Modeling economic behavior in Peru's informal urban retail sector

Policy, Research, and External Affairs WORKING PAPERS Women In Development Populatio,l and Human Resources Department The World Bank August 1990 WPS 469 Modeling Economic Behavior in Peru's Informal Urban Retail Sector J. Barry Smith and Morton Stelcner Small family businesses that operate outside the formal system comprise a large part of the economy in developing countries and more than half the Peruvian street vendors are women. This model of informal activity in Peru's urban areas elicits policy recommendations to improve productivity (especially women's) in the informal sector. Thw Policy,Research. and Extemal Affairs Complex distributes PRE Working Papers todisseminatethe findings of work in progress and to encourage the exchange of ideas among Bank staff and all others interested in development issues. These papers canry the names of the authors, reflect only their views, and should he used and cited accordingly. The findings, interpretations, and conclusions are the authors' own. They should not be attributed to the World Bank, its Board of Directors. its management, or any of its member countries.  LPolicy, Research, and External Affairs Women In Development | WPS 469 This paper - a product of the Women in Development Division, Population and lHuman Resources Department is part of a larger effort in PRE to determine if and how women's productivity (and thus family welfare) are improved when women are given more access to education, extension, training, credit, health care, and otherpublic resources. Copies are available free from the Wor!d Bank, 1818 H Street NW, Washington DC 20433. Please contact Maria Abundo, room S9-123, extension 36820 (87 pages with diagrams and tables). The informal sector is a collection of loosely Smith and Stelcner analyze Peru's urban organized, small-scale competitive family informal sector- particulary women's role in it businesses that rely little on nonfamily hired - based on a theoretical model of informal retail labor, use labor-intensive technologies, and trade (the dominant nonfarm family enterprises), operate largely outside of the legal, bureaucratic, using data from the Peru Living Standards and regulatory framework in terms of licenses, Survey (PLSS). taxes, and contractual obligations. They address these questions: What factors In Lima, Peru, the informal sector makes up explain differences in the performance of retail half the labor force, accounts for 61 percent of businesses? If these can be identified, what types the hours worked, and generates an astounding of policy initiatives might improve the perfor- 39 percent of GDP. More than half the street mance of firms, especially those run by women? vendors are women. Among their recommendations: In the informal sector, the free play of * Channeling crcdit to small businesses. market forces determiines returns to productive factors, especially labor. Informal enterprises are * Promoting cooperatives and self-help concentrated in low-income areas of urban associations, which provide credit, facilitate bulk centers, but rural households in Kenya and Peru, purchases, and establish markets for entrepre- among other countries, have joined. neurs. The informal sector is an important - if not * Providing technical assistance, such as the sole -- income opportunity for growing short-term instruction in basic management. numbers of the poor. International aid agencies have explored policies to make informal busi- * Making it easier and cheaper to get business nesses more profitable. But this surge of interest licenses. is not based on much empirical evidence about what determines the firms' performance. Nor is * Provide or facilitate cooperative childcare the value of women's entrepreneurial activities centers, facilities for preparing food, and neigh- reflected in the national accounts. borhood facilities for basic health care to reduce the heavy workload typical for women. The PRE Working Paper Series disseminates the findings of work undcr way in the Bank's Policy, Rcsearch. and Extemal Affairs Complex. Anobjective of thcseries is to get these findings outquickly, even if prescntations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the PRE Dissemination Center  Table of Contents page 1. Introduction ......................................... 1 2. Some Stylized Facts on the Informal Sector .... ....... 4 2.1 Magnitudes and other Characteristics .... ........ 4 2.2 The Role of Women ............................... 8 3. Model Formulation and Concepts ........................ 12 3.1 Assumptions ..................................... 13 3.2 Specific Aspects of the Model ................... 15 3.3 The Revenue Function of the Firm .... ............ 18 3.4 Marginal Revenue Products and Profit Maximization .................................... 19 4. Description of the Data and Variables .... ............. 22 5. The Empirical Model ................................... 28 5.1 Initial Specification ........................... 29 5.2 Preliminary Tests for Aggregation and Pooling .... 29 5.3 Diagnostic Analysis ............................. 31 5.4 Analysis of Nonlinearity ........................ 34 5.5 Distribution of the Errors ...................... 37 6. Empirical Findings and Interpretation .... ............. 38 6.1 Factor Productivities ........................... 42 6.2 Nonparametric Evidence .......................... 59 6.3 Simulation Experiments .......................... 67 7. Policy Implications ................................... 70 7.1 Enterprise-Specific Policies .................... 71 7.2 Women-Specific Policies .72 Appendix A: Nonparametric Analysis .74 Nonparametric Modeling .74 Defining Conditional Means and Their Properties 75 Estimating Conditional Means and Derivatives .76 Appendix B: The Rural Model .78 Bibliography .84 J. Barry Smith is a member of the Department of Economics, York University, Toronto, Ontario and Morton Stelcner is in the Department of Economics, Concordia University, Montreal, Quebec. The authors thank the foLlowing individuals: Jorge Castillo, Shahid Khandker, Valerie Kozel, and Marcia Schafgans of the World Bank; Richard Laferriere, Department of Economics, Concordia University; Georges Monette, Department of Mathematics, York University; Jeffrey Racine, Department of Economics, York University; Gwenn B. Hughes, Policy and Strategic Analysis Branch, Labour Canada, Ottawa, and Gail A. Spence, Canadian Pacific Forest Products Limited, Montreal. The authors also benefited from the opportunity to present an earlier version of the paper at a seminar sponsored by the Women in Development Division. Jeffrey Racine performed the nonparametric analysis and wrote Appendix A. The research was funded by the Women in Development Division, the World Bank.  Modeling Economic Behavior in the Informal Urban Retail Sector of Peru J. Barry Smith and Morton Stelcner 1. Introduction Few topics in the literature on development economics have inspired as much interest, controversy, and rhetoric as the informal sector in developing countries.1 Here the term is intended as a shorthand expression to describe the collection oJf loosely organized (but not necessarily fly-by-night) small-scale competitive family businesses. Such businesses rely little on nonfamily hired labor. Their technologies are labor intensive and they operate largely outside the legal, bureaucratic, and regulatory framework reg.rding such matters as licenses, taxes, and contractual obligations. An important characteristic of the informal sector is that the free play of market forces generally determines returns to productive factors, especially labor. The enterprises are usually concentrated in low-income areas of large metropolitan centers, but it is not uncommon to find rural households, for example in Kenya and Peru, that have joined the informal sector.2 There appears to be a consensus that the informal economy is a sizable and growing component of developing economies. It accounts for a substantial fraction 1 For more recent literature see Bromley (1978), Cornia (1987), Hallak and Caillods (1982), Hart (1987), House (1984), IDB (1987a), Mattera (1985), Moser and Marsie-Hazen (1984), Sethuraman (1981), Stewart (1987), and Tokman (1978). 2 For a discussion see Stelcner and Moock (1988), Moock, Musgrove, and Stelcner (1989), and Freeman and Norcliffe (1985). 1  of the labor force, especially in urban areas, and provides an important -- if not the sole - - income opportunity for growing numbers of the poor. But the debate about its role in economic development continues. There is considerable disagreement aoout whether measures should be taken to promote informal activities as an impetus to economic growth and a strategy for improving the earnings of low-income households. The place of the informal sector in development is all the more imnortant because of the severe economic crisis in most third world economies. As Cornia (1987) discusses, households faced with sharply reduced employment and incom3 prospects in the formal (or modern) sectors -- manufacturing, services, mining, and government - - tend to seek employment and income opportunities in the informal economy. The first to shift are those who have lost jobs in the formal sector. Next, employed formal sector workers, especially government employees, resort to moonlighting activities, most of which are informal. As household incomes decline, married women and children who previously did not work in the market are drawn into informal market activities, and soon new entrants to the labor force begin to find jobs in the informal rather than the formal sector. 3 The rapid growth of the informal sector has led international aid agencies and governments to explore policies to improve the profitability of such businesses. This surge of interest, however, is not based on much empirical evidence about the underlying determinants of the performance of the firms. Informals perform a remarkable array of activities, ranging from vending foodstuffs and prepared foods to consumer goods and services, including 3 Several recent studies have identified these general patterns. See Cornia (1987), IDB (1987b), PREALC (1985), van der Gaag, Stelcner and Vijverberg (1989), and Tokman (1986). 2  carpentry, tailoring, barbering, shoe-repair, domestic work, vehicle and tool repairs, and transport. In addition, small-scale entrepreneurs manufa re textiles, garments, footwear, household utensils, musical instruments, metal products, furniture and wood products, and leather goods. They process foods and beverages, and recycle junk. Some firms are also involved in such illicit ventures as smuggling and processing alcohol and cocaine. This research analyzes the informal sector in Peru, particularly the role of women, based on a theoretical model of informal retail trade that uses data from the Peru Living Standards Survey (PLSS). Retailing is the domirnant nonfarm family enterprise. The central questions are: What factors explain differences in the performance of retail businesses? Assuming that these cc.nsiderations can be identified, what types of policy initiatives might improve the performance of firms, particularly those run by women? The analysis is confined to urban areas where most of these bus.nesses are located .4 4We attempted to provide an analysis of nonfarm enterprises in both urban and rural areas. But the limited size of the sample in rural areas precluded the estimation of a mod-l that we were confident of using. The resu- ts for rural areas are in appendix L. 3  2. Some Stylized Facts on the Informal Sector 2.1 Magnitudes and Other Characteristics Informal economic activity is a mainstay of the Peruvian economy. Its sustained growth stands in marked contrast to formal activity, which has deteriorated in the last decade at an alarmingly rapid rate. There is extensive evidence that a high proportion of the labor force, especially the Lemale component, is in informal activities. As Glewwe and de Tray (1989) and Suarez- Berengue]a (1987) discuss, the majority of the bottom socioeconomic strata in urban areas earns a livelihood from self-employment in the informal sector. Peru's shadow economy has recently attracted worldwide attention5 as a result of the recent publication of El Otro Sendero: La Revolucibn Informal6 by Hernando de Soto, a businessman and president of Lae Instituto Libertad y Democracia (ILD) in Lima. He concludes that the informal sector is the dominant and most dynamic part of the economy, and believes that removing the burdensome obstacles to legitimacy (such as bureaucratic red tape) would considerably improve Peru's economic malaise. Several other recent studies corroborate de Soto's view ,hat a sizable portion, perhaps a majority, of the labor force is in the informal sector.7 Surveys by the ILD in 1985 and 1986 show that the informal sector in Lima makes up almost half the labor force, accounts for 61 percent of the hours worked and s See The Economist, February 18 and September 23, 1989; and Leaders, March 1989 (12) 1. 6 The English translation is The Other Path: The Invisible Revolution in the Third World. (See de Soto 1989). 7See Althaus and Morelli (1980), Kafka (1984), Litan and others (1987), Stelcner and Moock (1988), Moock, Musgrove, and Stelcner (1989), Strassmann (1987), Suarez-Berenguela (1987), Vargas Llosa (1987), Webb (1977), and World Bank (1987). 4  generates an astounding 39 percent of GDP (1984). For such sectors as commerce mind personal services this share exceeds 60 percent. Litan and others (1987) report that the official national accounts estimate of the informal sector's contribution to GDP it, 3'84 resulted in an understatement of total GDP of 23 percent. Perhaps even more striking is the estimate that 439,000 Lima residents depend on the underground commercial economy, and almost three fourths (314,000) of these individuals depend on street sales. According to de Soto, these activities generated about $25 million a month in gross sales in 1985 and an average ret per capita profit of $58 a month, about 40 percent more than the legal minimum wage. The 314,000 street merchants include 91,455 street vendors, are 42 percent of the Lima work force involved in commerce. The ILD surveys show that women make up 54 percent of the street vendors. Eighty-six percent of the street merchants occupy curbside sites, while the remaining 14 percent rove the streets. Business is also conducted in (illegal) cooperative markets -- collections of kiosks, stalls and booths. Of the 331 markets in Lima, 274 were put up illegally; only 57 were built by the government. It is estimated that $41 million has been invested in these illegal markets, which employ about 125,000 people (including sc - 40,000 vendors). More than 80 percent of the street vendors and 64 percent of the informal markets are found in low-income districts. According to recent studies that used the Peru Living Standards Survey8, half of the 5,100 households surveyed owned at least one nonagricultural family enterprise. Of the 27,000 individuals in the survey, 13,600 were in the labor force and 97 percent were employed. More than 4,500 worked in nonfarm family 8 See Stelcner and Moock (1988), and Moock, Musgrove, and Stelcner (1989). 5  businesses, and 3,100 worked in family enterprises as their main occupation. About 6,200 worked on family farms. Metropolitan Lima accounted for 34 percent of n3nfarm family businesses, other urban areas for 44 percent, and rural areas only 22 percenc.. Of course, a large fraction of rural households also operate farms. These proportions correspond closely to the distribution of family workers and households across regions. The average number of enterprises per household is 1.25. These family businesses are dominated by retail trade, manufacturing (especially textiles), and personal services (mainly in urban areas). Retail trade encompasses small shops, inns and cafes, kiosks, stalls, and street vending. Nontextile manufacturing includes food, beverages, pottery, furniture, toys,novelties, and musical instruments. The textile sector includes spinning, weaving, and tailoring. Personal services range from laundries and hairdresser. and barbers to entertainment, auto and electrical repairs, and cleaning services. Most businesses rely on just one or two family workers; the use of hired labor is negligible. Many firms do not own any capital or inventory and often have no operating expenses. In the retail sector selling is often on consignment or commission. Large factories, wholesalers, and stores in the formal sector often provide the goods and perhaps the cart, stall, or kiosk. The goods are sold either on straight commission, or on consignment: the sellers pay only for what they sell and return the unsold goods. Factories often subcontract textile, clothing, leather and footwear manufacturing to family enterprises, which they provide with materials and equipment. In the labor-intensive personal services sector, little use is made of capital equipment. Thus it is not surprising that many family 6  businesses in the dominant sectors reported little or no capital, inventories, or operating expenses. How much credit do informals use and obtain? The Peru Living Standards Survey provided information on the current debt position of each household, and on the source and termns of loans obtained in the past year. Only 10 percent of the households that operated businesses reported that they received loans or were in debt. This is not surprising for several reasons. First, the PLSS was conducted when inflation rates were extremely high (June 1985-July 1986). (During the first half of 1985 the annual inflation was 200 percent, and in the first half of 1986 monthly inflation was 4 to 5 percent. Such high rates of inflation are unlikely to foster a willingness to lend, except at interest rates so high that few households would choose to borrow. Second, given a sensitivity to questions about indebtedness, the Peru Living Standards Survey probably underestimated the debt among respondents. (To preserve good will, questions on debt and credit were last in a long questionnaire". Third, given the uncertain legal position of family businesses, their ability to obtain credit from formal lending institutions is very restricted at best. As Carbonetto (1984), Mescher (1985), and Kafka (1984) document, only a minute fraction of informal sector firms in Lima borrow from financial intermediaries in the formal sector. Households that need to borrow must depend on loan sharks and pawnbrokers, or resort to a "pandeiro". This is a revolving fund to which members make a weekly contribution. A lottery determines the winner of the week's contributions .9 9 See Mescher(1985), World Bank (1989). 7  In Lima, retail trade, personal services, and manufacturing account for more thati 80 percent of the businesses and almost 85 percenc of the family workers (Table 1). The single most important activity is retail trade. The pattern is the same in the rest of Peru. In other urban areas the proportions are 86 percent (retail trade, 52 percent) and 90 percent. In rural areas 80 percent of entrepreneurial families and workers are in retail trade (43 percent) or manufacturing, particularly textiles. These are typical informal sector endeavors in most developing countries. What distinguishes Peru from other countries is the unusually large proportion of women and households in these activities. 2.2 The Role of Women The Peru Living Standards Survey shows that women dominate the informal economy. Schafgans (1989) notes t'at women make up about 45 percent of the labor force. The vast majority work in family-owned firms and farms. Table 2 shows that 82 percent of the 5,952 employed women worked in family-owned nonfarm businesses (28 percent) or on family farms (54 percent). The remaining 18 percent were wage earners, mostly in urban areas. In Lima 45 percent of employed women were wage earners in contrast to 19 percent in other urban areas. In rural areas only 5 percent of employed women worked as salaried employees, usually as school- teachers and clerks in the public sector. There is practically no labor market for women in rural areas. In Lima, 38 percent of the women were employed in nonfarm businesses and 17 percent in agricultural activities; in other urban areas the corresponding proportions were 44 and 37 percent, and in rural areas 11 and 84 percent. 8  TabTh. Percendfbgttribuf5 forbEhitysineAqtctofictivity Metrbima OthdirbaAreas AlUrbabreas RuraUreas Peru No.oWDbservationlO45 1387 2432 692 3124 Retdilade 40.7 51.5 46.8 43.1 46.0 Persofekvices 19.8 12.4 15.6 4.5 13.1 TextM1mnufacturiI*.9 12.2 11.6 26.4 14.9 Othd4anufacturidrW.4 9.7 10.0 11.9 10.4 Subtotal 81.8 85.8 84.0 85.9 84.4 Othdefctors 18.2 14.2 16.0 14.1 15.6 Note0thdfanufactubtraidoedbvera6irxii rcorduebis cel1laiieems . OthLactdcswlud6nstruc inanporbdtmdi i:adhnintishfrnhunting. SourcberIiiviSgandaSldviey. TABLE 2. Percentfbtribuflmpfoy&dbo]Pordvfrypefmployment Metrbima OthdfrbaAreas AlUrbaAreas Rurareas Peru Females Males Females Males Females Males Females Males Females Males No.o9Dbservatl8fll 1747 1978 2450 3215 4197 2737 3041 5952 7238 NonFarSector 38.2 31.5 44.1 32.5 41.9 32.1 10.6 5.3 27.5 20.8 Farigector 16.5 3.4 37.0 25.1 29.0 16.1 84.0 77.0 54.4 41.7 Wageector 45.3 65.1 18.9 42.4 29.1 51.8 5.3 17.7 18.1 37.5 Notefhemonfaumfarwectae