Report No. 43737-PE PERU Trajectories towards Formality JUNE 16, 2008 Poverty Reduction and Economic Management Bolivia, Ecuador, Peru and Venezuela Country Management Unit Latin America and the Caribbean Region Document of the World Bank REPUBLIC OF PERU FISCAL YEAR January 1 to December 31 CURRENCY EQUIVALENTS Currency Unit = Soles 1 US Dollar = S/. 2.85 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS AAA Analytical and Advisory Activities LAC Latin America and Caribbean Region BCRP Central Reserve Bank of Peru (Banco MSEs Micro and Small Enterprises Central de Reserva del Perú) MEF Ministry of Economy and Finance BONOPYME Training and technical assistance METR Marginal Effective Tax Rate vouchers MTPE Ministry of Labor and Employment CIT Corporate Income Tax Promotion (Ministerio del Trabajo y CNC National Competitiveness Council Promoción del Empleo) (advisory body to the PCM) MYPEs Micro and Small Enterprises (Micro y COFIDE Development Financial Corporation Pequeñas Empresas) (Corporación Financiera de Desarrollo NRS National Registry of Suppliers to the State SA) (Registro Nacional de Proveedores) COFOPRI Comission to Formalize Informal OLS Ordinary Least Squares Property (Comisión de Formalización de PCM Office of the Prime Minister (Presidencia la Propiedad Informal) del Consejo de Ministros) CONFIEP Peruvian Private Business Association PRODAME Self-Employment and Micro enterprise (Confederación Nacional de Promotion Program (Programa de Instituciones Empresariales Privadas) Autoempleo y Microempresa) CONSUCODE Council of State Procurement (Consejo PRODUCE Ministry of Production (Ministerio de la Superior de Contrataciones y Producción) Adquisiciones del Estado) PROMPEX Export Promotion Agency (Comisión para la CTS Compensation for time in service Promoción de Exportaciones) (compensación por tiempo de servicio) PROMPYME Comission for Promotion of the Small and DNMYPE National Directorate for Micro and Micro Enterprise (Comisión para la Small Companies (Dirección Nacional Promoción de la Pequeña y Micro Empresa) de la Micro y Pequeña Empresa) RER Special Tax Regime (Régimen Especial de ENAHO National Household Survey (Encuesta la Renta) Nacional de Hogares) RUC Tax identification Lumber (Registro Único FIAS Foreign Investment Advisory Service de Contribuyente) GDP Gross Domestic Product RUS Simplified Single Regime (Régimen Unico GOP Government of Peru Simplificado) GNP Gross National Product SBS Superintendency of Bank and Insurance ICA Investment Climate Assessment (Superintendencia de Banca y Seguros) IDB Inter-American Development Bank SMEs Small and Medium Enterprises IFC International Finance Corporation SUNARP National Registry (Superintendencia IGV General sales tax (Impuesto General a Nacional de las Ventas) Registros Públicos) ILO International Labor Organization SUNAT Government Tax Agency (Superintendencia IMF International Monetary Fund Nacional de Administración Tributaria) INDECOPI National Institute for Consumer and UIT Income Tax Unit Intellectual Property Defense VAT Value Added Tax INEI National Statistics Institute (Instituto Nacional de Estadística e Informática) Vice President, LCR: Pamela Cox Director, LCC6C: C. Felipe Jaramillo Director, LCSPR: Marcelo Giugale Sector Director, LCSPE: Rodrigo Chaves Sector Leader, LCSPE: Carlos Silva-Jauregui Task Team Leader: Oscar Calvo-Gonzalez / Rossana Polastri ACKNOWLEDGMENTS This report was prepared by a team led by Oscar Calvo-Gonzalez (LCSPE) and Rossana Polastri (Senior Country Economist, LCSPE) under the overall supervision and guidance of Rodrigo Chaves (Sector Manager, LCSPE) and Carlos Siva-Jauregui (Lead Economist and Sector Leader, LCSPR). C. Felipe Jaramillo (Country Director, LCC6C) linked the team to the Bank’s overall strategy and steered them in that direction. Mauricio Carrizosa (Adviser, IEGCR) and Vicente Fretes-Cibils (IDB) provided initial guidance on the study. The team also included Luis Barrantes (LCSPE), Sebastian James (CICRS), Jan Loeprick (CICKM), David McKenzie (Senior Economist, DECRG), Rashmi Shankar (LCSPE), and Rich Stern (Program Coordinator for Business Taxation, FIAS). Contributions were also received from Cesar Baldeon (CFOMR) and Gladys Triveño (Proexpansión). Ipsos APOYO Opinión y Mercado conducted the survey of micro and small businesses and CONECTA carried out the focus groups and in-depth interviews with micro and small entrepreneurs. Maria Ines Thorne provided valuable logistical and production support. The peer reviewers for the report were Alvaro Escribano (Universidad Carlos III), Pablo Fajnzylber (Senior Economist, LCRCE), Luke Haggarty (General Manager, CLALA), Vincent Palmade (Lead Economist, AFTFP), and Jamele Rigolini (Economist, EASHD). The report also benefited from substantive comments from a variety of people during various stages of this project, including Jacqueline Coolidge (Lead Policy Investment Officer, CICAF), Gladys Lopez-Acevedo (Senior Economist, LCSPP), Juan Carlos Mendoza (Senior Financial Economist, LCSPF), and Ian Walker (Lead Social Protection Specialist, LCSHS). The team received valuable guidance through meetings with various authorities. Important insights and contributions were gained through interviews with a wide variety of individuals and groups representing officials of the Peruvian Government, institutions, and private sector (micro and small) firms. Particular thanks to the entrepreneurs who generously granted us their time through numerous interviews. The team would like to thank the Peruvian authorities for their continued cooperation, especially the support of the Ministry of Economy and Finance (MEF), the Ministry of Labor (MTPE), the Ministry of Production (PRODUCE), the National Institute of Statistics (INEI), the Superintendence of Tax Administration (SUNAT), the Superintendence of Banks and Insurance (SBS), the National Council of Competitiveness (CNC), and the Central Bank of Reserve (BCRP). 3 4 Trajectories towards Formality in Peru Table of contents Executive summary and policy implications ...................................................................... 9 1. Introduction .............................................................................................................. 19 Informality in the sectors selected for the survey .................................................... 26 2. Determinants of informality..................................................................................... 31 A. Who is our source of information? ..................................................................... 32 B. What do micro and small business owners tell us? ............................................. 35 C. Econometric analysis of the determinants of informality ................................... 61 D. Dynamics out of informality ............................................................................... 65 3. The impact of informality ........................................................................................ 69 A. Is informality a cause of low profitability? ......................................................... 69 B. What is the impact of informality on access to credit? ....................................... 73 List of references............................................................................................................... 78 Annex A. Measures of informality based on the ENAHO using different definitions . 80 Annex B. Econometric results on the choice of tax regime and determinants of tax inspections ....................................................................................................................... 98 Annex C. Econometric results on the determinants of informality ............................. 103 Annex D. The impact of informality – methodology and impact on profits and on access to credit ................................................................................................................ 104 Annex E. Trajectories towards formality .................................................................... 114 Annex F. Database of sector-specific labor regulations - methodology ..................... 115 Annex G. The Marginal Effective Tax Rate Methodology ......................................... 116 Annex H. Data for calculating the marginal effective tax rate .................................... 125 List of boxes Box 1: What do we mean by informality? ........................................................................ 19 Box 2: A dedicated survey to study informality in Peru ................................................... 24 Box 3: Understanding the incentives of tax regimes – the Marginal Effective Tax Rate of the different corporate income tax regimes in Peru .......................................................... 44 Box 4: What do businesses know about the Special Labor Regime for micro firms?...... 51 Box 5: Severance payments de jure and de facto ............................................................. 56 Box 6: Sector-specific labor regulations and the prevalence of informality .................... 58 Box 7: Case studies – formality is also a ‘refuge’ ............................................................ 65 5 List of tables Table 1.1: Measures of informality based on the household survey in the sectors selected for the survey and in the economy .................................................................................... 26 Table 2.1: Number of businesses surveyed....................................................................... 32 Table 2.2: Selected informality characteristics of surveyed businesses ........................... 33 Table 2.3: Benefits and costs of formalizing perceived by micro and small business owners ............................................................................................................................... 35 Table 2.4: Reasons for not having a municipal license .................................................... 39 Table 2.5: Overview of corporate income taxation .......................................................... 41 Table 2.6: Results from econometric analyses of the choice of tax regime ..................... 47 Table 2.7: Definitions of micro and small enterprises in the labor code and correspondence with tax regulations ................................................................................. 48 Table 2.8: Business owners’ estimates of how common is for workers to earn less than the minimum wage and not being on the payroll .............................................................. 54 Table 2.9: Reasons for self-employment .......................................................................... 56 Table 2.10: Variables that help predict having a municipal license and an RUC number 62 Table 2.11: Variables that help predict having both a municipal license and an RUC number .............................................................................................................................. 63 Table 2.12: Summary of steps towards formalization ...................................................... 67 Table 3.1: Impact of different formality characteristics on firm profits ........................... 72 Table A.1: Special fiscal regimens for small taxpayers in Latin America ....................... 88 Table A.2: Regulatory labor burden in comparative perspective ..................................... 88 Table A.3: Main informality characteristics among surveyed businesses, by city ........... 89 Table A.4: Main informality characteristics among surveyed businesses, by sector ....... 89 Table A.5: Most important problem affecting micro and business owners ...................... 90 Table A.6: Obstacles for micro and small businesses - detailed questionnaire ................ 91 Table A.7: Sales of micro and small businesses surveyed, by customer type and business size .................................................................................................................................... 92 Table A.8: Elements of formality ..................................................................................... 92 Table A.9: Factors affecting the choice of tax regime ...................................................... 93 Table A.10: Participation in state-supported programs .................................................... 94 Table A.11: Key factors behind the decision not to have a license, by business size ...... 95 Table A.12: Knowledge of where to formalize ................................................................ 95 Table A.13: Share of workers earning less than minimum wage, by business size ......... 96 Table A.14: Reasons for not paying the minimum wage, by business size ...................... 96 Table A.15: Percent of workers not on the payroll, by business size ............................... 97 Table A.16: Reasons for not having workers on payroll, by business size ...................... 97 Table B.1: Choice of tax regime - probit regression ......................................................... 98 Table B.2: Choice of tax regime - multinomial logit (general regime) ............................ 99 Table B.3: Choice of tax regime - multinomial logit (RUS) .......................................... 100 Table B.4: Choice of tax regime - multinomial logit (RER) .......................................... 101 Table B.5: Probit regression on the determinants of tax inspections .............................. 102 Table D.1: "Impact" of having an RUC on log profits ................................................... 105 Table D.2: "Impact" of having a municipal license on log profits ................................. 106 Table D.3: OLS Regression of log profits on both municipal license and an RUC number ......................................................................................................................................... 107 6 Table D.4: Businesses with and without bank accounts ................................................. 108 Table D.5: Sources of finance among businesses that have borrowed in 2007 or 2006. 108 Table D.6: Use of credit to purchase inputs .................................................................... 109 Table D.7: Sources of start-up capital............................................................................. 110 Table D.8: Borrowing by sector...................................................................................... 110 Table D.9: Sources of finance for most recent loan........................................................ 111 Table D.10: Terms of loan (annual or monthly rates) by source of finance ................... 111 Table D.11: Average length of loan in months ............................................................... 112 Table D.12: Home ownership - percent of entrepreneur owning their home ................. 112 Table D.13: Business taking credit cards and checks, percent ....................................... 112 Table D.14: Impact on getting a loan of having a license and RUC number ................. 113 Table E.1: Time lapsed from opening business to getting RUC number and municipal license ............................................................................................................................. 114 List of figures Figure 1.1: Informality in Peru and other countries in Latin America and the Caribbean 21 Figure 1.2: Informality by firm size, selected sectors and economy-wide ....................... 27 Figure 1.3: Map of informality prevalence by region ....................................................... 28 Figure 1.4: Informality prevalence by region and income per capita ............................... 29 Figure 1.5: Informality rate and economic growth, 2003 to 2006 .................................... 30 Figure 2.1: Share of surveyed businesses at different degrees of formality ..................... 34 Figure 2.2: Benefits from having a municipal license and reasons not to have one ......... 38 Figure 2.3: Benefits from having an RUC number and reasons not to have one ............. 40 Figure 2.4: Distribution by turnover of businesses filing under different tax regimes ..... 43 Figure 2.5: Relationship between the number of workers and those with health coverage and pension ....................................................................................................................... 50 Figure 2.6: Kernel distributions of monthly income for private salaried workers – Metropolitan Lima ............................................................................................................ 53 Figure 2.7: Minimum wage in Peru and selected countries in the region......................... 53 Figure 2.8: Informality prevalence for workers, by pay system (piece-rate or not) ......... 55 Figure 2.9: Formality characteristics among surveyed businesses, by age of business .... 68 Figure 3.1: Difference in profitability between the business with a municipal license and those without ..................................................................................................................... 70 Figure 3.2: Difference in profitability between the business with an RUC number and those without ..................................................................................................................... 71 Figure 3.3: Financial depth ............................................................................................... 73 Figure A.1: Calculating the productive definition of informality based on the ENAHO . 80 Figure A.2: Map of informality prevalence by region ...................................................... 81 Figure A.3: Map of informality prevalence by region ...................................................... 82 Figure A.4: Map of informality prevalence by region ...................................................... 83 Figure A.5: Informality prevalence by region and income per capita .............................. 84 Figure A.6: Informality prevalence by region and income per capita .............................. 85 Figure A.7: Informality prevalence by region and income per capita .............................. 86 Figure A.8: Informality rate and economic growth, 2003 to 2006 ................................... 86 Figure A.9: Informality rate and economic growth, 2003 to 2006 ................................... 87 Figure A.10: Informality rate and economic growth, 2003 to 2006 ................................. 87 7 8 Executive summary and policy implications Context ES. 1. Informality in Peru is high compared to other countries in the Latin America and the Caribbean region. Measuring informality is notoriously difficult, in large part because informality is not easily defined. Alternative measures of informality include a so-called ‘productive’ definition focused on the characteristics of the productive unit and a ‘legalistic’ definition focused on the coverage of workers by mandated social protection. According to the so-called productive definition of informality, 76 percent of the economically active population in Peru is informal (the third highest share in the Latin America and the Caribbean region). Taking the legalistic definition of informality, around 86 percent of the economically active population lack access to a pension, 86 percent work without a contract, and 80 percent lack healthcare insurance. Peru reports the third-highest rate of informality in Latin America and the Caribbean as measured by the percent of workers without a contract and without health insurance, while Peru is in the middle of the table in the region with regard to access to a pension. ES. 2. Addressing informality is a priority in the policy agenda of the Government of Peru. The Peruvian economy has enjoyed a relatively long period of stable and sustained growth. Poverty numbers are coming down and formal employment is picking up. However, informality is still prevalent and this raises questions about the sustainability of growth in the medium term. The country is investing efforts in access to larger markets through trade agreements and if micro and small firms find it difficult to remain formal due to an onerous state, little can be expected for them in this context. More broadly, informality means that most people are not covered by labor protections. Without a pension, the living standards in old age of many Peruvians are at risk. Without health insurance workers face significant risks to their livelihoods. President García himself referred in April 2008 to informality as the “slavery of the twenty-first century.� But informality also erodes the social contract between the population and the state, making the provision of public services more difficult by reducing the available revenue base. In addition, informality may have a negative impact on business performance – though the cross-country evidence is mixed. It is in this context that the Government of Peru has set the goal of reducing informality by one-third by 2011. ES. 3. This study is timely, as the authorities’ interest in analyses of informality has increased. This report is the outcome of very close cooperation with the authorities of the Government of Peru. The authorities have been involved in all stages of the process, going back to the first phase of this programmatic study. The authorities proved instrumental in helping design the focus of this report, including the scope of a survey of businesses conducted for this study, the survey questionnaire itself, they also suggested specific topics to be explored – which are presented throughout the report mainly in the form of boxes. 9 ES. 4. ‘Formality’ is better understood as having different dimensions. In this report we explore four of those dimensions: business licensing, tax, labor relations and pensions, and business incorporation. It is well known that informality is not a straightforward concept. As a result, it is not easy to agree on a single definition of informality or to measure it easily. The important conclusion to draw from this is that informality can be better understood as being a multi-dimensional concept. In this report we take a pragmatic approach by exploring four dimensions of ‘formality’: business licensing, tax, labor relations and pensions, and business incorporation. Of course, a business person can legally operate with only service providers and no employees; or as a natural person and not a business with a separate legal entity. However, the choice of these aspects of what ‘formality’ may be stems from the result of focus groups with micro and small business owners – these are the very issues they have in mind when they think of ‘formality.’ ES. 5. This report focuses on the trajectories towards formality of micro and small businesses, drawing insight by polling those businesses directly. For the purpose of this report, micro businesses are those with up to 10 workers and small businesses those with 11 to 50 workers. This working definition is a practical choice that also reflects the criteria regarding maximum number of workers in the existing Peruvian legislation. It also helped us to design a dedicated survey of 802 micro and small businesses conducted for this study. Informality is widespread across all sectors of the economy and affects all types of firms but it is particularly high among small businesses. Drawing from the day-to-day experiences of micro and small businesses is at the core of this report. The analysis adds value to the authorities’ information set by presenting new evidence on what are the most critical steps in the formalization process and what are the key considerations at those junctures. By focusing on practices and decision-making of businesses, this report complements the analysis of labor market and labor legislation that was a key part of the first phase of the programmatic study. ES. 6. Informality represents not only the exclusion of many workers but also reflects a decision by economic agents to exit the formal sphere. The approach followed here is informed by the analytical work for the World Bank’s Informality. Exit and Exclusion flagship report of the Latin American and Caribbean Vice-Presidency of the World Bank in 2007. In particular, this approach recognizes that informality not only undermines the social contract but is also a consequence of the limited value for the population of the services provided by the state. In other words, informality results also from decisions of economic agents about the optimal level of engagement with the state. Micro and small entrepreneurs use implicit cost-benefit analysis as to whether to remain informal or to formalize. Digging deeper into those choices and cost-benefit considerations is the subject of this report. 10 Main findings The study finds no significant impact of certain characteristics of formality, such as having a municipal license or being registered with the tax authorities, on the profitability of businesses or on their probability of obtaining a loan. These results must be accompanied, however, by the caveat that establishing the impact of informality econometrically is notoriously difficult. The methodology used in this report does not attempt to capture the negative effects that informality may have on the performance of formal firms, or the effects that informality may have in undermining social trust and the overall business environment. The study also finds informality to be particularly persistent in the area of labor relations and access to pension, driven largely by high costs of being formal, limited enforcement of labor relations, and a preference for flexible labor relations. These two findings combined suggest that: (i) efforts to increase the access to health insurance and pension coverage of workers through a shift from informal arrangements to existing formal labor relations will prove difficult; and (ii) alternative methods to increase access to health and pension coverage are well-worth exploring. The rest of these main findings concentrate on the conclusions of the study regarding the determinants of informality and what prompts businesses to move towards formality. ES. 7. What is an informal business? Micro and small entrepreneurs consider themselves to be formal if they pay (some) taxes, have a tax ID number and a municipal business license. Micro and small business owners do not typically consider that not having workers on the payroll is being informal. At some level business owners know that having workers on the payroll is mandated by law but the practice of not doing so is so widespread that it does not register with them as an informal practice. This implicit definition shared by micro and small entrepreneurs of what makes a business formal or informal is an important starting point. It helps to identify that any policy aimed at formalization of labor relations (or any aspect absent from the implicit definition) will face key communication and enforcement challenges. ES. 8. As businesses grow they tend to become more formal – but not in the area of labor relations. The first step of formalization among micro and small businesses is to obtain a tax ID number (Registro Único de Contribuyente, RUC) and a municipal license. Most of these businesses pay taxes on some of their actual income. As micro and small businesses grow businesses consider incorporating the business into a separate legal entity with limited liability. However, at the same time informal labor relations persist even as the business grows. Businesses with 11 to 50 workers display a very complex mix of labor relations among their workers. But these complex relations are largely informal even among the core group of stable workers. ES. 9. The trajectory of businesses towards formality is driven by how the costs and benefits across different dimensions of informality change as the business grows. The study draws on both a quantitative survey of 802 micro and small businesses as well as on qualitative data gathered from seven focus groups, interviews, and case- studies. Overall, the different sources of evidence presented in this study present a consistent pattern of how micro and small entrepreneurs make their choices regarding 11 whether to cross into formality or to remain informal. This pattern, which we now discuss in more detail, can be characterized in broad terms as reflecting largely the costs and benefits faced by business owners. At the same time, some evidence was also found that certain behaviors may be partly explained by a persistent ‘culture of informality’ by which micro and small businesses would remain informal simply because this is the societal norm. ES. 10. In many cases, the main driver behind the formalization process is the fact that enforcement makes it costly to remain informal. In the areas of business licensing and paying taxes the costs of remaining informal increase notably as the business grows. Enforcement is the critical factor affecting the decision to obtain a municipal license which otherwise is seen as conferring no benefits to the business. In addition, businesses perceive the process to get a municipal license as unnecessarily burdensome and arbitrary. Thus, micro and small businesses exploit any opportunity to avoid compliance. Businesses are less likely to get the municipal license if they operate in sectors such as textiles because they can operate workshops without being detected. In some cases, businesses prefer to pay off inspectors rather than obtain the license, but as the business grows so also does the number of visits by inspectors, thus making the strategy of paying them off not worthwhile, most businesses end up obtaining a municipal license. ES. 11. Businesses perceive benefits from formalizing only in the area of paying taxes and business incorporation. While enforcement of taxes is a key factor in ensuring compliance, businesses do see benefits in being registered with the tax authorities and regularly paying taxes (at least on some of their actual income). Benefits from formalizing in the area of paying taxes include an expanded client base since the business can issue receipts and can therefore sell to larger clients that require receipts and also to the state. In addition, there is some limited evidence to show that being registered with the tax authority and in a position to show regular payment of taxes may improve access to credit. The analysis also suggested that incorporating the business may be an important step in the formalization process. Micro and small business owners consider that incorporating the business under a separate legal entity with limited liability provides not only a way to preserve the individual patrimony but also business opportunities as the business gains credibility in the marketplace. ES. 12. But the existing tax regimes may need to be fine-tuned as they currently do not encourage formalization. This applies when special tax regimes do not encourage micro and small businesses to move to the general regime as they grow and to incorporate as separate legal entities. Few businesses choose to file taxes under the special regime for corporate income tax (Régimen Especial de la Renta, RER). Instead businesses prefer to remain under the much simpler, but more onerous, simplified regime (Régimen Único Simplificado, RUS) or the general regime – despite the fact that the latter imposes the full set of requirements designed for larger enterprises. These filing patterns suggest that compliance costs may be high under the RER – discouraging businesses from moving from the RUS to the RER. In addition, the constraint under the simplified tax regime (RUS) by which most legal entities are prevented from filing taxes under that regime, may result in businesses choosing to remain operating as natural persons for 12 longer than they otherwise would, which may have a negative impact in business access to credit and market opportunities. Overall, current practices contrast with the goal of having small businesses under the special regimes and larger businesses under the general regime. ES. 13. In the absence of enforcement, businesses keep informal labor relations and explain them partly on the grounds of a need for greater flexibility. The very fact that businesses do not consider this an aspect of informality is indicative of how prevalent it is. Based on the survey conducted for this study more than 80 percent of micro and small businesses have no workers with a contract. This high number is unsurprising given what we know from the household survey about the economically active population in Peru (again around 80 percent of the labor force without a pension). What is more remarkable is that the proportion of businesses with no workers on the payroll decreases only slightly among businesses with 11 to 50 workers, of which approximately two-thirds have no workers on their payrolls. As for the reasons why labor relations are so informal, the analysis from the focus groups suggested that in part this stems from a need to keep flexible labor relations in what is a rapidly changing and highly cyclical marketplace. While this may be true for smaller businesses, this reason is also claimed among larger businesses of 11 to 50 workers, which typically have a stable group of core workers yet often those core workers are still reported to be service providers who simply issue receipts for their services (recibos por honorarios). ES. 14. Micro entrepreneurs claim that the minimum wage and holiday pay impose not only a high burden but are also hard to reconcile with their preference for piece-rate pay arrangements. In some cases the foremost concern of micro and small business owners with the formal requirements regarding minimum wage and holiday pay does not appears to be their level but the fact that they perceive it to be in opposition to the piece-rate pay arrangement (destajo). Micro entrepreneurs are often confused when it is suggested that a minimum wage could be paid as a floor for their workers’ salaries. This perception is all the more puzzling given the complex mix of labor arrangements that micro and small businesses actually have. One interpretation of this finding is that minimum wages would be in fact binding for a large proportion of the workforce (for more than half of the workforce if the survey results are taken at face value). Hence the minimum wage may truly have the disincentive effects on effort that micro entrepreneurs fear. The survey conducted for this study provides some supportive evidence that two-thirds of all businesses surveyed consider that their business is not productive enough to allow paying the minimum wage. At the same time, the survey also provides evidence that there is a group of workers who earn more than the minimum wage but who are not on the payroll, for which the preference for piece-rate pay arrangements (on the side of the employer) and for higher take-home pay instead of benefits (on the part of the worker) seems to play a role. ES. 15. Firing costs are also perceived to be high and a deterrent to include workers on the payrolls. Most micro and small entrepreneurs consider that the compensation that workers receive upon been fired is not only too high but also a more important factor than the actual wage to be paid in explaining why workers are not on the payroll. Firing is also perceived to be complicated and the process prone to legal disputes 13 and costs. In this regard, and somewhat paradoxically, there may be an incentive to incorporate the business into a legal separate entity as a way to protect the entrepreneur from labor-related litigation. In addition to the level of firing costs, the study uncovered evidence that the de facto severance payments differ considerably from the de jure severance payments. To the extent that this results in increased uncertainty about the true costs of firing, this may further discourage the formalization of labor relations. ES. 16. Overall, micro and small business owners have limited information about the regulatory frameworks affecting them. Most notably, the Special Labor Regime, which provides for a much reduced burden of social benefits on the part of the employer, is virtually unknown among micro and small entrepreneurs. The limited knowledge of the Special Labor Regime for micro businesses highlights the communication challenges implied in addressing the issue of labor informality. In addition, there is limited information among micro and small entrepreneurs about the new regulatory framework for municipal business licensing. The interaction between the entrepreneur and the state is also made more difficult by the existence of different definitions of what constitutes a micro or a small business for tax and labor purposes and, more importantly, the fact that those definitions overlap only partially with each other. ES. 17. Micro and small businesses typically obtain information about the costs and benefits from their peers or from an accountant to formalize their business. This finding, which emanated from the qualitative research conducted for this study, helps to explain why, for example, factors such as owner education do not affect the choice of tax regime. The choice of tax regime is effectively not made by the owner but by the accountant. This qualitative research suggests that a similar pattern may also affect other decisions to formalize, particularly the decision to incorporate the business as a separate legal entity. Policy implications The following conclusions are meant to serve as input for the ongoing policy dialogue with the authorities of the Government of Peru (GOP). Policy implications that can be implemented in a relatively shorter time-frame will be discussed here first, followed by those that will pose greater challenges to implement. ‘Quick-wins’ ES. 18. Unify the definition of a micro and small enterprise for tax and labor purposes. While there is no standard international practice for what the criteria should be to consider a business to be a micro and small enterprise (MSE), the guiding principle should be one of clarity. The preferred solution is to have a single definition agreed by all different ministries and applied consistently. Alternatively, the number of criteria that a business has to meet to qualify under the different regulations could be reduced. One possibility would be to define eligibility for certain tax regimes on the basis of turnover and to define eligibility for the special labor regime on the basis of the number of employees. The decision to include additional criteria as a way of preventing certain businesses from ‘getting through’ under a more simple framework (e.g., a capital- 14 intensive and high-turnover business with few employees would qualify for the special labor regime) ought to be carefully weighed against the disincentive that complexity and partially overlapping criteria pose to formalization. ES. 19. Eliminate the register of eligible suppliers to the state in order to expand the benefits of formalization. Peru mandates that 40 percent of all state purchases to be sourced from MSEs. In practice less than 20 percent of purchases are sourced from MSEs. In part, the effectiveness of this legal requirement is weakened by the additional bureaucratic and cost burden that MSEs have to face in order to compete for state contracts, as all businesses wanting to supply the state have to pay to be registered on a list of eligible providers. A way to reduce the bureaucratic burden on micro and small business would be to eliminate the register of eligible suppliers, establishing that all formal businesses (e.g., those with a RUC number) are eligible to bid in public procurement tenders by default. If necessary the GOP could replace the register of eligible suppliers with a listing of ineligible suppliers: those fined by SUNAT or the Ministry of Labor. This change would have some impact on expanding the benefits of formalization without unduly compromising the integrity of the system since the register of eligible suppliers does not carry inspections. Overall, improved market access must be emphasized as a key benefit of formalization, given that micro and small business owners already perceive this a benefit of formalizing in terms of paying taxes and of business incorporation. ES. 20. Continue with simplification efforts in municipal business licensing. In Lima the Tramifácil project, with support from the IFC, reduced the time to obtain a license from a month to a week and cut the number of procedures from 33 to five. Such efforts are being expanded across an increasing number of municipalities and are expected to alleviate somewhat the concerns of micro and small business owners regarding the bureaucratic burden posed by municipal business licensing. To ensure that the costs of the municipal license remain affordable, municipalities need to ensure that, as provided by the regulatory framework, the fees charged to businesses reflect only cost recovery considerations. The cost of the municipal license in 2007 ranged from 1.7 percent of per capita GDP in Lima to 9.7 percent of per capita GDP in Huancayo. These figures are relatively high compared to a selection of 65 municipalities in the Latin America and the Caribbean region monitored by the IFC Municipal Scorecard, in fact Huancayo was the lowest ranked municipality of all 65, and Lima, while being the highest ranked in Peru, was only the nineteenth of the regional sample. ES. 21. Review the costs of complying with the different tax regimes to identify ways to reduce them, and re-calibrate the simplified and special tax regimes. Costs of complying with the requirements in the different tax regimes may be a significant factor affecting the choice of tax regime under which businesses are filing their taxes. However, more analysis is needed as the drivers of the choice of tax regime remain unclear. The restriction that only natural persons can file under the simplified regime: RUS, may discourage businesses from incorporating into separate legal entities, this study suggests that this is an important step in the formalization process. 15 ES. 22. Step up communication efforts, possibly in partnership with associations of accountants, those who are close to the micro and small entrepreneurs. The finding that micro and small entrepreneurs often based their decisions to formalize on the information they gather from other MSEs poses a challenge for effectively communicating with them. However, the finding that MSEs also rely heavily on the accountant that files their taxes opens up an opportunity for the GOP to partner with associations of accountants to improve the effectiveness of efforts to relay relevant information, such as that regarding the labor regime available to micro enterprises. ES. 23. Target businesses that already participate in some state programs (purchases, etc) to provide information relevant to those businesses. An additional opportunity to improve the reach of information to MSEs is to make the most of the fact that some MSEs are already in contact with the state through participation in a number of state-sponsored programs. Yet the survey evidence suggests that even among these MSEs knowledge of the Special Labor Regime was poor. The GOP may therefore want to explore making use of the opportunity to be in contact with the MSEs to introduce this information and to present options like the Special Labor Regime as an opportunity to reduce costs compared to the general labor regime. An additional point of contact with the MSEs would be through the SUNAT, though this is likely to be a less efficient way to relay information since the set of businesses dealing the state-sponsored programs is in effect a self-selected group. For municipalities to efficiently perform as a link in the relays of information, an even bigger effort would be required given the still broadly negative view that micro and small entrepreneurs have of the municipalities. Longer-term challenges ES. 24. Consider ways to expand access to health care insurance and pensions that do not necessarily involve including workers in the general labor regime. Given the intrinsic difficulties in establishing the link between informality and business performance measures, the econometric results presented in this study must be interpreted with great caution. However, they suggest that the key negative aspects of informality relate to the lack of social protection of workers. Given the tough hurdles that need to be overcome to make the labor regulatory environment more conducive to the expansion of formal employment in MSEs (see challenges below), it seems worthwhile to explore the feasibility of broad access schemes that will help expand health care insurance and pension coverage. In exploring policy options the potential impact of reforms on the incentives to contribute under the existing system needs to be carefully examined. A particular concern to bear in mind is that super-imposing non-contributory programs on top of a contributory system may actually create incentives for informality. ES. 25. Alleviate the burden of labor regulations on micro and small businesses, especially regarding social benefits and firing costs. The burden of labor regulation in Peru is among the highest in Latin America, as the first phase of this study documented. For example, in the Doing Business indicators 2008, Peru had a difficulty of hiring index of 44 (compared to 37 for the region on average). The new evidence uncovered in preparation of this second phase provides further support to the conclusion that labor regulations remain a barrier to formalization. The evidence compiled for this study 16 suggests that benefits such as holiday pay and high firing costs are perceived to be among the most important factors which deter businesses from including workers in their payrolls. In addition, severance payments that workers receive de facto are substantially (and somewhat unpredictably) lower than what workers are entitled to de jure. This adds to the uncertainty about what the actual firing costs may turn out to be, and hence firing costs and procedures may act as a deterrent to formal employment. Reducing firing costs to bring them in line with the average de facto severance payments would reduce uncertainty. An important caveat, however, applies to the case of micro businesses since most of these entrepreneurs do not know about the reduced benefits and firing costs that the Special Labor Regime for micro businesses allows. Expanding the coverage of this special regime to larger businesses may provide an alternative, though crucially reform efforts would be in vain if information failed to reach the MSEs. ES. 26. Consider ways to improve the minimum wage regime so that they become less of a binding constraint, including considering the appropriateness of region- specific minimum wage rates. The survey of micro and small businesses conducted for this study provides evidence that the minimum wage is perceived to be as too high, especially for the smaller businesses. In fact, the statutory minimum wage is slightly below the median wage for the economy as a whole. The level of the minimum wage ranks high for businesses among the reasons why workers are not paid the minimum wage. More than 70 percent of businesses with 1 to 5 workers consider that their business is not productive enough to pay the minimum wage to workers. This proportion is lower for larger businesses but even of those with 11 to 50 workers, more than half of surveyed businesses consider that they are not profitable enough to pay the minimum wage to workers. ES. 27. Once an adequate regulatory labor regime is in place it is important to enforce it firmly and consistently. The combination of weak enforcement and overly strong de jure labor regulations may undermine the rule of law and cultivate a social norm of non-compliance. This calls for improvements in the regulatory regime, as noted above. It also calls for a more effective enforcement once an adequate regime is in place. The role of weak enforcement was also highlighted by the results of the survey. The finding that the share of workers outside the payroll is higher than the share of workers earning the minimum wage indicates that there are workers for which the minimum wage is not the binding constraint impeding formalization. In turn, this finding suggests that opportunistic behavior may be a driving force of labor informality, at least for some workers. Moreover, the analysis found some support for the role of labor inspection rates in affecting the probability of a business to employ workers under formal labor relations. Improving the enforcement of labor laws will likely require an increase in resources given the number of inspectors that the Ministry of Labor (MTPE) currently has at its disposal. In this regard, the addition of 104 new labor inspectors to the MTPE, bringing the total to 450, brought the number of inspectors to around 1 per 40,000 members of the economically active population. This is still a low figure as Peru would need approximately three times as many inspectors to reach the benchmark suggested by ILO (2006) for industrializing countries of 1 inspector per 15,000 workers. Peru would still need to almost double its number of inspector to reach the ratios observed in Argentina, Brazil, or Chile, all of which have ratios of 1 inspector per 20,000 to 25,000 workers. 17 18 1. Introduction 1.1 This report, the second part of a programmatic study, focuses on the trajectories of Peruvian businesses towards formality. This is the second phase of a programmatic study on informality in Peru the objective of which is to analyze the causes and effects of informality so as to provide policy options for a reform agenda. The first phase took stock of the characteristics of informality, and focused on how the labor market and labor legislation in Peru may be associated with informality (World Bank, 2007a). While the first phase of the study noted strong incentives to remain informal, particularly due to labor and tax codes, it also argued the need to better understand the businesses’ perspective on the costs and benefits of informality and to map the steps that businesses need to take towards formalization. 1.2 ‘Formality’ is better understood as having different dimensions. It is well known that informality is not a straightforward concept (Box 1). As a result, it is not easy to agree on a single definition of informality or to measure it easily. The important conclusion to draw from this is that informality can be better understood as being a multi- dimensional concept covering areas such as business licensing, tax, labor relations, and (as discussed below) also business incorporation. The focus of this study is to document the trajectories of businesses towards formality along these different dimensions. The interest in the trajectories towards formality is rooted in the fact that, regardless of their size, most businesses have some characteristics that are to a degree informal. Hence, answers to questions about which dimensions of formality are achieved first, and why, can help the policy discussion on how to reduce different aspects of informality. Box 1: What do we mean by informality? The term informality means different things to different people. Some are concerned about workers not enjoying the protection of existing labor laws, with the corresponding impact on equity and welfare. Others are concerned about efficiency and productivity, as informality is seen as being associated with firms that are too small. Informality is also a source of concern for those who see it as eroding the legitimacy of public institutions and as suggestive of poor policy regimes. Reflecting in part these different concerns, multiple measures of informality have been advanced in the literature, commonly classified in two broad groups: (i) The productive definition of informality, which focuses on the characteristics of the productive unit. One such productive measure of informality is the one traditionally used by the International Labor Organization, which defined the informal sector as economic units “with scarce or even no capital, using primitive technologies and unskilled labor, and then with low productivity� (ILO 1991). To make this definition operational, it is common to assume that the self-employed, family units, and micro-entrepreneurs and their employees fit under the description of activities with low capital and low productivity and are therefore considered to represent informal employment. (ii) The legalistic definition of informality, which focuses on coverage of workers by mandated social protections. A common example of a measure of informality inspired by the legalistic definition of informality is to consider that a worker without rights to a pension is an informal worker. This approach has the advantage of recognizing that informal employment can be found 19 in firms of all sizes and in all sectors of the economy. It is particularly useful in the case of Peru since participation in the state pension system is mandatory for all workers on the payroll of a firm (strictly speaking, this participation is optional for workers under the special labor regime for micro firms with less than 10 employees, but participation in this regime is limited to only around thirty thousand micro firms, see Box 4). Informality covers many aspects, and in practice many indicators of informality are widely applied. While helpful, the use of other proxies creates the potential for ambiguity in the meaning of: ‘informality’. To that end, we will now make clear which proxy measure we are using for informality. This box draws from Chapter 1 of World Bank (2007b). 1.3 In this report we explore four of those dimensions of formality: business licensing, tax, labor relations and pensions, and business incorporation. In this report we take a pragmatic approach by exploring four dimensions of ‘formality’: business licensing, tax, labor relations and pensions, and business incorporation. Some of these dimensions are more easily understood as referring to the formality-informality divide: to be formal a business has to comply with business licensing requirements and pay the required taxes. Regarding labor relations, an entrepreneur could in principle have no employees on the payroll and still be in full compliance with the law, as it is possible to pay for services through a contractual relation that is not of an employment nature. However, such a situation goes against the spirit of the labor protection legislation and, in the legalistic definition of informality, is typically considered to be one of the dimensions of informality. Finally, the report also considers one more dimension of formality: the incorporation of a business as a separate legal entity. A business person can legally operate as a natural person. But, as will be shown below, entrepreneurs themselves consider this an important aspect of the formalization process. It is also worth stressing that this report is not only concerned with illegal economic activity, which by definition, cannot be conducted in a formal way. 1.4 The report has been designed and prepared in close cooperation with the authorities to help draw implications for the Government’s priority of reducing the size of the informal sector. This suggests that there is much to gain from a detailed analysis of what is formal and informal at different stages of the development of a business. Documenting the reasons why and when a business chooses to formalize and the reasons why it may do so with regard to one facet of informality (say tax) but not another (say labor) is of interest to Government authorities as they consider policies that will help increase the speed with which businesses traverse the trajectories toward formality. It is also an area where, as identified in the first phase of the study and the in concept stage of this report, the World Bank can add value to the existing literature on informality in Peru. 20 Figure 1.1: Informality in Peru and other countries in Latin America and the Caribbean Percent of workers that are informal, according to different definitions Informal workers (productive definition) 100 80 60 40 20 0 El Salvador Paraguay Argentina Uruguay Nicaragua Venezuela Dominican Rep. Honduras Guatemala Ecuador Bolivia Panama Jamaica Brazil Colombia Costa Rica Haiti Chile Mexico Peru Workers without a pension 100 80 60 40 20 0 Uruguay Salvador* Paraguay Nicaragua Argentina* Venezuela* Guatemala Ecuador Bolivia Brazil Colombia Chile Mexico* Peru El Workers without health insurance 100 80 60 40 20 0 El Salvador Paraguay Nicaragua Argentina Venezuela Guatemala Ecuador Jamaica Colombia Costa Rica Haiti Chile Mexico Peru Workers without a contract 100 80 60 40 20 0 El Salvador Nicaragua Guatemala Ecuador Bolivia Panama Dominican Chile Mexico Peru Rep. Note: Data for different countries do not all refer to the same year. * Refers only to salaried workers. Source: Gasparini and Tornarolli (2006) and ENAHO (2006) for Peru. 21 1.5 The interest of the authorities in understanding the trajectories towards formality stems from their development objective of reducing informality. The Peruvian authorities’ interest is high compared to other countries in Latin America and the Caribbean according to some measures (Figure 1.1). The Government has a formally stated goal of reducing informality by one-third by 2011. Most recently in April 2008 President García referred to informality as the “slavery of the twenty-first century� and stressed that, without access to a pension, the living standards in old age of many Peruvians are at risk.1 1.6 The first phase of the study raised a number of questions about informality in Peru that required additional data and further analysis. In broad terms, the first phase of the study argued that informality is driven by: (i) strong incentives to remain informal, particularly related to the high cost of labor and tax codes; (ii) bureaucratic obstacles to formalization; (iii) insufficient incentives to become formal, including greater access to credit, markets, and government-funded business services, combined with enforcement and penalties for those who do not formalize; (iv) a pervasive “culture of informality� that encourages regulation evasion; and (v) a frequently-changing array of laws and regulations which inhibits long-range planning and acts as a disincentive to formality. However, the first phase of the study also noted that it was often not possible to identify which factors were most critical given the limited availability of data. 1.7 The path towards formality reflects the choices made by businesses to stay informal or to formalize. This report explores why and how firms make those choices. Regardless of the measure with which one proxies informality, a common thread is the fact that informality reflects a lack of compliance with legal norms. This raises the question as to why agents are not in compliance with the law. The answer to this question has often been framed in terms of the exclusion of workers, which are involuntarily left outside formal institutions and are also without the protections given to formal workers. However, a second view emphasizes that there is an exit dimension to the actions of agents, by which firms and workers may simply opt into informality not finding any net benefit of interacting with the state (World Bank 2007b). A key objective of this report is to contribute to a better understanding of this exit dimension of informality in Peru from the perspective of the small business owner or self-employed person. To do so the World Bank has gathered new firm-level data to complement existing information.2 1 The goal of reducing informality was stated by President García in his latest annual address on the state of the nation on July 28, 2007. The reference to informality as the new slavery of the twenty-first century is from remarks by President García to the press on April 23, 2008. 2 Phase I of the study was largely based on data from national household survey (ENAHO). The literature on informality in Peru has been based mainly on the ENAHO and other household surveys. For example, De la Roca and Hernández (2004) draw on the Encuesta Nacional de Niveles de Vida conducted in May 2000 by the Instituto Cuánto. To the best of our knowledge the survey commissioned as part of this report constitutes the first business-level survey with a focus on decisions regarding formalization issues. 22 1.8 To this end, this report provides new business-level evidence on the determinants of informality in Peru. The main value added of this study is to draw from new business-level data to shed light on the determinants of informality in Peru and in the policy debate. Therefore, the focus throughout the text will be to present and analyze this new information. Design and analysis of the research has been greatly influenced by recent efforts to explain the phenomenon of informality in Latin America, most notably the World Bank’s Informality. Exit and Exclusion flagship report of the Latin American and Caribbean Vice-Presidency (World Bank, 2007b). For the sake of brevity, references to broader regional trends and to theoretical considerations are kept to a minimum throughout the text. The interested reader is referred to the flagship report for a broader context and for a discussion of theoretical issues. For an international perspective on informality in Peru the reader is referred to the report produced during the first phase of this programmatic study. 1.9 By focusing on business-level data the report helps to fill a gap in the literature. As noted in the first phase of this programmatic study there are only a limited number of studies that address the determinants of informality in Peru using business- level information. Jaramillo (2004) asked a sample of firms in the garment industry in Lima for the reasons that explained their decision to become formal: 85 percent answered access to more suppliers and customers; 70 percent answered access to credit from a formal source; 63 percent to avoid the payment of fines and 40 percent to avoid the payment of bribes. Yamada and Chacaltana (2007) document six successful cases of firms expanding and generating employment. Access to exports markets was an important factor for firm expansion in the cases studied. The Municipal Scorecard report for 2007 asked firms that had obtained their business operating licenses and construction permits the main reasons why they decided to go formal. These firms identified not paying fines and law enforcement as the main reasons, while access to the judicial system and better credit were ranked as the least important reasons. While informative, these studies do not always provide systematic treatment of the different dimensions of informality nor do they have a broad coverage in terms of the number of businesses polled. 1.10 The report also studies the impact of informality on businesses’ profitability and access to credit. The first contribution of the report is to study the factors affecting informality controlling for a large number of business characteristics. A second contribution of the report is to analyze the impact of those factors. This requires econometric analyses that can establish the ‘causal’ links between the determinants of informality and variables of interest such as profitability or access to finance. While these analytical techniques have shortcomings, the estimates provide a contribution to the discussion of informality in Peru as a careful econometric analysis of the impacts of informality that was not available previously. 1.11 A key building block for the analysis in this report is a dedicated survey of 802 micro (up to 10 workers) and small (11 to 50 workers) businesses, both formal and informal. The National Household Survey (Encuesta Nacional de Hogares, ENAHO) provides a good basis for determining the extent of informality in Peru. It allows us to provide a clear answer to the extent of informality according to the legalistic 23 definition, since it contains information on whether the respondents to the survey are covered by a pension or not. There is also information that can be used to construct a measure of informality according to the productive definition (as was done in phase I of this programmatic study). However, the objective of this study is to move one step further and explore the impact and determinants of informality. To answer these questions we need a firm-level data set of both formal and informal businesses that will allow us to control for a large number of business characteristics such as size, age, education of the owner, etc., that may be thought to affect the choices of firms on their trajectory towards formality. To this end the World Bank commissioned a dedicated business survey, as detailed in Box 2.3 While the survey conducted provides us with a cross-section of data by including questions such as ‘how long did the business operate before obtaining a license?’ it allowed us also to obtain information about the trajectories of businesses towards formality. Box 2: A dedicated survey to study informality in Peru To better understand the factors behind informality in Peru the World Bank commissioned a dedicated survey of businesses, both formal and informal. A total of 802 businesses were interviewed face-to-face in five cities: Arequipa, Cusco, Huancayo, metropolitan Lima, and Trujillo. These five cities were selected to reflect a cross-section of urban centers, accounting for more than 10 million inhabitants in total or around 36 percent of Peru’s estimated population. We restricted the survey to cities since urban informality has been the focus of the debate in Peru, largely on account that it is perceived to be in cities where the potential for reducing informality is greatest. In addition, informality in urban and rural areas may not share common characteristics. Conducting a survey of both formal and informal businesses poses the challenge that there is no readily available census data to use as a frame from which to draw a random sample of businesses. To ensure as much as possible a random sample of respondents, the sampling was designed as a three step process. First, on the basis of the population census, a certain geographical area within a city was selected at random. Second, an enumerator was sent to that particular geographical area and was asked to list all businesses that could be found. While this necessarily introduces some selection bias, it is a pragmatic approach for addressing the complex issue of how to construct a frame from which a random sample can be drawn with limited resources. It is also an approach that has been successfully used in other cases (De Mel et al., 2007). Third, a random sample of businesses to interview was drawn from that list. For businesses in metropolitan Lima (except those in transport) the first and second steps of the methodology were unnecessary since the survey firm already had a proprietary census carried out in 2007 of more than 100,000 firms covering the sectors of activity of interest. A pure random sample of informal and formal businesses would have resulted in a very large number of single-person businesses and very few businesses of 6 to 10 employees and of 11 to 50 employees. Therefore stratification quotas by city, industry, and firm size were applied to ensure that the sample covered sufficient observations to enable a statistical analysis of the results by different firm sizes and other categories (see Chapter 2 for descriptive statistics). 3 The design of the survey has benefited from the experience with a similar survey recently undertaken for World Bank (2008a), Republic of Bolivia: Policies for Increasing Firms’ Formality and Productivity (Report No. 40057-BO). 24 Only businesses operating in seven sectors of the economy were surveyed. The selected sectors were chosen to represent a broad cross-section of manufacturing and services activities: (i) shoe and leather manufacturing; (ii) textile and apparel manufacturing; (iii) wood products and furniture manufacturing; (iv) metal products manufacturing; (v) retail sales of foodstuffs; (vi) transport by land; and (vii) restaurants and hotels. In 2006 the combined seven groups of activity accounted for around 30 percent of Peru’s economically active population, and a somewhat higher figure of employment in micro and small firms. Transport by land accounted for 4.6 percent of GDP in 2006, restaurants and hotels for 3.7 percent, textile and leather manufactures for 2.2 percent, metal products manufacturing for 1.2 percent, and wood products and furniture manufacturing for 0.5 percent of GDP. However, it is not possible to exactly match the weight of the retail sales of foodstuffs in GDP (retail activities as a whole account for around 14 percent of GDP). 1.12 Focus groups, in-depth interviews, and case studies complemented the quantitative data gathered from the quantitative survey. The dedicated survey of businesses was complemented with qualitative information gathered from seven focus groups and 15 in-depth interviews with micro and small entrepreneurs. Participants in focus groups were micro and small entrepreneurs with at least two years in business. This allowed us to probe focus group participants about their trajectory towards formality. Participants in focus groups were grouped according to the size of their businesses. In particular, two focus groups were dedicated to entrepreneurs with up to 4 workers, two more for entrepreneurs with 5 to 10 workers, and three groups with owners of businesses with 11 to 50 workers. Participants in the focus groups undertook their business in one of the following areas of activity: manufacturing (textiles and apparel, wood and furniture, leather and shoes, metal products), retailing of foodstuffs, transport, hotels and restaurant, and construction. The in-depth interviews were targeted at business owners who also took an active role in an association of micro and small businesses (gremios), which tend to be better informed. This ensured that the data gathering effort polled those better informed micro and small business owners while avoiding that their presence in a focus group would affect the group dynamic and that they would come to dominate the focus group discussion. In-depth interviews were held with leaders representing associations of micro and small businesses in a range of activities: manufacturing of wood products and furniture (four interviews); manufacturing of metal products (three); textile and apparel manufacturing (three); shoe and leather manufacturing (one); retail of foodstuffs (one); transport (one); and construction (one). Different research guides were prepared for the focus groups and for the in-depth interviews. Both the focus groups and in-depth interviews took place in January 2008 in metropolitan Lima. 25 Informality in the sectors selected for the survey 1.13 The degree of informality observed in the sectors selected for the survey is close to the one seen for the economy as a whole (see Table 1.1). In addition to allowing us to compare measures of the legalistic and productive definition of informality, the ENAHO captures other characteristics of a business, where a respondent works, that can be thought of as reflecting how formal or informal the business is. In particular, the ENAHO contains questions on whether the worker has healthcare coverage; has a contract or not; whether the employer is a business that is registered with its own legal personality; and whether the business where the respondent works keeps any accounting books. As noted in the first phase of the study, informality in Peru is high regardless of the definition or aspect of it that one chooses to examine. As shown in Table 1.1, the same is also the case for the sectors selected to be included in our dedicated survey. However, the ENAHO is not a good source of information for understanding the trajectories of businesses towards formality since it does not provide us with information as to the reasons why businesses choose to remain informal or to formalize. Table 1.1: Measures of informality based on the household survey in the sectors selected for the survey and in the economy Percent of economically active population … ... informal ...without ...without ...working …working in …working according to a pension health without a firms without in firms that the productive (legalistic coverage contract legal keep no definition* definition) personality accounting books Aggregate of selected sectors 85.7 85.6 79.9 85.9 88.1 82.3 for the survey All economic 75.9 80.8 71.7 73.4 81.5 76.9 activities Notes: Data for 2006. * See Appendix A for details on how this measure was constructed. Source: World Bank staff calculations based on ENAHO. 1.14 Informality affects all sectors of the economy selected for the survey but is particularly high among retail firms. Informality is a phenomenon that affects all sectors of the economy. The widespread prevalence of informality across sectors is reflected in the sectors selected for the survey.4 Among the seven sectors selected for the survey, the share of workers with a pension ranges from 71 percent in the case of metal and machinery manufacturing to 89 percent in the case of retailers of foodstuffs and 4 The inclusion of retail and transport – two sectors traditionally highly informal – is also justified on the basis that these are dynamic sectors. In fact, Chong et al. (2007) argue that one reason for the increase in informality observed during the 1990s is precisely the structural change towards retail and transport. 26 hotels and restaurants. Other measures of informality display a similar pattern, with retail of foodstuffs having the highest share of workers without a contract (92 percent) and the highest share of employees working in firms that are not registered with their own legal personality (93 percent). Among the manufacturing sectors included in the survey the share of workers without a pension is highest in footwear and leather manufacturing and in furniture manufacturing (86 percent in both cases). 1.15 Informality is particularly high among micro firms (with 10 or less workers). In both the sectors selected for the survey and in the economy as a whole, informality is a phenomenon that is highly correlated with firm size. As shown in Figure 1.2 the share of workers without a pension is around 90 percent for firms with up to 10 employees. The share of workers without a pension declines among those working in firms with 11 to 50 employees, although it is still relatively high at more than 60 percent. These data make it clear that controlling for firm size, and any other observable characteristics of firms, is essential to draw conclusions about the determinants of informality. The analysis underpinning Figure 1.2, in addition to other tables and figures was repeated using the different indicators of informality from the household survey, as reported in Table 1.1. The results obtained using these additional indicators are reported in Figure A.2 to Figure A.10 in Annex A. Figure 1.2: Informality by firm size, for selected sectors and for the economy as a whole As measured by the ratio of workers without a pension in firms of different sizes (2006) Sectors selected for the survey All economic activities 1 1 .9 .9 Informality prevalence Informality prevalence .8 .8 .7 .7 .6 .6 .5 .5 1 to 5 6 to 10 11 to 50 1 to 5 6 to 10 11 to 50 Number of employees Number of employees Source: World Bank staff calculations based on ENAHO. 1.16 The five cities included in the survey are located in regions with a diverse incidence of informality – within the high rates seen across the country. By including businesses from Trujillo (La Libertad), Huancayo (Junín), and Cusco, the survey draws from cities in regions where the share of workers without a pension is relatively high (above 85 percent) compared to Lima and Arequipa (70 percent), as shown in Figure 1.3. 27 Figure 1.3: Map of informality prevalence by region Percent of workers without a pension (2006) 88.5 88.5 83.9 94.5 83.7 94.7 95.8 85.2 89.0 93.0 90.1 85.5 70.8 85.1 68.6 91.8 Legend 95.4 87.1 Five less informal departments 94.2 Second five less informal departments 60.4 96.0 88.4 Third five less informal departments Second five more informal departments 71.8 Five more informal departments 71.8 71.5 Note: Informality rates are shown at the departmental level since this is the most disaggregated geographical unit for which the annual ENAHO survey provides a representative sample of the population. Source: World Bank staff calculations based on ENAHO. 1.17 Even though urban areas are typically richer, the choice of cities for the survey reflects a range of average incomes and incidence of informality. Informality is highly correlated with income levels, as Figure 1.4 shows. This is an important factor when designing any research into informality and when interpreting the results. In our case, the inclusion of cities such as Cusco and Trujillo ensured that the dedicated survey did not only cover the richest areas of the country. Finally, the positive correlation 28 between average incomes and formality makes the importance of exploring possible directions of causality very clear: is it that formality contributes to higher incomes or that those with higher incomes self-select themselves into more formal arrangements? These questions are explored in Chapter 3, which includes an analysis of the impact of informality on the profitability of businesses. Figure 1.4: Informality prevalence by region and income per capita Informality as measured by the ratio of workers without a pension (2006) 1 San Martín Huanca- velica Ayacucho Amazonas Cajamarca Apurímac Huánuco Madre de Dios Ucayali .9 Puno Ancash Loreto Tumbes Informality prevalence Cusco Pasco Junín Piura La Libertad Lambayeque .8 Arequipa Moquegua Tacna Lima .7 Callao Ica .6 4000 5000 6000 7000 8000 9000 10000 11000 12000 Average annual income per capita, Nuevos soles Source: World Bank staff calculations based on ENAHO and INEI. 1.18 The sectors selected for the survey are also representative of the economy as a whole with regard to trends in informality. Informality, as measured by the share of workers with a pension, has declined in the sectors selected for the survey in line with the decline observed for the economy as a whole. As measured by the proportion of workers without a pension, informality declined from around 87 percent in 2003 to around 80 percent in 2006. Informality also declined in the sectors selected for the survey, as shown in the left panel of Figure 1.5. These declines in the level of informality have taken place in a context of accelerating economic growth since 2004, as shown in the right panel of Figure 1.5. This evidence suggests that while informality may be a persistent phenomenon, strong economic growth has contributed to its reduction in recent years. However, using alternative indicators of informality based on the ENAHO, there has been no significant decline in recent years in the share of workers without a contract or working in businesses without a separate legal personality or without accounting books (Figure A.8 to Figure A.10 in Annex A).5 In addition, one has to bear in mind that the 5 While the relationship between long-run and short-run informality has not been formally examined here, this evidence suggests the possibility that, while different measures of informality are typically highly correlated, different ‘dimensions’ of informality may respond faster than others to transient 29 decline in informality, as measured by the lack of access to a pension, has required ever increasing rates of economic growth, as shown in the right panel of Figure 1.5. Figure 1.5: Informality rate and economic growth, 2003 to 2006 Informality as measured by the percent of workers without a pension Informality rate for sectors selected for the Informality rate (all economic activities) survey and for all economic activities and economic growth 95 95 90 90 2004 2003 Informality rate Informality rate 85 85 2005 2006 80 80 75 All economy 75 Selected sectors 70 70 2003 2004 2005 2006 3.5% 4.5% 5.5% 6.5% 7.5% 8.5% Real GDP growth Source: World Bank staff calculations based on ENAHO and INEI. 1.19 The remainder of the report reviews, in turn, the determinants of informality (Chapter 2) and its impact (Chapter 3). As part of the analysis of the determinants of informality the study identifies the trajectories of micro and small businesses towards formality. In addition, a number of boxes in the report pay particular attention to: (i) taxation of micro and small enterprises; (ii) sector-specific labor regimes; and (iii) unfair dismissal costs. These topics were selected in close cooperation with the authorities of the Government of Peru to address specific queries that the authorities had. In this regard the report contributes to the knowledge base for policy-makers by: (i) quantifying the tax burden on new capital investment by micro and small businesses under different tax regimes: (ii) providing a detailed compilation of labor regulations affecting different sectors; and (iii) producing a new data set that allows for a comparison between the payments to which a dismissed worker is entitled de jure and the actual payments that the dismissed worker receives after a conciliation process. shocks. For a discussion of the link between long-run and short-run changes in informality see Loayza and Rigolini (2006). 30 2. Determinants of informality Scope: To explore why businesses choose to keep certain aspects of their operations informal we ask businesses themselves, through the dedicated survey and focus groups. We ask about businesses’ perceptions but also about their actual level of formalization along a range of dimensions and about a large number of characteristics of their businesses. This allows us to explore the determinants of informality econometrically. Structure: The chapter is structured in four sections. First, we provide information about the characteristics of the sample of businesses from which we are drawing information. Second, we report what our respondents tell us, through the survey and focus groups, about the factors that drive their choices regarding formality and informality along a range of issues. This section is structured along four broad dimensions of informality: (i) business licensing; (ii) paying taxes; (iii) labor relations and access to a pension; and (iv) business incorporation. Third, the chapter exploits the information from the survey to establish econometrically what factors affect the probability that a business is formal (as captured through some characteristics associated with formality). In addition, the chapter includes a number of boxes reviewing specific issues selected with the authorities as areas of particular interest: (a) taxation of micro and small enterprises; (b) sector-specific labor regimes; and (c) unfair dismissal costs. Fourth, the chapter sums up the information from the different data sources, quantitative and qualitative, to describe the common paths towards formality among micro and small businesses. 31 A. Who is our source of information? 2.1 Businesses surveyed cover a broad spectrum of micro and small enterprises... Details of the actual businesses surveyed in terms of their size, location, and sector of activity is provided in Table 2.1. The qualitative data-gathering effort also took account of the heterogeneity of micro and small businesses. Therefore different focus groups were arranged for businesses of different sizes. In total, four focus groups were conducted with businesses of 10 or less workers and three focus groups were conducted with businesses of 11 to 50 workers. An additional source of information gathered was through in-depth interviews with micro and small business owners who are also representatives in associations (gremios) of micro and small businesses. Table 2.1: Number of businesses surveyed By size No. of By city No. of By sector No. of businesses businesses businesses 1 – 3 workers 226 Arequipa 111 Shoe and leather 117 4 – 5 workers 195 Cusco 108 Textile and apparel 119 6 – 10 workers 201 Huancayo 93 Wood products 126 11 – 20 workers 149 Lima (met.) 368 Metal products 118 21 – 50 workers 31 Trujillo 122 Retail foodstuffs 105 Transport 95 Restaurants 122 Total 802 Total 802 Total 802 Source: World Bank survey of micro and small businesses. 2.2 … and a wide range of situations regarding how formal or informal they are. The ‘dimensions of informality’ that are explored in this chapter can be grouped into four broad categories: (i) business licensing; (ii) paying taxes; (iii) labor relations; and, (iv) incorporation of businesses as separate legal entities. Some of these dimensions of informality, like paying taxes and business licensing, are directly linked with the definition of informality as not being compliant with the laws and regulations. Others, like labor relations, have a direct link with the legalistic definition of informality discussed above (as it covers the extent to which workers have access to a pension). Finally, while a businessperson can operate formally as a natural person, the registration of the business as a separate legal entity emerged from the focus groups as a step that has important implications for a business. Table 2.2 shows the characteristics of the businesses surveyed along those four dimensions of informality, including a breakdown by size of business. 32 Table 2.2: Selected informality characteristics of surveyed businesses Percent of businesses surveyed All 1–5 6 – 10 11 – 50 businesses workers workers workers Business licensing Firm has municipal license 67 59 67 83 Firm has permanent municipal license 51 44 50 67 Taxes Firm pays taxes* 69 56 77 89 Percent of sales reported for tax purposes** 53 51 58 53 Firm has RUC number 78 65 90 96 Firm files under RUS 28 33 31 14 Firm files under Special Regime (RER) 11 7 19 11 Firm files under General Regime 38 24 38 70 Firm never gives facturas or boletas 19 29 11 6 Labor relations Firm has no workers with a labor contract 83 92 83 63 Firm has no workers with health insurance 71 83 71 45 Firm has no workers with pension coverage 83 93 84 59 Business incorporation Firm has limited liability (SA, SRL, SAC) 17 6 16 44 Sample size 802 421 201 180 Notes: * Businesses were asked to detail their outlays for a month and to include how much they pay in taxes in that month. The figure reported in the table represents the share of businesses which reported a positive amount being paid in taxes; ** The question asked was: “Let us suppose that the true sales of a business are 1,000 soles a month. How much do you think that a business owner like yourself would report for tax purposes?� Source: World Bank survey of micro and small businesses. 2.3 Bigger businesses are more formal across all dimensions of formality but are still largely informal regarding labor relations and often operate as natural persons. There are differences in the degree of informality along the different dimensions of informality analyzed and for businesses of different sizes. Bigger businesses, especially those with 11 to 50 workers, are more formal than smaller businesses. However, there is low formalization with regard to labor issues even among businesses with 11 to 50 workers: 59 percent of the businesses of that size surveyed have no workers with pension coverage and 63 percent have no workers with a written contract. Similarly, less than half of the businesses with 11 to 50 workers are incorporated as a separate legal entity with limited liability. Among businesses with 11 to 50 workers, formality along the other 33 dimensions is much greater: 83 percent have a municipal license and 89 percent pay taxes. However, the assessment of how much income goes unreported to the tax authorities is similar for businesses of different sizes. (Additional breakdowns by city and sector of activity can be found in Annex B.) 2.4 Smaller businesses are relatively formal with regard to business licensing and paying taxes. While micro businesses of up to 10 workers are largely informal with regard to labor relations, they are relatively formal with regard to getting a municipal license and paying taxes. Even among businesses with 1 to 5 workers, 59 percent have a municipal license; 56 percent pay taxes; and 65 percent have a tax (RUC) number. The share of businesses with a RUC number increases to 90 percent among businesses with 6 to 10 workers. Smaller businesses are relatively more sophisticated regarding tax issues than may have been expected: 24 percent of businesses with 1 to 5 workers report filing taxes under the general regime.6 In fact, some large businesses use the simplified tax regime (designed for small businesses) and some small ones use the general tax regime (designed for larger businesses), an issue which will be explored further below. 2.5 The survey data confirms that there are different stages along the continuum from informality to formality. The survey provides us with a large number of formality characteristics for our sample of firms. To study which combinations of formality are most common we use principal components analysis. This allows us to identify which characteristics of informality are interrelated and can be grouped together. The results from this exercise are provided in Figure 2.1, showing that formality along the dimensions of business licensing and paying taxes are most common. Figure 2.1: Share of surveyed businesses at different degrees of formality Percent of firms by groups of formality characteristic 100 Other combinations of formality 90 RUC, municipal license, all workers with 80 labor contract, and limited liability 70 RUC, municipal license, and all workers with labor contract 60 RUC, municipal license, and limited liability 50 Municipal licenseand RUC 40 30 RUC only 20 Municipal license only 10 Nothing 0 1 to 5 workers 6 to 10 workers 11 to 50 workers Note: The underlying data is reported in the appendix in Table A.8. Source: World Bank survey of micro and small businesses. 6 The breakdown of businesses filing under the different regimes in our sample (among those with a RUC number: 36% file under RUS, 14% under RER, and 49% under the general regime) is similar to the breakdown for all firms registered with SUNAT. Among the 834,000 firms with sales under 360,000 soles in 2006, 41% were under RUS, 10% under RER, and 49% under the general regime. 34 B. What do micro and small business owners tell us? 2.6 The pattern of selective formalization as businesses grow reflects the perception of costs and benefits associated with each dimension of formality. The focus group technique and the in-depth interviews with micro and business persons provide us with some insight into the decision-making process regarding formalization. Table 2.3 summarizes the key considerations when making those decisions as reported by micro and small entrepreneurs in the focus groups and interviews. Table 2.3: Benefits and costs of formalizing perceived by micro and small business owners Issue Benefits Costs Key rationale for Dominant decisions regarding force driving formalization formali- zation Munici- Informal - Avoid paying the - Possibility of fines Municipal license is Enforcement pal (not license fee and payments to perceived to convey no business having a - Avoid heavy inspectors benefits and as a heavy license license) bureaucracy bureaucratic burden, often unreasonably enforced. Formal - Avoiding fines - Annual fees As firms grow and are (having a - High cost of dealing more easily detected, they license) with the bureaucracy obtain the license to avoid - Fines still possible repeated fines. Firms due to unreasonable operating behind closed inspections doors further delay obtaining the license. Taxes Informal - Avoid paying - Possibility of fines if Registering with the tax Enforcement (not taxes detected authority (getting RUC and paying, - Selling without - Lost customer base number) is perceived to be expanding no RUC, receipts makes tax due to inability to easy and not expensive. customer no use of evasion easier issue receipts Very limited information base. receipts) about the tax regime – decisions which are left to Key role Formal - Avoiding fines - Paying the the accountant. played by (paying, - Slightly better accountant accountant having As firms grow the risk of (choosing tax access to credit - Tax payments, being detected increases RUC, for larger amounts perceived to be too regime for using and the need to issue smaller high (same tax rate receipts makes firms receipts) for all businesses firms). register with the tax draws criticism) authority and pay taxes. Labor Informal - Avoid paying - None mentioned Preference for pay by Lack of relations (no staff social benefits (some recognize the piece rate (on grounds of enforcement in pay- - Greater flexibility need for health care productivity) and desire and need for roll) in labor relations coverage but provide for flexibility dominate flexibility. for core group of among smaller firms. Being able to - More incentives workers privately) Firms claim that workers sell to larger for productivity prefer higher take-home firms/abroad Formal - None mentioned - Social benefits and pay instead of benefits. (though this is (staff in minimum wage are Enforcement is not a key generally a payroll) too costly and factor – as labor weak factor) discourage hard work inspections are rare. - Makes firing costly 35 Issue Benefits Costs Key rationale for Dominant decisions regarding force driving formalization formali- zation Business Informal - None mentioned - Limits customer base Large differences across Expanding incorpo- (work as (perhaps unaware (cannot sell to the firm size: smaller firms customer base ration natural of the state) see no benefits and fear (selling to the person) requirements for - Lack of a business high costs. Larger firms state and filing under RUS) brand limits already established as larger firms) customer base (large separate legal entities see and improved firms do not consider broad range of benefits access to you credible) and point to limited costs. credit. The requirement to be a Formal - Expand customer - Businesses with 1 to natural person to file taxes Key role (separate base (state, larger 5 workers fear high under the RUS was not played by legal firms, exporting) fixed and recurrent raised, possibly reflecting accountant entity) - Some perceive costs if setting up as the fact that choices about (providing slightly better a separate legal tax regimes are not made information access to credit entity directly by the business about what it owner. entails). - Larger firms note the benefit of protecting own capital Source: World Bank focus groups and in-depth interviews with micro and small businesses. Business licensing 2.7 The procedures for obtaining the municipal license draw heavy criticism from micro and small businesses. Evidence from the focus group analysis suggests that the municipal license is perceived to be an unnecessarily burdensome bureaucratic hurdle that conveys no benefits to the micro and small entrepreneur. The safety inspection (certificado de inspección técnica de seguridad en defensa civil) is perceived more as an opportunity for municipalities to obtain revenue from businesses than as a legitimate means to ensure the safety of operations. Checking compliance with zoning requirements is also perceived to be unnecessarily burdensome, as it is claimed that it may also result in an additional inspection. The business polled did not perceive the benefits of such regulations.7 As an interviewee put it: “my business is located in an industrial park and I still have to pay for them to come and check that this is an area where one can set up a manufacturing business.� There is also the perception that municipal licenses need to be renewed every year, especially among the micro business owners, that a business needs to get a different license for every type of activity that they conduct. Overall, the micro and small entrepreneurs that were polled consider that the municipal license conveys no benefits to the business owner. A small entrepreneur that participated in one of the focus groups for businesses with 5 to 10 workers put it most starkly when he commented on the municipal license: “it is an obligation; it does not give you any benefit.� 7 Public policy has a legitimate role to play in terms of zoning and planning to avoid negative outcomes such as: unrestricted growth that does not follow basic safety codes for construction or operation: or significant levels of pollution. 36 2.8 Businesses have not yet perceived the recent improvements in the municipal licensing framework. To some extent businesses’ perceptions of municipal licensing do not reflect the latest changes in the regulatory framework, by which municipal licenses are to be issued with indefinite validity and can encompass several areas of activity (Framework Law of Municipal Licenses, Law 28976 of January 20, 2007). Evidence from the municipal licensing simplification project Tramifácil, supported by IFC, suggests that improvements in the municipal licensing administration are taking place, as reported in the IFC Municipal Scorecard.8 The cost of the municipal license in 2007 ranged from 1.7 percent of per capita GDP in Lima to 9.7 percent of per capita GDP in Huancayo. These figures are relatively high compared to a selection of 65 municipalities in the Latin America and the Caribbean region monitored by the IFC Municipal Scorecard. In fact Huancayo was the lowest ranked municipality of all 65, and Lima, while being the highest ranked in Peru, was only the nineteenth among the regional sample. 2.9 The probability of being detected is the key factor affecting the decision to obtain a municipal license. Business owners state that the key consideration when deciding to obtain a municipal license is how likely they are to be inspected. Those operating behind closed doors, such as textile workshops that operate in relative quietness, are less likely to obtain a license even as the business grows. In some cases, especially among micro firms, the entrepreneur prefer to pay off inspectors, allegedly for as little as 20 nuevos soles according to information from the focus groups, rather than go through the hassle of obtaining the license. Those that expressed a preference to pay off inspectors said they do so partly because businesses fear that the license will become invalid should they change their activity, for example if they change the type of goods that they sell. As businesses grow, however, the costs of paying off inspectors increase and the cost-benefit calculus prompts business owners to obtain the license. 2.10 The survey results lend support to two conclusions regarding licenses: (i) they are perceived to have only one benefit: avoiding fines, and (ii) they are perceived to pose a heavy bureaucratic burden. Results from our dedicated survey of micro and small businesses confirms that most businesses perceive little benefit from having a municipal license other than avoiding fines and being in compliance with the law (left-panel in Figure 2.2). In addition, the perception that the process to obtain a license is overly burdensome is also reflected in the survey results. Thus, among businesses that do not have a license, factors that can be thought of as reflecting a ‘culture of informality’, such as the fact that no business like mine has a license, are not stated as the main reason for not having a license. This result holds even among the smallest businesses surveyed (right-panel in Figure 2.2) and is unsurprising given the definition of informality that respondents from the in-depth interviews and focus groups have in mind: an “informal� business is one that does not have a RUC number or municipal license and that does not typically give receipts to its customers. 8 A reform to make public zoning clearer was also introduced in the municipality of Lima in parallel with the operating licensing reform. The zoning reform aims to make it easier to find out what activities are permitted in different parts of the city. 37 2.11 There may be other channels, not explored in this report, whereby having a municipal license may convey benefits to businesses, particularly regarding greater access to credit. Not having a municipal operating license increases the probability for a business to be closed down and is therefore a risk factor that, in principle, creditors may want to take into account when allocating credit. We would therefore expect that businesses with a municipal license have greater access to credit. However, as the left- panel in Figure 2.2 shows, improving access to credit is perceived to be the main benefit of having a license only by very few entrepreneurs, and the econometric analysis shown in the next chapter does not refute this view. One possible explanation for these results is that having a municipal license may not affect the probability of getting a relatively small amount of credit. For larger amounts of credit the risks of a business being closed down due to not having a license may become too high and the creditor may start asking for evidence of a license after a certain loan amount. Given the design of our research it is not possible to corroborate or refute this hypothesis, simply because most businesses surveyed may not be sufficiently big to have crossed the threshold where not having a license may be a constraining factor for access to credit. These issues are explored further below. 38 B. What is the impact of informality on access to credit? Figure 2.2: Benefits from having a municipal license and reasons not to have one Percent of businesses Main benefit of having a license Main reason not to have a license (single answer) (single answer) 100 100 100 Other No business like No business like 80 a license mine has mine has a license To access credit 80 80 It is not It is not 60 60 60 compulsorycompulsory To issue receipts 40 40 40 It is too expensive It is too expensive To avoid fines 20 To comply with 20 20 It is too It is too difficult difficult the law to get to get 0 0 0 1 to 5 6 to 10 11 to 50 65 11 to 1 to 5 1 to to 10 6 to 10 50 11 to 50 workers workers workers workers workers workers workers workers workers Number of Number of 250 135 149 171 66 31 respondents respondents Note: Other includes ‘to avoid bribes’, ‘to attract new clients’, and ‘none’. Source: World Bank survey of micro and small businesses. 2.12 Cost-benefit calculations drive businesses’ decision to obtain a license, enforcement plays a key role. Besides asking about the main reason for not having a municipal license, the survey also questioned business owners without a license in more detail about whether certain reasons had played a role or not in their decision not to have a license. The results suggest that factors related to the costs of getting a license were consistently perceived as more important than lack of information or other factors that one can think of as reflecting a ‘culture of informality’ (Table 2.4). A detailed econometric analysis is presented below in section ‘C. Econometric analysis of the determinants of informality.’ For now, it is simply worth noting that the econometric results are consistent with the view that businesses in sectors of activity that are more visible to the authorities are more likely to obtain a license. However, as the regression analysis will show, a variable specifically capturing the probability of inspection by city and by industry fails to significantly explain the probability of a business obtaining a municipal license. This latter result could reflect difficulties in distinguishing econometrically between the information captured in the different variables (the sector- specific variables and the variable with the inspection rate). Additionally, it could also reflect the possibility that, as businesses internalize the higher probability of being detected by being more likely to obtain the license, the municipalities do not need to inspect those businesses all that often. Therefore, ex post, there would be no significant difference in inspection rates by sector. Table 2.4: Reasons for not having a municipal license In brackets the percent of businesses without a municipal license that reported that this factor was one reason why they do not have a license 39 Notes: Based on responses by the 268 businesses without a license. See Table A.11 in the annex for a breakdown by business size. * Cost-benefit factors ‘Culture of informality’ factors  The process is too time – consuming (69 percent)  The costs of operating a business with a More license are too high (63 percent) important  The process is too expensive  My business is too small to have a (56 percent) license (57 percent)  I see no benefits in getting it (55 percent)  The forms are too complicated  There is no obligation to get it (46 percent) (46 percent) Less  No point obtaining it since fines are rare  No business like mine has a license important (37 percent) (42 percent)  To get the license they ask for bribes  I do not know how to obtain it* (36 percent) (40 percent) This factor may also reflect genuine information constraints. Source: World Bank survey of micro and small businesses. 40 Paying taxes 2.13 Obtaining a RUC number is relatively easy, but business owners know little about tax issues and often rely on third-parties for decisions regarding taxes. Evidence from the focus groups suggests that obtaining a RUC number is straightforward and imposes a limited burden. Having a RUC number and paying taxes (on some of the income) are indeed considered to be core aspects of what constitutes a ‘formal’ business. As for paying taxes, focus group participants were in general unable to explain the reasons behind the choice of regime (this contrasted with the relatively clear views on other issues). There appeared to be a limited understanding of the tax obligations under different tax regimes. What emerged was that in many cases, decisions regarding tax issues were delegated by the business owner to an accountant or some other person who is perceived to be better-informed than the business owner. 2.14 Having a RUC number is perceived to have some benefits, mainly through expanding the customer base and improving access to credit. In contrast with the negative view on the municipal license, the focus groups and interviews with micro and small businesses suggest that having a RUC number conveys concrete benefits. Having an RUC number is perceived to allow the business to grow its customer base because one can issue receipts and hence sell to formal businesses. A small furniture manufacturer put it simply: “some customers place large orders and demand a receipt� while another noted that “I was working with firms and needed to issue receipts, that’s why I got the RUC.� In addition, some interviewees perceive that having a RUC number improves access to credit and overall, the positive aspects of having a RUC number are deemed to outweigh the costs, which are mainly the perception of a greater monitoring by the tax authorities. The result that having a RUC number is perceived to have the benefit of attracting clients because of the ability to issue receipts finds support in the survey results (Figure 2.3). Figure 2.3: Benefits from having a RUC number and reasons not to have one Percent of businesses Main benefit of having a RUC number Main reason not to have a RUC number (single answer) (single answer) 100 100 100 Other No business like No business like a RUC mine has mine has RUC no. 80 80 80 number To access credit It is too expensive It is too expensive 60 60 60 To avoid fines It is not 40 To issue receipts 40 40 It compulsoryis not compulsory 20 20 20 It is too difficult To comply with to get It is too difficult the law to get 0 0 0 1 to 5 6 to 10 11 to 50 65 11 to 1 to 5 1 to to 10 6 to 10 50 11 to 50 workers workers workers workers workers workers workers workers workers Number of Number of 273 181 173 120 18 6 respondents respondents Note: Other includes ‘to avoid bribes’, ‘to attract new clients’, and ‘none’. Source: World Bank survey of micro and small businesses. 41 2.15 The survey also provides information on the use of the different tax regimes available to micro and small businesses. Corporate income taxation is structured in a three-tier system: the general regime and two special regimes for MSMEs: the RUS and the RER. The simplified regime (RUS) is aimed mainly at natural persons. The RUS is a single tax (monotributo) that substitutes both the corporate income tax and the sales tax (Impuesto General a las Ventas, IGV). The RER does not replace one’s liability for IGV. The key features of these different regimes are provided in Table 2.5. An overview of tax regimes for small enterprises in Latin America is presented in Table A.1 in the annex. Table 2.5: Overview of corporate income taxation Simplified Regime Special regime General RUS (Régimen Único Simplificado) RER (Régimen Especial de Regime la Renta) Rationale Single tax (monotributo) that replaces Simplified corporate income Tax liability on both IGV and corporate income tax tax but liable for VAT net income  Can expenses be No No Yes deducted from corporate income?  IGV liability? Not liable. Cannot issue receipts Liable Liable (facturas) that generate IGV credit Qualifying criteria  Annual turnover Maximum 360,000 Nuevos soles Maximum 360,000 Nuevos None soles  Assets Maximum 70,000 Nuevos soles Maximum 87,500 Nuevos None soles  Legal entity Only for natural persons (and very Natural persons and legal All restrictions restricted legal entities) entities  Type of activity All, with some exceptions (e.g., transport, All, with some exceptions All show business, notaries) (e.g., transport, construction, medical) Tax liability Fixed monthly quota as follows: 1.5% of turnover for 30% of net Cate- If gross and total Monthly manufacturing and income gory monthly monthly quota commerce; turnover purchases (soles) 2.5% of turnover for services (soles) (soles) 1 < 5,000 < 5,000 20 2 < 8,000 < 8,000 50 3 <13,000 <13,000 200 4 <20,000 <20,000 400 5 <30,000 <30,000 600 Bookkeeping None RUC number. RUC number. requirements Register of sales and 14 accounting purchases, and balances books in total Note: The following changes were introduced In December 2006: (i) the maximum annual turnover to qualify for RUS and RER was raised from 240,000 soles to 360,000 soles; (ii) the tax rates under the RER were lowered for manufacturing and commerce businesses from 2.5% to 1.5% of gross income, and from 3.5% to 2.5% of gross income for businesses in the service sector; (iii) the physical parameters used to determine the eligibility of firms for RUS and RER (number of employees, locations, and electricity consumption) were removed. Source: Based on information provided by SUNAT. 42 2.16 Many businesses with a low turnover file taxes under the general regime. In addition, some businesses with a relatively high turnover file taxes under the simplified and special regimes, as shown in Figure 2.4. This adds to the evidence presented in Table 2.2, by which 24 percent of businesses with 1 to 5 employees file taxes under the general regime. The fact that some small businesses file under the general regime suggests a high level of sophistication in the decisions affecting tax issues. This contrasts, however, with the findings from the focus groups and interviews. How can these two be reconciled? An answer may be found in the role played by the accountant regarding tax filing issues. It appears that whoever makes decisions on taxes for the businesses, and often this may be the accountant, has a sophisticated understanding of the incentives and disincentives behind the different tax regimes. 2.17 Some businesses with a high turnover and employees are filing under the simplified regimes. For a few businesses filing under both the RUS and the RER, the turnover declared to the survey is actually higher than the actual limit for those regimes. Businesses may be choosing regimes for which they ought not to qualify. But, more importantly, the overall distribution by turnover for businesses filing under different regimes (Figure 2.4) is not much different across tax regimes. In our sample, 14 percent of businesses with 11 to 50 employees file taxes using the RUS. The RUS system is an appropriate instrument for taxing micro enterprises. However, the RUS imposes a relatively high tax burden on investment (see Box 3 for calculations of the marginal effective tax rate on incremental investment), which suggests that a factor for larger businesses remaining under the RUS is to pass off as smaller businesses and avoid paying that tax. 2.18 These filing patterns contrast with the goal of having small businesses under the special regimes and larger businesses under the general regime. The overlap of tax regimes for businesses of different sizes and turnovers presents an unclear picture of the underlying motivations of choosing one regime over another. In addition to tax evasion, mentioned above in the context of larger businesses’ filing under the RUS, other factors also appear to play9 a role. In fact the survey provides some evidence on the importance of ‘having always filed under that regime’ as a reason behind the choice of tax regimes.10 10 The reason “I have always been in that regime� is the second-most important reason for businesses filing under the RUS or the RER and the third-most important for those filing under the general regime. See the detailed results in Table A.9 in the annex. 43 Percent of businesses Percent of businesses Percent of businesses 10 15 20 25 30 35 40 45 50 10 15 20 25 30 35 40 45 50 10 15 20 25 30 35 40 45 50 0 5 0 5 0 5 up up up t o2 to to 0 20 20 20 20 20 to t o4 t o4 40 0 0 40 40 40 to t o6 t o6 60 0 0 60 60 60 to to to 80 80 80 80 80 80 to to to 10 10 10 0 0 0 10 10 10 0t 0 to 0 to o1 12 12 20 0 0 12 12 12 0t 0t 0t o1 o1 o1 40 40 40 14 14 14 0t 0t 0t o1 o1 o1 60 60 60 16 16 16 00 00 00 to t o1 t o1 18 0 80 80 18 18 18 0t 0t 0t o2 o2 o2 00 00 00 20 20 20 0 to 0 to 0 to 22 22 22 Source: World Bank survey of micro and small businesses. 0 0 0 22 22 22 0t 0 to 0 to o2 24 24 40 0 0 24 24 24 0t 0 to 0 to 44 o2 26 26 60 0 0 26 26 26 0t 0t 0t Percent of businesses filing within each of the tax regimes o2 o2 o2 80 Thousand Nuevos Soles 80 Thousand Nuevos Soles 80 Thousand Nuevos Soles 28 28 28 0t 0t 0t o3 o3 o3 00 00 00 30 30 30 0 to 0t 0t 32 o3 o3 0 20 20 32 32 32 0 to 0t 0t 34 o3 o3 0 40 40 34 34 34 0t 0t 0t o3 o3 o3 60 60 60 36 36 36 0t 0t 0t o3 o3 o3 80 80 80 38 38 38 0t 0t 0t o4 o4 o4 00 00 00 40 40 40 0t 0t 0t o4 o4 o4 20 20 20 42 42 42 0 to 0 to 0 to 44 44 44 Distribution of businesses filing under the General Regime (sample = 304) 0 0 0 Distribution of businesses filing under the special regime, RER (sample = 86) 44 44 44 0 to 0 to 0 to 46 Distribution of businesses filing under the simplified regime, RUS (sample = 228) 46 46 0 0 0 46 46 46 0t o4 0 to 0 to 48 48 Figure 2.4: Distribution by turnover of businesses filing under different tax regimes 80 0 0 48 48 48 0t 0t 0t o5 o5 o5 00 00 00 2.19 The low take-up rate of the special tax regime (RER) calls for a review of tax regimes, especially since the RER has some features that make it the most conducive regime for investment. The special tax regime (RER) is in place mainly to protect the VAT chain, as it requires businesses to comply with the VAT and to pay a presumptive turnover tax. The analysis of the marginal effective tax rate suggests that the RER is in fact the most conducive for investment (see Box 3). Yet, in our sample only 14 percent of businesses with a RUC number file under the RER. Why does it have such a low take-up rate? Box 3: Understanding the incentives of tax regimes: the Marginal Effective Tax Rate of the different corporate income tax regimes in Peru How does the tax system affect the decision by economic agents to invest in capital? This box addresses this question by providing a measure of the tax burden that a micro entrepreneur faces when making an incremental investment. It does so by calculating Marginal Effective Tax Rate (METR) on capital. The METR is a summary measure of the effective rate of tax imposed on the rate of return generated by the last, or marginal, unit of capital that a firm invests in. The METR therefore summarizes the effect of the total distortion in the rate of return to capital imposed by the entire business tax system (through the rates of tax, depreciation, capital taxes, special regimes, etc.). The METR exploits the fact that, at the margin, an investment must earn a rate of return after the payment of all business taxes which is equal to the hurdle rate of return required by the business stakeholders. The derivation of the METR formula can be quite complicated (see Annex G), but the idea can be conveyed with a simple example. Say the hurdle rate of return, i.e., the minimum rate of return required by the stakeholders in the project after the payment of all business taxes, is 10 percent. Say the business tax system is such that in order to earn a rate of return of 10 percent after business taxes, an investment must earn a rate of return of 15 percent before business taxes. The METR is then 33 percent (determined as (0.15-0.10)/0.15). The METR therefore measures the share of the investment’s pre-tax required rate of return needed to cover the tax costs associated with the investment. This box provides estimates of the METR for corporations resident in Peru for four broad classes of capital: equipment, building, land, and inventory. We can then calculate the METR across sectors with varying degrees on the use of different classes of capital: manufacturing, agriculture, finance, tourism and mining (the actual capital mix used is based on international survey data). The METR is calculated for small businesses (assumed to source their funds domestically) under the three possible tax regimes (RUS, RER, and general regime) as well as for large businesses (assumed to have access to funds in international financial markets). The table below provides the main results of the analysis (see Annex H for details of the data and assumptions used). 45 Marginal Effective Tax Rates for corporations in different sectors of activity Manufacturing Tourism Agriculture Mining Finance Small businesses RUS 44 28 27 48 39 RER 11 12 11 15 12 General Regime 28 35 19 31 32 Large businesses 24 32 15 36 28 Source: World Bank staff calculations. The first thing to note is that METRs vary widely from the standard corporate income tax rate of 30 percent. Taking the case of large businesses as the benchmark, the accelerated depreciation rate of 20 percent on straight line basis lowers the METRs for manufacturing. On the other hand the low depreciation rate for buildings of 3 percent on a straight line basis raises the METR for investments in the tourism sector as this sector is more building intensive in its asset pool. The METRs for investments in the mining sector are higher than the corporate tax rate due to the royalty rate of 3 percent for large mining business and despite the accelerated depreciation for buildings. The agriculture sector has an METR just equal to the lower corporate tax for this sector of 15 percent with depreciation benefits and penalty just about balancing each other. As large businesses can offset VAT on inputs from VAT on sales the VAT imposes no burden on capital. Second, the METRs for small businesses under the general regime are higher than for large businesses reflecting the different assumptions about the source of funds (though in the case of mining it is actually lower as they pay a lower mining royalty of 1 percent). Third, businesses filing under the RUS face the highest METRs in most sectors of activity. The METRs are particularly high in sectors that use relatively more machinery and equipment, which is assumed to be sourced from the formal sector and hence liable to VAT. In those cases, the burden on investment from not being able to offset VAT on inputs is quite high. In manufacturing, the VAT exemption accounts for 31 percentage points (ppts) of the overall 44 percent METR. The equivalent figures for the VAT exemption for the other sectors are: 15 ppts in tourism and agriculture; 31 ppts in mining; and 25 ppts in finance. Fourth, the METR is lowest for small businesses filing under the special regime (RER). This is primarily because businesses under this regime face lower input taxes on their machinery and equipment and also a lower turnover tax as compared to the RUS regime. Even though businesses cannot deduct expenses from corporate income (as they would under the general regime), the lower turnover tax results in more beneficial treatment than under the general regime. The METRs under the RER were somewhat higher prior to the December 2006 reform (on average 3 percentage points higher) but still the METRs were the most favorable of all regimes even before the turnover tax rates were lowered. 2.20 Compliance costs may be a reason for the low take-up rate of the special tax regime (RER), but no estimate of tax compliance costs exists. The requirements under the RER for book keeping and complying with IGV impose compliance costs on business. Given the advantageous tax rate treatment, it would appear that high compliance costs are offsetting the advantages of the RER. Businesses for which complying with accounting requirements and IGV filing may be too cumbersome would 46 remain under the RUS, thus discouraging businesses from progressing from one regime to another. Businesses that are able to meet the accounting and IGV requirements may also be in a position to meet the requirements under the general regime. 11 Little is known, however, about the actual compliance costs of meeting the accounting requirements vis- à-vis the costs of complying with the IGV filing requirements. Measuring the costs of compliance, a request from the Government on which the World Bank has already started to work, for each of the separate regimes will help to identify administrative measures to reduce those costs and to help define separate regimes. 2.21 The key factors driving the choice of tax regime are business characteristics such as size and being incorporated as a separate legal entity with limited liability. The choice of tax regime among the surveyed firms was explored econometrically through probit and multinomial logit models. These models help us to identify the factors that, controlling for a large number of characteristics such as firm size, age, etc., have a significant impact on the probability that a given business would choose one tax regime over another. The results from these analyses suggest that firm characteristics are more important than the characteristics of the owner such as the level of education. In particular, larger businesses and those incorporated with limited liability are more likely to file under more sophisticated regimes. The result that owner characteristics do not play a role is consistent with the insight from the focus groups that tax decisions affecting micro and small businesses are often made by an accountant. 2.22 Tax evasion motives may also be playing a role in the choice of tax regimes. The survey data also allows us to indirectly explore the impact of tax evasion on the probability of filing under one regime or another. We do so by testing whether proximity to a SUNAT office affects the choice of tax regime, on the basis that businesses located closer to a SUNAT office are in fact more likely to be inspected and hence their propensity to evade taxes maybe lower (see Annex B). In this regard, businesses that are closer to an office of SUNAT are relatively more likely to file taxes under the special regime (RER) (strictly speaking, businesses are less likely to file under the RUS and less likely to file under the general regime). One interpretation of this result is that, faced with a greater probability of being inspected, tax evasion motives may play a lesser role, diminishing the attraction of the RUS (prone to under-reporting of sales given limited inspections) and the general regime (where sales can be netted with expenses to end up with little or no profits on which taxes are levied). 11 Although the marginal effective tax rate under the RER is lower than under the general regime, the possibility of deducting expenses from corporate income may be particularly attractive to businesses as it may make tax evasion easier. The result from the econometric analysis discussed above would be consistent with this interpretation that the RER would not be the tax regime ‘of choice’ for evasion purposes. 47 Table 2.6: Results from econometric analyses of the choice of tax regime Based on Choice Main results analyzed - No effect of education of owner RER - More likely under RER if business has limited liability vs. - More likely under RER in wood products (compared to restaurants) RUS - More likely under RER in Arequipa, Trujillo, Cusco (compared to Lima) Marginal effects from - More likely under RER for larger businesses (cf. to 1 to 5 workers) probit - University educated more likely to be in the general regime (compared to analyses RER owners with less than secondary education) vs. - More likely under general regime if business has limited liability General Regime - Less likely under general regime in shoe/leather (compared to restaurants) - Less likely under general regime in Arequipa, Trujillo, Cusco (cf. to Lima) - No effect of any owner characteristics RUS - Less likely if business has limited liability vs. others - Less likely in businesses with 6 to 10 or 11 to 50 workers - Less likely the closer the business is to a SUNAT office Marginal - More likely in businesses with 6 to 10 workers than 5 or less effects from RER - More likely in Arequipa than Lima multinomial vs. others - No other factor is significant. The limited sample size (86 businesses) may logit model be a factor affecting results of choice among the - No effect of any owner characteristics three - More likely among larger businesses (11 to 50 workers) and those with alternatives General limited liability regime - Less likely the closer the business is to a SUNAT office vs. others - Less likely in Huancayo and Cusco than in Lima - More likely if business is in metal products, less likely in shoe/leather Note: See Annex B for estimation results. Source: World Bank survey of micro and small businesses. 2.23 The limited understanding by MSEs of the requirements under different tax regimes is not helped by a complex web of different definitions of what it means to be an MSE. At the moment there is no single definition across different ministries of the GOP of what a MSE is. In particular, the criteria used by the SUNAT and the MTPE are only partly consistent (see Table 2.7). The current co-existence of multiple and partly overlapping definitions of what an MSE is may lead to confusion and act as a further barrier to formalization. 48 Table 2.7: Definitions of micro and small enterprises in the labor code and correspondence with tax regulations Maximum Maximum Other criteria number of gross sales workers Microenterprise 525,000 Labor Ministry 10 - No nuevos soles* - Natural persons only (with few exceptions) Tax (RUS 360,000 - Limit on assets (70,000 nuevos soles) None requirements) nuevos soles - Restrictions on certain sectors (e.g., transport, show business, notaries) Small enterprise 2,975,000 Labor Ministry 50 - No nuevos soles* - Limit on assets (87,500 nuevos soles) - Restrictions on certain sectors (e.g., Tax (RER 360,000 None construction, medical, transport) requirements) nuevos soles - RUC number and keeping register of sales and purchases, and balances Note: * Strictly speaking the limit is set at 150 tax units (UITs) for micro enterprises and 850 UITs for small enterprises. The equivalent amount of a UIT in nuevos soles is announced annually by SUNAT. For 2008 the UIT amounts to 3,500 nuevos soles, up from 3,450 nuevos soles in 2007, and 3,400 nuevos soles in 2006. Source: SUNAT and MTPE. Labor relations and access to pension12 2.24 Businesses perceive only costs from formalizing their labor relations, while informal arrangements are valued in part for the flexibility they provide. Focus groups and interviews suggest that the costs of formalizing labor relations are high. The perception from businesses is that costs would be too high in terms of holidays, social benefits, as well as firing costs. In addition, businesses state that formal labor relations provide a disincentive to work. Conversely, informal labor relations not only provide the advantage of avoiding having to pay social benefits but also are valued in part because of the flexibility that they provide to the business owner. In particular, piece rate work is valued by many businesses as an arrangement that ensures the right incentives for the workers, an arrangement that they do not see how they could marry with entitlements such as minimum wage and holiday pay. A flexible contractual arrangement is also 12 This section deals with labor issues in so far as the focus groups and survey results provide new evidence. For a broader coverage of labor issues and informality the reader is referred to the report of the first phase of this programmatic study (World Bank, 2007a). 49 sought because of the high fluctuations in workload that small entrepreneurs face. As a small furniture manufacturer put it: “customer orders and output are not constant. It is rare that a workshop could bear having three or four permanent workers; in peak times it can reach up to 30 or 50 workers, and usually maybe eight to 10…� 2.25 Yet labor regulations are at the bottom of the list of obstacles for micro and small businesses, perhaps because respondents to the survey do not find these regulations binding. When asked about a long list of possible obstacles to their business, those most highly ranked are crime, corruption, economic instability, and political instability, the only four factors which are considered to be serious obstacles by more than half of our sample of micro and small businesses (Table A.6 in the annex).13 In contrast, labor regulations are ranked at the very bottom of a list of 15 factors. At first sight, this would appear to suggest some contradiction with the finding from the focus groups that micro and small entrepreneurs only see costs in formalizing labor relations. The apparent paradox can be explained by the fact that employers have most workers outside formal labor relations and therefore are de facto outside the scope of labor regulations. In other words, in practice labor regulations may not be perceived as a serious obstacle simply because they are largely bypassed. 2.26 Larger businesses include only a small share of their workers in their payrolls. The resistance to formalize labor relations does not decrease as businesses grow in size and become more complex. In short, larger businesses are more likely to have at least some workers on the payroll, with health insurance and pension rights, but the share of those on the payroll may be low. Businesses of 11 to 50 workers typically have a very complex mix of workers. They typically have a core group of stable workers but only a small portion of those core workers are in fact on the payroll. 2.27 The complexity in labor arrangements grows with firm size, as businesses report a preference for informal practices, even for workers nominally on the payroll. Among workers on the payroll, the officially reported pay is often supplemented with informal payments. Other core workers are systematically hired as providers of services, often issuing receipts to the business for the provision of their services as allegedly independent contractors. Finally, the truly non-core part of the workforce, which can be large in some sectors of activity such as textiles that are highly seasonal, – is typically employed on a piece-rate basis without any formal arrangement. Some of the business owners also emphasize that the informal labor relations are in fact demanded by workers, who prefer to increase the take-home pay rather than having access to certain benefits. In some cases, the provision of basic health care is perceived as a benefit for the business, since it helps to keep a core group of workers happy. However, business owners prefer to provide some support for those core workers when contingencies occur or, more rarely, provide for the purchase of health insurance in the private market. 13 It is noteworthy that economic instability ranks so highly, given real GDP growth of 9 percent and inflation of 3.9 percent in 2007. A possible reason for this result is that memories of past episodes of economic crises are still very much in the minds of micro and small business owners. 50 2.28 Survey evidence confirms that even businesses with 11 to 50 workers are mostly informal with regard to labor issues. Labor relations are by far the most informal dimension of informality found in this study. In addition, the process of firm growth does not provide for a set of circumstances that make formalization of labor relations more likely. As shown in Figure 2.5 below, the proportion of workers with health and pension coverage is not higher among larger businesses in our sample. It is therefore unsurprising that having workers on the payroll is not perceived to be a part of the implicit definition of what is a ‘formal business.’14 Figure 2.5: Relationship between the number of workers and those with health coverage and pension Each dot represents a business – those with all workers covered are on the 45 degree line Vertical axis: Vertical axis: number of workers with health coverage number of workers with pension 50 50 40 40 30 30 20 20 10 10 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Number of workers Number of workers Source: World Bank survey of micro and small businesses. 2.29 Survey evidence suggests that the high cost of providing social benefits is perceived to be a major factor in explaining informal labor relations. The dedicated survey of micro and small entrepreneurs queried respondents in some detail about the reasons for not having workers on the payroll (see detailed results in Table A.15 in the annex). In addition to the fact that there are plenty of workers willing to do so, the most highly ranked response was the fact that one would have to pay the workers’ holidays and other benefits. Among the lowest ranked responses were concerns about the paperwork that would be necessary to have workers on the payroll and factors capturing the ‘culture of informality’ (in particular ‘no business in my sector has its employees on the payroll’ and ‘you do not need to include family members on the payroll’). These results suggest that concerns about the costs of providing social benefits are among the most critical 14 There is also evidence that at least some workers without a contract, health insurance or a pension do not consider themselves to be informal: a survey of 400 individuals conducted by the Grupo de Opinión Pública de la Universidad de Lima in April 2008 indicated that only 43 percent of those employed reported having an informal job (compare this with the ENAHO survey results of more than 70 percent of informality under any measure used, see Table 1.1). While other factors may help to explain this result, it suggests that, as in the case of the entrepreneurs polled in our survey, the implicit definition of what constitutes informality among workers does not necessarily match the standard definitions. 51 factors preventing the formalization of labor relations. Yet, when the survey asked open- ended questions about the major obstacles to business growth, labor regulations are not generally perceived as a serious obstacle (see Table A.6 in the annex). However, this may reflect the fact that they simply ignore them. 2.30 Although few know about the true costs of providing social benefits under the Special Labor Regime for micro businesses. The dedicated survey of micro and small businesses allows us also to explore in detail what micro businesses really know about the Special Labor Regime for micro businesses. Few businesses know of the special regime, especially among those businesses with 10 employees or fewer, precisely those who can potentially benefit from this regime. In addition, the survey also showed that, among those that know of that special regime, they underestimate the savings that it could provide for them (see Box 4 for details). Box 4: What do businesses know about the Special Labor Regime for micro firms? A Special Labor Regime (SLR) was put in place in 2003 (Law 28015) to ease the labor regulatory burden of micro firms (i.e., those with up to 10 employees and sales of 150 tax units a year: equivalent to 525,000 nuevos soles in 2008). The key features of the regime are that employers only need to make contributions for health care and for 15 days of holiday a year. This compares with the employer contributions required under the general labor regime: family compensation; extraordinary bonus salaries paid in July and December; the compensation of time in service; and a higher 30-day holiday period. In addition, under the SLR signing up for a pension scheme is voluntary. As of December 2007 there were 33,289 businesses registered under the SLR for micro businesses. This figure is low given the large universe of businesses that could potentially sign up for the SLR. What helps to explain this low take up rate? Evidence from the survey suggests that lack of information about the SLR may help to explain this low take-up rate. Of the 802 businesses surveyed only 85 respondents (11 percent) knew about the SLR. The lack of information is even more pronounced among businesses with 10 workers or fewer, those that can potentially make use of the regime: out of the 622 businesses with 10 workers or fewer surveyed only 57 respondents (9 percent) knew about the SLR. Of those 57 respondents, there are 13 businesses which declare to have at least some workers under the SLR. In addition there are 11 businesses that in principle should not be able to qualify for the SLR (since they report 11 or more workers and therefore can no longer be considered micro businesses) and that have employees under the SLR. This is a notable result given that in our sample there are only 180 businesses with 11 or more workers. Thus, in our sample the take-up rate of the SLR is substantially higher among firms that fit into the SLR only through under-reporting of employees. In some cases they report having as many as 18, 21, 24, and 26 employees to the survey. However, the survey uncovered limited evidence that the SLR led to a benefit erosion, as only 5 of the 24 businesses that have workers under the SLR reported that they had been previously registered under the general labor regime. In most cases, the workers who were signed up under the SLR had not been previously registered with the Ministry of Labor. Knowledge of the SLR is greater among businesses that participate in some type of state- sponsored program (see Table A.10 in Annex A for a list of programs about which the respondents were asked). However, knowledge is far from universal across the businesses that participate in state programs: out of the 57 businesses that report participating in some state program only 14 knew of the SLR (around 25 percent, compared to around 11 percent of 52 businesses knowing of the SLR for the entire sample). This suggests that there is the potential for improving awareness about the SLR by providing information to micro and small businesses that come into contact with the state through other programs. Even those respondents that know of the SLR know little about it. Of the 85 business owners that know about the SLR only a third (28) provided an answer on the amount of the monthly contributions borne by the employer for a worker that gets paid the minimum wage. The mean response by these 28 business owners is that for a worker getting paid the minimum wage and registered under the SLR they would need to pay 184 soles per month. This figure is far below the 307 nuevos soles per month that they estimate they would need to pay under the general labor regime, but well above the 68 nuevos soles that they would need to pay in reality (calculations by the Ministry of Labor made for the previous minimum wage rate of 500 nuevos soles per month). While any result on the basis of such few observations has to be interpreted with great caution it may be noted that the estimate of the respondents on the monthly contribution by an employer under the general regime (307 nuevos soles) is much closer to the actual cost as calculated by the Ministry of Labor (303 nuevos soles). The results do not change if we consider only those businesses that actually have workers registered under the SLR. The above results suggest that decisions regarding the labor regime, as in the case of decisions about tax regimes, may not be made directly by the micro business owner but perhaps are implicitly made by the person to whom the administrative tasks are outsourced. This may have implications about the prospects of any information campaign and the appropriate channels for disseminating information. In this regard the focus groups also provided some insight regarding the lack of knowledge of the SLR. Among the micro business owners that participated in the focus groups none knew about the SLR. In addition, when the facilitator explained some characteristics about the regime, there was no interest among micro business owners to find out more about it. It appeared as if micro business owners have well-defined channels through which they get their information, mainly through peers and the accountant, and are not accustomed to using other sources or are reluctant to trust any other sources. Overall, the evidence from the survey and the focus groups confirms that there is very little knowledge of the SLR and the contributions that would need to be borne by employers. However, the evidence also cautions against any optimistic view that a general information campaign may increase the use of the SLR. Any strategy for disseminating information about the SLR would first need to identify more clearly how decisions regarding the compliance with regulatory issues are made by micro entrepreneurs and how the information would be used. 2.31 Survey results suggest that the high level of the minimum wage may be a constraint to the formalization of at least half of the informal workforce. The dedicated survey of micro and small entrepreneurs asks respondents about specific reasons why workers are paid less than the statutory minimum wage. While the most important factor is the availability of labor willing to work for less than the minimum wage, the second issue with the highest score is the claim that the business is not profitable enough to allow for payment of the minimum wage. This is particularly the case among the smallest businesses with 1 to 5 workers (where 71 percent of respondents agree that the business cannot afford to pay the minimum wage) but it is also true for the larger businesses (51 percent of businesses with 11 to 50 workers report that the business simply cannot afford to pay the minimum wage). (See detailed results in Table A.14 in the annex.) A comparison between minimum wage and distribution of earnings is 53 presented in Figure 2.6 below. The clear spikes in the wage distribution of both formal and informal workers that can be observed in 2003 may be indicative that the minimum wage has a wide-reaching effect, as the minimum wage may be used to lock in wages to the minimum (Kristensen and Cunningham, 2006). Figure 2.6: Kernel distributions of monthly income for private salaried workers – Metropolitan Lima Vertical line represents the minimum wage 2003 2005 1 Parzen Kernel Density Function .5 0 2 4 6 8 10 Logarithm of Worker's Monthly Income Informal Salaried Workers Formal Salaried Workers ENAHO 2005 Source: 2003 from Jaramillo (2004), 2005 World Bank (2007a). 2.32 The minimum wage may be indeed a binding constraint that would make formalizing these jobs non-viable, as argued in the first phase of this study (World Bank, 2007a). In this regard, the statutory minimum wage, measured in relation to the productivity of workers in a country, is higher in Peru than in other countries like Mexico, Uruguay or Brazil (see Figure 2.7). Evidence from Latin American countries suggests that the minimum wage has far-reaching effects, particularly in cases where it is sufficiently high so as to distort the wage distribution in the formal sector, as Figure 2.6 shows for the case of Peru (Maloney and Nuñez, 2001). Figure 2.7: Minimum wage in Peru and selected countries in the region Mandated minimum wage as a percent of the average value added per worker 70 60 50 40 30 20 10 0 Bolivia Colombia Brazil Venezuela Mexico Ecuador Argentina Peru Uruguay Paraguay Note: Data on the statutory minimum wage is collected annually. The chart reflects the latest data available for each country. Source: Doing Business 2008. 54 2.33 In some businesses and for some workers the minimum wage does not appear to be the most critical constraint that keeps labor relations from being formalized. The survey also asked respondents to provide an estimate of the proportion of workers that, in the area of activity of the respondent, earn less than the minimum wage. Additionally, respondents were also asked about the proportion of workers that they estimate are not in the payroll, again in their area of activity. The results show that the share of workers earning less than the minimum wage is indeed lower than the share of workers that are outside the payroll (see Table 2.8 below). This suggests that the minimum wage may not be the binding constraint preventing the formalization of the labor relation between the employer and the employee at least for those workers that earn above minimum wage but are not on the businesses’ payroll. Table 2.8: Business owners’ estimates of how common it is for workers to earn less than the minimum wage and not be on the payroll Percent of workers 1 to 5 6 to 10 11 to 50 All workers workers workers businesses Estimated share of workers earning less than the 62 52 45 56 minimum wage Estimated share of workers not on the payroll 75 68 61 70 Difference between estimated share of workers on the 13 16 16 14 payroll and earning less than the minimum wage Note: See Table A.13 and Table A.15 in the annex for the wording of the questions asked and further breakdowns by sector and city. Source: World Bank survey of micro and small businesses. 2.34 In some cases the minimum wage (and holiday pay) impose not only a high burden but businesses find them hard to reconcile with piece-rate pay arrangements as they fear disincentive effects from minimum wages. An insight from the focus group was the preoccupation among micro and small business owners with the disincentive effects of formal labor relations, and in particular minimum wages. Micro and small entrepreneurs also find it difficult to reconcile their preference for paying on a piece-rate basis (al destajo) with formal contracts. The ENAHO allows us to explore further this issue. To this end, we compare, for workers of a given income, the proportion of those that are informal within the sub-groups of piece-rate workers with workers under any other pay arrangement. To ensure greater comparability across the two sub-groups we only use blue-collar workers as our starting sample. As shown in Figure 2.8, the prevalence of informality, conditional on income level, among workers on piece-rate pay is higher than among non piece-rate workers, especially at the higher end of the income distribution. While no causal links are explored here, this evidence from the ENAHO is consistent with the view that this type of pay arrangement has an impact on whether the labor relation will be formalized or not. 55 Figure 2.8: Informality prevalence for workers, by pay system (piece-rate or not) Informality rate measured by having a pension. Non piece rate worker 1 .8 .6 .4 Informality prevalence .2 0 Piece rate worker 1 .8 .6 .4 .2 0 0 10 20 30 40 50 60 70 80 90 100 Income centile Notes: Workers at a given income level are classified as piece rate workers or not and the proportion of those without a pension is then calculated. The calculation is based only on blue-collar workers (categories 4 and 5 from the ENAHO – see Annex A). Source: World Bank staff calculations based on ENAHO. Data refer to 2006. 2.35 In some cases the worker may prefer an informal labor relation. The micro and small entrepreneurs that participated in the focus groups often suggested that workers also opt for informal labor relations to maximize take-home pay at the expense of social benefits. While this research was not designed to test such a hypothesis, evidence from the ENAHO on the reasons for self-employment suggests that the motivation for higher earnings and flexibility are key to explaining self-employment (and conceivably they may also help to explain the preference of some workers for informal labor relations to maximize take-home pay, see Table 2.9).15 This preference for an informal labor relation therefore suits both the worker and the employer. A small business owner interviewed summed it up as follows: “the worker wants to have the money cash in hand, he doesn’t want to contribute because he doesn’t believe in the state’s health benefits, and, on the other hand, the business owner is not concerned because he sees formal labor relations as increasing the costs.� 15 In the survey of the Universidad de Lima noted in footnote 14 the main reasons why respondents want a formal job have to do with: security (33 percent want it because of greater security and 15 percent to have fixed income); and higher pay (33 percent), more so than access to health insurance (18 percent). 56 Table 2.9: Reasons for self-employment 2002 2003 2004 2005 2006 Involuntary (could not find salaried employment) 40 40 34 33 28 Voluntary Higher earnings 30 26 29 29 25 Has more flexibility 16 21 26 23 27 Family tradition 5 4 5 7 7 Other 8 9 6 8 13 Sum of all voluntary reasons 60 60 66 67 72 Source: World Bank staff calculations based on ENAHO. 2.36 Firing costs are also perceived to be high and a deterrent to including workers on the payroll. The cost of firing features also prominently among the reasons that micro and small entrepreneurs provide for not having workers on the payroll. The cost and difficulty of firing are indeed the third-highest rated factor in explaining labor informality according to the respondents to our dedicated survey of micro and small businesses (see Table A.16 in Annex A for detailed results). These concerns are particularly acute among the micro entrepreneurs with 1 to 5 workers or 6 to 10 workers, for whom it is not only the cost of firing by also how complex it actually is to fire a worker. In this regard, the complexity of the process adds to uncertainty as to the actual costs of terminating an employee’s contract (an issue which is also explored in Box 5 in relation to the wedge between the severance payments to which dismissed workers are entitled and the actual payments that they receive). Moreover, if we take into account that many dismissal processes do not end in a conciliation but rather go on to the courts, the uncertain length of the litigation adds to the uncertainty about the actual severance payments that a business may have to pay a dismissed worker. Box 5: Severance payments de jure and de facto Dismissed workers are entitled to a variety of benefits ranging from back holiday pay to compensation of time in service (CTS). They may also include any accrued but yet unpaid bonuses for July and December, as well as redundancy compensatory payment. Upon being laid off the worker can request an estimate of these entitlements from the Ministry of Labor. Should the worker be dissatisfied with the resolution of the dismissal, the worker can make use of the services provided by the Ministry of Labor to arrange for a conciliation hearing between the worker and the employer. Attendance by the employer at these conciliation hearings is compulsory. The purpose of these conciliations is to reach a mutually satisfactory agreement so that the matter can be settled without recourse to the labor tribunal. Evidence from more than 1,700 conciliations in metropolitan Lima was compiled for this study and cross-referenced with the initial estimates provided by the Ministry of Labor’s workers’ entitlements. This data had not been previously processed (due to the fact that the two data sets were collected separately without a direct link to cross check them). Results indicate that on average the actual settlement received by dismissed workers amounts to around 64 percent of the payments to which the worker is entitled according to the initial estimate provided by the Ministry of Labor. Results by type of worker (no breakdown by sector was available) suggests no large differences between the share received by employees and by 57 workers. Unfortunately the outcome of the conciliations specifies only the total settlement but not a breakdown by type of benefit. The results that the severance payment that workers receive de facto is substantially lower than what they are entitled to de jure has implications for the debate on whether firing costs in Peru are high or not (an issue which was also touched upon in the report of the first phase of this programmatic study). To the extent that this adds to the uncertainty about what the actual firing costs may turn out to be, they may act as a deterrent to formal employment. Perhaps more relevant than the average ratio of the de facto to de jure severance payments is the distribution among cases. This is because the more uncertain the wedge between the two measures the more negative its consequences. The evidence suggests that there is indeed a wide dispersion of this wedge between entitlements and settlements, as shown in the chart below. This distribution is shown below for the two types of workers that are recorded in the system (‘employees’ or white-collar and ‘workers’ or blue-collar). This suggests some uncertainty about the eventual outcome of a dismissal, – even those that are settled by mutual accord through the conciliation process. Employees Workers All 20 Share of workers (percent) 15 10 5 0 0 25 50 75 100 125 150 0 25 50 75 100 125 150 0 25 50 75 100 125 150 Ratio of settlement to amount owed (percent) Source: World Bank staff calculations based on data from MTPE. This finding about the uncertainty of the outcome of conciliation processes is supported by the fact that the wedge between entitlements and settlement does not show any distinct pattern when mapped against any worker characteristics (gender, age, and length of tenure in the firm) 58 for which the dataset contains information. However, caution must be exercised in interpreting the wedge between the entitlement and the actual settlement. The initial estimate of the entitlements is prepared by the Ministry of Labor without prejudice and on the basis of information provided by the worker, information that cannot always be verified. In the conciliation process there is an element of discovery of facts, as the employer may provide documentary evidence challenging the basis for the initial estimate. In some cases, as shown in the chart, some workers actually receive more at the settlement than was initially estimated. While these cases are rare they underscore the fact that the process is not only a mere negotiation whose outcome is determined by the relative bargaining power of the two parties. 2.37 In addition, sector-specific labor regulations may add to the disincentives to formalize in particular sectors of activity. The burden of labor regulation in Peru is high even when compared with other countries in Latin America, as the first phase of this study documented (see Table A.2 in Annex). This burden is particularly high in a number of sectors which face a number of sector-specific constraints, in addition to the general regulatory framework. While no attempt to identify which of those specific constraints may be most constraining, this study took a first step towards exploring those issues by concentrating on documenting the actual number of legal instruments of a labor regulatory nature that are applicable to specific sectors of activity, and mapping that information with the actual informality rate observed in that sector of activity (Box 6). Box 6: Sector-specific labor regulations and the prevalence of informality This box provides a first look at how the number of labor regulations affecting specific sectors of activity may be related to the level of informality in that sector of activity. A total of 83 labor regulations were identified as being sector-specific (see Annex F for details). As shown in the charts below, this data calls for further examination of whether the relatively large number of sector-specific labor regulations on sectors that already have a high degree of informality, such as construction, may be contributing to informality. Informality rate across sectors and number of labor regulations issued during 2001 to 2007 specifically for a given sector of activity Informality rate as measured by percent of Informality rate as measured by percent of workers without a pension workers without a contract 15 Construction Construction 15 Other business Other business 10 10 Ports Ports Health Health Number of laws Number of laws Fishing Fishing Mining Other services Mining Other services 5 5 Private households Private households Hydrocarbons Recreation Energy Recreation Hydrocarbons R&D Energy Water transport Water transport R& Agriculture Agriculture D 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Informality rate Informality rate Note: See Annex F for details on how the number of laws affecting labor regulations in specific sectors was constructed. Source: World Bank staff calculations based on Ministry of Labor database of laws and ENAHO. 59 Business incorporation as a separate legal entity 2.38 Incorporating the business as a separate legal entity is perceived to be an important step in the formalization process. Participants in the focus groups often noted that moving from operating as a natural person to incorporating the business as a separate legal entity was a key step in what they consider their formalization process. Of particular interest in the focus groups was the fact that the perceptions of costs and benefits of incorporating a business were quite different between the group that had already incorporated and those that were still operating as natural persons. This difference in the assessments of the ‘haves’ and the ‘have-nots’ was a distinct feature of this dimension of informality. As an example of this contrast, a micro entrepreneur with five to 10 workers noted that incorporation “was good for nothing� while a number of small entrepreneurs noted many benefits of incorporation, including: “to be able to take part in public tenders� and “to protect your personal capital.� This difference in perceptions was in stark contrast with most other issues discussed in the focus groups where, despite there being businesses on both sides of the formality divide, businesses typically agreed on what were the main pros and cons of the different choices. The importance of incorporating the business as a separate legal entity was also underscored in three case studies of the trajectory of businesses towards formalization that were commissioned for this study (and which will be introduced and discussed in section D. Dynamics out of informality). 2.39 Businesses with legal personality perceive that they are better positioned to sell to large companies and to the state, and to have better access to credit. Micro and small entrepreneurs that have already incorporated their businesses consider that the separate legal personality has helped them in marketing terms. The perception is that this greater credibility allows businesses to sell to a different type of client, particularly bigger firms and also to the state. Evidence from the survey shows some positive correlation between the extent to which businesses are integrated in larger supply chains and measures of formality, including operating the business as a separate legal entity (Table A.8 in the annex). Selling to the state, however, is perceived to be particularly difficult as it involves additional fixed costs simply to try to do so. The fee required to appear in the Registry of Providers to the State (RPS) managed by CONSUCODE is non-negligible for micro and small entrepreneurs.16 In this regard, the perception among micro and small business owners that participated in the focus groups was that the state provided no service for this fee. Some small entrepreneurs report having had better access to credit from financial institutions following incorporation. As a small businessperson put it, “I changed to a separate legal personality so that they give me more credit for purchasing machinery.� Finally, it was also noted that operating the business as a separate legal entity served as a way of protecting the personal property. 16 The registration fee depends on the annual sales of the business as follows: 35 nuevos soles for businesses selling up to 45,500 nuevos soles; 280 nuevos soles for those with sales from 45,500 to 262,500 nuevos soles; 560 nuevos soles if sales between 262,500 and 525,000 nuevos soles; 1,295 nuevos soles if sales between 525,000 and 2,975,000; and 2,100 soles for those with sales above 2,975,000 nuevos soles. 60 2.40 Yet, few MSEs set up their business as a separate legal entity, in part due to limited knowledge about the costs and processes involved. Among the surveyed micro and small businesses, only 139 businesses, or 17 percent of the sample, are incorporated as businesses with separate legal personality. This begs the question why, on the face of the perceived benefits of incorporation, so few micro and small businesses choose to do so. A tentative answer based on the focus groups would revolve around gaps in the information available to micro and small entrepreneurs. In particular, micro businesses consider that setting up the business as a separate legal entity would be costly and, in addition, imply high recurrent costs. It appeared that many micro businesses simply did not know what incorporation entailed, nor their costs or benefits. In this regard the role of the accountant again appears to play a critical role in suggesting to micro business owners that they incorporate. It is therefore possible that the constraint by which only natural persons (with few exceptions) can file under the RUS may be a factor that the owner, if not the accountant, would have in mind as a factor to not incorporate. 61 C. Econometric analysis of the determinants of informality This section presents the results of estimating a series of probit equations on the likelihood of a firm’s being formal, i.e., having a municipal license, a RUC number, and workers with health coverage). This allows us to determine which variables are robust determinants of informality, in the sense that those variables help predict whether a business is formal or not. In particular we are able to control for: (i) characteristics of the business such as size and age; (ii) characteristics of the business owner such as education, gender or marital status; (iii) the city and industry in which the business operates; and even: (iv) enforcement as captured by inspection rates by city and industry. 2.41 Among business characteristics the older the business the more likely it is that it will have a municipal license or a RUC number. The impact is, however, relatively small. A business that is 10 years old is around 7 to 8 percent more likely to have a municipal license (Table 2.10). Larger businesses are more likely to have a license or a RUC number, again controlling for a wide range of other variables that could potentially also help predict whether a business is formal or not. The decision of jointly holding a license and a RUC number is also analyzed. The results of this additional analysis, shown in Table 2.11, suggest that the conclusions are robust to different specifications.17 2.42 The more educated the business owner, the more likely the business is to obtain a license or a RUC number. In contrast, no other personal characteristic of the owner such as gender, age, or marital status appears to be a robust determinant of informality. Similarly, whether the parent of the entrepreneur was also in business or the reasons for being in business: to care for the family or for prospect of growth, is not a significant determinant of whether a business obtains a license or a RUC number. 2.43 The sector of activity matters, with businesses in shoe and leather, and textiles and clothing less likely to obtain a municipal license and a RUC number. The effects are large. Being in the textile and clothing sector decreases the probability of having a municipal license (compared to the base case of restaurants) by around 40 percent. This result is fully consistent with the argument that those businesses operating in sectors of activity that are less visible to the authorities are less likely to obtain a license. Businesses in these less visible sectors are also less likely to have a RUC number, though the impact is much smaller than in the case of the municipal license. Businesses in transport are much less likely to have a RUC number. A possible interpretation of this result is that, faced with the requirement to file taxes under the general regime, as they are not allowed to use the RUS or RER, (see Table 2.5), transport operators simply fall back into greater informality in this area. 2.44 Businesses in Arequipa and Trujillo are less likely to have a municipal license (relative to businesses in Lima). A possible reason behind this result is that the observed differences by city may result from differences in the costs of dealing with the 17 The identification of these coefficients is determined by businesses that have a license but no RUC. There are few larger businesses like that, which may affect the results from this joint estimation. 62 authorities or, perhaps differential enforcement of the rules (more on this below). As expected, the differences across cities are greater regarding the probability for obtaining a municipal license than a RUC number. Table 2.10: Variables that help predict having a municipal license and a RUC number Marginal effects from probit estimation Municipal Licence RUC (1) (2) (3) (4) (5) (6) (7) (8) Female owner -0.00274 -0.00509 0.00554 -0.0257 -0.0208 -0.0131 -0.00292 -0.0241 (0.041) (0.041) (0.041) (0.046) (0.034) (0.034) (0.031) (0.031) Owner's Age -0.00164 -0.00223 -0.00109 -0.000428 -0.00231* -0.00260* -0.000659 0.000378 (0.0017) (0.0018) (0.0018) (0.0019) (0.0014) (0.0014) (0.0013) (0.0012) Owner is Married 0.0147 0.0267 0.0224 -0.00433 0.0506* 0.0518* 0.0398 0.0226 (0.038) (0.038) (0.038) (0.041) (0.030) (0.030) (0.027) (0.025) Owner has Tecnico education 0.0416 0.0426 0.0331 0.0418 0.0994*** 0.0917** 0.0707** 0.0596** (0.061) (0.061) (0.062) (0.067) (0.036) (0.037) (0.033) (0.028) Owner has University education 0.184*** 0.200*** 0.169*** 0.161*** 0.191*** 0.176*** 0.126*** 0.0775*** (0.046) (0.046) (0.048) (0.054) (0.030) (0.031) (0.030) (0.030) Owner has complete secondary education 0.00688 0.0114 0.00760 0.0307 0.0722** 0.0611* 0.0471 0.0400 (0.047) (0.048) (0.048) (0.051) (0.035) (0.035) (0.032) (0.029) Parent owned a business -0.0104 -0.00690 -0.00344 0.000235 -0.0455 -0.0471 -0.0440 -0.0492* (0.037) (0.037) (0.037) (0.040) (0.031) (0.031) (0.028) (0.026) Age of firm 0.00799*** 0.00814*** 0.00695*** 0.00677*** 0.00640*** 0.00670*** 0.00472*** 0.00519*** (0.0021) (0.0021) (0.0021) (0.0023) (0.0017) (0.0017) (0.0015) (0.0015) Municipal inspection rate in city*industry -0.0850 -0.134 -0.134 -0.0674 (0.24) (0.24) (0.24) (0.26) Shoe/Leather -0.301*** -0.316*** -0.325*** -0.302*** -0.193** -0.195** -0.206*** -0.158** (0.10) (0.10) (0.11) (0.11) (0.078) (0.078) (0.080) (0.080) Textiles/Clothing -0.433*** -0.455*** -0.459*** -0.442*** -0.265*** -0.272*** -0.264*** -0.221** (0.11) (0.11) (0.11) (0.12) (0.078) (0.079) (0.081) (0.086) Wood products -0.296*** -0.297*** -0.301*** -0.301*** -0.101 -0.0987 -0.0923 -0.176** (0.089) (0.089) (0.091) (0.097) (0.068) (0.069) (0.065) (0.081) Metal products -0.184** -0.171* -0.153* -0.155 -0.0186 -0.0248 0.000548 -0.0415 (0.092) (0.092) (0.093) (0.098) (0.064) (0.065) (0.055) (0.062) Food stuffs -0.0355 -0.0259 -0.0177 -0.0218 0.0554 0.0572 0.0604 0.0507 (0.078) (0.077) (0.077) (0.082) (0.053) (0.053) (0.043) (0.037) Transport -0.180* -0.207* -0.201* -0.192* -0.409*** -0.411*** -0.422*** -0.484*** (0.10) (0.11) (0.11) (0.11) (0.080) (0.081) (0.086) (0.098) Arequipa -0.160** -0.187*** -0.154** -0.131* -0.153*** -0.146** -0.106* -0.0391 (0.063) (0.065) (0.065) (0.070) (0.059) (0.059) (0.055) (0.045) Trujillo -0.279*** -0.309*** -0.288*** -0.326*** 0.0821 0.0788 0.0879** 0.0488 (0.072) (0.072) (0.074) (0.078) (0.055) (0.056) (0.042) (0.044) Huancayo -0.0390 -0.0565 -0.00894 -0.0287 -0.106 -0.1000 -0.0382 0.00844 (0.074) (0.076) (0.073) (0.079) (0.078) (0.078) (0.064) (0.048) Cusco -0.0425 -0.0587 0.00340 -0.0114 -0.0369 -0.0321 0.0260 0.0558 (0.075) (0.077) (0.074) (0.082) (0.067) (0.067) (0.050) (0.034) Log distance to the Municipality Office -0.0765*** -0.0740** -0.0797*** -0.0616** (0.029) (0.029) (0.029) (0.031) Log distance to City Center -0.0181 -0.0266 -0.0168 -0.0349 0.00730 0.00923 0.0216 0.00610 (0.028) (0.029) (0.029) (0.032) (0.029) (0.029) (0.026) (0.023) In business to care for family -0.0717* -0.0601 -0.0438 -0.0550* -0.0346 -0.0272 (0.039) (0.039) (0.043) (0.030) (0.028) (0.026) In business for flexibility 0.167*** 0.169*** 0.172*** -0.0121 -0.00185 0.0000304 (0.048) (0.049) (0.053) (0.036) (0.033) (0.032) In business because can't find wage job 0.0280 0.0381 0.0311 0.00506 -0.00174 0.00698 (0.041) (0.042) (0.044) (0.034) (0.030) (0.028) In business for prospect of growth -0.0632 -0.0839 -0.0779 0.131* 0.0978 0.0277 (0.065) (0.062) (0.069) (0.070) (0.066) (0.055) SUNAT inspection rate in city*industry -0.183 -0.178 -0.131 -0.156 (0.14) (0.14) (0.13) (0.11) Log distance to the SUNAT Office -0.0559* -0.0603* -0.0593** -0.0237 (0.032) (0.032) (0.028) (0.025) 6 to 10 workers 0.0805** 0.0293 0.158*** 0.0774*** (0.040) (0.048) (0.021) (0.024) 11 to 50 workers 0.216*** 0.186*** 0.205*** 0.110*** (0.037) (0.050) (0.020) (0.026) Log November Sales 0.0317* 0.0836*** (0.017) (0.013) Observations 801 801 801 694 801 801 801 694 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education. Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 63 Table 2.11: Variables that help predict having both a municipal license and a RUC number Marginal effects from probit estimation (1) (2) (3) (4) Female owner 0.0223 0.0242 0.0405 0.0158 (0.044) (0.044) (0.045) (0.051) Owner's Age -0.00211 -0.00254 -0.000629 0.000616 (0.0019) (0.0019) (0.0020) (0.0022) Owner is Married 0.0318 0.0401 0.0341 0.0169 (0.041) (0.041) (0.042) (0.046) Owner has Tecnico education 0.113* 0.108* 0.0943 0.0868 (0.063) (0.063) (0.065) (0.071) Owner has University education 0.263*** 0.265*** 0.222*** 0.195*** (0.048) (0.049) (0.053) (0.060) Owner has complete secondary education 0.0483 0.0462 0.0367 0.0445 (0.051) (0.051) (0.053) (0.057) Parent owned a business -0.00931 -0.00734 -0.00209 0.000505 (0.040) (0.040) (0.041) (0.044) Age of firm 0.0100*** 0.0101*** 0.00842*** 0.00936*** (0.0023) (0.0023) (0.0023) (0.0026) Municipal inspection rate in city*industry -0.0215 -0.0831 -0.106 -0.193 (0.28) (0.28) (0.29) (0.31) SUNAT inspection rate in city*industry -0.264 -0.218 -0.179 -0.0878 (0.20) (0.21) (0.21) (0.22) Shoe/Leather -0.374*** -0.388*** -0.416*** -0.400*** (0.097) (0.096) (0.097) (0.11) Textiles/Clothing -0.475*** -0.493*** -0.516*** -0.533*** (0.098) (0.096) (0.097) (0.099) Wood products -0.288*** -0.290*** -0.304*** -0.346*** (0.089) (0.089) (0.092) (0.096) Metal products -0.176* -0.173* -0.151 -0.178* (0.094) (0.094) (0.097) (0.10) Food stuffs -0.00170 -0.00206 0.00567 -0.0185 (0.087) (0.087) (0.088) (0.096) Transport -0.422*** -0.440*** -0.457*** -0.486*** (0.089) (0.087) (0.090) (0.089) Arequipa -0.199*** -0.219*** -0.186*** -0.155** (0.067) (0.067) (0.070) (0.075) Trujillo -0.148 -0.178* -0.141 -0.242** (0.10) (0.10) (0.11) (0.11) Huancayo -0.0909 -0.108 -0.0514 -0.0627 (0.086) (0.087) (0.087) (0.093) Cusco -0.0567 -0.0800 -0.00219 -0.0119 (0.091) (0.093) (0.092) (0.099) Log distance to SUNAT office -0.0416 -0.0454 -0.0603 -0.0372 (0.041) (0.041) (0.041) (0.044) Log distance to Municipal office -0.0987*** -0.0946*** -0.105*** -0.0913*** (0.032) (0.032) (0.032) (0.035) Log distance to city center 0.0110 0.00622 0.0313 -0.00709 (0.040) (0.040) (0.041) (0.044) In business to care for family -0.0714* -0.0536 -0.0495 (0.042) (0.044) (0.048) In business for flexibility 0.115** 0.128** 0.157*** (0.050) (0.052) (0.057) In business because can't find wage job -0.000140 0.0123 -0.00333 (0.044) (0.046) (0.048) In business for prospect of growth 0.0131 -0.0241 -0.0460 (0.076) (0.076) (0.081) 6 to 10 workers 0.181*** 0.105** (0.040) (0.050) 11 to 50 workers 0.325*** 0.254*** (0.036) (0.054) Log November Sales 0.0682*** (0.019) Observations 801 801 801 694 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education. Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 64 2.45 Results are mixed regarding the role that inspection rates may play in determining whether a business gets a license or a RUC number. Taking the regressions results at face value would suggest that enforcement, as captured by the inspection rate by city/industry, does not play a role in the decision of businesses to obtain a license or a RUC number. However, it is possible that other variables for which we are controlling may be capturing some information regarding enforcement. As noted above, the city dummies themselves may well be capturing part of the enforcement information. More importantly, the distance of a business to the municipality and to the SUNAT are robust determinants of the likelihood that a business obtains a license and a RUC number (the further you are the less likely you are to obtain the license and a RUC). In part, this negative correlation between formality and distance to the offices can be explained by the costs involved in dealing with the authority. However, as we have already seen regarding the case of the SUNAT, those businesses that are closer to the SUNAT offices are in fact more likely to be inspected (see Annex B), possibly affecting the results on the inspection rate variable. 2.46 Business characteristics such as age and size also help explain the probability that a business provides workers with health insurance. The econometric exercise conducted above was repeated to examine the determinants of the probability of a business having (any) workers with medical health coverage (see Annex C). Significantly, larger businesses are more likely to have some workers with health insurance but also to have a lower share of their workers covered. This result is fully consistent with the view that informality in labor relations does not decrease as businesses grow. Other results regarding the determinants of having workers with health coverage include the positive correlation with the owner of the business having a university education, being married, and also businesses in retail sales. 2.47 Labor inspections play a role in explaining the probability that a business provides workers with health insurance. In contrast with the mixed conclusions about the role of enforcement in the case of tax or business licensing, there is stronger evidence that the labor inspections rate by city/industry positively affect the probability that a business would have workers with health insurance (see Annex C). 65 D. Dynamics out of informality In addition to the survey and focus groups, three case studies were commissioned to provide further insights on the question of what the key steps are in the trajectories that businesses follow along the informality-formality continuum. This section introduces these case studies and also provides a summary of the evidence gathered regarding this question of the dynamics out of informality. Box 7: Case studies – formality is also a ‘refuge’ These case studies tracked the evolution of three agri-businesses that were selected to capture businesses which, while successful, were at different stages of growth and expansion. Following a summary table with their key characteristics, this box briefly reviews their experience with the formalization process. Overall, three main insights may be drawn from the business cases studied. First, many entrepreneurs have several businesses at any given point in time and, since those business have different age profiles and operate in different sectors, their informality characteristics are also likely to be different. A business person that has a highly formal business and decides to start a new business is likely to do so as informal. This conclusion underscores the importance the cost-benefit calculations that MSEs make regarding the decision to formalize or not. It also suggests that simply providing information to micro and small entrepreneurs is unlikely to have much of an impact since even those that know how to formalize typically choose not to when setting up a new venture. Second, the trajectory toward formality is not necessarily one-way and can certainly be bumpy. In two of the cases studied, the businessperson decided at some point in time to revert back to work as a natural person, effectively closing down the legal business. Dealings with the tax authorities account for a fair share of the bumpiness, mostly as a result of limited knowledge by the entrepreneur on how to file taxes; trying to exploit the system of duty drawbacks for exporters; as well as labor disputes with former employees. Third, formality – and in particular business incorporation can be a refuge. Typically one thinks of informality as the refuge where the non-payment of taxes gives the entrepreneur a fighting chance. In contrast to this view, these businesspeople found that formality provided them with a refuge to preserve their personal property and a conduit to transmit wealth; a way to protect themselves from labor lawsuits; and an immediate boost in their credit standing. Business Since Annual Activity Key steps in Key drivers sales formalization Ecological 1992 USD 0.3 Snacks, It reverted back to The initial drive to register Amazon million jams, operating as a natural in 1993 was enforcement Food SAC yoghurts person from 2000 to by the local authorities. (PERSA) 2005 to take In 2005 the return to advantage of certain operating as a legal entity incentives. was due to tax issues. Malakasí 1987 USD 0.6 Mango In 2005 incorporated To obtain duty drawback producer 66 SAC million and the business; from exports; exporter Repeated attempts to Self-confidence in formalize as previous capacity to grow the businesses had to business close Frutas y 1993 USD 0.7 Avocado In 2006 incorporated The link with larger Frutas million producer the business, this has exporters ‘forced’ him to SAC and allowed him to obtain formalize; exporter credit previously To reduce the risk of labor unavailable due to his lawsuits against him advanced age. Ecological Amazon Food SAC (PERSA) Rocío Torres set up this small jam manufacturer in Iquitos in 1992 in a completely informal manner, trying to make a living after she had been fired by her employer when she got pregnant. Within one year Rocío Torres and her husband had incorporated a business (PERSA SRL), obtained their RUC number and, in 1994, obtained the municipal license and sanitary certification for their products. In 1997 she was able to register the brand PERSA with the relevant authorities (Instituto Nacional de Defensa del Consumidor y la Propiedad Intelectual, INDECOPI) as well as receive some financing through the Financial Corporation for Development (Corporación Financiera de Desarrollo, COFIDE). The business continued growing, diversifying into new products like snacks, and in 2000 they decided to close down PERSA SRL to continue working as a natural person. The intention was to benefit from the tax concessions granted through the law of regional development of the Amazonas. However, this caused problems with SUNAT. After some financial difficulties in 2005, which prevented them from having access to credit, the company was once again incorporated (as Ecological Amazon Food SAC) and has continued to expand in the Iquitos market. Agroindustrias Malakasí SAC Armando Echevarría started to trade agricultural products in 1987. Since then he has started five different business ventures. He has operated businesses both formally, with different business partners, and informally, especially in fishing. The formal activities, however, were the ones that brought problems with the authorities, in particular related to duty drawbacks. Despite these problems with the authorities, Armando Echevarría’s business activities generated enough capital for him to be able to invest in a mango packing plant in Piura, as well as in land to grow mangos in 2003. Throughout this process, access to credit has been blocked to him and he has had to rely on retained earnings for every business he has undertaken. Following his decision to enter into this business, at the same time keeping his fishing business, Armando Echevarría decided to incorporate Malakasí in 2005 in order to be eligible for claiming the duty drawback from his exports as the rules require a certain amount of capital before being eligible for any payment back. Frutas y Frutas SAC Armando Bermúdez has been in business since 1968, working on auto parts and diesel engines. By the 1990s, however, competitive pressures from used spare parts and informal repair workshops made him consider a change of activity. In this case the accountant not only prepared the books but also offered him a business deal: a plot of land in Casma that he could purchase in installments. Thus, in 1993 Armando Bermúdez became an avocado producer and 67 soon-to-be exporter. Despite being a seasoned entrepreneur, Armando Bermúdez started off his new business venture as a natural person. Funding for the investments into the new activity was from the earnings of his previous businesses. But by 2000 that source of funds had dried up and Armando Bermúdez unsuccessfully looked around for financing. It took until 2003 before he obtained some funding, at high rates and mortgaging the land, from the Caja Rural. Apparently, a cause of concern from lenders was his advanced age (he is now 73). In 2002 the land finally started producing as expected and he channeled his sales through some local wholesalers and in 2004 he started to sell to an import-export trader directly. After 13 years of having to run the business as a natural person, the business was incorporated in 2006. The main reasons for doing so was to put an end to some problems that were building up with SUNAT and the MTPE as he had worked as a natural person under the agrarian regime but with continuous contracts for his laborers. In addition, as a separate legal entity he is now protected from lawsuits by former hired-hands. Finally, Armando Bermúdez noted that he is now able to draw credit from financial institutions as a legal person since his age is no longer the key consideration but rather the soundness of his business. 2.48 Based on the three data gathering efforts undertaken for this report: the survey, the case studies, and the focus groups and interview, a synthesis of the steps toward formality is presented in Table 2.12. Table 2.12: Summary of steps towards formalization Step Timing Key factors 1. Obtain RUC number, start - Shortly after starting - Business needs to issue paying taxes operations, typically within a receipts (boletas) to expand year customer base 2. Obtain municipal license - Shortly after getting RUC - Avoid fines and having to (if business visible from number or simultaneously pay-off inspectors the street) 3. Maintain workers as - As the business expands, can - To avoid paying high social service providers take several years benefits and eventual firing costs - Need for flexibility 4. Incorporate business as a - As the business expands - To expand customer base to separate legal entity several years before it happens larger firms and the state (if at all) - To gain better access to credit 5. Include workers on payroll - Businesses only do so when - Enforcement of labor they reach a much bigger size, regulations not as MSEs Source: World Bank staff elaboration. 68 2.49 The survey data also provide further evidence to support this stylized summary of businesses’ trajectories toward formality. We can corroborate these findings with our survey data in two ways. First, we can exploit the questions in the survey that ask about the process of formalization itself; we know that around 50 percent of businesses surveyed state that they have had a RUC number and a municipal license since they started the business (see further details in Table E.1 in Annex E). Second, even though the survey provides only a snapshot of a cross-section of businesses, we can examine the formality characteristics of businesses of different ages (Figure 2.9). This evidence suggests that while formalization with regard to the RUC number and the municipal license not only starts from a high level but also continues an upward trend over time, it also confirms that informality along the labor relations dimension does not actually take place; and only a few micro and small entrepreneurs choose to incorporate their businesses as separate legal entities. However, caution must be exercised when interpreting this evidence since we might expect more productive firms to be more likely to be formal and therefore also to survive. This would create an upward bias for older businesses when looking at the cross-section of the formality-age relationship. Given this possible bias, it is all the more remarkable how few businesses have workers with a contract or are incorporated as separate legal entities. Figure 2.9: Formality characteristics among surveyed businesses, by age of business Proportion among surveyed businesses, percent 100 With RUC 80 number With 60 municipal license 40 Incorporated 20 With some workers under contract 0 1 3 5 7 9 11 13 15 17 19 Age of business in years Source: World Bank survey of micro and small businesses. 69 3. The impact of informality Establishing empirically the effect of a variable like informality is not a simple task. The potential for estimation biases arising from the self-selection of businesses into different categories (formal/informal) is large. To minimize these potential biases this chapter makes use of three different methods, discussed in detail in Annex D and follows the methodology used in World Bank (2008a): (i) the OLS estimates as a first approximation; (ii) a treatment effects regression that tries to take account of the potential endogeneity of formality; and (iii) an estimate using propensity score matching. A. Is informality a cause of low profitability? 3.1 Any analysis of the impact of informality on business performance must be interpreted with great caution as there are many possible links that are not easy to capture. The methodology used in this report does not attempt to capture the negative effects that informality may have on the performance of formal firms, or the effects that informality may have in undermining social trust and the overall business environment. In addition, we must also bear in mind that to the extent that informality is a second-best reaction, forcing firms to become formal may actually hurt growth, under the assumption that there is no increase in the benefits of formality. Therefore, as a general starting point for the discussion on the impact of informality on business performance we must acknowledge the difficulty of establishing causal links. What is usually accepted is that the factors that drive informality (high regulation, excessive payroll-financed employment protection legislation) negatively affect business performance and reduce economic growth. 3.2 A naïve comparison would appear to suggest that businesses with a municipal license are more profitable than informal ones. A naïve comparison of the profitability of businesses with and without a municipal license shows an apparent ‘premium’ for formality, as businesses in our sample that have a municipal license are around 55 percent more profitable than those without a license. However, these correlations are o not accurately indicated the true effect of a municipal license on the profitability of a business. The survey data allows us to dig much deeper into this issue and to go beyond these naïve comparisons. 3.3 But controlling for exogenous characteristics of the business and its owner indicates that there is no robust impact of having a municipal license on profitability. The survey provides us with a large number of characteristics of the owner and the business which we can take into account when estimating the profitability premium of formal businesses. The results of this analysis applied to the profitability ‘premium’ of having a municipal license, after controlling for different sets of variables are shown in Figure 3.1. The figure starts with the naïve comparison, not controlling for any observable characteristic, and then reports a number of OLS regressions controlling by an increasing number of variables. We first account for differences in owner characteristics (gender, marital status, level of education attained, and whether a parent of the owner had a business), age of the business, and city and industry dummy variables. 70 Controlling for this first set of variables brings down the profitability ‘premium’ of having a municipal license to around 41 percent (as shown in the second bar from the left in Figure 3.1). The third bar in Figure 3.1 shows the estimated profitability ‘premium’ controlling for an even larger number of variables, now including the reasons as to why the owner went into business as well as a measure of self-efficacy. This is to control for the possibility that the causality chain would run from more able entrepreneurs to more profitable businesses to greater likelihood of having a license. When we then control for the number of workers in the business, – the fourth bar in the figure, we no longer obtain a statistically significant profitability ‘premium.’ The last two bars on the right of Figure 3.1 represent estimates of the ‘premium’ using different techniques. In the ‘matching’ methodology the premium is estimated by comparing those businesses that most closely resemble each other on all observable characteristics, other than having a municipal license. However, this technique does not account for the possibility that businesses may differ in characteristics other than those that can be observed by the variables in our dataset. This is a serious shortcoming since factors such as entrepreneurial or sales skills are typically difficult to fully capture. The final bar on the right in Figure 3.1 shows the results from a technique that allows for the underlying observations to be heterogeneous on unobservable characteristics. Overall, the last three bars on the right of Figure 3.1 represent alternative ways of having controlled for as many variables as possible. The results indicate that having a municipal license does not have a robust impact on the profitability of a business. Figure 3.1: Difference in profitability between the business with a municipal license and those without Percent 80 60 40 20 Statistically not significant 0 Statistically -20 not significant -40 no additional owner + reasons for being + number of Matching on all Controlling for variable (OLS) characteristics, age in business and workers (OLS) observable unobservable of firm, industry owner ability (OLS) characteristics characteristics and city (OLS) (2SLS) Source: World Bank staff calculations. See Table 3.1 below and Table D.1 and Table D.2 in Annex D. 3.4 Similarly, there is no robust impact of having a RUC number on the profitability of a business. As in the case of the municipal license, a naïve comparison suggests that businesses in our sample that have a RUC number have a profitability that is around 103 percent higher than the profitability of those businesses without a RUC 71 number. However, as in the case of the municipal license, this correlation is not robust to a different estimation technique, especially when we allow for the possibility that the underlying observations in our sample differ also on characteristics that are not captured by the variables in our dataset (Figure 3.2). Figure 3.2: Difference in profitability between businesses with a RUC number and those without Percent 120 80 Statistically 40 not significant 0 -40 -80 -120 -160 no additional owner + reasons for + number of + capital stock Matching on all Controlling for variable (OLS) characteristics, being in workers (OLS) (OLS) observable unobservable age of firm, business and characteristics characteristics industry and owner ability (2SLS) city (OLS) (OLS) Source: World Bank staff calculations. See Table 3.1 below and Table D.1 and Table D.2 in Annex D. 3.5 Overall, the econometric analysis provides no evidence that as a consequence of having a municipal license or a RUC number businesses are more profitable (Table 3.1). Evidence from the propensity score matching would appear to suggest a positive impact on profits of having a license or a RUC number, however, the result is not robust in the treatment effects regression (2SLS), suggesting that a self-selection of businesses on the basis of unobservable variables may be the driving force behind the OLS and the matching results (for additional results see Table D.1 and Table D.2 in Annex D). 72 Table 3.1: Impact of different formality characteristics on firm profits OLS First-stage F-Stat 2SLS Matching Municipal license All businesses 0.351*** 3.70 -0.124 0.435*** (0.13) (1.95) (0.134) 1 to 5 workers 0.138 8.55 0.240 0.237* (0.15) (0.66) (0.137) 6 to 10 workers -0.222 0.36 n.a. 0.003 (0.29) (0.218) 11 to 50 workers 0.536 0.19 n.a. 0.350 (0.39) (0.444) RUC All businesses 0.845*** 2.14 -1.635 0.886*** (0.13) (3.30) (0.144) 1 to 5 workers 0.490*** 1.73 -0.523 0.550*** (0.15) (2.09) (0.135) 6 to 10 workers 0.396 3.33 2.068 0.374 (0.36) (2.30) (0.314) 11 to 50 workers 0.459 0.02 n.a. -0.131 (0.47) (0.706) Note: Dependent variable is log profits. Robust Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 These regressions weigh observations differentially to account for the differential reporting of profits of businesses depending on distance to the SUNAT office. N.a.: the intended instrument (distance to municipal office / SUNAT) was not feasible in some cases due to missing data (reporting of profits by businesses closer to the offices was poor). Source: World Bank staff estimates. 73 B. What is the impact of informality on access to credit? This section examines the current context of how Peruvian informal businesses access sources of finance. It briefly reviews conceptual and measurement issues, and presents evidence based on the firm-level survey built for this study on the extent to which credit access is an obstacle to informal firms and entrepreneurs. In the background, as a motivation for our concern with this issue, lies the cross-country evidence that returns to capital are high for micro enterprises (World Bank, 2007c) suggesting that, to the extent that informality may be found to curb access to credit for micro and small businesses, this would provide an incentive to formalize. 3.6 While financial depth, as captured by the ratio of bank credit to the private sector, remains low, the growth of the performing loan portfolio has been high in recent years. The Peruvian financial system remains shallow relative to the region (Figure 3.3), but has experienced a rapid expansion of bank credit in recent years. The policy objectives in the context of rapid credit growth should be to consolidate the gains from improved financial depth by implementing policies that will limit the vulnerability of the real and financial sectors while at the same time prioritizing financial widening, or access to finance for all. The GOP has taken several steps to mitigate risks and continue promoting sustainable credit expansion.18 Widening access to the system is also a key policy priority and encompasses a broad range of issues pertaining to the overall business environment. Figure 3.3: Financial depth Private sector credit in Peru and selected Total credit to the private sector in Peru neighboring countries In USD million Percent of GDP, 2006 60,000 100 50,000 80 40,000 60 US$ MM 30,000 40 20,000 20 10,000 0 - Chile Brazil Colombia LAC Mexico Peru 9 99 9 0 00 0 M 1 01 1 M 2 02 2 M 3 03 3 M 4 04 4 M 5 05 5 M 6 06 6 M 7 07 7 e9 p9 e0 p0 e0 p0 e0 p0 e0 p0 e0 p0 e0 p0 e0 p0 e0 p0 ay ay ay ay ay ay ay ay ay average En Se En Se En Se En Se En Se En Se En Se En Se En Se M M Month Year Source: WDI. Source: BCRP. 3.7 Access to credit is conceptually distinct from the actual use of credit. Financial depth, as measured by domestic credit to the private sector or by the ratio of quasi-money to GDP, is a commonly used macro indicator of credit access. The efficiency of the formal financial system in resource mobilization is usually summarized 18 See World Bank (2008b) for specific details on policies implemented. 74 by a variable such as the share of deposits in GDP and summary indicators of the cost of credit. These measures have the advantage that they are readily available, and easy to update and compare across countries. However, they tend to capture actual use as opposed to improvements in the opportunity to access finance by all segments of the population.19 Measuring access to credit involves a qualitative assessment of the costs of and opportunities to potential users of obtaining finance. A good measure of credit access should therefore reflect the ability of the financial system to be accessible by all segments of the population. It should therefore capture the extent to which the system facilitates information flows and risk diversification; lowers transactions costs; and eases intermediation. It should also reflect the quality of the institutional environment. For example, a financially-developed system is associated with a regulatory and legal framework that supports credit expansion and protects both debtor and creditor rights.20 It encourages the creation of infrastructure that can allow technology to bring down transaction costs. 3.8 Obstacles to financial widening at the user-end include the need to provide proof of income, guarantees and/or collateral. It is harder to keep track of income and assets in the informal sector, especially at the small and micro end, which leads to financial institutions typically requiring both loan guarantees and collateral from borrowers. Giving every individual a national identification number and creating credit registries where lenders share information about their clients' repayment records can be very valuable in this context. All borrowers would then have an asset: their future access to credit, that they implicitly offer as collateral when they obtain a loan. Reducing the costs of registering or repossessing collateral, as well as the elimination of usury laws is another feature of accessible financial systems. Authorities should seek to strengthen creditors’ rights and facilitate access to secured credit through the full implementation of the new movable guarantees registry from the National System for Public Registry (Sistema Nacional de Registros Públicos, SUNARP). Movable collateral is a widely-held asset and using it as guarantee to obtain credit should enable an important sector of the population to become part of the financial sector. A new Movable Collateral Law has been enacted and SUNARP has taken the necessary steps to set up the infrastructure for implementing the law. 3.9 The high cost of credit is an additional constraint on widening access to credit. The per dollar transaction costs associated with small and micro loans is higher than those of large loans. Micro firms comprise 53 percent of the survey sample and over 90 percent of the population of informal firms in Peru, suggesting that the average size of financial transactions is likely to be small. The result is a high cost of credit to small, informal borrowers. Many medium- and long-term policies that enhance the overall effectiveness of the system will lower the cost of financial services for both large and small borrowers, for example: judicial reforms to strengthen creditor and debtor rights; improvements in information exchange;21 or the provision of credit registries; liens; and property ownership security. 19 World Bank (2008a) provides a database of micro measures of credit penetration. 20 See Demirguc-Kunt and Levine (2008), World Bank (2008d), and de la Torre et al. (2007) for a more detailed treatment of issues pertaining to credit access and market-friendly intervention. 21 Djankov, McLiesh, and Shleifer (2007); Haselmann and Wachtel (2006). 75 3.10 Policy reforms aimed at improving credit access should focus on both the supply and demand-side. There is strong cross-country evidence that security of contracts and private property rights significantly influence financial development. 22 In addition to increasing incentives to engage with formal financial institutions, these reforms reduce firms’ incentives to “circle around� the domestic system.23 Reducing transactions costs will also require the adoption of supply-side innovations that improve flexibility in use, such as mobile banks and electronic technology, along with demand- side measures such as strengthening consumer protection laws and services. There is also evidence that competition-enhancing policies are effective in stimulating the development of financial products targeted at poorer households, with the appropriate prudential precautions.24 As private sector banks find traditional businesses facing competition, they will seek out non-traditional business. Given the right environment, the private sector has the ability, the incentives, and the resources to develop innovative services that can strengthen access to finance for informal firms and entrepreneurs. 3.11 The data supports the hypothesis that firm size and geographic location matters for access to credit (Annex D). The access characteristics examined were firm participation in the formal financial system; terms of credit; availability of finance for start-up capital; and whether or not the proprietor of the firm holds the title to their own home. Data stratification was by sector, location (in Lima or outside Lima), firm size, and whether firms had acquired a RUC number from the very outset. The last factor turned out to be of significance only in the context of holding an account at a financial institution and borrowing for start-up capital, and was therefore dropped in subsequent tabulations. Significant differences emerged however across firms by size and location. 3.12 Less than half the firms surveyed had an account with a financial institution. Only 43 percent of the respondent firms had an account with any kind of financial institution, and only 15 percent of total firms had a bank account (Table D.4 in Annex D). Of firms with less than 6 workers, 70 percent had no account at all, while the corresponding figure for larger firms (with 11 to 50 workers) was about one-third suggesting significant differences based on firm-size. Firms outside Lima were more likely to have an account. The highest proportion of firms without accounts was in foodstuffs and shoe leather. Having had a RUC number from the outset had some impact in this context, especially for larger firms, but the differences are not significant. Moreover, most firms operate on a cash-only basis. 70 percent of the respondent firms accepted neither checks nor credit cards from their customers. While the proportion of cash-only transactions was highest for food stuffs and shoe leather, this trend holds across all sectors. 3.13 In order to start the business, 48 percent of businesses used only personal savings, which accounted for 64 percent of all capital needed at the outset. More than 90 percent of surveyed businesses reported having used personal savings as a source of finance while 12 percent of businesses reported having drawn from family and friends (as a business can use several sources of finance these percentages add up to more than 100 22 Acemoglu and Johnson (2005). 23 De la Torre and Schmukler (2004) and Qian and Strahan (2007) point to many such mechanisms, such as dollar contracts, short maturities with rollover clauses, and contracting under foreign jurisdiction. 24 Caprio and Honohan (2004). 76 percent). External finance accounts for less than a fourth of total borrowing for starting a business. Transport and textile clothing show the highest extent of borrowing from banks at close to 14 percent of start-up capital, however, inter-sectoral differences are minor (Table D.7 in Annex D). Small and mid-sized firms outside Lima were more likely to use formal financing sources: Banks, Cajas Municipales, Edpymes, and Microfinance institutions, with the opposite holding true for larger firms. Personal savings and support from family and friends provide the dominant share of start-up financing in Lima and outside Lima, and across all firm sizes and sectors. Formal finance was more likely to be used by firms that had always had a RUC number, this may be used as an indicator of “tending towards formality.� However the impact is small, suggesting that the general credit environment, including the full range of laws relating to creditor and debtor protection; comprehensiveness of credit registries; and collateral-related regulation, may be more important than informality per se in improving access to the formal financial system. 3.14 The picture is quite different when one examines whether firms have currently taken out a loan for any purpose at all. Overall, nearly half the firms in the sample took out a loan in either 2006 or 2007, and of these more than 93 percent relied on formal sources (Table D.9 in Annex D). These figures are relatively high. Using data from 71 investment climate surveys, an average of just over 20 percent of firms with 1 to 20 employees use external finance for new investment (World Bank 2008d:45). The loans were taken primarily for equipment and machinery and also to finance working capital, as out of the 392 businesses that took out a loan 235 reported it being at least partly used for financing machinery and equipment. 3.15 Slightly more than half of all loans are granted with the interest rate on a monthly basis. Around 53 percent of those businesses declaring having taken a loan are quoted monthly interest rates, although in the case of banks the share of loans with annual interest rates is around 55 percent (banks are the finance provider for which loans with annual interest rates are more common than those with monthly interest rates (Table D.10 in Annex D). Among loans from banks the most common interest rate is 30 percent per year (for loans with annual interest rates) and 4 per cent per month (for loans with monthly interest rates). Among loans from moneylenders the most common interest rate is 10 percent per month. 3.16 The maturity of loans is around 14 months. The maturity of loans does not vary much by location or firm size, hovering around 14 to 15 months in all cases (Table D.11 in Annex D). The results of the survey have to be interpreted cautiously, as limited financial literacy may affect the quality of survey responses. In this regard, it should be noted that 56 percent of total respondents outside Lima claimed to have taken a loan recently, compared to only 41 percent in Lima. This may have been due to some confusion between “commercial� borrowing and interest-free loans from friends and family. 3.17 Nearly two-thirds of the respondent firms had to provide loan guarantees. The proportion of firms providing guarantees was lower in Lima than in other regions. Larger firms were less likely to be asked for a guarantee than small or mid-sized firms. Overall, the proportion of firms that had to guarantee loans was 62.2 percent. This data 77 points to the potentially significant impact of implementing the new Movable Collateral Law and recent reforms aimed at improving coverage of credit registries. 3.18 There is no econometric evidence that having a municipal license or a RUC enhances the likelihood of obtaining a loan. The same econometric techniques applied to the impact on profits were also used to estimate the impact that having a municipal license or a RUC number has in terms of the probability of getting a loan. The evidence indicates that having a license or RUC number does not increase the probability of getting a loan (Table D.14 in Annex D) 78 List of references Acemoglu, D. and S. Johnson (2005), ‘Unbundling Institutions,’ Journal of Political Economy (113)5 (doi: 10.1086/432166). Caprio, G. and P. Honohan (2004), ‘Can the unsophisticated market provide discipline?’ World Bank Policy Research Working Paper No. 3364. Chong, A., Galdo, and J. Saavedra (2007), ‘Informality and Productivity in the Labor Market: Peru 1986–2001,’ IDB Research Department Working Paper No. 609 De La Roca, J. and M. Hernández (2004), ‘Evasión Tributaria e Informalidad en el Perú: Una aproximación a partir del Enfoque de Discrepancias en el Consumo,’ Lima: Consorcio de Investigación Económica y Social. De la Torre, A., J. C. Gozzi, and S. L. Schmukler (2007), ‘Financial Development: Maturing and Emerging Policy Issues,’ The World Bank Research Observer Advance Access published on May 4, 2007. De Mel, S., D. McKenzie, and C. Woodruff (2007), ‘Returns to Capital in Microenterprises: Evidence from a Field Experiment,’ World Bank Policy Research Working Paper No. 4230, May. Demirguc-Kunt, A. and R. Levine (2008), ‘Finance and Economic Opportunity,’ World Bank Policy Research Working Paper No. 4468. Djankov, S., C. McLiesh, and A. Shleifer (2007), ‘Private credit in 129 countries,’ Journal of Financial Economics (84)2:299-329(doi:10.1016/j.jfineco.2006.03.004). Gasparini L. and L. Tornarolli (2006), ‘Labor Informality in Latin America and the Caribbean: Patterns and Trends from Household Survey Microdata,’ background paper for World Bank (2007), Informality. Exit and Exclusion, Flagship report of the Latin American and Caribbean Vice-Presidency. González, D. (2006), ‘Regímenes especiales de tributación para pequeños contribuyentes en América Latina,’ Banco Interamericano de Desarrollo, Departamento de Integración y Programas Regionales. Haselmann, R.F. and P. Wachtel (2006), ‘Bank risk and bank management in transition,’ Policy Conference, Tokyo, European Bank for Reconstruction and Development. IFC (2008), Municipal Scorecard 2008. Midiendo las Barreras Burocráticas a Nivel Municipal. Reporte Perú. Washington, D.C.: World Bank - IFC. ILO (1991), Dilemma of the informal sector. Report of the Director-General presented at the 78th International Labour Conference, International Labour Office, Geneva. ILO (2006), Strategies and practice for labour inspection, Report of the Committee on Employment and Social Policy to the Governing Body of the International Labor Organization (GB.297/ESP/3), November. Jaramillo, M. (2004), ‘La Regulación del Mercado Laboral en el Perú,’ Lima: GRADE. Kristensen, N. and W. Cunningham (2006), ‘Do Minimum Wages in Latin America and the Caribbean Matter? Evidence from 19 Countries,’ World Bank Policy Research Working Paper No. 3870, March. 79 Loayza, N. and J. Rigolini (2006), ‘Informality Trends and Cycles,’ World Bank Policy Research Working Paper No. 4078, December. Maloney, W. and J. Núñez (2001), ‘Measuring the Impact of Minimum Wages: Evidence from Latin America,’ World Bank Policy Research Working Paper No.2597, April. Qian, J. and P.E. Strahan (2007), ‘How laws and institutions shape financial contracts: the case of bank loans,’ Journal of Finance (62)6:2803-2834 (doi:10.1111/j.1540- 6261.2007.01293.x) World Bank (2007a), Peru: Casting Light on the Shadow Economy (Report No. 40220-PE). World Bank (2007b), Informality. Exit and Exclusion, Flagship report of the Latin American and Caribbean Vice-Presidency. World Bank (2007c), Doing Business 2008. Washington, D.C.: World Bank. World Bank (2008a), Republic of Bolivia: Policies for Increasing Firms’ Formality and Productivity (Report No. 40057-BO). Washington, D.C.: World Bank. World Bank (2008b), Second Fiscal Management and Competitiveness Policy Development Loan – Project Appraisal Document. Washington, D.C.: World Bank. World Bank (2008c), Database associated with is the database associated with A. Demirguc- Kunt, A. and R. Levine, ‘Finance and Economic Opportunity,’ World Bank Policy Research Working Paper No. 4468. World Bank (2008d), Finance for All? Policies and Pitfalls in Expanding Access, A World Bank Policy Research Report. Yamada, G. and J. Chacaltana (2007), ‘Generación de Empleo en el Perú: Seis Casos Recientes de Éxito,’ Documento de Discusión, DD/07/03. Universidad del Pacifico, Lima, Perú. 80 Annex A. Measures of informality based on the ENAHO using different definitions Figure A.1: Calculating the productive definition of informality based on the ENAHO Information by respondent from ENAHO Classification Calssification ocu500 p507 niv_ed / pub_o_priv tamano Included in informality as per productive definition? 1. Primaria incompleta 4. Self employed non Yes 1. Empleador o 2. Primaria completa professional patrono 3. Secundaria incompleta 4. Secundaria completa 5. Superior incompleta 6. Educaci ón superior 1. Self employed 1. Ocupado completa professional 2. Trabajador 1. Primaria incompleta 4. Self employed non Yes independiente 2. Primaria completa professional 3. Secundaria incompleta 4. Secundaria completa 5. Superior incompleta Trabajador 6. Educaci ón superior 1. Self employed completa professional 3. Empleado Público 3. Salaried public - formal Privado 0. Hasta 10 5. Salaried private - Yes informal 1. M ás de 10 2. Salaried private - formal 4. Obrero Público 3. Salaried public - formal Privado 0. Hasta 10 5. Salaried private - Yes informal 1. Más de 10 2. Salaried private - formal 5. Trabajador 6. Family worker and others Yes familiar no remunerado 6. Trabajador 7. Domestic workers Yes 2. Desocupado abierto del hogar 3. Desocupado oculto 4. No pea 7. Otro 6. Family worker and others Yes 81 Figure A.2: Map of prevalence of informality by region Percent of workers without a contract (2006) 76.8 78.2 84.5 86.2 76.4 89.5 86.3 74.7 81.0 89.7 78.5 77.2 58.3 76.4 55.4 75.4 Legend 90.6 84.8 Five less informal departments 88.2 Second five less informal departments 63.7 89.0 85.4 Third five less informal departments Second five more informal departments 68.9 Five more informal departments 64.2 62.6 Note: Informality rates are shown at the departmental level since this is the most disaggregated geographical unit for which the annual ENAHO survey provides a representative sample of the population. Source: World Bank staff calculations based on ENAHO. 82 Figure A.3: Map of prevalence of informality by region Percent of workers working in businesses without legal personality (2006) 92.6 90.4 88.8 97.7 87.2 97.5 90.3 82.7 89.5 97.0 89.0 89.3 65.7 91.6 62.0 89.7 Legend 96.2 94.0 Five less informal departments 96.3 Second five less informal departments 70.6 98.0 95.9 Third five less informal departments Second five more informal departments 79.2 Five more informal departments 78.0 81.9 Note: Informality rates are shown at the departmental level since this is the most disaggregated geographical unit for which the annual ENAHO survey provides a representative sample of the population. Source: World Bank staff calculations based on ENAHO. 83 Figure A.4: Map of prevalence of informality by region Percent of workers working in businesses that do not keep accounting books (2006) 88.5 88.2 86.5 90.9 75.7 95.2 90.2 78.8 82.1 94.3 77.0 84.9 60.9 84.9 58.5 87.8 Legend 94.9 88.4 Five less informal departments 95.1 Second five less informal departments 66.9 93.7 92.3 Third five less informal departments Second five more informal departments 74.9 Five more informal departments 78.0 75.1 Note: Informality rates are shown at the departmental level since this is the most disaggregated geographical unit for which the annual ENAHO survey provides a representative sample of the population. Source: World Bank staff calculations based on ENAHO. 84 Figure A.5: Prevalence of informality by region and income per capita Informality as measured by the ratio of workers without a contract (2006) Huancavelica .9 Huánuco Cajamarca Ayacucho Apurímac Amazonas Puno San Martín Cusco Piura Ancash .8 Ucayali Informality prevalence Loreto Pasco Tumbes Junín Lambayeque Madre de Dios La Libertad .7 Arequipa Moquegua Ica Tacna .6 Lima Callao .5 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 1000010500 110001150012000 Average annual income, Nuevos soles Source: World Bank staff calculations based on ENAHO and INEI. 85 Figure A.6: Prevalence of informality by region and income per capita Informality as measured by the ratio of workers working in a firm without legal personality (2006) 1 Ayacucho Amazonas Huánuco Cajamarca Puno Apurímac Huancavelica Cusco Tumbes Junín San Martín Pasco Loreto .9 Ancash Madre de Dios Piura Ucayali Informality prevalence Lambayeque La Libertad Tacna .8 Arequipa Moquegua Ica .7 Lima Callao .6 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 1000010500 110001150012000 Average annual income, Nuevos soles Source: World Bank staff calculations based on ENAHO and INEI. 86 Figure A.7: Prevalence of informality by region and income per capita Informality as measured by the ratio of workers working in a firm without accounting books (2006) 1 Huancavelica Cajamarca Apurímac Huánuco Ayacucho Puno Amazonas San Martín .9 Cusco Tumbes Madre de Dios Informality prevalence Loreto Piura Pasco Junín Ancash .8 La Libertad Moquegua Ucayali Lambayeque Arequipa Tacna .7 Ica Lima .6 Callao 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 1000010500 110001150012000 Average annual income, Nuevos soles Source: World Bank staff calculations based on ENAHO and INEI. Figure A.8: Informality rate and economic growth, 2003 to 2006 Informality as measured by the percent of workers without a contract Informality rate for sectors selected for the Informality rate (all economic activities) survey and for all economic activities and economic growth 95 95 All economy Selected sectors 90 90 Informality rate Informality rate 85 85 80 80 2004 2005 75 75 2006 70 70 2003 2004 2005 2006 3.5% 4.5% 5.5% 6.5% 7.5% 8.5% Real GDP growth Source: World Bank staff calculations based on ENAHO and INEI. 87 Figure A.9: Informality rate and economic growth, 2003 to 2006 Informality as measured by the percent of workers working in businesses without legal personality Informality rate for sectors selected for the Informality rate (all economic activities) survey and for all economic activities and economic growth 95 95 90 90 Informality rate Informality rate 85 85 2004 2003 2005 2006 80 80 75 All economy 75 Selected sectors 70 70 2003 2004 2005 2006 3.5% 4.5% 5.5% 6.5% 7.5% 8.5% Real GDP growth Source: World Bank staff calculations based on ENAHO and INEI. Figure A.10: Informality rate and economic growth, 2003 to 2006 Informality as measured by the percent of workers working in businesses without accounting books Informality rate for sectors selected for the Informality rate (all economic activities) survey and for all economic activities and economic growth 95 95 All economy 90 Selected sectors 90 Informality rate Informality rate 85 85 80 80 2003 2004 2005 2006 75 75 70 70 2003 2004 2005 2006 3.5% 4.5% 5.5% 6.5% 7.5% 8.5% Real GDP growth Source: World Bank staff calculations based on ENAHO and INEI. 88 Table A.1: Special fiscal regimens for small taxpayers in Latin America Component Countries Have special tax regimes for small contributors. Argentina, Bolivia, Brazil, Colombia, Costa Rica, Chile, Ecuador, México, Nicaragua, Honduras, Paraguay, Peru, Dominican Republic and Uruguay. Do not have special tax regimes for small El Salvador, Panama y Venezuela. contributors Applies more than one special regime. Argentina, Bolivia, Brazil, Chile, México, Peru y Uruguay. Only includes individual persons. Argentina, Bolivia, Colombia, Chile, Ecuador, México Nicaragua, Honduras, Paraguay, Peru (RUS), Dominican Republic and Uruguay. Includes juridical persons. Brazil, Costa Rica, Chile, Mexico, Peru (RER) and Uruguay. Considers only gross income to determine Brazil, Chile, México, Paraguay, and Dominican Republic. included contributors. Considers assets, physical parameters or other Argentina, Bolivia, Colombia, Costa Rica, Chile, Ecuador, indicator to determine included contributors. Mexico, Nicaragua, Honduras, Paraguay, Peru and Uruguay. Applies fixed tariff technique. Argentina, Bolivia, Chile, México, Nicaragua, Peru (RUS)and Uruguay. Applies a percentage over gross income. Brazil, Chile and Peru (RER). Applies a regime that substitutes taxes only. Bolivia, Brazil, Colombia, Costa Rica, Chile, Ecuador, Mexico, Nicaragua, Honduras, Paraguay, Peru, Dominican Republic and Uruguay. Applies a regime that substitutes taxes and social Argentina, Brazil and Uruguay security. Applies a regime that substitutes only one tax. Brazil, Colombia, Costa Rica, Chile, Ecuador, Honduras, Paraguay, Peru (RER) and Dominican Republic Applies a regime that substitutes VAT and the Argentina, Bolivia, Mexico, Nicaragua, Paraguay, Peru (RUS) income tax and Uruguay. Source: Gonzalez (2006) Table A.2: Regulatory labor burden in comparative perspective Difficulty of Rigidity of Difficulty of Rigidity of Nonwage labor cost Firing costs (weeks Hiring Index Hours Index Firing Index Employment Index (% of salary) of wages) Peru 44 60 60 55 10 52 Argentina 44 60 20 41 26 139 Bolivia 78 60 100 79 14 n.a. Brazil 78 60 0 46 37 37 Chile 33 20 20 24 3 52 Colombia 22 40 20 27 29 59 Ecuador 44 60 50 51 12 135 Mexico 33 40 70 48 21 52 Paraguay 56 60 60 59 17 113 Uruguay 33 60 0 31 6 31 Venezuela 78 60 100 79 16 n.a. Latin America & Caribbean 36.9 35.5 24.5 32.3 12.7 56.1 Source: Doing Business 2008. 89 Descriptive statistics from the dedicated survey Table A.3: Main informality characteristics among surveyed businesses, by city Percent of surveyed businesses Arequipa Cusco Huancayo Lima Trujillo Firm has municipal license 60 72 80 71 43 Firm has permanent license 52 53 46 58 30 Firm paid taxes in November 2007 50 75 67 71 73 Percent of sales reported for tax purposes 62 58 44 52 51 Firm has RUC number 68 79 82 81 77 Firm files under RUS 21 40 37 24 32 Firm files under Special Regime (RER) 14 12 11 7 17 Firm files under General Regime 32 26 32 49 26 Firm never gives facturas or boletas 19 22 17 19 19 Firm has no workers with a labor contract 86 78 82 81 94 Firm has no workers with health insurance 71 78 72 66 79 Firm has no workers with pension coverage 86 87 81 79 88 Firm has limited liability (SA, SRL, SAC) 15 12 19 23 6 Sample size 111 108 93 368 122 Source: World Bank survey of micro and small businesses. Table A.4: Main informality characteristics among surveyed businesses, by sector Percent of surveyed businesses Shoes Retail Textiles Wood Metal Trans- Restau- and food- & apparel products products port rants leather stuffs Firm has municipal license 56 51 60 69 81 68 83 Firm has permanent license 43 37 44 51 63 47 70 Firm paid taxes in November 2007 64 54 68 78 85 55 75 Percent of sales reported for tax purposes 54 50 56 46 53 59 53 Firm has RUC number 74 71 80 87 90 54 88 Firm files under RUS 38 23 19 25 49 5 39 Firm files under Special Regime (RER) 16 10 14 10 6 11 7 Firm files under General Regime 21 38 46 52 36 35 37 Firm never gives facturas or boletas 21 21 12 7 13 52 15 Firm has no workers with a labor contract 90 89 80 82 88 80 74 Firm has no workers with health insurance 85 82 69 69 67 69 57 Firm has no workers with pension coverage 90 87 80 82 84 84 75 Firm has limited liability (SA, SRL, SAC) 9 15 16 25 11 29 17 Sample size 117 119 126 118 105 95 122 Source: World Bank survey of micro and small businesses. 90 Table A.5: Most important problem affecting micro and small business owners Percent of respondents Most important problem affecting the business 1 to 5 workers 6 to 10 workers 11 to 50 workers Total Economic situation 18 15 18 17 Taxes 11 21 16 14 Lack of capital, financing 12 15 14 13 Inflation 16 9 7 12 Competition 10 15 10 11 Lack of support from the authorities 5 4 7 5 Crime 5 4 5 5 Bureaucracy 4 1 5 4 Other social problems 3 1 5 3 Imports 3 1 3 3 Difficulties getting inputs 3 4 1 3 Unemployment 2 4 0 2 Corruption 2 1 3 2 Low purchasing power 2 1 2 2 Difficulties getting right employees 1 2 1 1 Informality 1 1 2 1 Lack of investment 1 0 1 1 Lack of information and advice 1 0 0 1 High cost of rents 1 1 0 1 Lack of infrastructure (transport, etc) 0 1 1 1 Not being located in a commercial area 0 0 1 0 Sample size 378 171 153 702 Second most important problem affecting the business 1 to 5 workers 6 to 10 workers 11 to 50 workers Total Inflation 17 17 10 16 Taxes 14 9 16 14 Unemployment 1 1 2 1 Crime 6 9 3 6 Lack of capital, financing 11 10 8 10 Economic situation 11 11 8 10 Competition 12 10 12 11 Bureaucracy 1 4 4 2 Informality 1 1 0 1 Lack of support from the authorities 5 4 6 5 Corruption 3 6 3 4 Imports 2 2 2 2 Difficulties getting right employees 2 4 11 5 Difficulties getting inputs 3 1 2 2 High cost of rents 0 0 0 0 Lack of infrastructure (transport, etc) 2 3 0 2 Low purchasing power 4 2 5 4 Not being located in a commercial area 1 0 0 1 Other social problems 3 2 5 3 Lack of information and advice 1 1 2 1 Sample size 306 139 128 573 Source: World Bank survey of micro and small businesses. 91 Table A.6: Obstacles for micro and small businesses - detailed questionnaire Percent of respondents. Issues ordered top-to-bottom, left-to-right according to share of respondents that consider an issue to be a serious obstacle. 1 to 5 6 to 10 11 to 50 1 to 5 6 to 10 11 to 50 Total Total workers workers workers workers workers workers Crime The limited size of the market No obstacle 4 4 3 4 No obstacle 10 21 23 16 Minor obstacle 9 10 10 9 Minor obstacle 17 17 19 17 Moderate obstacle 19 19 17 18 Moderate obstacle 32 37 36 34 Serious obstacle 68 68 69 68 Serious obstacle 41 25 22 33 Corruption Lack of qualified labor No obstacle 7 7 7 7 No obstacle 34 24 29 30 Minor obstacle 9 8 10 9 Minor obstacle 20 20 21 20 Moderate obstacle 17 17 15 17 Moderate obstacle 27 31 29 29 Serious obstacle 67 68 68 68 Serious obstacle 19 24 22 21 Economic instability Access to inputs No obstacle 4 5 5 4 No obstacle 31 33 38 33 Minor obstacle 9 13 13 11 Minor obstacle 19 22 19 20 Moderate obstacle 26 26 21 25 Moderate obstacle 29 27 22 27 Serious obstacle 61 58 62 60 Serious obstacle 21 19 20 20 Political instability Customs regulations No obstacle 7 7 8 7 No obstacle 45 40 40 43 Minor obstacle 15 15 15 15 Minor obstacle 16 15 17 16 Moderate obstacle 26 26 26 26 Moderate obstacle 22 20 24 22 Serious obstacle 52 52 51 52 Serious obstacle 18 25 19 20 Taxes Quality of transport infrastructure No obstacle 14 11 11 13 No obstacle 35 27 37 34 Minor obstacle 15 12 12 13 Minor obstacle 24 24 15 22 Moderate obstacle 24 25 23 24 Moderate obstacle 20 32 26 24 Serious obstacle 47 53 55 50 Serious obstacle 21 16 22 20 Cost of credit Enforcing contracts No obstacle 13 15 19 15 No obstacle 33 36 36 34 Minor obstacle 14 9 20 14 Minor obstacle 25 25 22 24 Moderate obstacle 27 24 25 26 Moderate obstacle 22 25 20 22 Serious obstacle 46 52 36 45 Serious obstacle 20 14 22 19 Unfair competition Labor regulations No obstacle 11 6 9 9 No obstacle 34 27 31 32 Minor obstacle 17 16 15 16 Minor obstacle 24 21 24 23 Moderate obstacle 31 27 29 29 Moderate obstacle 29 37 24 30 Serious obstacle 41 51 47 45 Serious obstacle 13 14 21 15 Requirements to obtain a loan No obstacle 16 23 31 21 Minor obstacle 16 12 19 16 Moderate obstacle 23 26 23 24 Serious obstacle 45 39 27 39 Source: World Bank survey of micro and small businesses. 92 Table A.7: Sales of micro and small businesses surveyed, by customer type and business size Percent of total sales 1 to 5 workers 6 to 10 workers 11 to 50 workers Total General public 90 78 73 83 Small businesses (less than 20 workers) 3 7 9 5 To intermediaries in Peru 3 6 5 4 Medium and large businesses (> 20 workers) 1 2 4 2 Exports directly 0 1 3 1 To intermediaries that export 0 2 2 1 State 0 1 2 1 Multinationals 0 1 0 0 Others 2 3 3 3 Source: World Bank survey of micro and small businesses. Table A.8: Elements of formality Percent of businesses displaying different elements of formality Business has a key Business has a key Business is linked to a Business belongs supplier providing client that buys more bigger firm through All businesses to an association more than 25 percent than 25 percent of its sub-contracting (gremio )? of inputs? output? arrangement? 1 to 5 6 to 10 11 to 50 Total No Yes No Yes No Yes No Yes workers workers workers Nothing 15 24 7 2 16 7 20 9 15 13 15 8 Municipal license only 6 11 2 1 6 7 9 2 7 2 6 11 RUC only 16 15 23 11 17 13 13 20 14 24 16 22 Municipal license and RUC 39 40 47 31 41 28 38 41 43 29 40 32 RUC, municipal license, and limited 10 4 9 26 8 22 8 12 9 14 10 12 liability RUC, municipal license, and all 6 4 6 11 6 6 6 6 6 6 6 5 workers with contract RUC, municipal license, all workers 5 1 5 15 4 11 5 5 4 8 5 5 with contract, and limited liability Other combinations of formality 3 2 4 4 2 5 2 4 3 4 3 6 Sample size 802 688 114 438 364 615 187 737 65 o/w businesses with 1 to 5 workers 421 421 - - 381 40 265 156 348 73 394 27 …businesses with 6 to 10 workers 201 - 201 - 172 29 94 107 145 56 177 24 …businesses with 11 to 50 workers 180 - - 180 135 45 79 101 122 58 166 14 Source: World Bank survey of micro and small businesses. 93 Table A.9: Factors affecting the choice of tax regime Percent of businesses Reasons for choosing the General Regime Reasons for choosing the RUS or RER regimes 1 to 5 6 to 10 11 to 50 1 to 5 6 to 10 11 to 50 Total Total workers workers workers workers workers workers To issue receipts (facturas ) You pay less taxes Not important 1 4 2 2 Not important 4 8 5 6 Somewhat important 10 6 5 7 Somewhat important 14 13 11 14 Fairly important 27 27 26 27 Fairly important 37 30 32 34 Very important 62 62 67 64 Very important 44 48 52 47 Sample size 101 77 126 304 Sample size 166 99 44 309 To avoid fines I have always been in a special regime Not important 5 11 4 6 Not important 14 6 17 12 Somewhat important 10 7 8 8 Somewhat important 23 9 7 16 Fairly important 23 26 29 26 Fairly important 33 39 41 36 Very important 62 57 59 60 Very important 30 46 34 36 Sample size 101 76 125 302 Sample size 153 93 41 287 I have always been in the general regime My sales are low Not important 5 9 10 8 Not important 8 12 11 10 Somewhat important 19 13 14 15 Somewhat important 26 27 23 26 Fairly important 31 24 25 27 Fairly important 29 29 30 29 Very important 44 53 51 50 Very important 37 32 36 35 Sample size 99 75 125 299 Sample size 167 97 44 308 All businesses in my area are in this regime All businesses in my area are in the special regimes Not important 10 14 11 11 Not important 16 7 5 12 Somewhat important 22 28 22 23 Somewhat important 24 21 21 23 Fairly important 26 23 25 25 Fairly important 29 42 44 35 Very important 43 35 43 41 Very important 31 30 31 31 Sample size 94 69 114 277 Sample size 147 81 39 267 My sales are too high for a special regime I do not need to issue receipts (facturas) to my clients Not important 24 18 14 19 Not important 15 12 18 15 Somewhat important 33 38 20 29 Somewhat important 28 24 16 25 Fairly important 20 20 26 23 Fairly important 28 29 39 30 Very important 22 24 39 30 Very important 29 35 27 31 Sample size 90 71 125 286 Sample size 166 100 44 310 Source: World Bank survey of micro and small businesses. 94 Table A.10: Participation in state-supported programs Percent of businesses unless otherwise indicated 1 to 5 6 to 10 11 to 50 1 to 5 6 to 10 11 to 50 Total Total workers workers workers workers workers workers Do you know about? Do you take part or ever took part? Credit lines by COFIDE Yes 28 31 44 33 1 1 2 1 No 72 69 56 67 99 99 98 99 Preferences to micro businesses in state purchases (PROMPYME program) Yes 24 30 39 29 2 5 4 3 No 76 70 61 71 98 95 96 97 Market information by MTPE or PRODUCE ministries Yes 10 15 24 15 1 2 4 2 No 90 85 76 85 99 98 96 98 BONOPYME Yes 8 14 17 12 2 4 3 2 No 92 86 83 88 98 96 97 98 Support for exporting by PROMPEX Yes 14 20 30 19 2 3 3 2 No 86 80 70 81 98 97 97 98 1 to 5 6 to 10 11 to 50 Total workers workers workers Number of businesses participating in… At least one program 20 19 18 57 Any two programs 13 11 10 34 Credit lines by COFIDE 6 2 4 12 Preferences to micro businesses in state purchases (PROMPYME program) 9 10 7 26 Market information by MTPE or PRODUCE ministries 4 4 7 15 BONOPYME 7 8 5 20 Support for exporting by PROMPEX 7 6 5 18 Memo item: survey sample 421 201 180 802 Source: World Bank survey of micro and small businesses. 95 Rationale for decision-making regarding business licensing Table A.11: Key factors behind the decision to not have a license, by business size Percent of businesses without a license that answered yes when asked if a particular factor had played a role or not in their decision not to have a license 1 to 5 6 to 10 11 to 50 Factor Total workers workers workers The process is too time consuming 71 65 65 69 The costs of running a business with a license are too high 67 56 61 63 My business is too small to have a license 64 47 35 57 The process is too expensive 59 53 48 56 I see no benefits in getting it 56 55 45 54 The forms are too complicated 50 39 42 46 There is no obligation to get it 47 44 45 46 No business like mine has a license 43 45 29 42 I do not know how to obtain it 39 36 52 40 No point obtaining it since fines are rare 36 36 45 37 Number of respondents 250 135 149 534 Source: World Bank survey of micro and small businesses. Table A.12: Knowledge of where to formalize Percent of businesses 1 to 5 6 to 10 11 to 50 All workers workers workers Do you know where is the office to get the municipal license? Yes 91 97 93 93 Do you know where is the office to get the RUC number? Yes 87 94 91 90 Do you know where is the office of SUNAT? Yes 87 94 94 90 Source: World Bank survey of micro and small businesses. 96 Rationale for decision-making regarding labor relations Table A.13: Share of workers earning less than minimum wage, by business size Percent of workers. Mean of responses to the question: In your area of activity, what percentage of workers do you estimate earn less than the minimum wage? 1 to 5 workers 6 to 10 workers 11 to 50 workers Total All 62 52 45 56 Shoe and leather manufacturing 61 41 43 52 Textile and apparel manufacturing 63 54 53 58 Manufacturing of wood products and furniture 60 53 46 55 Manufacturing of metallic products 58 53 37 52 Retail of foodstuffs 73 67 57 68 Transport by land 43 36 40 41 Restaurants and hotels 72 63 42 62 Arequipa 64 49 48 57 cusco 64 57 35 59 Huancayo 75 70 63 72 Lima 54 49 43 50 Trujillo 64 47 46 56 Source: World Bank survey of micro and small businesses. Table A.14: Reasons for not paying the minimum wage, by business size Percent of businesses that answered yes when asked if a factor explains why workers are paid less than the minimum wage 1 to 5 workers 6 to 10 workers 11 to 50 workers Total There are many people willing to work for less than the 81 80 81 81 minimum wage The business is not productive enough to allow paying the 71 66 51 65 minimum wage It is not compulsory to pay the minimum wage to workers 53 46 54 51 that are not on the payroll No firm in my sector pays the minimum wage to its 55 42 41 49 workers You do not have to pay family members to work in the 40 38 36 39 family business Source: World Bank survey of micro and small businesses. 97 Table A.15: Percent of workers not on the payroll, by business size Percent of workers. Mean of responses to the question: In your area of activity, what percentage of workers do you estimate that are not on the payroll? 1 to 5 workers 6 to 10 workers 11 to 50 workers Total All 75 68 61 70 Shoe and leather manufacturing 76 76 80 77 Textile and apparel manufacturing 74 64 63 69 Manufacturing of wood products and furniture 72 70 59 69 Manufacturing of metallic products 67 70 51 64 Retail of foodstuffs 79 71 63 74 Transport by land 79 61 62 71 Restaurants and hotels 77 61 52 67 Arequipa 75 70 65 72 cusco 76 73 44 72 Huancayo 76 73 63 73 Lima 73 67 60 68 Trujillo 76 59 68 69 Source: World Bank survey of micro and small businesses. Table A.16: Reasons for not having workers on payroll, by business size Percent of businesses that answered yes when asked if a factor explains why workers are not on the payroll 1 to 5 workers 6 to 10 workers 11 to 50 workers Total There are many people willing to work without 88 90 89 89 being on the payroll One would have to pay the worker holidays and 89 84 81 86 other benefits if they were on the payroll It would be very expensive to fire a worker that is 77 72 64 73 on the payroll It would be very complicated to fire a worker that is 76 69 59 70 on the payroll You have to pay them a higher salary 77 67 55 70 Having an employee on the payroll creates more 71 64 60 67 paperwork No busineses in my sector has its employees on the 58 53 46 54 payroll You do not need to include family members on the 50 39 37 44 payroll Source: World Bank survey of micro and small businesses. 98 Annex B. Econometric results on the choice of tax regime and determinants of tax inspections Table B.1: Choice of tax regime - probit regression Marginal effects from probits looking at choice of 1. RER vs RUS 2. RER vs Regimen General In both cases RER is coded as 1. There are only 86 firms with RER. (1) (2) Female owner 0.00889 0.0629 (0.056) (0.053) Owner's Age 0.000767 0.000842 (0.0025) (0.0022) Owner is Married 0.0698 0.0180 (0.053) (0.046) Owner has Tecnico education 0.0479 -0.0299 (0.098) (0.075) Owner has University education 0.0253 -0.105* (0.083) (0.059) Owner has complete secondary education 0.132* 0.0124 (0.071) (0.061) Parent owned a business 0.0424 0.00680 (0.053) (0.044) Age of firm 0.000199 0.00313 (0.0028) (0.0022) SUNAT inspection rate in city*industry -0.210 -0.342 (0.22) (0.21) Shoe/Leather 0.0870 0.239* (0.097) (0.12) Textiles/Clothing 0.108 0.0459 (0.11) (0.095) Wood products 0.309*** 0.0504 (0.12) (0.088) Metal products 0.164 0.00268 (0.11) (0.084) Food stuffs -0.0470 0.00991 (0.088) (0.094) Transport 0.272** 0.0647 (0.14) (0.100) Arequipa 0.279** 0.196** (0.11) (0.090) Trujillo 0.221* 0.163 (0.13) (0.13) Huancayo 0.0150 0.290** (0.100) (0.13) Cusco 0.0863 0.344*** (0.11) (0.13) 6 to 10 workers 0.228*** 0.0688 (0.061) (0.055) 11 to 50 workers 0.233** -0.0526 (0.099) (0.054) Log distance to the SUNAT Office -0.00218 0.0611 (0.049) (0.043) Log distance to City Center -0.0232 -0.0577 (0.045) (0.040) Firm has limited liability 0.318** -0.111** (0.13) (0.045) Observations 345 389 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 99 Table B.2: Choice of tax regime - multinomial logit (general regime) Multinomial Logit Marginal effects of being under the Regimen General. variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+ ----------- --------------- ------ ------- ---------- ---------- --------- female*| -0.045 0.053 -0.850 0.396 -0.150 0.059 0.299 ownerage | -0.002 0.002 -0.740 0.461 -0.007 0.003 41.664 married*| 0.027 0.051 0.530 0.595 -0.073 0.127 0.472 ed_tec~o*| 0.008 0.088 0.100 0.923 -0.164 0.181 0.131 ed_uni~y*| 0.098 0.074 1.310 0.190 -0.048 0.243 0.307 ed_sec~e*| 0.033 0.069 0.470 0.635 -0.102 0.168 0.384 parent~s*| 0.055 0.049 1.120 0.264 -0.041 0.150 0.407 firmage | -0.004 0.003 -1.430 0.153 -0.009 0.001 12.116 inspec.. | 0.211 0.241 0.870 0.382 -0.262 0.683 0.206 shoele~r*| -0.187 0.087 -2.150 0.032 -0.358 -0.016 0.140 textil~g*| 0.115 0.088 1.300 0.193 -0.058 0.287 0.133 woodpr~s*| 0.134 0.085 1.580 0.113 -0.032 0.300 0.160 metalp~s*| 0.161 0.082 1.950 0.051 -0.001 0.322 0.164 foodst~s*| -0.039 0.089 -0.440 0.661 -0.215 0.136 0.153 transp~t*| 0.093 0.098 0.950 0.342 -0.099 0.285 0.091 arequipa*| -0.082 0.080 -1.020 0.308 -0.240 0.076 0.116 trujullo*| -0.042 0.116 -0.360 0.718 -0.270 0.186 0.150 huancayo*| -0.255 0.082 -3.100 0.002 -0.416 -0.094 0.117 cusco*| -0.262 0.087 -3.000 0.003 -0.433 -0.091 0.144 mediana1*| 0.061 0.055 1.120 0.264 -0.046 0.169 0.279 grande1*| 0.301 0.055 5.510 0.000 0.194 0.408 0.262 ldista~T | -0.110 0.044 -2.500 0.013 -0.196 -0.024 1.391 ldista~y | 0.100 0.045 2.220 0.027 0.012 0.189 1.312 limite~y*| 0.378 0.050 7.500 0.000 0.279 0.477 0.202 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 100 Table B.3: Choice of tax regime - multinomial logit (RUS) Multinomial Logit Marginal effects for being under the RUS ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+ ----------- --------------- ------ ------- ---------- ---------- --------- female*| 0.029 0.050 0.590 0.558 -0.069 0.128 0.299 ownerage | 0.001 0.002 0.530 0.595 -0.003 0.006 41.664 married*| -0.048 0.048 -1.000 0.318 -0.142 0.046 0.472 ed_tec~o*| -0.017 0.080 -0.210 0.831 -0.174 0.140 0.131 ed_uni~y*| -0.059 0.067 -0.880 0.381 -0.191 0.073 0.307 ed_sec~e*| -0.079 0.060 -1.300 0.193 -0.197 0.040 0.384 parent~s*| -0.060 0.045 -1.310 0.189 -0.149 0.029 0.407 firmage | 0.002 0.003 0.870 0.382 -0.003 0.007 12.116 inspec.. | -0.012 0.218 -0.060 0.956 -0.440 0.416 0.206 shoele~r*| 0.081 0.087 0.930 0.354 -0.090 0.251 0.140 textil~g*| -0.148 0.068 -2.180 0.029 -0.281 -0.015 0.133 woodpr~s*| -0.224 0.056 -3.990 0.000 -0.334 -0.114 0.160 metalp~s*| -0.196 0.060 -3.270 0.001 -0.314 -0.078 0.164 foodst~s*| 0.059 0.083 0.710 0.476 -0.104 0.222 0.153 transp~t*| -0.183 0.067 -2.720 0.007 -0.314 -0.051 0.091 arequipa*| -0.091 0.071 -1.290 0.198 -0.229 0.048 0.116 trujullo*| -0.113 0.088 -1.290 0.196 -0.285 0.058 0.150 huancayo*| 0.169 0.097 1.740 0.082 -0.022 0.360 0.117 cusco*| 0.119 0.097 1.230 0.219 -0.071 0.308 0.144 mediana1*| -0.164 0.044 -3.770 0.000 -0.250 -0.079 0.279 grande1*| -0.324 0.042 -7.620 0.000 -0.407 -0.240 0.262 ldista~T | 0.092 0.041 2.270 0.023 0.012 0.172 1.391 ldista~y | -0.076 0.041 -1.830 0.067 -0.157 0.005 1.312 limite~y*| -0.366 0.041 -8.970 0.000 -0.446 -0.286 0.202 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 101 Table B.4: Choice of tax regime - multinomial logit (RER) Multinomial Logit Marginal effects for being under the RER ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+ ----------- --------------- ------ ------- ---------- ---------- --------- female*| 0.016 0.034 0.470 0.639 -0.051 0.082 0.299 ownerage | 0.001 0.001 0.390 0.698 -0.002 0.003 41.664 married*| 0.021 0.031 0.660 0.508 -0.040 0.082 0.472 ed_tec~o*| 0.009 0.058 0.150 0.883 -0.106 0.123 0.131 ed_uni~y*| -0.039 0.043 -0.900 0.366 -0.122 0.045 0.307 ed_sec~e*| 0.046 0.043 1.060 0.287 -0.039 0.131 0.384 parent~s*| 0.005 0.030 0.170 0.864 -0.054 0.065 0.407 firmage | 0.002 0.002 1.010 0.314 -0.001 0.004 12.116 inspec.. | -0.198 0.137 -1.450 0.147 -0.467 0.070 0.206 shoele~r*| 0.107 0.080 1.340 0.180 -0.049 0.262 0.140 textil~g*| 0.033 0.067 0.500 0.620 -0.099 0.165 0.133 woodpr~s*| 0.090 0.074 1.220 0.222 -0.054 0.234 0.160 metalp~s*| 0.035 0.066 0.530 0.596 -0.095 0.165 0.164 foodst~s*| -0.020 0.061 -0.330 0.742 -0.139 0.099 0.153 transp~t*| 0.090 0.084 1.060 0.288 -0.076 0.255 0.091 arequipa*| 0.173 0.077 2.240 0.025 0.021 0.324 0.116 trujullo*| 0.155 0.104 1.490 0.135 -0.049 0.359 0.150 huancayo*| 0.085 0.084 1.020 0.309 -0.079 0.250 0.117 cusco*| 0.143 0.093 1.540 0.123 -0.039 0.325 0.144 mediana1*| 0.103 0.041 2.490 0.013 0.022 0.184 0.279 grande1*| 0.023 0.042 0.550 0.581 -0.059 0.105 0.262 ldista~T | 0.018 0.029 0.600 0.551 -0.040 0.075 1.391 ldista~y | -0.024 0.028 -0.890 0.374 -0.078 0.029 1.312 limite~y*| -0.013 0.037 -0.340 0.737 -0.086 0.061 0.202 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 102 Table B.5: Probit regression on the determinants of tax inspections Who does SUNAT inspect? Marginal effects from probit estimation of probability of receiving an inspection (1) (2) (3) (4) Log distance to SUNAT office -0.0337*** -0.0289** -0.0390** -0.0121 (0.012) (0.012) (0.017) (0.026) Has a RUC 0.169*** 0.161*** 0.140*** (0.027) (0.026) (0.028) 6-10 workers 0.0613* 0.0552 0.0496 (0.037) (0.036) (0.037) 11-50 workers 0.0901** 0.124*** 0.0838** (0.040) (0.042) (0.042) Arequipa 0.113** 0.0975* (0.055) (0.057) Trujillo 0.295*** 0.246*** (0.058) (0.089) Huancayo 0.0402 0.0437 (0.061) (0.063) Cusco 0.232*** 0.225*** (0.064) (0.072) Female owner -0.0430 (0.029) Owner's Age -0.000396 (0.0014) Owner is Married -0.0455 (0.029) Owner has Tecnico education 0.132* (0.070) Owner has University education 0.144*** (0.055) Owner has complete secondary education 0.0674 (0.044) Parent owned a business 0.0268 (0.029) Age of firm 0.00215 (0.0015) In business to care for family -0.0103 (0.031) In business for flexibility 0.0304 (0.032) In business because can't find wage job -0.0507 (0.033) In business for prospect of growth 0.00293 (0.056) Shoe/Leather -0.0305 (0.045) Textiles/Clothing -0.0763* (0.039) Wood products -0.00140 (0.049) Metal products -0.0404 (0.044) Food stuffs 0.136** (0.063) Transport 0.00379 (0.055) Log distance to the city center -0.0245 (0.026) Observations 801 801 801 801 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 103 Annex C. Econometric results on the determinants of informality Determinants of Public Medical Coverage of Workers Columns 1-4: marginal effects from probit of any worker being covered Column 5: results dropping family enterprises Column 5: OLS regression of share of workers covered Column 6: OLS regression of share of workers covered conditional on having any worker covered Column 7: Fractional logit model of share of workers covered, to account for proportion Any worker covered Share covered (1) (2) (3) (4) (5) (6) (7) (8) Female owner 0.0114 0.0151 0.0221 0.0422 0.0417 0.0298 -0.0107 0.331 (0.039) (0.039) (0.039) (0.043) (0.055) (0.027) (0.061) (0.25) Owner's Age -0.00244 -0.00262 -0.00116 -0.00131 -0.00228 -0.000559 -0.00169 -0.00606 (0.0017) (0.0017) (0.0017) (0.0018) (0.0023) (0.0011) (0.0024) (0.013) Owner is Married 0.108*** 0.108*** 0.109*** 0.111*** 0.142*** 0.0682*** 0.0321 0.635*** (0.036) (0.036) (0.036) (0.038) (0.047) (0.025) (0.055) (0.24) Owner has Tecnico education 0.0923 0.0950 0.0837 0.114 0.116 0.0896** 0.303*** 1.126*** (0.073) (0.073) (0.074) (0.081) (0.10) (0.044) (0.10) (0.40) Owner has University education 0.258*** 0.258*** 0.201*** 0.167** 0.223*** 0.106*** 0.256*** 1.211*** (0.062) (0.064) (0.066) (0.070) (0.080) (0.038) (0.092) (0.37) Owner has complete secondary education 0.0517 0.0461 0.0548 0.0317 0.0328 0.0538 0.371*** 0.838** (0.052) (0.052) (0.053) (0.055) (0.071) (0.033) (0.094) (0.34) Parent owned a business -0.0182 -0.0183 -0.0232 -0.0292 0.0289 -0.00720 -0.00438 -0.0954 (0.035) (0.035) (0.035) (0.036) (0.047) (0.024) (0.058) (0.22) Age of firm 0.00764*** 0.00758*** 0.00607*** 0.00564*** 0.00771*** 0.00281** 0.000110 0.0268** (0.0018) (0.0018) (0.0018) (0.0020) (0.0025) (0.0013) (0.0026) (0.011) Labor inspection rate in city*industry 0.829** 0.860** 0.823** 0.729** 0.651 0.443* -0.123 4.073* (0.33) (0.33) (0.33) (0.35) (0.44) (0.23) (0.52) (2.37) Shoe/Leather -0.183*** -0.177*** -0.185*** -0.181*** -0.199*** -0.0749 0.130 -0.932** (0.042) (0.044) (0.040) (0.039) (0.059) (0.048) (0.12) (0.41) Textiles/Clothing -0.142*** -0.135*** -0.132*** -0.0828 -0.141** -0.0286 0.0844 -0.264 (0.050) (0.051) (0.050) (0.058) (0.071) (0.051) (0.12) (0.46) Wood products -0.0528 -0.0449 -0.0473 -0.0535 -0.0767 0.00253 0.0843 0.0487 (0.059) (0.060) (0.058) (0.058) (0.078) (0.047) (0.094) (0.37) Metal products -0.0926* -0.0908* -0.0757 -0.0847 -0.0620 -0.0470 0.0101 -0.461 (0.053) (0.053) (0.055) (0.053) (0.078) (0.047) (0.095) (0.39) Food stuffs -0.0377 -0.0252 -0.0116 -0.0371 0.000893 -0.0305 -0.0260 -0.365 (0.061) (0.063) (0.065) (0.063) (0.095) (0.049) (0.10) (0.43) Transport -0.0859 -0.0867 -0.0701 -0.0660 -0.113 -0.0135 0.106 -0.0605 (0.056) (0.056) (0.057) (0.058) (0.072) (0.050) (0.11) (0.38) Arequipa -0.131*** -0.133*** -0.0913** -0.0264 -0.131** -0.0536 -0.176* -0.602* (0.040) (0.040) (0.044) (0.054) (0.057) (0.040) (0.093) (0.32) Trujillo -0.145** -0.147*** -0.102 -0.0783 -0.109 -0.0267 -0.00731 -0.430 (0.057) (0.057) (0.063) (0.068) (0.086) (0.053) (0.14) (0.56) Huancayo -0.182*** -0.177*** -0.134*** -0.0943* -0.103 -0.0934* -0.197* -0.900** (0.038) (0.039) (0.046) (0.055) (0.077) (0.050) (0.12) (0.45) Cusco -0.221*** -0.228*** -0.185*** -0.152*** -0.170*** -0.0804* 0.127 -0.897* (0.032) (0.031) (0.037) (0.043) (0.063) (0.045) (0.12) (0.53) Log distance to the SUNAT Office -0.0309 -0.0305 -0.0465 -0.0296 -0.0387 -0.0226 -0.00683 -0.132 (0.032) (0.032) (0.032) (0.033) (0.043) (0.023) (0.053) (0.19) Log distance to City Center -0.0386 -0.0429 -0.0253 -0.00961 -0.00907 -0.00674 0.00942 -0.0912 (0.033) (0.033) (0.033) (0.034) (0.045) (0.022) (0.052) (0.22) In business to care for family -0.0634 -0.0468 -0.0561 -0.0549 -0.0274 0.0966 -0.231 (0.040) (0.040) (0.042) (0.052) (0.026) (0.063) (0.21) In business for flexibility 0.0113 0.0144 0.0259 0.0111 -0.0261 -0.115* -0.0785 (0.041) (0.041) (0.042) (0.055) (0.029) (0.064) (0.25) In business because can't find wage job 0.00885 0.0366 0.0224 0.0339 -0.000835 -0.0867 0.0372 (0.038) (0.037) (0.038) (0.050) (0.026) (0.062) (0.25) In business for prospect of growth -0.0932 -0.145* -0.162* -0.128 -0.0611 -0.0876 -0.560 (0.076) (0.082) (0.091) (0.10) (0.047) (0.11) (0.39) 6 to 10 workers 0.158*** 0.0331 0.184*** -0.0442 -0.279*** -0.224 (0.047) (0.048) (0.062) (0.030) (0.074) (0.29) 11 to 50 workers 0.386*** 0.177*** 0.434*** 0.0633* -0.196** 0.323 (0.052) (0.064) (0.060) (0.037) (0.079) (0.32) Log November Sales 0.0766*** 0.0523*** 0.0181 0.442*** (0.015) (0.010) (0.022) (0.11) Observations 713 713 713 626 504 609 159 609 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 104 Annex D. The impact of informality – methodology and impact on profits and on access to credit Impact on profitability - methodology We consider three approaches for measuring the impact of formality on profitability. Our methods include: (i) an ordinary least squares regression (OLS); (ii) a Treatment Effects regression that tries to take account of the potential endogeneity of formality; and (iii) an estimate using propensity score matching. Methodology for the OLS regression. The ordinary least squares regression is based on the following equation: (1) lnPROFITS  �  �Firms'Characteri  �Owners'Characteri  �Industry  �Location  �Formal  � i stics i stics i i i i The outcome of interest here is θ, which measures the average increase in log profits associated with being formal, conditional on the other variables included in the regression. Assigning a causal interpretation to formality based on this estimation requires assuming selection on the observable variables included in this equation, as well as assuming that the linear functional form adequately captures profits. A key concern with OLS estimation of equation (1) is that the error term ε is correlated with formality. There are several possible reasons for this concern. First, if we exclude firm size from equation (1), we would expect a positive bias, due to larger firms earning higher profits and also being more likely to be formal. In the OLS results we should therefore expect the estimate of θ to become less positive as we include firm size. Second, a concern is that ε may include unmeasured ability of the firm, with higher ability leading to more profits and also affecting the decision to become formal. If ability is a complement to formality, we should again expect this to lead to an upward bias in θ. Methodology for the Treatment Effects regression. A second approach to estimation is to take account of the potential endogeneity of formality by instrumenting for formality status when estimating equation (1). Since formality is a binary variable, we use STATA’s treatreg command to fit a maximum-likelihood Treatment Effects model. The instrument we rely on is the distance from the firm to the location of the office where the firm would have to register. The exclusion restriction we rely on here is that, conditional on the distance of a firm to the city center, and the average tax enforcement in a firm’s city and industry pair, distance to the SUNAT office has no direct impact on profitability. One possible reason this assumption could be violated would be if firms choose where to operate with the location of the SUNAT office in mind. However, within a large city it seems that the location of the tax office is not a main concern when deciding where to locate. Methodology for the propensity score matching. Finally, to complement the OLS and Treatment Effects regressions, we also provide estimates of the impact of being formal on profits using propensity score matching. Propensity score matching assumes that all selection occurs on observables, but does not require assuming a linear function form. We use the same variables as in the OLS regressions, along with higher-order interaction terms in carrying out the match. 105 Table D.1: "Impact" of having a RUC on log profits Columns 1-7 OLS regression, Columns 8-9 2SLS (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) OLS OLS OLS OLS OLS OLS OLS 2SLS 2SLS W-OLS W-2SLS RUC 1.030*** 0.960*** 0.840*** 0.532*** 0.299** 0.371* 0.380* -1.181 0.166 0.531*** -0.129 (0.11) (0.11) (0.12) (0.11) (0.12) (0.21) (0.21) (3.40) (2.12) (0.13) (1.99) Female owner -0.334*** -0.327** -0.286** -0.266** -0.222 -0.220 -0.402** -0.293** -0.310** -0.316** (0.13) (0.14) (0.13) (0.13) (0.19) (0.19) (0.20) (0.13) (0.13) (0.13) Owner's Age -0.0250*** -0.0234*** -0.0156*** -0.0110** -0.0172* -0.0170* -0.0266*** -0.0154*** -0.0141** -0.0134** (0.0049) (0.0054) (0.0053) (0.0051) (0.0095) (0.0098) (0.0085) (0.0056) (0.0055) (0.0060) Owner is Married 0.204* 0.176 0.126 0.114 0.175 0.191 0.219 0.130 0.149 0.150 (0.11) (0.12) (0.11) (0.11) (0.17) (0.17) (0.16) (0.11) (0.11) (0.11) Owner has Tecnico education 0.0699 -0.197 -0.147 -0.145 0.394 0.457 0.0296 -0.111 -0.0339 0.0366 (0.21) (0.24) (0.22) (0.22) (0.32) (0.32) (0.45) (0.29) (0.23) (0.30) Owner has University education 0.450** 0.271 0.114 0.0738 -0.0806 -0.0653 0.754 0.179 0.144 0.267 (0.18) (0.19) (0.18) (0.18) (0.30) (0.30) (0.84) (0.41) (0.20) (0.41) Owner has complete secondary education 0.199 -0.0647 -0.0354 0.00969 0.259 0.253 0.0955 -0.00934 0.0156 0.0634 (0.15) (0.16) (0.15) (0.15) (0.26) (0.26) (0.33) (0.21) (0.17) (0.22) Parent owned a business 0.163 0.201* 0.192* 0.151 -0.113 -0.117 0.124 0.178 0.158 0.134 (0.11) (0.12) (0.11) (0.11) (0.20) (0.19) (0.19) (0.13) (0.11) (0.14) Age of firm 0.0104* 0.0152** 0.00927 0.00834 0.00333 0.00473 0.0290 0.0112 0.00981 0.0129 (0.0059) (0.0064) (0.0061) (0.0058) (0.0099) (0.010) (0.024) (0.013) (0.0066) (0.011) SUNAT inspection rate in city*industry 0.320 0.219 0.464 0.235 -0.0830 0.0649 0.0277 0.448 0.799 0.749 (0.52) (0.61) (0.57) (0.54) (0.91) (0.87) (0.72) (0.53) (0.67) (0.66) Shoe/Leather -0.244 -0.242 -0.297* -0.235 -0.526* -0.509* -0.422 -0.335 -0.282* -0.344 (0.18) (0.19) (0.17) (0.17) (0.30) (0.30) (0.39) (0.29) (0.17) (0.26) Textiles/Clothing -0.0542 -0.0939 -0.155 -0.175 -0.139 -0.135 -0.422 -0.220 -0.0952 -0.197 (0.18) (0.18) (0.17) (0.17) (0.28) (0.28) (0.60) (0.42) (0.17) (0.35) Wood products 0.0379 0.107 0.0730 0.0386 -0.108 -0.0816 -0.0736 0.0401 0.0657 0.0180 (0.17) (0.17) (0.16) (0.17) (0.30) (0.30) (0.39) (0.27) (0.16) (0.22) Metal products -0.206 -0.261 -0.210 -0.210 -0.736** -0.727** -0.229 -0.201 -0.189 -0.169 (0.18) (0.19) (0.18) (0.18) (0.30) (0.30) (0.24) (0.19) (0.19) (0.20) Food stuffs -0.374 -0.316 -0.337 -0.222 -0.573 -0.570 -0.219 -0.324 -0.519* -0.493 (0.25) (0.29) (0.26) (0.25) (0.42) (0.43) (0.31) (0.23) (0.31) (0.31) Transport 0.330 0.392 0.232 0.194 -0.0861 0.0395 0.0174 0.154 0.187 0.0350 (0.22) (0.26) (0.25) (0.24) (0.39) (0.37) (0.69) (0.50) (0.24) (0.49) Arequipa -0.351** -0.489*** -0.370** -0.340** 0.0331 0.0855 -0.708* -0.402 -0.366** -0.423* (0.16) (0.16) (0.15) (0.15) (0.23) (0.23) (0.42) (0.25) (0.15) (0.23) Trujillo -0.529*** -0.647*** -0.549*** -0.574*** -0.215 -0.220 -0.729*** -0.556*** -0.581*** -0.593*** (0.19) (0.21) (0.19) (0.19) (0.26) (0.26) (0.28) (0.20) (0.20) (0.20) Huancayo -0.555*** -0.696*** -0.439** -0.458** -0.0400 0.0241 -0.819** -0.443** -0.438** -0.443** (0.19) (0.20) (0.20) (0.19) (0.28) (0.28) (0.33) (0.20) (0.20) (0.20) Cusco -0.726*** -0.770*** -0.495** -0.489** -0.213 -0.217 -0.877** -0.492** -0.609** -0.593** (0.21) (0.25) (0.24) (0.23) (0.41) (0.41) (0.35) (0.24) (0.27) (0.27) Log distance to City Center 0.0592 0.0803 0.113* 0.100 0.125 0.137 0.0508 0.110 0.121* 0.113 (0.067) (0.070) (0.068) (0.067) (0.10) (0.11) (0.099) (0.071) (0.070) (0.074) In business to care for family -0.219* -0.149 -0.118 -0.199 -0.204 -0.382 -0.170 -0.120 -0.167 (0.13) (0.13) (0.13) (0.22) (0.23) (0.31) (0.17) (0.14) (0.19) In business for flexibility -0.224 -0.237 -0.159 -0.268 -0.304 -0.202 -0.231* -0.250 -0.235 (0.17) (0.16) (0.15) (0.31) (0.32) (0.17) (0.14) (0.17) (0.17) In business because can't find wage job -0.0528 -0.0507 0.0216 0.174 0.211 -0.0519 -0.0535 -0.0134 -0.0179 (0.12) (0.12) (0.11) (0.19) (0.19) (0.15) (0.12) (0.12) (0.12) In business for prospect of growth 0.391 0.301 0.279 0.353 0.376 0.532 0.324 0.451 0.485 (0.30) (0.29) (0.28) (0.43) (0.42) (0.39) (0.28) (0.35) (0.34) Entrepreneurial Self-efficacy 0.106*** 0.0763*** 0.0581** 0.0662* 0.0641 0.146** 0.0800** 0.0669*** 0.0739** (0.027) (0.025) (0.025) (0.040) (0.039) (0.073) (0.032) (0.025) (0.034) 6 to 10 workers 0.562*** 0.643 0.603*** 0.743* (0.12) (0.49) (0.13) (0.43) 11 to 50 workers 1.201*** 1.293** 1.219*** 1.377*** (0.17) (0.55) (0.17) (0.53) Log capital stock 0.0758 (0.061) Dummies for each worker category no no no no yes yes yes no no no no First stage F-statistic 1.19 2.11 3.54 Observations 615 614 520 520 520 205 205 520 520 520 520 R-squared 0.09 0.22 0.26 0.35 0.42 0.50 0.49 0.35 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education Robust Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 106 Table D.2: "Impact" of having a municipal license on log profits (1) (2) (3) (4) (5) (6) (7) (8) (9) OLS OLS OLS OLS OLS 2SLS 2SLS W-OLS W-2SLS Municipal Licence 0.546*** 0.409*** 0.324** 0.157 0.107 -0.756 -0.147 0.180 0.238 (0.11) (0.12) (0.13) (0.11) (0.11) (1.85) (1.49) (0.12) (1.46) Female owner -0.375*** -0.374*** -0.308** -0.279** -0.351** -0.299** -0.328** -0.330** (0.13) (0.14) (0.13) (0.13) (0.14) (0.13) (0.14) (0.14) Owner's Age -0.0266*** -0.0237***-0.0148*** -0.0103** -0.0258*** -0.0150*** -0.0127** -0.0126** (0.0051) (0.0057) (0.0054) (0.0051) (0.0071) (0.0056) (0.0056) (0.0060) Owner is Married 0.240** 0.190 0.133 0.116 0.219 0.139 0.148 0.147 (0.11) (0.12) (0.11) (0.11) (0.14) (0.12) (0.11) (0.12) Owner has Tecnico education 0.132 -0.114 -0.101 -0.118 -0.0170 -0.0727 0.0131 0.00768 (0.21) (0.24) (0.22) (0.22) (0.28) (0.24) (0.23) (0.26) Owner has University education 0.602*** 0.412** 0.186 0.105 0.601 0.230 0.229 0.221 (0.18) (0.19) (0.18) (0.18) (0.38) (0.28) (0.20) (0.27) Owner has complete secondary education 0.253* -0.00335 0.00306 0.0342 0.0123 0.00855 0.0651 0.0649 (0.15) (0.17) (0.15) (0.15) (0.17) (0.15) (0.17) (0.16) Parent owned a business 0.111 0.164 0.175 0.136 0.221 0.191 0.142 0.139 (0.11) (0.12) (0.11) (0.11) (0.16) (0.13) (0.11) (0.14) Age of firm 0.0144** 0.0184*** 0.0110* 0.00922 0.0241** 0.0123 0.0113* 0.0111 (0.0060) (0.0064) (0.0062) (0.0057) (0.012) (0.0085) (0.0067) (0.0087) Municipal inspection rate in city*industry 0.0817 0.943 0.899 0.466 1.400 1.018 0.936 0.918 (0.73) (0.79) (0.72) (0.67) (1.15) (0.94) (0.72) (0.85) Shoe/Leather -0.290 -0.0232 -0.124 -0.135 -0.0936 -0.149 -0.123 -0.118 (0.26) (0.27) (0.23) (0.22) (0.33) (0.30) (0.24) (0.25) Textiles/Clothing -0.0747 0.195 0.0686 -0.0472 -0.0779 -0.0122 0.129 0.144 (0.31) (0.32) (0.29) (0.28) (0.59) (0.50) (0.30) (0.47) Wood products 0.0443 0.253 0.195 0.100 0.134 0.159 0.189 0.195 (0.21) (0.23) (0.20) (0.20) (0.33) (0.29) (0.21) (0.25) Metal products -0.165 -0.0608 -0.0502 -0.126 -0.0910 -0.0590 -0.0385 -0.0368 (0.22) (0.22) (0.22) (0.21) (0.26) (0.23) (0.22) (0.22) Food stuffs -0.289 -0.179 -0.180 -0.130 -0.169 -0.177 -0.316 -0.316 (0.24) (0.26) (0.23) (0.24) (0.23) (0.20) (0.27) (0.27) Transport 0.0574 0.466 0.339 0.251 0.549 0.356 0.279 0.278 (0.29) (0.35) (0.32) (0.30) (0.35) (0.29) (0.31) (0.30) Arequipa -0.381** -0.558*** -0.411*** -0.347** -0.784* -0.466 -0.379** -0.369 (0.17) (0.17) (0.15) (0.16) (0.43) (0.32) (0.15) (0.30) Trujillo -0.325* -0.547*** -0.431** -0.503*** -0.939 -0.534 -0.389** -0.369 (0.18) (0.19) (0.18) (0.17) (0.70) (0.53) (0.18) (0.53) Huancayo -0.546*** -0.718*** -0.423** -0.431** -0.818*** -0.437** -0.405** -0.403** (0.19) (0.20) (0.19) (0.19) (0.29) (0.22) (0.19) (0.20) Cusco -0.687*** -0.844*** -0.470** -0.475** -0.976*** -0.490** -0.498** -0.496** (0.19) (0.22) (0.20) (0.20) (0.33) (0.24) (0.21) (0.21) Log distance to City Center 0.0673 0.0850 0.113 0.102 0.0201 0.0961 0.118 0.121 (0.071) (0.075) (0.071) (0.069) (0.14) (0.11) (0.072) (0.11) In business to care for family -0.252* -0.159 -0.123 -0.353 -0.183 -0.126 -0.121 (0.13) (0.13) (0.13) (0.22) (0.17) (0.15) (0.18) In business for flexibility -0.276 -0.261 -0.176 -0.106 -0.214 -0.282 -0.292 (0.18) (0.16) (0.15) (0.33) (0.27) (0.17) (0.30) In business because can't find wage job -0.0662 -0.0629 0.0162 -0.0451 -0.0561 -0.0260 -0.0275 (0.12) (0.12) (0.11) (0.14) (0.13) (0.12) (0.13) In business for prospect of growth 0.442 0.343 0.293 0.468 0.345 0.499 0.499 (0.30) (0.29) (0.27) (0.29) (0.25) (0.35) (0.34) Entrepreneurial Self-efficacy 0.118*** 0.0816*** 0.0582** 0.139*** 0.0861*** 0.0730*** 0.0720** (0.027) (0.025) (0.025) (0.046) (0.033) (0.025) (0.036) 6 to 10 workers 0.681*** 0.699*** 0.713*** 0.709*** (0.12) (0.16) (0.13) (0.15) 11 to 50 workers 1.293*** 1.347*** 1.296*** 1.286*** (0.17) (0.30) (0.17) (0.31) Observations 615 614 520 520 520 520 520 520 520 R-squared 0.03 0.16 0.22 0.34 0.42 0.34 Note: omitted dummies are Restaurants, Lima, <5 workers, Male, and Less than Secondary Education Robust Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 107 Table D.3: OLS Regression of log profits on both municipal license and RUC number Dependent variable: Log profits (1) (2) (3) RUC 0.975*** 0.850*** 0.532*** (0.12) (0.13) (0.13) Municipal Licence 0.202 0.129 0.0524 (0.13) (0.13) (0.12) Female owner -0.375*** -0.325** (0.14) (0.14) Owner's Age -0.0212*** -0.0131** (0.0058) (0.0056) Owner is Married 0.212* 0.158 (0.12) (0.11) Owner has Tecnico education -0.0235 0.0318 (0.25) (0.23) Owner has University education 0.325 0.157 (0.21) (0.20) Owner has complete secondary education 0.0357 0.0566 (0.18) (0.16) Parent owned a business 0.161 0.161 (0.12) (0.11) Age of firm 0.0130* 0.00828 (0.0071) (0.0067) Municipal inspection rate in city*industry 0.671 0.473 (0.93) (0.86) SUNAT inspection rate in city*industry 0.478 0.741 (0.82) (0.77) In business to care for family -0.153 -0.0942 (0.15) (0.14) In business for flexibility -0.261 -0.259 (0.18) (0.17) In business because can't find wage job -0.00247 -0.000332 (0.13) (0.12) In business for prospect of growth 0.569 0.440 (0.37) (0.35) Shoe/Leather -0.00553 -0.131 (0.28) (0.25) Textiles/Clothing 0.261 0.0923 (0.34) (0.31) Wood products 0.285 0.186 (0.24) (0.21) Metal products -0.0628 -0.0703 (0.23) (0.23) Food stuffs -0.429 -0.488 (0.39) (0.36) Transport 0.514 0.290 (0.37) (0.34) Arequipa -0.460*** -0.361** (0.17) (0.16) Trujillo -0.551** -0.505** (0.23) (0.22) Huancayo -0.589*** -0.365* (0.20) (0.20) Cusco -0.903*** -0.611** (0.29) (0.27) Log distance to city center 0.0834 0.114 (0.074) (0.071) 6 to 10 workers 0.652*** (0.13) 11 to 50 workers 1.264*** (0.17) Constant 6.560*** 6.872*** 6.538*** (0.10) (0.63) (0.59) Observations 520 520 520 R-squared 0.09 0.24 0.35 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 108 Access to credit Table D.4: Businesses with and without bank accounts Percent of businesses No account With bank With savings association With microfinance institution All businesses 57 35 6 2 By sector Shoe leather 62 27 8 3 Textile clothing 46 46 5 3 Metal works 52 43 3 2 Wood products 53 38 8 1 Foodstuffs 76 20 2 2 Restaurant 53 39 7 3 Transport 60 33 7 0 No. of workers Overall Lima Other regions 1–5 29 26 32 6 – 10 51 45 55 11 – 50 66 65 68 Source: World Bank survey of micro and small enterprises. Table D.5: Sources of finance among businesses that have borrowed in 2007 or 2006 Percent of businesses Personal From Banks Caja Edpyme Micro- Money Client Supplier Other Savings family and Municipal finance lender advance credit friends All businesses 77 14 11 3 1 1 1 2 2 5 By sector Shoe & leather 67 14 7 3 1 1 0 1 4 8 Textile clothing 62 12 14 3 1 0 1 2 1 5 Metal manuf. 84 13 9 3 0 0 2 3 1 3 Wood products 87 12 12 4 1 0 0 4 1 1 Foodstuffs 66 17 14 3 1 1 1 1 4 6 Restaurant 95 15 11 3 0 2 0 0 1 5 Transport 73 13 14 2 0 1 2 1 1 6 Source: World Bank survey of micro and small enterprises. 109 Table D.6: Use of credit to purchase inputs Percent of businesses 1 to 5 6 to 10 11 to 50 Total workers workers workers Did you buy inputs on credit? Yes 28 40 57 37 No 72 60 43 63 Total 100 100 100 100 Only businesses from manufacturing sectors Yes 22 43 59 35 No 78 58 41 65 Total 100 100 100 100 Percent that bought on credit by business size and sector Shoe and leather manufacturing 16 48 48 32 Textile and apparel manufacturing 30 33 72 39 Manufacturing of wood products and furniture 24 43 57 36 Manufacturing of metallic products 18 46 62 34 Retail of foodstuffs 64 78 68 69 Transport by land 29 25 42 31 Restaurants and hotels 16 10 52 24 Percent that bought on credit by business size and city Arequipa 22 47 39 32 Cusco 28 43 45 34 Huancayo 38 29 69 40 Lima 27 39 64 41 Trujillo 24 42 38 32 If purchasing on credit, what percentage of purchases was bought on credit? All 44 46 45 45 Shoe and leather manufacturing 47 47 48 47 Textile and apparel manufacturing 49 51 52 51 Manufacturing of wood products and furniture 39 38 41 39 Manufacturing of metallic products 45 50 52 49 Retail of foodstuffs 39 45 50 43 Transport by land 64 49 41 54 Restaurants and hotels 31 40 30 31 Arequipa 53 40 50 47 Cusco 41 48 19 41 Huancayo 50 34 43 45 Lima 40 51 49 46 Trujillo 49 46 29 43 If purchasing on credit, how many days to pay were you given? All 67 50 51 57 Shoe and leather manufacturing 23 33 34 31 Textile and apparel manufacturing 120 95 75 97 Manufacturing of wood products and furniture 86 66 61 71 Manufacturing of metallic products 105 58 36 63 Retail of foodstuffs 20 20 24 20 Transport by land 127 86 163 127 Restaurants and hotels 18 37 21 22 Arequipa 88 49 44 64 Cusco 72 53 163 78 Huancayo 39 30 52 40 Lima 75 51 49 58 Trujillo 54 59 16 46 Source: World Bank survey of micro and small businesses. 110 Table D.7: Sources of start-up capital Percent of businesses 1 to 5 workers 6 to 10 workers 11 to 50 workers All cities Personal savings 76 66 90 From family and friends 13 16 12 Formal finance* 14 18 15 Other** 8 11 10 Lima Personal savings 85 63 89 From family and friends 12 20 14 Formal finance* 9 15 16 Other** 6 14 7 Other cities Personal savings 69 69 92 From family and friends 14 13 10 Formal finance* 18 21 14 Other** 9 9 16 * Banks, Caja Municipal, Edpyme, Micro finance institutions; ** Money lender, client advance, supplier credit, and others. Source: World Bank survey of micro and small enterprises. Table D.8: Borrowing by sector Percent of businesses Businesses who took loan in Among those businesses that took a loan, share 2006 or 2007 that borrowed from formal sources All businesses 49 93 By sector Shoe & leather 60 91 Textile clothing 60 96 Metal manuf. 47 96 Wood products 51 94 Foodstuffs 53 91 Restaurant 37 87 Transport 41 90 Source: World Bank survey of micro and small businesses. 111 Table D.9: Sources of finance for most recent loan Percent of businesses, relative to businesses that took a loan 1 to 5 workers 6 to 10 workers 11 to 50 workers All cities Banks 51 66 80 Caja Municipal 32 18 4 Micro finance 5 5 3 Edpyme 5 7 4 Money lender 3 3 8 Other 3 2 1 Memo item: share of businesses that borrowed in 2007 or 2006 43 59 52 Lima Banks 82 88 89 Caja Municipal 7 0 0 Micro finance 2 0 0 Edpyme 0 5 2 Money lender 6 5 9 Other 4 2 0 Memo item: share of businesses that borrowed in 2007 or 2006 32 49 48 Other cities Banks 38 54 68 Caja Municipal 43 27 10 Micro finance 7 8 8 Edpyme 7 8 8 Money lender 2 3 5 Other 3 1 3 Memo item: share of businesses that borrowed in 2007 or 2006 50 67 59 Source: World Bank survey of micro and small businesses. Table D.10: Terms of loan (annual or monthly rates) by source of finance Number of businesses receiving loans under annual and monthly rates of interest Caja Micro- Money Bank Edpyme Other Total municipal finance lender Number of businesses that have taken a 126 30 6 9 1 2 174 loan with annual interest rate Number of businesses that have taken a 103 52 12 12 16 5 200 loan with monthly interest rate Sum of the above 229 82 18 21 17 7 374 Total number of businesses that have 245 82 18 21 17 9 392 taken a loan Note: the sum and the total number of businesses having taken a loan does not necessarily match since there are businesses which report having taken a loan but do not answer the question on whether the interest rate they received was on an annual or monthly basis Source: World Bank survey of micro and small businesses 112 Table D.11: Average length of loan in months No. of workers Overall Lima Other 1-5 14.3 14.6 14.2 6 - 10 15.1 16.5 14.4 11 - 50 14.0 11.5 17.4 Overall 14.5 14.0 14.8 Source: World Bank survey of micro and small businesses Table D.12: Home ownership - percent of entrepreneur owning their home Sector Overall Lima Other Regions Overall 53 48 57 Restaurant 57 39 67 Foodstuffs 59 56 63 Wood products 56 52 60 Shoe leather 56 51 60 Metal products 54 56 53 Textile clothing 43 41 45 Transport 40 36 43 No. of workers Overall Lima Other regions 1–5 48 42 52 6 – 10 52 41 60 11 – 50 63 62 66 Source: World Bank survey of micro and small enterprises. Table D.13: Business taking credit cards and checks, percent Sector Both Only Checks Only Credit Cards None Overall 6 19 5 69 Foodstuffs 2 7 3 89 Transport 4 15 0 79 Restaurant 8 2 15 75 Shoe leather 4 14 8 74 Textile clothing 4 30 2 64 Wood products 10 26 5 60 Metal works 9 37 4 49 Source: World Bank survey of micro and small enterprises. 113 Table D.14: Impact on getting a loan of having a license and RUC number Municipal license RUC OLS First-stage 2SLS Matching OLS First-stage 2SLS Matching F-stat F-stat All firms -0.0381 8.12 -0.245 -0.037 0.0862* 5.65 -0.680 0.117** (0.045) (0.42) (0.047) (0.052) (0.71) (0.055) 5 or fewer workers -0.0136 11.28 0.137 -0.008 0.00444 4.48 -0.532 -0.025 (0.060) (0.34) (0.060) (0.066) (0.58) (0.064) 6 to 10 workers -0.150* 0.69 . -0.185** 0.172 1.02 -0.973 0.231* (0.090) (0.086) (0.13) (1.68) (0.130) 11 - 50 workers 0.0534 0.02 0.015 0.297 0 0.498** (0.15) (0.151) (0.18) (0.210) 114 Annex E. Trajectories towards formality Table E.1: Time lapsed from opening business to getting RUC number and municipal license Percent of businesses Timeline for getting a RUC number 1 to 5 6 to 10 11 to 50 Total workers workers workers Since starting the business 48 55 56 52 Less than a year after starting the business 15 17 14 15 1 to 2 years after starting the business 15 12 10 13 3 - 4 years after starting the business 5 8 8 7 5 or more years after starting the business 16 8 12 13 Sample size 273 181 173 627 Timeline for getting a municipal license 1 to 5 6 to 10 11 to 50 Total workers workers workers Since starting the business 53 53 57 54 Less than a year after starting the business 19 22 14 19 1 to 2 years after starting the business 14 11 8 12 3 - 4 years after starting the business 4 8 7 6 5 or more years after starting the business 9 6 13 9 Sample size 249 135 149 533 Source: World Bank survey of micro and small businesses. 115 Annex F. Database of sector-specific labor regulations - methodology To prepare a dataset of labor regulations affecting specific sectors, a listing available in the Labor Ministry's website containing 833 legal instruments was used as a basis. In addition, to ensure that no legal instruments were overlooked, other sources were consulted. In this sense, other listings - available in the Labor Ministry's website - were used. That is the case of - which provided 81 legal instruments - - which provided 101 legal instruments - and - which provided 76 legal instruments to the counting. Combining all these sources the total number of legal instruments issued by the Ministry of Labor totaled 865. However, this total number contained very broad legal instruments since it included all regulations issued by the Labor Ministry and other legal instruments that were not really labor-specific. For example, appointments of officers in several positions in the Ministry and procurement decisions were included. This total number of legal instruments had to be filtered to create a long list of legal instruments that were truly affecting labor regulations. This long list of legal instruments affecting labor regulations had 284 legal instruments as 581 laws were excluded on the basis that they did not cover labor regulations. The excluded laws dealt largely with administration of the Labor Ministry (279), pensions (118), administrative sanctions for rehiring in the public sector (28), holidays (15), survey approvals (15), income taxation (11), and a group of other diverse legal instruments (115). Once non-labor regulations were removed from the listing, a second selection process involved identifying those labor regulations that applied to specific sectors. Out of the long list of 284 legal labor instruments, 162 legal instruments were removed because they were not specific to any determined sector. Out of the remaining 122 legal instruments affecting only specific sectors, 2 were excluded because they were specific not to sectors but to professions that cut across different sectors of activity,25 and 37 were excluded because they were specific to public sector activities. A final list of 83 legal instruments was identified as referring to labor regulation, including those that were identified as specific to private sector activities and were mapped to those areas of activity. 25 This is the case of the “law for the work of the biologist�, and the “law for the work of the psychologist�, which although are specific to certain professionals, is not specific to any activity. 116 Annex G. The Marginal Effective Tax Rate Methodology Conceptual Explanation of the METR There are two main "stakeholders" that hold an interest in firms: debt holders and equity holders (owners or "shareholders"). In order to satisfy these stakeholders, an investment must earn a rate of return after the payment of all business taxes which is greater than or equal to the (weighted average) "hurdle" rate of return required by these stakeholders. The hurdle rate is the minimum rate of return acceptable to the stakeholders. For example, say the (weighted average) hurdle rate of return is 10%. This means that all capital projects that earn a rate of return greater than 10%, after the payment of all business taxes, will be undertaken. So, if a firm invests in all projects with a rate of return greater than the hurdle rate, and it invests in projects with the highest rate of return first and then moves down the “menu� of capital projects available to it, the last (marginal) project undertaken will earn an after-tax rate of return of 10% exactly. The derivation of the METR formula can be quite complicated, but the idea can be conveyed in a simple example. As above, say the hurdle rate of return is 10% - i.e., the minimum (weighted average) rate of return required by the stakeholders in the project after the payment of all business taxes is 10%. Say the business tax system is such that in order to earn a rate of return of 10% after business taxes, an investment must earn a rate of return of 15% before business taxes. The METR is then 33.33% (determined as (.15-.10)/.15). The METR therefore measures the share of the investment’s pre-tax required rate of return needed to cover the tax costs associated with the investment. Note that the 33 1/3% METR could differ substantially from the statutory CIT rate because of the existence of other taxes, deductions, credits, allowances, etc. The METR will equal the statutory tax rate in the case of no debt financing, no investment incentives, no turnover or other implicit or explicit capital taxes, and if the tax depreciation rate is equal to the economic rate of depreciation.26 In general, the METR is higher (the tax distortion greater),  the higher is the statutory CIT rate and other tax rates on capital - capital taxes, business taxes, property taxes, turnover taxes;  the lower is the tax depreciation rate on capital relative to the "economic" depreciation rate;  the lower are various tax incentives - investment tax credits, investment allowances, etc,  the lower the proportion of investment financed by tax deductible debt Several things should be kept in mind when interpreting the METRs calculated for this study. First, the METR should not be confused with the average effective tax rate (AETR). The AETR measures total taxes as a share of total income while the METR measures taxes paid by a marginal investment as a share of the required before-tax rate of return on that investment. The METR is thus a forward-looking measure which measures the impact of taxes on the incentive to invest at the margin. The AETR is a backward looking measure which reflects previous investment decisions and taxes levied on the income generated by those past investments. METRs and AETRs can differ significantly. In particular, it is quite possible for the AETR to be quite high and the METR quite low, or vise-versa. Both measures are useful for tax policy analysis, but serve quite different purposes. The METR provides a measure of the incentive effects of the tax system on investment. The AETR provides a measure of the burden of the business tax system 26 This also presumes no personal taxes or, a small open economy with internationally mobile capital. See below. 117 and is a useful indicator of the distribution of the business tax burden across companies or sectors. Second, it is important to understand that the METR measures the impact of taxation on the incentive to invest in capital inputs. Taxes that corporations pay on other inputs that do not impinge on the rate of return to capital, in particular various taxes imposed on labor such as payroll taxes, are not reflected in the METR on capital. Third, it is important to emphasize that METR analysis is only one of several tools that should be employed in an analysis of the tax system. The calculations are based upon a hypothetical marginal capital investment subject to a rough caricature of the tax system. While several important aspects of the tax system can be incorporated into the calculations: the corporate income tax rate; tax depreciation allowances; investment allowances; tax credits; property and capital taxes; tax holidays; implicit sales taxes on inputs; the treatment of inventories; etc., several simplifying assumptions must be made, and important nuances of the tax code cannot be captured in METR calculations.27 The methodology thus captures only the “big picture� aspects of the tax code and should be interpreted in this light. Finally, it is important to note that the METR calculations reflect the statutory provisions of the tax system. As a rule, METR calculations do not reflect various administrative and compliance issues that are extremely important for determining how a tax system works “on the ground.�28 Taxes on capital can lead to several types of distortions in an economy. First, a high overall METR on capital is indicative of a tax system that discourages investment, generating what might be called an inter-temporal distortion. Second, differences in METRs across assets suggest a tax system that distorts the allocation of investment across different types of capital: inter-asset distortions. Third, differences in METRs across sectors is indicative of a tax system that introduces distortions in the allocation of investment across industries, leading to inter-sectoral distortions. All of these types of distortions: inter-temporal, inter-asset and inter-sectoral, can lead to an inefficient allocation of resources in the economy, and therefore to an economy that does not produce to its capacity; does not generate jobs to its capacity; and therefore does not grow to its capacity. More Formal Explanation of the METR Neoclassical investment theory tells us that a firm will invest in capital up to the point where, at the margin, the present value of the after-tax cash flow from the last dollar invested equals one dollar. Therefore the marginal unit of capital just breaks even in the sense that the present value of the cash flows after the payment of taxes just equals the one dollar cost. Thus, for a simple corporate tax system: (1) 1 = ∫ C(1-u)e-(rf-π-δ)tdt – t + uZ(1+t) where C is the asset’s pre-tax rental rate, u is the corporate income tax rate, rf is the firm’s after- tax nominal discount rate (defined in more detail below), π is the inflation rate, δ is the economic depreciation rate, Z is the present value of the tax depreciation deductions on $1 of capital 27 An example is loss carryforward provisions. While it is possible to incorporate loss carryforwards, this requires detailed information on the anticipated future profile of income and taxes, which is typically not available in reliable form. 28 An attempt to illustrate the importance of these issues in the METR framework is undertaken below. 118 (explained in more detail below), and t is the implicit tax rate on $1 of capital (for example a sales tax imposed on capital, an asset transfer tax, a turnover tax, etc.). The left hand side of the equation is the cost of $1 in capital while the right hand side is the present value of the after-tax returns to that marginal unit of capital. The first term on the right hand side is the present value of the after-tax cash flows (net of depreciation) generated by the marginal unit of capital; the second term is the sales taxes paid on that unit of capital; the third term is the value of the tax depreciation allowances. The term Z, which is the present value of the tax depreciation deductions on a $1 capital expenditure, depends upon the form of the tax depreciation allowances (i.e., straight line or declining balance), whether or not tax depreciation allowances are indexed for inflation, and can incorporate tax incentives such as tax allowances and investment tax credits. For example, under a declining balance approach to tax depreciation with no allowance for inflation, Z is equal to: Z = α/(rf+α) where α is the tax depreciation rate (i.e., 20%, in which case α=.2). If inflation indexing of tax depreciation allowances is allowed the nominal discount rate rf is replaced with the real discount rate rf-π. Under straight-line depreciation, where an asset is written off over T years, and again assuming no inflation adjustment, Z = ∑t=1 to T (1/T)/(1+rf)t An investment allowance granted at rate θ, for example in the presence of a declining balance approach, can be incorporated as follows: Z = θ + (1-fθ)α/(rf+α) Where f=1 if the investment allowance reduces the subsequent depreciation allowances and f=0 if it does not. An investment tax credit granted at rate φ can be incorporated by replacing θ with φ/(1-u). Integrating equation (1) to determine C, and subtracting the economic depreciation rate δ, gives the before-tax, after depreciation, rate of return required to cover the firm’s opportunity cost of funds and the taxes associated with a marginal investment, commonly referred to as the user cost of capital: (2) rg ≡ C - δ = (1+t)(rf-π+δ)(1-uZ)/(1-u) - δ This is akin to the before-tax hurdle rate of return discussed above. The user cost of capital, as written above, clearly reflects the corporate income tax and any sales taxes imposed on capital, which are levied on the demand side of the capital market. Personal taxes levied on the supply side of the capital market can enter the user cost of capital expression through the firm’s discount rate rf. The precise expression for rf (and the impact of personal taxes on the user cost of capital) depends upon several assumptions concerning the marginal source of funds to the firm and the extent to which capital is internationally mobile. The extent to which capital is mobile internationally, and therefore whether the domestic capital market should be thought of as closed or open, or some combination of the two, is a complicated question. It turns out that the assumption that one makes 119 in this regard has a big impact on the expression for rf and important implications for the effect of the domestic tax system on investment. To deal with this issue, we present several sets of calculations. First, we assume that large corporations operate in a small open economy with internationally mobile capital. This is a common assumption that is made in many METR studies. For small corporations, we take a slightly different approach. First, in order to compare the calculations directly with those of large corporations, we present a set of METRs under the small open capital market assumption. Second, we present another set of METRs for small businesses under the closed capital market assumption. The idea here is that small businesses do not have access to international financial markets, and may therefore be thought of as operating in a segmented, closed market for capital. One might think of the METRs for the closed economy case as the METRs on “entrepreneurial� firms. As will be shown below, an important difference between the open and closed capital markets scenarios is that in the open economy, domestic personal taxes levied on the return to saving have no impact on the level of domestic investment. In the closed economy case, on the other hand, domestic taxes levied on the return to savings at the personal level can affect domestic investment undertaken by small firms. In order to understand the distinction more easily, it is useful to start with the closed economy model and utilize a simple diagram. Diagram 1 (below) depicts the market for a homogeneous investment good in a closed economy. The investment demand schedule, I(r g), gives the level of investment as a function of the before-all-taxes (gross-of tax) rate of return. The savings supply schedule, S(rn) gives the level of savings as a function of the required after-all-taxes (net-of-tax) rate of return. All taxes mean both corporate and personal taxes. In terms of the conceptual framework laid out above, rg is the before-all-taxes hurdle rate of return and rn is the after-all- taxes hurdle rate of return. Diagram 1: Closed Capital Market Diagram 2: Open Capital Market with Internationally Mobile Capital S(rn) S(rn) rg rg ri rn rn I(rg) I(rg) Ie=Se Se Ie Of course in a closed economy the level of saving must equal the level of investment. This means that there is an equivalence between taxes imposed on savings (the supply side) and taxes imposed on investment (the demand side); taxes imposed on both sides of the market impact upon both investment and savings. In equilibrium, given the taxes imposed on both sides of the market, the level of investment and savings in the diagram is Ie=Se, the gross-of-tax rate of return is rg and the net-of-all-taxes rate of return is rn. 120 In the closed economy case the nominal cost of debt to the firms is the return to debt required to yield rn after personal taxes, given by: (3) rd = (rn+π)/(1-Ti) where Ti is the domestic PIT rate on interest. Similarly, the nominal cost of equity is given by: (4) re = (rn+π)/(1-Te) where Te is the domestic PIT rate on equity, which is a weighted average of the tax rate on dividends and the accrual equivalent tax rate on capital gains.29 The discount rate facing the firm, r f, is then a weighted average of the cost of debt and equity, given by: (5) rf = βrd(1-u) + (1-β)re where β is the debt/asset ratio. This reflects the fact that debt interest costs are deductible for CIT purposes, while the cost of equity is not. The marginal effective tax rate on capital is the share of the before-tax-rate of return (rg) required to pay the taxes on the investment (rg-rn): (6) METR = (rg – rn)/rg The METR therefore determines the share of the investment’s pre-tax required rate of return needed to cover the tax cost. In the closed economy model it is important to note that there is an equivalence between supply side taxes on savings and demand side taxes on investment: both lead to a decrease in investment (and savings). If capital is perfectly mobile internationally and the domestic capital market is small in the sense that savings and investment decisions in the country have no impact on the “world� interest rate, the link between domestic savings and domestic investment is broken. This means that the equivalence between taxes levied on investment and taxes levied on savings no longer holds. An important implication of this is that domestic personal taxes imposed on savings may then have very little impact on domestic investment, though these taxes can still affect domestic savings. This is most easily seen with reference to Diagram 2 (above). In an open economy the after- corporate-tax but before-personal-tax rate of return is fixed by world financial markets. Denote this fixed rate of return in real terms by ri. Demand side domestic taxes imposed on corporate capital increase the before-corporate-tax rate of return required to generate this after-corporate- tax rate of return to rg. Domestic investment in this case is Ie and the METR on capital in an open economy is given by the tax wedge (rg-ri) divided by the gross-of-tax rate of return rg. Supply side taxes imposed on domestic savings decrease the after-personal-tax rate of return to rn, yielding 29 More precisely, the tax rate on equity is T e=γTd+(1-γ)Tc, where Td is the tax rate on dividends and T c is the accrual equivalent tax rate on capital gains. The accrual equivalent tax rate on capital gains reflects the fact that capital gains are taxed on realization and not on accrual. T e thus takes the deferral of capital gains taxes until realization into account. If one presumes a ten year holding period for shares, this typically means that the accrual equivalent tax rate on capital gains is approximately half of the realized, statutory rate. 121 domestic savings of Se. In the case depicted, the excess of domestic investment over domestic savings (Ie-Se) is provided by foreign investment inflows. Importantly, while taxes on domestic savings affect savings and the share of domestic investment financed by foreigners, they have no impact on domestic investment itself (unlike in the closed economy model). In terms of the user cost of capital framework laid out above, the internationally determined cost of funds determines the firm’s discount rate, rf in the user cost of capital expression given in equation (2). An important consideration here is the identity of the marginal international investor. We assume that the marginal investor is an “average� of investors from G7 countries. If rd is the nominal cost of debt facing the firm, fixed by international markets, then the nominal cost of equity is given by the arbitrage condition that the after-tax return to debt equals the after-tax return to equity, rd(1-Td*)= re(1-Te*) where Td* is the tax rate on debt for the marginal international investor and Te* is the tax rate on equity. Then, in equilibrium the nominal cost of equity facing the firm fixed by international financial markets is: (7) re = rd(1-Td*)/(1-Te*) and the weighted average nominal discount rate facing the firm is: (8) rf = βrd(1-u) + (1-β)re The after-corporate-tax, before-personal-tax, real rate of return fixed by international financial markets is then: (9) ri = βrd + (1-β)re - π and the marginal effective tax rate on capital in a small open economy is then: (10) METR = (rg – ri)/rg All of the above formulae were developed for physical capital, in particular for depreciable capital like equipment and buildings. Non-depreciable capital, such as land, is easily accommodated by setting economic depreciation (δ) and the present value of tax depreciation allowances (Z) to zero in the expression for the user cost of capital, rg (equation (2)), giving: (11) rg = (1+t)(rf-π)/(1-u) For inventory capital the expression for the cost of capital can be shown to be: (12) rg = (1+t)(rf-π+uπf)/(1-u) where the parameter f is equal to 0 if LIFO inventory valuation is used for tax purposes and equal to 1 if FIFO is used. Thus, the use of FIFO inventory accounting is seen to increase the cost of holding inventory capital by virtue of the taxation of the inflationary increase in the cost of inventories. Issues Illuminated by METR Analysis The primary applications of METR analysis are twofold. First, the results of a METR analysis show the net effect of all components of the tax system on the level of the taxation of capital income generated by the marginal investment analyzed. Thus, a METR provides a measure of the actual tax burden on a prospective investment attributable to the existing (or proposed) tax system. Moreover, an appropriately weighted average of the METRs on specific types of 122 investments can be constructed to provide a measure of the overall level of taxation of capital income in the economy, showing how the tax system distorts investment decisions (and, if individual level taxes are considered, saving decisions as well) and thus introduces inefficiencies or “excess burdens� into the economy.30 Second, by considering a wide variety of investments that differ by asset, method of finance, investor or economic circumstances, METR analysis provides an indicator of the tax differentials that arise across different types of investments, that is, it shows how taxes affect the composition of investment. In particular, a METR analysis shows how the tax system results in a variety of distortions of investment decisions, thus creating additional efficiency losses, beyond those associated with simply taxing capital income at a uniform effective tax rate. The most commonly cited distortion is across types of assets, as differential taxation of different types of assets induces businesses to invest too heavily in tax-advantaged assets and too little in tax- disadvantaged assets. This of course translates into distortions across business sectors, as the tax system favors sectors with production processes that use tax-favored assets intensively and penalizes businesses that use relatively heavily taxed assets intensively. METRs provide an indication of the overall level of taxation of various forms of capital income and thus indicate how the tax system affects investment and saving decisions. Because they consider many aspects of the tax system, METR analyses often give very different results regarding the effects of the tax system on investment decisions than would a simple examination of statutory tax rates (or special preferences) in isolation. Effective tax rates that are far above or below the statutory rate indicate potential areas for reform, as relatively high positive rates act as a deterrent to investment, while negative METRs suggest that the tax system stimulates investments that are socially undesirable because they earn a return lower than the opportunity cost of funds. METRs are also very useful in identifying the extent to which the tax system distorts investment allocation decisions by asset and by business sector (given the benchmark level of taxation of capital income in the tax system). Apart from the arguments for differential taxation noted above, most public finance economists would argue that competitive markets are generally efficient in allocating resources. The implication of this view is that tax differentials are generally undesirable because the associated distortions of investment allocation decisions result in reduced productivity of investment; that is, a disproportionate amount of capital is allocated to those sectors and assets in which tax treatment is relatively favorable rather than to those sectors and assets where investment would be most productive in the sense of generating output valued by 30 "Distortions" of investment decisions must be measured relative to a benchmark. In general, a tax system would not distort investment decisions if the METR were zero on all types of investment; this would occur, for example, under an ideal consumption-based tax (Zodrow and McLure, 1991). In this case, METR differentials – and the associated distortions of investment decisions – would be measured relative to a benchmark tax rate of zero. However, under an income-based tax, the benchmark level of taxation of capital income is typically the statutory income tax rate. In this case, the distortion of saving/investment decisions implied by the taxation of capital income at the statutory rate is in a sense taken as given, and the distortions attributable to tax differentials are measured relative to the statutory income tax rate. In addition, note that this discussion assumes that efficiency requires a tax system that is neutral across assets. This need not be true. For example, tax differentials may be desirable to correct for negative production externalities (e.g., pollution) or to offset other inefficiencies in the economy (e.g., inefficiencies in the taxation of labor income). These complications are ignored in the analysis, as they are best addressed with specific tax policies as needed (e.g., taxes on effluents or reform of the system of labor income taxation) rather than through the ordinary income tax system applied to capital income; for further discussion, see Gugl and Zodrow (2004). 123 consumers. In other words, the tax system should generally be characterized by "economic neutrality" with respect to investment allocation decisions, or METRs that do not vary according to the type of asset or business sector. Limitations of the METR analysis The METR analysis of capital, while being a very useful parameter for measuring the burden on investment across capital assets and sectors, has its limitations. First, it only measures the burden on capital investment. It does not measure the burden of other taxes such as payroll taxes, custom duties, etc. that do not affect the calculation of the METR on capital. Second, it is an analysis of the tax code as it exists in the statute books and does not incorporate the actual implementation of the tax code which might vary widely due to compliance reasons and the way the tax administration operates. Third, the METR analysis does not incorporate government revenue considerations. While a lower METR is good for encouraging investment, this is only a one sided view, as lower taxes also imply lost revenue for the government. The economics literature suggests that the optimal tax on capital is zero; but this holds only when economic efficiency is the sole consideration. When equity is also a consideration a zero METR need not be the optimal choice. Also, the METR analysis does not incorporate issues of incidence. It is assumed that the burden on capital is borne by capital, though in reality, part of this could be passed on to labor. Likewise taxes, other than those imposed on capital such as payroll taxes can also be passed on to capital. For the purpose of this study incidence issues are not taken into consideration. Fourth, it must be noted that, in order for the METR analysis to be used to provide broad information of the tax burden faced by different sectors of the economy, it limits the input parameters to those mentioned above. Hence highly nuanced tax benefits faced by a small number of taxpayers are generally not incorporated into the METR analysis, though it can be done in these individual cases. It must be noted that the METR is highly sensitive to non-tax parameters such as the interest rate. The METR is most useful when used as a tool to compare the burden imposed by the tax system across different capital assets or across different sectors of the economy. When comparing across countries, several simplifying assumptions are made in order to make a comparison of only the tax systems. As a result, non-tax parameters are kept constant as far as possible. For example, the interest rates might vary across countries due to various reasons, but only the tax aspects that cause this difference are incorporated into the METR. When a comparison of the METRs is made, the interest rate in the local country is taken to be inflation plus the international real interest rate, when the country operates as a small open economy. This is a valid assumption because any interest rate that is different from this can be arbitraged away in the financial market through the open economy. When the country is considered a closed economy, the interest rate in the local country less the tax imposed on it, when reduced by inflation is assumed to be equal to the international interest rate. As a result, in the closed economy model, the tax system imposes a wedge between the international interest rate and, the local interest rate less inflation. As this translates to a higher interest rate in the local country, the METRs calculated using the closed economy model take into consideration the differential interest rates across countries that are explained by the taxation of interest income. All other non-tax factors that impose a wedge between the after-tax interest rate and inflation plus international real interest rate are not a part of the calculation. As a result local credit constraints, country risk premia, etc. that might result in a high cost of doing business in a country are not included in the analysis. 124 The two input parameters in the model: the debt equity ratio and the sector-wise composition of assets are very important for determining the METRs. For the purpose of the calculation, these ratios are taken as exogenous. In reality, these parameters are also a function of the tax and economic parameters. Hence there is need to fix this at an arbitrary but reasonable value. The debt-equity ratio, for example, is a function of the extent of risk that an investor faces and his expected returns from the investment, both of which depend on the tax system. The same applies to the sector-wise composition of capital assets. While the capital asset composition depends on the nature of the business, the differential effective tax rates on asset types distorts the investment in different assets. (This only affects the total METR which is a weighted average of the separate asset-wise METRs). When the sector-wise ratio of debt to equity and the capital asset ratios for the country being studied are available, it is not unreasonable to use this ratio as an input into the model when one is analyzing the METR across sectors within an economy. But, when comparing across countries these ratios are kept fixed and the tax system is analyzed from that stand point.31 Based on actual surveys of the asset ratios it has been found that the asset ratios in general follow the composition used, though the actual numbers are not exactly the same. Manufacturing in general has high proportion of equipment while Tourism is much more building-intensive. Despite these limitations, the Marginal Effective Tax Rate analysis is one of the very few tools that allows us to quantitatively summarize the impact of the complicated tax system on investment. It allows us to isolate the tax impact from all the other factors that affect investments. As a result, it is used on a regular basis to analyze tax systems all over the world and is very useful for comparing tax systems. The approach used in this report uses the analysis developed by Hall & Jorgenson (1967), King and Fullerton (1984) and Mintz (1995).32 31 FIAS has used the asset ratios based on a survey in Canada for all its country analysis. This keeps this fixed and allows us to show the only effect the Tax factors. 32 Mintz, Jack M. (1995). “Tax Holidays and Investment,� in Anwar Shah (ed.), FiscalIncentives for Investment and Innovation. Oxford: Oxford University Press, for the World Bank. King, Mervyn and Don Fullerton (1984). The Taxation of Income From Capital. Chicago: University of Chicago Press. 125 Annex H. Data for calculating the marginal effective tax rate Domestic Corporate Income Tax rate 30% of income before dividends Capital Gains Tax Of a natural person If the net income is up to 27 UIT: 15%. For the excess over the 27 UIT: 21%. For the excess over the 54 UIT: 30%. Of a legal person 30% Tax on Interest Currently 0%. However, it has already been decided that a 6.25% tax rate will start to be effective in 2009. Corporates will be able to include interests as part of their net income paying the regular corporate tax rate. Tax on Dividend 4.1% of net income after taxes are paid Sector specific tax (Agri/Manufact/tourism/Mining) Agr. 15% of income [compared with the 30% of income for non agriculture activities] Domestic inflation rate 3.90% Tax depreciation rate equipment 20%, except when the machine is leased, in which case the machine is depreciated with an special rate Tax depreciation rate building 3%, except when leased, in which case it is depreciated with an special rate Customs Duty on machinery-equipment 0% Land transfer tax/Stamp Duty Property registration 3/1000 of the real estate property's value for "registration rights" plus a fee of 1.94% of UIT for "qualification rights" Transfer registration 1.5/1000 of the transfer value when the value does not exceed S/ 35 000 or 3/1000 of the transfer value when the value exceeds that amount plas a fee of 0.8% of UIT for "qualification rights" Capital tax Net Assets Temporal Tax Currently 0% for net assets worth of up to 1 million Nuevos Soles and a 0.5% tax rate for the excess over the 1 million Nuevos Soles threshold. However, it has already been ruled that starting 2009, the tax rate for the excess over the 1 million Nuevos Soles threshold will be lowered to 0.4%. Rebatable. Mining royalty For annual sales of up to 60 US$ millions: 1%. For annual sales between 60 y 120 US$ millions: 2%. For annual sales of more than 120 US$ millions: 3% Selective Consumption Tax For fuels: an specific tax currently averaging 2.11 Soles per galon. For the purchase of vehicles: 10% for new vehicles, 30% for used vehicles and 10 % for pick up trucks. For sodas: 17%. For alcoholic beverages: wines, whisky, rum, gin are taxed at 20%, Pisco at 1.50 soles per liter, and beer at 27.8% of the import price or the recommended retail price. For tobacco: common cigarettes are taxed at 50%, cigarettes of black and blond tobacco at 30%. For rice: 4% of the price of the first sale in national territory. VAT 19%, of which 2 percentage points correspond to the Municipal Promotion Tax, which is transferred to local governments VAT Refund (Delay in years) Legally 45 working days Tax holiday (years): H No Investment Tax credits No Small Business Tax regime / Micro As described in Table 2.5 in the main text Business Tax Regime (RUS and RER) 126