades in the 1980s and the 1990s when growth averaged about 4-5 percent a year. Greater invest- ment, increased productivity, and deeper integration in the world economy have been the key growth drivers. The current challenge is to maintain medium-term growth potential and provide benefits, as well as economic opportunities, to ever larger segments of the population. Strong and more inclusive growth should form the basis for India's economic sta- bility going forward, and will be critical for garnering political support for growth- oriented reforms, given concerns over widening inequality. Nonetheless, three main challenges remain: insufficient job creation in the formal sector; large and growing disparities between states; and increasing, but still low, productivity. The current global financial crises, and downward revisions of India's growth forecast, only high- light the need to bolster internal reform efforts in order to increase the resilience of the economy. This book identifies key investment climate bottlenecks that slow down growth and poverty reduction. Based on face-to-face surveys of owners and managers of firms, combined with extensive dataset analysis, backed by secondary sources, the study analyses the critical factors that influence day-to-day decisions by firms on how to invest. As a result, it identifies growth-enabling reforms that cover macroeco- nomic policies, governance, institutions, and infrastructure. To get an understanding of investment climate constraints for the entire economy, the book focuses on four key sectors. The manufacturing sector, both organized and unorganized, is important in that it not only contributes a significant share of the over- all GDP, but it is also critical for employment generation and to the growth of the"lag- ging" states. Similarly, the retail sector is an important part of the overall economy in its contribution to GDP and to employment. Finally, the ICT sector, though a small part of GDP, has been leading India's services success story and serves as a good example from which we may learn. The book presents illustrative case studies of the key con- straints identified,and the efforts to address them,along with policy recommendations. xiii Foreword Furthermore, the book identifies a number of issues that require further analysis; in those instances the book should be viewed as a first step to bringing the issues into the forefront. The World Bank Group stands ready to work with the government, private sector, and other stakeholders in taking forward this work to help generate real impact on the ground. Rachid Benmessaoud Ernesto May India Acting Country Director Sector Director, Poverty Reduction, South Asia Region Economic Management, Finance and Private Sector Development South Asia Region xiv Acknowledgments This report was prepared by a World Bank team led by Aurora Ferrari and Inderbir Singh Dhingra. Marina Wes was the main author of chapter 2, Inderbir Singh Dhingra and Ashish Narain of chapter 3, Ashish Narain of chapter 4, Aurora Ferrari of chapter 5, Aurora Ferrari and Suhail Kassim of chapter 6, and Giuseppe Iarossi of chapter 7. The team was composed also of Angelica Salvi, Alvaro Escribano, Taye Mengistae, Manuel de Orte Ramírez, Jorge Pena, Shanthi Divakaran, Pooja Churamani, and Giovanni Tanzillo. Background papers for this report were prepared by Evaluserve New Delhi, Alvaro Escribano and Manuel de Orte Ramírez, Taye Mengistae and Maddalena Honorati, Mahesh Uttamchandani and Raha Shahid-Saless, Rohit Mittal, Surbhi Goyal, Kwawu Mensan Gaba, and Ashish Khanna. The Investment Climate Surveys that underpin the analysis were conducted in cooperation with the Confederation of Indian Industries (manufacturing and infor- mation communications technology), ACNielsen India (retail), and Marketing & Development Research Associates (unorganized manufacturing). Administrative support was provided by Syed Hye and Vinod Satpathy. The report was edited by Linda Moll. Overall guidance for the project was provided by Simon Bell, Priya Basu, John Speakman, Ernesto May, Fayez Omar, and Isabel Guerrero. The team also benefited greatly from the extensive advice and guidance of the peer reviewers Mary Hallward- Driemeier, Lili Liu, and Vincent Palmade. In addition, the report incorporates suggestions and comments received from Ahmad Ahsan, Nagavalli Annamalai, Andreas Blom, Hooman Dabidian, Klaus Deininger, Mark Dutz, Rajni Khanna, Mani Khurana, Leora Klapper, William Mako, Jana Malinska, Varsha Marathe, Vikram Menon, Deepak Mishra, Rita Ramalho, Binyam Reja, Deepa Sankar, Stefka Slavova, Natalya Stankevich, Randeep Sudan, Zenaida Hernández Uriz, and Niraj Verma. Useful suggestions and materials provided by Asitava Sen from the consulting firm Technopak have also been incorporated. The report has also benefited from pre-survey consultations with the National Investment Commission, National Commission for Enterprises in the Unorganized Sector, and National Manufacturing Competitiveness Council of the government of India. xv Abbreviations AETR average effective tax rate APDRP Accelerated Power Development and Reform Program ASI Annual Surveys of Industry CIBIL Credit Information Bureau (India) Limited FDI foreign direct investment FMCG fast-moving consumer goods GDP gross domestic product ICA Investment Climate Assessment ICI Investment Climate Index ICOR incremental capital output ratio ICS Investment Climate Survey ICT information and communication technology IT information technology ITES information technology­enabled services METW marginal effective tax wedge NSSO National Sample Survey Organisation OECD Organisation for Economic Co-operation and Development OLS ordinary least squares R&D research and development SEB State Electricity Board SRFAESI Securitization and Reconstruction of Financial Assets and Enforcement of Security Interest Act 2002 TFP total factor productivity VAT value added tax xvi Chapter 1 Overview The Indian Economy: Robust Growth but Challenges Remain India has emerged in the past few years as one of the world's fastest growing economies. Gross national income per capita has nearly doubled since 2001/02, going from about US$400 to about US$800. Over the past two decades, the growth of the economy has increasingly been driven by the services sector and the industry sector. This trend has changed the structure of the economy. Currently, the services sector accounts for 53 percent of gross domestic product (GDP), industry for 29 per- cent, and agriculture for 18 percent. The key growth drivers remain the high level of investment and domestic private consumption, but increased productivity and a deeper integration in the global economy have also played a role. Despite considerable achievements, India will need to overcome three challenges to maintain growth: pressure on formal sector employment from demographics, rising inequalities between states, and future growth's increasing dependency on productivity gains. 1. Creating employment growth in the formal sector. Employment growth has not kept pace with GDP growth, and the composition of employment has been slow to change even as the economy has seen relatively significant structural changes. As a result, most new jobs have been created in the informal services sector, and agriculture still employs 57 percent of the population, although it contributes only 18 percent of GDP. Given the volatility and low growth of agriculture, the current employment composition has significant implications for income distri- bution. Jobs in the informal sector, which in India are called "unorganized" jobs, are less secure, give no right to social security, and are not under the purview of the labor laws. 1 Overview 1 2. Reducing inequalities between states. The disparities in the growth performance of Indian states are large and growing. If these differences were to be sustained over time, they could translate into vast differences in material well-being. A decom- position of the growth of per capita income reveals that labor productivity growth accounts for most of the differences in growth rates between richer and poorer states. This finding highlights the importance of the investment climate. 3. Accelerating productivity growth. Overall economic growth will depend on either (a) capital accumulation through greater investment or (b) productivity enhancements. However, private savings and investment are already at a record high, and public savings have benefited from an impressive fiscal consolidation process. Hence, productivity is key to sustaining current growth rates, especially as sector studies have revealed that the potential to increase productivity is high. This report seeks to show how investment climate improvements in key sectors will assist the Indian economy in overcoming these three challenges. The key sectors are organized and unorganized manufacturing, retail, and software and information technology­enabled services (ITES). The report focuses on external factors and does not include operational analysis of the sectors. Moreover, the analysis is based on datasets of new and existing firms and not of firms that exited the market. Manufacturing: Investment Climate Obstacles to Productivity Growth and Employment Generation Organized manufacturing is the largest subsector contributor to GDP after agricul- ture. However, its share as a percentage of GDP has been decreasing and is much lower than that of other developing countries, suggesting potential for growth. Moreover, unusual for a developing country, excess labor from agriculture is being absorbed more by the services sector--especially the informal services sector--than by manufacturing, which usually plays a bigger role in creating jobs for semiskilled labor. Productivity is important for the growth and the competitiveness of the manufacturing sector, but the sector is characterized by low productivity. Indian manufacturing performs poorly both compared with other sectors of the Indian economy--average total factor productivity (TFP) is 4.3 in the manufacturing sector and 35.6 in the software and ITES sector--and with other developing countries. (With a US$7,011 value added per worker in 2005, Indian manufacturing was well below that of China at US$13,182 and that of Turkey at US$10,186.) To understand what holds back the productivity of the manufacturing sector, the World Bank conducted two nationally representative surveys--the Manufacturing Investment Climate Surveys (ICSs)--the first in 2003 and the second in 2006. Overall, firms perceived the investment climate to be less of a problem in 2006 than in 2003. Perceptions worsened, however, with respect to the specific areas of electricity, taxes, corruption, and workforce skills. 2 Overview 1 Electricity shortages, tax rates and administration, and corruption were the biggest obstacles identified by manufacturing firms, being rated as the most impor- tant obstacle by 32, 23, and 10 percent of the respondents, respectively. Half of all manufacturing firms had a backup generator because of the poor supply of electric- ity, and 5 percent of annual sales were lost because of power outages. In terms of high tax rates and burdensome tax administration, data from Doing Business 2008 (World Bank 2007a) confirm that a representative limited liability company has to pay almost 71 percent of its profits as taxes--much more than in other developing coun- tries. Tax administration is also costly: a representative company needs to make 60 payments a year and will spend 271 hours doing so. Informal payments are wide- spread, and dealing with government officials is also time consuming: firms pay, on average, 4.9 percent in bribes, and managers spend, on average, 12.6 percent of an average week dealing with government officials. The econometric analysis supports the firm identification of investment climate obstacles: 76 percent of the explainable intrafirm variability in TFP can be explained by differences in investment climate. Red tape, corruption, and crime had the largest negative impact on firm productivity, real wages, and exports. It accounted for per- centages ranging from 16 to 28 percent. Poor infrastructure was the second-largest investment climate obstacle, affecting the most productive firms relatively more. It accounted for as much as 23 percent of explained intrafirm productivity differentials in the manufacturing sector. Labor skills, quality, and innovation have a smaller effect on productivity but are more important for efficient firms. Improving the labor quality available to enterprises and stimulating innovation has the potential to raise productivity by as much as 15.5 percent. There are large differences across states in the investment climate variables that firms find constraining. Firms in low-income states were more likely to complain about the power situation.Firms in middle-income states,in contrast,were more likely to complain about access to land, access to finance, labor regulations, and worker skills. In high-income states, relative to the national average, the major issues were corrup- tion and the possibly closely related area of high taxes and tax administration. The dif- ferences are most pronounced in areas that are more directly affected by state-level policies, such as power, in that firms in low-income states complained more than national average and those in high-income states complained less. When comparing firms' perceptions and objective indicators for 2003 and 2006, one finds that the investment climate at the state level has improved for middle- income states on all fronts. The investment climate has worsened moderately for high-income states, except for corruption, transport, and licensing, and it has wors- ened substantially, on all fronts, for low-income states, especially with respect to power, tax administration, corruption, and transport. India's export performance in manufactured goods has been modest compared with the growth of overall exports, which instead have been dominated by the serv- ices sector. Moreover, the majority of India's exports are concentrated in lower- growth segments worldwide, and the sectors in which India has the largest share of world exports are losing market share worldwide or improving only moderately. 3 Overview 1 When the competitiveness of the manufacturing sector of India is compared with that of China by measuring the value added per unit labor cost, China is found to be more competitive than India both in the sectors in which India is gaining world mar- ket share and in those in which India is not. The poor performance of the Indian manufacturing sector can partly be explained by the very high indirect cost that Indian firms have to face--much higher, for example, than Chinese enterprises. Unorganized Manufacturing: Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity Unorganized manufacturing employs the vast majority of India's manufacturing workforce. In 2000/01, almost 99 percent of manufacturing enterprises and 74 per- cent of the manufacturing workers were in the unorganized sector. In addition, unorganized manufacturing is more evenly distributed across low-, middle-, and high-income states. To understand what hampers the productivity of the sector and whether these firms avoid becoming organized and growing, the World Bank conducted the Unorganized Manufacturing ICS in 2006 in five major industrial clusters. There is significant scope for improving productivity and wages in unorganized manufacturing, where firms are characterized by low productivity relative to organ- ized manufacturing. Differences in performances between organized and unorgan- ized manufacturing are to be expected, but in India they are large, on the order of 8 to 1, compared with 3 to 1 in Japan before its era of post­World War II growth or in the Republic of Korea during its early stages of development. It is estimated that, after human capital attributes are controlled for, workers in larger enterprises earn three times more than similar workers in the unorganized sector. Unorganized manufacturing firms are not integrated into the supply chain, thus limiting the transfer of technology. Possibly because of the legacy of the small-scale reservation policy, much of the manufacturing activity in the unorganized sector is geared toward producing final products for the consumer market, rather than inter- mediate products and parts for the organized sector. Being part of the organized sector--often referred to as the formal sector-- increases a firm's bargaining power (for example, the firm has easier access to finance). However, international experience suggests that despite the advantages, unorganized enterprises often wish to remain informal because organized firms are subject to more regulations. In India the 10-worker threshold is especially important because labor laws on wages and benefits apply to units above this size. Firms can avoid being part of the organized sector in two ways: (a) by operating "under the radar" by simply not registering or (b) by not growing. Nonetheless, fear of formalization has not prevented firms from growing. A large number of unorganized manufacturing firms were registered with some government authority and had declared most workers and sales for tax purposes.A relatively large number also crossed the threshold into organized production; around 46 percent of enterprises interviewed in the context of the (formal) Manufacturing ICS started off 4 Overview 1 as unorganized manufacturing firms. Unorganized firms stand to benefit in terms of productivity and bargaining power from raising their scale, and investment climate variables were identified as obstacles to growth in the sector. Access to finance was identified as the single biggest obstacle to firm growth. Although most firms had a bank account, very few had a loan. Getting a loan from a bank took almost a month on average after all the documents had been submitted. Moreover, it required significant immovable assets, whereas unorganized firms had mainly movable assets. Firms that were able to borrow from formal financial insti- tutions had 37 percent higher labor productivity than firms that were not. Borrow- ing from informal sources is not strongly associated with any similar positive effects, maybe because the bulk of informal loans are short term and are used for financ- ing working capital and maintaining operations rather than for making fixed investments. Power availability is identified as the second-biggest constraint for unorganized manufacturing. Nonetheless, power outages appear to have no effect on productiv- ity, possibly because they push enterprises to develop coping strategies based on employing more labor. Retail Sector: Facilitating the Emergence of Large Players to Boost Productivity and Employment The retail sector is the fourth-largest contributor to GDP, after agriculture, manu- facturing, and the financial sector, and its share is growing. Moreover, the structure of the retail sector affects the performance of the domestic supplying industries and agricultural producers. Finally, given the limited capital, land, and skills required, small-scale retailing is often the default employer for people who cannot find employment in the formal manufacturing sector. The retail sector is made up of two sharply contrasting segments: the large chain outlets and a multiplicity of small traditional retail stores. Despite fast growth among large chain outlets, traditional retail stores still dominate and accounted for 97 per- cent of total retail in 2005. Although the retail sector has been growing, its growth has been modest relative to the whole services sector. Employment growth has also been limited. Moreover, the retail sector in India has been characterized by low labor productivity when com- pared with the retail sector in other countries. To understand the sector, the World Bank conducted a nationally representative survey in 2006, the Retail ICS. The survey data show that low labor productivity is attributable to the large pres- ence of small outlets, which have a labor productivity that is much lower than that of modern-format stores. The labor productivity decomposition confirms this find- ing by showing that the most productive outlets in India are the single stores with the largest market share. Employment data for the past three years show that modern-format stores also create more jobs (albeit from a very low base) than do small outlets, and the growth of modern retailers has not had an adverse impact on 5 Overview 1 the latter in the medium term. To increase productivity and promote employment generation in the retail sector, policy makers need to facilitate the growth of the largest market players. In recent years, central and state governments have adopted a number of policies to facilitate the growth of the retail sector--particularly that of the modern-format stores. Despite these efforts, the investment climate is a bigger problem for large retailers than for small ones. This outcome is mainly the result of two facts. First, the existing legal framework is more favorable to small stores. For example, the cost structure analysis for the two segments revealed that small retailers are charged lower (residential) rates for key inputs, such as power and land. Hence, although the cost of goods is lower for large chains, the profit margins of small retailers are higher. Sec- ond, even when the legal framework is the same, small stores can operate more eas- ily outside the framework. The labor productivity analysis supports this finding by showing that the investment climate and control variables explain 44 percent of the variability of labor productivity for large chains versus 22 percent for small retailers. Electricity shortages, corruption, tax rates, and lack of access to land are the biggest obstacles to large chain outlets. Power shortages occur daily, and even though these stores have generators, the suffer 3.1 percent loss of sales. Although informal payments are small (less than 1 percent of total annual sales), dealing with govern- ment officials is time consuming: on average, retail managers spend 5 percent of their time dealing with officials. The issue of taxes needs to be analyzed in terms of the uneven playing field between organized and unorganized retail segments. This discrepancy is caused by different tax rates applicable to the two segments and by a complex tax system that, although improving, still penalizes large stores that operate across different states. Finally, only half of the stores that tried to expand were suc- cessful, mainly because of land unavailability. The perceptions of the managers are supported by productivity analysis: infrastructure problems and problems with red tape and corruption each accounted for 15 percent of the impact of the investment climate on the average labor productivity for modern-format stores. Software and ITES Sector: Productivity Improvements Key to Continued Growth and Competitiveness Software and ITES has been the fastest growing sector and is the largest contributor to exports. In a sense, the sector represents the success story of the liberalization of the 1990s. Hence, an understanding of what has supported this success is crucial to supporting enhanced productivity in the other sectors. To understand the sector's success and the future challenges, the World Bank conducted a national survey, the Software/ITES ICS 2006. The success of the sector is based on a rare set of market conditions: a favorable domestic investment climate and a large demand from Western companies seeking to reduce costs, although their computer systems still required labor-intensive sup- port. The favorable investment climate translated into cost advantages. India had 6 Overview 1 relatively large numbers of English-speaking engineers, who were less expensive than their Western counterparts. The sector had been almost completely liberalized, allowing for technology transfer. Moreover, companies located in software technol- ogy parks enjoyed a full tax holiday on profits from exports. Finally, unlike the man- ufacturing industry, software and ITES were not subject to stringent labor laws, and because of its features, the sector was less reliant on infrastructure. However, these favorable conditions have allowed productivity in the software and ITES sector to lag its potential. Part of the explanation for lagging productivity can be traced to operational factors, such as individual organization of functions within firms, product mix, and lack of brand name. There have also been external barriers to productivity growth, such as lack of a sophisticated end-user market and low wages. The large supply of English-speaking information technology engineers has, until recently, kept wages in the sector relatively low compared with the world average. This trend has allowed India to compete effectively in the outsourced serv- ices market, but it has also meant that most companies continued to focus on these low-value segments using fairly basic processing engineers. Firms have tended to respond to increased demand for their products by their increasing staffs. Conse- quently, productivity gains, which are harder to achieve, were not essential to main- taining growth. Changing conditions, on both the supply and the demand sides, have the poten- tial to hurt India's competitiveness. On the supply side, the shortage of trained professionals is beginning to bite and can be seen in the wage data. Moreover, the benefits conferred by the tax breaks for companies located in software technology parks are due to expire. All these factors are contributing to a decrease in profits. On the demand side, given the weight of the U.S. market in the firms' exports, a medium-term slowdown of the U.S. economy is likely to adversely affect the Indian software and ITES sector. As cost advantages are being eroded, productivity gains will become more impor- tant. Higher wages will erode the sector's competitiveness in the low-value segment of the world software and ITES market, which has been the focus of Indian firms. This development highlights the importance of increasing productivity so that firms can operate in high-value segments. The investment climate has a significant effect on the potential productivity growth of firms in the sector. The productivity analysis shows that, for the overall software and ITES sector, investment climate and control variables explain 27 per- cent of the variation in TFP. Limited skills in the workforce, electricity shortages, and corruption are the biggest obstacles for exporters, whereas corruption, electricity shortages, and access to land are the biggest obstacles for nonexporters. The perceptions of the managers are supported by productivity analysis: infrastructure accounts for 38 percent of the variations in intrafirm TFP; quality, innovation, and labor force skills account for 23 percent; and red tape and corruption account for 20 percent. The workforce shortage is likely to become more acute as India further positions itself as major player in the high-value segment. In the past three years, a shortage of 7 Overview 1 skilled labor has caused wages to increase by significantly more than nominal GDP. Regarding corruption, though informal payments are small (2.2 percent of sales), the time managers spend dealing with government officials is substantial (8.1 percent of an average week). As far as power is concerned, a third of the firms rely on a power backup source (with a much higher incidence among exporters), and 18 percent of all energy requirements of the sector are provided by generators. Of the 30 percent of exporters and 16 percent of nonexporters who tried to acquire land in the three years before the survey, only 65 and 61 percent of them, respectively, were successful. Unavailability of land, unclear ownership, and difficulties involved in registering or obtaining permits were given as significant reasons. Reducing Inequalities between States by Improving the Investment Climate Investment climate reforms by Indian states have yielded tangible results. Liberaliza- tion policies of the early 1990s have allowed states to play a larger role in determin- ing their development path. As a result, growth accelerated in all the states, but at a faster pace in middle-income states. This trend has resulted in increasing divergence between low-income states on the one hand and middle- and high-income states on the other. In economic matters, states have substantial autonomy over the reform agenda. Broadly speaking, land and licensing matters are within the realm of state jurisdic- tions. Infrastructure, labor laws, and vocational education are part of the joint jurisdiction. Telecommunications, the financial sector, customs, environmental reg- ulations, and judicial reforms are under the purview of the union. States can undertake only so many reforms at a given time; hence, it is crucial to understand which aspects of the investment climate matter most. A comprehensive Investment Climate Index (ICI) was created to help pinpoint the crucial obstacles. The ICI was created as a composite indicator of microlevel investment climate vari- ables drawn from the ICS datasets. These variables measure both the cost and the perception of the investment climate as experienced by small and medium manu- facturing and retail entrepreneurs operating in the state. The ICI compared the 16 states on the basis of their investment climate features and identified three groups of states: best performers (with lower value in the index), average performers, and worst performers. According to the ICI ranking, the states with the best investment climate are Karnataka and Kerala. The average performers are Gujarat, Andhra Pradesh, Haryana, West Bengal, Maharashtra, and Delhi. The worst investment climate is found in Bihar, Uttar Pradesh, and Rajasthan. The decomposition of the ICI also shows that the main drivers of a better busi- ness environment in India are represented by variables associated with infrastructure and institutions. Within infrastructure, power outages emerge as the single biggest constraint, followed by transport problems. Corruption, tax regulations and admin- istration, and security are the biggest institutional constraints. The impact of each 8 Overview 1 investment climate variable varies by type of state; infrastructure matters most in low-growth and low-investment states, whereas institutions matter equally for low- growth and low-investment states and for high-growth and investment states. Conclusions The analysis of manufacturing, retail, and software and ITES firm-level data identi- fied power shortages, lack of access to land and finance, taxes, and skilled labor shortages as the major investment climate obstacles to increased productivity and employment (see table 1.1). These obstacles have a direct bearing on the three challenges facing the Indian economy: 1. Formal sector employment would be greatly enhanced by removing obstacles to productivity and growth in the unorganized manufacturing sector, which employs the vast majority of India's workforce. The analysis suggests that these firms would be willing to become formal if growth were facilitated. The greatest obstacles in this respect are lack of access to finance and unreliable power supply. Moreover, facilitating the emergence of larger retail stores by tackling access to power and land, reducing corruption, and leveling the playing field with respect to taxes will accelerate job creation in the formal sector. 2. The rising inequalities between states can be traced to the positive effects of liber- alization and investment climate improvement in the fastest-growing states, which have attracted a higher concentration of productive firms. Mirroring these reforms in backward states will be an important step in reducing the inequality. The decomposition of the state-level ICI confirms that variables associated with infrastructure (power, in particular) are key drivers of a better investment climate in India. The analysis also shows that the effect of investment climate variables varies by type of state; infrastructure is particularly important in low-growth states. Moreover, facilitating the formalization of unorganized manufacturing, which is present in most regions, will help reduce disparities. Table 1.1 Main Investment Climate Obstacles, Identified and Ranked by Sector Obstacles Tax rates and Skilled Access to Access to Sectors Power administration Corruption labor land finance Organized manufacturing 1 2 3 Unorganized manufacturing 2 1 Retail 2 3 1 Software and ITES 2 3 1 Sources: Manufacturing ICS 2006; Unorganized Manufacturing ICS 2006; Retail ICS 2006; Software/ITES ICS 2006. 9 Overview 1 3. Productivity gains among small, unorganized firms are limited because of the firms' inability to access finance for productivity-enhancing investment. In the software and ITES sector, a shortage of skilled labor and infrastructure are the greatest barriers to moving up the value chain and enhancing productivity. For formal manufacturing firms, red tape, tax and corruption, followed by infra- structure, had the biggest impact on TFP, accounting for 27 percent and 23 per- cent of the observed variability respectively. 10 Chapter 2 The Indian Economy Robust Growth, but Challenges Remain Trends in the Economy India has emerged in the past few years as one of the world's fastest-growing economies. Gross national income per capita has nearly doubled since 2001/02, going from about US$400 to about US$800 (measured using the Atlas method). India has transformed from the world's 50th-ranked economy in nominal U.S. dol- lar terms in 1980 to the 10-largest economy today. In terms of purchasing power par- ity, the Indian economy occupies fourth place in the world, after the United States, Japan, and the Russian Federation. Real gross domestic product (GDP) growth for 2006/07 was 9.7 percent, and the advance estimate for 2007/08 is 8.7 percent (fig- ure 2.1). Various observers (for example, Poddar and Yi 2007) estimate that India's potential growth rate is now 8 percent, following a rate of 5 to 6 percent during the previous two decades. A number of important characteristics of India's rapid growth are discussed in this chapter. The growth of the Indian economy has increasingly been driven by the services and industry sector over the past two decades. Between 2002 and 2006, the real growth of the services and industry sectors averaged around 10 percent, higher than real GDP growth of 9 percent. In the 1990s the services sector grew at an average annual rate of 9 percent, contributing nearly 60 percent of the overall growth of the economy. This growth was led by a continued uptrend in telecommunications sub- scribers, financial services, tourism, software and information technology­enabled services (ITES), and freight traffic. The services sector now accounts for more than half of total value added and GDP in India, well ahead of China and of most other countries with a similar income level. Industrial growth has also accelerated and is now around 10 percent. 11 Robust Growth, but Challenges Remain 2 Figure 2.1 Real GDP Growth, by Sector, in India, 1990/91­2007/08 45,000 40,000 35,000 30,000 software and ITES rupees) 25,000 (billion 20,000 trade GDP 15,000 other services real 10,000 manufacturing other industry 5,000 agriculture, forestry, and fishing /04 /07 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 2000/01 2001/02 2002/03 2003 2004/05 2005/06 2006 2007/08 1999/2000 fiscal year Source: Authors' calculations based on Central Statistical Office 2007. Industry's growth has been led by positive trends across a diverse range of sectors and by the emergence of many new capacities. Following the slowdown of the late 1990s--and especially since 2003--manufacturing growth has resurged within the industry sector, supported by domestic as well as export demand, strong corporate profitability, and resilient business confidence. From 2004 to 2005, manufacturing has posted a growth of more than 10 percent, a substantial improvement from the growth of less than 5 percent in the 1990s. Although the services sector growth rate has been most stable, industrial growth has also become significantly less volatile in recent years. The industry and the services sectors now account for more than 80 percent of GDP. The growth of services and industry has changed the structure of the economy (figure 2.2). Advanced estimates for 2007/08 show that services (including trade and software and ITES) account for around 53 percent of GDP (figure 2.3). Agriculture is expected to account for 18 percent of GDP in 2007/08; industry (including man- ufacturing) is expected to account for 29 percent. As the importance of manufactur- ing and services increases and foreign demand becomes more significant, the growth of the Indian economy depends less and less on the performance of the agricultural sector (see figure 2.2). The agricultural sector, though, cannot be ignored because it still employs more than half of the labor force. Key Growth Drivers in the Economy India's overall GDP growth has so far been resilient--with falling volatility--in the face of higher oil prices, rising interest rates, and slowing U.S. growth. The key 12 The Indian Economy 2 Figure 2.2 Structural Change in India, 1950/51­2005/06 60 50 40 GDP of 30 percentage 20 10 0 1950/51 1955/56 1960/61 1965/66 1970/71 1975/76 1980/81 1985/86 1990/91 1995/96 2000/01 2005/06 fiscal year agriculture industry services Source: Authors' calculations based on Central Statistical Office 2007. Figure 2.3 Real GDP Composition, 2007/08 organized unorganized manufacturing, manufacturing, 11% 5% other industry, 13% other services, 38% agriculture, forestry, and fishing, 18% software trade, and ITES, 10% 5% Source: Authors' calculations based on Central Statistical Office 2007. 13 Robust Growth, but Challenges Remain 2 Figure 2.4 Savings and Investments, 1950/51­2005/06 40 35 30 GDP 25 of 20 15 percentage 10 5 0 1950/51 1955/56 1960/61 1965/66 1970/71 1975/76 1980/81 1985/86 1990/91 1995/96 2000/01 2005/06 fiscal year gross domestic savings gross domestic capital formation Source: Authors' calculations based on Central Statistical Office 2007. growth drivers remain the high level of investment and domestic private consump- tion, but increased productivity and a deeper integration in the global economy have also played a role. Capital resources committed to India's growth have been on the rise because investments as a share of GDP have increased continually from 2001/02. Following a period of corporate restructuring, investment has boomed on the back of buoyant corporate profits; profitability has recorded growth in the 25 to 60 percent range on a year-to-year basis since 2002/03. The investment rate increased to 33.8 percent of GDP in 2005/06, up from 22.9 percent in 2001/02 (figure 2.4). Sectorally, infrastruc- ture is dominating the scene, with its share in total capital investment being about one-third in recent years. Investment in the manufacturing sector remains strong, and the share of services sector industries in total capital expansion is increasing and accounted for about 15 percent in 2005/06. Investments continue to be driven by the private and household sectors, which now account for 24 percent of GDP, whereas public sector investments have stayed roughly constant at 7 percent of GDP. Investments have been supported by domestic savings, which are estimated at 32.4 percent of GDP in 2005/06, up from 23.4 percent in 2001/02. This increase is explained largely by the buoyancy of private corporate sector savings (which rose from 3.4 percent in 2001/02 to 7.8 percent in 2006/07). The increase in corporate sector savings is linked to sustained improvements in corporate profitability and 14 The Indian Economy 2 Figure 2.5 Savings as a Percentage of GDP, 1965/66­2005/06 30 25 20 GDP of 15 10 percentage 5 0 5 1965/66 1970/71 1975/76 1980/81 1985/86 1990/91 1995/96 2000/01 2005/06 fiscal year household savings private corporate sector savings government savings Source: Authors' calculations based on Central Statistical Office 2007. internally generated resources. Household savings, which were already among the highest in the world, have also grown during the past decade, and savings in bank deposits are rising rapidly; they are estimated to account for nearly 50 percent of all savings in 2006/07, up from 37 percent in 2005/06 because of higher interest rates and the deepening of financial markets (figure 2.5). Investments in equities as a share of financial assets are also rising, although ample room remains for further increases. A turnaround in public savings (reflecting the fiscal consolidation process) has also played an important role, with a reduction of the budget deficit from 9.5 percent to 6.2 percent of GDP between 1999 and 2007. India's fiscal position remains tight but has shown solid improvements. Although at 6 percent of GDP (2006/07 estimates), the consolidated fiscal deficit1 is well below its peak of 10 percent (figure 2.6), the debt overhang implies that one-third of revenue is diverted to debt servicing. Fiscal con- solidation appears to have taken hold in India through both revenue enhancements and selected expenditure restructuring--at both the center and the state levels--and the Fiscal Responsibility and Budget Management Act targets are on track. Monetary policy has also supported growth, and since mid 2006, all three major credit-rating agencies have upgraded India to investment grade. Indicative analysis suggests that the acceleration in economic growth in India is increasingly owed to greater productivity growth, rather than just greater use of inputs. There is evidence of a turnaround in productivity levels in manufacturing 15 Robust Growth, but Challenges Remain 2 Figure 2.6 Trends in Consolidated Fiscal Deficit and Public Debt Stock, 1991/92­2006/07 90 12 85 10fiscal GDP) deficit of 80 8 % a (as 75 6 (as a % of stock 70 4 GDP) debt 65 2 60 0 1991/921992/931993/941994/951995/961996/971997/981998/99 2000/012001/022002/032003/042004/052005/062006/07 1999/2000 fiscal year debt stock fiscal deficit Sources: Data from the Reserve Bank of India, World Bank Live Database, and Ministry of Finance of India. Note: Consolidated fiscal deficit combines states and center budget financing. Data for 2006/07 are estimated. since 2003. This turnaround is related to the increased efficiency of private sector firms in the face of heightened competition, which has been crucial in raising India's underlying trend growth rate (see International Monetary Fund 2006). Bosworth, Collins, and Virmani (2007) suggest that total factor productivity (TFP) accounted for the bulk of the increase in output per worker in India during 1980 to 2000, higher than in all other regions of the world except China. From a comparative perspective, India has enjoyed better growth in output per worker than have many parts of the world in recent decades. Other studies, summarized in table 2.1 and in box 2.1, also suggest that productivity increases are playing an increasing role in driving India's growth. Despite recent improvements, productivity in many sectors in India is still low compared with other countries and with viable best-practice potential. Goldman Sachs (2007) suggests that the turnaround in manufacturing produc- tivity has been central to the ratcheting up of productivity growth and that the increased efficiency of the private sector was the principal driver of this turnaround in the face of increased competition and a decade of reforms.2 There is also evidence of a steadily declining ratio of net capital stock to value added in industry and of a widening and deepening of the financial sector. Growth in India has been less capital intensive than in many other countries, partially because it has been led by the services sector, which relies more heavily on labor inputs and less on capital. The productivity of capital use, as measured by the 16 The Indian Economy 2 Table 2.1 India Total Factor Productivity Estimates Source Annual average growth (%) Acharya and 1950/51­ 1967/68­ 1981/82­ 1991/92­ others 2003 1966/67 1980/81 1990/91 1999/2000 GDP 3.8 3.4 5.3 6.5 TFP 1.4 0.7 2.0 2.6 Proportion of GDP growth explained by TFP 38 21 38 40 Virmani 2004 1950­80 1980­2003 Net domestic product per worker 1.3 2.4 TFP 0.7 2.4 Proportion of per worker net domestic product growth explained by TFP 54 69 Bosworth and Collins 2003 1960­70 1970­80 1980­90 1990­99 GDP per worker 1.9 0.7 3.9 3.3 TFP 0.7 0.5 2.5 1.6 Proportion of per worker GDP growth explained by TFP 40 negative 64 48 Rodrik and Subramanian 2004 1960­70 1970­80 1980­90 1990­99 GDP per worker 1.8 0.9 3.7 3.3 TFP 1.2 0.5 2.9 2.4 Proportion of per worker GDP growth explained by TFP 67 56 78 73 Sources: Ahmad 2006 and other sources as listed in table. incremental capital output ratio (ICOR), has been high by Asian standards but is now also showing a downward trend, which reflects an improvement in the produc- tivity of capital.3 India is benefiting from its increasing integration into the global economy. The share of exports and imports4 in GDP increased from 17 percent in 1990/91 to around 50 percent in 2006/07. Trade barriers have come down significantly, and Indian exports have increased rapidly over the past 10 years. Exports in India have been growing strongly since 2000/01, posting a 19 percent growth in 2005/06.5 Over the same period, the composition of export has also changed; although growing in absolute terms, the share of manufacturing goods in total exports has been declin- ing, from 56 percent to 44 percent because of the rise in service exports. Despite increased integration into the global economy, the share of exports and imports in India's GDP today is only 2.5 percent, compared with China's 10.5 percent. In the future, provided that some of the constraints in this report are addressed, India is well placed to be a much more significant presence in the world market for exports. Capital flows to India have increased very rapidly, especially in the past two years. Inward foreign direct investment--channeled mainly into financial services, manufacturing, banking and information technology (IT) services, and 17 Robust Growth, but Challenges Remain 2 Box 2.1 Reforms and TFP during the 1990s Several studies have found that the effect that the market opening after 1991 had on TFP of firms was positive.a In a study of 131,558 factories, based on the Annual Survey of Industries, Tata Services Limited (2003) found that TFP growth increased from 0.68 percent per year dur- ing the prereform period (1981/82­1992/93) to 0.97 percent during the postreform period (1993/94­2001/02). Similar findings were reported in Unel (2003). The contribution of TFP growth to real output growth has increased as has labor productivity growth. Pattnayak and Thangavelu (2003) found that reforms in India in the 1990s raised TFP in 10 out of 13 key industries, such as leather and leather products, chemical and chemical products, metal products and parts, machinery and equipment, and electrical and related parts. In com- parison, TFP growth in traditional industries, such as food, beverage, basic metal and alloys, wool, silk and humanmade fibers, textiles, and nonmetallic minerals, has either remained con- stant or declined. Pattnayak and Thangavelu's comparison of technical change before and after the 1991 reforms suggests that the TFP trend in most of the sectors is significantly higher after the 1991 reforms. There are also indications that productivity has increased most in those sectors that have opened up to the global markets. Topalova (2004) and Goldar and Kumari (2003) found that reductions in trade protectionism led to higher growth of firm productivity. Interestingly, Topalova (2004) showed that while this effect is robust and highly statistically significant for pri- vate companies, there is no evidence that trade liberalization leads to any productivity improve- ments for government-owned companies. This finding suggests that private companies are bet- ter able to take advantage of the new opportunities. Further evidence of the links between productivity and reform has been provided in Dollar, Iarossi, and Mengistae (2002) and in Veeramani and Goldar (2004), who analyzed the influence of investment climate on TFP in Indian states. Using a survey of more than 1,000 manufactur- ing establishments and controlling for establishment size and industry, Dollar, Iarossi, and Mengistae found that value added per worker is about 45 percent lower in states that business managers consider to possess a relatively poor investment climate. Most of this gap is a result of the lower TFP of firms in these states. The average rate of net fixed investment is less than 2 percent for firms from states with a poor investment climate in contrast to 8 percent for firms in states with a good investment climate.b Sources: Dollar, Iarossi, and Mengistae 2002; Goldar and Kumari 2003; Pattnayak and Thangavelu 2003; Topalova 2004; Unel 2003; Veeramani and Goldar 2004. a. Goldar (2004) and a number of other authors found that TFP growth in the 1990s decelerated. Goldar and Kumari (2003) suggested that gestation lags in investment projects and slower agricultural growth in the 1990s had an adverse effect on productivity growth. The analysis reveals that underuse of industrial capacity was an important cause of the productivity slowdown. b. Dollar, Iarossi, and Mengistae (2002) found that one-fourth of the TFP gap can be traced to inferior power supply and weak Internet connectivity in states with a poor investment climate. About one-tenth of the gap is due to a higher regulatory burden in those states. 18 The Indian Economy 2 construction--touched nearly US$20 billion in 2006/07, up from its historic US$3 billion to US$4 billion range. Foreign institutional investors made record purchases in the Indian stock market 2007 as the Bombay Stock Exchange's Sensex continued to attain new highs (it was the fourth best-performing market globally in 2007). Challenges in the Way Forward Creating employment in the formal sector, reducing inequalities between states, and accelerating the growth of productivity are the key challenges ahead. Despite its con- siderable achievements, India will need to overcome significant challenges to main- tain growth. Formal sector employment is under pressure from demographics, and inequalities between states are on the rise. Given the record-high investment levels, future growth will depend increasingly on productivity gains. Creating Employment in the Formal Sector Employment growth has not kept pace with GDP growth, and most newly created jobs are in the informal sector. Although GDP increased at about 6 percent per year in real terms between 1999/2000 and 2004/05, the annual growth rate in employ- ment for the same period was around 3 percent (NSSO 1984, 1994, 2000, 2005). Moreover, the informal sector continues to generate more jobs than the formal one, with 94 percent of the jobs in the informal sector (figure 2.7). The formal sector share of jobs has been decreasing because of a reduction in public sector employ- ment and sluggish growth in the formal private sector. Jobs in the informal sector, which in India are called "unorganized" jobs, are less secure, give no right to social security, and are not under the purview of the labor laws. Figure 2.7 Formal and Informal Labor Force, 1990 and 2003 2003 94 2 4 1990 92 2 6 0 10 20 30 40 50 60 70 80 90 100 percentage of labor force informal formal private sector formal public sector Source: OECD 2007. 19 Robust Growth, but Challenges Remain 2 Increases in the working-age population and the overall number of employed women put additional pressure on the labor market. Over the past 20 years, the working-age population has increased by more than 60 percent. During the same period, the number of employed women has increased by 47 percent. It is estimated that more than 10 million people will join the workforce between 2005 and 2015. The composition of employment in India has been slow to change even while the economy has seen relatively significant structural changes. The decline in the contri- bution of agriculture to the GDP has not been accompanied by a comparable decline in the number of people employed by the sector (figure 2.2). In fact, agriculture is only 18 percent of GDP, but it still employs 57 percent of the working population. In addition, the decline in the share of agricultural employment (from 68 percent to 57 percent) has not been accompanied by an increase in the share of manufacturing employment (figure 2.8; World Bank 2007b). Volatility and low growth of the agri- cultural sector remain particularly worrisome because almost two-thirds of India's people still depend on the rural sector for their livelihoods. The reallocation of the labor force from the agricultural sector to other sectors has resulted in increased total productivity because productivity of labor in the informal agricultural sector was far lower than in all other sectors (see table 2.2). In particular, employment growth picked up in the construction and trade sectors as well as in the financial sector, which has high productivity but also limited effect on the overall productivity of Indian labor because it accounts for a small share of employment (see figure 2.9). Much of the labor force, however, moved to informal employment, where the average productivity is still lower than that of formal employment (see table 2.2). More specific information about reallocation of the Figure 2.8 Evolution of the Labor Force Distribution, by Sector, 1983/84­2004/05 100 18 21 24 25 80 11 force 11 3 11 12 4 60 5 6 labor of 40 68 64 60 57 percentage 20 0 1983/84 1993/94 1999/2000 2004/05 fiscal year agriculture industry (excluding manufacturing) manufacturing services Sources: Authors' calculations based on Guruswamy and others 2005; National Commission for Enter- prises in the Unorganised Sector 2007; NSSO 1984, 1994, 2000, 2005. 20 The Indian Economy 2 Table 2.2 Labor Productivity and Employment Shares Level of productivity Share of employment (%) (1999/2000 Rs) 1993 2003 1993 2003 Informal sector: 92.6 94.0 20,721 27,958 Agriculture 63.9 57.7 14,745 17,431 Informal other 28.8 36.3 33,989 44,674 Formal sector: 7.4 6 139,663 308,099 Private companies 2.1 1.9 174,220 431,699 Public enterprises 2.7 2 148,215 319,883 Public services 2.5 2.1 101,756 186,891 Whole economy 100.0 100.0 29,475 44,706 Source: OECD 2007 (table 2.1). Note: Productivity is computed as net product per person. labor force in nonagricultural sectors is provided by figure 2.9, which shows that from 1993 to 2002 growth in employment became concentrated in sectors with higher productivity. Reducing Inequalities between States The disparities in the growth performance of Indian states are large and growing. If these differences are to be sustained over time, they could translate into vast differ- ences in material well-being. India's average growth performance conceals very dif- ferent growth experiences across states. Introduced in the 1990s, the reforms (which liberalized the choice of location for private Indian firms and dismantled investment licensing) were accompanied by a significant shift in growth patterns, which led to a large gap in per capita income growth between the high- and middle-income states on the one hand and the low-income states on the other (figure 2.10). Although growth in the poorer states as a whole did not change much, growth picked up in the high-income states and accelerated significantly in every middle-income state, thanks to continued reforms. In fact, these reforms resulted in a concentration of skill-intensive firms within manufacturing and services in those states that had a bet- ter investment climate and that already had industrial agglomerations. Such states thus experienced acceleration in their economic growth (Kochhar and others 2006; World Bank 2008). Lagging states, in contrast, continue to be characterized by low- productivity and small-scale activities, subsistence agriculture, and little urbaniza- tion. If these large differences in growth rates between rich and poor states were to be sustained over long stretches of time and low-income states kept falling behind, however, disparities in growth could eventually translate into vast differences in material well-being (most socioeconomic indicators are explained overwhelmingly by income) and possibly social unrest. 21 2 22 Figure 2.9 Productivity Growth and Labor Reallocation in India, 1983­2002 40.0 35.0 33.3 30.0 25.0 20.0 18.0 15.5 15.0 10.1 10.0 9.7 percentage 10.0 8.8 8.0 6.9 7.3 4.9 5.8 4.6 5.0 3.6 3.0 2.6 3.2 3.5 1.0 1.8 2.1 4.5 0.9 0.5 0.0 Robust 0.3 0.2 5.0 3.6 Growth, 10.0 agriculture construction trade, hotels, manufacturing community, transport, mining and electricity, financial, and restaurants social, and storage, and quarrying gas, and insurance, personal communications water supply real estate, and but services business services Challenges increase in labor force, 1983­93 increase in labor force, 1993­2002 labor productivity relative to agriculture Source: World Bank 2007b. Remain The Indian Economy 2 Figure 2.10 States' Growth in Per Capita Income, 1980s and 1990s high-income states middle-income states low-income states 0 1 2 3 4 5 6 growth in per capita income (%) 1980s 1990s Source: World Bank 2008. Note: Low-income states are Bihar, Jharkhand, Madhya Pradesh, Orissa, and Rajasthan, and Uttar Pradesh. Medium-income states are Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and West Bengal. High-income states are Delhi, Gujarat, Haryana, Maharashtra, and Punjab. The growth of labor productivity accounts for most of the differences in growth rates between richer and poorer states; hence, the investment climate matters. If the growth of per capita income is decomposed into its sources, labor productivity growth is found to account for 80 percent of the differences between lower- and higher-income states for the period from 1983 to 2004 (figure 2.11). The effects of the employment rate growth and demographic changes on per capita income growth have been much smaller. A comparison of the income gap between rich and poor states or provinces in various countries is provided in box 2.2. Accelerating the Growth of Productivity Accelerating productivity growth will be a key factor in sustaining current growth rates. Overall economic growth will depend on either (a) capital accumulation through greater investment or (b) productivity enhancements. However, private sav- ings and investment are already at a record high, and public savings have benefited from an impressive fiscal consolidation process. Nonetheless, the potential to increase productivity is high (figure 2.12). Focus of the Report This report focuses on four sectors that are critical to the challenges of increasing employment in the formal sector, reducing inequalities among states, and improving productivity: 23 Robust Growth, but Challenges Remain 2 Figure 2.11 Decomposition of Differences in Per Capita Income Growth across Groups of States, 1983­2004 a. difference between high-income states and low-income states 2.5 2.0 1.5 1.0 0.5 percentage 0.0 0.5 1.0 growth in per labor employment proportion of capita income productivity rate working-age people in population b. difference between middle-income states and low-income states 2.5 2.0 1.5 1.0 0.5 percentage 0.0 0.5 1.0 growth in per labor employment proportion of capita income productivity rate working-age people in population Source: World Bank 2007b. 1. Organized manufacturing. Manufacturing is the largest subsector contributor to GDP after agriculture. However, its share as a percentage of GDP is much lower than that of other developing countries (for example, 33 percent in China and 35 percent in Thailand), suggesting potential for growth. Moreover, unusual for developing countries, excess labor from agriculture is being absorbed more into services, especially informal services, than into manufacturing, which usually plays a bigger role in creating jobs for semiskilled labor. 24 The Indian Economy 2 Box 2.2 International Comparisons of Income Gaps between States The comparison between India and other regions is interesting. Even with the increase in income, the current gap in income between the richest and poorest Indian states is much smaller than that of other large federal countries. A comparison of the standard deviation in per capita output among internal provinces, states, or counties within Brazil, China, India, Indonesia, the United States, and the 12 states of the European Union as of 1986 (that is, Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, and the United Kingdom)--all regions with more than 200 million people--reveals that, while the varia- tion in output across states increased in India, it is still substantially smaller than in Brazil, China, and Indonesia, but strikingly larger than the cross-state differences in the United States or the cross-country differences in the European Union (see accompanying figure). Source: World Bank 2006b. Standard Deviation of GDP Per Capita in India and Comparator Countries, 1980s and 2000s 0.8 0.7 0.6 0.5 deviation 0.4 0.3 standard 0.2 0.1 0.0 Indonesia China Brazil India EU United States 1980s 1990s Source: World Bank 2006b. Note: EU Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, and the United Kingdom. 2. Unorganized manufacturing. Unorganized manufacturing (firms employing fewer than 10 workers--fewer than 20 if the firm does not use power) is particularly important for employment in the manufacturing sector; 74 percent of the manufacturing workers are in the unorganized segment (figure 2.13). In addition, unorganized manufacturing is more evenly distributed across low-, middle-, and high-income states. In 2005, 10 districts alone accounted for 49 percent of 25 Robust Growth, but Challenges Remain 2 Figure 2.12 Relative Sector Productivities and Potential 100 95 95 90 90 84 80 80 75 76 75 60 53 55 53 value 44 40 index 35 25 24 20 16 11 12 6 8 0 retail housing steel retail apparel automotive software construction banking sector average local best practice viable best-practice potential Source: McKinsey Global Institute 2001. Note: The viable local good practice is based on an estimate of how much the productivity of the cur- rent good practice in India could be improved by further improving the organization of functions and tasks, larger scale of operations, and so forth. Reported values are indexed to the U.S. productivity level set at 100 for each sector. Figure 2.13 Labor Force Distribution, 2004 organized manufacturing, 3% industry, excluding unorganized manufacturing, manufacturing, 6% 9% other services, 17% retail, 7% agriculture, 58% software and ITES, 0.3% Sources: Authors' calculations based on Guruswamy and others 2005; NASSCOM 2007; National Com- mission for Enterprises in the Unorganised Sector 2007; NSSO 2005; OECD 2007. 26 The Indian Economy 2 national investment in manufacturing, while in the same year, high-income states accounted for less than half of estimated investments in unorganized manufac- turing across all the states. Facilitating the growth of this subsector to organized manufacturing could help promote growth and employment in the poorer states. 3. Retail. The retail sector is the fourth-largest contributor to GDP after agriculture, manufacturing, and the financial sector, and its share is growing. Moreover, the structure of the retail sector affects the performance of the domestic supplying industries and agricultural producers. Finally, given the limited capital, land, and skills required, small-scale retailing is often the default employer for people who cannot find employment in the formal manufacturing sector. 4. IT and ITES. The fastest-growing sector in recent years has been IT and ITES, and average labor productivity in the sector is much higher than in the other main sectors. Thanks to this growth, the sector is the largest contributor to exports, and in 2005, India became the world's largest exporter of IT- and ITES-related serv- ices. In a sense, the sector represents the success story of the liberalization of the 1990s. Hence, an understanding of what has supported this success is crucial to supporting enhanced productivity in the other sectors. This report focuses on ITES and a subsector of IT--namely, software. Notes 1. Combined for the central government and states. 2. According to Bosworth and Collins (2007), faster growth during 1993/94 was mainly attributa- ble to more rapid accumulation of physical capital stocks (growth at 1.8 percent in 1993/94 com- pared with 1.0 percent in 1978 to 1993) and stronger TFP growth (2.3 percent compared with 1.1 percent). 3. See Reddy (2006). ICOR is defined as the ratio of incremental capital to incremental output. 4. That is, the sum of export and imports of goods and nonfactor services over GDP at factor cost, measured in current rupees. 5. That is, exports of goods and nonfactor services measured in constant rupees. 27 Chapter 3 Manufacturing Investment Climate Obstacles to Productivity Growth and Employment Generation The contribution of manufacturing to employment and gross domestic product (GDP) remains limited in India, and the sector is characterized by growing but still low productivity when compared to the situation in other countries and in other sec- tors in India. Improving the investment climate is key to increasing productivity. Although the investment climate has improved in recent years, with middle-income states seeing the largest improvement, enterprises nevertheless identified four key constraints to growth: electricity, taxes, tax administration, and corruption. The impact of investment climate obstacles on growth varies across groups of states; dif- ferences are most pronounced in areas that are more directly affected by state-level policies, such as power. The investment climate also limits the competitiveness of Indian export manufacturing firms, which face much higher indirect costs than, for example, their Chinese competitors. The analysis draws from the results of the 2003 and 2006 Manufacturing Investment Climate Surveys (ICSs) and data from second- ary sources.1 The Manufacturing Sector: A Potential Driver of Growth The low share of manufacturing in GDP and manufacturing's limited contribution to employment remain a concern. Within industry, manufacturing's share has decreased from 18 percent of GDP in 1995/96 to 16 percent in 2005/06. This share is much lower than that in other developing and East Asian countries (table 3.1), sug- gesting room for further growth that could lead to more jobs and greater productiv- ity. The share of manufacturing employment is also low, at less than 15 percent of all employment. Atypically for developing countries, excess labor from agriculture is being absorbed more into the services sector--especially informal services--than 29 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Table 3.1 Contribution of Manufacturing to GDP in Selected Countries and Regions, 2005 Country or Region Manufacturing, value added (% of GDP) China 33 India 16 Mexico 18 Thailand 35 East Asia and the Pacific 32 South Asia 16 Sources: Ministry of India 2008; World Bank 2007c. into manufacturing, which usually plays a much bigger role in creating jobs for the semiskilled. The National Strategy for Manufacturing (National Manufacturing Competitiveness Council 2006) estimates that if manufacturing sector growth can be pushed up to 12 percent, it can generate some 1.6 million jobs annually directly and two to three times this number indirectly. Productivity is important for increased growth in the manufacturing sector. According to economic theory, economic growth can come from increases in the fac- tors of production, such as labor supply and capital, or from increased total factor productivity (TFP). Virmani (2004) shows that the contribution of TFP growth to economic growth in India fell progressively from 85 percent in 1951/52 to a little more than 25 percent in 1991/92 and then rose sharply to almost 50 percent in 2003/04. As explained by Easterly and Levine (2001), a large body of literature indicates that inter- nationally the bulk of economic growth differences are explained by differences in TFP rather than by differences in physical and human capital accumulation. To understand what holds back the growth of the manufacturing sector, the World Bank conducted two nationally representative surveys in 2003 and in 2006. The Manufacturing ICS 2003 covered 1,827 firms across 11 industries (garments, textiles, leather goods, pharmaceuticals, consumer electronics, white goods, machin- ery, auto parts, metal products, chemicals and plastics, and food processing) and 40 cities in 12 states (Andhra Pradesh, Delhi, Gujarat, Haryana, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Punjab, Tamil Nadu, Uttar Pradesh, and West Bengal). The 2006 ICS covered the same sectors and cities and expanded the sample to include more than a dozen new cities from four additional states (Bihar, Jharkhand, Orissa, and Rajasthan) for a total of nearly 2,300 firms. All the data presented in this chapter refer to data collected through this survey except when otherwise indicated. (Appendix A describes the survey methodology.) The sector is characterized by low productivity. This situation is particularly problematic for exporters that need to be globally competitive. The labor productiv- ity of the Indian manufacturing industry is low when compared to U.S. levels: the labor productivity of the Indian apparel, steel, and automotive sectors is at 26, 11, and 24 percent, respectively, of that of the United States (McKinsey Global Institute 30 Manufacturing 3 2001: exhibit 1.10). The Indian manufacturing sector also performs poorly compared with that in other developing countries: in 2006, the value added per worker in India (US$8,274) was well below the value added per worker in China (US$12,367) and in Turkey (US$10,186). The productivity of the manufacturing sector is also very low when compared with that of the Indian software and infor- mation technology­enabled services (ITES) sector: average TFP in the manufactur- ing sector is 4.3, while average TFP in the software and ITES subsector is 35.6.2 Investment climate matters for productivity growth. It is now well documented that the investment climate of a country can significantly affect productivity: an unfavorable investment climate results in less value added being produced for a given level of human and physical capital (see, for example, Dollar, Hallward-Driemeier, and Mengistae 2005; Escribano and Guasch 2005; Rodrik and Subramanian 2004). Returns to inputs are larger when the investment climate is better, stimulating fur- ther accumulation of capital. A favorable investment climate therefore promotes economic growth through a positive influence on productivity. Firms' Perceptions of Investment Climate Constraints: Power, Taxes, and Lack of Skilled Workforce With a few exceptions, firms responding to the 2006 survey were less likely to report most investment climate variables as severe problems than those responding to the 2003 survey. This finding suggests that progress has been made on such issues as cost of and access to finance and telecommunications--probably the outcome of reforms in the telecommunications sector and the gradual opening up of the financial sector (figure 3.1). Improvements in access to finance are encouraging because analysis sug- gests that ability to obtain financing has a large effect on both employment genera- tion and the probability of receiving foreign direct investment (FDI). In 2006, busi- nesses were less likely to be worried about economic policies and macroeconomic stability. They were also, on average, less concerned about corruption and crime. Backing this perception was the fact that the incidence of firms reporting giving "gifts" or "informal" payments to public officials to facilitate business fell from 62 percent in 2003 to 49 percent in 2006. This trend is consistent with that from other sources such as Transparency International's Corruption Perceptions Index. India's score on this index improved from 2.7 (out of 10.0) in 2002 to 3.5 in 2007. Nonetheless, perceptions have worsened with respect to electricity, tax rates, and workforce skills. Electricity was rated a major or severe constraint by 36 percent of respondents in 2006--a sharp increase from 2003. Backing the perception of more significant power shortages is the fact that, for the firms in the survey, as much as 22 percent of all energy requirements came from generator-captive power in 2006, up from 19 percent in 2003. This problem is widespread and not confined to a few states. High taxes were a close second to electricity in terms of obstacles; in 2006, 35 percent of respondents believed taxes were a major problem for growth.Although only 15 per- cent of businesses listed workforce skills as a major obstacle, this figure represents a 31 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Figure 3.1 Proportion of Firms Reporting Each Obstacle as Major or Severe, 2003 and 2006 40 37 36 35 35 30 29 28 28 27 26 firms 25 of 21 20 19 17 17 16 16 16 15 15 15 14 14 13 12 12 percentage 10 10 10 9 9 9 8 8 7 5 5 4 0 taxes trade skills land crime finance to electricityhigh corruption and to licensing behaviorinstability worker regulations uncertainty administration access transportation tax labor customs access policy telecommunications anticompetitive macroeconomic 2006 2003 Source: Manufacturing ICS 2003; Manufacturing ICS 2006. steep increase from 2003; 24 percent of respondents reported unavailability of required skills in the market, and 34 percent of all firms claimed that finding an ade- quately skilled employee is a lengthy process. These results bring into focus that skill shortages are becoming more pressing with the rapid growth of the Indian economy. Finally, more than a quarter of businesses identified corruption as a major obstacle (figure 3.2). Electricity, tax rates and tax administration, and corruption are the biggest obsta- cles identified by manufacturing firms. When asked to name the most important obstacle to growth, 32 percent of managers identified electricity, 15 percent high taxes, 8 percent tax administration, and 10 percent corruption (figure 3.2). A study by McKinsey Global Institute (2001) also identified similar issues as obstacles to growth.3 The key challenges facing the power sector in India are insufficient generation and inefficient and poorly governed distribution, which together lead to low access. On average, firms reported facing power outages 122 times a year, and the median firm reported losing 5 percent of its sales revenue because of this problem. More than half of all enterprises relied on a backup energy source. The situation was worse in the poorer states of India; on average, these firms lost almost 10 percent of their sales because of power shortages, and more than a quarter of their electricity require- ment came from their own or shared generators. Public grid power outages raise costs for firms because private sector backup sources cost almost twice as much. Because they invest less in mitigating measures, small and medium enterprises 32 Manufacturing 3 Figure 3.2 Firms Perception of Investment Climate Variables as the Most Important Obstacle or as a Major or Very Severe Obstacle, 2006 40 36 35 35 32 30 28 27 firms 25 of 20 16 15 15 15 10 15 14 percentage 10 10 10 9 9 8 8 8 7 5 5 5 4 4 4 3 3 2 2 2 2 1 1 1 1 1 1 0 0 0 land trade skills to taxes crime finance system foreign cations electricity high and to regulations licensingto uncertaintyinstability corruption behavior regulations telecommuni- legal transportation access administration workers technology macroeconomic anticompetitive tax labor customs accesspolicy access most important obstacle major or very severe obstacle Source: Manufacturing ICS 2006. suffered higher losses from power outages and surges (only 35 percent of firms with fewer than 20 workers owned generators compared with 85 percent of firms employ- ing more than 100 workers). The average small firm lost 9 percent of its sales because of this problem alone, compared with about 5 percent for the typical large firm. Firms cite high tax rates and inefficient tax administration as a key obstacle to growth. Doing Business 2008 (World Bank 2007a) provides an additional reason for concern. It finds that taxes are high in India in an international context. A represen- tative limited liability company in India has to pay almost 71 percent of its profits as taxes4--a much higher percentage than found in other developing countries, such as Pakistan (40.7 percent) and the Arab Republic of Egypt (47.9 percent). Tax admin- istration is also costly for firms. A representative company in India needs to make 60 payments a year and will spend 271 hours doing so, compared with just 12 hours in a country like the United Arab Republic. Notably, in the Manufacturing ICS 2006, enterprises that complain about tax administration are also very likely to complain about corruption. There has been little reform on this front, yet the issue is impor- tant because international evidence at the firm level shows that higher taxes and poor tax administration have an adverse effect on growth and distort financial and invest- ment decisions. High tax rates contribute to the growth of a shadow economy, which carries costs in terms of forgone tax receipts and lower productivity. Red tape and corruption affect businesses both in terms of how often and how much they have to pay and in terms of the amount of management time spent in 33 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 dealing with officials. Informal payments are common for accessing basic services, permits, and licenses from the government. Gifts or informal payments were required in 39 percent of cases for an electrical connection and 27 percent of cases for a water connection. For permits and licenses, informal payments were made in as much as 68 percent of cases for construction-related permits, 46 percent of cases for import licenses, and 55 percent of cases for a main operating license. The typical value of the gift or informal payment expected to secure any government contract stood at 6.6 percent of the contract value. Gifts or informal payments were requested or expected during inspections by 54 percent of tax inspectors, 56 percent of labor officials, 48 percent of fire and building safety officials, 54 percent of sanitation and epidemiology officials, and 59 percent of police officials. Overall, manufacturing firms paid on average 4.9 percent of total sales in bribes. Additionally, senior man- agement spends a substantial amount of time dealing with government officials. Overall, managers of manufacturing firms spent on average 12.6 percent of their work week dealing with government officials. Corruption hits smaller firms harder. First, though smaller firms are less likely to be inspected, when they are, they are more likely to end up paying bribes (figure 3.3). Second, informal payments often need to be made to "get things done." Smaller firms are less likely to opt for this route, compared with larger firms, because they are more financially constrained. Their smaller size also means that, relative to sales, the impact of bribes is larger. In the ICS, though larger firms reported paying about 2 percent of their annual sales as bribes, the impact on smaller firms was more than three times that percentage (6 percent). Figure 3.3 Impact of Corruption, by Firm Size, 2006 70 63 60 57 55 55 49 50 46 40 30 20 13 13 13 9 10 5 6 6 4 2 0 average number percentage of average % of percentage of average % of of inspections firms that paid senior manage- firms that paid annual sales paid from officials bribes during ment time spent bribes to get as bribes to get inspections dealing with things done things done regulations fewer then 20 employees 20­99 employees 100 or more employees Source: Manufacturing ICS 2006. 34 Manufacturing 3 Figure 3.4 Percentage of Firms Rating Labor Regulations as a Major or Very Severe Obstacle, 2006 40 firms 23 of 20 17 13 10 percentage 0 fewer than 100 employees 100 or more employees firms with low labor intensity firms with high labor intensity Source: Manufacturing ICS 2006. Only 4 percent of the firms identify labor regulations as a major obstacle, but the problem is more acute for labor-intensive industry and large firms. On the whole, the proportion of respondents rating labor regulations as the biggest obstacle was low, at 4 percent in 2006. However, almost one-fifth of the smaller (fewer than 100 employees) labor-intensive firms and almost one-quarter of the larger labor- intensive firms reported labor regulations as a major concern (figure 3.4). These regulations also seem to work to restrict workforce growth. Although a very small percentage (less than 3 percent) of firms reported wanting to downsize, almost 16 percent of respondents wanted to expand their current workforce. For enterprises in labor-intensive sectors, this figure was as high as 21 percent. However, among the firms that wanted to hire workers, almost a quarter said they were deterred from doing so by laws and regulations relating to hiring and firing of workers. A further 8 percent were deterred by the mandatory benefits they would have to offer new workers. It thus appears that, in periods of upturn in manufacturing activity, labor regulations serve to dampen employment growth. Productivity Analysis of the Investment Climate and Firms' Perceptions of Key Obstacles Econometric analysis provides strong support to the firms' identification of invest- ment climate obstacles. An econometric estimate of the effect of investment climate variables on TFP, employment, wages, probability of exporting, and probability of receiving FDI highlights the importance of investment climate (see appendix B for the methodology). The econometric analysis of the effect of investment climate on firms' performance suggests that a large part (76 percent) of the explainable intrafirm 35 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Figure 3.5 Proportional Impact of Investment Climate Variables on Firm TFP, Employment, Real Wages, Exports, and FDI, 2006 FDI 18.0 15.8 19.1 9.3 2.7 35.1 exports 12.1 30.4 12.1 21.9 2.3 21.1 real wages 7.4 18.3 12.3 35.8 13.5 12.6 employment 7.1 11.8 19.5 16.0 22.1 13.9 9.6 productivity 22.8 27.5 10.2 15.5 23.9 0 10 20 30 40 50 60 70 80 90 100 proportion of firms (%) infrastructures red tape, corruption, and crime finance and corporate governance quality, innovation, and labor skills other control variables real wages productivity Source: Manufacturing ICS 2006. Note: The proportional impact of the investment climate and control variable groups is measured by evaluating each estimated equation (productivity, employment, wages, exports, and FDI) at the sample mean while taking the absolute value of the esti- mated coefficients. variability in TFP can be explained simply by the differences in the investment climate that firms face. Firm-level characteristics are much less important (figure 3.5). Red tape, corruption, and crime had the largest negative impact on firm produc- tivity, real wages, exports, and possibility of getting FDI. These factors also are the hardest to circumvent. Red tape and corruption are effectively like taxes, which all businesses end up paying. The costs of red tape and corruption can be high; these investment climate variables explained 27.5 percent of observed productivity differ- entials among manufacturing establishments (figure 3.5). Productivity was nega- tively affected by informalities in the relations between firms and the government, such as making informal payments and incomplete tax declarations. Crime also hurts productivity: firms that reported having to hire security generally had lower productivity. Economic theory shows that taxes on firms also end up affecting work- ers; the analysis of this study confirms that corruption ends up hurting real wages. To the extent that it raises the costs of doing business, corruption also reduces the competitiveness of Indian products abroad and hurts exports. In fact, it is the single biggest investment climate variable affecting export possibility. Red tape also reduces the possibility of attracting FDI. Poor infrastructure was the second-largest investment climate obstacle, and it affected the most productive firms relatively more. Poor infrastructure accounted for 36 Manufacturing 3 as much as 23 percent of explained intrafirm productivity differentials in the manu- facturing sector. Key variables included the number of days taken to clear customs, power shortages, and inadequate water supply from public sources. However, unlike red tape and corruption, which affected all firms equally, poor infrastructure hits the more efficient firms even harder. It also damages the economy because FDI is less likely to flow to areas with poor infrastructure. Labor skills, quality, and innovation have a smaller effect on productivity but are more important for the more efficient firms in the economy. Improving the labor quality available to enterprises and stimulating innovation have the potential to raise productivity by as much as 15.5 percent. Firms that trained their workforce, used computers, undertook research and development activities, or accessed technology licensed by foreign companies were on average more productive. This fact was also reflected in the finding that firms better endowed in this respect paid higher wages to workers and were more likely to be exporters. Poor access to finance and corporate governance are associated with relatively small productivity impacts, but they have a large impact on employment growth and the possibility of attracting FDI. The productivity of firms is affected by whether the firms belong to a trade association, have been given a loan, or have had an external audit done. The effects of access to finance and corporate governance on employment gen- eration and on the probability of receiving FDI were large compared with those of other investment climate factors. Firms with better access to finance and corporate governance on the whole had higher employment and were more likely to receive FDI. Impact of Investment Climate Obstacles on Growth Across Groups of States The surveys showed large differences across states in the investment climate variables that firms find constraining. Firms in low-income states were more likely to complain about the power situation and less likely to complain about corruption, worker skills, taxes, and access to land than the national average (figure 3.6). Firms in middle- income states, in contrast, were less likely than those in low-income states to complain about power: 14 percent and 58 percent, respectively, make such complaints. They were, however, more likely to complain about issues such as access to land, access to finance, labor regulations, and worker skills. In fact, firms in middle-income states were more likely to complain about tax administration issues and were about as likely to consider labor regulations and labor skills as obstacles as the power situation. In high-income states, relative to the national average, the major issues were corruption and the possibly closely related issues of high taxes and tax administration. The differences are most pronounced in areas that are more directly affected by state-level policies, such as power. Issues for which there are large differences in per- ceptions among states are mainly areas that are the domain of the states (figure 3.6). These areas include power, access to land, tax administration, and corruption, which is possibly related to tax administration and inspections. Substantial differences in 37 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Figure 3.6 Deviation from the National Average of Percentage of Firms Reporting Each Obstacle as Major or Very Severe in 2006, by Income Level of States, 2006 15 10 5 (%) 0 deviation land taxes trade skills crime to finance instability 5 electricity licensing high and to regulations uncertainty corruption regulations worker transportation access administration tax labor customs access policy telecommunications 10 macroeconomic 15 high-income states middle-income states low-income states Source: Manufacturing ICS 2006. access to land and workers' skills are also explainable on the basis of the fast growth of middle-income states in recent years. Investment climate improvements in the middle-income states drove many of the perceived national improvements. When the 2003 and 2006 surveys are compared, it is clear that all states have benefited from macroeconomic stability. Perceptions have also improved across the board in terms of corruption, crime, theft, and disorder. The effect of financial deregulation--a central subject--has been more mixed; although high- and middle-income states appear to have benefited, enterprises in low-income states still complain more about access to finance. Differences between states were less stark in 2003. However, between 2003 and 2006, middle-income states saw the largest improvements in investment climate variables amenable to state policy. For example, the percentage of firms reporting electricity to be a major issue in middle-income states more than halved, from 31 percent in 2003 to 14 percent in 2006. Perceptions also improved significantly in other areas, such as transport, taxes and tax adminis- tration, licensing, labor regulations, and corruption (figure 3.7). In contrast, percep- tions on all these factors, except for corruption, worsened (or increased only margin- ally) both for the low-income and the high-income states. Objective indicators support the perceived changes in most investment climate variables reported by firms across states. In middle-income states, firms reported 38 Manufacturing 3 Figure 3.7 Change in the Percentage of Firms Identifying Factor as a Major or Very Severe Obstacle between 2003 and 2006, by Income Level of States 30 25 20 15 10 (%) 5 change 0 5 10 15 20 power transport taxes tax licensing labor corruption administration regulations low-income states high-income states middle-income states Source: Manufacturing ICS 2003; Manufacturing ICS 2006. improvements in power, transport, tax, and corruption obstacles, while in low- income states, they perceived a worsening power situation and modest improve- ments in tax administration. These perceptions are supported by objective indicators that show that the number of power outages faced by interviewed firms in middle- income states decreased from 2003 to 2006 (67 versus 57 percent for national average) and the reliance on generators also decreased (35 percent versus 12 percent for the national average). In low-income states, the average number of power outages decreased by less than the national average (that is, 52 percent) and percentage of electricity provided by generators increased substantially (68 percent). In terms of transport, over the period from 2004 to 2006, middle-income states had the largest highway development, with an average of 316 kilometers added per state.5 In tax administration, the number of tax inspections faced each year by interviewed firms in middle-income states decreased from six to three between 2003 and 2006. The average number of tax inspections per year in low-income states also decreased, but by less than the national average (from four to three). The data on informal pay- ments were not comparable across time; however, data on number of inspections, which are often associated with informal payments, show a clear improvement for middle-income states (from 10 to 5 inspections per year). The average amount of management time spent dealing with bureaucracy also decreased substantially in these states, from 27 to 15 percent. Improvements in the middle-income states in infrastructure are a direct outcome of policy initiatives. Improvements in the power situation of middle-income states 39 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 are most likely the result of the unbundling of some State Electricity Boards (SEBs) in such states as Andhra Pradesh and Karnataka and because of the thorough imple- mentation of the Accelerated Power Development and Reform Program (APDRP) in states such as Kerala, Tamil Nadu, and West Bengal. The unbundling of SEBs into separate entities has helped each entity to concentrate on a particular function, to increase efficiency, and to cut commercial losses. Although the APDRP is a national program aimed at increasing investment in power generation and at providing incentives to states that reform the sector, some states have progressed much faster and more thoroughly than others with the reform. As a result of these new policies, and over the period from 2003 to 2006, Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and West Bengal have recorded a decrease in aggregate technical and com- mercial losses higher than the national average: from 36.1 percent to 27.2 percent, from 45.7 percent to 34.7 percent, from 36.2 percent to 32.1 percent, from 20 per- cent to 18.8 percent, and from 26.6 percent to 23.9 percent, respectively. Improve- ments in the transport system of middle-income states can be attributed to an increase in capital spending in the road sector.6 Middle-income states have spent on average 27 percent of the state's total capital spending in the road sector (as opposed to 12 percent and 13 percent for low- and high-income states, respectively) during the period from 2000 to 2005. In absolute terms, most middle-income states have invested twice as much as low-income states in the road sector during the same period. Such states as Andhra Pradesh, Karnataka, and Tamil Nadu have even higher capital investments in the road sector than some high-income states. The Investment Climate for Exporting Manufacturing Firms India's integration in the global economy has increased in the past two decades but is lower than in comparator countries. Exports posted a strong 19 percent growth between 2000/01 and 2005/06. However, integration, as measured by the trade to GDP ratio, is still lower than in other comparator countries, such as China, Mexico, and Turkey, for which trade accounted for 72, 65, and 64 percent of GDP, respectively. India's integration is also the result of reduced trade barriers. Reforms of the early 1990s reduced licensing requirements, small-scale reservation lists, and tariffs. FDI requirements have been lowered by introducing automatic clearing of FDI inflow and by allowing 100 percent foreign ownership in a number of manufacturing sec- tors. As a result, FDI inflows had a threefold increase between 2003/04 and 2006/07. The number of sectors reserved to small firms (that is, the so-called small-scale reser- vation policy) was also significantly reduced--to 35 items in 2006 from 873 in 1990. Finally, tariff rates have been progressively lowered; the highest standard tariff rate for nonagricultural products, for example, was brought down from 35 percent in 2001 to 10 percent in 2007. Despite these improvements, India has still relatively high import tariff rates. The reduction in trade barriers has resulted in a worsened perception of the investment climate in terms of customs and trade (22 percent 40 Manufacturing 3 of exporters reported trade and customs to be a major or severe obstacle in 2006, while 17 percent did in 2003) because firms are less protected from international competition. India's export performance in manufactured goods has been modest when com- pared with the growth of overall exports. Although exports in India have posted strong growth rates since 2000/01, the growth of manufactured goods exports decel- erated sharply, posting a 4 percent per year growth in the 2000/01 to 2004/05 period, compared with 14 percent in the 1990/91 to 2000/01 period. As a result, the share of manufactured goods in total exports declined from 56 to 44 percent (World Bank 2007c). Over the same period, exports of nonfactor services grew by 28 percent a year; the segment posting the highest growth was software, which increased its share from 10 to 13 percent of total exports. The composition of the Indian manufacturing sector changed considerably over the past two decades. Table 3.2 presents the 15 most important manufacturing seg- ments for Indian exports, based on share of total exports in 2006/07; overall they accounted for 70.7 percent of total exports. The sectors with the largest share of world exports in 2006/07 were mineral fuels, pearls and gems, and textile articles (the last sector includes both apparel categories described in the table and "other made textiles," which are not included in the top 15 export manufacturing sectors). Table 3.2 Top 15 Manufacturing Exports, 2006/07 Exports share of Exports share of Industry total exports (%) world exports (%) Mineral fuels, oils, distillation products, and so forth 14.96 1.07 Pearls, precious stones, metals, coins, and so forth 12.72 6.81 Organic chemicals 4.54 1.92 Iron and steel 4.43 1.70 Articles of apparel and accessories, not knitted or crocheted 4.18 3.32 Boilers, machinery, nuclear reactors, and so forth 4.03 0.32 Ores, slag, and ash 3.86 5.09 Electrical and electronic equipment 3.25 0.25 Cottona 3.11 7.86 Vehicles other than railways, tramways 2.98 0.38 Articles of apparel and accessories, knitted or crocheted 2.86 2.48 Articles of iron or steel 2.69 1.64 Pharmaceutical products 2.52 1.10 Copper and copper articles 2.41 2.25 Plastics and plastic articles 2.17 0.71 Source: United Nations Comtrade Database, http://comtrade.un.org/db/. Note: The graph is limited to the top 15 segments in terms of share in total exports. a. This category includes cotton products from uncarded cotton, not combed to cotton yarn. It does not include any textile or apparel products woven with cotton yarn or thread. 41 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Since 2003/04, the mineral fuels and oil sector has gained shares dramatically (from 6 to 15 percent) thanks to the global increase in demand for oil. In contrast, pearls, gems, and precious stones declined appreciably, from 17 to 13 percent, over the same period. The mix of the top Indian manufacturing exports is dispersed and not focused on unskilled labor-intensive products, where the country has a comparative advantage. India's top manufacturing exports are distributed across many sectors. In fact, India's top nine exporting sectors account for more than 3 percent of total exports; by com- parison, China's exports are much more specialized, with five sectors accounting for more than 3 percent of exports (see Kumar and Sen Gupta 2008; Panagariya 2007). Moreover, India's main exports are either capital intensive (for example, mineral fuels or iron and steel) or skilled labor-intensive (such as pearls and precious stones), and most products have low value added. The majority of India's exports are concentrated in lower growth segments. Figure 3.8 compares each sector's performance in terms of export growth for India with its performance in terms of world exports. The size of each circle represents the sector's export share of India's total export and, therefore, the importance of the sec- tor for the country's exports. The figure shows that most exports, including many important ones, are concentrated in sectors where global trade is growing more slowly, except for mineral fuels and oil. Figure 3.8 Trade Competitiveness: Growth of Indian Exports and Worldwide Exports for the Sectors with the Largest Share of Total Indian Exports, 2001/02­2006/07 ores, slag, and ash 45 copper 40 35 value iron and steel in 30 pearls, precious stones, and so on 25 pharmaceutical products mineral fuels, oils, exports apparel articles of iron or steel plastics 2001/02­2006/07), 20 (knit or and so on electrical and electronic crochet) world equipment 15 of year organic vehicles per chemicals 10 apparel boilers, (% (not knit growth machinery, and so on 5 or crochet) cotton 0 5 0 10 20 30 40 50 60 70 growth of Indian exports in value (% per year, 2001/02­2006/07) Source: UN Comtrade database, http://comtrade.un.org/db/. Note: The graph is limited to the top 15 sectors in terms of share in total exports. Growth refers to the 2001/02 to 2006/07 period. 42 Manufacturing 3 Figure 3.9 Importance of Largest Manufacturing Export Sectors in the Indian Export Basket and Growth of Sectors' Market Share in the World Export Market, 2001/02­2006/07 35 mineral fuels, oils, and so on 30 cooper exports 25 vehicles world 20 in electrical and boilers, machinery, and so on 2001/02­2006/07), electronic 15 equipment share year organic chemicals of plastics 10 articles of iron or steel per cotton (% ores, slag, and ash growth 5 iron and steel pharmaceutical apparel (not knit or crochet) products 0 apparel (knit or crochet) pearls, precious stones, and so on 5 0 2 4 6 8 10 12 14 16 exports as a share of total Indian exports, 2006 (%) Source: UN Comtrade database, http://comtrade.un.org/db/. Note: The graph is limited to the top 15 sectors in terms of share in total exports. Growth refers to the 2001/02 to 2006/07 period. Moreover, the sectors for which India has the largest share of world exports are losing market share worldwide or improving only moderately. The pearl and pre- cious stone sector and the knitted apparels sector are losing market share; the cotton sector and the ore, slag, and ash sector are increasing at the pace of the sector's world- wide growth or by much less (figure 3.9). Other segments in which India has increased its share in world exports have a relatively small share in domestic exports (copper and vehicles). The competitiveness of Indian manufacturing firms is assessed for sectors that are performing well in exports and for sectors that are not performing well by com- paring their performance with that of the same sectors in other countries. Produc- tivity and costs of exporting manufacturing firms in India are compared with those of other countries to explore whether India's export performance is determined by low competitiveness. The data are drawn from the Manufacturing ICS for India for 2006 and the Manufacturing ICSs for China for 2002 and 2005. The sectors have been selected on the basis of two criteria: sectors in which India's share in world export has been growing (or falling) and sectors that were also included in the ICS of other comparator countries. According to this definition, India is gaining market share in electrical equipment, chemicals, plastics, and vehicles, but not in apparel and leather. When compared with China's manufacturing sectors, India's sectors are not com- petitive. Labor productivity and labor costs are summarized in table 3.3 by the ratio 43 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Table 3.3 Value Added per Unit Labor Cost Sector Ratio (China/India) Sectors not gaining market shares in world exports for India Garments 1.4 Leather and leather products 2.4 Textiles 81.7 Sectors gaining market shares in world exports for India Electric appliances 1.5 Pharmaceuticals 5.0 Plastics and plastic products 14.4 Other chemicals 50.2 Source: Manufacturing ICS 2006 for India; Manufacturing ICS 2005 for China. Table 3.4 Indirect Costs for Exporting and Nonexporting Firms in India and China Indirect cost India (2006) China (2002) Losses from power outages (% of sales) 5.0 1.9 Losses in transit (% of average cargo value) 0.5 1.0 Unofficial payments to get things done (% of sales) 3.0 1.1 Losses from theft, vandalism, and arson (% of sales) 1.3 0.1 Cost of providing security (% of sales) 0.5 0.5 Source: Manufacturing ICS 2006 for India; Manufacturing ICS 2002 for China. of value added per worker to unit labor cost. This ratio accounts for the fact that value added per worker may not be the only indicator of the competitiveness of a country because a certain level of productivity could be matched by an even lower unit labor cost, which would make the country a competitive location for production. By this measure, China is found to be more competitive in all sectors--both those in which India is gaining world market share and those in which India is not. The poor performance of labor-intensive Indian exports can be explained by the small scale of operations of the vast majority of firms and by stringent labor regula- tions. The virtual absence in India of the large-scale labor-intensive manufacturing firms that are often found in China results in much lower productivity. The reason for this absence is found in the legacy of small-scale industry reservation and in India's labor laws. In fact, labor regulations are reported as major obstacles by 26 per- cent of India high labor-intensity exporting firms, but only 12 percent of other firms. Higher indirect costs also constrain Indian export manufacturing firms more than Chinese firms. Costs borne by exporting manufacturing firms in terms of percentage of sales lost to power outages, losses from theft or vandalism, and bribes paid by firms are larger in India than in China (table 3.4). The investment climate therefore ham- pers the competitiveness of India's export-oriented manufacturing industry. 44 Manufacturing 3 Further analysis of competitiveness of Indian exports is needed. The analysis pre- sented in this section is only preliminary, and it aims to draw attention to the impor- tant challenges that face the Indian manufacturing sector and that need more detailed study. Conclusions Action to improve the investment climate is necessary for manufacturing growth that will support India's growth trajectory and create employment for an expanding labor force. This growth is possible only if both investment and productivity growth in the sector can be increased. In turn, such growth depends on fostering a better investment climate. Investors will be less willing to enter if the general environment is not conducive to allowing them to function in a normal and productive manner. The analysis suggests that almost 76 percent of explained productivity differentials among firms can be traced to the investment climate. Most of these factors, such as infrastructure, the regulatory environment, and the supply of inputs like finance and labor, are beyond the control of individual firms. Improvements in these areas require public sector action. Despite improvements, businesses continue to complain about poor power avail- ability, high taxes, inefficient tax administration, and red tape and corruption. Com- paredwith2003,in2006enterpriseswerelesslikelytocomplainaboutmacroeconomic instability, corruption, and crime. Their perceptions about telecommunications and the availability of finance also improved. This improvement shows that deregulation and other reforms in these sectors have started to have an influence. However, percep- tions have worsened with respect to electricity and tax rates. Another area in which deterioration has been sharp is workforce skills; 24 percent of respondents reported unavailability of required skills in the market, and 34 percent of all firms claimed that it takes too long to find an adequately skilled employee. These areas should now be the focus of attention for policy makers. Improvements in the investment climate can also help in promoting more inclusive growth by narrowing interstate differences.Differences in manufacturing growth have been important drivers of the divergence in economic growth seen in the 1990s among Indianstates.Theanalysisshowsthatimprovementsinthemiddle-incomestatesdrove much of the improvement of the overall perception of the investment climate. The experience of the middle-income states (Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and West Bengal) suggests that this type of improvement is indeed possible. A better investment climate could also enhance the competitiveness of exporting Indian manufacturing firms. When compared with the manufacturing sector in China, the Indian manufacturing sector is not competitive. The lower competitive- ness of Indian firms can partly be attributed to a worse investment climate, as shown by the very high indirect costs faced by Indian firms. The productivity impacts of a poor investment climate are large. Although poor infrastructure often dominates the public discourse, the analysis shows that red tape, 45 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 corruption, and crime had the largest negative impact on firm productivity, real wages, exports, and the possibility of getting FDI. Poor infrastructure came second and accounted for as much as 23 percent of intrafirm productivity differentials. Notably, poor infrastructure availability affected the more productive firms relatively more. Similarly, labor skills, quality, and innovation had smaller effects on produc- tivity but were more important for the more efficient firms in the economy. Poor access to finance and corporate governance had a large impact on employment growth and the possibility of attracting FDI. Notes 1. The 2006 ICS was conducted in connection with this study. See appendix A for more information. 2. These figures are authors' calculations based on data from the 2006 Manufacturing ICS for India, the 2002 and 2005 Manufacturing ICS for China, and the 2005 Manufacturing ICS for Turkey, as well as from the 2006 software and ITES ICS for India. 3. McKinsey Global Institute (2001) looked at labor productivity rather than TFP and identified red tape, corruption, and crime; labor regulations; and infrastructure as major constraints across the four manufacturing sectors covered by the study. Other obstacles were government owner- ship (in the automotive and steel sectors), poor governance of state-owned banks, and lack of enforcement of minority shareholder rights (steel sector). 4. Taxes and contributions are measured at all levels of government and include corporate income tax, turnover tax, all labor contributions paid by the company (including mandatory contribu- tions paid to private pension or insurance funds), property tax, property transfer tax, dividend tax, capital gains tax, financial transactions tax, vehicle tax, and other small taxes (such as fuel tax, stamp duty, and local taxes). A range of standard deductions and exemptions are also recorded. 5. This figure is from Indiastat.com, accessible at http://www.indiastat.com. 6. Spending in the road sector typically represents 75 to 90 percent of the total spending in the transport sector. 46 Manufacturing 3 Annex 3.A Obstacle Case Study--Power The key challenges facing the power sector in India are insufficient generation, limited inter- state and regional transmission capacity, and inefficient and poorly governed distribution lead- ing to low access. Although GDP growth has been impressive, inadequate electricity services and power infrastructure compromise India's poverty alleviation efforts, its economic compet- itiveness, and private investment in energy-dependent industries. Currently, total generating capacity connected to the grid is 140,000 megawatts (out of which 50,000­60,000 megawatts need rehabilitation), but this capacity is insufficient to meet the connected demand for elec- tricity. Captive generation capacity is estimated to be 40,000 megawatts and is generally owned by large firms. Transmission bottlenecks limit interstate trade to 11,000 megawatts, meaning that surplus power from one region cannot easily be transferred to deficit regions. State distri- bution systems are unable to meet demand. Moreover, there are high system losses (around 40 percent of electricity produced is "lost"). As a result, consumers experience significant short- ages and poor supply. Reported peak and shortages (13.8 percent and 9.6 percent, respectively) are higher than five years ago and understate actual shortages (for example, scheduled load shedding is not included). In addition, many people are still not connected; 30 percent of vil- lages and 44 percent of the population still lack electricity. Lack of investment, unpredictability of fuel supplies and pricing, and limited competition have led to erratic and insufficient power supply, which is cited as the biggest bottleneck to gen- eration capacity expansion. Against a target of capacity addition of 41,110 megawatts during the 10th five-year plan (2002­07), only about 21,000 megawatts were added, and the private sector added only 27 percent of what was expected. In addition, unpredictability in quantity of reserves (attributable to a lack of reliable data and the absence of recent geological surveys) and long time frames involved in the commercial supply of fuels, particularly of coal and gas, are serious barriers to scaling up generation investments. Similarly, long-term clarity on principles for pricing these fuels, determined through a transparent regulatory process, is needed. As a consequence, 60 percent of Indian firms rely on captive or backup generation (compared with 21 percent of firms in China). In India, the distribution business is largely state owned and undermaintained because of inadequate financial resources. Only in Orissa and Delhi has distribution been privatized. The distribution business suffers from complex challenges of high system losses, political economy issues in tariff revisions, and agriculture metering. As a result, it is unable to fully recover costs, leading to undermaintenance of assets and insufficient new investments. The public sector util- ities depend on subsidies from the government budget, thereby adding to the fiscal strain. In 2006/07, the losses in the sector were around US$6.2 billion. In addition, weak governance, poor cost recovery, and lack of autonomy are holding back modernization of the energy sector. The system has high distribution losses, and even among relatively advanced reformers (Andhra Pradesh, Karnataka) only 55 to 60 percent of supply is metered. Tariffs are above cost for several consumer categories (industry and commercial entities). However, tariffs are significantly below cost for other categories (agriculture and 47 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 households), which represent the biggest share of the customer base. The subsidy benefits are largely untargeted and enjoyed by all consumers within these subsidized categories, irrespective of their income level; this problem has contributed to the deterioration of the financial posi- tion of the utilities. Moreover, the majority of the power sector continues to be under state ownership (more than 80 percent across the value chain), with existing state-owned entities not empowered to act autonomously (for example, they cannot set up performance-based remu- neration systems or provide market-linked compensation to new employees). The government of India has recognized the severity of the problems in the sector and has begun to design and implement policies to address them. It has underpinned these policies with financial incentives to (a) improve the efficiency, accountability, and quality of electricity distri- bution services (contributing to notable improvements in several states); (b) improve the relia- bility of old thermal power plants; (c) expand investment in renewable energy (growth in wind power is an example); and (d) expand rural access. It has also initiated the preparation of ultra- mega coal-based plants (4,000 megawatts each) for competitive tendering to the private sector. Going forward, the Indian government intends to increase the rate of investment in the sec- tor, enhance market mechanisms, and modernize distribution. It is framing an ambitious plan of attracting Rupees 120,000 crore (US$30 billion) of investments per year from 2007/08, seek- ing to draw an increased portion from the private sector in the form of equity and debt. Under the 11th five-year plan (2007­12), the government expects to facilitate the addition of 78,000 megawatts of generation capacity (including 16,000 megawatts of hydropower), expand inter- state transmission capacity from 11,000 megawatts to 37,000 megawatts, assist states in expand- ing and modernizing their distribution networks, and improve sector governance and finances. Progress in setting up an electricity trading market and in enhancing the entry conditions for large-scale project development (especially the issuance of necessary clearances and easements) before projects are awarded to the successful bidder has lowered the risk for project developers and investors. Moreover, the presence of private distribution licensees in select areas (Ahmed- abad, Greater Noida, Kolkata, Mumbai, and Surat) and the recent introduction of franchisees signal a shift in electricity delivery mechanisms, with an added focus on service quality and reduction of commercial losses. These efforts are beginning to yield results, and the sector's financial performance is showing signs of improvement. Well-structured market transactions for three of the ultra- mega power plants have resulted in significantly more competition than in the past. More vil- lages have been electrified under the 10th five-year plan than under the 8th and 9th plans combined. Availability-based tariffs and tighter performance norms have improved usage factors (plant load factors) of private and central public sector generators. System losses in several states have been brought below 20 percent. First-generation reforms--unbundling, corporatization, and independent regulation--have increased transparency, quality of data, and public awareness of sector performance. Although the level of commitment varies, state authorities are taking the lead in advancing the reform of state utilities and helping state elec- tricity regulatory commissions to strengthen their capacity. Although some states have brought system losses down and improved commercial practices, others have relied more on tariff measures and growth in profitable sales to commercial customers. 48 Manufacturing 3 However, more needs to be done if India's impressive growth rate is to be sustained. The fol- lowing steps are essential: 1. Enhance human capacity across the energy sector value chain. Power sector investments are expected to increase from current levels of around US$7 billion per year to around US$30 billion to US$40 billion per year by 2012. Efficient project execution of such a quan- tum leap, backed by an ability to collect money for these additional investments, requires a skills upgrade in the sector. This upgrade is also important given that a number of utilities have had a recruitment freeze since the 1980s, and senior managers in these utilities are likely to retire in the near future, leaving a management void. 2. Strengthen utility governance and institutional capacity. Most states have unbundled their power sector or corporatized previously vertically integrated power utilities, set up autonomous regulatory agencies, and in some cases even privatized the distribution business (Orissa and Delhi). However, many newly formed utilities (as a result of the unbundling process) are still not empowered to act autonomously and suffer from severe institutional capacity constraints. In addition, their accountability, efficiency, and customer service are well below industry benchmarks. 3. Improve targeting of subsidies. With 40 percent technical and commercial losses, tariff meas- ures need to be accompanied by improved efficiency and accountability, plus a better tar- geting of subsidies to allow distribution companies (especially private ones) to recover their costs and earn an acceptable return on investments. Low return on investments is one of the reasons (apart from the broader issue of political economy) private companies have shown limited interest in this sector, especially in the distribution segment. Source: World Bank staff. 49 Investment Climate Obstacles to Productivity Growth and Employment Generation 3 Annex 3.B Obstacle Case Study--Corruption Although numerous administrative commissions have discussed the problem of corruption, there has been little rigorous assessment of its extent and scope. Microlevel studies that exam- ine corruption across India are relatively few. A recent study of the public distribution scheme cited by India's former central vigilance commissioner found that Rs 50 billion of food grain and sugar meant for distribution was lost to corruption, which amounts to just over one-third of the total food grain subsidy for India. India ranks 88th out of 158 countries on the Transparency International's Corruption Per- ception Index for 2002 through 2005 and 10th among 15 Asian countries, above only Nepal, Vietnam, Pakistan, Indonesia, and Bangladesh. The report (Transparency International 2005) is the first national-level study on the topic. It assesses the perception of corruption across a range of government services (including police, judiciary, land administration, municipal serv- ices, government hospitals, electricity, public distribution systems, income tax, water supply, schools, and rural financial institutions) and tries to quantify the amount of corruption. The survey on which the report is based found that citizens pay bribes amounting to Rs 210.68 mil- lion while accessing one or more of these services. As many as 62 percent of the citizens reported firsthand experience of paying bribes and using "contacts" to complete a task at a government office, and three-fourths of citizens believed that corruption had increased in public life over the previous year. Need-based services are more corrupt than basic services. The study concludes that need- based services (such as income tax, judiciary, and police) are more corrupt than basic services (such as health and education), thanks to the monopolistic nature of provisioning and the fact that officials connected with need-based services are vested with substantial powers, which makes the repercussions of not paying bribes more serious. According to the study, of the serv- ices included in the survey, the police in India are seen as the most corrupt, with 80 percent of respondents claiming they had paid bribes. In the judiciary, 41 percent of those who paid bribes did so to influence judgments, 31 percent to speed up or delay judgments, and 28 percent to get routine jobs done, such as listing of cases and gaining copies of documents. According to the survey, the total monetary value of petty corruption is more than Rs 210 billion annually, or approximately 0.75 percent of current national GDP. For perspective, consider that this amount, if distributed equally among the poor, could raise the annual income of each poor per- son by approximately Rs 850 and approximately halve the national poverty headcount ratio. Many incentives in the system facilitate corruption. Although human greed is the obvious driver of corruption, structural incentives, such as the following, facilitate it: · Complex and nontransparent administrative systems of command and control. For example, official bodies responsible for investigating corruption allegations are typically part of the administration and thus are widely perceived to lack independence and autonomy; investi- gation of senior officials requires permissions that amount to interference; and resources invested against corruption are too limited. 50 Manufacturing 3 · Poor enforcement systems. For example, the Central Vigilance Commission, which is the main agency of enforcement in the central government, remains constrained by the limitations imposed on it by the Central Vigilance Act and the jurisdiction of the government of India under the constitution, wherein law and order is a state subject. One result is that the Central Bureau of Investigation has filed more than 6,000 cases, of which about 50 percent have been pending for more than five years, while the criminal rate of conviction in India is about 6 percent. · Weak notion of citizens' rights. For example, laying down a statutory right to information has been a very significant reform in public administration in India, but its implementation across the country is facing a number of constraints, given the vested interests at every level. Sources: Chand 2006; Transparency International 2005. 51 Chapter 4 Unorganized Manufacturing Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity The unorganized manufacturing sector in India employs a vast majority of the man- ufacturing workforce, but it is characterized by low productivity and wages. A survey conducted for this study shows that there is considerable potential to improve both employment and productivity in this sector. Access to finance remains the major constraint to the operation of firms. As a result, enterprises are credit constrained and have difficulty investing in machinery and equipment. Analysis of the survey results shows that firms that are able to borrow from formal sources have 37 percent higher labor productivity than firms that cannot. Lack of a reliable power supply is the second-largest constraint. Firms get around this problem by employing more workers but pay a price in terms of profitability. Importance of Unorganized Manufacturing Unorganized manufacturing employs the vast majority of India's manufacturing workforce. In 2000/01, almost 99 percent of manufacturing enterprises and 74 per- cent of the manufacturing workers were in the unorganized sector (see box 4.1 for a definition of unorganized sector in India). Employment and wages in the unorgan- ized sector grew over the 1990s, while they stagnated in the organized sector (Malhotra and others 2006). Unorganized manufacturing is also more evenly distributed regionally than organized manufacturing. Large-scale organized manufacturing activity is highly agglomerated in a few areas and has become even more concentrated since the reforms of the 1990s. In 2005, 10 districts alone accounted for 49 percent of national investment in manufacturing (Chakravorty and Lall 2006). New investment has favored states with large manufacturing bases (Bhaumik, Gangopadhyay, and 53 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 Box 4.1 What Constitutes Unorganized Manufacturing in India? The most commonly used definition of the unorganized sector--the definition used in this report--flows from the Indian Factories Act, 1948, which requires that an enterprise register with the state government if it employs 10 or more workers and uses power or employs 20 or more workers and does not use power. Units that do not come under the purview of this law constitute the unorganized sector. Businesses that are organized are required to comply with health, safety, and welfare require- ments. Firms are required to contribute toward insurance against sickness, disability, and mater- nity and to deposit linked provident funds or pension schemes. The state monitors compliance through a system of inspections. Unorganized firms largely fall outside this system, except for a few omnibus regulations pertaining to use of child labor, minimum wages, and workplace safety. Krishnan 2006), existing industrial clusters, and locations with easy access to the coast (Chakravorty and Lall 2006). Low Productivity, Low Wages, Weak Skill Levels, and Limited Integration with Large Supply Chains in Unorganized Manufacturing Unorganized manufacturing performs worse than organized manufacturing in terms of productivity and, hence, jobs and wages. Differences in performance between organized and unorganized firms are to be expected, but the peculiarity of India is the large productivity differential between the largest and the smallest groups. These differentials are of the order of eight to one, compared with three to one in Japan before its post­World War II era growth or in the Republic of Korea during its early stage of development (World Bank 2007b). Productivity is often a good proxy for assessing how well enterprises use existing knowledge. Firms with higher productivity presumably have absorbed or developed superior production and management technology. Thus, absorption needs appear greatest among small enterprises. Productivity differentials are also reflected in wage differentials. Little, Mazumdar, and Page (1987) estimate that after controlling for human capital attrib- utes, workers in larger enterprises earn wages three times higher than those of simi- lar workers in the unorganized sector. Per capita output across states varies more in organized manufacturing than in unorganized manufacturing (figure 4.1). This finding may be related in part to the fact that the relatively lower capital requirements for starting an unorganized enter- prise allow a larger pool of entrepreneurs to start businesses. It also may be that agglomeration externalities (that is, the gains from concentration in a cluster) are less important for the smaller manufacturing firms. Nonetheless, interstate differences in the size of unorganized manufacturing are still quite large, suggesting 54 Unorganized Manufacturing 4 Figure 4.1 Per Capita Manufacturing Output, Organized and Unorganized Manufacturing, 2004 6,000 5,000 4,000 capita 3,000 per Rs2,000 1,000 0 organized unorganized low-income states middle-income states high-income states Sources: Central Statistical Office and World Bank estimates. that state-level policies and the prevailing investment climate have an effect even on this sector. To understand what hampers the productivity of the segment and whether firms avoid becoming organized and growing, an enterprise survey was conducted in 2006. The Unorganized Manufacturing Investment Climate Survey (ICS) covered 1,500 enterprises in five major industrial clusters in India: the Delhi­national capital region, Howrah, Hyderabad, Ludhiana, and Mumbai (see appendix A for a full description of the data and methodology). Enterprises have been divided into three subcategories: household (enterprises operating out of a home and employing hired workers), tiny (enterprises employing fewer than 6 workers), and micro (enterprises employing 6­10 workers). All the data presented in this chapter refer to data col- lected through this survey except when indicated otherwise. The average age of firms in the sample was around 13 years, though there was some variation across industries. At the same time, more than 50 percent of firms in the sample were less than 10 years old, which suggests that a large number of firms tend to fail within the first 10 years. The data also show that a significant proportion of older, more established enterprises tend not to grow above the 10-worker thresh- old. More than 90 percent of the firms in the sample were single proprietorships, and the remaining 10 percent were partnerships. The incidence of female ownership was low in the sample group--only around 7 percent--and was highest among house- hold enterprises.1 Most enterprises were operated out of fixed premises, with unclear land titles being less of an issue. In India, a nationwide survey of unorganized manufacturing in 2001 conducted by the National Sample Survey Organisation (NSSO) found that almost 94.5 percent of urban unorganized manufacturing enterprises operated out of permanent structures and from fixed locations (NSSO 2001). The current ICS 55 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 also found that most enterprise owners either owned (51.3 percent) or leased (47.5 percent) their premises. Out of the total sample of 1,500 enterprises, there were only six cases in which ownership of land was reported to be unclear. Small enterprises catered mostly to other small firms and were not well connected with larger firms. Links between large and small enterprises can lead to productivity gains through technology transfer. Traditionally, however, the links between the organized and unorganized segments in India have been weak. Much of the manu- facturing activity in the unorganized sector is geared toward producing final prod- ucts for the consumer market rather than intermediate products and parts for the organized sector (Papola 1991). ICS results show that only 4.6 percent of firms reported that their main clients were larger firms. For 80 percent of firms, the main clients were either other small firms or individuals. This situation could also be explained in part by the small-scale reservation policy, which aimed to expand employment opportunities and production of consumer goods in the small sector. On the basis of this policy, the government restricted the entry of large players for as many as 873 items. The government has been phasing out this policy, and 35 items are still on the reserved list. Nevertheless, reversing the impact of the policy on links between firms will take some time. With respect to education, most owners had not received formal training, the workforce in the enterprises was reasonably well educated, and part-time workers were used to deal with seasonality. In the ICS, 42 percent of owners were high school educated, and another 19 percent had general graduate or postgraduate degrees. Only 3 percent had professional graduate degrees. Formal vocational training was not common: only 12 percent of owners had received any formal training, mostly in technology-intensive industries, such as electrical goods and chemicals (figure 4.2). At the same time, the average manager reported around 16 years of experience, which did not vary much by industry. Years of managerial experience in the sector generally exceeded the age of the firm, suggesting that most owners had some expe- rience before they set up their own enterprise. Half of the enterprises (50.3 percent) reported that their average worker had 7 to 12 years of education (that is, a second- ary school­level education). On average, female, full-time employees constituted less than 5 percent of the total workforce, with the highest percentage in the garments and electronics industries. Family labor was important only in the garments and leather industries, and it was lowest in the chemicals industry (figure 4.3, panel a). Activity in many sectors varied over the year, and firms seemed to use part-time workers to deal with this seasonality.2 There is a very close correlation between reported output variability and the use of part-time workers (figure 4.3, panel b). Benefits of Becoming Organized and Why Some Firms Remain Unorganized Being part of the organized sector, often referred to as the formal sector, has its benefits. However, international experience suggests that, despite the advantages, 56 Unorganized Manufacturing 4 Figure 4.2 Management Qualifications and Experience, by Industry, 2006 30 30 25 25 20 20 managers of 15 15 experience of 10 10 years percentage 5 5 0 0 auto food and goods leather tools textiles chemicals garments electronics metal components processing electrical machine managers with professional graduate degree managers with formal vocational training managers' years of experience in sector Source: Unorganized Manufacturing ICS 2006. unorganized enterprises often wish to remain informal. Firms that are not registered with government authorities often suffer because of lack of legal status and bargain- ing power. Consequently, they may have less access to finance or can get it only at a high cost. They may also face difficulties in accessing business services, such as insur- ance or legal services. However, firms often prefer not to register or grow because they must then pay taxes and be subject to environmental and health and safety reg- ulations. Sometimes larger firms are also more subject to bribes as they become more visible. In India, the 10-worker threshold is especially important because labor laws on wages and benefits are applied to units above this critical size. Enterprises gradu- ating out of the unorganized sector thus face extra costs and lower flexibility, which encourage them to remain under the limit. Firms can avoid being part of the organized sector in two ways: (a) by operating without detection by simply not registering or (b) by not growing. In India, firms with more than 10 employees must register with the state government if they are to comply with labor laws. In addition, all businesses need to register with the munici- pality to get industrial and electrical connections and to pay excise taxes. Nonetheless, a large number of firms were registered with some government authority and hence were not strictly beyond detection. Registration with tax or municipal authorities confers a measure of visibility to firms. About 41 percent of surveyed household enterprises reported being registered with a municipal agency, and 38 percent reported being registered with a tax authority. Almost three-fourths of microenterprises were registered with some municipal agency, with the relevant tax authorities, or with both (figure 4.4). This finding is at variance with the NSSO survey, which found overall that 80 percent of unorganized enterprises were 57 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 Figure 4.3 Full-Time and Part-Time Workforce, by Industry, 2006 a. Full-time workforce 7 6.5 6 5.5 5 workers of 4.5 4 number 3.5 3 2.5 2 goods and leather textiles garments tools components chemicals electronics processing metal auto electrical food machine male paid workers female paid workers working family members b. Ratio of part-time to full-time workers 120 100 (%) firms 90 100 80 reporting workers 70 80 60 variable full-time 60 50 to 40 40 activity 30 part-time 20 20 of (%) 10 ratio 0 0 and goods tools leather textiles chemicals metal garments components processing electronics auto electrical machinefood ratio of part-time workers to full-time workers firms reporting variable activity Source: Unorganized Manufacturing ICS 2006. 58 Unorganized Manufacturing 4 Figure 4.4 Surveyed Firms Reporting Themselves as Registered, 2006 100 90 80 70 firms 60 of 50 40 percentage 30 20 10 0 household tiny micro enterprises enterprises enterprises registered with any municipal agency registered with tax authorities Source: Unorganized Manufacturing ICS 2006. unregistered (NSSO 2001). The apparent difference between the two surveys may be because the ICS covered enterprises in major industrial clusters in urban locations, whereas the NSSO survey also covered rural enterprises. The chosen clusters were also often in states that did better even in the NSSO survey. Interestingly, most of the enterprises that reported themselves as not registered (83 percent) claimed that they did not think that they were required to do so. These two points suggest that a sig- nificant number of eligible enterprises are willing to be registered. Moreover, the vast majority of unorganized establishments declare most workers and sales for tax purposes. Between 64 percent and 78 percent of unorganized firms in the ICS reported that a typical establishment declares 80 to 100 percent of sales, and between 77 percent and 89 percent of those firms reported that a typical estab- lishment declares 80 to 100 percent of the workers. This result, which is somewhat at odds with the common perception of the unorganized sector, may reflect, in part, the fact that the sample was drawn from large industrial clusters. The surveyed firms may be more growth oriented and less survivalist than the average unorganized firm--and possibly more visible than enterprises located outside such clusters. Firms that did not register failed to do so mainly because they lacked information or because the process took too long. Despite knowing that they were required to register, the largest percentage of firms attributed their unregistered status to lack of information about the process. Firms also did not register when it was costlier for them to do so, in terms of either resources or time (the average time to obtain an operating license varied from a low of five days in Ludhiana to about 67 days in Delhi). Registration rates were strongly negatively correlated with this timing (cor- relation of 0.58). Finally, a small proportion also reported that they failed to regis- ter because they wanted to avoid paying taxes. 59 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 A large number of enterprises started on a small scale and grew into organized manufacturing. The survey instrument for organized manufacturing firms asked respondents the workforce size when the enterprise began its operations, as well as the current workforce size. This information provides an opportunity to track firms' growth: a relatively large number of organized enterprises started life as small units, around 30 percent of organized enterprises started as microenterprises, 16 percent began as tiny enterprises, and 1 percent began as household enterprises. Unorganized enterprises also stand to benefit from scaling up. An analysis of Indian firms' productivity and employment from the Unorganized Manufacturing ICS shows that a 1 percent increase in labor productivity, if the rest of the variables are kept constant, will generate a 0.91 percent increase in employment hours (see appendix B). The implication is that firms that become more productive tend to expand and employ more people. The survey shows that older, more established enterprises do tend to grow (figure 4.5). Around 58.5 percent of the interviewed firms reported having grown in terms of employment from the time they started operations.3 On average, growing enterprises added 2.6 full-time, paid workers to their payroll. The investment climate is an important factor affecting the potential productiv- ity growth of unorganized manufacturing. The productivity analysis shows that investment climate and control variables explain 19 percent of the variation in labor productivity. This relatively low contribution of investment climate variables reflects the fact that the explanatory power of the model is limited because unorganized manufacturing firms, thanks to their size, can operate outside of the investment climate and because unobservable factors, such as labor and managerial quality, play Figure 4.5 Enterprises Reporting Employment Growth, by Age, 2006 70 60 (%) 50 growth 40 30 reporting 20 firms 10 0 young mid-age established (1­5 years) (6­15 years) (more than 15 years) number (%) percentage growing Source: Unorganized Manufacturing ICS 2006. 60 Unorganized Manufacturing 4 a big role in the productivity of the sector. Although these factors are important for the sector, they are unobservable and are covered by other reports. Therefore, they are not discussed in the subsequent analysis, which instead focuses on the investment climate. Access to Finance and Power Supply: Largest Constraints to Higher Productivity and Growth By a wide margin, firms surveyed identified access to finance as the single biggest obstacle to their growth (box 4.2). More than a third of respondents identified this issue as their single, largest concern (figure 4.6). The availability of electricity was the Box 4.2 Understanding the Problem of Access to Finance The vast majority of the enterprises surveyed for the Unorganized Manufacturing ICS 2006 had a bank account, but very few had a loan and, of those that did, very few received it from a bank. Indeed, 89 percent of small enterprise owners had accounts in banks, and 58 percent of them used those accounts for business purposes. However, only 20 percent of unorganized enter- prises borrow from either formal or informal sources. Of these enterprises, fewer than 10 per- cent had taken either a term loan or an overdraft from a bank (see the accompanying table). Proportion of Enterprises with a Bank Account, by Enterprise Size, 2006 Enterprises Enterprises with loans with no Enterprises from any Enterprises loans from Borrowing with financial with loans Type of any source enterprises informal institution from banks enterprise (%) (%) loans (%) (%) (%) Household 79 21 15 8 6 Tiny 83 17 11 8 7 Micro 78 22 13 10 9 Source: Unorganized Manufacturing ICS 2006. Getting a loan from a bank is a lengthy endeavor, and it requires large amounts of immovable collateral. On average, it took surveyed enterprises 25 days to get a loan from a bank after all the documents had been submitted. Usually, an immovable asset of the enterprise (60 percent of cases) or personal asset of the owner (25 percent of cases)--and likely immovable assets as well--was required as collateral. At the same time, immovable assets accounted for a relatively small percentage of total asset value of the unorganized firms. Moreover, in about half of these cases, the collateral was more than double the amount of the loan. (continued) 61 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 Box 4.2 continued Informal finance is preferred for its flexibility. Interviews with a focus group suggested that informal borrowing offers features that the formal banking sector in India does not offer. Infor- mal loans do not require any sort of collateral, and even where it is required, the loan is usually made against accounts receivable and inventories. Most important is the flexibility that informal sources provide in determining the payment schedule (small fixed installments collected daily or weekly, rather than large monthly payments); in structuring or restructuring repayment schedules and amounts; and in taking account of past credit history with the informal lender. Flexibility is especially important for small enterprises with seasonal activities and cash flow. However, neither formal nor informal loans sufficiently meet the needs of the unorganized sector. Formal credit markets are difficult for small firms to access, while informal loans are based on personal relationships, are short term, and often attract high interest rates. Most enterprises surveyed preferred not to expand their activities until they had built up internal resources: 93 percent of enterprises said they usually used internal funds for making fixed investments, and 99 percent financed their working capital from the same source. Only 7 per- cent of enterprises had borrowed any amount from banks for working capital requirements, and just 9 percent had borrowed for a fixed investment. The inability of the formal financial system to meet the needs of this sector has implications. First, it places a limit on firms' ability to invest in productive growth opportunities. Second, tax laws in India are such that the highest burden falls on investment financed out of own funds. Hence, inability to access debt from formal institutions places small enterprises at a significant disadvantage. second, with 18 percent of enterprises saying that this was the most significant con- straint. These findings contrast with the responses from larger firms in the manu- facturing sector (as discussed in chapter 3), where availability of electricity is the largest issue. Among other constraints highlighted by the unorganized sector, mar- ket conditions, such as competition and volatility of demand, were considerable problems. Other factors appeared less important, though 7 percent of firms surveyed felt that tax rates and tax administration posed a major problem for them. The find- ings of the ICS echo those emerging from the NSSO (2001) survey. Access to formal finance is associated with significantly higher labor productivity, and informal finance is a poor substitute for productivity- and growth-enhancing investments. The econometric analysis of the effect of investment climate on unorganized firms also shows that firms that were able to borrow from financial institutions had 37 percent higher labor productivity than firms that could not, after controlling for other factors (see appendix B). This factor possibly reflects that cash- constrained unorganized enterprises use these short-term financial instruments to supplement their own resources for financing new productivity-enhancing fixed investments. Borrowing from informal sources is not strongly associated with any similar positive effects.4 The difference may be because the bulk of informal loans 62 Unorganized Manufacturing Figure 4.6 Major Constraints to Normal Functioning Identified by Unorganized Enterprises 45 infrastructure red tape and crime inputs market conditions 40 38 35 30 firms of 25 22 20 20 18 16 14 15 15 percentage 10 9 7 5 6 7 5 2 3 3 3 2 0 0 0 0 1 1 0 0 0 1 1 0 to in and theft, and land labor to electricity transport permits corruption crime, disorder access functioning courts rates finance regulationsvolatilitydemandcompetition business of inadequately and tax access educated administration licensing workforce telecommunications firms report as a major obstacle firms report as the largest obstacle Source: Unorganized Manufacturing ICS 2006. 63 4 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 Figure 4.7 Effect of Power Outages on Unorganized Firms 60 100 90 50 80 70 40 60 30 50 hours 40 percentage 20 30 20 10 10 0 0 Delhi Ludhiana Mumbai-Thane Howrah Hyderabad firms with recourse to backup (%) sales lost (%) hours of power outage per month Source: Unorganized Manufacturing ICS 2006. (93 percent) are used for financing working capital and maintaining operations rather than for making fixed investments. A variety of reasons may limit their use- fulness in financing the latter. Informal loans are likely to be of shorter duration, and they may not be available at the time they are needed or in the amount required. Consequently, their effect on labor productivity is lower. The result is in line with international evidence that suggests that informal lending can only partially substi- tute for formal lending in very poor countries. Power availability is another major constraint for small enterprises, with arguably greater impact on growth and productivity. As with the larger manufacturing firms in the ICS, unorganized firms also had to contend with power outages. Power outages averaged as high as 89 hours (about nine working days) a month in Ludhiana. The relatively small scale of operation probably limited recourse to expensive backup options, such as generators and invertors, and few enterprises reported having them (figure 4.7). Outages were costly: on average, enterprises reported huge losses--11.3 percent of annual sales--because of outages. One must also keep in mind that unreliable power supply is likely to deter firms from growing and investing in new technology to the extent that such technology requires assured power supply. Hence, unreliable power sources might keep firms trapped in a lower- scale, low-productivity, and labor-intensive mode of production. Nonetheless, power outages appear to have no effect on productivity, possibly because they push enterprises to adopt coping strategies based on employing more labor. Anecdotal evidence suggests that firms deal with outages by either employing more workers or enforcing longer shifts in order to finish orders. This hypothesis is borne out in the data to the extent that power outages are not negatively correlated with employment: more power outages lead enterprises to employ more labor, if other factors are held fixed. However, as discussed earlier, the loss of sales as a result of power outages is still quite significant at more than 11 percent. 64 Unorganized Manufacturing 4 Almost no enterprise identified transportation, telecommunications, law and order, the functioning of the courts, or labor force skills as a severe problem in the operation of their establishments. In some areas, such as telecommunications, this identification owes to the rapid progress India has made, especially in urban areas. In other areas, such as transportation or courts, it may reflect that these factors are not very important in the context of small firms. Unorganized firms probably do not sup- ply across large distances and may rely only on local transport infrastructure. Simi- larly, very few firms ever use courts to settle disputes (only about 1 percent). Unlike larger firms, unorganized firms also do not rate labor regulations as a major problem. Again, this finding is to be expected because small enterprises are outside the purview of most labor laws. More than 54 percent of firms did not know or could not answer a question about which labor regulations they found most difficult to follow. Although transport and corruption are not identified as a constraint by firm management, empirical analysis shows that both the short-term impact of transport problems and the long-term impact of corruption on productivity are considerable. Although only about 9 percent of firms reported difficulties in making timely deliv- eries to customers or receiving supplies because of transport-related issues, those that did report such difficulties had about 13 percent lower labor productivity. These firms also employed more workers, possibly to cope with these difficulties, but suf- fered nonetheless. Similarly, only 3 percent of firms ranked corruption as their fore- most problem, and 5 percent ranked it as one of their three biggest problems; bribes were a small percentage of annual sales. This finding was expected because invisibil- ity is one known advantage of small firms. However, corruption may also induce firms to remain small, which could have long-term impacts on their productivity. The amount of bribes a firm pays can be expected to be closely related to the num- ber of inspections it faces, which, in turn, is closely related to its annual turnover, the number of workers it employs, and whether it is registered. This situation creates incentives for the firm to remain small, unregistered, and inconspicuous. The analy- sis also shows that when the reverse link between bribes and productivity is con- trolled for, paying bribes has a negative effect on firm productivity. A 0.1 percent increase in the amount of bribes paid (as a ratio to annual sales) is associated with a 3.4 percent decrease in the firm's labor productivity. Conclusions There is significant scope for improving productivity and employment in unorgan- ized manufacturing. An analysis of survey results shows that firms in this sector face increasing returns from scaling up and that they thus have a strong incentive to grow. Consequently, many firms appear willing to grow but are held back by investment climate obstacles. Some successful firms, in fact, do grow--some even beyond the 10-worker limit that pushes them into the organized sector. At the same time, a very large number do not. The survey provides evidence of why they do not grow. Access to finance is the biggest investment climate obstacle facing firms. This finding is apparent both from the firms' own perceptions and from the analysis of 65 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 their productivity and employment. Firms that are not able to access financing from formal financial institutions have almost 37 percent lower productivity than those that can access such financing. Interest rates are not the problem. Enterprises are deterred from seeking such loans because of stringent collateral requirements, nec- essary documentation, and the time it takes to process their application. Loans from informal sources are inadequate substitutes; they are more suited to meeting work- ing capital requirements than to making fixed investments. Enterprises are thus forced to rely mostly on internal funds for such investments. Improving the situation does not require the government's direct intervention in the form of subsidies, but rather its facilitation of the development of an enabling financial infrastructure and its efforts to build good institutions. Enterprises are also hampered by poor infrastructure and by their limited ability to invest in mitigating mechanisms. Expectedly, power is a problem, although a much smaller one than finance. In this respect, the unorganized sector differs markedly from organized manufacturing. Enterprises cope sometimes by investing in costly backup sources, but mostly by employing more workers and presumably enforcing longer working hours. The transport infrastructure is less problematic, and almost no firm in the survey identified it as a major constraint. However, econo- metric analysis shows that firms that reported being unable to make timely deliver- ies had significantly lower labor productivity. The high losses incurred because of poor infrastructure (more than one-tenth of annual sales because of power alone) reduce profitability and, consequently, firms' incentives to invest and to improve scale and productivity. The various constraints discussed in this chapter may interact with each other, and action may need to be taken simultaneously on many fronts. For example, unless enterprises receive adequate financing, they will be limited in their ability to make productivity-enhancing investments. However, unless they also have access to a reli- able power supply, the gains from investing in technology will also be lower. Improv- ing access to finance thus will have only limited effect. If such investments make enterprises more vulnerable to rent-seeking, their incentive to grow will be reduced. A better understanding of these factors, the ways they interact with each other, and their impact on technology adoption and productivity is thus very necessary. Notes 1. These results match the National Sample Survey 56th Round Survey, which found that 97 per- cent of urban, nonhousehold enterprises were proprietary, and of these only 5 percent had female owners. These results are also consistent with surveys in other developing countries, which found that firms operating out of the home were overwhelmingly female owned and that enterprises in commercial districts, industrial sites, and traditional marketplaces were mostly male owned (Ingram, Ramachandran, and Desai 2006). 2. Part-time workers are both those who work fewer than eight hours per day on all workdays and those who work for only a part of the year. 3. Employment refers to the number of paid, full-time employees. 4. The effect is positive but is not statistically significant at the 10 percent level. 66 Unorganized Manufacturing 4 Annex 4.A Obstacle Case Study--Access to Finance The government of India's fiscal policies often crowd out lending to the private sector, hence limiting firms' access to finance. Although access to finance is an issue for all businesses, the problem is more acute for small businesses. High fiscal deficits have led the government to appropriate a large share of financial savings for itself, thus displacing credit to the private sec- tor. As of 2006, total loans and advances for scheduled commercial banks amounted to 54.4 per- cent, with most of the other assets kept in liquid assets, including government securities. As a result, the ratio of private credit to gross domestic product is less than 40 percent (compared with more than 100 percent for such countries as the Republic of Korea and Malaysia), and much of this private credit takes the form of relatively safe lending to large corporations and consumer loans to high-income individuals. In fact, although banks are required to direct 40 percent of their lending to the "priority sector," which includes small firms, most get around this requirement by subscribing instead to other eligible instruments (for example, bonds issued by apex banks, such as the National Bank for Agriculture and Rural Development and the Small Industries Development Bank of India) (Reserve Bank of India 2005). Competition in the small business segment is still quite limited, and financial institutions have not yet devel- oped appropriate products and procedures to serve the segment profitably, which requires large volumes, efficiency, and very high portfolio quality. In addition, the legal regulatory regime, although improving, still does not facilitate lending to the extent that it could, especially for small businesses. Access to finance is also hampered by the high costs and lengthy enforcement procedures for immovable collaterals. Stamp duties for mortgage registration vary across states, ranging from 3 percent to 14 percent. At the same time, enforcement procedures for registered mortgages can take up to four years.As a result, the majority of the mortgages are unregistered (so-called equi- table mortgages). Under an equitable mortgage, the debtor deposits the title deed with the cred- itor and this act creates the security. The mortgage might be a purely oral arrangement, and in most states no registration or stamp duty is required. However, because the title deeds are phys- ically deposited with the creditor, the debtor can have only one mortgage on the entirety of the assets. Hence, businesses cannot fully leverage their collateral, leading de facto to a higher level of collateralization. Despite improvements in the legal regulatory framework, it is still difficult for small busi- nesses to use movable assets as collateral. The movable collateral regime for financial institu- tions is regulated by the Securitization and Reconstruction of Financial Assets and Enforcement of Security Interest Act 2002 (SRFAESI), which substantially improved the enforcement of securities over movables because it introduced out-of-court execution. SRFAESI also mandated the creation of a central registry for liens on movable assets, thereby making it easier, in theory, for the creditor to conduct searches. In practice, no central registry has been created. Hence, as in the past, liens on movable assets of incorporated businesses are registered in the company registry, except for liens on assets that must be registered in a particular registry (for example, 67 Facilitating Entry to the Formal Sector by Removing Obstacles to Growth and Productivity 4 cars). SRFAESI has not improved the situation for small businesses, many of which are unin- corporated and have mainly movable assets (as opposed to immovable assets). In fact, SRFAESI does not mandate any registration for assets pledged by unincorporated businesses (for exam- ple, partnerships and sole proprietorships), which make up the bulk of small businesses. This circumstance means that the registration is not a requirement for perfection of a security inter- est over a movable asset of an incorporated business. Hence, a lender has no means of conclu- sively determining whether the asset being offered as collateral has already been pledged. The effect of this ambiguity is to leave small businesses underleveraged because their assets cannot safely be offered as security. The establishment of the Credit Information Bureau (India) Limited (CIBIL) and the approval of the Credit Information Companies Act have substantially improved the credit information system. However, much remains to be done to enhance the quality of the data col- lected and to facilitate a more extensive use of the bureau. A number of facts explain the lim- ited reliability of the data collected by CIBIL. First, the data provided to CIBIL by financial insti- tutions, especially state-owned organizations, are of variable quality and are often submitted a few months after the loan has been issued. Many state-owned banks, which account for the greater part of the retail financial sector, are not yet fully computerized. As a result, transcrip- tion errors occur often. Lack of computerization also affects the timeliness of the data reported by this segment of banks, which often report on only a quarterly basis. Second, some operating rules of the registry make it difficult to verify information. For example, there is a very high level of anonymity in the credit reports (for instance, the name of the reporting bank is not included in the report). Unlike other credit bureaus around the world, CIBIL does not verify information using third-party sources, such as courts, government registration departments, or company registries. Finally, in India, there is no unique identification number for customers. Although these issues affect the reliability of the data on all borrowers, the problem is particu- larly acute for small businesses, which are mainly served by state-owned banks. Under the Credit Information Companies Act, credit bureaus can collect information only from financial institutions. This provision disadvantages small businesses, which often have limited credit histories with financial institutions. Unlike other bureaus, CIBIL can collect information only from financial institutions and not from utilities or mobile phone companies. This provision excludes important sources of credit history data, especially for small businesses, which often have limited credit history with financial institutions but are users of mobile phones and clients of utilities companies. Source: World Bank 2004a. 68 Chapter 5 Retail Sector Facilitating the Emergence of Large Players to Boost Productivity and Employment This chapter analyzes productivity and employment growth in India's retail sector. The sector is composed of small retail stores at one extreme and large modern- format stores at the other. The analysis shows that larger stores are more productive and have had a faster rate of growth in employment in recent years. Thus, removing obstacles--especially power outages, taxes, corruption, and lack of access to land-- will facilitate the productivity of the largest stores and contribute to growth and employment generation for the entire sector. The chapter presents the results of the nationwide Retail Investment Climate Survey (ICS) 2006 and data from secondary sources. Two Sharply Contrasting Segments: Small Stores and Large Chains The retail sector comprises two sharply contrasting segments: the large chain outlets and a multiplicity of small, traditional retail stores. Retail in India is highly fragmented and is dominated by the traditional retail sector--that is, retail shops run by the owner or the caretaker of the store. These shops include most of the fast-moving consumer goods (FMCG) stores.1 India has one of the highest retail densities in the world-- around 6 percent,or 12 million retail shops for 209 million households (Ernst &Young 2006). The dominance of the traditional retail sector is also evident from its share in total retail sales in India, accounting for 97 percent of such sales in 2005 (Ernst & Young 2006). Henceforth, the terms traditional retail sector and FMCG stores will be used interchangeably, as will modern-format stores and large chain outlets. To understand what hampers the growth of the sector, the World Bank conducted a nationally representative survey in 2006 in connection with this report. The Retail ICS 2006 covered 1,433 outlets in 41 cities (see appendix A for a full description of 69 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 the data and methodology). All the data presented in this chapter refer to data col- lected through this survey except when indicated otherwise. Small stores are mainly sole proprietorships (more than 93 percent). Small stores have, on average, two workers, usually the owner and an employee. The average small store has been operating for 13 years. This period coincides with the average num- ber of years of experience of managers of small stores, suggesting that such stores have been managed since their inception by the same individual. Although still limited in size, organized retail is growing fast. Modern-format stores began appearing in South India in the late 1980s. Though such stores accounted for only 3 percent of total sales in 2005, the segment is expanding rap- idly (Ernst & Young 2006): between 2003/04 and 2006/07, total sales by small retail- ers grew by 11 percent per year, while sales of modern-format stores grew by about 20 percent per year (Joseph and others 2008). The Retail ICS shows that growth in sales is strongly correlated with growth in employment, with a correlation between the two of 0.88. Modern-format stores in urban areas employed, on average, 37 employees, and 37 percent of such stores are publicly listed companies. Private com- panies, partnerships, and sole proprietorships represent the balance of the modern- format stores. Sector Performance: Low Productivity and Modest Employment Growth Although the retail sector has been growing, its growth has been slightly below aver- age when compared with that of the wider services sector. The trade sector, includ- ing both retail and wholesale, has been growing at 7.7 percent over the period from 1990/01 to 2006/07, while the average rate for the services sector has been 7.9 per- cent (figure 5.1). Moreover, the growth of employment in the trade sector has been around 2.7 percent per year over the period from 1999/2000 to 2004/05--a modest rate when compared with that of the entire services sector (7.6 percent). Labor productivity in the retail sector is low. A study (McKinsey Global Institute 2001) comparing labor productivity in the Indian retail sector with that in the U.S. retail sector found that Indian levels of productivity were only 6 percent of U.S. lev- els (see figure 5.2).2 The United States was used as a comparator because labor pro- ductivity in the services sector there is among the highest in the world. In addition, India's average labor productivity in retail, as a percentage of U.S. productivity in the same sector, fares badly when compared with other Indian services sectors, such as telecommunications, software services and products, and banking (figure 5.3). Low labor productivity is attributable to the large presence of small, less produc- tive stores. In line with secondary data, the results of the labor productivity analysis for the sample show that traditional FMCG stores are the least productive types of outlets. The productivity of the average FMCG store is only about half that of the average modern-format store, or US$3.60 per hour worked versus US$7.60 (figure 5.4). 70 Retail Sector 5 Figure 5.1 Growth Rate of Selected Services Sectors, 1990/01­2006/07 communication 20.0 hotels and restaurants 9.7 banking and insurance 9.8 transport by other means 8.1 trade 7.7 storage 2.3 railways 4.8 services (overall) 7.9 0.0 5.0 10.0 15.0 20.0 25.0 growth rate (%) Sources: World Bank 2004b; Central Statistical Office 2007. Note: The figure reflects constant prices and compound annual growth rates. Figure 5.2 Labor Productivity in Retail Relative to the United States, 2001 United States 100 Poland 24 Brazil 16 India 6 0 20 40 60 80 100 labor productivity (purchasing power parity per labor hour) Source: McKinsey Global Institute 2001. Note: For developing countries, years of productivity vary, depending on the year in which the McKin- sey Global Insight study was conducted in the country. The labor productivity decomposition shows that the most productive outlets in India are the single stores with the largest market share. These findings are consistent with the results of the productivity analysis. The aggregate productivity, which meas- ures the total productivity of the sector, is the sum of the efficiency terms and the average productivity. The efficiency term measures the efficiency in the allocation of resources: a large and positive efficiency term (61 percent of the aggregate produc- tivity for the retail sector) indicates that in India the most productive retailers are the individual stores with the largest market share. The overall productivity of the retail sector can therefore be increased by letting the most productive outlets acquire an even larger market share. 71 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Figure 5.3 Overall Productivity in Selected Segments of the Indian Services Sector Relative to the U.S. Sector, 2001 software services 50 telecommunications 48 software services 44 and products banking services 12 housing construction 8 retail distribution 6 0 10 20 30 40 50 percentage Source: McKinsey Global Institute 2001. Figure 5.4 Olley and Pakes Decomposition of Aggregate Labor Productivity for the Retail Segments and for the Retail Sector, 2006 a. retail segments 12 10 worked 8 6 hour per 4 US$ 2 0 traditional FMCG stores modern-format stores average productivity b. retail sector 12 10 worked 8 6 hour per 4 US$ 2 0 aggregate average efficiency productivity productivity term Source: Retail ICS 2006. 72 Retail Sector 5 Employment data for the past three years show that modern-format stores also have a higher average growth rate in employment. Although starting from a much lower base, the modern-format stores' average employment growth has been five times higher than that of FMCG stores in the past three years. Thus, to increase productivity and promote employment generation in the retail sector overall, the Indian government should promote an environment that facilitates the growth of the largest market players, which are the most productive ones. Hence, the analysis of the effect of investment climate variables on the sector productivity and employment generation focuses on modern-format stores, which are predomi- nantly those that have the largest market share. A description of the effect of invest- ment climate variables on labor productivity and employment generation for FMCG stores is presented in box 5.1. The recent report on the effect of modern-format retailers on the traditional retail segment (Joseph and others 2008) finds a marginal increase in employment in the latter for the period during which modern-format stores covered by the survey existed (on average, 21 months), indicating that an expansion of larger stores does not have to come at the cost of unorganized retailing. Government Reform Efforts to Improve the Investment Climate In recent years, state governments have adopted a number of policies to facilitate the growth of the retail sector, particularly of the modern-format store. These policies cover such issues as real estate, opening hours for shops, agricultural marketing, and foreign direct investment (FDI) in the sector. Examples of state governments' efforts to facilitate the growth of the retail sector include the following: · The state governments of Delhi, Karnataka, and Maharashtra have increased the land available for commercial purposes by allowing land formerly designated for residential use to be used for commercial purposes and by releasing unused state- owned company land. · In Bangalore, Delhi, Kolkata, and Mumbai, the state governments have permitted flexibility in the use of labor and have allowed retail outlets to operate seven days a week. · States such as Haryana, Himachal Pradesh, Madhya Pradesh, Punjab, and Rajasthan have amended the Agricultural Produce Marketing Committee Act to permit implementation of "modern forms of distribution" for agricultural pro- duce. The amendment allows farmers to sell directly to companies instead of requiring them to sell in the mandis (market centers). The central government has partially liberalized FDI in the retail sector and intro- duced the value added tax (VAT). To facilitate FDI inflow, the central government has allowed FDI of up to 51 percent, with prior government approval, in "single brand" products and FDI of up to 100 percent, without government approval, for cash-and-carry wholesale trading and export trading. Although not specifically aimed at promoting organized retail, the VAT could indirectly facilitate its growth. 73 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Box 5.1 Effect of Investment Climate on Labor Productivity and Employment Generation for Traditional Retail Stores Overall, investment climate matters less for the labor productivity of traditional retailers than for that of modern format stores. All the investment climate and control variables together are able to explain 22 percent of the observed variability of labor productivity for the FMCG segment (compared with 44 percent for the modern-format segment). In terms of the factors identified by store managers as severe obstacles to growth, the main difference noted is that 18.09 percent of managers of traditional stores identified access to finance as a severe obstacle, whereas a much smaller percentage of modern-format stores did so. The findings of the productivity analysis confirm this difference, with finance and accounta- bility having the largest effect on average productivity among the investment climate variables for FMCGs. In particular, possession of a current account in formal financial institutions com- prises 28 percent of average productivity. For traditional stores, internal funding is by far the most important source of funding for working capital (81 percent of working capital value) and even more so for fixed assets (83 per- cent of fixed asset value). Because of the very high immovable collateral requirement (190 per- cent of the loan value), only 7 percent of traditional stores applied for a loan in the before year the survey was conducted. When stores borrow from formal sources, state banks are the most frequent providers. Borrowing from informal sources has a very small effect on productivity, probably because informal sources are used primarily for financing working capital and main- taining operations, rather than for maintaining fixed investments. Improvements in the investment climate are also associated with an increased demand for labor by the traditional retail sector, with the finance and accountability category being the most important. Retail outlets that have a current account and an external audit implement more workforce hours than firms without, 11 percent and 29 percent, respectively. Presumably, this finding is because these outlets are able to access the financing necessary to invest and expand. Also, firms that have an external audit have a reduced incentive to use informal labor, which may be reflected in higher observed employment. Stores that finance themselves with internal funds are associated with lower labor demand. Source: Retail ICS 2006. By levying a tax on value added at each stage of sale, the VAT system should bring all retailers, including small ones, under the tax regime. However, if not evenly imple- mented, the VAT could create a very uneven playing field between the modern- format and small store segments.3 Effect of the Investment Climate on Modern-Format Stores Despite these efforts, investment climate is a bigger problem for large retailers than for small stores. Modern-format stores perceive the investment climate as more of an obstacle to their business than do unorganized stores (figure 5.5) in all areas 74 Retail Figure 5.5 Perceived Major or Very Severe Investment Climate Obstacles, 2006 Sector 35 30 29 25 25 24 22 22 20 stores 20 20 20 of 18 17 16 16 15 15 14 14 13 percentage 11 11 11 11 11 10 10 10 9 9 8 6 6 5 5 4 4 3 1 0 0 crime courts hours rates land skills of pricing finance to electricity licensing tax corruption of of regulations behavior to operation regulations transportation anticompetitive of administration access workforce labor business trade regulation tax access telecommunications functioning regulation infrastructure red tape and crime finace labor skills traditional FCMG stores modern-format stores Source: Retail ICS 2006. 75 5 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Figure 5.6 Differing Cost Structures for Traditional and Modern-Format Food Retail Stores 100 3.9 13 1.9 7 1 3 80 17.5 7 1 5 2.2 7.5 60 percent 40 20 60 70 0 modern-format stores FMCG stores cost of goods sold rent electricity others employee costs tax profit margin Source: Evalueserve data. except the following: trade regulations, practices of informal competitors (defined as street sellers or sellers of spurious goods), access to finance, and to a lesser extent, electricity. This difference in perception exists mainly because the current regulations are more favorable for small retailers. Because of their different legal status and location, small retailers enjoy lower rates for inputs, such as power and land. In fact, small retailers normally pay lower land rents (because of their location in residential areas, which have lower rentals) and lower power prices (because most of them pay resi- dential rates, which are not quite half of the commercial rates paid by the modern retail chains). Small stores are also subject to a lower tax. Such differences affect the profitability of the larger stores (figure 5.6). In addition, given their size, small retailers can operate outside of the investment climate. The labor productivity analysis shows that the effect of the investment cli- mate is larger for modern-format stores than for the small retailers. The economet- ric analysis confirms a strong association between investment climate factors and labor productivity for the retail sector (see appendix B for a full analysis of produc- tivity and employment). According to the results of this regression, the investment climate and control variables explain 22 percent of the observed variability of labor productivity of small retailers and 44 percent of the observed variability of the modern-format stores. This difference can be attributed to the fact that small stores, thanks to their size, can operate outside of the investment climate more easily. 76 Retail Sector 5 Obstacles to Increased Productivity and Employment Generation for Modern-Format Stores Electricity, corruption, tax rates, and access to land remain the biggest obstacles to increased productivity and employment. When asked, 23.28 percent of the man- agers of modern-format stores identify electricity as the main obstacle to their growth and 10.58 percent identify both tax rates and corruption (figure 5.7). Although access to land was identified as the biggest obstacle by only 8.47 percent of the store managers, its magnitude was underrepresented in the survey, as explained subsequently. The obstacles identified by store managers are relevant also in the productivity analysis. The infrastructure and the red tape, corruption, and crime categories of obstacles each accounted for roughly 15 percent of the total effect of investment climate on the average labor productivity of modern-format stores.4 Power shortages occur daily, disrupting the business operations for two hours on average, and despite the presence of generators, a considerable part of sales are lost. Obtaining a power connection is a lengthy endeavor; retail stores wait an average of 53 days to obtain it. Having a connection does not, however, ensure receiving elec- tricity. In fact, 80 percent of the respondent stores experienced power outages in the year prior to the survey--averaging more than one per day and two hours in length. To cope with the shortages, 70 percent of large stores own a generator. Despite the alternative arrangements, modern-format stores estimate that they lose 3.3 percent of their annual sales because of power outages. Tax rates and corruption is a major obstacle. It is identified by almost 11 percent of the surveyed firms as the most severe obstacle to their growth. Stores report that they regularly pay informal gifts to get connected to the electricity grid, water system, and so forth and to pass inspections. Although the amounts paid are small (on aver- age, equivalent to less than 1 percent of annual sales), dealing with government agen- cies is also time consuming: managers of organized retail stores spend, on average, 5 percent of the workweek dealing with government officials, and almost 60 percent of the modern-format stores are visited, on average, six times per year. Managers of modern-format stores spend more time dealing with government officials than do their counterparts in small stores. The issue of tax must be analyzed in terms of a level playing field between the organized and unorganized segments. First, the majority of modern-format stores are subject to the corporate income tax regime, which is higher than the individual income tax that is paid by unorganized retailers. Second, because organized retailers are larger, it is more difficult for them to evade taxes.When asked to estimate the per- centage of total annual sales reported for tax purposes by stores similar to their own, unorganized retailers reported 70 percent on average, whereas managers of modern- format stores reported 85 percent. Third, although the elimination of sales tax and the octroi5 and the introduction of VAT have simplified the tax regime and could bring many unorganized retailers into the tax net, the existing tax regime neverthe- less creates difficulties for stores operating across various states. In fact, various states 77 5 78 Facilitating Figure 5.7 Perceived Most Serious Investment Climate Obstacles for Modern Format Store, 2006 25 23.28 the Emergence 20 stores 15 of of Large 10.58 10.58 10 8.47 8.47 Players percentage 6.35 5 4.23 4.76 3.17 2.12 2.12 2.65 2.65 to 1.59 0.53 0.53 1.06 1.06 Boost 0 of Productivity and crime policy rates courts hours land pricing finance to transport of electricity licensing tax of corruption and of to customs regulation practices operation regulationsinadequately workforce uncertainty administration competitors of access tax labor access trade business telecommunications regulation regulatory functioning regulation educated and Employment infrastructure red tape, corruption, and crime labor skills finance Source: Retail ICS 2006. Retail Sector 5 have different VAT rates, and the central sales tax, a central levy on interstate sales, has not yet been abolished.6 Access to land was identified as a major problem by 8.47 percent of the modern- format stores. However, this statistic underrepresents the magnitude of the prob- lem.7 Of the survey respondents who tried to acquire land or buildings during the year prior to the survey, only 60 percent were successful.8 The majority attributed the failure to expand to the impossibility of obtaining land in the desired zone, while a smaller percentage reported that the municipality did not provide the required infra- structure for the desired site.9 Limited land availability in urban areas forces larger stores out of urban areas, which are difficult for consumers to access because of poor roads. Even when suitable land is identified, the transaction process is very long. On average, the survey respondents who tried to expand found that it took four months to formally take ownership of the property once they had identified it. Because of the lengthy and complex land transfer process, informal payments to get transactions completed are considerable, estimated at 8 percent of the total costs. Conclusions To increase productivity and promote employment generation in the retail sector over- all, policy makers should facilitate the growth of the largest players. In line with secondary data, the results of the labor productivity analysis for the sample show that traditional FMCG stores are the least productive types of outlets. In fact, the labor pro- ductivity decomposition shows that the most productive outlets in India are the indi- vidual stores that have the largest market share--namely, the modern-format stores. Employment data for the past three years for the sample also show that with the same growth in sales, modern-format stores generated more employment than other types of stores. Hence to increase productivity and promote employment generation in the retail sector, the Indian government should facilitate the growth of the largest players. The labor productivity analysis shows that the effect of the investment climate on the modern-format segment is considerable. In particular, both the perceived obsta- cles and the productivity analysis show that unreliable infrastructure (power short- ages in particular), tax rates, corruption, and land access are the largest obstacles to the growth of the sector. The issue of tax rates must be analyzed in terms of uniform application and a level playing field with small retailers, which, in practice, are sub- ject to more favorable conditions. (Reforms to the tax regime that took place after the ICS was conducted, however, have reduced the discrepancy between the FMCG stores and modern-format tax regimes.) Finally, with respect to land, the problem lies both with land availability and with the complex and costly land transfer process. Notes 1. FMCG stores include grocers, general stores, pharmacies, food stores, and cosmetic stores. 2. McKinsey Global Institute (2001) measured labor productivity as value added per number of hours worked. Value added, in turn, was calculated as value of sales net of value of goods sold. 79 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Productivity estimates were obtained through interviews with more than 250 retailers in four cities and in both the food and the nonfood sectors. 3. See, for example, the analysis of the effects of VAT introduction on the retail sector in Brazil (McKinsey Global Institute 2003). 4. For practical reasons, the investment climate obstacles have been grouped into five categories: infrastructure; red tape, corruption, and crime; finance and accountability; labor skills; and other variables. Labor skills is the investment climate variable with the largest effect on aggregate labor productivity (32 percent contribution to average productivity), with experience of the manager being the largest contributor. The discrepancy between the perceived obstacle (labor skills--especially an inadequately educated workforce--was identified as a major obstacle by only 6.35 percent of the respondents) and the results of the productivity analysis can be explained by two factors. First, the perception variable for labor skills referred to the education of the workforce, whereas the variable used in the productivity analysis referred to managerial skills because there were not enough observations in the dataset for the education of the work- force. Second, because the store managers were the survey's respondents, it is unlikely that they would report themselves as lacking managerial skills. 5. Octroi is a tax on movement of goods from one district to another, but it has largely been abolished. 6. Tax rates could not be studied in the productivity analysis because the dataset does not report quantitative information on tax rates. 7. The percentage is probably an underestimation because, again, the store manager was the respondent for the survey and might not have been aware of the chain's expansion plans. 8. Land availability could not be used in the labor productivity analysis because there were not enough observations (that is, too few modern-format stores attempted to acquire land or buildings). 9. The other possible reasons were the government's unwillingness to sell the land and an exces- sively long time to obtain required zoning approval. 80 Retail Sector 5 Annex 5.A Obstacle Case Study--Taxes The authority to levy taxes in India is divided between the central government and the state governments. The central government levies direct taxes, such as personal income tax and cor- porate income tax; indirect taxes, such as customs, excise duties, and a service tax; and a sales tax on interstate transactions. States levy a VAT on goods, state sales taxes, and various local taxes. Despite reforms, the tax structure in India remains dominated by indirect taxes. Because of high tax rates, compliance is limited. Despite the high rates, average effective tax rates (AETRs) are low compared with those of advanced economies and higher-income emerg- ing markets in the region. The AETR is measured as the ratio of tax collections to the national tax base derived from national accounts, and it summarizes tax rates; the effective tax base (tak- ing into account tax evasion, exemptions, and the extent of informal activity); and the quality of tax administration. The low AETR, despite the high tax rate, confirms the idea that revenue could be increased by expanding the taxpayer net, lifting exemptions, and stepping up tax administration. Tax-induced distortions tend to be high, leaving scope for additional reforms. The mar- ginal effective tax wedge (METW)--the difference between the pretax and posttax return on capital--summarizes the various tax effects, including corporate and personal tax rates on capital income, depreciation allowances, and the inventory valuation method. It measures the potential cost of taxation to investors, which, in turn, affects their decision to invest. The marginal tax burden in India is low, reflecting low personal taxes, including the elimination of tax on long-term capital gains in 2004/05. However, tax-induced distortions tend to be high. Inventory investment is treated more harshly than investments in machinery and build- ings, so firms that need to carry more inventories--particularly large retail outlets--are penalized in India more than in other countries. Moreover, there is an uneven playing field across different economic actors in the same sectors, especially retail; the majority of modern-format stores are subject to the corporate income tax regime, whereas small stores are subject to the lower personal income tax regime. Personal income tax rates are 5, 20, and 30 percent, whereas the corporate income tax rate is 35 percent. Firms that rely on internal financing and have limited access to debt financing are especially penalized in that they face a high effective marginal tax rate. The METW on new investments varies greatly depending on financing sources. The government effectively subsidizes marginal debt-financed investments because the METW on them is negative. Similarly, investments financed by new equity face a below-average tax wedge, thanks to relatively low dividend taxa- tion. However, investments financed by retained earnings face a tax wedge in excess of 2.5 per- cent, compared with the Organisation for Economic Co-operation and Development (OECD) average of 2 percent. Smaller firms that face problems in borrowing and tend to depend more on internal sources of funds are thus disadvantaged in comparison with larger firms. Reforms are needed on a number of fronts. Direct tax measures include further reductions in statutory rates: a reduction of the corporate income tax rate, a reduction to 15 percent of the 81 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Figure 5.A.1 Tax Payments in India and OECD Countries: Number of Payments, Time Needed, and Tax Rates, 2006 300 271 250 200 183.3 150 100 70.6 60 46.2 50 15.1 0 number of time needed total tax payments (hours) rate (% of profit) India OECD Source: World Bank 2007a. general depreciation rate, and the elimination of the withholding tax on distribution of divi- dends. The introduction of a goods and services tax and further reductions in customs duties are important proposed indirect tax reforms. There is scope for a significant increase in tax pro- ductivity through removal of exemptions and incentives, expansion of the taxpayer net, and increased reliance on information technology to improve tax administration and compliance. The planned introduction of large taxpayer units in major cities in 2006 has helped reduce compliance and transaction costs for large taxpayers. The revenue administration system also places additional costs on firms that are not reflected in information on tax rates: a representative business in the manufacturing sector in India needs to make around 60 tax payments a year and spend 271 hours doing so (see fig- ure 5.A.1), compared with just 12 hours for a similar firm in the United Arab Emirates. Source: This annex draws extensively from Poirson 2006. 82 Retail Sector 5 Annex 5.B Obstacle Case Study--Access to Land The urban land market is constrained by limited supply and high fragmentation as a result of high population density, strict land ownership laws, and zoning laws. High land tenure frag- mentation is partly the result of high population density and especially of a deliberate policy to prevent concentration of land in the hands of a few people. The central government enacted the Urban Land Ceiling Act in 1976, and the act was subsequently approved by many state leg- islatures. The law imposed a ceiling on land that individuals could hold, and land in excess of the limit needed to be surrendered to the government in exchange for monetary compensation. The law was enacted to prevent concentration of urban land in the hands of a few people and to ensure distribution of land to low-income households. The act did not achieve the desired outcome, and the government repealed it in 1999. However, some states, such as Maharashtra and West Bengal, still follow the law, although there are proposals for its repeal. It is estimated that repeal of the act might bring down the land prices in Mumbai by 30 to 40 percent by mak- ing available an additional area of roughly 7,284 hectares. Finally, zoning laws limit land-use transformation from agricultural to commercial use. The rental market is restricted by stringent rent and tenancy laws that cap rents at very low levels and make eviction very difficult. In an effort to eliminate exploitation of tenants by land- lords, the government passed a rent control act in 1947. Very little reform has occurred since then. Because the rental caps have not kept pace with inflation and the realty boom in urban areas and because it is very difficult to evict tenants, these laws have the net effect of subsidiz- ing old tenants. Among the old tenants are many FMCG stores, which can overlook their oper- ational inefficiency because of subsidized rent. As a result, the cost of land is very high in India, compared with costs in other countries. The annual cost of land per square meter in Mumbai and New Delhi is more than three times gross domestic product (GDP) per capita, while in Shanghai it is only one-fourth of GDP per capita (see panel a of figure 5.B.1). Likewise, the ratio of annual rent of land per square meter in Delhi and Mumbai to GDP per capita is more than double that of Beijing (see panel b of figure 5.B.1). Uncertainty surrounding property titles drives up the cost of land acquisition by limiting the supply of reliable investments. Such uncertainty is partly explained by the fact that in many states, land records and registration of deeds have not been computerized. Manual transcrip- tions are more prone to errors, and hard-copy files are more likely to be misplaced. Adding to the problem, certain types of transactions--inheritance being the most common--do not need to be registered under the existing law. Finally, the very high transfer costs drive a number of transactions into informality. These costs vary by state and amount on average to 10.5 percent of the property. Of these costs, stamp duties are the most significant and account for the great- est discrepancy across cities. For example, stamp duty varies from 4 percent of property value in Ranchi to 11 percent of property value in Bhubaneswar (see figure 5.B.2, panel a). The magni- tude of informal transactions is considerable. In Mumbai alone, a recent reduction in stamp duty by half increased the revenue by 20 percent, most of which was generated by an increase in the number of transactions. 83 Facilitating the Emergence of Large Players to Boost Productivity and Employment 5 Figure 5.B.1 Cost and Rent of Land Relative to GDP Per Capita, Various Years a. cost of urban land per square meter b. ratio of annual rent of land per square meter to GDP per relative to GDP per capita capita (2006) index: New Delhi 100 400 120 361 350 100 307 100) 100 300 86 82 80 250 Delhi 200 180 60 (New 45 43 43 percentage150 40 40 32 100 value 21 20 20 50 24 index 0 0 Delhi Delhi Kong Beijing ChennaiBangkok Bangalore Mumbai New Shanghai New MumbaiShanghaiHong Hyderabad Bangalore Singapore Sources: For panel a, data are from http://www.globalpropertyguide.com, http://www.indianrealestateforum.com, http://www.indexmundi.com, and http://www.chinadaily.com. For panel b, data are from Evalueserve. Note: In panel a, Bangalore, Mumbai, and New Delhi data are from 2007, and Shanghai data are from 2003. In panel b, annual rents are for second quarter, 2007, and GDP per capita is for 2006. Although recent reforms at the state level have improved the transfer process in those states, in India overall it takes, on average, 87 days to register a property. Recent reforms include com- puterization of land records in Gujarat, Karnataka, Madhya Pradesh, Maharashtra, and Tamil Nadu and computerization of the registration of deeds in Andhra Pradesh, Karnataka, Maharashtra, Rajasthan, and Tamil Nadu. These reforms have led to substantial improvements. For example, in the cities of Bangalore and Hyderabad, it takes the least time to register a property--just more than one month. By comparison, in states where no reform has taken place, such as Kolkata, registration of a property can take more than five months (see fig- ure 5.B.2, panel b). Because of the lengthy and complex land transfer process, informal payments are consider- able. A survey from Karnataka estimates that the computerization of the land records saved Rs 800 million in bribes and Rs 66 million in waiting time. Similarly, stores responding to the Retail ICS 2006 paid an additional 8 percent of the total transaction cost to get things done. Source: World Bank staff. 84 Retail Sector 5 Figure 5.B.2 Cost and Time of Registering Property across India a. cost 200,000 150,000 (Rs) 100,000 cost 50,000 Patna Delhi Kolkata Jaipur Lucknow Chennai Mumbai Ranchi BangaloreHyderabad New Chandigarh Bhubaneswar clearances stamp duty preparation of deed registration of transfer Source: Doing Business in South Asia 2006. b. time 160 155 16 14 138 140 132 13 14 cost 12 123 12 12 120 119 11 11 (% 10 10 10 11 10 of 100 property 8 (days) 86 8 80 time 63 6 62 value) 60 6 56 4 43 40 35 35 2 20 0 Delhi Patna Kolkata Ranchi Jaipur Chennai Mumbai New Chandigarh LucknowBangaloreHyderabad Bhubaneswar time cost Source: World Bank 2006c. Source: World Bank 2006a. Note: Data represented in the panel a use a property value of Rs 1,590,269. 85 Chapter 6 Software and ITES Sector Productivity Improvements Key to Continued Growth and Competitiveness This chapter identifies the conditions that underpin the success story of the software and information technology­enabled services (ITES) sector--namely, a favorable domestic investment climate and a large demand from Western companies seeking to reduce costs. However, favorable conditions have allowed productivity to lag potential. As these favorable conditions erode, productivity gains become very important. Within the investment climate, shortages of labor skills and electricity, as well as corruption, are the biggest obstacles for exporters, while corruption, electric- ity shortages, and lack of access to land are the biggest for nonexporters. The analy- sis draws from the results Software/ITES Investment Climate Survey (ICS) 2006 and data from secondary sources. The Software and ITES Sector: An Indian Success Story The software and ITES sector in India witnessed strong growth between 2000 and 2007. It grew at a compound annual growth rate of approximately 29 percent and is estimated to generate revenues worth US$47.8 billion in 2007.As a result, the sector's contribution to India's gross domestic product (GDP) also increased, from 2.8 per- cent in 2002 to 5.4 percent in 2007 (NASSCOM 2007). To understand the sector's success and the challenges going forward, the World Bank conducted a nationally representative survey in 2006. The Software/ITES ICS 2006 covered 273 firms in the software and ITES sector. The firms were located mainly in the seven largest cities, which had the highest concentration of software and ITES firms (see appendix A for a full description of the data and methodology). Most of the data presented in this chapter refer to data collected through this survey, although the analysis also draws on secondary sources. 87 Productivity Improvements Key to Continued Growth and Competitiveness 6 The software and ITES sector is characterized by export-driven growth and con- siderable market concentration. The average annual growth rate of export revenues between 2000 and 2007 (34 percent) has been 1.6 times that of domestic market rev- enues in the same period. Revenues from exports (US$31.9 billion in 2007) are almost double those of the domestic market (US$15.9 billion in the same year) (NASSCOM 2007). Most companies that export rely on the export component of their business to generate more than half of their revenues, and in fact, 54 percent of them earn more than 75 percent of their revenues from exports. The software and ITES sector is highly concentrated in terms of contribution to the total industry rev- enue: 10 percent of the enterprises generate about 73 percent of all the revenues in the sector, and the top four software companies and some of their subsidiaries account for approximately 22 percent of total revenues (figure 6.1). Employment in the sector has also grown fast, albeit starting from a very low base. The sector's contribution to employment in India has been steadily rising, and the number of individuals employed in the Indian software and ITES sector increased by 26 percent between 2006 and 2007. Because sector employment growth started from a very low base, employment in the sector as a percentage of total employment is still only at 0.3 percent of the total labor force. However, it is esti- mated that for every job created in the software and ITES sector, another three jobs are created in support services (transportation, catering, and so forth). As a result, India has become the world's largest exporter of software and ITES. The share of services in total exports reached 37 percent in 2006, driven primarily by software and ITES, which accounted for 86 percent of total services exports (UNCTAD 2007). The main export markets for the Indian software and ITES indus- tries are the United States, the United Kingdom, and--to a lesser extent--Western Europe. Traditionally, the sector has capitalized on its cost advantage and focused on the low-end, low value added segment. More recently, large exporting firms have Figure 6.1 Distribution of Software and ITES Firms and Contribution to the Overall Sector Revenue, 2006 80 70 60 total of 50 40 30 20 percentage 10 0 small medium large firms revenues Source: Software/ITES ICS 2006. Note: Small firms are those with revenues of less than Rs 60 million, medium firms are those with revenues between Rs 60 million and Rs 400 million, and larger firms are those with revenues of Rs 400 million or more. 88 Software and ITES Sector 6 started moving up the value chain and have begun focusing on high value added products and services (NASSCOM and Everest Group 2008). Favorable Conditions Underpinning the Software and ITES Success Story The success of the sector is based on a rare set of market conditions: a favorable domes- tic investment climate and a large demand from Western companies seeking to reduce costs. The favorable investment climate translated into large cost advantages. India had relatively large numbers of English-speaking engineers, who were less expensive than their Western counterparts. The sector had been almost completely liberalized--there is virtually no barrier to foreign direct investment (FDI)--allowing for technology transfer. Moreover, companies located in software technology parks enjoyed a full tax holiday on profits from exports.Unlike the manufacturing industry,software and ITES were not subject to stringent labor laws, and because of its features, the sector was less reliant on infrastructure (see box 6.1). In fact, telecommunications, which is the infra- structure on which the sector relies the most, was liberalized in the mid 1990s; since Box 6.1 Comparing Perceived Obstacles for Software and ITES Exporters and Manufacturing Exporters A comparison of the perceived investment climate obstacles by manufacturing exporters and software and ITES exporters revealed that the latter are less exposed to red tape and regula- tions. For example, because software and ITES firms export through the airwaves or optical fibers, customs and trade regulations are less of an issue. The decisions of software and ITES firms are also less affected by labor regulations than those of their manufacturing counterparts. Tax rates are less of a problem for software and ITES exporters, probably because 20 percent of the software and ITES firms in sample were located in software technology parks and, as such, were exempt from corporate income tax on exports. Understandably, telecommunications failures are a bigger problem for software and ITES firms than for manufacturing ones. Although the shortage of skilled labor is starting to be an issue for the manufacturing sector, the magnitude of the problem is greater for the software and ITES sector because the latter has grown much faster than the former. The managers of software and ITES firms find transportation a bigger problem than do their manufacturing coun- terparts, probably because the former are located in the seven largest cities in the country (see appendix A for sample locations), which are affected by worse traffic. Such firms do not face problems transporting their goods; rather they have difficulty ensuring that their employees can reach the workplace. Finally, software and ITES firms complain more about business licensing than do their manufacturing counterparts, probably because the sample included a higher per- centage of young firms, which have a more recent memory of the licensing process. (continued) 89 Productivity Improvements Key to Continued Growth and Competitiveness 6 Box 6.1 continued Major and Very Severe Obstacles for Software and ITES Exporters and Manufacturing Exporters, 2006 50 45 40 36 37 37 34 firms 35 31 31 32 of 30 26 27 25 22 19 20 15 15 13 13 11 1010 11 percentage 9 9 9 10 10 8 8 8 8 4 5 5 3 3 0 land finance to to workers regulationselectricity transporttechnologyregulationsof behaviorlicensingcorruption disorderconflicts and and access access trade administration labor foreign business theft, tax and telecommunicationsto education system and anticompetitive crime, legal customs access skills software and ITES exporters manufacturing exporters Source: Software/ITES ICS 2006. A summary comparison of the obstacles facing software and ITES exporters and manufac- turing exporters is provided in the accompanying figure. Source: Software/ITES ICS 2006. then, the industry has grown fast, and the quality of the services has increased sub- stantially. On the demand side, the Indian software and ITES industry has benefited from growing demand from Western companies that needed to cut costs, while their computer systems still required a lot of labor-intensive support. Software and ITES firms benefit from a better investment climate than either the retail or the manufacturing sector. To analyze the investment climates for the dif- ferent sectors, the study conducted a one-to-one comparison of the average value of key investment climate indicators for each sector.1 The comparison between the software and ITES and retail sectors revealed that the investment climate is more favorable to the former in all but one of the investment climate categories.2 The comparison between the software and ITES sector and the manufacturing sector has been done separately for the exporters and nonexporters because these two segments usually face different issues. For exporters, the investment climate for software and ITES is better in almost all dimensions.3 The comparison is slightly 90 Software and ITES Sector 6 less clear cut for nonexporters, but the dimensions in which the investment climate is better for software and ITES exceed the dimensions where the investment climate favors manufacturing.4 The Potential for Greater Productivity The productivity of software and ITES companies lags that of the best performers. The favorable investment climate, liberalization of the sector, and subsequent expo- sure to other business methods and technology have allowed software and ITES firms to make greater productivity gains than other sectors over recent years (see fig- ure 6.2). Consequently, the best Indian performers in this sector have productivity levels comparable to the U.S. software sector. Nonetheless, the average productivity of this sector is still significantly below its potential and below the averages in other countries (see figure 6.3). A comparison with other countries is particularly relevant with respect to information and communication technology (ICT) and ITES, because this segment competes with the rest of the world. Part of the explanation for lagging productivity can be traced to operational fac- tors, such as individual organization of function within firms, product mix, and lack of brand name. Three main internal operational factors affect productivity in the software service market:5 1. The organization of functions within Indian firms does not lead to efficient use of capacity. This problem is exacerbated by high turnover and a relative lack of senior information technology (IT) professionals. Figure 6.2 Labor Productivity Growth across Services Sectors, 1982­90 and 1992­2002 12 10 (%) 10 8 8 productivity 6 6 6 6 5 labor 4 4 4 4 of 4 2 growth 0 construction wholesale trade, transport, finance, community, retail trade, storage, and insurance, social, and and so on communication real estate, personal (including and so on services software and ITES) 1982­90 1992­2002 Source: Kohli and Mohapatra 2007. 91 Productivity Improvements Key to Continued Growth and Competitiveness 6 Figure 6.3 Overall Labor Productivity in Software Sector Indexed to U.S. Average, 1998 100 100 90 90 80 value 70 60 index 50 44 38 40 30 20 productivity 10 0 Russian India Germany United Federation States Source: McKinsey Global Institute 2001. 2. Indian companies have a high proportion of low value added services in their product mix. The ability of many Indian companies to move up the value added chain is constrained by the fact that international clients find it easier to outsource the lower parts of the value chain, such as maintenance of large mainframe and legacy systems, and so contracts can be more easily defined. The shortage of sen- ior personnel also affects the product mix because companies are unable to pro- vide high value added products, such as system design and strategy development. 3. Indian companies have found it hard to sell themselves as high-value producers because of weak sales teams and the perception of risk by firms in the West. Low productivity of companies in the goods market was largely attributed to their small size, which means they do not benefit from economies of scale (McKinsey Global Institute 2001). There have also been external barriers to productivity growth, such as low wages and the lack of a sophisticated end-user market. The large supply of English-speaking IT engineers has kept wages in the sector relatively low compared with the world aver- age, until recently. This situation has allowed India to compete effectively in the out- sourced services market, but it has also meant that most companies have continued to focus on these low-value segments using fairly basic processing engineers. Firms have tended to respond to increased demand for their products by increasing their staffs, as confirmed by the 0.73 correlation between revenues and change in the number of permanent employees. Consequently, productivity gains, which are harder to achieve, have not been essential to maintaining growth. Moreover, the Indian end-user industry (for example, the financial sector) is not very sophisticated because of a relatively slow adoption of IT. The low-value requirements of domestic industries and government, such as basic data processing, further complicate the ability of the software and ITES companies to diversify their products up the value chain.Although some Indian companies have reached productivity levels comparable to global standards, the low barriers to entry in the Indian market mean that many 92 Software and ITES Sector 6 new entrants focus on low value added services and products, thus bringing the aver- age sector productivity down. The Importance of Productivity Gains as Favorable Conditions Erode Changes in favorable conditions on both the supply and the demand sides have the potential to hurt India's competitiveness. On the supply side, the shortage of trained professionals is beginning to affect the industry and can be seen in the wage data, dis- cussed later. Moreover, the benefits conferred by the tax breaks for companies located in software technology parks are due to expire. All these factors are con- tributing to a reduction in profits. On the demand side, given the weight of the U.S. market in the sector's exports, a medium-term slowdown of the U.S. economy is likely to have an adverse effect. Although a short-term slowdown might actually favor India--because U.S.-based companies will explore opportunities to outsource work to low-cost destinations, such as India--a longer slowdown can have a serious impact, especially the small and medium-size companies. Moreover, increased domestic demand from real income growth and government expenditure have led to an appreciation of the rupee that may adversely affect India's competitiveness and may exacerbate the effect of labor shortages on wages. It is estimated that an increase of 1 percent in the value of the rupee versus the dollar leads to a decrease of 20 to 30 basis points in the profit margins of companies in this sector.6 As cost advantages are being eroded--especially wage advantages--productivity gains will become more important. Higher wages will erode the sector's competi- tiveness in the low-value segment of the world software and ITES market, which has been the focus of Indian firms. This development highlights the importance of increasing productivity so that firms are able to operate in high-value segments. As cost advantages are being reduced, part of the sector has already started refocusing on higher value-added segments. Thanks to the evolution of the domestic demand (attributable to the growth of the financial services and similar industries), the value added of products for the domestic market should also grow. This productivity growth will assist in offsetting the impact of the exchange rate on competitiveness. The investment climate is an important factor affecting the potential productiv- ity growth of firms in the software and ITES sector. The productivity analysis shows that for the overall software and ITES sector, investment climate and control vari- ables explain 27 percent of the variation in total factor productivity (TFP). This rel- atively low contribution of investment climate variables reflects the fact that the explanatory power of the model is limited by a number of unobservable factors. For example, technological spillovers play a big role in the productivity of the sector; the relative productivity of firms focusing on exports and the domestic market is influ- enced by the sophistication of the end user. Organizational functions and tasks, which are hindered by high attrition rates of employees, also affect productivity. These factors are important for the sector, but they are unobservable and are covered 93 Productivity Improvements Key to Continued Growth and Competitiveness 6 Figure 6.4 Most Important Perceived Obstacle, 2006 25 23 20 17 firms of 15 13 13 11 11 10 9 7 7 7 7 7 5 5 5 5 percentage 5 4 4 4 4 3 3 3 2 2 2 2 3 3 3 2 2 1 1 1 1 0 0 0 0 0 trade crime rates skills land system finance to and electricity behaviorlicensing tax corruption limitations to instability legal uncertainty technology regulations transportation transportation FDI administration workforce access air labor macro customs access road business policy tax foreign telecommunications to anticompetitive access infrastructure red tape, corruption, and crime labor skills finance nonexporters exporters Source: Software/ITES ICS 2006. by other reports; thus, they are not discussed in the subsequent analysis, which instead focuses on the investment climate. Limited skills in the workforce, electricity shortages, and corruption are the biggest obstacles for exporters, while corruption, electricity shortages, and lack of access to land are the biggest obstacles for nonexporters. When asked, 17 percent of the exporters identified shortage of qualified workforce as the biggest obstacle to growth, 13 percent said electricity shortages, and 13 percent said corruption (see fig- ure 6.4). In contrast, 23 percent of nonexporters identified corruption as the biggest obstacle, 11 percent identified electricity shortages, and 9 percent lack of access to land. The perceptions of firms' management are supported by the productivity analysis. Infrastructure accounts for 38 percent of the variations in intrafirm TFP; quality, innovation, and labor force skills accounts for 23 percent; and red tape and corruption account for 20 percent. The advantages the sector held in terms of workforce skills and low costs have already been eroded by labor shortages. According to the Software/ITES ICS 2006, 35 percent of the firms report facing shortages of workers with the necessary skills, but the proportion is much larger among exporters (55 percent) than among nonexporters (17 percent). This finding is reflected in the perception of obstacles to growth, where 17 percent of exporters reported lack of workforce skills as the most important obstacle, compared with 3 percent of nonexporters. Software and ITES firms need 4.4 weeks on average to fill managerial positions and 2.7 weeks to fill skilled worker positions (figure 6.5). Exporters need more than one extra week than nonexporters to fill professional and skilled worker positions and 2.4 extra weeks for 94 Software and ITES Sector 6 Figure 6.5 Average Number of Weeks to Fill Position, 2006 6 5.5 5 4.4 4.5 weeks 3.8 of 4 3.1 3.3 3.0 3 2.7 number 2.0 2 1 average 0 managerial professional skilled workers' positions positions positions nonexporters exporters total Source: Software/ITES ICS 2006. Figure 6.6 Average Annual Percentage Change in Average Wages by Type of Skills, 2003­06 20 17.3 14.7 15.2 14.8 15 14.1 13.4 (%) 12.3 12.5 12.3 11.2 10.2 9.4 9.3 9.3 change 10 8.7 annual 5 0 basic skills standard office industy-specific hardware and advanced software software maintenance programming skills skills skills skills nonexporters exporters total Source: Software/ITES ICS 2006. managerial positions. As a consequence of labor shortages, firms report wages hav- ing increased on average between 11 and 14 percent per year between 2003 and 2006, and increases have been higher for more specialized classes of workers (figure 6.6). These averages hide large differences between exporters and nonexporters; the for- mer have faced increases of between 13 and 17 percent per year in the past three years. By comparison, wages in the manufacturing sector over the 2002 to 2003 period grew by 5.5 percent. To counteract the increasing high cost and to diversify its client base, the indus- try is repositioning itself in the high-end segment. This effort, however, is likely to make the workforce shortage more acute. In particular, as India further positions 95 Productivity Improvements Key to Continued Growth and Competitiveness 6 Table 6.1 Corruption for Exporters and Nonexporters, 2006 Manager time spent with government Share of total sales Type of firm officials on a weekly basis (%) paid as bribe (%) Nonexporter 5.8 3.0 Exporter 9.9 1.6 Total 8.1 2.2 Source: Software/ITES ICS 2006. itself as major player in the high value added segment, the competition for highly skilled staff will increase even further. Informal payments are small, but dealing with government officials is time con- suming. The survey shows that the proportion of time that managers have to spend with government officials and the average number of inspections are larger for exporters than for nonexporters (table 6.1). The percentage of sales paid as bribes is larger for nonexporters than for exporters. However, because sales are on average lower among nonexporters than among exporters, the latter pay higher informal payments than nonexporters in absolute terms. Generators provided a fifth of the energy required by the industry. A much larger share of exporters own a generator (75 percent) than nonexporters (31 percent), but the percentage of power coming from the generator is about the same--18 percent of all energy requirements. A bit more than half of the firms that tried to expand managed to acquire the required land. Problems obtaining access to land are particularly acute for the soft- ware and ITES sector because most of these firms are located in large cities, where land is particularly scarce. Thirty percent of exporters and 16 percent of nonex- porters tried to acquire land in the three years before the survey. Only 65 percent and 61 percent of them, respectively, were successful. When asked why they did not suc- ceed in expanding, exporters mainly reported unavailability of land, while nonex- porters identified problems concerning unclear ownership and difficulties involved in registration or obtaining permits as equally significant reasons. It is estimated that between 2006 and 2010, a workforce of 1 million people will join the IT and ITES industries in tier 1 Indian cities, such as Bangalore, Chennai, Hyderabad, Mumbai, and New Delhi (NASSCOM and McKinsey & Company 2005). This workforce will exert additional pressure on the already inadequate urban infrastructure facilities. Conclusions Favorable conditions have allowed productivity in the software and ITES sector to lag its potential. However, as these conditions erode, productivity gains take center stage. On the supply side, the shortage of trained professionals and the expiration of the tax breaks for companies located in software technology parks are contributing 96 Software and ITES Sector 6 to a reduction in profits. On the demand side, a medium-term slowdown of the U.S. economy is likely to adversely affect the Indian software and ITES sector. Moreover, increased domestic demand from real income growth and government expenditure have led to an appreciation of the rupee that may adversely affect India's competi- tiveness and may exacerbate the effect of labor shortages on wages. These develop- ments highlight the importance of increasing productivity. The investment climate is an important factor affecting the potential productiv- ity growth of firms in the sector. The productivity analysis shows that, for the overall software and ITES sector, investment climate and control variables explain 27 per- cent of the variation in TFP. When asked, exporters identified a shortage of qualified workforce, electricity shortages, and corruption as the biggest obstacles to growth. Nonexporters identified corruption, electricity shortages, and problems gaining access to land as the biggest obstacles. Notes 1. These indicators include the proportion of firms that paid for security, the percentage of monthly private sales that were never paid and the number of days it took to resolve an overdue payment, the proportion of firms with an overdraft facility, the proportion of firms that spent on research and development, the proportion of firms that paid for royalties, the proportion of firms that gave in-house training, the proportion of firms involved in court cases and the num- ber of weeks it took to resolve cases in courts, the proportion of firms experiencing power out- ages, the percentage of firms that failed to acquire land, the number of weeks it took to complete an acquisition of land, the percentage of senior management time spent dealing with govern- ment officials, the number of inspections received, and the time required to clear customs for imports. 2. The only exception is percentage of management time spent dealing with bureaucracy. 3. Manufacturing has a better investment climate only in terms of percentage of firms that failed to acquire land, the proportion of firms that paid for security, and the number of weeks it took to resolve cases in courts. With respect to access to land, note that the software and ITES firms were sampled in the seven largest cities (which had the highest concentration of ICT firms). Hence, it is understandable that access to land for software and ITES firms is more problematic. 4. In addition to the variables listed in note 3, the manufacturing investment climate seems to be better for the following variables: percentage of monthly private sales that are never paid, pro- portion of firms with an overdraft facility, and time required to clear customs for imports. 5. Software services include maintaining mainframe systems at the low-value end and developing IT strategies or specifying design parameters at the high-value end. The software goods market includes programs such as operating systems, network systems, utility programs, and application software. 6. The estimate was provided by Evalueserve. 97 Productivity Improvements Key to Continued Growth and Competitiveness 6 Annex 6.A Obstacle Case Study--Skill Shortages Skilled labor, which has been key to India's growth, is in increasingly short supply. Rising levels of education and skills have been critical factors in India's sustained economic growth; India's rapid growth over the past few years has been led by the services sector in general and by rela- tively skill-intensive services in particular. Recent industrial growth has also come mainly from relatively skill-intensive and capital-intensive sectors. Not surprisingly, therefore, the demand for educated and skilled workers is rising, and the premium on these workers has increased. Returns to secondary education are particularly high. However, the supply of these workers has not kept pace with demand, and this shortage has the potential of becoming a constraint to fur- ther growth. The Federation of Indian Chambers of Commerce and Industry and PHD Cham- ber of Commerce and Industry estimate that there will be a shortage of 5 million to 6 million skilled professionals from 2007 to 2012 and that the biggest shortage will be concentrated in the manufacturing and services sectors (FICCI 2007). Skill shortages are more acute in sectors that have been growing fast and whose growth has been led by relatively skill-intensive services. Firms across the board are now reporting diffi- culty recruiting skilled workers: the IT sector reports a shortage of half a million engineers, the banking and finance sector faces an estimated 50 to 80 percent personnel shortage, and the food processing industry is dealing with a 65 to 70 percent shortage of mechanics and food safety personnel (FICCI 2007). Signs of skill shortages can also be seen in wage data; recent real wage growth in managerial and executive ranks--at about 10 percent--has been the highest in the region. Lack of facilities, shortage of quality human resources, and a weak accountability framework result in poor service delivery and uneven quality of elementary education. These problems with the education system lead to a large number of children dropping out, and those who stay often do not attain the desired learning levels. For example, India has a shortage of almost 1 million classrooms, teacher attendance stood at 82 percent in 2007, and the student dropout rate at primary level was 29 percent in 2004/05. As a result, India is home to more than a third of the world's illiterate population. Moreover, of the small percentage of people who completed postsecondary education (just 5 percent of the adult population in 2004), many are not well trained because of a secondary education system often characterized by rote learning and out- dated curricula that preclude innovative methods of teaching. In addition, India's vocational education and training system is not aligned to emerging market needs. Despite the high dropout rate from secondary education, fewer than 3 percent of students in grades 11 and 12 are enrolled in vocational education. Generally, the students who enter vocational education programs are those who perform poorly academically. Conse- quently, the relevance and quality of India's vocational training programs is questionable. Except for islands of excellence in higher education, India's education and vocational system is not producing enough workers with key research and engineering skills. Rigid curriculum policies and a lack of incentives to modernize curricula have led to limited innovation in the education system. Universities find it hard to retain good faculty and the best researchers, given 98 Software and ITES Sector 6 competitive private sector salaries and an inclination to go abroad for postgraduate studies (87,900 postgraduate students went abroad to study in 2002/03, up from 51,400 in 1999/2000). The market gap is not filled by private providers, because a number of regulations limit the financial viability of private colleges. By law, private higher education should not amount to profiteering or commercialization of education. The supreme court has interpreted this statute as "private colleges can earn a reasonable surplus on invested capital," leaving the state to inter- pret what "reasonable surplus" means. As a result, each year, private colleges need to submit a business plan (including fees that they intend to charge) to the state audit committee, which then decides whether the proposed fees generate a "reasonable surplus on invested capital." The business plan of a potential private provider needs also to take into account myriad norms for providing private education. These regulations determine other key aspects of education pro- vision, such as permissible annual intake, land requirements, classroom areas, computer and software requirements, library resources, and requirements for faculty. In contrast, important outcomes, such as graduation rates, student learning, scientific output, accreditation, and employment rates are not measured. 99 Chapter 7 Reducing Inequalities between States by Improving the Investment Climate This chapter explores the business climate characteristics that explain the differences in performance of individual states in terms of investment and growth. To this end, an Investment Climate Index (ICI)--a first for India--has been built. The index summarizes the different aspects of the business environment that entrepreneurs look at when deciding whether to invest. The analysis shows that infrastructure (mainly power and transportation) and institutions (corruption, tax, and security) are the main bottlenecks to private sector development in India. However, while institutions affect high-growth and low-growth states almost equally, infrastructure affects low-growth states much more than it affects high-growth states. Importance of Measuring the Investment Climate at the State Level Investment climate reforms by Indian states have yielded tangible benefits.As a result of state-level reforms, some states have grown faster than others. Thanks to the lib- eralization policies of the early 1990s, states have been allowed to play a larger role in determining their development paths.As a result, growth has accelerated in all states, but at a faster pace in middle-income states. This trend has brought an increasing divergence between the low-income states and the middle- and high-income states. In economic matters, states have substantial autonomy over the reform agenda. Broadly speaking, land and licensing matters are within the realm of state jurisdic- tion. Infrastructure, labor laws, and vocational education are part of the joint juris- diction. Regulation of the telecommunication and financial sectors, customs activi- ties, environmental regulations, and judicial reforms are the responsibility of the union. 101 Reducing Inequalities between States by Improving the Investment Climate 7 A comprehensive ICI can help pinpoint the crucial obstacles. The investment cli- mate covers a broad spectrum, but governments have limited resources, and India's states can undertake only so many reforms at a given time. Therefore, it is crucial to understand which aspects of the investment climate matter most. Such an under- standing requires unbundling the investment climate into its components and deter- mining which components are key to growth. By building a comprehensive ICI and comparing the states' performance on the basis of different ICI rankings, one can identify the key constraints for the worse-performing states. The ICI: Methodology The ICI is a composite indicator of microlevel investment climate variables drawn from the Investment Climate Survey (ICS) datasets. These variables measure both cost and perception. When making an investment decision, entrepreneurs look at a host of factors, ranging from cost of inputs, to reliability of infrastructure, to quality of institutions.1 A sound investment climate index should take into account as many of these factors as possible. The ICI has been built on approximately 45 investment climate variables collected in the Manufacturing ICS 2006 and Retail ICS 2006.2 These variables measure inputs, infrastructure, and institutions and include both objective (cost) and subjective (perception) indicators. They reflect the perception of mainly small and medium entrepreneurs because those entrepreneurs constitute most of the sample. Appendix C provides a full description of the methodology used to build the index. The index has also been tested for reliability (see box 7.1). ICI State Rankings The ICI ranks the 16 states on the basis of their investment climate features and iden- tifies three groups of states: best performers, average performers, and worst per- formers. The best performers have the lowest score value. When reading this rank- ing, one must remember that the ICI is useful in highlighting broad patterns but should not be taken as an indicator of the exact ordinal position of any individual state in the ranking. Because the ICI is a linear combination of factors estimated from a sample of the population, the value of each index has a margin of error. Con- sequently, values that are very close to each other need not be considered as repre- senting the exact position of a state. For instance, the differences among Jharkhand, Madhya Pradesh, and Orissa in the rankings are so small that it may not be correct to assume that the exact ranking of these three states is as the ICI shows. Rather, it is more appropriate to conclude that these three states have similar levels of investment climate. State rankings based on the ICI have also been compared with rankings based on the World Bank's Doing Business study (World Bank 2007a) and the Infra- structure Quality Index of the Government of India (Government of India 2000) (box 7.2). 102 Reducing Inequalities between States by Improving the Investment Climate 7 Box 7.1 ICI Reliability The composite ICI has been successfully tested for reliability: states with good investment cli- mates have higher private domestic investment as a share of gross domestic product (GDP), higher GDP growth, and higher per capita GDP growth. Because there is no theoretical model on the estimation of the weights used in the construction of the index, the reliability of the ICI as a predictor of investment climate in India was tested. To this end, the ICI for each state has Correlation of the ICI with Other Indicators a. share of GDP 0.8 Karnataka Gujarat 0.6 Madhya Pradesh Andhra Pradesh (%) Maharashtra Jharkhand GDP Tamil Nadu 0.4 Kerala Bihar of Haryana share Punjab Rajasthan 0.2 West Bengal Uttar Pradesh 0.0 1 2 3 4 ICI value b. GDP growth 12 10 Bihar Karnataka Tamil Nadu 8 Kerala (%) Orissa Haryana Maharashtra 6 West Bengal growth Andhra Pradesh GDP 4 Punjab Gujarat 2 Uttar Pradesh Madhya Pradesh 0 1 2 3 4 ICI value (continued) 103 Reducing Inequalities between States by Improving the Investment Climate 7 Box 7.1 continued c. foreign direct investment 10 8 Maharashtra Gujarat flows Delhi Karnataka Punjab 6 Andhra Pradesh ent Tamil Nadu Kerala West Bengal 4 investm Haryana Uttar Pradesh 2 Orissa Rajasthan direct Jharkhand 0 foreign Bihar log 2 Madhya Pradesh 4 1 2 3 4 ICI value Sources: Manufacturing ICS 2006; Retail ICS 2006. been correlated with the state performance in terms of level of domestic private investmenta and GDP growth.b The ICI showed clear and significant association with all these indicators at the state level (see the accompanying figure). These tests provide a degree of confidence that the composite ICI is a reliable indicator of the investment climate in India.c States with better investment climates also attract more foreign investment. Even though the sample interviewed in the Manufacturing ICS 2006 and Retail ICS 2006 included domestic firms, states with better investment climates attract more direct foreign investment (see panel c of the accompanying figure). Foreign investors share the view of domestic investors on the quality of investment climate in Indian states. Sources: Manufacturing ICS 2006; Retail ICS 2006. a. The correlation is significant at the 5 percent level. Delhi is omitted because data are not available, and Orissa is eliminated because it is an outlier. b. The correlation is significant at the 10 percent level. Ideally, one would like to test causality by correlating the ICI with subsequent growth. Unfortunately, at the time this book was finalized, data on 2006 GDP growth for the 16 Indian states analyzed were not available. In earlier work, in a set of 24 European and Central Asian countries, this relationship was shown to be significant (see Iarossi, Saliola, and Tanzillo 2007). c. Similar significant results were obtained between the ICI and the "Investment Environment" ranking of Indian states reported in the September 26, 2006, issue of India Today. According to the ICI ranking, the states with the best investment climate are Karnataka and Kerala, followed by Gujarat, Andhra Pradesh, Haryana, West Bengal, Maharashtra, and Delhi. The worst investment climate is found in Bihar, Uttar Pradesh, and Rajasthan (figure 7.1). 104 Reducing Inequalities between States by Improving the Investment Climate 7 Box 7.2 Comparing the ICI with the World Bank's Doing Business Study and the Government of India's Infrastructure Quality Index Investment Climate under the Doing Business Study and Infrastructure Quality Index better investment climate worse investment climate doing business study infrastructure quality index poor good Sources: Government of India 2000; World Bank 2007a. ICI rankings are compared with those of the World Bank's Doing Business study and the Gov- ernment of India's Infrastructure Quality Index. Investigating the relationship requires classify- ing the 16 Indian states into two categories: (a) by ease of regulatory environment according to their rank in the Doing Business study (World Bank 2007a) and (b) by quality of infrastructure in accordance with their ranking in the Infrastructure Quality Index (Government of India 2000). For each category, the average value of the states' ICI ranking has been estimated and compared. Consequently, the bars in the accompanying figure represent the average quality of the invest- ment climate, as measured by the ICI, for the top and bottom performers in each of these two categories (the ICI is calculated for the top eight states and the bottom eight states). The comparison reveals that an ICI ranking is more closely associated with the quality of infrastructure than with the quality of the regulatory environment as measured by the Doing Business study. According to the Infrastructure Quality Index developed by the government of India, states that show a good quality of infrastructure have a much better investment climate than do states with poor infrastructure. In contrast, states that perform poorly on the invest- ment climate composite indicators show the same average ranking in the quality of their regu- latory environment as do states that have a good investment climate. Sources: Government of India 2000; World Bank 2007a. Although it is not surprising to see Karnataka as one of the states with the best investment climate, the ranking of Kerala within the top tier merits further analysis.3 A closer examination of the data shows that Kerala consistently scores better than average on most objective and perception indicators. In objective indicators, Kerala 105 Reducing Inequalities between States by Improving the Investment Climate 7 Figure 7.1 ICI Ranking of Business Climate in 16 Indian States, 2006 Karnataka Kerala Gujarat Andhra Pradesh Haryana West Bengal Maharashtra Delhi Tamil Nadu Punjab Jharkhand Orissa Madhya Pradesh Bihar Uttar Pradesh Rajasthan 1 2 3 4 ICI ranking Sources: Manufacturing ICS 2006; Retail ICS 2006. performs better than average in four out of five indicators, and in perception measures, Kerala performs better than average in all four indicators. In objective measures, only Karnataka and West Bengal perform better than Kerala in all five indicators. Punjab also performs above average in all indicators, but its performance is not as high as Kerala's. Within the perception variables, Kerala performs above aver- age for all indicators, as do Karnataka, Andhra Pradesh, Tamil Nadu, and Orissa. Objective and perception data on the same investment climate dimensions in Kerala also overlap.4 Objective and perception data contradict themselves in only one of the indicators: transport. The reason for such apparent contradiction lies in the fact that 40 percent of firms in Kerala have their own means of transport--compared with 11 percent in the rest of India--so its firms perceive it as less of a problem simply because they are not relying on outside transport. However, objective data show that firms in Kerala suffer losses because international transport costs are three times higher for such firms than for other firms in India. So in the objective rating, Kerala rightly performs worse.5 West Bengal ranks high because of its good performance on objective indicators of the business environment. Objective indicators show a much lower incidence of power interruption, corruption, and government disservices in West Bengal than in the other 15 states in India. More specifically, West Bengal performs well on reliabil- ity of power and on cost of corruption, where it is the best performer. This good 106 Reducing Inequalities between States by Improving the Investment Climate 7 rating compensates for its relative lower performance on perception indicators, such as access to land, crime, access to finance, and transport. On the other side of the spectrum, Delhi and Tamil Nadu are ranked relatively low. Although both states have a similar performance with regard to infrastructure (corresponding to the average performance of the 16 states), Delhi scores poorly on institutions while Tamil Nadu scores very poorly on inputs. In particular, Delhi has the worst performance in terms of corruption among all 16 states, whereas Tamil Nadu shows a poor performance on access to finance, lack of skills, and availability of technology. In terms of infrastructure, both states have an average performance, but while Delhi performs badly on reliability of power, Tamil Nadu is among the top five states in the country. The states at the bottom of the ranking--Bihar, Uttar Pradesh, and Rajasthan-- perform poorly on all indicators of infrastructure, institutions, and inputs. Although Bihar and Uttar Pradesh perform worse mainly in infrastructure and less so in insti- tutions and inputs, Rajasthan performs better than the other two on infrastructure but much worse on institutions and inputs. In Uttar Pradesh, the main bottleneck is power reliability, whereas in Bihar, telecommunications and transportation seem to be more of an obstacle. Rajasthan performs well on losses attributable to power fail- ures and on losses attributable to transport failures, but particularly poorly on cor- ruption, customs, access to technology, tax administration, and labor regulations. Identifying the Key Investment Climate Differences between Low-Growth and High-Growth States The decomposition of the ICI shows also that infrastructure and institutions are the main drivers of a better business environment. Not only do these variables explain most of the variance of all the 46 variables representative of the investment climate in India (almost 60 percent), but infrastructure has the largest unweighted difference in indices between best and worst states. This result implies that infrastructure and institutions are the variables in which best- and worst-performing states most often differ and consequently explain most of the variation in the business environment between them. Power emerges as the single most important infrastructure constraint, followed by transport. A further decomposition of the infrastructure subindexes to identify the indicators that have the highest weight shows that power--more specifically, the number of power outages and the losses attributable to power outages--is the single most important infrastructure constraint. This finding is true for both objective and subjective indicators. The second most important infrastructure constraint is trans- port. In the perception questions on transport, the bottom six states in the ICI rank- ing perform the worst. Rajasthan is an exception because it performs average on elec- tricity; however, transport remains a key bottleneck there. Corruption, tax regulations, tax administration, and security are the biggest insti- tutional constraints. This finding is true for both perception and cost indicators in 107 Reducing Inequalities between States by Improving the Investment Climate 7 Figure 7.2 Relative Effect of Infrastructure Indicators on the Overall ICI and on Low- Investment and Low-Growth States higher transportation transportation power power ICI on effect higher effect on low-investment (low-growth) states in comparison to high-investment (high-growth) states investment states growth states Sources: Manufacturing ICS 2006; Retail ICS 2006. Note: Low-investment states are Bihar, Haryana, Kerala, Punjab, Rajasthan, Uttar Pradesh, and West Bengal. Low-growth states are Andhra Pradesh, Gujarat, Madhya Pradesh, Punjab, Rajasthan, and Uttar Pradesh. both high- and low-growth states. Security, however, remains problematic for the states showing a poor investment climate ranking. The effect of each investment climate variable varies across typology of states: infrastructure matters most in low-growth and investment states, whereas institu- tions matter equally for low- and high-growth and investment states. Figures 7.2 and 7.3 show the effect of each investment climate variable in low-investment or low- growth states, as compared with the effect in high-investment or high-growth states, as well as the weight associated with each variable in the construction of the com- posite index. The y axis represents the effect of such indicators on the overall ICI, while the x axis reports the difference in magnitude of such indicators between low- and high-growth (or low- and high-investment) states. The point in figure 7.2 labeled "power" shows that entrepreneurs in states with lower levels of private domestic investment complain more about the cost of power than those in states with higher levels of domestic investments.6 Similarly, the point labeled "transporta- tion" shows that transportation appears to be more of a constraint to businesses in states that show lower levels of growth than in fast-growing states. Both of these indi- cators present the highest values on the y axis, meaning that they are the most impor- tant indicators in the construction of the overall ICI. In contrast, figure 7.3 shows that the most important institution bottlenecks in the construction of the ICI, such as corruption, crime, and tax regulations, are considered equally binding by man- agers in both slow-growth and low-investment states. 108 Reducing Inequalities between States by Improving the Investment Climate 7 Figure 7.3 Relative Impact of Institution Indicators on Overall ICI and on Low- Investment and Low-Growth States higher ICI on effect higher effect on low-investment (low-growth) states in comparison to high-investment (high-growth) states investment states growth states Sources: Manufacturing ICS 2006; Retail ICS 2006. Conclusions Infrastructure is the most binding constraint on low-growth states, whereas institu- tions equally affect high-growth (or high-investment) states and low-growth (or low- investment) states. The ICI identifies the two most critical investment climate causes of interstate differences as infrastructure and institutions.A closer analysis also reveals that within these two indicators, power, transport, corruption, tax regulations, and theft remain the major obstacles to growth. It is important to note that poor, infra- structure is particularly binding on states that have a low level of investment and growth, whereas institutional difficulties appear to equally affect high-growth (or high-investment) states and low-growth (or low-investment) states. This finding has clear implication for policy makers, as it can guide them in the prioritization of reforms and investments. Notes 1. Inputs include labor, access to finance, and raw materials. 2. The standardized methodology used to collect such information makes comparison across states possible and meaningful. 3. In an analysis of the distribution of investments by the private sector in 15 Indian states, the Centre for Monitoring Indian Economy ranked Karnataka 4th as of April 2006, while Kerala ranked 14th. See the Centre for Monitoring Indian Economy database for 2006 at http://www.cmie.com/. 109 Reducing Inequalities between States by Improving the Investment Climate 7 4. In terms of the power indicator where in the objective data Kerala performs worst, the percep- tion is the lowest of all states. 5. One could argue that there is no contradiction because two separate conditions are being measured. 6. Figure 7.2 shows on the x axis the difference in the level of each constraint between low-investment (or low-growth) and high-investment (or high-growth) states and on the y axis the importance of the indicators in the construction of the ICI (represented by the share of variance explained by each factor). Consequently, values at the top right of the figure represent business environ- ment indicators that have the highest weight in the ICI and for which each category of states performs differently than the other. 110 Chapter 8 Conclusions India's progress has been remarkable. Because of the reforms of the 1980s and 1990s, India has emerged as one of the world's fastest-growing economies in recent years. Despite its considerable achievements, India will need to overcome three main challenges to continue to grow: (a) Pressure on formal sector employment from demographics; (b) rising inequalities between states and (c) given the record-high investment levels, future growth's increasing dependency on productivity gains. The analysis of manufacturing, retail, and software and information technology­ enabled services (ITES) firm-level data identified power shortages, lack of access to land and finance, taxes, and shortages of skilled labor as the major investment cli- mate obstacles to increased productivity and employment (table 8.1). Infrastructure is a bigger problem for lagging states than for high-growth ones, whereas institutions matter equally for low- and high-growth states. Power Most firms rely on generators, and a substantial amount of sales are lost because of power outages. Half of the manufacturing firms had a generator, 22 percent of all energy requirements for manufacturing came from generators, and 5 percent of annual sales were lost because of power outages. Unorganized manufacturing firms suffered larger losses (11 percent of sales) because on average fewer of them had gen- erators (20 percent). In retail, although the majority of modern-format stores had generators (70 percent), they still lost, on average, 3.3 percent of annual sales. Among the software and ITES firms, 75 percent of exporters owned a generator, whereas only 31 percent of the nonexporters did, but the percentage of power coming from the generators is about the same: 18 percent of all their energy requirements. 111 Conclusions 8 Table 8.1 Main Investment Climate Obstacles, Identified and Ranked by Sector Obstacles Tax rates and Skilled Access Access to Sectors Power administration Corruption labor to land finance Organized manufacturing 1 2 3 Unorganized manufacturing 2 1 Retail 2 3 1 Software and ITES 2 3 1 Sources: Manufacturing ICS 2006; Unorganized Manufacturing ICS 2006; Retail ICS 2006; Software/ITES ICS 2006. Access to Land Access to land is a constraint on firms' expansion, especially in the larger cities. In retail, only 40 percent of the modern-format stores that tried to acquire land suc- ceeded; in 60 percent of those cases, land was not available in the desired sites. Even when the land was located, the transfer process was complex and lengthy (on aver- age, it took firms four months to formally take ownership of a property once the land had been identified). Because of the lengthy and complex transfer process, modern- format stores spent on average 8 percent of the transaction cost on payments to offi- cials to get things done. Software and ITES firms also identify access to land as a major problem. Thirty percent of exporters and 16 percent of nonexporters tried to acquire land in the three years before the survey, and only 65 percent and 61 percent of them, respectively, were successful. When asked why they did not succeed in expanding, exporters mainly reported unavailability of land, whereas nonexporters identified problems concerning unclear ownership and difficulties involved in regis- tration or obtaining permits as equally significant reasons. Access to Finance Access to finance is a particular obstacle to the growth of unorganized manufactur- ing. Although the majority of unorganized manufacturing firms have a bank account, only 6 percent had a loan from a bank. Getting a loan from a bank takes on average 25 days, and it requires large amounts of immovable collateral, whereas unorganized forms have mainly movable assets. Taxes High taxes, uneven policies, and inefficient tax administration are a major problem for visible firms that are incorporated. Tax administration was identified as the main problem by 15 percent of manufacturing firms (the second-largest obstacles overall) 112 Conclusions 8 and 20 percent of modern-format stores. High taxes were considered the main obsta- cle by 8 percent of manufacturing firms (the fourth-largest obstacle overall) and 29 percent of modern-format stores. Most exporting software and ITES firms receive tax holidays on profits from export, provided that they are located in software tech- nology parks. Indian firms need to make on average 60 tax payments a year. Total taxes (central, state, and local) for an average firm amount to 71 percent of profit. The playing field is not level for modern-format and small retailers. This differential treat- ment is due in part to enforcement and in part to different tax regimes (rates and applicable taxes), which penalize large retail chains operating across different states. Informal payments are widespread, and dealing with government officials is also time consuming. Manufacturing firms pay on average 4.9 percent of annual sales in bribes, software and ITES firms pay 2.2 percent, and large retail stores pay 1 percent. Corruption also has costs in terms of senior management time spent dealing with government officials. Managers of manufacturing firms spend on average 12.6 per- cent of the workweek dealing with government officials, software and ITES man- agers spend 8.1 percent, and large retail store managers spend 5 percent. Skilled Labor Skilled labor shortages are eroding the competitiveness of software and information and communication technology (ICT) firms and are starting to emerge as an issue for manufacturing firms. More than one-third of software and ITES firms report a shortage of labor with the necessary skills. This shortage translates into high wage increases, especially for specialized personnel (up to 46 percent on average in the 2003­06 period), and replacing employees takes a long time. Although shortage of skilled labor is not yet one of the three main obstacles for the manufacturing sector, it is a growing problem: 15 percent of firms identify it as a major or very severe prob- lem, and 24 percent report unavailability of required skills in the market. As the sec- tor continues to grow, the problem is likely to become more acute. Three Challenges These obstacles have a direct bearing on the three challenges facing the Indian economy: · Formal sector employment would be greatly enhanced by removing obstacles to productivity and growth in the unorganized manufacturing sector, which employs the vast majority of India's workforce. The analysis suggests that these firms would be willing to become formal if growth were facilitated. The greatest obstacles in this respect are lack of access to finance and unreliable power supply. Moreover, facilitating the emergence of larger retail stores by tackling access to power and land, reducing corruption, and leveling the playing field with respect to taxes will accelerate job creation in the formal sector. 113 Conclusions 8 · The rising inequalities between states can be traced to the positive effects of liber- alization and investment climate improvement in the fastest-growing states, which have attracted a higher concentration of productive firms. Mirroring these reforms in backward states will be an important step in reducing the inequality. The decomposition of the state-level Investment Climate Index confirms that variables associated with infrastructure (power, in particular) are key drivers of a better investment climate in India. The analysis also shows that the effect of investment climate variables differs by type of state; infrastructure is particularly important in low-growth states. Moreover, facilitating the formalization of un- organized manufacturing, which is present in most regions, will help reduce disparities. · Productivity gains among small, unorganized firms are limited because of the firms' inability to access finance for productivity-enhancing investment. In the software and ITES sector, a shortage of skilled labor and infrastructure are the greatest barriers to moving up the value chain and enhancing productivity. For formal manufacturing firms, red tape, tax, and corruption, followed by infra- structure, had the biggest impact on total factor productivity, accounting for 27 percent and 23 percent of the observed variability, respectively. Sound technical analysis of the obstacles is a first step, but an ongoing dialogue on solutions is needed. The brief analysis of the investment climate obstacles pre- sented in this report is intended to bridge the gap between the public and the private sector in terms of understanding the scope of market distortions and the required public sector inputs, such as infrastructure and regulation. Detailed policy prescrip- tions on how to address these issues are beyond the scope of this report. Although progress has been made in many of these areas, continued improvement will depend on processes and institutions that facilitate monitoring of these investment climate obstacles and development of flexible, rapid responses to the needs of both public and private sectors. To ensure that the right priorities are selected and that the imple- mentation actually happens successfully, reformers around the world have focused their attention on creating institutions that engage the bureaucrats in ongoing dialogue of pertinent themes with the private sector (Hausmann, Rodrik, and Sabel 2007). 114 Appendix A Investment Climate Surveys As a background to this report, surveys of enterprises engaged in the manufacturing, retail, unorganized, and software and information technology­enabled services (ITES) sectors were conducted. The Investment Climate Surveys (ICSs) collected information on the investment climate constraints the sectors faced, such as infra- structure, access to land, relationship with the court system, crime, government administration, use of financial services, and labor force. In addition, the surveys col- lected basic information on the firms, such as ownership structure, number of years of operations, and revenues and costs. The surveys were conducted in 2006. Manufacturing ICS Sampling Frame and Selection Process The Manufacturing ICS 2006 covered 16 states: Andhra Pradesh, Bihar, Delhi, Gujarat, Haryana, Karnataka, Kerala, Jharkhand, Madhya Pradesh, Maharashtra, Orissa, Punjab, Rajasthan, Tamil Nadu, Uttar Pradesh, and West Bengal. Of these, only 12 were covered in the Manufacturing ICS 2003; Bihar, Jharkhand, Orissa, and Rajasthan were covered only in 2006. Across these 16 states, samples were drawn from 13 industries--namely, textiles, garments, drugs and pharmaceuticals, chemicals, plastics and rubber, electronic con- sumer goods, electrical appliances, automotive parts, food processing, mineral pro- cessing, metal products, machinery, and leather and leather products. Each of these industries was chosen for coverage because it represents a significant source of industrial exports. Together the industries also represent the largest manufacturing industries in India in terms of employment and output shares. 115 Investment Climate Surveys A Table A.1 Allocation of Manufacturing ICS 2006 Sample: Bihar, Jharkhand, Orissa, and Rajasthan Allocation 1 (based on Allocation 2 (based on manufacturing employment) manufacturing establishments) Number Share Number of Share of workers of establishments of Mean of under ASI total Implied under ASI total Implied allocations States 2001/02 (%) allocation 2001/02 (%) allocation 1 and 2 Bihar 51,375 12 70 617 15 88 79 Jharkhand 119,624 27 164 1,617 39 232 198 Orissa 89,349 20 122 807 19 116 119 Rajasthan 178,521 41 244 1,146 27 164 204 Total 438,869 100 600 4,187 100 600 600 Source: ASI 2001/02; Manufacturing ICS 2006. The sample methodology used in 2003 was replicated in 2006 for the new states. The distribution of the overall sample was based on the share of each state in India's manufacturing gross domestic product (GDP) for 1998/99. In view of the relatively small combined share of the states of Bihar, Jharkhand, Orissa, and Rajasthan in the share of national industrial output, it was decided to determine the allocation of the newly sampled enterprises, not on the basis of respective shares in manufacturing GDP, but on the relative sizes of the manufactur- ing sectors of the four states as inferred from the 2001/02 Annual Surveys of Indus- try (ASI) of the government of India. As shown in table A.1, one allocation of the sample of 600 firms between the four states of Bihar, Jharkhand, Orissa, and Rajasthan was based on relative share in manufacturing employment, and a second allocation was based on relative share in the number of manufacturing enterprises. Both allocations are based on the ASI. An average of the two allocations was taken because neither would be satisfactory on a stand-alone basis. On grounds of practicality, a decision was made to limit sampling to the largest three or four industrial cities of each state. The sampling frames were consequently drawn up from official lists and directories of business establishments of the selected cities but restricted to the 12 surveyed industries and to enterprises with a minimum of 10 full-time employees. The lists were obtained from government offices and busi- ness associations. Selection from each sample frame was done by systematically sampling one enterprise at a fixed interval, having arranged listings in descending order of employ- ment size. This method was used to ensure sampling from the entire range of the frame, rather than exclusively from the lower end of the distribution, at which busi- ness size distributions tend to concentrate. Sample Size The sample size for the Manufacturing ICS 2006 was set at 2,460 establishments for the manufacturing sector. This decision was based on budgetary considerations. 116 Investment Climate Surveys A Table A.2 Sample Allocation and Actual Coverage of Manufacturing ICS 2006, by State States Sample allocation Actual coverage Andhra Pradesh 179 160 Bihar 79 44 Gujarat 201 178 Haryana 103 107 Jharkhand 198 188 Karnataka 194 203 Kerala 78 66 Madhya Pradesh 138 123 Maharashtra 244 194 Orissa 119 122 Punjab 119 106 Rajasthan 204 174 Tamil Nadu 202 195 Uttar Pradesh 170 166 West Bengal 160 164 Delhi 88 92 Total 2,476 2,282 Source: Manufacturing ICS 2006. Final Sample Summary The distribution of the full (target) sample of the 2006 survey across the 16 states is shown in table A.2, along with the distribution of actual responses (coverage). The distribution of responses deviates substantially from the initial sample design because of low response rates in some areas being compensated for by planned coverage in high-response states. Unorganized Manufacturing ICS Sampling Frame and Selection Process The closest and most relevant data on the unorganized sector was the 56th round of data provided by the National Sample Survey Organisation (NSSO). Hence, these data have been chosen as the base frame for sampling. The data were used to (a) select the cities to be covered under the survey and (b) provide an estimate of the sample size for each manufacturing sector to be sampled from that city. On the basis of the National Sample Survey's 56th round, various manufacturing subsectors were identified for the study (table A.3). 117 Investment Climate Surveys A Table A.3 Industry Sectors Identified for the Unorganized ICS 2006 Compared with Industry Sectors Covered by National Sample Survey's 56th Round Industry sectors National Industry sector covered in National Industrial Sector identified by Sample Survey's Classification number World Bank 56th round code 1 Auto components Not covered 2 Drugs and pharmaceuticals Not covered 3 Chemicals Chemicals and chemical products 24 4 Electrical goods Electrical machinery, apparatus, and so forth 31 5 Electronics Radio, television, and communication equipment and apparatus 32 6 Food processing Food products and beverages 15 7 Garments Garments (manufacture of wearing apparel, dressing and dyeing of fur) 18 8 Leather Leather (tanning and dressing of leather; manufacture of luggage, handbags, saddles, harnesses, and footwear) 19 9 Metal and machine tools (a) Fabricated metal products, except machinery and equipments (a) 28 (b) Machinery, equipment, and so forth (b) 29 10 Sugar Not covered 11 Textiles Textiles 17 Source: NSSO 2001; Unorganized Manufacturing ICS 2006. In view of the objective of sampling enterprises in the unorganized sector (firms having 10 or fewer full-time paid employees) in urban areas, only data for urban areas were chosen. The process adopted to select the cities to be surveyed is given below: · Step 1: Identification of states. For each of the manufacturing subsectors, the seven states with the maximum number of manufacturing units were identified. Delhi, a city-state, was considered for selection despite having a relatively low propor- tion of units when compared with other, bigger states in the country. · Step 2: Selection of cities. In each identified state, districts having a large number of manufacturing units in each subsector were identified. · Step 3: Selection of cities for each manufacturing subsector. For final selection of districts for each manufacturing subsector, the following criteria were applied: (a) maximum availability of each specific manufacturing subsector in districts and (b) appropriate coverage of each region in the country--namely, north, east, west, and south. In this way, the following six cities across four regions were selected: north, Delhi and Ludhiana; west, Mumbai and Thane; east, Howrah; and 118 Investment Climate Surveys A south, Hyderabad. Because certain manufacturing subsectors were underrepre- sented in the selected cities (for example, chemicals and leather), two additional cities were identified where these subsectors are easily available--namely, Banga- lore and Kanpur. These two additional cities also served as reserve cities in case some manufacturing subsectors were not available in a sufficient number in the main cities. Each city was split into clusters that were designated as industrial clusters or were places where these types of enterprises were most likely to operate. This identifica- tion was done through discussions with local municipal bodies, government bodies, trade bodies, chambers of commerce, and other experts. Once the clusters were iden- tified, snowballing techniques were applied for identifying enterprises in various industry sectors. After discussions with local trade bodies and experts, a quota was fixed for each of the three types of enterprises based on size: household enterprises, tiny enterprises (five employees or fewer), and microenterprises (more than 5 but no more than 10 employees). Sample Size The sample size for each manufacturing subsector in each of the selected cities was determined as follows: · The proportion of each manufacturing subsector had to be met. · The proportion of each manufacturing subsector in each of the selected cities had to be met. · Some adjustments were made to ensure that there would be no oversampling in some cities and manufacturing subsectors and to include those subsectors that were not covered by the NSSO in the 56th round (auto components, drugs and pharmaceuticals, and so on). These adjustments were done on the basis of past experience and subjective judgments. Final Sample Summary The summary of the final sample is presented in table A.4. Retail ICS Sampling Frame and Selection Process The sampling frame for the Retail ICS was the list of retail stores regularly inter- viewed by ACNielsen for inventory verification on behalf of distributors of branded goods. This list covered retail stores in 41 cities across India for two industry seg- ments: fast-moving consumer goods (FMCG) stores and modern-format stores.1 119 A 120 Table A.4 Sample Allocation and Actual Coverage of Unorganized Manufacturing ICS 2006, by City and Manufacturing Subsectors Delhi Ludhiana Hyderabad Mumbai Howrah Total Industry Sample Actual Sample Actual Sample Actual Sample Actual Sample Actual Sample Actual Auto components 30 31 30 38 0 0 10 11 10 9 80 89 Drugs and pharmaceuticals 10 0 0 0 30 0 0 0 0 0 40 0 Chemicals 10 12 10 9 5 14 10 19 5 14 40 68 Electrical goods 60 57 20 13 5 6 25 29 10 12 120 117 Electronics 20 34 40 11 10 11 20 25 10 8 100 89 Food processing 85 81 25 31 45 49 80 86 55 47 290 294 Garments 90 84 100 113 30 37 50 53 30 33 300 320 Leather 25 25 30 31 19 19 30 34 16 12 120 121 Metal and machine tools 50 61 30 39 10 20 60 66 60 61 210 247 Textile 10 20 15 25 40 42 75 83 60 38 200 208 Investment Total 390 405 300 310 194 198 360 406 256 234 1,500 1,553 Source: Unorganized Manufacturing ICS 2006. Note: The drugs and pharmaceuticals industry does not have firms of fewer than 10 employees; hence, this quota was shifted to the related chemicals industry. Climate Surveys Investment Climate Surveys A FMCG stores include grocers, general stores, chemists, food stores, and cosmetic stores; most are small family businesses. Modern-format stores include outlets in the organized sector, such as Pantaloon and Reliance. The sample was stratified according to segment-specific criteria. FMCG stores were stratified on the basis of turnover, number of salespeople, number of FMCG products, and presence of cooling equipment. Consumer durables and modern- format stores were stratified on the basis of turnover; however, high-end modern format stores2 were completely enumerated. Sample Size The sample size was determined so as to minimize the standard error in the sample variables, given the available resources for each surveying stratum. Once the sample size was determined, the sample was allocated to strata using Neyman's allocation rule. The resulting sample size was 1,482 stores; it covered 41 towns for FMCG stores and 29 towns for modern-format stores.3 Final Sample Summary Because of the high nonresponse rate, the final sample consisted of 1,433 stores and was distributed across industry segments as shown in table A.5. Software/ITES ICS Sampling Frame and Selection Process The Software/ITES ICS 2006 covered seven states: Andhra Pradesh, Delhi, Haryana, Karnataka, Maharashtra, Tamil Nadu, and Uttar Pradesh. Across the seven states, samples were drawn from two subsectors: software and ITES. The frame used for the sampling was the National Association of Software and Services Companies business directory. Unlike the sample for the Manufacturing ICS, the software/ITES sample was allocated across major cities rather than across states. Table A.5 Sample Allocation, Actual Coverage, and Distribution of Retail ICS 2006, by Sector Type of store Sample allocation Actual coverage Actual sample distribution (%) FMCG 1,199 1,244 86.8 Modern format 283 189 13.2 Total 1,482 1,433 100.0 Source: Retail ICS 2006. 121 Investment Climate Surveys A Sample Size The target sample size was set at 364 firms across the seven states. The final number of acceptable survey returns was 273 (table A.6). Table A.6 Actual Coverage and Distribution of Software/ITES ICS 2006, by Subsector Subsector Actual coverage Actual sample distribution (%) Software 213 74.4 ITES 60 25.6 Total 273 100.0 Source: Software/ITES ICS 2006. Table A.7 Actual Coverage and Distribution of Software/ITES ICS 2006, by City City Actual coverage Actual distribution (%) Bangalore 29 10.6 Chennai 65 23.8 Delhi 45 16.5 Hyderabad 17 6.2 Kolkata 0 0.0 Mumbai 62 22.7 Pune 17 6.2 Other cities 38 13.9 Total 273 100.0 Source: Software/ITES ICS 2006. Table A.8 Actual Coverage and Distribution of Software/ITES ICS 2006, by State State Actual coverage Actual distribution (%) Andhra Pradesh 17 6.2 Delhi 45 16.5 Haryana 25 9.2 Karnataka 29 10.6 Maharashtra 79 28.9 Tamil Nadu 68 24.9 Uttar Pradesh 10 3.7 Total 273 100.0 Source: Software/ITES ICS 2006. 122 Investment Climate Surveys A Final Sample Summary The actual sample was allocated between the largest cities with the highest concen- tration of software and ITES firms (table A.7). The resulting distribution across states is reported in table A.8. Notes 1. The 41 cities covered by the survey are as follows: Ahmedabad, Bangalore, Bhopal, Bhubaneswar, Chandigarh, Chennai, Coimbatore, Cuttack, Delhi, Dhanbad, Faridabad, Ghaziabad, Greater Mumbai, Guntur, Gurgaon, Gwalior, Hubli-Dharwad, Hyderabad, Indore, Jaipur, Jalandhar, Jamshedpur, Kanpur, Kochi, Kolkata, Kota, Kozhikode, Lucknow, Ludhiana, Madurai, Mangalore, Mysore, Nagpur, Nashik, Noida, Patna, Pune, Surat, Vadodara, Vijayawada, and Visakhapatnam. 2. Those with sales exceeding Rs 500,000. 3. Modern-format stores are not covered in Dhanbad, Guntur, Gwalior, Hubli-Dharwad, Jalandhar, Jamshedpur, Kota, Mangalore, Mysore, Nashik, Noida, and Vijayawada. 123 Appendix B Econometric Analysis of the Effect of Investment Climate on Firm Performance The effect of investment climate on the performance of firms can be measured in a number of ways, depending on the chosen measure of performance. For the India Investment Climate Assessment (ICA), the effect of investment climate on produc- tivity was measured in terms of all or some of the following economic performance measures: total factor productivity (TFP) or labor productivity, employment demand, wages, exports, and foreign direct investments. Five categories of invest- ment climate variables are taken into account: (a) infrastructure; (b) red tape, cor- ruption, and crime; (c) finance and accountability; (d) labor skills;1 and (e) other control variables. The methodology adopted to evaluate the effect of investment climate on pro- ductivity is that proposed by Escribano and Guasch (2005), which seeks to provide elasticities or semielasticities of investment climate variables on productivity that are robust (equal signs and of similar magnitudes) to different measures of productiv- ity. The same methodology has been adopted for the productivity analysis of the ICA reports published by the World Bank for Chile, China, Colombia, Malaysia, Mexico, the Philippines, and Thailand.2 The alternative productivity measures consider (a) using different functional forms of the production functions (Cobb-Douglas and Translog); (b) using different assumptions on technology and market conditions to get consistent estimators based on Solow residuals, ordinary least squares (OLS), or random effects; and (c) restricting or not restricting input-output elasticities to be the same across industries (see table B.1). The choice of the econometric specification and the choice of the most appro- priate measure of productivity, for all the sectors considered, had to take into account two types of problems. The first is the endogeneity of the investment climate variables. When any of the production function inputs is influenced by common causes affecting productivity, such as investment climate variables or other plant 125 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.1 Definitions of Firm Performance Variables Measure of firm performance Definition Survey used Labor productivity Sales per employee Unorganized Manufacturing ICS, Retail ICS TFP--Solow residuals TFP estimated using Solow residuals Manufacturing ICS, Software/ITES ICS TFP--Cobb-Douglas OLS TFP estimated using Cobb-Douglas Manufacturing ICS, OLS estimator Software/ITES ICS TFP--Cobb-Douglas TFP estimated using Cobb-Douglas random Manufacturing ICS, random effects effects estimator Software/ITES ICS TFP--Translog OLS TFP estimated using Translog OLS estimator Manufacturing ICS, Software/ITES ICS TFP--Translog random TFP estimated using Translog random effects Manufacturing ICS, effects estimator Software/ITES ICS Employment demand Total number of full-time workers (permanent Manufacturing ICS, or temporary) Unorganized Manufacturing ICS, Software/ITES ICS Wages Total expenditures on personnel (deflated Manufacturing ICS, by using the Producer Price Indexes, base Software/ITES ICS 2002) divided by total number of permanent and temporary workers Exports Dummy variable that takes value 1 if Manufacturing ICS, exports are greater than 10 percent Software/ITES ICS Foreign direct Dummy variable that takes value 1 Manufacturing ICS, investments if any part of the capital of the firm Software/ITES ICS is foreign Source: Authors' compilation. characteristics, one should expect the productivity to be correlated with the produc- tion function inputs. Inputs should therefore be treated as endogenous regressors. The second problem is missing variables. A large number of individual investment climate observations were missing at the plant level, which represents one of the most important problems of the Investment Climate Surveys (ICSs). The solution for the first problem has been to use the industry-region-size aver- age of the investment climate variables that are thought to be endogenous on the basis of an a priori hypothesis supported by the findings of the relevant literature. Such variables are the greatest number of days to clear customs to export, a dummy for research and development (R&D), and an unskilled workforce (see note [b] to table B.2). Substituting the industry-region-size for the missing investment climate observations has also been the adopted solution for the issue of missing variables. The industry-region-size averages have also been used to replace missing values in the productivity function variables, with the exception of cost of rental capital in the retail sector regressions, which was replaced with values obtained by regressing the cost of capital on selling area plus sector and size controls for the available data. 126 Econometric Table B.2 Manufacturing Investment Climate Elasticities and Semielasticities with Respect to Productivity, Unrestricted Estimation Two-step estimation Single-step estimation Solow residuals Cobb-Douglas Translog Analysis Categories of Random Random Random ICA variables Explanatory ICA variables OLS effects OLS effects OLS effects Infrastructure Greatest number of days to clear of customs to exporta 0.040*** 0.033 0.041*** 0.030 0.017 0.001 the Dummy for own generator 0.084*** 0.082** 0.094*** 0.179*** 0.112*** 0.189*** Effect Water supply from public sourcesb 0.002*** 0.002*** 0.002*** 0.001** 0.001*** 0.001** of Shipment losses in the domestic marketb 0.0003 0.0002 0.0030 0.0060 0.0040** 0.0050 Investment Dummy for own transport 0.041 0.056 0.058 0.145** 0.060 0.148*** Dummy for Web page 0.023 0.016 0.035 0.077* 0.051** 0.056 Red tape, corruption, Dummy for security 0.072*** 0.058 0.065*** 0.117*** 0.067*** 0.093** Climate and crime Sales reported to taxesb 0.0001 0.0001 0.0001 0.0020* 0.0010 0.0020* Workforce reported to taxesb 0.001 0.001 0.001 0.002 0.002*** 0.002** on Dummy for payments to speed Firm up bureaucracy 0.088*** 0.095*** 0.099*** 0.083** 0.065*** 0.058* Dummy for interventionist labor Performance regulation 0.031 0.037 0.024 0.027 0.025 0.010 Absenteeismb 0.032** 0.021 0.019 0.028 0.032*** 0.040* Finance and corporate Dummy for trade association 0.032 0.032 0.030 0.095** 0.074*** 0.140*** governance Working capital financed by domestic private banksb 0.0004 0.0010 0.0010 0.0020** 0.0001 0.0010 Dummy for loanb 0.046* 0.044 0.053** 0.065 0.063*** 0.086** Dummy for external audit 0.094*** 0.101* 0.040 0.102* 0.079*** 0.133** (continued) 127 B B 128 Table B.2 (continued) Two-step estimation Single-step estimation Solow residuals Cobb-Douglas Translog Econometric Categories of Random Random Random ICA variables Explanatory ICA variables OLS effects OLS effects OLS effects Quality, innovation, Dummy for R&Da 0.227** 0.183 0.192** 0.379*** 0.156** 0.313** and labor skills Dummy for product innovation 0.037* 0.040 0.022 0.032 0.014 0.010 Analysis Dummy for foreign licenseb 0.145** 0.167** 0.156** 0.218** 0.141*** 0.179** Dummy for internal trainingb 0.062* 0.051 0.076** 0.034 0.095*** 0.078 of Unskilled workforcea 0.004* 0.004 0.003* 0.003 0.002 0.003 the Workforce with computer 0.002** 0.002** 0.002** 0.003*** 0.002*** 0.003*** Effect Other control variables Dummy for incorporated company 0.041 0.032 0.036 0.152*** 0.058** 0.127*** of Age 0.054*** 0.052** 0.057*** 0.100*** 0.047*** 0.066*** Investment Share of exportsb 0.001 0.001 0.001 0.003*** 0.001* 0.002*** Trade unionb 0.001 0.001 0.001* 0.002*** 0.001 0.002*** Strikesb 0.071** 0.085* 0.019 0.015 0.003 0.009 Climate Dummy for medium firms 0.071*** 0.083** 0.067*** 0.095** 0.030 0.029 Number of observations 5,230 5,230 5,230 5,230 5,230 5,230 on R-squared 0.07 0.07 0.88 0.88 0.90 0.90 Firm Source: Authors' calculations. Performance Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. These variables are instrumented with the industry-region-size average. b. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). Econometric Analysis of the Effect of Investment Climate on Firm Performance B Outliers have been dropped from the sample. More details about the measure of firm performance and about the estimation technique are provided within the sector- specific results. Once the elasticities or semielasticities of investment climate variables on per- formance have been estimated, the partial direct effect of each investment climate variable on each measure of productivity can be obtained by evaluating the effect of the average investment climate and control variables on the sample average value of the performance measures: productivity, employment, wages, exports, and foreign direct investment (FDI). This evaluation can be made at different aggregation levels (for example, for the whole sector, by industry, or by size of the firm). Through use of the Olley and Pakes (1996) decomposition method, aggregate productivity can be decomposed into two terms: the productivity of the average firm and a covariance term that provides information on whether the economy is able to efficiently reallocate resources from less productive establishments to more produc- tive firms. When the efficiency term is positive, the larger it is, the higher the share of sales that goes to more productive firms and the higher the total productivity of the sector. When the efficiency term is negative, the result is allocation inefficiencies because larger shares of sales are going to less productive firms. Manufacturing Sector An empirical analysis of firm performance has been conducted for the manufactur- ing sector in India at the plant level, taking advantage of the data available in the Manufacturing ICS 2006 for the years 2002, 2003, and 2004. Six dimensions of per- formance have been measured (total factor productivity, labor productivity, exports, FDI, wages, and employment), each through three different approaches (Solow residuals, Cobb-Douglas, and Translog production functions). The performance analysis is done for the whole manufacturing sector as well as by industry. Industries covered by the survey are food, apparels, textiles and leather, chemicals and chemical products, plastics and rubbers, nonmetallic products, struc- tural metal and metal products, and machinery and equipment. The India Manufacturing ICS 2006 data represent a panel short in the time dimen- sion--three years of observations--but long in the cross-section dimension. How- ever, information about the investment climate variables is available for 2004 only; in the empirical application, it is assumed that the investment climate characteristics for these three years (2002, 2003, and 2004) are constant at the plant level, and they are therefore treated as observable fixed effects. The productivity equations are not esti- mated in first differences to avoid losing all the information on ICA variables (which are constant over time by construction); they are instead estimated in levels (logs), adding dummy variables to control for the three years and the eight industries. Because any empirical evaluation on the productivity effect of investment climate variables might critically depend on the particular way productivity is measured, several productivity measures are employed. Each measure is used to evaluate the 129 Econometric Analysis of the Effect of Investment Climate on Firm Performance B effect of the investment climate variables on productivity for both the entire manu- facturing sector and the individual regions. The first measure is the Solow residuals.3 This method involves a two-step proce- dure by which, in the first step, the productivity measure (logPi) is estimated and, in the second step, the investment climate elasticities are obtained. The production function estimated in the first step is logYj ,it sj logLj ,L ,it sj logMj ,M ,it sj logKj ,K ,it logPj . ,it Having estimated the productivity P from the above equation, one can estimate the investment climate elasticities and semielasticities by OLS and by random effects from the following two-step estimator: log Pj ,it aICICi aCCi aDsDj aDTDt aP uj . ,it The second productivity measure is obtained through a parametric estimation of an extended Cobb-Douglas production function by OLS. Next, the well-known prob- lem of endogeneity of the inputs must be addressed, so the unobserved firm-specific fixed effects are proxied by a long list of firm-specific observed fixed effects coming from the investment climate information. The estimated equation is the following: log Yj,it aj,LlogLj,it aj,M logMj,it aj,Klog Kj,it aICICi aCCi aDsDj + aDTDt + ap + ej,it. The effect of investment climate variables on productivity is then estimated as the extended residuals of the production function: log P^j,it aICICi aCCi aDsDj aDTDt aP jj . ,it The third measure consists of the same procedure described earlier, but a translog rather than Cobb-Douglas functional form is used for the production function: 1 log Yj,it aj log Lj ,L ,it aj,Mlog Mj ,it aj log Kj ,K ,it 2aj ,LL(log Lj )2 ,it 1 1 2 aj ,MM (log Mj )2 ,it 2aj ,KK (log Kj )2 ,it aj,LM (log Lj )(log Mj ) ,it ,it aj,LK (log Lj )(log Kj ) ,it ,it aj ,MK (log Mj )(log Kj ) ,it ,it aICICi aCCi aDsDj aDTDt ap mj,it The endogeneity of some investment climate variables is accounted for by using the region-industry-size average of the plant-level investment climate variables instead of the crude investment climate variables, which is a common solution in panel data studies at the firm level. By means of the Olley and Pakes decomposition, the aggregate log productivity (by industry, state, or firm size) is expressed as the sum of the productivity of the average establishment and the allocative efficiency--that is, the covariance between the share of sales and productivity. This information complements the information provided by the investment climate elasticities and semielasticities estimated in the productivity equations. Consider, for example, the case in which a given investment 130 Econometric Analysis of the Effect of Investment Climate on Firm Performance B climate variable has a low marginal effect on productivity but most of the firms suffer the effect. Then the overall effect of that investment climate variable on aver- age productivity could be large. If the firms that are not constrained by that variable are those with the largest market shares, the effect of that variable will also be large in terms of efficiency. The Olley and Pakes decomposition rule is log Pq a^ICICq a^CCq a^DsDq a^p u^aq ,i Nqa^ICc^ov(slog , ICq ) Y q,i ,i Nq a^Cc^ov(slog ,Cq ) Y q,i ,i Nqa^Dsc^ov(slog ,Dj) Y q,i Nqc^ov(slogY,u^aq ), q,i ,i where the set of parameters comes from the two-step estimation with the restricted Solow residuals as a dependent variable. For an estimate of the effect of investment climate and control variables on sev- eral measures of economic performance, the following simultaneous equations that estimate productivity, employment demand, real wages, probability of exporting, and probability of receiving FDI are used: logPj ,it aP aICICi aCCi aDsDj aDTDt (vP,i eP ) ,j,it LogLj,it gL gPlogPj,it gwlogWj,it gLICi gCCi gDsDj gDTDt (vL,i eL,j,it) LogWj,it bW bPlogPj,it bICICi bCCi bDSDj bDTDt (vW,i eW,j,it) yj Exp ,it dExp dPlogPj ,it dICICi dCCi dDSDj dDTDt (vExp,i eExp ) ,j,it yj FDI ,it rFDI rPlogPj ,it rICICi rCCi rDSDj rDTDt (vFDI,i eFDI ) ,j,it Because results for investment climate elasticities and semielasticities were robust to the choice of productivity measure, this part of the analysis concentrated only on Solow residuals for the restricted case and the unrestricted-by-industry case. The equations for export and FDI measure, respectively, the probability that the plant exports its production and the probability that any part of the capital of the firm is foreign. These equations are estimated through linear probability models, that approximate well the probit and logit nonlinear models when the variables are evaluated at their sample means and give easily interpretable parameters.4 Further- more, with this estimator the endogeneity of investment climate variables can be addressed by standard instrumental variable approaches such as the two-stage least squares and general method of moments approaches. Each equation is estimated by two-stage least squares using heteroskedasticity- robust standard errors because the endogeneity of certain investment climate vari- ables induces a correlation between those investment climate variables and the errors of the system of equations and creates simultaneous equation biases and inconsis- tencies in least squares estimators. This correlation is in general mitigated by replac- ing those plant-level investment climate variables by their region-industry averages. However, for some other explanatory variables, such as productivity, wages, exports, and FDI, the endogeneity is intrinsic because of the simultaneous structure of the system of equations. Tables B.3 to B.6 present the empirical results. 131 Table B.3 Manufacturing Two-Stage Least Squares Estimation of Employment Equation Unrestricted Solow residualsa Blocks of ICA variables Explanatory ICA variables Coefficient Contribution (%) Productivity 0.349** 11.3 Real wages 0.066** 14.9 Infrastructure Dummy for own generator 0.243*** 4.0 Shipment losses in domestic marketb 0.015*** 0.6 Dummy for own transport 0.115** 0.4 Dummy for Web page 0.291*** 2.9 Red tape, Sales reported for tax purposes 0.263*** 5.5 corruption, Dummy for securityb 0.002*** 6.1 and crime Dummy for payments to obtain a contract with governmentb 0.081** 0.8 Absenteeismb 0.097*** 1.0 Finance and Largest shareholderb 0.006*** 14.7 corporate Dummy for chamber of commerceb 0.204*** 5.4 governance Dummy for overdraft 0.093*** 1.6 Borrowing denominated in foreign currencyb 0.007** 0.3 Quality, innovation, Dummy for quality certificationb 0.253*** 1.7 and labor skills Dummy for R&D 0.282*** 2.3 Dummy for foreign licenseb 0.318*** 0.5 Skilled workforcec 0.005*** 10.0 Dummy for internal trainingb 0.262*** 1.2 Workforce with computer 0.005*** 2.3 Other control Dummy for incorporated company 0.537*** 5.1 variables Age 0.134*** 12.6 Dummy for exporter 0.571*** 2.6 Dummy for FDI 0.409*** 0.3 Trade unionc 0.018*** 4.2 Instruments First-stage R-squaredd 0.17 evaluation Partial R-squarede 0.01 Partial R-squared F-test (p-value)f 0.00 Hansen test (p-value)g 0.186 Number of observations 4,902 Source: Authors' calculations. Note: ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for product innovation (industry-region-size average), number of inspections (industry-region-size average), dummy for current account (industry-region- size average), dummy for interventionist labor regulation (industry-region-size average), number of competitors (industry-region- size average), and water supply from public sources. b. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). c. These variables are instrumented with the industry-region-size average. d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 132 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.4 Manufacturing Two-Stage Least Squares Estimation of Real Wages Equation Unrestricted Solow residualsa Blocks of ICA variables Explanatory ICA variables Coefficient Contribution (%) Productivity 0.363* 5.3 Infrastructure Dummy for own generator 0.255*** 2.1 Shipment losses in the domestic marketb 0.027*** 0.4 Dummy for Web page 0.101** 0.5 Red tape, Sales declared for tax purposes 0.004*** 4.8 corruption, Dummy for interventionist labor regulation 0.129** 1.7 and crime Dummy for payments to speed up bureaucracy 0.126*** 1.0 Finance and Largest shareholder 0.002** 1.9 corporate Dummy for overdraft 0.199*** 1.7 governance Dummy for external audit 0.103 1.4 Quality, innovation, Dummy for quality certificationb 0.140** 0.5 and labor skills Dummy for R&D 0.127** 0.5 Dummy for product innovation 0.205*** 1.9 Skilled workforcec 0.011*** 8.9 Experience of the managerb 0.088*** 2.7 Other control Number of competitorsc 0.055* 2.2 variables Share of exportsb 0.003*** 0.5 Dummy for medium firms 0.364*** 1.9 Dummy for large firms 0.425*** 0.9 Instruments First-stage R-squaredd 0.070 evaluation Partial R-squarede 0.020 Partial R-squared F-test (p-value)f 0.000 Hansen test (p-value)g 0.293 Number of observations 4,786 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes staff with computer (industry- region-size average), dummy for current account, dummy for internal training, dummy for foreign license, and water supply from public sources. b. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). c. These variables are instrumented with the industry-region-size average. d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 133 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.5 Manufacturing Two-Stage Least Squares Estimation of Probability of Exporting Equation Unrestricted Solow residualsa Blocks of ICA variables Explanatory ICA variables Coefficient Contribution (%) Productivity 0.164** 149.4 Infrastructure Number of days to clear customs to exportb 0.013** 30.0 Dummy for own generator 0.033** 14.1 Shipment losses in the domestic marketc 0.002* 2.2 Dummy for own transport 0.060*** 5.8 Dummy for e-mail 0.048*** 23.9 Dummy for Web page 0.068*** 16.5 Red tape, Dummy for security costc 0.033*** 18.3 corruption, Dummy for interventionist labor regulationb 0.237*** 201.2 and crime Dummy for payments to speed up bureaucracy 0.026** 12.6 Finance and Working capital financed by domestic corporate private banksc 0.001*** 31.2 governance Dummy for current account 0.046** 45.3 Borrowing denominated in foreign currencyb 0.008*** 11.0 Quality, innovation, Dummy for R&D 0.030** 5.9 and labor skills Dummy for outsourcing 0.068*** 11.4 Dummy for external trainingc 0.051** 2.2 Experience of manager 0.071*** 138.3 Other control Dummy for medium firms 0.075*** 11.3 variables Dummy for large firms 0.172*** 5.4 Instruments First-stage R-squaredd 0.050 evaluation Partial R-squarede 0.010 Partial R-squared F-test (p-value)f 0.000 Hansen test (p-value)g 0.484 Observations 5,246 Source: Authors' calculations. Note: * = significant at 10 percent; ** = significant at 5 percent; *** = significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for foreign license, dummy for external audit, dummy for product innovation, and number of competitors (industry-region-size average). b. These variables are instrumented with the industry-region-size average. c. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 134 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.6 Manufacturing Two-Stage Least Squares Estimation of Probability of Receiving FDI Unrestricted Solow residualsa Blocks of ICA variables Explanatory ICA variables Coefficient Contribution (%) Productivity 0.043* 276.5 Infrastructures Number of power outages 0.004*** 83.5 Dummy for own generator 0.008* 26.6 Shipment losses in the domestic marketb 0.0003* 2.8 Dummy for Web page 0.010*** 18.3 Red tape, Dummy for conflicts with court involvedc 0.075*** 61.8 corruption, Dummy for illegal payments for protectionc 0.088*** 56.7 and crime Finance and Working capital financed by internal fundsb 0.0001* 45.4 corporate Dummy for current account 0.014*** 94.5 governance Quality, innovation, Dummy for foreign licenseb 0.028 8.5 and labor skills Unskilled workforce 0.0003*** 50.1 Dummy for external trainingb 0.026** 8.7 Other control Dummy for exporter 0.020** 15.4 variables Dummy for large firms 0.029*** 4.3 Instruments First-stage R-squaredd 0.050 evaluation Partial R-squarede 0.010 Partial R-squared F-test (p-value)f 0.000 Hansen test (p-value)g 0.195 Number of observations 5,007 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for foreign license, dummy for external audit, dummy for product innovation, and number of competitors (industry-region-size average). b. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). c. These variables are instrumented with the industry-region-size average. d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. Unorganized Manufacturing Sector An empirical analysis of labor productivity at the firm level was also conducted for the unorganized manufacturing sector, along the lines of the analysis for the retail sector. This analysis was based on the Unorganized Manufacturing ICS 2006 for India. 135 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Productivity Analysis The productivity analysis is based on the concept of labor productivity, instead of TFP,5 because of a lack of adequate information on the capital stock. Labor produc- tivity is computed as value added divided by total employment. The effect of the investment climate is estimated by regressing estimated labor productivity on the log of wages, log of estimated rental, various investment climate variables, and other industry and firm-size control variables: Y log L g0 g1 log wit g2 log rit g3ICi g4Cit ei . ,YL it As in other surveys, one problem is that of missing variables. The solution again has been to use industry-region-size averages of the investment climate variables. An exception has been made for the cost of rental capital. For all firms, this variable has been replaced with values predicted by regressing the cost of capital on the net worth of land and machinery and on industry and regional controls for the available data. Outliers were dropped from the sample. In the first cut, an OLS estimation is used because after controlling for observable fixed effects (investment climate), the investment climate variables can be treated as observable fixed effects. Robust standard errors are used to control for the typical heteroskedasticity of firm-level data. The results for these estimates are reported in table B.7. However, these estimates are likely to be inconsistent given that some of the reported investment climate variables are themselves likely to be endogenous. Better estimates can be obtained by using an instrumental variable method. This method is used in the second set of regressions, which instrument for the bribes that the firm reports paying. The instruments chosen are whether the firm is registered, the num- ber of inspector visits, and regional dummies. The results are presented in the last column of table B.7. Employment Demand A firm's conditional employment demand can be represented as follows: Y log Lit b0 b1 log L b2 log wit b3 log rit b4ICi b5Cit eL ,it it As explained earlier, this equation should be estimated through two-stage least squares because labor productivity (Y/L) is correlated with the error term. The instruments used in this estimation are losses due to theft, whether the legal system was used to settle disputes, the number of inspections the firm faced in a year, its age, and whether it faces seasonally varying demand. Estimation results for the condi- tional employment demand equation are reported in table B.8. The analysis on employment demand shows that in the unorganized manufac- turing sector there are strong increasing returns to scale.A 1 percent increase in labor productivity generates a 0.9 percent increase in the employment hours. 136 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.7 Unorganized Sector Investment Climate Elasticities with Respect to Log of Labor Productivity Instrumental Explanatory variables OLS variable Log of wages 0.672*** 0.693*** Log of rent 0.115*** 0.126*** Infrastructure Power outages 0.0001 0.0001 Difficulty in transport 0.133* 0.132* Red tape, corruption, and crime Losses from theft 0.023 0.023 Disputes requiring third-party arbitrage 0.092 0.076 Bribe payments 10.710** 33.913* Finance and accountability Loan or overdraft from financial institution 0.371*** 0.372*** Borrowing of funds from informal source 0.131** 0.108 Difficulty in complying with labor regulations 0.086* 0.090** Lack of ownership of land occupied by building 0.011 0.007 Experience of manager 0.002 0.002 Manager's formal vocational training 0.001 0.009 Other control variables Age of enterprise 0.003* 0.004* Tiny enterprisesa 0.400*** 0.415*** Microenterprises 0.422*** 0.421*** Constant 7.981*** 7.271*** R-squared 0.186 0.174 Number of observations 1,328 1,328 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes industry-specific dummy variables. a. "Household enterprises" was the excluded category. Retail Sector A detailed empirical analysis of labor productivity at the firm level was conducted for traditional fast-moving consumer goods (FMCG) stores and for modern-format stores in India. The productivity analysis is based on the concept of labor productivity, instead of TFP because of a lack of information on some of the basic inputs of the production functions, such as the capital stock. Labor productivity is computed as value added divided by total employment. 137 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.8 Unorganized Sector Investment Climate Elasticities with Respect to Employment Demand Explanatory variables Instrumental variable Log of productivity 0.905*** Log of wages 0.534*** Log of rent 0.047 Infrastructures Power outages 0.001** Difficulty in transport 0.0285*** Red tape, corruption, and crime Losses from theft 0.005 Bribe payments 0.015 Finance and accountability Loan or overdraft from financial institution 0.131 Borrowing of funds from informal source 0.243*** Difficulty in complying with labor regulations 0.159*** Lack of ownership of land occupied by building 0.178*** Other control variables Age of enterprise 0.004** Direct sales to a large company 0.101 Constant 5.651*** F-test (p-values) 0.000 Number of observations 1,328 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes industry specific dummy variables. The effect of the investment climate on labor productivity is estimated by regress- ing the log of labor productivity on the log of wages on the investment climate vari- ables and on other control variables, including industry and firm size: Y log L g0 g1 log wit g2 log rit g3ICi g4Cit ei . ,YL it Consistent estimates of the coefficients of this equation can be obtained through an OLS estimator, after controlling for observable fixed effects (investment climate) because the investment climate variables are treated as observable fixed effects. Robust standard errors are used to control for the typical heteroskedasticity of firm- level data. Table B.9 shows the results. The Olley and Pakes decomposition rule for labor productivity is Y Y Njt Y~ , L s~Yj ,it jt L jt L j,it i 1 where the first term on the right-hand side is the average labor productivity of indus- try j, and the second term on the right-hand side measures the allocative efficiency or covariance between the share of sales and labor productivity. 138 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.9 Retail Investment Climate Elasticities and Semielasticities with Respect to Log of Labor Productivity Traditional FMCG Modern-format stores Explanatory variables coefficient coefficient Log of wages 0.380*** 0.131* Infrastructures Average duration of power outages n.s. n.s. Losses from power outages 0.049*** n.s. Computer 0.715*** 0.329* Red tape, corruption, and crime Criminal attempts n.s. n.s. Security 0.187** 0.373** Store visited by any agency 0.237*** n.s. Third-party arbitrage 0.468*** 0.783*** Finance and accountability Credit line 0.239** n.s. Current account 0.391*** n.s. External audit 0.211** n.s. Financing--internal funds n.s. n.s. Financing--family sources 0.003*** 0.011* Financing--informal sources 0.008*** n.s. Labor skills Experience of manager 0.102*** 0.229*** Other control variables Inventory 0.068** n.s. Microenterprises 1.195*** 2.163*** Small firms 1.157*** 1.164*** Medium firms 0.739*** 1.180*** Extra-large firms -- 0.982*** Industry dummies Modern-format stores -- -- Consumer durable stores -- -- Constant 3.062*** 1.003* Number of observations 1,182 1,812 R-squared 0.22 0.44 Source: Authors' calculations. Note: -- = not available; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent; n.s. = not significant. 139 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.10 Software and ITES Investment Climate Unrestricted Elasticities and Semielasticities with Respect to Productivity Two-step estimation Single-step estimation Blocks of Solow Cobb- ICA variables Explanatory ICA variables residuals Douglas Translog Infrastructure Dummy for own generatora 0.599** 0.442* 0.321 Dummy for high-speed Internet connectiona 0.734* 0.604 0.518 Dummy for unavailability of quality Internet connectiona 0.512* 0.596** 0.492* Dummy for Internet connection used to deliver services to the establishment's clientsa 0.847** 0.752** 0.736** Dummy for conflicts in courtsa 0.698* 0.616* 0.599* Red tape, Sales never repaidb 0.278** 0.357*** 0.389*** corruption, Manager's time spent in and crime bureaucratic issuesa 0.014 0.017 0.023* Working capital financed by state-owned banksa 0.012** 0.015** 0.014** Finance and Dummy for overdrafta 0.427* 0.449* 0.403 corporate R&D expendituresb 0.142** 0.190*** 0.219*** governance Quality, innovation, Advertising expendituresb 0.078* 0.123*** 0.136*** and labor skills Dummy for incorporated companya 0.816* 0.692* 0.772* Other control Dummy for microfirms 0.465 0.647 0.553 variables Number of observations 175 175 175 R-squared 0.27 0.76 0.79 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry and year dummies and a constant term. a. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). b. These variables are instrumented with the industry-region-size average. Software and Information Technology­Enabled Services Sector An empirical analysis of firm performance was conducted for the software and infor- mation technology­enabled services (ITES) sector in India on the basis of the data collected in the Software/ITES ICS 2006. Five measures of performance have been measured (productivity, employment, wages, exports, and FDI) following the same empirical strategy described for the manufacturing sector. Tables B.10 to B.14 report the results. 140 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.11 Software and ITES Two-Stage Least Squares Estimation of Employment Equation Unrestricted Solow residualsa Blocks of ICA Contribution variables Explanatory ICA variables Coefficient (%) Productivity 0.363* 22.12 Real wages 0.248*** 54.30 Infrastructure Dummy for own generatorb 0.525** 7.59 Red tape, Sales never repaidc 0.179*** 18.65 corruption, Manager's time spent in bureaucratic issuesb 0.020*** 4.39 and crime Finance and Working capital financed by private commercial corporate banksb 0.010* 1.66 governance Dummy for loanb 0.902*** 2.85 Rental landb 0.008*** 14.34 Quality, innovation, Dummy for quality certificationb 0.513** 4.81 and labor skills Dummy for external trainingb 0.390* 5.30 Dummy for payments for royaltiesb 0.767* 1.45 Other control Age 0.548*** 33.14 variables Percentage of direct exportsb 0.008*** 8.69 Dummy for importerb 0.414* 3.28 Instruments First-stage R-squaredd 0.520 evaluation Partial R-squarede 0.086 Partial R-squared F-test (p-value)f 0.002 Hansen test (p-value)g 0.255 Number of observations 175 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry and year dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for high-speed Internet connection, dummy for Internet connection used to deliver services to the establishment's clients, working capital financed by state-owned banks, advertising expenditures (industry-region-size averages), and dummy for incorporated company. b. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). c. These variables are instrumented with the industry-region-size average. d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 141 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.12 Software and ITES Two-Stage Least Squares Estimation of Real Wages Equation Unrestricted Solow residualsa Blocks of ICA Contribution variables Explanatory ICA variables Coefficient (%) Productivity 0.788*** 22.70 Infrastructure Number of power outagesb 0.739*** 47.78 Red tape, Sales never repaidb 0.180* 8.12 corruption, Security costb 0.155*** 14.12 and crime Finance and Largest shareholderb 0.027** 22.95 corporate Working capital financed by family or friendsc 0.035** 1.16 governance Working capital financed by informal sourcesc 0.064** 0.30 Quality, innovation, Dummy for new productc 0.545* 2.64 and labor skills Other control Age 0.480** 13.76 variables Instruments First stage R-squaredd 0.479 evaluation Partial R-squarede 0.099 Partial R-squared F-test (p-value)f 0.007 Hansen test (p-value)g 0.548 Number of observations Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry and year dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for high-speed Internet connection, dummy for Internet connection used to deliver services to the establishment's clients, working capital financed by state-owned banks, advertising expenditures (industry-region-size averages), and dummy for incorporated company. b. These variables are instrumented with the industry-region-size average. c. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 142 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.13 Software and ITES Two-Stage Least Squares Estimation of Probability of Exporting Equation Unrestricted Solow residualsa Blocks of ICA Contribution variables Explanatory ICA variables Coefficient (%) Productivity 0.145** 62.88 Infrastructure Number of power outagesb 0.078* 77.72 Dummy for Internet connection uses to make purchases for this establishmentc 0.149** 17.14 Red tape, Security costb 0.030*** 45.22 corruption, Manager's time spent in bureaucratic issuesc 0.005** 8.24 and crime Finance and Working capital financed by family or friendsc 0.003** 1.67 corporate Working capital financed by informal sourcesc 0.024*** 1.79 governance Quality, innovation, Staff for whom knowledge of English is criticalc 0.002* 19.95 and labor skills Dummy for external trainingb 0.116* 11.50 Other control Age 0.123** 52.84 variables Dummy for microfirms 0.312** 9.40 Instruments First-stage R-squaredd 0.862*** evaluation Partial R-squarede 0.115 Partial R-squared F-test (p-value)f 0.005 Hansen test (p-value)g 0.675 Number of observations 171 Source: Authors' calculations. Note: * = significant at 10 percent; ** = significant at 5 percent; *** = significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for high-speed Internet connection, dummy for Internet connection used to deliver services to the establishment's clients, working capital financed by state-owned banks, advertising expenditures (industry-region averages), and dummy for incorporated company. b. These variables are instrumented with the industry-region-size average. c. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instruments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 143 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Table B.14 Software and ITES Two-Stage Least Squares Estimation of Probability of Receiving FDI Unrestricted Solow residualsa Blocks of ICA Contribution variables Explanatory ICA variables Coefficient (%) Productivity 0.107* 124.57 Infrastructure Number of power outagesb 0.129** 340.97 Dummy for own transportb 0.516* 55.86 Red tape, Dummy for cybercrimec 0.377*** 7.58 corruption, and crime Finance and Working capital financed by informal sourcesc 0.009** 2.25 corporate Rental landc 0.002*** 60.35 governance Quality, innovation, Dummy for payment of royaltiesc 0.224* 6.77 and labor skills Other control Share of direct exportsc 0.002*** 46.48 variables Dummy for small firms 0.125* 51.92 Instruments First-stage R-squaredd 0.143 evaluation Partial R-squarede 0.130 Partial R-squared F-test (p-value)f 0.001 Hansen test (p-value)g 0.886 Number of observations 166 Source: Authors' calculations. Note: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. Each regression includes a set of industry dummies and a constant term. a. Productivity is endogenous, and the list of variables used as excluded instruments includes dummy for high-speed Internet connection, dummy for Internet connection used to deliver services to the establishment's clients, working capital financed by state-owned banks, advertising expenditures (industry-region averages), and dummy for incorporated company. b. These variables are instrumented with the industry-region-size average. c. These variables are approximated with a proxy (only missing values replaced by the industry-region-size average). d. The first-stage R-squared is from the regression of productivity on both the included instruments and the excluded instru- ments. e. The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. f. The F-test of joint significance of the excluded instruments corresponds to the partial R-squared. g. The Hansen test is a test of over-identifying restrictions. The null hypothesis is that the instruments are valid instruments-- that is, uncorrelated with the error term--and therefore the excluded instruments are correctly excluded from the estimated equation. 144 Econometric Analysis of the Effect of Investment Climate on Firm Performance B Notes 1. This category also includes information on innovation for the manufacturing sector. 2. Although this methodology has been widely applied and tested, a number of critiques have been made. The first issue under discussion is the replacement strategy for the missing values in pro- duction function variables, which could result in inefficient estimates if missing observations are random. The second matter concerns whether it would be appropriate to test for the validity of the restrictions on the input-output elasticities. The third point of discussion has been the choice of endogenous variables, which, in the chosen specification of the model, do not include vari- ables that are sometimes considered to be endogenous. The clustering of standard errors has also been suggested as an improvement on the methodology because it would prevent inflation of the significance of variables. 3. The advantage of the Solow residuals method is that it requires neither the inputs to be exoge- nous nor the input-output elasticities to be constant or homogeneous. The drawback is that it requires hypotheses of constant returns to scale and competitive input markets. 4. The parameters measure the change in probability when one of the explanatory variables changes, holding the rest of the explanatory variables constant. 5. Notice that in most of the ICAs done on the manufacturing sector, the analysis was based on TFP and not on labor productivity. See, for example, Escribano and Guasch (2005). 145 Appendix C Investment Climate Index Methodology The microdataset is sourced from the World Bank Investment Climate Survey conducted in 2006 and is based on face-to-face interviews of a representative sample of more than 3,700 entrepreneurs in both manufacturing and retail establishments in 16 Indian states. Forty-six variables that describe the business environment in the 16 states have been identified. For simplicity, these variables have been grouped into three categories: infrastructure, inputs, and institutions (see table C.1). Each of these categories is classified under two dimensions: objective values (cost) and subjective indicators (perception). As a result of this classification, the 46 variables are grouped in six sets that represent the backbone of the Investment Climate Index (ICI) and aim at measuring the cost and quality of infrastructure services, input markets, and institutions. The ICI is constructed by means of a series of three separate aggregations (see box C.1).1 First, to ensure that each component has equal weight, the 46 variables have been grouped in six sets (subindexes) that represent the backbone of the ICI and aim at measuring the cost and quality of infrastructure services, input markets, and institutions. These six subindexes are then further aggregated into three subindexes, one for each category: infrastructure, inputs, and institutions. Finally, these three subindexes are combined into the ICI. At each stage of the aggregation process, principal component and geometric aggregation is employed. The composite ICI has been successfully tested for reliability. Because there is no theoretical model on the estimation of the weights used in the construction of the index, the reliability of the ICI as a predictor of a good investment climate in India was, prior to the analysis, tested by correlating it with other performance indicators. Level of domestic private investment and gross domestic product growth for the 16 states were selected as the benchmark performance indicators. The ICI showed clear and 147 Investment Climate Index Methodology C Table C.1 Variables Used in the Construction of the Composite ICI Cost Perception Infrastructure Hours of power outages last year Perception of electricity Hours of telephone outages last year Perception of telecommunications Percentage of sales lost in transit Perception of transport Percentage of sales lost because of Perception of access to land power outages Days of inventories kept for main input (proxy for quality of transportation) Inputs Excess labor Share of short-term finance obtained from banks Cost of finance: value of collateral required Share of long-term finance obtained from banks to obtain a loan Proximity to raw materials (share of inputs Short-term finance represented by trade finance bought by domestic sources) Proximity to domestic customers Duration of loan Share of firms using new technology Perception of access to finance Trade credit: share of sales sold on credit Perception of labor regulations Trade credit: share of inputs bought on credit Perception of customs Perception of availability of skills Institutions Law and order: security cost Perception of law and order: crime Law and order: losses from theft Perception of corruption Manager's time spent dealing with regulations Perception of licensing and permits Days spent with officials to deal with regulations Quality of administration: consistent interpretation of rules Tax evasion (percentage of sales not declared) Perception of tax administration: rates Days to obtain a telephone connection Perception of tax administration: administration Days to obtain an electrical connection Perception of functioning of judicial system Days to obtain a construction permit Bribes to "get things done" Share of firms reporting officials' request of gifts Share of firms reporting gifts requested to obtain a power connection Share of firms reporting gifts requested to obtain a telephone connection Share of firms reporting gifts requested to obtain a construction permit Share of firms reporting gifts requested to obtain a main operating license Average time to reach a court judgment (weeks) Source: Authors' compilation. 148 Investment Climate Index Methodology C Box C.1 Formulas Used to Construct the ICI Step 1: Factor Variables Analysis Aggregation Index Hours of power outages Hours of telephone outages Factor 1 Sales lost in transit Factor 2 (Factor 1)W1 · (Factor 2)W2 · (Factor 3)W3 Infrastructure cost subindex Sales lost to power outages Factor 3 Days of inventories where, Factor 1 a Hours of power outages b Sales lost to power outages Step 2: Factor Factor Variables Analysis Aggregation Index Analysis Aggregation Index Infrastructure cost index Factor 1 (Factor 1)W1 · (Factor 2)W2 Infrastructure Infrastructure quality index Factor 2 subindex Factor 1 Investment Input cost index Factor 1 (Factor 1)W1 · (Factor 2)W2 Inputs Factor 2 F1W · F2W · F3W Climate Input quality index Factor 2 subindex Factor 3 Index Infrastructure cost index Factor 1 Institutions (Factor 1)W1 · (Factor 2)W2 Infrastructure quality index Factor 2 subindex Source: Authors' construction. significant association with both indicators at the state level. These tests provide a degree of confidence that the composite ICI is a reliable indicator of the investment climate in India. Note 1. Through the use of a stricter distinction between perceptions and objective indicators, four vari- ables classified in table C.1 as inputs perceptions (share of short-term finance obtained from banks, share of long-term finance obtained from banks, trade finance, and loan duration) were reclassified as inputs costs. The ICI was then recalculated to test its robustness. The results were very similar to those of the original ICI. The correlation coefficient among the two indexes was 90 percent, with a significance level of 1 percent. 149 References Acharya, Shankar, Isher Ahluwalia, K. L. Krishna, and Ila Patnaik. 2003. India: Economic Growth, 1950­2000. New Delhi: Indian Council for Research on International Economic Relations. Ahmad, Sadiq. 2006. India's Long-Term Growth Experience: Lessons and Prospects. New Delhi: Sage. Bhaumik, Sumon K., Shubhashis Gangopadhyay, and Shagun Krishnan. 2006."Reforms, Entry, and Productivity: Some Evidence from the Indian Manufacturing Sector." IZA Discussion Paper 2086, Institute for the Study of Labor, Bonn, Germany. Bosworth, Barry P., and Susan M. Collins. 2003."The Empirics of Growth: An Update."Brook- ings Papers on Economic Activity, Brookings Institution Press, Washington, DC. Bosworth, Barry P., Susan M. Collins, and Arvind Virmani. 2007. "Sources of Growth in the Indian Economy." NBER Working Paper 12901, National Bureau of Economic Research, Cambridge, MA. Central Statistical Office. 2007. National Account Statistics. New Delhi: Government of India. Chakravorty, Sanjoy, and Somik V. Lall. 2006. Made in India: The Economic Geography and Political Economy of Industrialization. New Delhi: Oxford University Press. Chand, Vikram A., ed. 2006. Reinventing Public Service Delivery in India: Selected Case Studies. Washington, DC: World Bank. Dollar, David, Mary Hallward-Driemeier, and Taye Mengistae. 2005."Investment Climate and Firm Performance in Developing Economies."Economic Development and Cultural Change 54 (1): 1­31. Dollar, David, Giuseppe Iarossi, and Taye Mengistae. 2002."Investment Climate and Economic Performance: Some Firm-Level Evidence from India." Working Paper 143, Center for Research on Economic Development and Policy Reform, Stanford University, Stanford, CA. Easterly, William, and Ross Levine. 2001. "It's Not Factor Accumulation: Stylized Facts and Growth Models." Working Paper 164, Central Bank of Chile, Santiago. Ernst & Young. 2006. "The Great Indian Retail Story." Ernst & Young, New Delhi. http://www.ey.com/Global/assets.nsf/Sweden/The_Great_Indian_Retail_Story/$file/ The%20Great%20Indian%20Retail%20Story.pdf. Escribano,Alvaro, and J. Luis Guasch. 2005."Assessing the Impact of the Investment Climate on Productivity using Firm-Level Data: Methodology and the Cases of Guatemala, Honduras, and Nicaragua." Policy Research Working Paper 3621, World Bank, Washington, DC. FICCI (Federation of Indian Chambers of Commerce and Industry). 2007. Survey of Emerging Skill Shortages in the Indian Industry, July 2007. New Delhi: FICCI. 151 References Goldar, Bishwanath. 2004."Productivity Trends in Indian Manufacturing in the Pre- and Post- reform Periods." ICRIER Working Paper 137, Indian Council for Research on Internation- al Economic Relations, New Delhi. Goldar, Bishwanath, and Anita Kumari. 2003. "Import Liberalization and Productivity Growth in Indian Manufacturing Industries in the Late 1990s."Developing Economies 41 (4): 436­60. Goldman Sachs. 2007. BRICs and Beyond. New York: Goldman Sachs. Government of India. 2000. Report of the Eleventh Finance Commission for 2000­2005. New Delhi: Ministry of Finance, Government of India. Guruswamy, Mohan, Kamal Sharma, Jeevan Prakash Mohanty, and Thomas J. Korah. 2005. "FDI in India's Retail Sector: More Bad Than Good?" Centre for Policy Alternatives, New Delhi. Hausmann, Ricardo, Dani Rodrik, Charles F. Sabel. 2007. "Reconfiguring Industrial Policy: A Framework with an Application to South Africa."Working paper, Republic of South Africa National Treasury, Pretoria. Iarossi, Giuseppe, Federica Saliola, and Giovanni Tanzillo. 2007."Business Climate, Productiv- ity, and Competitiveness in Armenia: 2002­2005." Armenian Journal of Public Policy 2 (2): 153­91. IMF (International Monetary Fund). 2006. World Economic Outlook, Fall 2006. Washington, DC: IMF. Ingram, Michael,Vijaya Ramachandran, and Vyjayanti Desai. 2006."Why Do Firms Choose to Be Informal? Evidence from the Africa Investment Climate Surveys." Regional Program on Enterprise Development Working Paper 134, World Bank, Washington, DC. Joseph, Mathew, Nirupama Soundararajan, Manisha Gupta, and Sanghamitra Sahu. 2008. Impact of Organized Retailing on the Unorganized Sector. New Delhi: Indian Council for Research on International Economic Relations. Kochhar, Kalpana, Utsav Kumar, Raghuram Rajan, Arvind Subramanian, and Ioannis Tokatlidis. 2006. "India's Pattern of Development: What Happened, What Follows?" IMF Working Paper 06/22, International Monetary Fund, Washington, DC. Kohli, Renu, and Sudip Mohapatra. 2007. "What Explains India's Real Appreciation?" IMF Working Paper 07/268, International Monetary Fund, Washington, DC. Kumar, Rajiv, and Abhijit Sen Gupta. 2008. "Towards a Competitive Manufacturing Sector." ICRIER Working Paper 203, Indian Council for Research on International Economic Relations, New Delhi. Little, Ian M. D., Dipak Mazumdar, and John M. Page. 1987. Small Manufacturing Enterprises: A Comparative Analysis of India and Other Economies. New York: Oxford University Press. Malhotra, Mohini, Yanni Chen, Alberto Criscuolo, Qimiao Fan, Iva Ilieva Hamel, and Yevgeniya Savchenko. 2006."Expanding Access to Finance: Good Practices and Policies for Micro, Small and Medium Enterprises." World Bank, Washington, DC. McKinsey Global Institute. 2001. India: The Growth Imperative. New Delhi: McKinsey & Company. ------. 2003. New Horizons: Multinational Company Investment in Developing Economies. San Francisco, CA: McKinsey & Company. Ministry of Finance. 2008. Economic Survey 2007­2008. New Delhi: Government of India. NASSCOM (National Association of Software and Services Companies). 2007. "Indian IT Industry: NASSCOM Analysis." NASSCOM, New Delhi. http://www.nasscom.in/upload/ 5216/Indian%20IT%20Industry%20Factsheet%20Feb%202007.doc. 152 References NASSCOM (National Association of Software and Services Companies) and Everest Group. 2008. India BPO Study: Roadmap 2012--Capitalizing on the Expanding BOP Landscape. New Delhi: NASSCOM. NASSCOM (National Association of Software and Services Companies) and McKinsey & Company. 2005."Extending India's Leadership in the Global IT and BPO Industry."NASS- COM, New Delhi. National Commission for Enterprises in the Unorganised Sector. 2007. Report on Comprehen- sive Legislation for Minimum Conditions of Work and Social Security for Unorganised Workers. New Delhi: Government of India. National Manufacturing Competitiveness Council. 2006. The National Strategy for Manufac- turing. New Delhi: National Manufacturing Competitiveness Council. NSSO (National Sample Survey Organisation). 1984. Employment-Unemployment National Sample Survey on Employment and Unemployment, 38th Round. New Delhi: Ministry of Statistics and Programme Implementation, Government of India. ------. 1994. Employment-Unemployment National Sample Survey on Employment and Unemployment, 50th Round. New Delhi: Ministry of Statistics and Programme Implemen- tation, Government of India. ------. 2000. Employment-Unemployment National Sample Survey on Employment and Unemployment, 55th Round. New Delhi: Ministry of Statistics and Programme Implemen- tation, Government of India. ------. 2001. National Unorganised Manufacturing Survey 2000/01. New Delhi: Ministry of Statistics and Programme Implementation, Government of India. ------. 2005. Employment-Unemployment National Sample Survey on Employment and Unemployment, 61st Round. New Delhi: Ministry of Statistics and Programme Implemen- tation, Government of India. OECD (Organisation for Economic Co-operation and Development). 2007. Economic Survey of India 2007. Paris: OECD. Olley, G. Steven, and Ariel Pakes. 1996. "The Dynamics of Productivity in the Telecommuni- cations Equipment Industry." Econometrica 64 (6): 1263­97. Panagariya, Arvind. 2007. "Why India Lags Behind China and How It Can Bridge the Gap." World Economy 30 (2): 229­48. Papola, Trilok Singh. 1991. "Industry and Employment: Recent Indian Experience." ISID Foundation Day Lecture, Institute for Studies in Industrial Development, New Delhi. Pattnayak, Sanja Samirana, and Shandre M. Thangavelu. 2003. "Economic Reforms and Pro- ductivity Growth in Indian Manufacturing Industries: An Interaction of Technical Change and Scale Economies?" Working Paper 0307, National University of Singapore, Singapore. Poddar, Tushar, and Eva Yi. 2007. "India's Rising Growth Potential." Global Economics Paper 152, Goldman Sachs, New York. Poirson, Hélène. 2006."The Tax System in India: Could Reform Spur Growth?" IMF Working Paper 06/93, International Monetary Fund, Washington, DC. Reddy, Y. Venugopal. 2006. "Asian Perspectives on Growth: Outlook for India." Address at the G30's International Banking Seminar, Singapore, September 18. Reserve Bank of India. 2005. "India's Financial Sector: Recent Reforms, Future Challenges." World Bank, Washington, DC. Rodrik, Dani, and Arvind Subramanian. 2004. "From Hindu Growth to Productivity Surge: The Mystery of the Indian Growth Transition." NBER Working Paper 10376, National Bureau of Economic Research, Cambridge, MA. 153 References Tata Services Limited. 2003. "Reforms and Productivity Trend in Indian Manufacturing Sec- tor." Tata Services Limited, Mumbai, India. Topalova, Petia. 2004. "Trade Liberalization and Firm Productivity: The Case of India." IMF Working Paper 04/28, International Monetary Fund, Washington, DC. Transparency International. 2005. India Corruption Study 2005. New Delhi: Centre for Media Studies. UNCTAD (United Nations Conference on Trade and Development). 2007. Information Econ- omy Report 2007­2008: Science and Technology for Development--The New Paradigm of ICT. Geneva: UNCTAD. Unel, Bulent. 2003. "Productivity Trends in India's Manufacturing Sectors in the Last Two Decades." IMF Working Paper 03/22, International Monetary Fund, Washington, DC. Veeramani, Choorikkad, and Bishwanath Goldar. 2004. "Investment Climate and Total Factor Productivity in Manufacturing: Analysis of Indian States." ICRIER Working Paper 127, Indian Council for Research on International Economic Relations, New Delhi. Virmani, Arvind. 2004. "Sources of India's Economic Growth: Trends in Total Factor Produc- tivity." ICRIER Working Paper 131, Indian Council for Research on International Economic Relations, New Delhi. World Bank. 2004a. India: Scaling-up Access to Finance for India's Rural Poor. Report 30740-IN. Washington, DC: World Bank. ------. 2004b. "Sustaining India's Services Revolution." World Bank, Washington, DC. ------. 2006a. "Doing Business in South Asia 2006." World Bank, Washington, DC. ------. 2006b. Inclusive Growth and Service Delivery: Building on India's Success--India Development Policy Review. Washington, DC: World Bank. ------. 2007a. Doing Business 2008. Washington, DC: World Bank. ------. 2007b."India's Employment Challenge: Creating Jobs, Helping Workers."World Bank, Washington, DC. ------. 2007c. World Development Indicators 2007. Washington, DC: World Bank. ------. 2008."India Accelerating Growth and Development in the Lagging Regions of India." World Bank, Washington, DC. 154 Index Boxes, figures, notes, and tables are indicated by b, f, n, and t, respectively. A econometric analysis of. See econometric Accelerated Power Development and Reform analysis Program (APDRP), 40 electrical power. See electrical power AETRs (average effective tax rates), 81b finance, access to, 112­13 age of firms engaged in unorganized in organized manufacturing, 37, 38 manufacturing, 55, 60f in unorganized manufacturing, 61­62b, agricultural employment in India, 20­21f 61­64, 63f, 67­68b Agricultural Produce Marketing Committee inequalities between states, factors leading Act, 73 to, 107­8, 108f, 109f APDRP (Accelerated Power Development and land, access to, 112 Reform Program), 40 in retail sector, 78f, 79, 83­85b Arab Republic of Egypt, 33 software and ITES sector, 96 average effective tax rates (AETRs), 81b in organized manufacturing, 112t compared to software and ITES sector, B 89­90b Bangladesh, 50b firms' perceptions of constraints, 31­35, Brazil, 25b, 71f, 80n3 32­35f bureaucracy. See red tape, corruption, red tape, corruption, and crime. See red and crime tape, corruption, and crime in retail sector. See under retail sector C security. See security issues capital use, productivity of, 16­17 skilled labor, lack of. See skilled labor chain stores. See under retail sector shortages Chile, 125 in software and ITES sector, 89­90b, 93­96, China 94­96f imports and exports as share of GDP, 17 taxation. See taxation inequalities between states in, 25b in unorganized manufacturing, 61­62b, organized manufacturing in, 24 61­65, 63f, 64f, 67­68b, 112t electrical power issues, 47b corporate governance in organized exporting firms, 40, 42, 43, 44 manufacturing, 37 productivity in, 31 corruption. See red tape, corruption, as share of GDP, 30t and crime productivity analysis of ICA reports, 125 credit, access to, 112­13 services sector in, 11 in organized manufacturing, 37, 38 TFP in, 16 in unorganized manufacturing, 61­62b, CIBIL (Credit Information Bureau (India) 61­64, 63f, 67­68b Limited), 68b Credit Information Bureau (India) Limited Cobb-Douglas estimators, 125, 126t, 130 (CIBIL), 68b Colombia, 125 Credit Information Companies Act, 68b constraints on investment climate in India, crime. See red tape, corruption, 112t, 1111 and crime 155 Index D F debt reduction and public savings in India, family member workers in unorganized 15, 16f manufacturing, 56, 58f Doing Business study (World Bank), ICI fast-moving consumer goods (FMCG) stores. compared with, 102, 105b See under retail sector FDI. See foreign direct investment E finance, access to, 112­13 East Asia and Pacific region, 29, 30t in organized manufacturing, 37, 38 econometric analysis, 125­45 in unorganized manufacturing, 61­62b, Cobb-Douglas estimators, 125, 126t, 130 61­64, 63f, 67­68b endogeneity of investment climate variables, Fiscal Responsibility and Budget Management 125­26 Act, 15 methodology, 125­29, 126­28t FMCG (fast-moving consumer goods) stores. missing variables, problem of, 126 See under retail sector OLS, 125, 126t, 130 foreign direct investment (FDI) organized manufacturing, 35­37, 36f, econometric analysis and, 126t, 129, 127­28t, 129­31, 132­35t 135t, 144t retail sector, 137­40, 139t in Indian economy generally, 17­19 software and ITES sector, 140, 140­44t in organized manufacturing, 31, 35­37, 40 Solow residuals, 125, 126t, 130, 145n3 in retail sector, 73 Translog estimators, 125, 126t, 130 in software and ITES sector, 89, 144t unorganized manufacturing, 135­36, formal sector 137t, 138t employment growth, creating, 1, 9, 19­21, education. See skilled labor shortages 19f, 20f, 21t, 22t, 113 Egypt, Arab Republic of, 33 manufacturing in. See organized electrical power manufacturing conclusions regarding, 111 full-time versus part-time workers in in organized manufacturing, 31­33, 32­33f, unorganized manufacturing, 56, 58f 36­40, 38­39f, 47­49b in retail sector, 77, 78f G in software and ITES sector, 96 GDP. See gross domestic product states, inequalities between, 107, 108f Germany, 92t in unorganized manufacturing, global economy, integration of Indian 61­62, 64f economy into, 17 employment demand. See also labor force GNI (gross national income) per capita in definition of employment, 66n3 India, 11 in econometric analysis, 126t, 129, 132t, 136, gross domestic product (GDP) 138t, 141t agriculture and, 20­21f formal sector, creating growth in, 1, 9, composition of, 12, 13f 19­21, 19f, 20f, 21t, 22t, 113 employment growth in formal sector and, 19 in organized manufacturing, 132t growth in, 11, 12f in retail sector, 70­73, 74b imports and exports as share of, 17 shortage of skilled labor. See skilled investments and savings as share of, 14­15f labor shortages organized manufacturing's share of, 29, 30t in software and ITES sector, 88, 141t software and ITES sector contribution to, 87 in unorganized manufacturing, 60f, gross national income (GNI) per capita in 136, 138t India, 11 endogeneity of investment climate variables, 125­26 I European Union (EU), 25b, 88 ICA (Investment Climate Assessment). See exports econometric analysis in econometric analysis, 126t, 129, ICI (Investment Climate Index) measuring 134t, 143t inequalities between states. See under GDP, as share of, 17 inequalities between states organized manufacturing, 40­45, 41t, 42f, ICOR (incremental capital output ratio), 17 43f, 44t ICSs. See Investment Climate Surveys software and ITES sector, 88, 143t imports, as share of GDP, 17 156 Index incremental capital output ratio (ICOR), 17 manufacturing in. See unorganized Indian economy, 1­2, 11­27 manufacturing challenges faced by, 18­23 percentage of Indian workers in, 25, 26f constraints on. See constraints on investment information technology-enabled services climate in India (ITES). See software and ITES sector employment growth in formal sector, infrastructure importance of creating, 1, 9, 19­21, 19f, electrical power. See electrical power 20f, 21t, 22t inequalities between states, measuring, FDI in, 17­19 107­8, 108f GDP. See gross domestic product organized manufacturing, as obstacle to, global economy, integration into, 17 36­37, 39­40 GNI per capita, 11 unorganized manufacturing, as obstacle growth of, 11­12, 12f, 13f to, 65 key growth drivers, 12­19, 14­16f, 17t, 18b Infrastructure Quality Index (Government of manufacturing sector. See organized manu- India), ICI compared with, 102, 105b facturing; unorganized manufacturing institutional indicators productivity growth and, 15­17, 17t corporate governance in organized public savings and debt reduction, 15, 16f manufacturing, 37 retail. See retail sector inequalities between states, 107­8, 109f savings and investments, increases in, 14­15f red tape, corruption, and crime. See red services sector, 11 tape, corruption, and crime software and ITES. See software and ITES security issues, 107­8 sector taxation. See taxation states, inequalities between. See inequalities Investment Climate Assessment (ICA). See between states econometric analysis structural changes in, 12, 13f investment climate in India, 1­10, 111­14 Indian Factories Act, 54b constraints on. See constraints on Indian investment climate. See investment investment climate in India climate in India econometric analysis of. See econometric Indonesia, 25b, 50b analysis inequalities between states, 8­9, 101­10 Indian economy and. See Indian economy conclusions regarding, 9, 109, 114 inequalities between states. See inequalities electrical power, 107, 108f between states ICI, 8, 101­2 manufacturing. See organized manufactur- compared with Infrastructure Quality ing; unorganized manufacturing Index and Doing Business study, obstacles to, 9t 102, 105b retail. See retail sector formulas used to construct, 147, 149b software and ITES. See software and methodology, 102, 147­49, 148t, 149b ITES sector ranking of states by, 102­7, 106f Investment Climate Index (ICI) measuring reliability testing, 102, 103­4b, 147 inequalities between states. See under variables used in, 147, 148t inequalities between states importance of reducing, 2, 21­23, 23f, 24f Investment Climate Surveys (ICSs), 115­23 infrastructure, 107­8, 108f organized manufacturing, 29, 115­17, institutional indicators, 107­8, 109f 116t, 117t international comparison of income gaps retail sector, 69­70, 119­21, 121t between rich and poor states, 25b software and ITES sector, 87, 121­23, 122t investment climate at state level, importance unorganized manufacturing, 55, 117­19, of measuring, 101­2 118t, 120t key differences between high and low rank- ITES (information technology-enabled ing states, 107­8, 108f, 109f services). See software and ITES sector organized manufacturing, 37­40, 38f, 39f red tape, corruption, and crime, 107 J security issues, 107­8 Japan, 4, 11, 54 taxation, 107 informal sector K employment growth in, 19­20, 19f, 21t Korea, Republic of, 4, 54, 67b 157 Index L infrastructure, 36­37, 39­40 labor force. See also employment demand labor force regulation, 35, 38­39f regulation of, organized manufacturing labor force skills, 31­32, 32­33f, 37 affected by, 35, 38­39f as potential growth driver, 29­31 skilled labor. See skilled labor shortages productivity issues, 30­31, 35­37, 43­44, wages 44t, 127­28t in econometric analysis, 126t, 129, reasons for focus on, 2­4, 24 133t, 142t red tape, corruption, and crime, 31­34, in software and ITES sector, 93, 95f, 142t 32­34f, 36, 38­39f, 50­51b land, access to, 112 security issues, 36 in retail sector, 78f, 79, 83­85b unorganized firms' avoidance of organized software and ITES sector, 96 status, 56­61, 59f outsourcing of software and ITES sector jobs M to India from U.S., 93 Malaysia, 67b, 125 mandis (market centers), 73 P manufacturing. See organized manufacturing; Pakistan, 33, 50b unorganized manufacturing part-time workers marginal effective tax wedge (METW), 81b defined, 66n2 Mexico, 30t, 40, 125 in unorganized manufacturing, 56, 58f modern-format stores. See under retail sector pay municipal authorities, unorganized manufac- in econometric analysis, 126t, 129, 133t, 142t turers registered with, 57­59, 59f in software and ITES sector, 93, 95f, 142t Philippines, 125 N power, as investment constraint. See Nepal, 50b electrical power productivity O of capital use, 16­17 obstacles to investment climate in India. See conclusions regarding, 10, 114 constraints on investment climate differentials between large and small firms, in India 54­55 octroi, 77, 80n5 econometric analysis of. See econometric OECD (Organisation for Economic analysis Co-operation and Development) growth of Indian economy attributable to, countries, 81­82b 15­17, 17t ordinary least squares (OLS), 125, 126t, 130 importance of accelerating, 2, 23, 114 Organisation for Economic Co-operation and inequalities between states and, 23, 24f Development (OECD) countries, organized manufacturing sector and, 30­31, 81­82b 35­37, 43­44, 44t, 127­28t organized manufacturing, 29­46 potential for accelerating, 23, 26f changes in products, 41­42, 41t reallocation of labor force and, 20­21, 22t conclusions regarding, 45­46 in retail sector, 70­73, 71f, 72f, 74b, 139t constraints on, 112t services sectors, across, 91f compared to software and ITES sector, in software and ITES sector. See under 89­90b software and ITES sector firms' perceptions of, 31­35, 32­35f TFP. See total factor productivity corporate governance, 37 unorganized manufacturing sector and, econometric analysis of, 35­37, 36f, 127­28t, 54­55, 55f, 60, 61­65, 136, 137t 129­31, 132­35t public savings and debt reduction in India, electrical power, 31­33, 32­33f, 36­40, 15, 16f 38­39f, 47­49b exporting firms, 40­45, 41t, 42f, 43f, 44t R FDI, 31, 35­37, 40 red tape, corruption, and crime finance, access to, 37, 38 inequalities between states, 107 GDP share, 29, 30t organized manufacturing, as obstacle to, ICS for, 29, 115­17, 116t, 117t 31­34, 32­34f, 36, 38­39f, 50­51b inequalities across states, 37­40, 38f, 39f retail sector, as obstacle to, 77, 78f 158 Index in software and ITES sector, 96f security issues unorganized manufacturing, as obstacle to, 65 inequalities between states, 107­8 registration of unorganized manufacturers with organized manufacturing, 36 tax or municipal authorities, 57­59, 59f software and ITES sector, 97n1, 97n3 Republic of Korea, 4, 54, 67b services sector in India, growth of, 11 retail sector, 69­80 size of firms conclusions regarding, 79 retail sector, 69­70. See also under constraints on growth and investment retail sector identification of, 77­79, 78f unorganized manufacturing, 55 labor force skills, 80n4 skilled labor shortages large chain stores more impacted by, conclusions regarding, 113 74­76, 75f, 76f organized manufacturing, as obstacle to, econometric analysis of, 137­40, 139t 31­32, 32­33f, 37 electrical power as obstacle to, 77, 78f in retail stores, 80n4 employment growth in, 70­73, 74b software and ITES sector, shortage of skilled FDI in, 73 workers in, 94­96, 95f, 98­99b government reform efforts, 73­74 unorganized manufacturing, 56, 57f, 65 growth of, 70, 71f software and ITES sector, 87­97 ICS for, 69­70, 119­21, 121t conclusions regarding, 96­97 land, access to, 78f, 79, 83­85b constraints on, 89­90b, 93­96, 94­96f large chain stores (modern-format stores), econometric analysis of, 140, 140­44t 69­70 electrical power in, 96 cost structure for, 76f employment growth in, 88 effect of investment climate on, 74­76, export market, 88, 143t 75f, 76f favorable conditions obstacles identified by, 77­79, 78f leading to success in, 89­91 productivity and employment growth, 73 productivity gains to offset erosion of, productivity issues in, 70­73, 71f, 72f, 93­96 74b, 139t FDI, 88, 144t reasons for focus on, 5­6, 27 growth of, 87­89, 88f red tape, corruption, and crime as obstacle ICS for, 87, 121­23, 122t to, 77, 78f labor force small individual stores (traditional retail or rising wages for, 93, 95f, 142t FMCG stores), 69­70 shortage of skilled workers, 94­96, 95f, cost structure for, 76f 98­99b effect of investment climate on, 74­76, land, access to, 96 75f, 76f outsourcing to India from U.S., 93 land, access to, 83b productivity productivity and employment growth, 70, econometric analysis of, 140t 73, 74b gains to offset erosion of favorable types of stores included in, 79n1 conditions, 93­96 taxation potential for growth in, 81f, large chains, taxation as obstacle to, 91­93, 92f 77­79, 78f, 81­82b reasons for focus on, 6­8, 27 VAT, 73­74, 77­79 red tape, corruption, and crime in, 96f Russian Federation, 11, 92t security issues, 97n1 TFP for, 93 S Solow residuals, 125, 126t, 130, 145n3 salaries South Asia region, 30t in econometric analysis, 126t, 129, 133t, 142t South Korea (Republic of Korea), 4, 54, 67b in software and ITES sector, 93, 95f, 142t SRFAESI (Securitization and Reconstruction of savings, increases in, 14­15f Financial Assets and Enforcement of SEBs (State Electricity Boards), unbundling Security Interest Act 2002), 67­68b of, 40 State Electricity Boards (SEBs), unbundling Securitization and Reconstruction of Financial of, 40 Assets and Enforcement of Security states, inequalities between. See inequalities Interest Act 2002 (SRFAESI), 67­68b between states 159 Index T unorganized manufacturing, 53­66 taxation age of firms engaged in, 55, 60f AETRs, 81b characteristics of firms, owners, and workers inequalities between states, 107 involved in, 55­56, 57f, 58f octroi, 77, 80n5 constraints on, 61­62b, 61­65, 63f, 64f, organized manufacturing, as obstacle to, 31, 67­68b, 112t 32­33f, 33, 38­39f defined, 54b retail stores econometric analysis of, 135­36, 137t, 138t large chains, taxation as obstacle to, electrical power, 61­62, 64f 77­79, 78f, 81­82b employment demand in, 60f, 136, 138t VAT, 73­74, 77­79 finance, access to, 61­62b, 61­64, 63f, unorganized manufacturers registered with 67­68b tax authorities, 57­59, 59f ICS for, 55, 117­19, 118t, 120t telecommunications. See software and importance to India's economy, 53­54 ITES sector infrastructure obstacles, 65 TFP. See total factor productivity labor force skills, 56, 57f, 65 Thailand, 24, 30t, 125 organized sector, avoidance of, 56­61, 59f total factor productivity (TFP) productivity issues, 54­55, 55f, 60, 61­65, in econometric analysis, 125, 126t 136, 137t Indian economy and, 16, 17t, 18b reasons for focus on, 4­5, 25­27, 26f in organized manufacturing, 30­31, 35­36 red tape, corruption, and crime as obstacle in software and ITES sector, 93 to, 65 traditional retail stores. See under retail sector registration with tax or municipal Translog estimators, 125, 126t, 130 authorities, 57­59, 59f transportation. See infrastructure unorganized sector. See informal sector Turkey, 2, 31, 40, 46n2 V U value-added tax (VAT) and retail stores, 73­74, United Arab Emirates, 82b 77­79 United Arab Republic, 33 Vietnam, 50b United Kingdom, 88 United States W inequalities between states in, 25b wages organized manufacturing productivity in, 30 in econometric analysis, 126t, 129, purchasing power of India compared to, 11 133t, 142t retail sector in, 70, 71f in software and ITES sector, 93, 95f, 142t software and ITES sector women in unorganized manufacturing, 55, exports to U.S. from India, 88 56, 58f outsourcing, 93 workforce. See employment demand; productivity levels compared, 91, 92f labor force 160 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. The · 10 trees Office of the Publisher has chosen to print India's · 7 million BTU of Investment Climate on recycled paper with 30 per- total energy cent postconsumer fiber in accordance with the · 866 lb. of net green- recommended standards for paper usage set by the house gases Green Press Initiative, a nonprofit program sup- · 3,596 gal. of waste porting publishers in using fiber that is not water sourced from endangered forests. For more infor- · 462 lb. of solid waste mation, visit www.greenpressinitiative.org. ndia has emerged in the past few years as one of the world's fastest growing economies. I While a high level of investment and domestic private consumption continue to drive growth, it is necessary to address the bottlenecks that prevent sustainable economic growth in the future. India's Investment Climate: Voices of Indian Business identifies key investment climate bottlenecks that slow down growth and poverty reduction. Investment climate refers to factors that influence day-to-day decisions by firms on how to invest. It includes macroeconomic policies, governance, institutions, and infrastructure. In particular this book aims to answer the following three questions insofar as they relate to the investment climate: How can productivity be improved? How can employment generation be increased? How can inequality among states be reduced? India's Investment Climate highlights the significance of the challenges facing India and demonstrates how investment climate improvements in key sectors will assist the Indian economy in overcoming these three challenges. It provides national and local policy-makers a number of recommendations based upon well-researched analysis backed by sound economic theory. ISBN 978-0-8213-7757-4 SKU 17757