46435 v2 Report No. 46435-PK PAKISTAN’S INVESTMENT CLIMATE LAYING THE FOUNDATION FOR GROWTH (In Three Volumes) Volume II: Annexes December 2009 Poverty Reduction, Economic Management Department Finance and Private Sector Unit South Asia Department Document of the World Bank TABLE OF CONTENTS Annex A: References ..................................................................................................................... 1 Annex B: Private Sector Background .......................................................................................... 11 Annex C: Weighted Investment Climate Variables ...................................................................... 23 Annex D: Sampling Methodology ................................................................................................ 29 Annex E: Econometric Methods for Productivity and Investment Climate Analysis .................. 35 Annex F: Definitions of Variables and Econometric Results ....................................................... 43 ANNEX A: REFERENCES __________ 2004a. “Doing Business in 2005: Removing Obstacles to Growth”. Washington, D.C. World Bank. __________ 2004b. “2003 Annual Review of Development Effectiveness: The Effectiveness of Bank Support for Policy Reform. Report 28290”. Washington, D.C.: World Bank Operations Evaluation Department. __________ 2005. “World Development Report 2005: A Better Investment Climate for Everyone”. World Bank and Oxford University Press. Washington, D.C. Addison, J. and P. Teixeira (2003). "The economics of employment protection." Journal of Labor Research 24(1): 85-129. Ahn, S. (2001), "Firm Dynamics and Productivity Growth: A Review of Micro Evidence from OECD Countries", OECD Economics Department Working Papers, No. 297, OECD Publishing. Alexander, W. Robert J., Bell, John D., and Knowles, Stephen (2004). “Quantifying Compliance Costs of Small Businesses in New Zealand”. Discussion paper, University of Otahgo. Alfaro, L., A. Charlton and F. Kanczuk (2007), “Firm-size Distribution and Cross-Country Income Differences”. Unpublished manuscript. Andres. L., J.L.Guasch and S.Straub, 2007, “Do Regulation and Institutional Design Matter for Infrastructure Performance?”, Policy Research Working Paper No. 4378, The World Bank, Washington DC. Asian Development Bank 2005, Electricity Sectors in CAREC Countries: A Diagnostic Review of Regulatory Approaches and Challenges, Private Public Infrastructure Advisory Facility. Aslam, M. (2009). "Education Gender Gaps in Pakistan: Is the Labour Market to Blame?" Economic Development and Cultural Change, Forthcoming. Baily, M. N., C. Hulten, D. Campbell, (1992), "Productivity Dynamics in Manufacturing Establishments," Brookings Papers on Economic Activity, 187-249. Bari, F,. Cheema, A,. and Haque, E. “SME Development in Pakistan: Analyzing the constraints to Growth”. Asian Development Bank. Working Paper 8 Bari, F. A. Cheema and E. Haq, “SME Development in Pakistan: Constraints on Growth”, Pakistan Resident Mission, Asian Development Bank. Bari, Faisal, Ali Cheema and Ehsan ul Haq (2005), “SME Development in Pakistan: Analyzing the Constraints to Growth,” Working Paper No. 3, Asian Development Bank. Bartel, A. (1994). "Productivity gains from the implementation of employee training programs." Industrial relations: a journal of economy and society 33(4): 411-425. 1 Bartelsman E. J. and M. Doms (2000). “Understanding Productivity: Lessons from Longitudinal Microdata”. Journal of Economic Literature, Vol. 38, Nº 3, 569-594. Bartelsman, E, J. Haltiwanger and S. Scarpetta, (2006) “Cross-Country Differences in Productivity: The Role of Allocative Efficiency,” Unpublished manuscript. Bassanini, A and Stefano Scarpetta. 2002. Oxf Rev Econ Policy; 18:324-344. Besant-Jones, J. E., 2006, “Reforming Power Markets in Developing Countries: What have we learned?”, Energy and Mining Sector Board Discussion Paper No. 19, The Energy and Mining Sector Board, The World Bank, Washington DC. Besley, T. and R. Burgess (2004). "Can labor regulation hinder economic performance?" Quarterly Journal of Economics 119 (1): 91-134. Blunch, N-H and D. Verner, 1996. “Sector Growth and the Dual Economy Model: Evidence from Côte d’Ivoire, Ghana, and Zimbabwe”. Policy research Working Paper Series, The World Bank. Board of Investment.Government of Pakistan. 2006. IT & Telecom Sector Profile. Boeke, J H, 1953. Economics and Economic Policy of Dual Societies as Exemplified by Indonesia (New York, 1953) Bosworth, Barry and Susan Collins (2003). “The Empirics of Growth: An Update”. The Brookings Institution. Washington, D.C. Processed. Branstetter, L. Goldberg, I. and Kuriakose, S. 2008. “The Links amongst Knowledge Absorption, Trade and FDI” in Globalization and Technology Absorption in Europe and Central Asia, Working Paper 150, The World Bank, Washington D.C. Briscoe, J. and U. Qamar 2005, Pakistan’s Water Economy: Running Dry, Oxford University Press and The World Bank, Washington DC. Calderon, C. and L. Serven, 2008, “Infrastructure and Economic Development in Sub-Saharan Africa”,Policy Research Working Paper No. 4712, The World Bank, Washington DC. Cameron, A.C. and P.K Trivedi, 2005. “Microeconometrics: Theory and Applications”. Cambridge University Press. Caselli, F., (2005) “Accounting for Income Differences Across Countries,” chapter 9 in the Handbook of Economic Growth Vol. 1A, P. Aghion and S. Durlauf, eds., North Holland. Central Board of Revenue (2003), “Pakistan Tax Administration Reform; Comprehensive Medium and Long Term Strategy,” Government of Pakistan. Islamabad. Central Board of Revenue (2005a), “Year Book, 2004-2005,” Ministry Of Finance, Government of Pakistan, Islamabad. Central Board of Revenue (2005b), “Income Tax Ordinance,” Government of Pakistan, Islamabad. 2 Central Board of Revenue (2006a), “Taxation Structure of Pakistan,” Government of Pakistan, Islamabad. Central Board of Revenue (2006b), “Tax Policy Study-First Phase,” Government of Pakistan, Islamabad. Chaudhry, M. 2004. Strategic Partnership in Promoting Technology Incubation System for SMEs. Chaudry, S. 2007. Increasing Global Competitiveness: A Case for the Pakistan Economy. Lahore Journal of Economics. Papers presented at the Third Annual Conference on Management of the Pakistan Economy. Cohen, W., and D. Levinthal 1990. “A New Perspective on Learning and Innovation”. In special issue on “Technology, Organizations and Innovations,” Administrative Science Quarterly 35(1): 128-52. Cole, H. L., L. E. Ohanian, A. Riascos and J. A. Schmitz Jr. (2004). "Latin America in the Rearview Mirror". National Bureau of Economic Research WP #11008, December. Djankov, Simeon, La Porta, Rafael, Lopez-de-Silanis, Florencio, and Shleifer, Andrei (2002). “The Regulation of Entry”. Quarterly Journal of Economics 117, February 2002. 1-37. Dollar, D., M. Hallward-Driemeier and T. Mengistae, 2004, “Investment Climate and International Integration” World Bank Policy Research Working Paper 3323, The World Bank, Washington DC Dollar, David, Anqing Shi, Shuilin Wang and L. Colin. Xu (2004). “Improving City Competitiveness through the Investment Climate: Ranking 23 Chinese Cities”. Washington, D.C., World Bank. Eamets, R. and J. Masso (2004). "Labour market flexibility and employment protection regulation in the Baltic states." IZA Discussion Paper No. 1147. Easterly, W. (2000) “The Lost Decades: Explaining Developing Countries Stagnation 1980-98.” Mimeo. World Bank: Washington D.C. Easterly, W. and L. Serven (2003) The Limits of Stabilization: Infrastructure, Public Deficits and Growth in Latin America. Stanford University Press and TheWorld Bank. Edquist, C and McKelvey, M. 2000. Systems of Innovation: Growth, Competitiveness and Employment. Escribano A. and J. L. Guasch (2008). “Robust Methodology for Investment Climate Assessment on Productivity: Application to Investment Climate Surveys from Central America”. Working Paper # 08-19 (11), Universidad Carlos III de Madrid. 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, The World Bank, June. 3 Escribano, A. and J. Pena 2008. “Empirical Econometric Evaluation of Alternative Methods of Dealing with Missing Values in Investment Climate Surveys”. Mimeo. Escribano, A., J. L. Guasch and J. Pena. 2008. “Assessing the Impact of Infrastructure Quality on Firm Productivity in Africa” Mimeo. Escribano, A., J. L. Guasch, M. de Orte and J. Pena. 2008a. “Investment Climate and Firm’s Performance: Econometric and Applications to Turkey’s Investment Climate Survey.” Working Paper # 08-20 (12), Universidad Carlos III de Madrid. Escribano, A., J. L. Guasch, M. de Orte and J. Pena. 2008b. “Investment Climate Assessment Based on Demean Olley and Pakes Decompositions: Methodology and Applications to Turkey’s Investment Climate Survey.” Working Paper # 08-20 (12), Universidad Carlos III de Madrid. Escribano, A., M. de Orte and J. Pena (2008). “Econometric Methods applied in the Identification and Assessment of the Investment Climate Effects on Productivity, Allocative Efficiency and other Economic Performance Measures of the Manufacturing and Services Sectors of Pakistan”. Background paper prepared for the Pakistan Investment Climate Assessment, The World Bank, Washington, DC. Escribano, A., M. de Orte and J. Pena (2008). “Econometric Methods for Investment Climate Assessment on Economic Performance in Pakistan”. Background paper prepared for the Investment Climate Assessment of Pakistan, The World Bank, Washington DC, 2008. Escribano, A., M. de Orte, J. Pena and J.L. Guasch (2008). “CROATIA: Investment Climate Assessment (ICA) on Productivity, Allocative Efficiency and other Economic Performance Measures of the Manufacturing Sector: Analysis Based on Firm Level Data for FY2006”. Background paper prepared for the Croatia Competitiveness Report, The World Bank, Washington, DC. ESMAP (Energy Sector Management Assistance Program) 2008, “Potential and Prospects for Regional Energy Trade in the South Asia Region” Formal Report 334/08, Essakali, M. D., 2005, “Rural Access and Mobility in Pakistan: A Policy Note”, Transport Note No. TRN-28, World Bank, Washington DC. Estache, A., M.A. Rossi, and C.A.Ruzzier, 2002. “The Cae for International Coordination of Electricity Regulation: Evidence from the Measurement of Efficiency in South America”, World Bank Policy Research Working Paper 2907, The World Bank, Washington DC. Estaches, A. and M. Fay, 2007, “Current Debates on Infrastructure Policy”, Policy Research Working Paper No. 4410, The World Bank, Washington DC. Foster L., J. Haltiwanger and C.J. Krizan (1998). “Aggregate Productivity Growth: Lessons from Microeconomic Evidence”. NBER Working Paper W6803. Foster, V. 2008, “Overhauling the Engine of Growth: Infrastructure in Africa”, Africa Infrastructure Country Diagnostic Report, The World Bank, The African Development Bank, the African 4 Union,the Agence Française de Développement, the European Union, the New Economic Partnership for Africa’s Development, the Public-Private Infrastructure Advisory Facility, and theU.K. Department for International Development. Fraser, J. M., 2005, “Lessons from the Independent Private Power Experience in Pakistan”, Energy and Mining Sector Board Discussion Paper No. 14, The Energy and Mining Sector Board, The World Bank, Washington DC. Frenz, M. and Oughton, C. 2005 ‘Innovation in the UK Regions and Devolved Administrations: A Review of the Literature.’ General Accounting Office (1996), “Tax Administration: Alternative Filing Systems,” GAO/GGD- 97-6. Washington, D.C.: U.S. General Accounting Office. GNSED (Global Network on Energy for Sustainable Development) 2006, “Consolidated Report: Regional Workshops on Electricity and Development in Africa, Asia and Latin America”, UNEP et al. Government of Pakistan 2002, Policy for Power Generation Projections, Ministry of Water and Power, Islamabad. Government of Pakistan 2003, “Accelerating Economic Growth and Reducing Poverty: The Road Ahead”, Poverty Reduction Strategy Paper, Ministry of Finance. Government of Pakistan 2005, Power Sector Profile, Board of Investment, Islamabad. Government of Pakistan, 2007, Pakistan Highway Rehabilitation Project, National Highway Authority. Government of Pakistan, Planning Commission. 2007. Vision 2030. Government of Pakistan, Trade Policy Speech. 2008-09. Griliches Z. and J. Mairesse (1997). “Production Functions: The Search for Identification”. In S. Strom (ed.) Essays in Honor of Ragnar Frisch, Econometric Society Monograph Series, Cambridge University Press, Cambidge. Groot, W. (1997). "Productivity effects of enterprise-related training." Applied Economics Letters 6: 369-371. Hall, R. E. and C. I. Jones (1999) “Why Do Some Countries Produce So Much More Output Per Worker Than Others?” Quarterly Journal of Economics 114: 83-116. Haltiwanger, John (2002). “Understanding Economic Growth”. The Need for Micro Evidence”. New Zealand Economic Papers 36 (1), 33-58. He, Kathy S., Morck, Randall, and Yeung, Bernard (2003). “Corporate Stability and Economic Growth”. William Davidson Working Paper No. 553. 5 Hou, X. (2008). "Challenges in youth employment in Pakistan - are those youth specific?" Background paper for Poverty Assessment in Pakistan. Howitt, P. (2000) “Endogenous Growth and Cross-Country Income Differences,” American Economic Review 90: 829-846. http://www.business.otago.ac.nz/econ/research/discussionpapers/DPO406.pdf. Ilahi, N., Z. Naqvi, et al. (1995). "Policy Induced Distortions in the Demand for Unskilled Agricultural and Industrial Labour in Pakistan. Paper prepared for the World Bank." IT & Telecommunications Division, Ministry of Science & Technology. 2000. National IT Policy and Action Plan. Jamasb, T., R. Mota, D. Newbery and M. Pollitt, 2005. “Electricity Sector Reform in Developing Countries: A Survey of Empirical Evidence on Determinants and Performance”, World Bank Policy Research Working Paper 3549, The World Bank, Washington DC. Kemal, A.R. 2007. Industrial Competitiveness of Pakistan (2000-10). The Lahore Journal of Economics. Papers presented at the Third Annual Conference on Management of the Pakistan Economy. Kemal, M. Ali. (2003), “Underground Economy and Tax Evasion in Pakistan: A Critical Evaluation,” Research Report No. 184, Pakistan Institute of Development Economics. Islamabad. Kingdon, G. and M. Soderbom (2008). "Education, skills, and labor market outcomes: evidence from Pakistan." Education working paper series number 11. Klenow, P. J. and A. Rodríguez-Clare, (1997) “The Neoclassical Revival in Growth Economics: Has It Gone Too Far?” in B. Bernanke and J. Rotemberg, eds., NBER Macroeconomics Annual (MIT Press, Cambridge) 73-103. KPMG (2008), “KPMG’s Corporate and Indirect Tax Rate Survey 2008,” KPMG International. Krishnaswamy, V. and G. Stuggins, 2007, “Closing the Electricity Supply-Demand Gap”, Energy and Mining Sector Board Discussion Paper No. 20, The Energy and Mining Sector Board, The World Bank, Washington DC. Krueger, A., M. Schiff, and A. Valdes. 1992. The Political Economy of Agricultural Pricing Policy: A Synthesis of the Economics of Developing Countries. Baltimore, Md: Johns Hopkins University Press.Gopinath, Roe and Shane, 1996 Lewis, W. Arthur (1954). “Economic Development with Unlimited Supplies of Labour,” Manchester School, 22: 139-191.Hirschman, 1958 Loayza, N.V., A. M. Oviedo and Luis Serven (2004). “Regulation and Macroeconomic Performance”, World Bank, Washington DC. 6 Magnus J. R. and A. L. Vasnev (2007). “Local Sensitivity and Diagnostic Tests”. Econometrics Journal, V. 10, 166-192. Marschak J. and W. H. Andrews (1944). “Random Simultaneous Equations and the Theory of Production”. Econometrica, 12 (3,4), 143-205. Martin, W., and D. Mitra. 1998. “Productivity Growth and Convergence in Agriculture and Manufacturing,” Development Research Group, World Bank, and Dept. of Economics, Florida International University, Washington, D.C., and Miami, Fla. Martinez-Vazquez, Jorge (2006), “Pakistan: A Preliminary Assessment of the Federal Tax System,” Working Paper. Martinez-Vazquez, Jorge and Robert McNab (2000), “The Tax Reform Experiment in Transitional Countries.” National Tax Journal 53 (2): 273-298. Mauro, Paolo (1995), “Corruption and Growth,” Quarterly Journal of Economics, 110(3), pp. 681- 712. Mc Millan, John (2004), “A Flexible Economy? Entrepeneurship and Productivity in New Zeland”, Working Paper, Graduate Scholl of Business, Stanford University, Stanford, CA. McMillan, John, (1998), “Managing Economic Change: Lessons from New Zeland”, The World Economy 21, 827-43. Ministry of Industries, Production & Special Initiatives, Government of Pakistan. 2007. SME Policy 2007: SME Led Economic Growth – Creating Jobs and Reducing Poverty. Moghal, S. Y., 2007, “Public Private Partnership in Industrial Infrastructure Development Projects in Pakistan”, The World Bank, Washington DC. Olley G. S.and A. Pakes (1996). “The Dynamics of Productivity in the Telecommunications Equipment Industry”. Econometrica, Vol. 64, 6, 1263-1297. Organization for Economic Cooperation and Development (2001). Businesses’ Views on Red Tape, Paris, OECD. http://www1.oecd.org/publications/e-book/4201101E.PDF. Pakistan Council for Science and Technology, Ministry of Science & Technology. 2006. Science and Technology, Medium Term Development Framework 2005-2010. Pakistan Council for Science and Technology, Ministry of Science & Technology. 2007. Pakistan Science and Technology Data Book 2006-07. Pakistan Institute of Development Economics, Higher Education Commission (Draft). Technology- Based Industrial Vision and Strategy for Pakistan’s Socio-Economic Development. Planning Commission (2007), “Pakistan in the 21st Century: Vision 2030,” Planning Commission, Government of Pakistan, Islamabad. 7 Prescot, Edward C. (1998) "Needed: A Theory of Total Factor Productivity". International Economic Review, 39, 525-552. Prescot, Edward C. (1998) "Needed: A Theory of Total Factor Productivity". International Economic Review, 39, 525-552. Rodrik, Dani (2004), “Industrial Policy for the Twenty-First Century,” Paper prepared for UNIDO and available online at http://ksghome.harvard.edu./~drodrik/UNIDOSep.pdf Rodrik, Dani, and Arvind Subramanian (2004). “From ‘Hindu Growth to Productivity Surge: The Mystery of the Indian Growth Transition”. Harvard University, Cambridge, Mass. Processed. Sargana, M. A., 2006, “What made Telecom a success story: Lessons learnt to be replicated in other sectors of the economy”, Pakistan Telecommunications Authority(PTA), August 29. Schafer, J, 1997. “Analysis of Incomplete Multivariate Data”. London: Chapman and Hall. Schneider, Friedrich. (2002), “The Value Added of Underground Activities: Size and Measurement of the Shadow Economies of 110 Countries over the World,” Working Paper. Seethepalli, K., M.C.Bramati, and D.Veredas, 2008, “How Relevant is Infrastructure to Growth in East Asia?”,Policy Research Working Paper No. 4597, The World Bank, Washington DC. Shahab Khawaja. 2006. Unleashing the potential of the SME sector with a focus on Productivity Improvements. Shahab, K. and Moghal, S. 2006 The Role of Business Development Agencies in Developing Economies for Improving the Business Environment for SMEs” SMEDA. 2003. Pakistan: Land of Opportunity- An Investor’s Guide. South Asia Regional Cooperation Program, The World Bank, Washington DC. Straub. S., 2008, “Infrastructure and Growth in Developing Countires: Recent Advances and Research Challenges”,Policy Research Working Paper No. 4460, The World Bank, Washington DC. Svensson, Jakob and Raymond Fisman (2007), “Are Corruption and Taxation Really Harmful to Growth? Firm Level Evidence,” Journal of Development Economics, 83(1), pp. 63-75. Swiss Agency for Development and Cooperation 2003, “Pakistan: National Round Table on Water and Sanitation in Transition – Proceedings”, Peshawar Feb 3-4, Water and Sanitation Program. The World Bank (2006). Pakistan labor market study: regulation, job creation and skill formation in teh manufacturing sector. Washington DC, the World Bank Group. The World Bank Group (2005). Pakistan Country Gender Assessment 2005: Bridging the Gender Gap Opportunities and Challenges. Washington DC, the World Bank Group. 8 UNCTAD. 2004. The Impact of FDI on Development: Globalization of R&D by Transnational Corporations and Implications for Developing Countries. UNCTAD. 2006. Globalization of R&D and Developing Countries. USAID. 2008. Annual Progress Report. Competitiveness Support Fund. USAID. 2008. State of Pakistan’s Competitiveness Report 2008. Competitiveness Support Fund. Vollrath 2008, How important are dual economy effects for aggregate productivity? Wilkinson, Bryce (2001). “Constraining Government Regulation”. NZ Business Roundtable. http://www.nzbr.org.nz/documents/publications/publications-2001/constraining_govt.pdf. World Bank (2002), “Pakistan Development Policy Review: A New Dawn?” World Bank, Washington, DC. World Bank (2002), “Private Sector Strategy for Pakistan,” World Bank, Washington, DC. World Bank (2009), “Comparing Regulation in 181 Countries,” Doing Business, World Bank, Washington, DC. World Bank 1994, Infrastructure for Development, World Development Report. World Bank 2002, “Country Assistance Strategy for the Islamic Republic of Pakistan, Report No. 24399-PAK, South Asia Region. World Bank 2002, “Implementation Completion Report for Power Sector Development Project”, Report No.22889, South Asia Region. World Bank 2002, “Pakistan Development Policy Review: A New Dawn?”, Report No. 23916-PAK, South Asia Region. World Bank 2003, “Telecommunications Regulation and Privatization Support Project Implementation Completion Report”, Report No.27060, South Asia Region. World Bank 2003. “Doing Business in 2004” Understanding Regulation. Washington, D.C. World Bank. World Bank 2004, “Country Assistance Strategy Progress Report for the Islamic Republic of Pakistan”, Report No. 28262-PAK, South Asia Region. World Bank 2004, Pakistan: Doing Business in 2005-Regional Profile, Investment Climate Department. World Bank 2005, “Pakistan: North West Frontier Province (NWFP) Economic Report: Accelerating Growth and Improving Public Service Delivery in the NWFP – The Way Forward”, Report No. 32764-PK, South Asia Region. 9 World Bank 2005, “Pakistan: Punjab Economic Report: Towards a Medium-term Development Strategy”, Report No. 29373-PK, South Asia Region. World Bank 2005, A Better Investment Climate for Everyone, World Development Report. World Bank 2006, “Country Assistance Strategy for the Islamic Republic of Pakistan for the period FY 06-09”, Report No. 35718-PAK, South Asia Region. World Bank 2006, “Pakistan Growth and Export Competitiveness”, Report No. 35499-PK, South Asia Region World Bank 2006, “Pakistan Trade and Transport Facilitation Project Implementation Completion Report and Results”, Report No. ICR000002, South Asia Region. World Bank 2006, “Pakistan Value Chain Analysis”, FIAS, Washington DC. World Bank 2006, “Transport Competitiveness in Pakistan: Analytical Underpinnings for National Trade Corridor Improvement Program”, Report No. 36523-PAK, South Asia Region World Bank 2007, “Pakistan Infrastructure Implementation Capacity Assessment”, Report No. 41630-PK, South Asia Region. World Bank 2007, “Pakistan Rural Telecom and E-Services Project Appraisal Document”, Report No. 40715-PK, South Asia Region. World Bank 2008, “Pakistan Electricity Distribution and Transmission Improvement Program – Project Appraisal Document, Report No. 41328-PK, South Asian Region. World Bank 2008, “Pakistan Water Sector Capacity Building and Advisory Services Project Appraisal Document”, Report No. 43784-PK, South Asia Region. World Bank 2008, “Trade and Transport Facilitation in South Asia: Systems in Transition”, Report No.44061 -SAS, South Asia Region. World Bank, 2006 “Pakistan: Growth and Export Competitiveness”, Washington D.C. World Bank, 2007, Doing Business in South Asia 2007, Doing Business Project. World Bank, 2007, Pakistan: Promoting Rural Growth and Poverty Reduction, Report No. 39303- PK, Sustainable and Development Unit, South Asia Region. World Bank, 2008, Doing Business 2008: Pakistan, Doing Business Project. 10 ANNEX B: PRIVATE SECTOR BACKGROUND PAKISTAN PRIVATE SECTOR STRUCTURE: KEY CHARACTERISTICS TABLE OF CONTENTS A. Structure of Private Sector ....................................................................................................... 13 C. Ownership structure ................................................................................................................. 14 D. Firm Size by Employment ....................................................................................................... 14 E. Location .................................................................................................................................... 15 F. Sectoral Classification .............................................................................................................. 15 G. Investment and Sales ................................................................................................................ 16 H. Manufacturing Sub Sector Analysis ........................................................................................ 17 I. Services Sector .......................................................................................................................... 18 J. Services Sector Contribution to Real GDP Growth .................................................................. 20 K. Services Sub-sectors ................................................................................................................ 20 11 ANNEX B: PRIVATE SECTOR BACKGROUND PAKISTAN PRIVATE SECTOR STRUCTURE: KEY CHARACTERISTICS A. Structure of Private Sector 1. As in many other developing nations, Pakistan has an enormous informal private sector estimated at over 62% of GDP.1 It is estimated that of the 3.249 million businesses operating in Pakistan, 99% are categorized as small with a low capital and revenue base. Around 5% economic entities belong to the Government and the rest operate in the private sector. There are only 15,840 firms with more than 50 employees. The 2005 Pakistan Economic Census indicates that 84 % of enterprises reported sales below 0.5 million rupees (US$ 6,250) and 93% reported sales below 1 million rupees (US$ 12,500). 2. There are substantial differences between the way private sector operates in Pakistan from elsewhere in the world. The level of business constraints varies with their size, ownership structure and location. Over 90% firms are operating without having been registered and most also do not pay taxes.2 These are not inspected by the tax inspectors, police authorities, or other government institutions. Their owners advertise, place signboards on their businesses and hire employees in addition to opening bank accounts and seeking financial services.3 The district government is responsive to their requests and training is also available to them.4 3. Most businesses in private sector show little or no growth and are also reluctant to try the new technology.5 A recent World Bank survey reported that only 3% of Pakistan’s enterprises (employing 19 or less) have innovated their product or services in the past three years and there is only 2% increase in the number of exporting firms in the past five years. 4. The structure of private sector in Pakistan is complex due to the high degree of sectoral informality. However, below is the brief snapshot of private sector characteristics and major constraints faced by this sector. 1 It is estimated that the underground economy ranged between Rs 2.91 trillion and Rs 3.34 trillion (54.6 percent of GDP to 62.8 percent of GDP respectively) in 2005 and tax evasion ranged between Rs 302 billion and Rs 347 billion (5.7 percent of GDP to 6.5 percent of GDP respectively) in 2005. Source: PIDE Working Papers: 2007:13, A Fresh Assessment of the Underground Economy and Tax Evasion in Pakistan: Causes, Consequences, and Linkages with the Formal Economy, M. Ali Kemal, Pakistan Institute of Development Economics, Islamabad, 2007. 2 There are estimated 1.9 million active tax filers in Pakistan, the number of salaried taxpayers was 440,000 and those filed under old self-assessment scheme were 275,000; source: interviews with World Bank personnel in- charge of tax reforms program. 3 Assuming that they have the assets to secure and a reasonably good set of books. However, with the informal sector, it is rare that they meet either condition. The conditions for loans for the unregistered firm are precisely the same as for registered firms. The difference usually is that registered firms maintain better books in order to be compliant with the law, and are therefore better prepared for the loan application process.. 4 From interviews with unregistered firms. 5 From 1968-2001, Korean Industrial Value-Addition increased by 40 times, Malaysia 27 times, Thailand’s by 17 times and Pakistan’s only by 7 times; Source: A Strategy For Rapid Industrial Growth, Ministry of Industries Production & Special Initiatives, 2005. 13 C. Ownership structure 5. Ownership structure plays an important role in the development and growth of firm. Ideally, more formalized structure leads to better business services delivery from the business support institutions and should result in better business conditions for firms. 6. Ownership structure of these establishments is depicted in table: 1. Almost 91.3% of total enterprises are sole proprietor ship and 1.95% are partnerships. The ownership structure is very important in determining the degree of constraints. Sole proprietary concerns can more easily escape from the regulatory radar with out losing any advantages in term of access to business or resources. However, for the government, adequate planning and provision of effective business services have become more difficult to facilitate this larger segment of the economy. Informal economy has its roots in ownership structure of the firms and any effort to develop the private sector will help improve the investment climate. D. Firm Size by Employment 7. The 2005 Census of Establishments Figure B1: Percentage of SME Establishments provides the basis for statistical analysis of the by employment size 3.249 million economic establishments operating in the private sector. Around 91.1% enterprises (2.95 million) provided complete information. Data for remaining 8.9% i.e. 291,161 economic establishments could not be obtained due to reasons such as closure, refusal, non-contact etc. Out of these, 94.46% economic establishments have an employee base of 1-5. The employment pattern of these Establishments is such that of a total of 2.95 million, 2.78 million Units (94.46%) employed 1-5 persons, 0.14 million Units (4.6%) employed 6-10 persons, 0.02 million (0.89%) 11-50 persons, and 15,840 (0.048%) over 50 persons. 8. In the case of Manufacturing Establishments, 91.0% establishments have an employee base of 1-5 persons, 6.62% employ 6-10 persons, over 2.05% have 11-50 employees, and only 0.19% have over 50 workers. 98% of wholesale, retail, trade, and hotels and restaurants employed 1-5 persons. Similarly, 95% establishments in Community, Social and Personal Services sectors are also placed in 1-5 employees group. Due to their sheer numbers, size and nature of operations, this sector of the economy promotes indigenous sources of growth, employment generation and improving standard of living in rural as well as urban areas. 9. These SMEs are more vulnerable to the external shocks and cannot adjust as quickly as the large scale firms and hence face greater challenges than the large scale enterprises. SMEs in Pakistan are dealing with a plethora of issues that directly affect their productivity such as electricity breakdowns, expensive raw material, corruption, rising tariffs of utilities, demands of laborers to increase wages and so on. Large firms are in a better position to obtain government licenses and have greater investment incentives. They also have greater access to finance and modern technology as compared with SMEs. 14 E. Location 10. The Provincial distribution of firms across the country yet again points towards the nature and economic activity prevalent in the region. While Punjab houses 65.27% of firms, Sindh is home to 17.83% economic establishments followed by NWFP with 14.21% and Balochistan 2.10%. (Figure B2). Most of the enterprises related to manufacturing, mining, agriculture, whole sale and retail, construction and transportation are located in Punjab and Sindh (Figure B3). Figure B2: Distribution of Figure B3: No. of Establishments Establishments by division and provinces by provinces 100 Agri, Forestry, Hunting& Balochistan, 90 fishing 2.1 Mining & Querrying 80 NWFP, 14.2 70 Manufacturing Punjab, 65.3 60 Electricity, Gas & Water 50 Construction 40 Sindh, 17.8 Wholesale&retail Trade, 30 Resturants & Hotels 20 Transport, Storage & Communication 10 0 Punjab Sindh NWFP Balochistan Source: Economic Census 2005. 11. Although the constraints may vary from region to region, generally certain locations in Pakistan pose extra burden on firms due to lack of infrastructure, electricity, access to industrial land, customs and trade regulations, establishment of industrial zone and provision of infrastructure. F. Sectoral Classification 12. The nature of the economic activity largely determines the character of the economy. 52% of economic activity is associated with wholesale, retail trade, hotels and restaurants. While large hotels are few in number, the small street side hotels and restaurants provide employment to a large number of work force. Figure B4: Title Source: SME Development Report, SMEDA. 15 13. The prominent manufacturing sectors are textile and garments, sports goods, leather products, cutlery, surgical instruments, food processing industry, electronics manufacturers and automobile. These industries are growing and have vast potential to absorb foreign investment. 14. The workforce employed across the 2.95 Million6 economic establishments is approx 6.5 million out of which 5.9 million are employed in the shops/establishments industry. Figure B5: No of Workers Source: SME Development Report, SMEDA. G. Investment and Sales 15. The investments made in establishments excluding land & building by various investment groups is less than PKR 1 million. The figure does not change much with the inclusion of land & building. Investment in private sector has not reached its capacity due to constraints and concerns with the business environment. The cost of compliance with the multiple types of tax filings, the lack of enforcement of contracts, and the uncertainty associated with titles has inhibited firms from fully investing themselves in the Pakistan market. Similarly in terms of Sales Revenue, the threshold of revenue generated is low (Table B1). Table B1: Number of Establishments (by major industry division & investment groups, excluding land & building, in million Rupees) Investment group 2000001 Major Industry Division Total Below 10 10 to 20 to 30 Agriculture, forestry, hunting and fishing 46,378 46,096 189 93 Mining and quarrying 709 639 33 37 Manufacturing 583,126 578,212 2,569 2,345 Electricity, gas and water 122 116 3 3 Construction 1,410 1,388 9 13 Wholesale and retail trade & restaurants and hotels 1,566,711 1,565,358 868 485 Transport, storage and communication 51,548 50,756 410 382 Financing, insurance, real-estate and business services 48,436 48,115 209 112 Community, social and personal services 659,619 658,404 785 430 Source: SME Development Report, SMEDA. 6 Establishments that provided complete information in the Census of Economic Establishments 2005, conducted by the Federal Bureau of Statistics. 16 Table B2: Number of Establishments (by major industry division and sale revenue groups) Sale revenue groups Below 0.5 0.5 million 1000001 to 20000001 to Major Industry Division Total million to 1 million 20 million 250 million Agriculture, forestry, hunting & fishing 46,170 39,746 4,113 1,720 591 Mining & quarrying 659 214 117 181 147 Manufacturing 577,715 512,585 35,897 18,143 11,090 Electricity, gas & water 119 113 1 3 2 Construction 1,386 1,197 96 66 27 Wholesale, trade & restaurants & hotels 1,547,471 1,209,229 208,257 87,261 42,724 Transport, storage & communication 51,255 46,680 2,882 1,003 690 Financing, insurance, real-estate & business services 47,757 41,150 4,023 1,444 1,140 Community, social & personal services 658,987 636,253 16,165 4,838 1,731 Source: SME Development Report, SMEDA, Economic Census 2005. H. Manufacturing Sub Sector Analysis 16. A deeper analysis of the Census of Economic Establishments provides interesting insight into the manufacturing sector. At the sub-sectoral level, out of a total of 583,329 manufacturing establishments, textile apparel and leather industries account for 252,111 economic establishments (43.22%), followed by 121,875 establishments (20.09%) of food and beverages manufacturers, and tobacco makers. Interestingly, the number of manufacturers in the fabricated metal products, and machinery and equipment sectors are 58,476 and in the basic metal industries are 4,511. In terms of numbers, only 10.80% manufacturing enterprises provide inputs to other industries for significant value addition. Given below is the sub-sectoral provincial distribution of manufacturing establishments. It is interesting to note that despite Sindh housing 17.82 percent of all economic establishments, NWFP has a greater number of economic establishments in the manufacturing sector i.e. 15.93 percent. Table B3: Sector and Provincial Distribution of Manufacturing Establishments Federal Manufacturing Industry Division Total Punjab Sindh NWFP Balochistan capital Total 583,329 399,219 80,868 92,975 7,951 2,316 Food, beverages and tobacco 121,875 70,329 14,652 35,030 1,413 451 Textile, apparel & leather 252,111 167,141 43,331 35,863 4,684 1,092 Wood & wood products including furniture 63,087 46,766 5,182 10,045 890 204 Paper & paper products, printing & publishing 10,141 6,291 2,912 761 101 76 Chemicals and chemical products 5,781 4,170 1,185 362 22 42 Non-metallic mineral products 15,476 11,755 2,093 1,516 46 66 Basic metal industries 4,511 3,456 496 504 21 34 Fabricated metal products, & machinery 58,476 46,850 6,631 4,092 599 304 Other manufacturing industries 51,871 42,461 4,386 4,802 175 47 Source: SME Development Report, SMEDA. 17. In the manufacturing sector, 20% are household establishments. At the sub-sectoral 2 digit level, the proportion between household versus other than household establishments in textile wearing apparel and leather industries is 31.50 percent, which is greater than the overall percentage of household establishments in the total manufacturing sector. As opposed to this, in 17 the manufacturing of food, beverages and tobacco only 2.45 percent establishments are household. This shows a greater concentration of small businesses involved in the textile wearing apparel and leather industries sectors. 18. Within wholesale, retail trade & restaurants and hotels, percentage of number of employees is 42.93% of the total 6.58 million, while in the manufacturing sector the percentage is 28.79. The value of raw material used in the manufacturing sector is PKR 217.53 billion, the value added is PKR 439.17 billion. In the same sector, value of energy/fuel consumed is PKR 26.45 billion. The Gross Capital Investment in the same sector is PKR 165.44 billion that is 39.31 percent of total gross capital investment in the economy. Table B4: Distribution of Establishments According to Building Status Major Economic Activity Ownership Rent free Rented Total 955,235 259,432 1,728,598 Agriculture, Forestry Hunting & Fishing 35989 5563 4643 Mining & Quarrying 207 52 442 Manufacturing 225456 90469 264261 Electricity, Gas & Water 69 33 16 Construction 291 68 1047 Wholesale & Retail Trade and Restaurants & Hotels 530529 87581 944438 Transport Storage & Communication 8559 2759 40055 Financing, Insurance, Real Estate & Business Services 9097 6832 31127 Community, Social & Personal Services 145038 66075 442569 Source: SME Development Report, SMEDA 19. 32.48 percent economic establishments operate in own premises/buildings whereas 58.50 percent establishments operate in rented premises/buildings. The split of establishments vis-à-vis status of building is such that of the total owned premises, share of Punjab is the highest at 72.83 percent followed by Sindh & NWFP with 14.3 percent and 10.84 percent respectively. 20. Analysis of data at the District level also provides interesting insights into the physiognomy of the SME sector. In the province of NWFP, Peshawar is home to 2.34% manufacturing establishments followed by Mardan (1.71%), Swabi (1.51%) and Charsadda (1.28%) respectively. A look at the manufacturing establishments spread across Punjab identifies Faisalabad (7.88%), Sialkot (6.08%) and Lahore (5.77%) as the hub of manufacturing activity. Karachi (5.69%) and Quetta (0.41%) in Sindh and Balochistan respectively have a greater concentration of manufacturing firms in each of the province. I. Services Sector 21. A look at the service sector in Pakistan reveals that the sector grew by 8.5% in 2006-07 as against 9.6% in the previous year. The finance and insurance sector remained the major driver of growth, while better performance in the wholesale and retail trade as well as the transport and communications sectors made their contributions.7 22. All components of the services sector registered strong growth in 2006-07, with the exception of ownership of dwellings, and public administration and defense. The finance and 7 Trade in Services – An ITC - SMEDA Joint publication. 18 insurance sector posted remarkable growth in 2004-05, 2005-06 and 2006-07 with growth of 30.8%, 33.0% and 18.2%, respectively. The wholesale and retail trade, and transport, storage and communications sectors also registered strong growth of 7.1% and 5.8%, respectively. 23. The growth performance of the various economic sectors over time can be seen in the following table: Table B5: Components Growth Performance of GNP (percentage growth at constant factor cost) Economic group 1980s 1990s 2002-03 2003-04 2004-05 2005-06 2006-07 Commodity Producing Sector 6.5 4.6 4.2 9.3 9.5 3.4 6.0 1. Agriculture 5.4 4.4 4.1 2.4 6.5 1.6 5.0 Major crops 3.4 3.5 6.8 1.7 17.7 -4.1 7.6 Minor crops 4.1 4.6 1.9 3.9 1.5 0.4 1.1 Livestock 5.3 6.4 2.6 2.9 2.3 7.5 4.3 Fishing 7.3 3.6 3.4 2.0 0.6 20.5 4.2 Forestry 6.4 -5.2 11.1 -3.2 -32.4 -43.7 -3.8 2. Mining & Quarrying 9.5 2.7 6.6 15.6 10.0 4.6 5.6 3. Manufacturing 8.2 4.8 6.9 14.0 15.5 10.0 8.4 Large scale 8.2 3.6 7.2 18.1 19.9 10.7 8.8 Small scale* 8.4 7.8 6.3 6.2 6.3 8.3 7.7 4.Construction 4.7 2.6 4.0 -10.7 18.6 5.7 17.2 5. Electricity & Gas Distribution 10.1 7.4 -11.7 56.8 -5.7 -23.8 -15.2 Services Sector 6.6 4.6 5.2 5.8 8.5 9.6 8.0 6. Transport 6.2 5.1 4.3 3.5 3.4 6.9 5.8 7. Wholesale & Retail Trade 7.2 3.7 6.0 8.3 12.0 8.7 7.1 8. Finance & Insurance 6.0 5.8 -1.3 9.0 30.8 33.0 18.2 9. Ownership of Dwellings 7.9 5.3 3.3 3.5 3.5 3.5 3.5 10. Public Administration & Defense 5.4 2.8 7.7 3.2 0.6 10.0 6.9 11. Services 6.5 6.5 6.2 5.4 6.6 6.3 8.5 12. GDP 6.1 4.6 4.7 7.5 9.0 6.6 7.0 13. GNP 5.5 4.0 7.5 6.4 8.7 6.4 6.9 Source: Economic Survey of Pakistan, 2006-07. 24. Value added in the wholesale and retail trade is based on the margins taken by traders on the transaction of commodities traded in the wholesale and retail market. In 2006-07, the gross value added in wholesale and retail trade increased by 7.1% over the previous year, compared to 8.6% growth in 2005-06. Value added in the transport, storage and communications sector is based primarily on the profits and losses of Pakistan Railways, Pakistan International Airlines and other airlines, Pakistan Post & Courier Services, Pak Telecom and motor vehicles of different kinds on the road. In 2006-07, this sector grew by 5.7% compared to 6.9% in 2005-06. The increase resulted primarily from strong consumer demand for mobile phones, Internet services of Pak Telecom, and motor vehicles on road. 25. Public administration and defense posted growth of 6.9% in 2006-07, while ownership of dwellings grew by 3.5% and the social services sector by 8.5%. The construction sector continued its strong showing, partly helped by activity in the private housing market, spending on physical infrastructure, and reconstruction activities in earthquake affected areas. 19 26. Telecom, energy (oil, gas and power), financial services, trade, construction, chemicals, food and personal services were the major recipients of FDI, accounting for almost 88% or US$3.7 billion. The telecom sector was the single largest recipient of FDI with US$1.4 billion, followed by the financial services (US$871 million), energy sector (US$585 million), food services (US$492 million), wholesale and retail trade (US$133.9 million), construction (US$117.1 million), personal services (US$74.1 million) and cement (US$15.2 million). J. Services Sector Contribution to Real GDP Growth 27. Pakistan’s economic growth is broad-based and is shared by all the major sectors of the economy. However, a major contribution towards growth has come from the services sector, which has emerged as a growth powerhouse over the past few years. The commodity producing sectors (agriculture and industry) have contributed two-fifths while the services sector contributed the remaining three-fifths of the 2006-07 real GDP growth of 7.0%. Agriculture and industry contributed 30.2%, or 2.9% percentage points, while the remaining 59.8%, or 4.2 percentage points, came from the services sector. Within the commodity producing sectors, agriculture contributed 1.1 percentage points or 15.1% to overall growth, while industry contributed 1.8 percentage points or 22.7% (see Table 6 for details). Table B6: Sectoral Contribution to GDP Growth (percentage points at constant factor cost) Sector 2002-03 2003-04 2004-05 2005-06 2006-07 Agriculture 1.0 0.6 1.5 0.4 1.1 Industry 1.0 3.8 3.1 1.3 1.8 Manufacturing 1.1 2.3 2.7 1.8 1.6 Services 2.7 3.1 4.4 4.9 4.2 Real GDP 4.7 7.5 9.0 6.6 7.0 Source: Economic Survey of Pakistan, 2006-07. 28. Pakistan’s reliance on agriculture is minimizing with the passage of time. It is encouraging to note that the contribution of wholesale and retail trade is increasing. It contributed 19.4% or 1.4 percentage points to GDP in 2006-07. This sector is highly labour- intensive and this higher growth may have contributed to the rise in employment and income level of people attached to the sector. K. Services Sub-sectors 29. A closer look on the composition of sub-sectors that contribute most to the services sector shows that retail trade, wholesale trade, research and educational consultancy services, restaurants and hotels, financial, construction, and some business services, such as computer and information technology (IT) services, and professional services, such as engineering, legal and accounting services are the most significant in terms of scope and scale of export markets and the extent to which these services are currently being delivered. Professional services, such as legal and accounting, is a promising sector. While it is rare to find a Pakistani architectural firm with an office abroad, a handful of engineering companies have extended their presence outside Pakistan. Medical and health services are delivering primarily to foreign nationals in the domestic market. 20 Table B7: Statistics by Economic Sector 2005 Median Median sales No. of Total sales Total Contrib. to Economic sector employment PKR ’000 firms PKR billion employment employment Retail Trade 2 216 1,439,340 697.299 2,371,419 28.47% Research Services & Institutions 2 72 311,545 43.919 1,360,015 16.33% Personal & Household services 2 72 463,728 51.576 820,281 9.85% Government (excl. Defense) 6 108 21,207 5.073 456,445 5.48% Restaurants & Hotels 2 180 145,922 43.151 344,654 4.14% Wholesale Trade 2 480 89,574 274.603 212,207 2.55% Real Estate & Business services 2 180 49,926 22.838 130,828 1.57% Communication 1 144 49,892 12.706 117,340 1.41% Recreation & Cultural services 1 84, 52,994 6.851 88,378 1.06% Financial Institutions 6 720 8,252 332481 69,445 0.83% Transport Storage 3 288 12,151 10.517 54,047 0.65% Manufacturing, Paper & Paper 3 300 10,735 19.554 42,169 0.51% Production/ Printing & Publishing Water Works & Supplies 2 62 3,155 30.532 11,682 0.14% Insurance 4 480 874 1.453 6,235 0.07% Const., Rep & Maintenance 2 108 965 0.285 2,486 0.03% Drainage, Hydraulics Building Construction 2 120 653 1.215 2,065 0.02% Construction/Repair/Maintenance 2 360 288 0.58 529 0.01% of Streets, Roads Electricity Gas & Water 6 23 30 0.1 421 0.01% Financing, Insurance, Real Estate 5 108 102 0.1 387 0.00% & Business Groups Community Social & Personal 181 - 2 - 181 0.00% services Construction Repair, 3 204 10 0.2 27 0.00% Maintenance of Railway P Construction Group 2 90 14 0.018 26 0.00% Transport Storage & 4 42 6 0.00710 14 0.00% Communications Construction, Repair, 3 480 1 0.005 3 0.00% Maintenance of Pipe Lines Construction, Repair, 2 - 5 - 3 0.00% Maintenance of Sports Premises Source: Economic Census of Pakistan, Federal Bureau of Statistics (FBS) – tabulated by SMEDA. 30. The courier services sector stands out in terms of scale, and the Internet Service Providers distinguish themselves in terms of scope. Environmental services have significant experts in the area of environmental impact assessments. The banking sector is very prominent, while securities and asset management firms are also actively involved in international transactions mainly to solicit funds from foreign institutional and private investors. The insurance sector lags behind compared to the rest of the financial sector. Communications, distribution and transport services are significant in terms of scale of services traded but their supply is limited. 21 ANNEX C: WEIGHTED INVESTMENT CLIMATE VARIABLES Table C1: Summary of Investment Climate Indicators (FY02 and FY07) FY07 Size State Balu- All Small Medium Large Punjab Sindh chistan NWFP FY02 Infrastructure Power supply: Average number of power outages suffered by 33.6 34.2 36.0 18.3 40.8 23.9 18.0 20.3 14.6 month1 Average duration of power outages in hours1 2.2 2.2 2.4 1.6 1.8 2.8 1.4 2.0 . Losses due to power outages as a percentage of 10.1 10.5 10.0 4.6 8.7 13.0 3.7 8.4 5.5 total annual sales1 Percentage of establishments owning a 20.1 9.8 46.7 87.7 6.6 41.6 11.5 18.8 41.3 generator Percentage of electricity used from a generator 5.9 2.6 10.0 40.3 1.0 14.0 0.9 3.2 6.9 Water supply: Percentage of establishments experiencing 20.5 14.6 33.2 65.8 8.8 39.3 12.6 12.3 . insufficient water supply No. of water outages by month2 2.0 1.6 3.4 4.5 0.9 4.0 0.1 0.6 4.4 Average duration of water outages in hours2 2.5 1.7 5.3 5.3 1.0 5.0 0.7 1.8 . Percentage of water used from public sources 37.4 37.0 40.0 36.2 26.8 51.6 65.4 78.3 . Transportation: Percentage of establishments owning transport 16.5 11.3 31.8 45.2 5.6 32.3 27.3 8.9 . to ship products to costumers Percentage of products shipped with own 11.0 8.0 22.3 22.1 3.6 22.7 8.8 4.9 . transport Percent of the value of products shipped to 0.3 1.2 0.3 0.0 0.7 0.2 0.4 0.0 . supply domestic markets lost because of theft in international sales Percent of the value of products shipped to 0.5 0.5 0.3 0.7 0.4 0.5 0.2 1.1 . supply domestic markets lost because of theft in domestic sales Trade: Average number of days to clear customs in 3.7 2.0 3.9 4.2 7.2 2.4 5.7 . 9.9 exports Average number of days to clear customs in 7.3 4.7 10.9 4.6 17.6 5.4 5.0 3.5 17.6 imports Other infrastructures: Percentage of establishments located in an 28.5 20.7 47.6 83.0 15.2 49.4 25.7 8.4 . industrial park Average number of days that main intermediate 21.4 16.3 34.4 55.0 15.5 29.5 30.1 24.0 26.7 material is kept in stock Managers' perceptions: Percentage of managers considering electricity 79.7 80.5 78.0 73.6 77.2 85.6 60.0 41.0 7.7 as a major or very sever obstacle for operation and growth Percentage of managers considering 14.9 14.2 16.6 20.1 8.2 25.8 12.2 6.0 10.2 transportation as a major or very sever obstacle for operation and growth Percentage of managers considering customs 5.8 4.6 10.9 5.1 1.3 10.1 11.4 10.9 24.4 and trade regulations as a major or very sever obstacle for operation and growth Notes: In fiscal year 2007 all are weighted averages correcting for oversampling of large firms. "." no information available. 1 Conditional on the number of firms reporting having power outages. 2 Conditional on the number of firms reporting having water outages. 23 Table C1: Summary of Investment Climate Indicators (FY02 and FY07) – Cont Pakistan FY077 Size State Balu- Pakistan All Small Medium Large Punjab Sindh chistan NWFP FY02 Economic governance Courts and crime: Perc. of establms. having conflicts with 4.7 3.2 7.7 16.9 3.0 5.7 23.5 5.3 16.3 clients with a third part involved Perc. of establms. conflicts with clients 1.4 0.5 1.7 13.1 0.8 2.6 0.7 2.1 4.9 with a court involved Security expenses as a percentage of 4.0 5.4 1.1 2.8 2.6 6.5 1.7 8.4 1.4 total annual sales Crime losses as a percentage of total 0.5 0.5 0.4 0.5 0.4 0.7 0.0 0.9 0.2 annual sales Red tape and regulatory burden: Percentage of manager's time spent in 2.2 1.7 3.9 4.5 1.7 3.0 0.9 1.5 9.9 bureaucratic issues Number of tax inspections by year 1.5 1.3 1.7 3.7 0.9 2.2 3.0 1.4 12.7 Number of labor inspections by year 1.3 0.9 3.1 2.6 0.8 2.2 1.6 0.7 10.1 Percentage of establishments enjoying 16.7 8.2 33.9 85.6 7.9 31.1 8.0 12.1 27.4 a tax exemption Average number of days waiting to 29.6 34.4 16.5 35.2 66.3 26.2 . . 7.9 obtain a construction permit Average number of days waiting to 19.2 37.7 6.0 3.0 71.6 9.4 5.9 . 17.6 obtain an operation license Average number of days waiting to 91.8 100.6 85.8 26.6 55.5 126.2 168.2 . 31.8 obtain a power supply Average number of days waiting to 234.4 446.6 31.0 11.5 107.4 258.2 17.7 3.0 . obtain a water supply Average number of days waiting to 40.2 48.1 14.3 32.6 75.3 13.4 15.9 125.4 25.3 obtain a phone connection Average number of days waiting to 14.2 60.0 11.0 30.0 29.9 5.9 . . 6.4 obtain an import license Corruption: Perc. of establims. for which a gift 28.2 25.2 41.6 33.4 19.3 41.5 40.1 22.4 26.7 were requested in tax or labor inspections Payments to obtain a contract with the 0.8 0.8 0.7 1.6 1.8 0.5 0.0 0.9 . government as a percentage of contract value Payments to 'get things done' with 34.4 33.0 38.1 42.3 41.1 25.9 16.5 11.1 2.0 regard to bureaucratic issues as a percentage of total annual sales Perc. of establms. for which a gift were 0.6 0.1 1.4 4.9 0.3 1.1 0.0 0.0 0.2 requested to obtain a construction permit3 Percentage of establishments for which 0.4 0.5 0.0 0.1 0.3 0.7 0.0 0.0 0.1 a gift were requested to obtain an operating license3 Percentage of establishments for which 1.4 1.2 3.1 0.5 0.5 2.9 0.0 5.8 3.0 a gift were requested to obtain a phone connection3 Percentage of establishments for which a 1.6 0.6 4.4 7.6 0.4 3.5 0.0 1.2 . gift were requested to obtain a water supply3 24 Pakistan FY077 Size State Balu- Pakistan All Small Medium Large Punjab Sindh chistan NWFP FY02 Percentage of establishments for which a 4.8 4.4 7.6 3.5 3.5 7.2 1.6 2.5 1.5 gift were requested to obtain a power supply3 Percentage of establishments for which 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 a gift were requested to obtain an import license3 Perc. of establishments for which a gift 7.1 5.7 11.1 15.9 4.5 11.6 1.5 9.3 3.8 were requested to obtain any connection, supply, permit or license Informality: Percentage annual sales 'typically' 94.7 93.6 98.2 98.4 95.2 94.5 90.0 90.2 . reported for tax purposes Percentage workforce 'typically' 75.0 71.4 83.7 91.5 66.2 88.2 89.9 62.4 . reported for tax purposes Managers' perceptions: Percentage of managers considering 32.6 28.6 52.0 35.4 26.7 44.8 2.5 44.3 21.4 crime, theft and disorder as a major or very severe obstacle for op. and growth Percentage of managers considering 36.4 34.1 49.1 51.0 23.4 70.3 45.5 49.0 . functioning of the courts as a major or very severe obstacle for operation and growth Percentage of managers considering 56.7 54.1 61.8 76.8 43.2 78.1 59.6 98.0 40.3 corruption as a major or very severe obstacle for operation and growth Percentage of managers considering 40.3 37.3 46.6 63.0 38.9 40.5 66.0 26.3 46.8 tax rates as a major or very severe obstacle for operation and growth Percentage of managers considering 23.2 21.5 25.4 39.9 20.9 24.8 48.5 16.7 47.0 tax administration as a major or very severe obstacle for operation and growth Perc. of managers considering anti- 14.1 11.0 28.2 14.0 7.0 26.1 19.6 10.7 21.4 competitive practices of formal firms as a major or very severe obstacle for operation and growth Perc. of managers considering business 16.9 18.1 13.5 11.9 16.3 18.1 15.4 11.0 14.7 licensing and permits as a major or very severe obstacle for operation and growth Notes: In fiscal year 2007 all are weighted averages correcting for oversampling of large firms. "." no information available. 3 Conditional on the number of firms having requested such a connection. 25 Table C1: Summary of Investment Climate Indicators (FY02 and FY07) – Cont Pakistan FY07 Size State Balu- Pakistan All Small Medium Large Punjab Sindh chistan NWFP FY02 Finance Percentage of working capital financed 84.3 86.4 76.6 76.2 85.6 81.6 91.0 95.0 65.3 by internal founds Percentage of working capital financed 1.7 0.9 2.8 9.1 1.7 1.8 2.7 0.0 5.1 by private banks Percentage of working capital financed 0.2 0.1 0.3 0.0 0.2 0.1 0.0 0.0 0.1 by state-owned banks Percentage of working capital financed 0.1 0.1 0.2 0.1 0.1 0.0 0.0 0.0 6.9 by family/friends Percentage of working capital financed 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 . by non-bank financial institutions Percentage of working capital financed 14.1 12.8 20.7 14.8 12.1 18.3 6.3 5.1 4.9 by trade credit Percentage of working capital financed 0.1 0.0 0.2 0.0 0.0 0.1 0.8 0.0 1.4 by informal founds Percentage of establishments having a 63.3 56.2 86.1 99.2 58.3 69.6 78.6 57.1 . checking or saving account Percentage of establishments having 13.4 5.4 29.5 77.7 7.8 22.6 11.4 13.4 21.8 access to a credit line Percentage of establishments having a 8.5 3.4 11.7 67.8 3.1 17.3 3.5 13.2 19.6 loan Percentage of establishments having a 5.7 2.0 8.7 57.1 2.9 10.4 3.5 13.2 16.0 loan with collateral Average value of the collateral as a 9.5 4.6 13.3 65.3 3.7 19.0 3.5 13.8 11.7 percentage of the total value of the loan4 Percentage of establishments with 17.9 7.2 47.3 87.8 5.8 36.6 19.8 4.9 40.1 external auditory Percentage of lands owned by the 78.8 76.3 84.7 96.6 81.5 77.1 51.5 68.6 93.6 establishment Percentage of purchases paid before 6.8 5.5 10.3 15.4 5.2 9.8 0.1 39.4 10.7 delivery Percentage of sales paid for before delivery 4.4 3.7 5.8 9.6 2.3 8.1 0.0 31.6 . Managers' perceptions: Percentage of managers considering 17.6 20.2 9.1 9.0 16.5 16.0 48.8 25.2 38.3 access to financing as a major or very sever obstacle for operation and growth Percentage of managers considering 26.4 25.4 31.0 27.9 16.2 40.9 51.6 50.9 21.2 access to land as a major or very sever obstacle for operation and growth Innovation and competition Percentage of establishments with quality 8.7 1.4 22.2 70.7 3.6 16.7 9.0 7.0 17.4 certification Percentage of establishments using web 94.7 98.9 87.3 99.9 100.0 92.9 100.0 100.0 14.5 page to communicate with clients and suppliers Percentage of establishments using e-mail 25.6 12.7 63.0 98.8 11.3 47.5 29.6 1.5 30.4 to communicate with clients and suppliers Percentage of establishments having a 2.7 1.3 4.5 17.1 0.4 6.2 3.7 0.7 . foreign licensed technology Percentage of establishments that 6.4 3.5 15.5 21.6 2.1 13.2 4.4 9.0 . 26 Pakistan FY07 Size State Balu- Pakistan All Small Medium Large Punjab Sindh chistan NWFP FY02 introduced a substantial product innovation during FY06 Percentage of establishments that 7.0 2.6 19.3 33.2 1.7 15.8 2.6 3.2 . introduced a substantial process innovation during FY06 Percentage of establishments that agreed 0.6 0.1 2.2 3.6 0.4 1.0 0.2 0.0 . a joint venture Percentage of establishments outsourcing 1.6 1.6 1.2 2.8 0.2 4.0 0.0 0.0 . any part of the production activity Percentage of firms subcontracting R&D 0.7 0.0 3.6 2.2 0.2 1.6 0.0 0.0 . activities Percentage of computer controlled 3.0 0.8 9.4 15.6 0.4 7.2 1.0 1.1 . machinery used Percentage of machinery and equipment 23.2 20.7 26.1 47.5 13.1 39.0 23.4 33.5 13.2 that is less than 5 years old Percentage of establishments receiving 0.8 0.1 2.3 6.4 0.3 1.7 0.0 0.8 1.3 FDI inflows Percentage of establishments importing 17.4 10.6 33.5 63.6 12.8 25.7 14.6 8.9 14.4 Percentage of establishments exporting 11.5 2.0 35.5 75.3 4.1 28.4 7.5 0.0 18.0 directly Average number of years of exporting 2.0 0.2 6.5 14.1 0.7 4.1 0.2 0.0 2.7 experience Percentage of establishments operating in 2.6 3.1 0.0 1.4 0.6 6.0 0.0 0.7 0.7 a local monopoly Percentage of establishments competing 11.1 11.1 12.1 7.7 10.8 11.0 13.6 46.2 7.9 with less than 5 competitors Percentage of establishments competing 86.3 85.8 87.9 90.8 88.7 83.0 86.4 53.1 91.4 with more than 5 competitors Notes: In fiscal year 2007 all are weighted averages correcting for oversampling of large firms. "." no information available. 4 Conditional on the number of firms having a loan. 5 Conditional on the number of firms providing training. 27 Table C1: Summary of Investment Climate Indicators (FY02 and FY07) – Cont Pakistan fiscal year 2007 Size State Balu- Pakistan All Small Medium Large Punjab Sindh chistan NWFP FY02 Labor markets and skills Percentage of production workers in 78.8 78.5 78.5 83.0 79.5 77.5 79.7 85.5 70.1 staff Percentage of female workers in staff 1.2 0.5 3.2 5.1 0.5 2.0 5.7 0.0 2.5 Percentage of skilled workers in staff 64.5 65.7 62.3 54.5 66.7 60.7 68.0 75.5 76.7 Percentage of staff using computer at 3.3 1.3 8.8 16.4 0.9 7.4 2.4 0.3 6.2 work Percentage of staff with university 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 education Percentage of establishments providing 4.5 1.7 11.3 24.2 0.6 9.8 14.3 4.5 7.8 (beyond on the job) training to employees Percentage of production workers 2.4 0.4 6.9 16.9 0.3 5.0 10.7 0.9 1.0 receiving trainng5 Percentage of non-production workers 1.3 0.0 3.7 12.8 0.2 2.8 3.9 0.1 0.1 receiving training5 Number of years of experience of the 20.5 19.7 23.6 23.1 20.7 20.3 19.9 14.5 6.6 manager Percentage of managers having at least 34.1 21.5 73.5 99.6 23.8 50.4 29.1 27.2 72.6 university education Managers' perceptions: Percentage of managers considering 5.9 5.7 6.7 6.2 0.3 15.6 0.6 0.0 15.8 labor regulations as a major or very sever obstacle for operation and growth Percentage of managers considering 7.1 5.8 13.7 8.4 2.3 15.0 10.7 0.0 13.0 inadequately educated workforce as a major or very sever obstacle for operation and growth Corporate governance Percentage of publicly listed 2.5 0.5 5.5 21.8 1.4 4.5 0.1 0.0 3.3 companies Percentage of limited liability 9.6 4.1 23.4 46.7 5.2 16.1 14.1 0.8 50.8 companies Percentage of state-owned companies 0.5 0.0 2.7 1.0 0.7 0.3 0.0 0.0 0.9 Percentage of firms' share owned on 85.5 91.8 66.6 51.0 88.8 79.9 89.4 92.8 . average by the largest shareholder Managers' perceptions: Percentage of managers considering 56.5 55.6 58.9 62.9 53.0 65.0 30.6 46.0 34.5 macroeconomic instability as a major or very sever obstacle for operation and growth Percentage of managers considering 46.8 45.0 54.3 51.6 33.5 69.3 38.6 61.8 40.4 political instability as a major or very sever obstacle for operation and growth Notes: In fiscal year 2007 all the averages are computed according to probability weights to correct for oversampling of large firms. "." no information available. In fiscal year 2007 all are weighted averages correcting for oversampling of large firms. 28 ANNEX D: SAMPLING METHODOLOGY 1. Background. Through a statistically robust and representative sampling methodology, the Pakistan Enterprise Survey (ES) enables researchers to construct around xxx indicators which characterize various aspects of the Pakistan’s business environment from the perspective of the entrepreneur. The indicators, collected and calculated in a standardized fashion, measure subjective and objective dimensions of the investment climate, covering key dimensions, including infrastructure, market governance (red tape, bureaucracy, and corruption), courts & crime, competition & innovation, and factor markets (land, labor & finance). In addition, in order to enable analysis of the impact of the investment climate indicators on enterprise performance data is collected to measure sales, employment and productivity growth. 2. The indicators will allow researchers to benchmark investment climate of individual economies across the world and conduct firm performance analysis focusing on constraints to productivity and job creation in selected sectors. In addition, for this ES in Pakistan, in order to assess the impact of changes in the business environment on firm-level productivity and job creation over time, close to a third of the FY07 sample were drawn from the list of firms surveyed in 2002 (panel firms). 3. Sampling Goals. The sampling goals for the Pakistan ES were therefore to generate: (i) statistically significant and representative investment climate indicators which can be analyzed at level of the nation, four provinces, size category and industry breakdown, (ii) a dynamic panel of manufacturing firms surveyed in 2002 and 2006, and (iii) a complete list with address, sectors, size and other establishment characteristics of as a basis for future follow-up panels using the 2006 survey.8 To accomplish these goals, the Economic Census of Pakistan 2005, carried out by the Federal Bureau of Statistics (FBS), Government of Pakistan,—the first and only comprehensive firm level census ever conducted in Pakistan - was used as the sampling frame representing the population from which was drawn a stratified random sample. 4. Institutional Management. As the Economic Census of Manufacturing is proprietary to the Government of Pakistan and kept by FBS to protect the confidentiality of the respondents, FBS agreed to carry out the sampling procedure itself based on World Bank specified guidelines for sampling procedures for Enterprise Surveys. Based on Government business policy, another Government agency needed to be the fee based requestor of the sampling and recipient of the enterprise lists generated by the sampling process. 5. The Small and Medium Enterprise Development Authority (SMEDA) as the technical counterpart for the Investment Climate Assessment (ICA-II) Survey, managed the primary interface with the FBS and entered into a contractual arrangement with FBS to provide four products: (i) a stratified sample, representative at the national, provincial and sectoral levels drawn from the 2005 Economic Census, (ii) replacement samples to account for refusals, 8 The survey covered manufacturing sector and services and generated a large enough sample size for selected industries to conduct statistically robust analyses. With level of precision at a minimum 7.5 percent for 90 percent confidence intervals about estimates of population proportions and mean of log sales at the national, provincial and industry level. 29 closures and non-responses, (iii) sampling weights to maintain representativeness, and (iv) documentation of the sampling process, methodology, calculations, adjustments and caveats. 6. Global Standards. The Pakistan ES is based on guidelines for the uniform sampling methodology prescribed by the World Bank’s Investment Climate Guide9 to allow calculation of investment climate indicators that minimize sampling and non-sampling errors, are nationally and provincially representative and comparable across the world’s economies. 7. The survey adhered strictly to the global guidelines in terms of target firms and sectors, as well as sampling requirements regarding sample design, field work and post survey adjustments through appropriately designed sample weights. In particular, formal, urban firms with employment size of at least 5 were targeted. Minimum stratification requirements and sample sizes were exceeded in the sampling strategy as well as minimum iterations were imposed regarding replacements for non-locatable and refusing firms. Table D1: Pakistan ICA II Sampling Frame Distribution of Firms and Employment (by size and provincial distribution) Number of firms Manufacturing Services Total <50 emp > 50 emp Total <50 emp > 50 emp Total <50 emp > 50 emp Total Punjab 13,882 177 14,059 1,182 12 1,194 15,064 189 15,253 Sindh 4,665 429 5,094 3,258 13 3,271 7,923 442 8,365 Baluch 123 12 135 167 0 167 290 12 302 NWFP 993 36 1,029 574 4 578 1,567 40 1,607 Total 19,663 654 20,317 5,181 29 5,210 24,844 683 25,527 Reported formal employment Manufacturing Services Total <50 emp > 50 emp Total <50 emp > 50 emp Total <50 emp > 50 emp Total Punjab 137,697 41,509 179,206 9,470 1,724 11,194 147,167 43,233 190,400 Sindh 50,438 101,329 151,767 25,527 4,500 30,027 75,965 105,829 181,794 Baluch 986 3,053 4,039 1,440 0 1,440 2,426 3,053 5,479 NWFP 8,518 5,485 14,003 4,839 219 5,058 13,357 5,704 19,061 Total 197,639 151,376 349,015 41,276 6,443 47,719 238,915 157,819 396,734 8. Sampling Frame. The sampling frame used for manufacturing and services was drawn from the 2005 Economic Census of Pakistan, which included close to 3 million establishments. As the target population was formal, urban manufacturing and services establishments with more than 5 full time employees, the census identified 583,329 manufacturing firms and 1,566,722 establishments in Wholesale/Retail trade & Restaurants. 9 The Pakistan Enterprise Survey II, followed the global standards set by the Investment Climate Department of the World Bank, as described in the publication, xxxxx. 30 9. For manufacturing, 86.7 percent of firms employ 1-5 Table D2: Frame Statistics Sector workers, reducing the frame to 77,375 to cover firms with 6 or (by share of total) more employees and reduced further to 20,317 when including No. Employ. only firms from within the urban areas. For services, Food 7.0 10.4 establishments in those services sectors with at least 5 Textile 48.6 51.8 employees and an urban location numbered around 31,000. Chem 6.8 11.2 The cities covered were: Karachi, Lahore, Peshawar, and Metals 19.5 16.5 Quetta and the total number of firms in the frame numbered Sports 2.9 1.8 around 5,200 (Table D2).10 The employment coverage of the Other 15.2 8.3 sampling frame shows 350,000 formal, full time employees in Manufact 100.0 100.0 the manufacturing sector and another 45,000 employees in the Service 20.4 12.0 service activities targeted.11 Total 100.0 100.0 10. The distribution of the frame by size, province and sector is shown above in Table 1. A significant dimension of the frame concerns the small average size of firms—with size represented as number of reported full time employees. As Table 1 reveals, firms with less than 50 workers account for the bulk of the frame—over 97 percent of firms and 60 percent of total reported full time employment. This is true for all provinces apart from Sindh, where larger firms (greater than 50 employees) make up a greater proportion of the total (6 percent) and account for almost 60 percent of employment, due to the higher prevalence of larger firms in Karachi. 11. The provinces of Punjab and Sindh account for the bulk of economic activity as represented by the sampling frame. Together they represent 94 percent of all firms and employment. However, owing to the small firm size in Punjab, Sindh firms account for almost half of employment, but only 25 percent of firms. 12. The breakdown of firms and employment into sector categories in Table 2 shows that about half of the manufacturing establishments in the frame involved textiles, garments and leather, with another fifth in metal working and electric machinery. The other significant sectors (Food and Chemicals) account for only 7 percent of firms each, but together cover over a fifth of employment. Services on the other hand, account for a fifth of total firms, but only 12 percent of employment. 13. Sampling Methodology. The sample followed a stratified random sampling approach; stratified probability proportional to size (PPS) sample design. In order to obtain statistically significance and representation at the provincial and sector level, eighty-four individual strata were used. In addition, due to the predominance of small firms in the frame, the non- proportional size (employment) variable was used to ensure the sampling process selected with due probability the entire range size categories. 10 Based on need for a statistically significant service stratum in each province and an analysis of the city-wise distribution of these services, the 4 provincial capitals were selected as having the largest concentration of activities, in the services sector in order to reach a sufficient number of firms for each city out of the 150 total firms planned. 11 The low share of frame as percent of the working population is due to (i) the census counting regular, formal employment as reported by firms, (with no cross-checking), (ii) the high employment which takes place in firms with less than 6 reported employees, (iii) the high number of firms in semi-urban areas which were not captured by the filter of urban firms. 31 14. In accordance with the size and make up of the economy, the manufacturing sector was stratified into five 2-digit Pakistan Standard Industrial Classification (PSIC) sectors: (i) food processing, (ii) textiles, apparel & leather, (iii) chemicals and products, (iv) metal and electric machinery, and (v) sports goods and handicrafts with a residual stratum based on the 14 largest cities from the four provinces of the country. Services establishments engaged in wholesale & retail trade, hotels & restaurants were grouped to constitute an independent stratum for each provincial capital. 15. In other words, within each industry, the total sample size was distributed to the provincial/city sub-strata based on proportional allocation in order to be representative of the nation, the industry groups and the urban areas of each of the four provinces. Given the domination of smaller firms in sample frame, a sampling approach which oversampled larger firms was employed to ensure a sufficient number of large enterprise which otherwise might be underrepresented. 16. The specific steps involved: (i) extracting from the frame and dividing into activity/industry groups with selection made in proportion to each group’s contribution to total industrial employment, (ii) allocating the establishments selected in to each industry group across the provinces/cities selected using a proportional allocation, and (iii) selecting the establishments for each province/city sub-stratum with a probability of selection which is inversely proportional to size (i.e. larger firms will be selected with a higher probability). Due to the oversampling of larger firms, weights were computed so that inferences about the population could be extrapolated from the sample. Table D3: 2006/07 Representative Sample Frame and Sample Distribution Number of firms Punjab Sindh Baluchistan NWFP Total Economy Population Sample Population Sample Population Sample Population Sample Population Sample Food 611 71 570 49 44 17 193 13 1,418 150 Textile 6,819 147 2,612 89 38 9 401 8 9,870 253 Chem 916 59 373 20 6 2 83 5 1,378 86 Metal 3,376 110 467 17 16 9 111 3 3,970 139 Sports 386 39 193 21 1 0 15 2 595 62 Other 1,951 53 879 22 30 11 226 8 3,086 94 All manuf 14,059 479 5,094 218 135 48 1,029 39 20,317 784 Service 1,194 35 3,271 56 167 30 578 30 5,210 151 Total 15,253 514 8,365 274 302 78 1,607 69 25,527 935 Formal full time employment Punjab Sindh Baluchistan NWFP Total Economy Population Sample Population Sample Population Sample Population Sample Population Sample Food 11,704 2,547 21,842 2,932 406 259 2,178 700 36,130 6,438 Textile 89,103 4,734 86,145 9,092 2,009 528 3,662 316 180,919 14,670 Chem 16,355 3,195 18,598 5,973 13 13 4,085 1,121 39,051 10,302 Metal 41,755 4,488 13,481 2,470 1,264 522 952 76 57,452 7,556 Sports 3,541 594 2,334 989 8 0 527 337 6,410 1,920 Other 16,748 2,521 9,367 874 339 259 2,599 931 29,053 4,585 18,07 22,33 All manuf 179,206 9 151,767 0 4,039 1,581 14,003 3,481 349,015 45,471 Service 11,194 1,329 30,027 1,517 1,440 327 5,058 589 47,719 3,762 19,40 23,84 Total 190,400 8 181,794 7 5,479 1,908 19,061 4,070 396,734 49,233 32 17. Sample Size. The sample was drawn to satisfy the minimum sizes at provincial and industry levels and allocated for each city based on the relative share of each selected city out of the total frame.12 The total number of manufacturing and service establishments to be included in the final sample was targeted at 1,350 taking into consideration the need for minimum sample at the sector and provincial level and to ensure a reasonable minimum number of panel observations. 18. The survey field work was able to complete, check and confirm 1,337 interviews, 935 of which were non-panel interviews making up the representative sample for 2006/07. Of this, the 784 manufacturing and 151 service firms making up the ‘Non-panel’ weighted sample exceeds the 674 and 117 minimum sample size, respectively, needed to accommodate the sampling goals and stratification variables. For individual stratum, Table 3 below shows the frame and sample distribution of Non Panel across individual strata used in the sampling methodology. Table D4: Panel Sample 2001/2-2006/7 (by distribution of frame and sample) Number of establishments Punjab Sindh Baluchistan NWFP Total Economy Population Sample Population Sample Population Sample Population Sample Population Sample Food 182 13 160 26 35 2 112 22 489 63 Textile 1,601 146 1,041 55 44 4 17 1 2,703 206 Chem 215 29 121 11 28 2 31 3 395 45 Metal 1,129 52 82 4 45 0 51 7 1,307 63 Sports 261 25 0 0 0 0 0 0 261 25 Other 172 0 96 0 0 0 2 0 270 0 Total 3,560 265 1,500 96 152 8 213 33 5,425 402 19. Panel Firms. The survey strategy attempted to take advantage of the 2002 Pakistan Enterprise Survey in order to develop a dynamic panel.13 The panel would enable analysis of responses from the same firm made in 2001/2 and 2006/7 in order to draw inferences about changes in the investment climate over time and to attempt to attribute causes to those changes. 20. However, important difference between the approach taken in 2002 and the 2006 survey limits the range of panel-type of analyses to some degree. In particular, as the frame in 2002 was not representative of the population, and as weighting was not done in 2002, inferences made from the panel analysis cannot be take as statistically representative of the population in 2001 nor 2006, but are rather indicative of changes in the investment climate over the 5 year period. 21. The Pakistan Enterprise Survey II sample was also designed to include up to 600 firms from the original sample of Pakistan Enterprise Survey I, out of a total of 846 establishments surveyed in 2002 (panel firms with location and other identifiers). The remaining 246 establishments were kept as potential replacements in case of non-response by an establishment 12 Based on the global ES guidelines, the minimum sample size per industry (and province) to conduct statistically robust analyses is met to make inferences for 4 manufacturing sectors and services as well as for provincial analysis at the minimum 7.5 percent precision for 90 percent confidence. 13 Some significant differences between the 2002 and 2006 ES limit the use of the panel approach, to some extent. In particular, (i) the common questions between the 2002 and 2006 sample are somewhat limited as the core survey instrument changed substantially over the period, (ii) the frame was constructed using non-representative lists from Provincial Governments, (totaling firms 11,000, of which 5,425 were firms with employment greater than 5), and (iii) the stratification, oversampling and weighting strategy was quite different from that employed in 2006. 33 of similar characteristics in the original panel sample. In the end, 402 firms were interviewed out of 795 firms contacted. The frame and sample distribution of the 402 panel firms is shown below in Table 4. 22. Survey Response Rates. Due to a high number of firms which could not be located, 3 replacement samples were drawn by FBS maintaining the original sampling parameters (firms were replaced from the same strata). Table 5 is drawn from survey field reports as reported by the Survey firm, to provide details on the sample, replacements and non-response rates. Table D5: Survey Field Report Added Attempted Original Sample Contact Not Used Completed Not Found Closed Refused Non Usable Panel 600 195 795 0 402 120 119 120 34 Non Panel 750 1,127 1,547 330 935 335 79 130 68 Total 1,350 1,322 2,342 330 1,337 455 198 250 102 23. The field work involved a sample of almost 2700 firms with more than 2,300 firms contacted in order to complete the survey of 1,337 firms - a seemingly low 57 percent success rate. However, of the 1,000 non-successful contacts, about 45 percent were not located due to poor contact information and 25 percent refused to participate.14 Of the rest, only 20 percent were closed and 10 percent either were either non-responsive or produced non-usable data.15 For the non-panel sample, the response rate was slightly higher at 60 percent, but of the 612 non- responding firms, 55 percent were not found due to insufficient contact information, 21 percent refused participation, 11 percent were non-usable and 13 percent were confirmed as closed. 24. Calculation of Sampling Weights. Based on the sampling approach described above, where the sample was drawn using stratification along two variables and the firms were chosen with a probability related to size, appropriate weights were needed to accommodate the oversampling of larger firms. Sampling weights were calculated based on the distribution of the frame across stratification variables (city and sector). Each establishment was weighted by the establishment’s share of employment in the stratum to which the firm belongs. In other words, the weight is calculated as the inverse of the probability of selection which in turn is calculated by taking the employment of the firm, dividing by the employment of the firm’s stratum and multiplying by the number of firms sampled from the firm’s stratum. 14 More detailed reports from the field, showing the distribution of success rates, indicates that the distribution of the non-located firms and firms refusing to participate was random and distributed proportionally across the sample, thereby having minimum impact on the sampling goals. 15 Post survey adjustments in the sampling weights were not made, as there has not been a second census and therefore no information to calculate entry and exit. Only 20 percent of non-responders (4.4 percent of total firms contacted) were confirmed to have been closed and this was assumed to be proportional to the number of firms which have entered during the period. In addition, the firms which could not be located were presumed to exist but had poor contact information. Field work with FBS regional offices confirmed this presumption. 34 ANNEX E: ECONOMETRIC METHODS FOR PRODUCTIVITY AND INVESTMENT CLIMATE ANALYSIS 1. In the identification of the statistically significant investment climate effects on economic performance (productivity, demand for labor, real wages, probability of exporting and probability of receiving FDI) our main concern is to use the information contained in the IC survey to perform the estimation of the system and, at the same time, to control for all the contingent problems we are going to find: measuring productivity, simultaneous determination of the dependent variables of the system, endogenous regressors, over- identification, weak instruments as well as data quality issues such as missing values, outliers or measurement errors. In this appendix we briefly describe the methodology used and how we deal with the various methodological problems found. We follow Escribano and Guasch (2005 and 2008), Escribano et al (2008a and b) and Escribano, Guasch and Pena (2008). For more details on the econometric methodology applied in the ICA of Pakistan see Escribano, de Orte and Pena (2008). 2. The information we have for the analysis consists of three different sets of data. The firs dataset is a sample with 784 manufacturing establishments. The second one consists of 151 services establishments. Finally, the last dataset is a panel with information for 402 manufacturing firms in years 2001 and 2006. 3. Both manufacturing and services surveys provides information for production function variables for 2 years, although the information for IC variables is fixed. Nonetheless, we use information only for 2006 because we think information in 2005 is flawless, full of missing data and therefore too noisy to use it in the analysis. The description of the variables used, both dependent variables (productivity, employment, real wages, exports and FDI) and independent (set of more than 140 IC) variables can be found in tables 1.1 to 2.7. 4. We implement a different econometric methodology for each one of the three datasets. We first describe the methodology for the manufacturing sector and we follow with services and the panel. Most of the econometric difficulties described for manufacturing apply with almost no change to services and the panel. 35 Box 1: Summary of the econometric methods used to assess the IC effects on economic performance Basically, the econometric methodology consists of two steps: a. Identification of statistically significant IC effects on productivity, employment, wages, exporting propensity and FDI propensity. For the identification of the statistically significant IC effects on economic performance, we propose a simultaneous equations system that relates the interactions between the investment climate with productivity, demand for labor, wages, exports and FDI inflows. In the estimation we always control for firms’ size, state and sector. From the estimation we get the IC elasticities and semi-elasticities with respect to productivity, employment, wages, exporting propensity and FDI propensity. b. Evaluation of relative IC contribution to aggregate (weighted average) productivity and the sample means (un- weighted average) of economic performance variables. The IC elasticities and semi-elasticities give us a measure of the sensitiveness of outcome variables when the IC changes marginally. In order to consider the level of the IC in the population we also evaluate the relative contribution of the different IC variables on the sample means of the dependent or outcome variables (productivity, employment, wages, exports and FDI). For the case of productivity we go one step forward and we evaluate the IC on the terms of the Olley and Pakes (O&P) decomposition of aggregate productivity (or weighted average using as weights the share of sales) into average productivity and allocative efficiency term. In what follows we describe the estimation procedure, and we address the main econometric difficulties as well as the solutions proposed. Box 2: Estimation of IC effects on productivity, employment, wages, exports and FDI The identification of the statistically significant IC impacts on the outcomes is done by a system of five equations as the following: log Pit  ICiP  yit   P  IC Exp  yit FDI  Dr   Ds  DR  Dm  (vP,i   P,it )  Dj  DM (b2.1) logLi   L   P log Pi   w log Wi  yi Exp  Dr   Ds  ICiL   DR  yiFDI   L  Dj   DM  Dm  (vL,i   L,i ) (b2.2) logWi   W   P log Pi  yiExp  yiFDI   IC  IC i   DR  Dr   Ds  Dm  ( vW ,i   W ,i )  D j   DM (b2.3) yiExp   Exp   P log Pi  y i FDI  IC   IC Exp i  Dr   Ds   DR  Dm  (vExp ,i   Exp ,i )  D j   DM (b2.4) yi  FDI  P log P FDI i y  IC  DR  IC i Exp  Dr  Ds  Dj  DM FDI i  Dm  (vFDI ,i   FDI ,i ) (b2.5) exp FDI e Where P is productivity, L is employment, W are wages per employee, y and y IC e=P,L,W,Exp and FDI are the vectors of significant IC variables in each equation, Dr, Dj and Dm are the vectors of region, sector and size dummies respectively and ve and e are random error terms. 5. In the estimation of the system (b2.1)-(b2.5), productivity equation is of great importance. Since productivity enters as endogenous covariate in the remaining equation, we first estimate (b2.1) and we use the specification obtained in order to estimate the remaining equations by IV techniques. TFP as a residual is a black box that may contain any factor that affect the way firms transform inputs into outputs. Therefore we do not know what TFP exactly is. In addition, from now on, there is no a unified TFP theory (Prescott, 2004). The philosophy implicit in the estimation of productivity equation is to use the information contained in the IC survey of Pakistan to bring the abstruse TFP concept to the empirical world. That is, we fill the initially empty TFP measure with hard data numbers reflecting the way firms transform inputs into output and the factors mapping the differences between the more and the less efficient ones. Productivity estimation is summarized in Box 3. 36 Box 3: Summary of econometric issues in productivity equation Productivity equation comes from a structural system of equations as the following lo g Yi   L lo g L i   M lo g M i   K lo g K i  lo g Pi (b3.1)  R Dr   D lo g Pi  a i   D  M D m   P  wi  sD j   D (b3.2)  IC P ,i   C a i   IC  C P ,i   i (b3.3) where, Y is firms’ output (sales), L is employment, M denotes intermediate materials, K is the capital stock, IC and C are time-fixed effect vectors of other investment climate and control time-fixed effects, and Dr, Dj and Dm are the vectors of state, industry and size dummies. Econometric issues in productivity equation:  No single salient productivity measure Since there is no single salient measure of productivity, any empirical evaluation of the productivity impact of the IC may critically depend on the particular productivity measure used. Solution: Escribano and Guasch (2005, 2008) who–following the literature on sensitivity analysis of Magnus and Vasnew (2006)–suggest looking for empirical results (elasticities and semi-elasticities) that are robust to several productivity measures. The different productivity measures come from: i) different functional forms of the production function (b3.1) Cobb-Douglas and Trasnlog; ii) different sets of assumptions (technology and market conditions) obtaining two different approaches the two step estimation of the system (b3.1)-(b3.3) by applying the Solow residual to estimate firm level productivity and the single step estimation by parametric techniques of (b3.1)- (b3.3); iii) different levels of aggregation in measuring input-output elasticities of equation (b3.1) (at the industry level or at the aggregate country level). The productivity measures used in this report are summarized in the next table (for more details see Escribano, de Orte and Pena (2008): Functional forms of Estimation Aggregation level of production function procedure coefficients of PF Result Two-step 1.1 Unrestricted coefficients 1. Solow´s Residual 2 (P) measures; 2 (IC) elasticities estimation 2.2 Unrestricted coefficient Single-step 2.1 Restricted coefficient 2. Cobb-Douglas 2 (P) measures; 2 (IC) elasticities estimation 2.2 Unrestricted coefficient Single-step 3.1 Restricted coefficient 3. Translog 2 (P) measures; 2 (IC) elasticities estimation 3.2 Unrestricted coefficient Total 6 (P) measures; 6 (IC) elasticities  Endogeneity of the inputs. There is an identification issue separating TFP from the production function (PF); when any PF inputs is influenced by unobserved common causes affecting productivity—such as a firm’s fixed effects—there is a simultaneous equation problem in equation (b3.1) and therefore in the single step estimation procedure (2.2). Solution: To address this well-known problem (Marschak and Andews, 1944, and Griliches and Mairesse, 1995), we follow the approach proposed by Escribano and Guasch (2005, 2008). That is, we proxy the usually unobserved firm- specific fixed effects (which are the main cause of inputs’ endogeneity) by a long list of observed firm-specific fixed effects coming from the investment climate surveys. Controlling for the largest set of IC variables and plant C characteristics, we can get—under standard regularity conditions—consistent and unbiased least squares estimators of the parameters of the PF and the IC elasticities (or reduce the degree of endogeneity in the data). The robust IC elasticities and semi-elasticities with respect to productivity (equation s b3.1, b3.2 and b3.3) can be found in Table 3. The IC elasticities and semi-elasticities with respect to the remaining economic performance measures (system b2.1-b2.5) are in Table 5. [please confirm reference] 37 Box 4: Summary of other econometric issues and solutions proposed General econometric issues (applying to any equation of the system (b2.1-b2.5):  Endogeneity of IC variables: For consistency in the estimation of (b2.1)-(b2.5) we want ve and e to be uncorrelated with any variable contained in the IC vector. It can be argued that the error terms might contain third party, unmeasured effects correlated with IC, what renders the OLS estimator inconsistent. o Solution 1: correction for observable fixed effects, by using the full set of information contained in the IC variable we are able to correct for more than 180 variables in the estimation, what eliminates a large degree of endogeneity and spurious correlations as we condition the expectation of the outcome variables in as much information as we can. o Solution 2: in spite of the observable fixed effect correction the exogeneity condition does not hold for all the variables in the model. Otherwise, we use the industry-region (or industry-region-size) counterparts of firm level IC variables, which is equivalent than applying a general IV estimator.  Endogeneity of outcome variables: For some other explanatory variables like productivity, wages, exports and FDI, the endogeneity is intrinsic due to the simultaneous structure of the system of equations. o Solution: In these cases we instrument the corresponding outcome variable with a list of exogenous IC variables. We take the list of IC variables from the explanatory variables of the corresponding equation and we estimate the system by 2SLS.  Selection of the relevant model. The population model is unknown and we have to approximate it starting from a broad set of more than 180 variables. o Solution: The econometric methodology applied for the selection of the variables goes from the general to the specific. The otherwise omitted variables problem that we encounter generates biased and inconsistent parameter estimates. We then start removing from the regressions—the less significant variables—one by one, until we obtain the final set of variables, all significant in at least one of the regressions and with parameters varying within a reasonable range of values. The set of significant IC variables is as well used as included instruments of the corresponding endogenous outcome variable in the remaining equations when estimated by IV techniques. Note that the set of explanatory variables are used as excluded instruments in the IV estimation as well.  Heteroskedasticity in the error terms. o Solution: the heteroskedasticity of the errors is addressed by using robust (white) standard errors. We also compute cluster standard errors, allowing for correlation within industry and region.  Identification of the system of equations. o Solution: In order to ensure the rank condition is satisfied in the system (b2.1)-(b2.5), we force the coefficient of certain IC variables to be 0 (exclusion restrictions) prior to start estimating the system.  Weak instruments and over-identification in IV estimation. o Solution: We select instruments (either included or excluded) seeking a partial R-squared –or ‘Shea’ partial R- squared—as high as possible with the restriction that our model cannot be over-identified. To test over- identification we use the Hansen test, a robust to general heteroskedasticity variation of classic Sargan test. In addition we take into account significance and the R2 of the first stage estimates when removing instruments. 38 Box 5. Summary of econometric issues regarding data quality Although the Investment Climate surveys are valuable instruments improving our understanding of the economic, social, political and institutional factors determining economic growth, particularly in emerging and transition economies, at the same time we have to overcome some difficult issues related with the quality of the information provided; measurement errors, outlier observations and missing data are frequently found in this datasets. The ICS of Pakistan is not an exception. Econometric issues regarding data quality:  Missing data. The number of missing values in 2006 (FY07) reduces the sample available from 784 to 358 observations in the complete case (45.6% of the sample). In 2005 the Operating with the complete case is only acceptable if missing data comprise a small percentage, say 5% or less, of the size of the sample (Schafer, 1996), and preserves the representativeness of the original sampling frame. In models with a large number of regressors missing data problem force us to leave out of the regression some explanatory variables with high proportion of missing values. As Cameron and Trivedi (2005) point out, this practice may lead to an omitted variables problem. o Solution: We input those missing values with the objective of preserving the sample representativeness, gaining efficiency in the estimation and retrieving for the analysis a large number of very expensive interviews. Basically, the imputation mechanism replaces the missing value by the expectation of the variable conditional on the information we have on sector, size and state (see Escribano and Pena, 2008 for more details). To check robustness we also estimate productivity equation under different inputation mechanisms and different assumptions on the missing data mechanism (MDM), see Escribano de Orte and Pena, 2008. After the imputation we are able to use 731observations, 93.2% of the sampling frame.  Endogenous missing data mechanism (MDM). As Escribano and Pena (2008) signals, the missing data problem may also have important implications in the consistency of the IC parameters estimates when the pattern of missing values is determined by the IC variables. o Solution: We need to control for any IC variable correlated with the MDM to achieve consistency in the estimation of the system (b2.1-b2.5).  Outliers. o Solution: We exclude from the analysis those outlier observations, understanding by outlier those observations with ratios of materials and/or labor cost to sales larger than one. Box 6. Evaluation of IC effects on the Olley and Pakes (1996) decomposition The Olley and Pakes (O&P) decomposition of aggregate productivity in logs is, lo g P  lo g P  N co Y ˆ v( s it , lo g Pit ) . (b6.1) Where logP is aggregate log-productivity (or weighted average productivity, where the weights are given by the share of sales), log P is un-weighted average log-productivity and the last term is the covariance between share of sales and firm level productivity, or allocative efficiency term, describing the ability of the markets to reallocate resources from less to more productive establishments. The useful additive property of equation (b6.1) in logarithms, allow us to obtain an exact closed form solution of the decomposition of aggregate log productivity according to (b2.2) and (b2.3). Following Escribano et al. (2008b), we can express aggregate log productivity as a weighted sum of the average values of the IC, C, dummy D variables, the intercept and the productivity average residuals; and, the sum of the covariances between the share of sales and investment climate variables IC, C, dummies D and the productivity residuals. log P   ˆ ´ IC IC P  ˆ ´C C P  ˆ ´D R D r  ˆ ´D s D j   ˆi  N  ˆp  u ˆ ˆ ´ IC cov( s iY , IC P , i )  N ˆ ´C cov( ˆ s iY , C P , i ) (b6.2)  N ˆ ˆ ´ D s cov( s iY , D j )  N ˆ ´ D R cov( ˆ s iY , D r )  N ˆ ´ D M cov( ˆ s iY , D m )  N cov( ˆ s iY , u ˆi ) Concretely, we compute the IC contributions relative to aggregate productivity as follows 100  100 [ˆ ´ IC IC P  ˆ ´C C P  ˆ ´D R D r  ˆ ´D s D j   ˆ it  N  ˆp  u ˆ v( s iY , IC P , i )  N  ˆ ´ IC co ˆ v( s iY , C P , i ) ˆ ´C co (b6.3) lo g P  N ˆ v( s , D j )  N  ˆ ´ D s co Y i ˆ v( s , D r )  N  ˆ ´ D R co Y i ˆ v( s , D m )  N co ˆ ´ D M co Y i ˆ v( s , u ˆ i )] Y i The whole results of equation (b6.3) are in Table A.2. Notice that from (b6.3):  We can compare net contributions by isolating the impact of IC variables from the impact of industry dummies, the intercept, and the residuals (demean log-productivity, see next box);  We can express what portion of aggregate productivity is explained by IC, and C variables (demean logP), and what proportion is due to the constant term, industry dummies and so on.  We can get rid of the different directions (positive or negative) of the various IC effects by simply taking the percentage contributions in absolute value.  Finally, we can compute the absolute percentage contributions to the average log-productivity by blocks of IC variables or by states, sectors or sizes. The entire results of equation (b6.3) are in Table 4. 39 Box 7. International comparisons: demean log-productivity Our knowledge of TFP is conditional on our knowledge of firms’ operating conditions. Any productivity measure is subject to measurement errors, unmeasured third party effects, differences in the deflators used, etc. To make cross-country comparisons based on IC impacts on productivity, to avoid the problem of comparing apples and oranges, it is desirable create an index (demean productivity). After subtracting the mean (that is, the constant term, time effects, industry effects and country-specific effects) from firm level log-productivity we can concentrate on the part of log-productivity explained by the IC variables. Thus, demean log-productivity at the firm level is simply D e m e a n lo g P  ˆ ´ IC I C P  ˆ ´C C P  N  Y ˆ o v ( s it ˆ ´ IC c , I C P ,i )  N q ˆ ´C c Y ˆ o v( sq , it , C P , i ) (b7.1) Expression (b7.1) is country by country comparable because we use almost the same set of IC variables in all countries (with some slight modifications depending on the specific characteristics of each economy) and we apply the same methodology to select the set of significant IC variables. In addition, we can easily compute the O&P decompositions based on this demean part of productivity to do international comparisons of IC impacts on aggregate productivity. The share of aggregate log-productivity associated with the IC is in Table 4. Box 8. Evaluation of IC contributions to the means of the economic performance variables The objective now is to measure the partial direct effect of each IC variable on each dependent variable. For that purpose, we evaluate the impact of the average IC variable on the sample average values of the dependent variables (employment, wages, exports, FDI,). In what follows, we substitute all the unknown parameters from the system (3.5) to (3.9) by their corresponding 2SLS estimated values. The contributions are given by: ˆ ˆ  IC ˆ D   ˆ D ˆ D 100  100  L 100  IC 100  100 DR r 100 Ds j (b7.1) DM m lo g P lo g P lo g P lo g P lo g P ˆL ˆP log P ˆW logW  IC ˆIC C ˆC  D ˆDR  D ˆDs  D ˆDM 100  100  100 100  100  100  r 100  j 100  m 100 (b7.2) lo g L lo g L lo g L lo g L lo g L lo g L lo g L lo g L ˆ ˆ lo g P ˆ  IC ˆ C ˆ D ˆ D ˆ D 100  W 100  P 100  IC 100  C 100  DR r 100  Ds j 100  DM m 100 (b7.3) lo g W lo g W lo g W lo g W lo g W lo g W lo g W ˆ ˆ logP ˆ IC ˆ C ˆ D ˆ D ˆ D (b7.4) 100  100  100  100  100    100 . Exp P IC C DR r Ds j DM m ˆ ( Exp 0) P ˆ ( Exp 0) P ˆ ( Exp 0) P ˆ ( Exp 0) P ˆ ( Exp 0) P ˆ ( Exp 0) P ˆ ( Exp 0) P  ˆ FDI  ˆ P log P   IC ˆ IC  ˆCC   D ˆ DR   D ˆ Ds   D ˆ DM 100  100  100  100  100  r 100  j 100  m 100 . (b7.5) ˆ ( FDI 0) P ˆ ( FDI 0) P ˆ ( FDI 0) P ˆ ( FDI 0) P ˆ ( FDI 0) P ˆ ( FDI 0) P ˆ ( FDI 0) P The IC percentage contributions of equations (b7.1)-(b7.2) are in Table 6. 40 Box 9: Summary of econometric methodology for services sector The econometric analysis of the services sector is based on the concept of labour-productivity, instead of total factor productivity. In particular, the log of labour productivity is explained by the prices of the inputs, the investment climate (IC) variables and other control variables, including dummy variables for industry, region and size. The reduced form equation of labour productivity in terms of the input prices (w, r), investment climate (IC) variables and other control (C) variables is given by; Y  lo g     0   1 lo g w it   2 lo g rit   3 C i   5  IC i   4  D st   7  D jt   6  D m t   it (b9.1)  L  it where i=1,…, 150 an t=FY06 and FY07. Econometric issues: Most of the econometric issues commented and treated in the manufacturing probability model hold for the case of the services:  In the endogeneity of (5.1) we are treating the investment climate (IC) variables as observable fixed effects, as was suggested in Escribano and Guasch (2005). After controlling for observable fixed effects. The possible endogeneity of certain IC variables.is addressed with the corresponding region-industry average.  To control for heteroskedasticity we use robust standard errors allowing for clustering by sector and region.  In the cleaning of data and replacement of missing values we apply the same procedure than in the manufacturing sector.  The main difference with respect to the manufactures case is the way robustness is addressed here. We test robustness to the input prices used in (b8.1): wages, rental cost of capital; to the panel data dataset and to the cross- sectional dataset; and by using restricted and unrestricted by industry, region and industry/region input prices (we allow the elasticities of the input prices to vary by industry, region or both industry and region). Evaluation of IC effects on labor log productivity: In order to evaluate the contribution of each explanatory variable to labor-productivity we apply the O&P decomposition to the concept of labor productivity and we compute the contribution of each IC variable to the aggregate log labor productivity through average log labor productivity and allocative efficiency. IC elasticities and semi-elasticities estimates of equation (b9.1) are in Table 7. Box 10: Summary of econometric methodology for FY02-FY07 panel The evolution of the TFP between 1999 and 2006 shows an increase in the aggregate productivity of the manufacturing sector of Pakistan. In order to have additional insight on which are the IC factors associated with the increase in productivity observed, we use the information available for the 402 panel manufacturing firms. As we have already pointed out the information is far of being a random sample representative of the real population, so careful is required in the interpretation of the results. Concretely, we propose a model that relates the probability that productivity increase from 2001 to 2006 conditional on the investment climate establishments faced in 2001. We also control for productivity in 2001, and for region, size and sector. The population model is, P ( yit  1 | log Pit  5 ,IC it  5 ,Cit  5 , Drt  5 , D jt  5 , Dmt  5 ;  ,  IC ,  C ,  DS ,  DR ,  DM )  (b10.1)  Cit  5   DS  IC it  5   C G ( log Pit  5   IC  D jt  5   DM  Drt  5   DR  Dmt  5 ) As usual, in the estimation of (b9.1) we use the classical specifications for the unknown function G(.). Concretely, we select the set of significant IC variables going from the general to the specific with a LPM. Once we have selected the significant IC variables we use Logit and Probit models to test robustness. Most of the econometrics issues of the manufacturing econometric model also apply here. We address endogeneity of IC variables controlling for observable fixed effects and by using industry-region-size averages. The heteroskedasticity of the errors is addressed by using robust standard errors, allowing for clustering by region and size. There are some additional caveats on this model. We do not attempt to model causal relations between the IC and the increase of productivity. The advantage of the model is that we get rid of some simultaneous relations that may arise otherwise when we use contemporaneous explanatory variables. The estimation of IC effects on the probability of productivity increase according to equation (b10.1) is in Table 8. 41 ANNEX F: DEFINITIONS OF VARIABLES AND ECONOMETRIC RESULTS Table F1: General information on plant level and production function (productivity) variables for the analysis of manufacturing sector General Industrial a) Food Processing, Beverages, and Tobacco; b) Textile and Leather; c) information at classification Wood and Wood Products, other than Pulp; d) Pulp, Paper, and Printing; e) plant level Chemical, including Petrochemical; f) Non-Metallic Mineral Products; g) Iron, Steel, and Non-Ferrous Metal; h) Machinery; i) Miscellaneous Manufacturing Industries. Regional classification a) Punjab; b) Sindh; c) Baluchistan; d) NWFP. Size classification a) Small (< 25 employees); b) medium (>=25 & <100); c) large (>=100). Production Sales Used as the measure of output for the production function estimation. Sales function are defined as total annual sales. The series are deflated by using the variables Producer Price Indexes (PPI), base 2002. Employment Total number of permanent and temporal workers. Total hours worked Total number of employees multiplied by the average hours worked per per year year. Materials Total costs of intermediate and raw materials used in production (excluding fuel). The series are deflated by using the Producer Price Indexes (PPI), base 2002. Capital stock Net book value of machinery and equipment. The series are deflated by using the Producer Price Indexes (PPI), base 2002. User cost of capital The user cost of capital is defined in terms of the opportunity cost of using capital; it is defined as a 15% of the net book value of machinery and equipment. Labor cost Total expenditures on personnel. The series are deflated by using the Producer Price Indexes (PPI), base 2002. Dependent Exports Dummy variable that takes value 1 if exports are greater than 10%. variables in equation Foreign Direct Dummy variable that takes value 1 if any part of the capital of the firm is regressions Investment foreign. and linear Wages Real wage is defined as the total expenditures on personnel (deflated by probability using the Producer Price Indexes (PPI), base 2002.) divided by the total models number of permanent and temporal workers. Employment Total number of permanent and temporal workers. Source: Staff calculations with Pakistan ICS data. All series figure in US dollars, data obtained from WDI, The World Bank, 2008. 43 Table F2: General information on plant level and production function (productivity) variables for the analysis of services sector General Industrial classification a) Wholesale and retail trade; b) other services including IT, Information construction and transport. at Plant Level Regional classification a) Punjab; b) Sindh; c) Baluchistan; d) NWFP. Size classification a) Small (< 25 employees); b) medium (>=25 & <100); c) large (>=100). Labor Sales Used as the measure of output. Sales are defined as total annual productivity sales. The series are deflated by using the Producer Price Indexes variables (PPI), base 2002. Employment (demand Total number of permanent and temporal workers. for labor) Total hours worked per Total number of employees multiplied by the average hours worked year per year. Electricity cost Total annual costs of electricity. The series are deflated by using the Producer Price Indexes (PPI), base 2002. Communication cost Total annual costs of communications services. The series are deflated by using the Producer Price Index (PPI), base 2002. Rental cost of capital Total annual cost of rental of land/buildings, equipment, furniture. The series are deflated by using the Producer Price Indexes (PPI), base 2002. Labor cost Total annual cost of labor (including wages, salaries, bonuses, social payments). The series are deflated by using the Producer Price Index (PPI), base 2002. Source: Staff calculations with Pakistan ICS data. All series figure in US dollars, data obtained from WDI, The World Bank, 2008. Table F3: Definition of IC variables: infrastructure Name Definition Days to clear customs to export Average number of days to clear customs when exporting directly. Longest number of days to clear Longest number of days to clear customs when exporting directly. customs to export Days to clear customs to import Average number of days to clear customs when importing. Longest number of days to clear Longest number of days to clear customs when importing. customs to import Power outages Total number of power outages suffered by the plant in 2005. Average duration of power Average duration of power outages suffered in hours, conditional on the pant reports outages having power outages. Total duration of power outages Total duration of power outages suffered by the plant by month, in hours, by year conditional on the plat reports having power outages. Losses due to power outages Losses due to power outages as a percentage of total annual sales, conditional on the plant reports having power outages. Wait for a power supply Number of days waiting to obtain an electricity supply, conditional on submit an electrical connection. Dummy for gifts to obtain a Gifts expected or requested to obtain an electrical connection, conditional on submit power supply an electrical connection. Dummy for own generator Dummy variable taking value 1 if the firm has its own power generator. Electricity from a generator Percent. of the electricity used by the plat provided by a own generator. Dummy for insufficient water Dummy variable that takes value 1 if the firm has experienced Insufficient water supply supply for production during 2005. Water outages Total number of water outages suffered by the plant in 2005. Average duration of water Average duration of water outages suffered in hours, conditional on the plant reports outages having water outages. Total duration of water outages Total duration of water outages suffered by the plant by month in hours, conditional by year on the plant reports having water outages. 44 Name Definition Water from public sources Percentage of water supply from public sources. Wait for a water supply Number of days waiting for a water supply, conditional on submit a water supply. Dummy for gifts to obtain a Gifts expected or requested to obtain a water supply, conditional on submit a water water supply supply. Wait for a phone connection Number o days waiting to obtain a phone connection, conditional on submit a phone connection. Dummy for gifts to obtain a Gifts expected or requested to obtain a phone supply, conditional on submit a phone phone connection connection Dummy for webpage Dummy variable taking value 1 if the plant uses its own web page to communicate with clients and suppliers. Dummy for e-mail Dummy variable taking value 1 if the plant uses the electronic mail to communicate with clients and suppliers. Dummy for own transport Dummy variable taking value 1 if the plant uses its own transport to make shipments to its costumers. Products with own transport Percentage of shipments to costumers that were transported by the establishment own transport as a percentage of annual revenue. Conditional on the plant reports having it own transport. Shipment losses, exports Percentage of the consignment value of the products shipped for direct export lost while in transit because of theft, breakage or spoilage. Shipment losses, domestic Percentage of the consignment value of the products shipped for import lost while in transit because of theft, breakage or spoilage. Days of inventory of main input Average number of days (measured in production days) that the main input is available on stock. Wait for an import license Total days to obtain an import license, conditional on submit an import license. Dummy for gifts to obtain an Gifts expected or requested to obtain an import license, conditional on submit an import license import license. Dummy for industrial zone Dummy 1 if the plant is located in an industrial zone. Low quality supplies * Percentage of supplies that are of lower than agreed upon quality Sales lost due to delivery delays Percentage of sales lost due to delivery delays of key inputs. * Dummy for broadband internet Dummy variable taking value 1 if the establishment has a high-speed, broadband connection** Internet connection on its premises. Number of internet outages** Average number of times per month the establishment experienced unavailability of Internet connection. Average duration of internet In a typical month last fiscal year, average duration of unavailability of internet outages** connection. Dummy for security expenses Dummy variable taking value 1 if security of Internet connections or authentication for internet** of parties in a transaction affects the volume and/or nature of purchases that the establishment makes over the Internet. Source: Staff calculations with Pakistan ICS data. * Available only in 2002 ICS, ** Available only in services survey. 45 Table F4: Definition of IC variables: economic governance Name Definition Dummy for conflicts with Dummy taking value 1 if the plant has conflicts with clients with a third part involved. clients with a third part Dummy for conflicts with Dummy taking value 1 if the plant has conflicts with clients with a court involved clients with a court involved (conditional on having conflicts with clients with a third part involved). Weeks to judgment Number of weeks that took the court to come to judgment in the last conflict with clients (conditional on having conflicts with clients with a third part involved). Dummy for security expenses Dummy taking value 1 if the plant has security expenses. Security expenses Security expenses as a percentage of annual total sales. Dummy for crime losses Dummy taking value 1 if the plant has experienced losses due to criminal attempts in 2005. Crime losses Crime losses as a percentage of annual total sales in 2005. Manager's time spent in bur. In typical week percentage of manager's time spent dealing with bureaucratic issues. issues Weeks to bureaucracy Total number of weeks spent by management dealing with bureaucracy in 2005. Number of tax inspections Total number of inspections of tax officials received by the plant in 2005. Dummy for gifts in tax Gifts expected or requested in inspections with tax officials. inspections Dummy for payments to obtain Dummy that takes value 1 if firms operating in the same sector of the surveyed plant a contract with the government have to offer informal payments to obtain a contract with the government. Payments to obtain a contract Payments to obtain a contract with the government as a percentage of contract value. with the government Payments to deal with bur. Gifts or informal payments to public officials to “get things done” with regard to issues customs, taxes, licenses, regulations, services etc, as a percentage of annual sales. Sales reported to taxes Percentage of total annual sales that a typical firm operating in plant's sector reports for tax purposes. Workforce reported to taxes Percentage of total work force that a typical firm operating in plant's sector reports for tax purposes. Dummy for interventionist labor Dummy variable that takes value 1 if the labor regulation has affected plant's regulation employment decisions. Wait for a construction permit Days waiting to obtain a construction permit (conditional on submit a construction permit). Dummy for gifts to obtain a Gifts expected or requested to obtain a construction permit, conditional on submit a construction permit construction permit. Wait for an operation license Days waiting to obtain a main operating license (conditional on submit a operating license). Dummy for gifts to obtain an Gifts expected or requested to obtain an operating license, conditional on submit a operating license operating license. Dummy for tax exemption Dummy taking value 1 if the establishment is currently using or benefiting from: customs duty drawback, export rebate, sales tax refunds and profit tax exemption. Number of labor inspections Total number of inspections of labor officials received by the plant in 2005. Days of production lost due to Total number of production days lost due to worker absenteeism. absenteeism * Illegal payments in protection * Total amount of illegal payments to prevent violence (e.g. organized crime). Dummy for informal Dummy taking value 1 if the establishment competes against unregistered or informal competition** trading firms. Dummy for bribes to public Dummy variable taking value 1 if the establishment has paid any informal payment officials** (payment required/expected) to the police, political party, etc. (i.e. Bhatha) to ensure security to your establishment (protection from robbery, arson, or any other crime). Dummy for policy discussion Dummy taking value 1 if the establishment participated, directly or through its with government** representative body, in policy discussions with the local, provincial or federal government bodies during the last Fiscal year. Source: Staff calculations with Pakistan ICS data. * Available only in 2002 ICS, ** Available only in services survey. 46 Table F5: Definition of IC variables: finance Name of the variable Definition Purchases paid before delivery Percentage of annual purchases paid for before the delivery. Purchases paid on delivery Percentage of annual purchases paid for on delivery. Purchases paid after delivery Percentage of annual purchases paid for after the delivery. Sales paid before delivery Percentage of annual sales paid for before the delivery. Sales paid on delivery Percentage of annual sales paid for on delivery. Sales paid after delivery Percentage of annual sales paid for after the delivery. Working capital financed by Percentage of firm's working capital financed with internal funds. internal founds Working capital financed by Percentage of firm's working capital financed with funds from private commercial private banks banks. Working capital financed by Percentage of firm's working capital financed with funds state owned banks. state-owned banks Working capital financed by Percentage of firm's working capital financed with family/friends funds. family/friends Working capital financed bi Percentage of firm's working capital financed with funds from non-banking non-bank financial institutions financial institutions. Working capital financed by Percentage of firm's working capital financed with credits from suppliers. trade credit Working capital financed by Percentage of firm's working capital financed with funds from informal sources. informal founds New fixed assets financed by Percentage of investments in new fixed assets financed with internal funds. internal founds New fixed assets financed by Percentage of investments in new fixed assets financed with funds from private private banks commercial banks. New fixed assets financed by Percentage of investments in new fixed assets financed with funds state owned state-owned banks banks. New fixed assets financed by Percentage of investments in new fixed assets financed with family/friends funds. family/friends New fixed assets financed by Percentage of investments in new fixed assets financed with funds from non- non-bank financial institutions banking financial institutions. New fixed assets financed by Percentage of investments in new fixed assets financed with credits from suppliers. trade credit New fixed assets financed by Percentage of investments in new fixed assets financed with funds from informal informal founds sources. Dummy for checking or saving Dummy taking value 1 if the plant has a checking or saving account. account Own of the land Percentage of the lands in which the plant operates owned by the firm. Dummy for credit line Dummy that takes value 1 if the firm has access to a credit line or overdraft facility Dummy for loan Dummy that takes value 1 if the firm has access to a loan line. Dummy for loan with collateral Dummy that takes value 1 if the firm has access to a loan line with collateral (conditional on having a loan line). Value of the collateral Value of the collateral as a percentage of the loan value (conditional on having a loan with collateral) Dummy for debt Dummy taking value 1 if any of the number of rejected loan applications is larger than the number of applications for a loan. Dummy no loan because of Dummy that takes value 1 if the firm did not apply for loan because of its complexity complexity. Dummy no loan because of cost Dummy that takes value 1 if the firm did not apply for loan because of its cost. Dummy no loan because of Dummy that takes value 1 if the firm did not apply for loan because of its collateral. collateral Rejected credit applications Percentage of rejected credit applications. Accepted credit applications Percentage of accepted credit applications. Dummy for external auditory Dummy that takes value 1 if the firm has its annual statements externally audited. Dummy for trade association* Dummy 1 if the establishment belongs to any trade association. Percentage of credit line unused* Percentage of the credit line currently unused. Borrows in foreign currency* Percentage of establishment’s borrows denominated in foreign currency. 47 Name of the variable Definition Wait to clear a check* Days that it takes to clear a check with the establishment’s banking institution. Charges to clear a check* Charge to clear a check with establishment’s banking institution. Wait to clear a domestic Days that it takes to clear a domestic currency wire with establishment’s banking currency wire* institution. Charges to clear a domestic Charges to clear a domestic currency wire with establishment’s banking institution. currency wire* Wait to clear a foreign currency Days that it takes to clear a foreign currency wire with establishment’s banking wire* institution. Charges to clear a foreign Charge to clear a foreign currency wire with establishment’s banking institution. currency wire* Dummy for clear title for owned Dummy taking value 1 if out of the lands owned, the establishment has a clear title. land** * Available only in 2002 ICS, ** Available only in services survey. Source: Staff calculations with Pakistan ICS data. 48 Table F6: Definition of IC variables: innovation and competition Name Definition Dummy for quality certification Dummy taking value 1 if the firm has any kind of quality certification. Dummy for foreign technology Dummy taking value 1 if the plant uses technology licensed from a foreign-owned company. Dummy for product innovation Dummy taking value 1 if the plant has introduced any product innovation in the last 3 years. Dummy for process innovation Dummy taking value 1 if the plant has introduced any production process improvement in the last 3 years. Dummy for joint venture Dummy taking value 1 if the plant has agreed any new joint venture with a foreign company. Dummy for outsourcing Dummy taking value 1 if the plant subcontracts any part of the activity. Dummy for R&D Dummy that takes value 1 if the firm performed R&D activities during last year. Computer controlled machinery Percentage of plant’s machinery that is controlled by computer. Staff with computer Percentage of staff using computer at job. New equipment Percentage of plant’s equipment that is less that 5 years old. Dummy for FDI Dummy that takes value 1 if any part of firm's capital is foreign. Dummy for importer Dummy taking value 1 if the firm imports more than 10% of the total purchases of intermediate materials. Share of imports Share of imported inputs over total purchases of intermediate materials. Dummy for exporter Dummy taking value 1 if the firm exports more than 10% of the total annual sales. Exporting experience Number of years of exporting experience. Share of exports Share of exports over total annual sales. Dummy for local monopoly Dummy taking value one if the firm is a local monopoly. Dummy for less that Dummy taking value one if the plant has more or equal than 5 competitors in the local 5competitors market. Dummy for more than 5competitors Dummy taking value one if the plant has less than 5 competitors in the local market. Number of competitors* Total number of competitors within plant’s main product line. Dummy for new service Dummy taking value 1 if during the last 3 years the establishment introduced to the offered** market any new or improved service. Dummy for new methods of Dummy taking value 1 if during the last 3 years the establishment introduced to the providing services** market any new or improved methods of providing services. Dummy for investment in IT** Dummy taking value 1 if during the last year the establishment invested in information technologies. Source: Staff calculations with Pakistan ICS data. * Available only in 2002 ICS, ** Available only in services survey. Table F7: Definition of IC variables: labor markets and skills Name Definition Staff - production workers Percentage of production workers in staff. Staff - female workers Percentage of female workers in staff. Staff - skilled workers Percentage of skilled production workers in staff. Staff - university education Dummy taking value 1 if the typical production worker has at least one year of university education. Dummy for training Dummy taking value one if the firm provides formal (beyond on the job) training to its employees. Training to production workers Percentage of production workers receiving formal (beyond on the job) training Training to non-production workers Percentage of non-production workers receiving formal (beyond on the job) training University education* Manager experience in years. Tenure* Average tenure of the staff of the plant. Experience of the manager Number of years of experience of the manager in the establishment’s sector. Education of the manager Dummy taking value 1 if the manager has at least a bachelor degree. Education of the manager (post- Dummy taking value 1 if the manager has a post-grade (MA, PhD). grade) ** Source: Staff calculations with Pakistan ICS data. * Available only in 2002 ICS, ** Available only in services survey. 49 Table F8: Definition of IC variables: corporate governance Name Definition Largest shareholder Percentage of firm's capital owned by the largest shareholder. Dummy for incorporated company Dummy that takes value 1 if the firm is an incorporated company. Dummy for limited company Dummy that takes value 1 if the firm is a limited company. Dummy for state-owned firm Dummy that takes value 1 if any part of firm's capital is public. * Available only in 2002 ICS, ** Available only in services survey. Source: Staff calculations with Pakistan ICS data. Table F9: Definition of IC variables: other control variables Name Definition Age Age of the firm in 2005. Trade union Percentage of workforce unionized Capacity utilization Percentage of capacity utilized. Dummy for increased sales Dummy taking value 1 if the plant has increased its sales Dummy for decreased sales Dummy taking value 1 if the plant has decreased its sales Dummy for help from SMEDA Dummy 1 if the plant is receiving support from SMEDA Dummy for help from EPB Dummy 1 if the plant is receiving support from EPB Dummy for help from BOI Dummy 1 if the plant is receiving support from BOI Days of production lost due to Total number of working days lost due to strikes strikes* Days of production lost due to Total number of production days lost due to civil unrests. civil unrests* Dummy for materials from rural Dummy 1 if the direct purchase at the source is the mechanism of supply used for villages with direct purchase raw materials originated from rural villages. Dummy for materials from rural Dummy 1 if a regular supplier located in the rural area is the mechanism of villages with local supplier supply used for raw materials originated from rural villages. Dummy for materials from rural Dummy 1 if a regular supplier located establishment’s city is the mechanism of villages with supplier from supply used for raw materials originated from rural villages. Dummy for materials from rural Dummy 1 if the sub-contractual arrangement is the mechanism of supply used for villages with sub-contractual raw materials originated from rural villages. Materials from rural villages Percentage of the inputs used by the establishment originate from rural villages Local area in square feet Total area of the local occupied by the establishment in squared feet. Dummy for initiatives to address Dummy taking value 1 if the firm overcame any initiative to address AIDS among AIDS its employees during last fiscal year. Source: Staff calculations with Pakistan ICS data. * Available only in 2002 ICS, ** Available only in services survey. 50 Table F10: Robust IC elasticities and semi-elasticities with respect to productivity (manufacturing sector FY07) Two steps Single step estimation Solow residual Cobb-Douglas Translog Blocks of ICA variables Explanatory ICA variables Restricted Unrestricted Restricted Unrestricted Restricted Unrestricted Infrastructure Number of power outages (log)(b) -0.050* -0.050* -0.050* -0.036 -0.044* -0.026 Dummy for own generator (b) 0.071 0.058 0.108 0.092 0.095 0.074 Products with own transport 0.001 0.001* 0.001 0.002** 0.001* 0.001 (%)(b) Days of inventory of main 0.057** 0.060** 0.068*** 0.075*** 0.058** 0.065*** intermediate material (log)(b) Dummy for industrial zone 0.128* 0.131* 0.202** 0.199** 0.172** 0.159** Economic Dummy for conflicts with clients 0.543*** 0.501*** 0.542*** 0.352** 0.503*** 0.545*** governance with a court involved (b) Dummy for security expenses (b) 0.085* 0.084* 0.099** 0.081* 0.090* 0.105** Crime losses (%)(b) -0.005*** -0.005** -0.003 -0.003* -0.004** -0.004** Payments to obtain a contract -0.0002 -0.0002 -0.0003 -0.0001 -0.0003 -0.0004** with the government (b) Sales reported to taxes (%)(b) 0.001 0.002* 0.001 0.001* 0.001* 0.001 Dummy for gifts in tax -0.078 -0.079* -0.070 -0.042 -0.068 -0.025 inspections (b) Finance Purchases paid before delivery -0.002 -0.002 -0.002 -0.001 -0.002* -0.002* (%)(b) Working capital financed by -0.001 -0.001 -0.001 -0.002** -0.001 -0.002* internal founds (% (b) Working capital financed by 0.004* 0.004* 0.005** 0.002 0.004* 0.002 private banks (%)(b) Working capital financed by -0.013* -0.012* -0.011* -0.011* -0.010** -0.010 family/friends (%)(b) Working capital financed by -0.019 -0.018 -0.025** -0.026** -0.022* -0.025** informal founds (%)(b) Dummy for checking or saving 0.075 0.063 0.090 0.133** 0.085 0.111 account (b) Innovation and Dummy for process innovation (b) 0.300*** 0.291** 0.311** 0.265*** 0.310** 0.267** competition New equipment (%)(b) 0.001** 0.001** 0.001* 0.001 0.001* 0.001 Dummy for FDI (b) 0.191 0.238 0.260 0.267 0.343* 0.295 Labor markets Staff - female workers (%)(b) -0.005* -0.006* -0.006** -0.007*** -0.005* -0.005* and skills Dummy for training (b) 0.186* 0.179 0.262*** 0.192** 0.268*** 0.219* Corporate Largest shareholder (%)(b) -0.002** -0.002** -0.002** -0.002** -0.002** -0.001 governance Other control Dummy for help from BOI (b) 0.272** 0.263** 0.324*** 0.318*** 0.302** 0.363*** variables Dummy for materials from rural 0.144 0.108 0.217* 0.107 0.138 0.146 villages with local supplier (b) Observations 727 727 727 727 727 727 R-squared 0.22 0.21 0.92 0.93 0.93 0.94 Source: Staff calculations with Pakistan ICS data. Notes: (1) Two steps estimation: in the first step estimation of equation (b2.1) by non-parametric techniques to compute productivity (Solow residual), in the second step estimate (b2.2) and (b2.3) by OLS using as dependent variable the Solow residual from the first step, either restricted or unrestricted. (2) Single step estimation: estimae (b2.1), (b2.2) and (b2.3) in a single step by OLS, where (b2.1) can be a Cobb-Douglas Production function or a Translogarithmic. (3) Restricted: equal input output for all the establishments in the country. (4) Unrestricted: equal inoput-output elasticites for all the establishments in the same sector. (5) *significant at 10%; ** significant at 5%; *** significant at 1% given by robust standard errors corrected for correlation between cluster (industry and region). (6) Each regression includes a set of industry, size and region dummies and a constant term. (a) Variables instrumented with the industry-region-size average. (b) Variables approximated with a proxy (only missing values replaced by the industry-region-size average). 51 Table F.11: IC and other variables percentage contributions to aggregate log-productivity, FY07 (manufacturing) Aggregate log- Average log- Allocative productivity productivity efficiency Infrastructure Number of power outages (b) -2.42 -4.93 2.51 Dummy for own generator (b) 2.29 0.65 1.65 Products with own transport (b) 0.50 0.46 0.04 Days of inventory of main intermediate material (b) 8.99 5.60 3.39 Dummy for industrial zone 2.53 1.45 1.08 Economic Dummy for conflicts with clients with a court involved 2.18 0.55 1.63 governance (b) Dummy for security expenses (b) 2.50 1.61 0.89 Crime losses (b) -0.02 -0.10 0.09 Payments to obtain a contract with the government (b) 0.00 -0.07 0.06 Demean log-productivity Sales reported to taxes (b) 4.61 4.44 0.17 Dummy for gifts in tax inspections (b) -0.37 -0.58 0.21 Finance Purchases paid before delivery (b) -0.19 -0.50 0.31 Working capital financed by internal founds (b) -2.56 -3.25 0.69 Working capital financed by private banks (b) 4.05 0.61 3.44 Working capital financed by family/friends (b) 0.00 -0.06 0.06 Working capital financed by informal founds (b) 0.00 -0.02 0.02 Dummy for checking or saving account (b) 2.52 1.91 0.61 Innovation and Dummy for process innovation (b) 5.67 0.93 4.74 competition New equipment (b) 2.66 1.01 1.65 Dummy for FDI (b) 1.85 0.15 1.70 Labor markets Staff - female workers (b) -1.24 -0.33 -0.91 and skills Dummy for training (b) 4.04 0.44 3.60 Corporate Largest shareholder (b) -2.90 -5.84 2.94 governance Other control Dummy for help from BOI (b) 0.97 0.40 0.57 variables Dummy for materials from rural villages with local 0.07 0.17 -0.10 supplier (b) Total contribution of IC (demean log-productivity) 35.75 4.71 31.05 Other stuff Industry/region/size controls -13.10 -4.40 -8.70 Constant term 41.01 41.01 0.00 Residual 36.34 0.00 36.34 Total contribution of other stuff 64.25 36.61 27.64 Total 100.00 41.32 58.68 Source: Staff calculations with Pakistan ICS data. Notes: Results from equation (b6.3). The contribution of IC to aggregate log-productivity is equal to the sum of the contributions to average log-productivity and to the allocative efficiency. Demean log-productivity is the part of productivity associated with the investment climate The productivity measure used is the restricted Solow residual. (a) Variables instrumented with the industry-region-size average. (b) Variables approximated with a proxy (only missing values replaced by the industry-region-size average). 52 Table F12: IC elasticities and semi-elasticities on economic performance measures, 2SLS estimation, FY07 (manufacturing sector) Exporting FDI Productiv. Employm. Wages probability probability Productivity 0.264** 0.823*** 0.106*** Real wages -0.157*** Infrastructure Days to clear customs to export (log) -0.067 (a) Days to clear customs for exports - interaction with -0.159* (a) -0.018** (a) firms that do export (log) Number of power outages (log) -0.05 (b) Losses due to power outages (%) -0.020* (a) Dummy for own generator 0.071 (b) Electricity from a generator (%) 0.003* (b) 0.002** (b) Dummy for insufficient water supply -0.195* Number of water outages (log) -0.008* (b) Products with own transport (%) 0.001 (b) 0.003*** (b) -0.002* (b) 0.001*** (b) -0.0003** (b) Shipment losses, domestic (%) -0.018** (a) Shipment losses, exports (%) -0.023*** (a) Days of inventory of main intermediate material (log) 0.057 (b) 0.099*** (b) Dummy for industrial zone 0.128 Economic Dummy for conflicts with clients with a court 0.543 (b) governance involved Dummy for security expenses 0.085 (b) 0.396*** (b) Security expenses (%) 0.007 (b) 0.001 (b) 0.007** (a) Dummy for crime losses -0.239* (b) Crime losses (%) -0.005 (b) -0.007** (b) -0.001* (b) Payments to obtain a contract with the government (%) -0.0002 (b) -0.0004** (b) Sales reported to taxes (%) 0.001 (b) Dummy for gifts in tax inspections -0.078 (b) Payments to deal with bur. issues (%) 0.001***(b) Number of inspections (log) -0.046** (b) Number of labor inspections (log) -0.011** (b) Dummy for tax exemption 0.174*** (b) Finance Purchases paid before delivery (%) -0.002 (b) -0.003* (b) -0.0004** (b) Sales paid after delivery (%) 0.002* (b) Working capital financed by internal founds (%) -0.001 (b) 0.001* (b) Working capital financed by private banks (%) 0.004 (b) -0.011* (b) 0.001* (b) Working capital financed by state-owned banks (%) 0.010** (b) 0.009*** (b) Working capital financed by family/friends (%) -0.013 (b) -0.025*** (b) Working capital financed by informal founds (%) -0.019 (b) Dummy for checking or saving account 0.075 (b) 0.159*** (b) 0.145* (b) Dummy for credit line 0.167* 0.378*** (b) Own land (%) 0.002*** (b) 0.001** (b) Dummy for external auditory 0.331*** 0.204 (b) Innovation and Dummy for quality certification 0.267** (b) 0.184* (b) 0.157*** (b) 0.032* (b) competition Dummy for foreign technology 0.148*** (b) Dummy for product innovation 0.221* (b) Dummy for process innovation 0.3 (b) Computer controlled machinery (%) 0.007* (b) New equipment (%) 0.001 (b) Staff with computer (%) 0.004* (b) 0.005* (b) 0.001* (b) Dummy for e-mail 0.236** 0.358* (b) 0.118*** (b) Dummy for joint venture 0.182* (b) Dummy for outsourcing -0.206** (b) Dummy for FDI 0.191 (b) Exporting experience (log) 0.166*** (b) Dummy for more than 5 competitors 0.223** (b) 0.037* (b) 53 Exporting FDI Productiv. Employm. Wages probability probability Labor markets Staff - production workers (%) 0.003*** (b) and skills Staff - female workers (%) -0.005 (b) 0.015*** (b) Staff - skilled workers (%) 0.005*** (b) Dummy for training 0.186 (b) Training to non-production workers (%) 0.006** (b) 0.003* (a) Experience of the manager (log) 0.082** (b) Education of the manager 0.288*** (b) Corporate Largest shareholder (%) -0.002 (b) governance Dummy for incorporated company 1.267*** (b) 0.173** Dummy for limited company 0.550*** Dummy for state-owned firm -0.194* Other control Trade union (%) 0.005** (b) -0.001* (b) variables Capacity utilization (%) 0.003** (b) Dummy for help from EPB 0.244** (b) 0.176** (b) Dummy for help from BOI 0.272 (b) Dummy for materials from rural villages with local 0.144 (b) supplier Dummy for materials from rural villages with supplier -0.431*** (b) from firm's city Dummy for materials from rural villages with sub- 0.393 (b) -0.131* (b) 0.293* (b) contractual arrangement Observations 727 690 724 673 725 R-squared 0.22 Instruments First stage R-squared1 0.35 0.20 0.35 0.35 evaluation Partial R-squared: productivity2 0.21 0.05 0.22 0.29 Partial R-squared F test (p-value)3 0.00 0.00 0.00 0.00 Hansen test (p-value)4 0.69 0.78 0.88 0.28 Source: Staff calculations with Pakistan ICS data. Notes: Estimation by 2SLS of the system (b2.1)-(b2.5). * significant at 10%; ** significant at 5%; *** significant at 1% (robust standard errors corrected for clustering by industry and region). Significance of IC variables in productivity equation is in Table 3. The productivity measure used is the restricted Solow residual. Robustness was tested to other productivity measures. Each regression includes a set of industry, region and size dummies and a constant term. (a) Variables instrumented with the industry-region-size average. (b) Variables approximated with a proxy (only missing values replaced by the industry-region-size average). 1 Productivity, real wages, probability of exporting and probability of receiving FDI are endogenous and the list of variables used as excluded instruments comes from the list of explanatory variables from their corresponding equations. 2 First stage R-squared from the regression of productivity on both the included and the excluded instruments. 3 The partial R-squared measures the squared partial correlation between the excluded instruments and the productivity. 4 F-test of joint significance of the excluded instruments that corresponds to the partial R-squared. 5 The Hansen test is a test of overidentifying 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. 54 Table 6.13: IC percentage contributions to the means of economic performance measures, FY07 (manufacturing sector) Exporting FDI Productiv. Employm. Wages probability probability Productivity 16.5 94.1 46.6 174.6 Real wages -51.4 Infrastructure Days to clear customs to export (log) -38.0 Days to clear customs for exports - -1.9 -24.8 interaction with firms that do export (log) Number of power outages (log) -104.5 Losses due to power outages (%) -14.9 Dummy for own generator 13.7 2.4 Electricity from a generator (%) 0.9 Dummy for insufficient water supply -2.1 Number of water outages (log) -19.0 Products with own transport (%) 9.8 1.9 -2.9 3.9 -28.7 Shipment losses, domestic (%) -2.1 Shipment losses, exports (%) -0.5 Days of inventory of main intermediate 118.7 13.6 material (log) Dummy for industrial zone 30.7 Economic Dummy for conflicts with clients with 11.7 -2.5 governance a court involved Dummy for security expenses 34.2 11.8 1.0 77.6 Security expenses (%) 1.5 Dummy for crime losses -0.3 Crime losses (%) -2.2 -2.6 Payments to obtain a contract with the -1.4 -57.4 government (%) Sales reported to taxes (%) 94.0 Dummy for gifts in tax inspections -12.3 Payments to deal with bur. issues (%) 8.7 Number of inspections (log) -9.5 -34.7 Number of labor inspections (log) Dummy for tax exemption 12.2 Finance Purchases paid before delivery (%) -10.6 -2.0 -17.2 Sales paid after delivery (%) 4.4 Working capital financed by internal -69.0 24.5 founds (%) Working capital financed by private 12.9 -5.8 18.0 banks (%) Working capital financed by state- 0.3 1.8 owned banks (%) Working capital financed by -1.3 -0.3 family/friends (%) Working capital financed by informal -0.4 founds (%) Dummy for checking or saving account 40.5 5.8 9.7 Dummy for credit line 1.7 8.7 Own land (%) 8.4 13.4 Dummy for external auditory 4.2 6.2 55 Exporting FDI Productiv. Employm. Wages probability probability Innovation Dummy for quality certification 2.3 3.8 7.2 38.8 and Dummy for foreign technology 49.0 competition Dummy for product innovation 2.8 Dummy for process innovation 19.7 Computer controlled machinery (%) 1.3 New equipment (%) 21.4 37.2 Staff with computer (%) 1.0 2.9 Dummy for e-mail 3.8 10.8 Dummy for joint venture 0.6 Dummy for outsourcing -33.6 Dummy for FDI 3.2 0.7 Exporting experience (log) 3.0 Dummy for more than 5 competitors 9.7 11.8 Labor Staff - production workers (%) 11.4 markets and Staff - female workers (%) -6.9 1.6 skills Staff - skilled workers (%) 14.9 Dummy for training 9.4 Training to non-production workers 0.6 45.2 (%) Experience of the manager (log) 12.0 Education of the manager 6.2 Corporate Largest shareholder (%) -123.5 governance Dummy for incorporated company 1.6 40.0 Dummy for limited company 4.2 Dummy for state-owned firm -0.3 Other control Trade union (%) 0.7 -10.2 variables Capacity utilization (%) 12.5 Dummy for help from EPB 0.8 3.3 Dummy for help from BOI 8.6 Dummy for materials from rural 3.7 villages with local supplier Dummy for materials from rural -1.0 villages with supplier from firm's city Dummy for materials from rural 0.2 -0.4 villages with sub-contractual arrangement Source: Staff calculations with Pakistan ICS data. Notes: Percentage contributions compute according to equations (b7.1)-(b7.5). IC elasticities and semi-elasticities used are those from Table 5. The productivity measure used is the restricted Residual. 56 Table F.14: IC elasticities and semi-elasticities with respect to labor productivity, FY07 (services sector) Restricted estimation Unrestricted estimation Unrestricted Dependent variable: log of labor Panel data Cross-sectional Unrestricted Unrestricted by industry productivity log(sales/employment)] 2004-2005 data 2005 by industry by region and region Input prices Log of real wages 0.400(c) 0.407(c) 0.393(c) 0.389(c) Log of rental cost of capital 0.06 0.055 Infrastructures Number of power outages -0.806(b) -0.799(b) -0.839(b) -0.821(b) -0.796(a) -0.792(a) -0.709 (log)(e) Dummy for own generator 0.359(a) 0.352(a) 0.33 0.342(a) 0.356(a) 0.302(a) 0.3 (e) Red tape, Dummy for bribes from -0.268 -0.288(a) -0.35(a) -0.375(a) -0.297 -0.248 -0.284 corruption and public officials (e) crime Manager's time spent in -0.046(b) -0.048(b) -0.038(a) -0.041(a) -0.045(b) -0.042(b) -0.042(b) bur. issues (%)(e) Number of inspections -0.237(a) -0.23 -0.216 -0.203 -0.221 -0.233(a) -0.221 (log)(e) Sales reported to taxes -0.002 -0.002 -0.001 -0.001 -0.001 -0.002 -0.001 (%)(e) Dummy for informal 0.459(b) 0.486(b) 0.380(a) 0.410(b) 0.428(b) 0.423(b) 0.382(b) competition (e) Finance and Working capital financed -0.027(c) -0.026(b) -0.026(b) -0.025(b) -0.028(c) -0.027(c) -0.029(c) corporate by trade credit (%)(e) governance Working capital financed 0.008 0.011(a) 0.004 0.009 0.006 0.001 -0.001 by family/friends (%)(e) Owner of the lands (%)(e) -0.003 -0.003 -0.002 -0.002 -0.003 -0.003 -0.003 Dummy for loan (e) 0.743 0.806 0.804 0.821(a) 0.831 0.988(a) 1.096(a) Value of the collateral -0.004(b) -0.005(b) -0.004(b) -0.004(b) -0.004 -0.006(b) -0.007(b) (%)(e) Innovation and Dummy for quality 0.412(a) 0.445 0.419 0.458(a) 0.415 0.521 0.537 competition certification (e) Staff with computer (%)(e) 0.008 0.008 0.008(a) 0.008 0.008(a) 0.006 0.006 Exporting experience 0.357 0.381(a) 0.331 0.376(a) 0.315 0.422(b) 0.385(b) (log)(e) Other control Area of the local in square 0.136(a) 0.152(a) 0.123(a) 0.137(b) 0.127(a) 0.133(a) 0.124(a) variables feet (log)(e) Observations 277 277 143 143 277 277 277 R-squared 0.63 0.62 0.64 0.64 0.63 0.65 0.65 Source: Staff calculations with Pakistan ICS data. Notes: Estimation of equation (b9.1) Restricted: equal input prices for all the establishments in the country. Panel data estimation uses data for FY06 and FY07. Cross-sectional estimation uses data for FY07. Unrestricted: equal input prices for all the establishments in the same sector, region or region/sector. (a)significant at 10%; (b) significant at 5%; (c) significant at 1% given by robust standard errors corrected for correlation between cluster (industry and region). Each regression includes a set of industry, size and region dummies and a constant term (also a set of time dummies in the panel data estimation). (d) Variables instrumented with the industry-region-size average. (e) Variables approximated with a proxy (only missing values replaced by the industry-region-size average). 57 Table F.15: IC effects on the probability of productivity increase between FY02 and FY07 Linear Logit Model Probit Model Dependent variable: probability of productivity Probability Marginal increase Model Coeffs. Odd ratios Coeffs. effects Infrastructure Days to clear customs for imports - -0.043 -0.23 0.795 -0.13(a) -0.049(b) interaction with firms that do import (log)(d) Total number of power outages -0.044(a) -0.282(b) 0.754(b) -0.170(b) -0.064(b) (log)(e) Low quality supplies (%)(d) -0.009(a) -0.043(b) 0.958(b) -0.026(b) -0.010(b) Economic Security expenses (%)(e) 0.018 0.091(a) 1.095(a) 0.055 0.021 governance Crime losses (%)(e) -0.024(c) -0.144(b) 0.866(b) -0.091(c) -0.034(c) Manager's time spent in bur. -0.003(a) -0.014 0.986 -0.008 -0.003 Issues(%) Total number of inspections -0.118(a) -0.516 0.597 -0.335(a) -0.126(a) (log)(d) Payments to deal with bur. issues 0.011(b) 0.119(b) 1.127(b) 0.073(b) 0.027(b) (%)(e) Total number of labor inspections 0.059(c) 0.305(c) 1.356(c) 0.181(c) 0.068(c) (log)(e) Illegal payments in protection -0.044(a) -0.246(a) 0.782(a) -0.147(a) -0.055(a) (%)(e) Finance Dummy for credit line 0.075 0.407 1.503 0.263(a) 0.096(a) Dummy for loan 0.165(b) 1.167 3.213 0.656(a) 0.225(a) Dummy for loan with collateral -0.263(c) -1.630(a) 0.196(a) -0.934(b) -0.359(b) Dummy for trade association 0.072 0.4 1.492 0.249(a) 0.094(a) Labor Staff - female workers (%)(e) -0.009(c) -0.042(b) 0.959(b) -0.027(c) -0.010(c) markets Staff - university education (%)(e) 0.194 1.312 3.715 0.787 0.239 and skills Training to non-production workers 0.149(a) 0.715 2.045 0.392 0.148 (%)(d) Corporate Dummy for state-owned firm 0.304(b) 1.444(a) 4.239(a) 0.894(a) 0.261(a) governance R-squared/Pseudo R-squared 0.17 0.14 0.14 0.14 0.14 Observations 352 352 352 352 352 Source: Staff calculations with Pakistan ICS data. Notes: stimation of equation (b10.1) (a)significant at 10%; (b) significant at 5%; (c) significant at 1% given by robust standard errors corrected for correlation between cluster (industry and region). Each regression includes a set of industry, size and region dummies and a constant term. (d) Variables instrumented with the industry-region-size average. (e) Variables approximated with a proxy (only missing values replaced by the industry-region-size average). 58