The poor in Latin America during adjustment : a case study of Peru

urnI Living Standards Measurenment Study Working Paper No 56 The Poor in Latin America during Adjustment A Case Study of Peru  The Poor in Latin America during Adjustment A Case Study of Peru  The Living Standards Measurement Study The Living Standards Measurement Study (LSMS) was established by the World Bank in 1980 to explore ways of improving the type and quality of household data collected by statistical offices in developing countries. Its goal is to foster increased use of household data as a basis for policy decisionmaking. Specifically, the LSMS is working to develop new methods to monitor progress in raising levels of living, to identify the consequences for households of past and proposed government policies, and to improve communications between survey statisticians, analysts, and policymakers. The LSMS Working Paper series was started to disseminate intermediate products from the LSMS. Publications in the series include critical surveys covering different aspects of the LSMS data collection program and reports on improved methodologies for using Living Standards Survey (LSS) data. More recent publications recommend specific survey, questionnaire, and data processing designs, and demonstrate the breadth of policy analysis that can be carried out using LSS data.  LSMS Working Paper Number 56 The Poor in Latin America during Adjustment A Case Study of Peru Paul Glewwe Dennis de Tray The World Bank Washington, D.C.  Copyright X) 1989 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing July 1989 This is a working paper published informally by the World Bank. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to Director, Publications Department, at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to photocopy portions for classroom use is not required, though notification of such use having been made will be appreciated. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list and indexes of subjects, authors, and countries and regions; it is of value principally to libraries and institutional purchasers. The latest edition is available free of charge from the Publications Sales Unit, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. Paul Glewwe is an economist in the Welfare and Human Resources Division of the World Bank's Population and Human Resources Department. Dennis de Tray is the research administrator of the World Bank's Policy, Planning, and Research Department. Library of Congress Cataloging-in-Publication Data Glewwe, Paul, 1958- The poor in Latin America during adjustment. (LSMS working paper ; no. 56) 1. Poor--Government policy--Peru. 2. Economic stabilization--Social aspects--Peru. 3. Peru--Economic policy. I. De Tray, Dennis N. II. Title. III. Series. HV223.G57 1989 362.5'0985 89-16577 ISBN 0-8213-1262-6  AB3STRACT This paper provides a general overview of the effects of structural adjustment programs on the poor and then turns to a specific country setting, Peru, as a means of illustrating the discussion. It is a companion paper to a parallel study of an African country, the C6te d'Ivoire. The depth of Peru's economic problems makes it likely that the Peruvian economy will have to undergo major adjustments in the near future. Among the options policy makers will consider are: realignment of the various exchange rates, trade liberalization, removal of price distortions, and reduction of government expenditures. The analysis presented here shows how household level data can be used to assess the effect of various policy options on the welfare of the poor. Several features of Peru's poor have important implications for programs designed to reduce the effect of adjustment on those groups. First, most poor households are located in rural areas; second, most subsidies that affect rural areas don't reach the rural poor; third, the rural poor often produce much of what they consume. These facts suggest that correcting many of the distortions affecting Peru's agricultural sector may not directly harm the poor. For Peru's urban poor, the case is more complicated, but the conclusions much the same. Regarding food subsidies and taxes, were all of these dropped the net (immediate) effect may well be favorable as the prices of wheat and wheat products, a major staple, would decline substantially. Other government programs, including those related to health and education, could be changed in ways that would benefit the poor and, in some cases, reduce overall public sector expenditures.  - vi - ACKNOWLEDGMENT We are grateful for discussions and comments on earlier drafts from Jacques van der Gaag, Welfare and Human Resources Department, The World Bank, Washington, D.C.; Ricardo Lago-Callego, Latin-America and the Caribbean, Country Operations, The World Bank, Washington, D.C.; Elaine Zuckerman, Consultant, Country Economics Department, Macroeconomic and Growth Division, The World Bank, Washington, D.C.; Javier Escobal, Grupo de Analisis para el Desarrollo (GRADE), Peru, and Carlos Paredes-Lanatta, Brookings Institution.  }: ~~~~~~~~~~~- viil - TABLE OF CONTENTS -- ~~I. Introduction ...........................l II. Structural Adjustment and the Poor ..................3 $ ~~~~~~What are the Issues? ... The Nature of Adjustment Programs ....5 Effects on the Poor.. 7 Requirements for Successful Intervention*..** ..... ...7 III. Adjustment and the Poor in Peru ..........0 Overview of the Peruvian Economy ............................10 The Poor in Peru in 1985-86. ............15 Potential Effects of Recommended Policies on the Poor .......18 1. Incomes from employment..... ... l 8 -. . .... 18 Adjustment policies and the rural poor.. ............21 Adjustment policies and the urban poor .............26 2. Effect on consumption patterns ..28 3. The provision of public services..... .............33 Identifying the Poor: Shortcut Methods .....................39 IV. Conclusion 42........... *..... * .. ........ ..4 RSeferences ..... .......... . . ...000.*....45 LIST OF TABLES Table 1: The Need for Adjustment in Peru ...........................13 Table 2: Selected Potential Sources for Reducing Peru's Fiscal Deficit .. ...... .. . . ........ . . . . .14 Table 3: Expenditure Levels of the Poor and Non-Poor: Peru 1985-86. .. ........ .... . 16 Table 4: Location of the Poor by Region: Peru 1985-86 ............. 17 Table 5: Occupation and Employer of Heads of Household: Peru 1985-86 .... ................ .. 19 Table 6: Agricultural Work and Self-Employment Among Poor Households in Peru . ........ ............. .... 20 Table 7. Rates of Effective Protection for Selected Crops: Peru 1986 .............. ...........23 Table 8: Crops Grown by the Poor and the Non-Poor: Peru 1985-86 .....25 Table 9: Occupation and Employer of Urban Heads of Household... ....27 Table 10: Effect of Consumer Subsidies and Taxes on Food Prices Peru 1986 ......... ....... ..... ....29 Table 11: Budget Shares in Household Consumption: Peru 1985-86 ..*..30 Table 12: School Attendance and Scholarships: Peru 1985-86.... .... 36 Table 13: Medical Facilities Used by the Poor: Peru 1985-86 .........36 Table 14: Budget Shares of Selected Non-Food Items ................. 40 Table 15: Inequality Decompositions by Socioeconomic Groups.... ... .41  I. INTRODUCTION Since the inception of "structural adjustment" policies governments and donor agencies alike have expressed concern about the distributional consequences of macroeconomic policies aimed at fostering economic growth. There are two distinct parts to this concern: first, the poor may not benefit from policy-induced increases in economic growth; and, second, the poor may be especially hard hit by transitional costs associated with changing macroeconomic policies, especially if the transition is longer than initially expected. This paper focuses on the second issue by concentrating on the immediate, first round effects of adjustment policies on the poor. We start by isolating those features of adjustment packages that, when first introduced, are likely to affect either the consumption or income of poor households. In doing so we draw on a previous paper (Glewwe and de Tray, 1987) that used an African country, C6te d'Ivoire, as a case study. This paper extends the analysis to a Latin American country, Peru, to gauge the extent of the problem and indicate directions for policy interventions. Although Peru does not yet have a structural adjustment program, its present situation is one in which some kind of "adjustment" must occur. As in the C6te d'Ivoire, only some of the poor in Peru are at risk of being adversely affected by adjustment policies, at least immediately upon their introduction. Yet it appears that the poor are more at risk in Peru than in C6te d'Ivoire because dismantling Peru's agricultural price support system will reduce the real income of many poor households in both urban and rural Peru. In both countries protecting the poor during adjustment will depend  - 2 - critically on the government's ability to target interventions to those poor households actually hurt by adjustment policies. In the next section we develop a taxonomy of policy issues related to the distribution effects of adjustment programs to illustrate the question we address in this paper. Following this we review the generic features of structural adjustment programs to provide a framework for discussing their possible short term effects. We then examine how these effects may reduce the living standards of the poor and describe the essential characteristics of interventions aimed at reducing the vulnerability of the poor to changes in economic conditions. Section III moves from the abstract to the concrete; recent household survey data from Peru are examined to assess the effects of a typical structural adjustment program on the poor. We also demonstrate how alternative, "shortcut" methods for identifying the poor may reduce policy effectiveness. Section IV concludes the paper.  - ~ ~ ~ ~ ~ ~ ~ 3 II. STRUCTURAL ADJUSTMENT AND THE POOR 1! What are the Issues? At the risk of oversimplifying, analyzing adjustment and poverty can be divided into three tasks. The first, and most difficult, is to determine the final effect of adjustment policies on economic growth and the distribution of that growth. This is an exceedingly important topic, but three barriers stand in our way. First, many adjustment programs are still in their early stages; their short and medium term effects, not to mention the steady state effects, are still in the making. Second, even in those countries where adjustment programs have been underway for some time, reliable information on changes in the distribution of welfare is in woefully short supply. Third, even where reliable empirical information does exist, such information is only a starting point in the process of relating changes in policy to changes in growth. The issue of causation -- which macroeconomic policies cause what percentage change in economic growth -- is likely to remain in serious contention for some time. The second task is to provide an overall assessment of who gains and, more importantly, who loses during the adjustment transition period. To give one example, under many structural adjustment programs the middle class may bear the brunt of adjustment transition costs, and it is this middle class that would adamantly oppose adjustment programs. The challenge posed by this group is to provide "safety nets" that reduce their opposition to adjustment / This section is taken from Section II in Glewwe and de Tray (1987), as are parts of later sections. Readers already familiar with that paper should turn directly to Section III.  - 4 - policies but allow needed adjustments to take place. This is the issue foremost in the minds of many government policy-makers because it lies at the heart of the political sustainability of adjustment policies. This is a more tractable area for analysis than assessing the final effects of adjustment policies since dynamic (that is, long-run) analysis is not required, but as it may focus more on the non-poor than on the poor we do not take it up here. Finally there remains the task of analyzing how the poor fare during an adjustment program's transition phase. Emphasis here is on those families who were poor before adjustment began -- those least able to cope with even temporary downturns in their economic status -- and whose welfare will be affected by the policy changes.21 A poor family whose absolute level of consumption is not affected by adjustment policies during the transition period would not be included in this group since policies aimed at them can be implemented at any time and are not necessarily a part of a structural adjustment program. Our decision to focus on poor households who are directly and immediately affected by adjustment programs draws the distinction between a country's "steady state" level of poverty and adjustment-induced changes in the welfare of the poor. We concentrate on the latter because it calls for special consideration above and beyond that given to the poor generally. 2/ One could also add to this people who are not poor before the program of structural adjustment but become poor as adjustment takes place. The most likely scenario is that of an employee who loses his or her job.  5- The Nature of Adjustment Programs Programs to redress macroeconomic distortions are often complex and multifaceted. Yet, while details differ considerably from country to country, most such programs have the same basic objective: to promote economic efficiency and set the economy on a stable and sustainable growth path by: (1) removing (artificially induced) price and other distortions, such as barriers to trade; and (2) increasing private sector investment in part by reducing public sector expenditures. Adjustment programs are usually preceded by stabilization programs, which are designed to correct balance of payments or domestic disequilibria in the short term. The immediate effects of structural adjustment programs on households, both rich and poor, can come about through one of three avenues: changes in the prices of goods and services consumed by households, changes in the employment status and labor incomes of household members, (including changes in producer prices) and changes in the provision of public services (including government transfers) to households. Changes in the prices of goods and services are probably the most common and disruptive immediate effect of structural adjustment. Adjustment programs often require reductions in, or the complete removal of, food subsidies, as well as subsidies for certain non-food commodities, such as petroleum products. Removing such subsidies will in almost all cases result in higher prices to consumers. Another common recommendation which affects domestic prices is the realignment of the exchange rate to promote exports and reduce foreign exchange deficits. Exchange rate realignment inevitably takes the form of a devaluation, which increases the prices of imported goods households consume.  -6- Structural adjustment programs also affect employment of household members, most immediately through reduced public sector employment, but later also through changes in private employment opportunities as entrepreneurs (including the self-employed) attempt to bring their production activities in line with a new sset of profit incentives. Public sector employment effects are the most immediate and predictable. They stem from government attempts to reduce central budgets and from the need to revitalize, divest or close down failing parastatals. Employees in the private sector are likely to be negatively affected in industries whose protection from import competition is removed. For households who are self-employed, particularly farm households, changes in producer prices of both inputs and outputs can have a substantial effect on the incomes they receive from their employment activities. Finally, changes in the provision of public sector services, including government transfer payments, are a common feature of structural adjustment programs. Public funds for health and education may be cut to help reduce government deficits, which is likely to reduce the quantity and/or quality of those services used by the poor. Reductions in transfer payments are also sometimes recommended to keep government expenditures in line with revenues.3- 3/ These policies may give the impression that structural adjustment programs lead to an abandonment of the poor, but it is important to keep in mind that the poor may well be worse off over the long term if macro-economic imbalances are left unchecked. Whether or not this is the case is often extremely difficult to answer and will not be taken up in this paper.  -7- Effects on the Poor The immediate effects of adjustment programs on the poor depend on what the poor were doing before the program began. The employed poor may become unemployed or may find that their real wages have declined. Price increases, especially on food items, may make it more difficult for those poor who previously consumed subsidized goods to meet minimum consumption needs, but they can also benefit poor farmers who produce for the market. The poor who rely on government services may find themselves having to cope with fewer services or services of lower quality, and reductions in government transfers may hit some poor families particularly hard. The activities and characteristics of the poor differ considerably depending on the country in question. The bulk of countries undergoing structural adjustment can be classified into one of two "prototypical" groups, the countries of sub-Saharan Africa, where the poor are overwhelmingly rural and often little affected by government decisions, and the middle income countries of Latin America, whose poor are less frequently rural and more likely to be affected by reductions in government services. The empirical section of this paper concentrates on the Latin American poor. A previous paper (Glewwe and de Tray, 1987) examined the effect of adjustment on the poor in sub-Saharan Africa. Requirements for Successful Intervention Successful - and financially feasible - interventions to reduce the impact of structural adjustment on the poor must be based on some mechanism for targeting assistance to the poor. The form targeting takes depends on the ability of governments to identify the vulnerable poor. If the poor can be  - 8 - identified on a household or individual level, transfer payments or some other form of direct assistance can reduce their vulnerability to adjustment.4/ This approach we label direct targeting. For example, government employees who lose their jobs may be provided with severance pay or opportunities to participate in retraining programs. Another important example is the provision of food or medical care to households who display clear signs of malnutrition, or to individuals who have special needs, such as pregnant and lactating women. A serious problem with direct targeting is that the "screen" needed to identify the poor is expensive to construct. If less costly "shortcut" methods are used the screen is often too porous, in which case many non-poor households may receive benefits. This would greatly increase the costs of poverty reduction programs. An example of this is in Sri Lanka, where self-reported income was used to determine household eligibility for food rations - three times the number of households estimated to be eligible for the program received rations. (Freeman, 1981). If providing assistance directly to the vulnerable poor is not feasible, intervening on the basis of the characteristics of the poor may be required. This we refer to as characteristic targeting. Examples of this are food and wage subsidies. If the vulnerable poor consume large amounts of certain food items that are rarely purchased by wealthier households, the use of subsidies to lower the price of such items could protect the vulnerable poor with little leakage to better-off households. If poor farmers produce 41 This course of action will, of course, be limited by the need to reduce government spending. The international donor community is an alternative source of funds.  -9- certain marketed crops which are rarely cultivated by more prosperous households, steps could be taken to raise the market price of those crops. Or, if the poor are highly concentrated in certain regions or districts, the provision of public services to those areas could be increased. As a practical matter characteristic targeting has two potential drawbacks. First, some non-poor households will almost always possess the same characteristic or characteristics as the poor; second, not all poor households may possess the selected characteristic and may not, therefore, benefit from the intervention. These are both forms of leakage, one of resources away from the poor, and the other of the poor away from the program. The success of characteristic targeting depends on the ability of program designers to minimize these leakages, an issue to which we turn in the next section. Finally, interventions to protect the poor must be so designed as to not seriously disrupt or delay the structural adjustment process. An obvious implication of this consideration is that interventions cannot require "large" and sustained government expenditures, since such expenditures would exacerbate the macroeconomic imbalances from which the government wants to escape. Another implication is that any subsidization of consumer goods or producer inputs and outputs must be carefully controlled to avoid inefficient allocation of a country's productive capacity. Of course, some inefficiencies may have to be accepted to achieve distributional objectives, especially during the period of adjustment but these should be temporary and must be kept to a minimum.  - 10 - III. ADJUSTMENT AND THE POOR IN PERU The previous section discussed in general terms the nature of structural adjustment programs and their likely effect on the poor. It also discussed the necessary features of policies aimed at protecting the poor in countries where structural adjustment programs are under consideration. In this section the discussion applies this framework to a specific example; we examine the vulnerability of the poor in Peru to typical structural adjustment policies and discuss possible safeguards to protect them during transition. Peru is an especially useful example because: (1) Peru has not undertaken a structural adjustment program but it suffers from serious budget and trade deficits that such programs aim to address; (2) its difficulties are similar to those faced by other Latin American countries, many of which are now implementing or designing adjustment programs; and (3) there exist recent comprehensive household survey data, the Peru Living Standards Survey (PLSS), on which to base a thorough investigation.5/ Overview of the Peruvian Economy The Peruvian economy since the 1950's has experienced both periods of sustained growth and periods of stagnation. Webb (forthcoming) provides an indepth analysis of Peru's post-war economy. To summarize, from 1950 to 1966 the economy grew at a healthy rate, an average of 3.1% growth in GDP per capita per annum. However, since 1967, and especially since the mid-1970's, economic stagnation has set in. From 1967 until 1985 GDP per capita has, on average, declined by 0.1% per capita per annum. Among the causes were 5/ See Grootaert and Arriagada (1986) for a description of these data.  - 11 - declining terms of trade for Peru's exports, low or non-existent growth in aggregate investment in part due to government intervention in the private economy, a declining agricultural sector, and inappropriate government policies (cf. Webb, 1987). In the late 1970's and early 1980's Peru's economic decline accelerated rapidly - in 1980 GDP per capita was 5% lower than in 1975, and by 1985 it was 19% lower than in 1975.61 Webb traced this decline to several factors. First, since 1980 Peru's terms of trade have fallen sharply, primarily due to declining international prices for its traditional exports (oil, copper, zinc and other minerals). Second, adverse weather conditions reduced agricultural and fishery output, especially in 1983. Third, high interest rates on international capital markets increased Peru's debt payment burden and contributed to substantial declines in private investment. Government policies during this period proved ineffective as a means of cushioning the negative effects of these external shocks, and, in fact, may have exacerbated the problem. In the summer of 1985 a newly elected government instituted a variety of controls on many aspects of the Peruvian economy in an attempt to revive economic growth. Among the steps taken since 1985 are: 1. Placing an upper limit of 10% on the amount of export earnings to be used for external debt service. 2. Enactment of legislation to raise wage rates and lower taxes, while fixing controls on prices, interest rates and imports of goods and services. 3. A moderate devaluation of the currency followed by a strict 61 These figures are derived from International Monetary Fund (1987).  - 12 - adherence to a fixed exchange rate. 4. Raising effective protection rates on a variety of locally produced foods by increasing import barriers while subsidizing agricultural inputs (cf. Escobal, 1988). These policies appear to have succeeded in the short run as real GDP grew by 8.5% in 1986, but they have not proven to be sustainable. As seen in Table 1, by 1987 the fiscal deficit had grown to 5.5% of GDP as government revenues dropped precipitously. Further, the external deficit reached 2% of GDP in 1987, in part due to a severe decline in export earnings, and foreign exchange is now in extremely tight supply as sources of external financing have dried up. Finally, public investment has declined markedly, as a percent of GDP, since 1980, which reduces the ability of future governments to provide services. Government financing of the growing deficit by monetary expansion (i.e. printing money) has lead to enormous increases in the rate of inflation; the annual rate for the first eight months of 1988 was 535% and accelerating rapidly. In order to reduce the fiscal deficit through means other than monetary creation, large cuts in government spending, probably accompanied by tax increases, will be needed. Table 2 shows the cost of certain government expenditures, again measured by percent of GDP. Various schemes designed to reduce the domestic prices of gasoline, diesel fuel, electricity, rice and fertilizer result in a total revenue loss of about 4% of GDP. Their cost is almost as large as the fiscal deficit in 1987. Such subsidies are likely targets of structural adjustment programs, and removing them could have a major impact on the living standards of the poor. Other targets for cutting the fiscal deficit include reduced government payrolls (cutting wages of  - 13 - governments workers, reducing their number, or both) and reduced government spending on social services. These costs are also given in Table 2. The government payroll amounts to about 6% of GDP while spending on health and education amount to 0.6% and 2.6% of GDP, respectively. TABLE 1: The Need for Adjustment in Peru 1980 1985 1986 1987 A. Fiscal Deficit (as % of GDP) Central Government Revenues 17.1 14.1 11.9 8.7 Central Government Expenditures Current 15.1 13.6 12.5 12.0 Capital 4.4 2.5 2.9 2.2 Overall Deficit -2.4 -2.0 -3.5 -5.5 B. Balance of Payments (as % of GDP) Current Account -0.5 0.7 -4.3 -4.3 Of Which: Exports 19.0 16.4 10.1 7.5 Imports -15.0 -9.9 -10.3 -8.8 Capital Account 2.2 3.8 2.5 2.3 Overall Balance 1.7 4.5 -1.8 -2.0 C. Public Investment 6.1 5.8 4.9 3.9 (as % of GDP) Source: World Bank, 1988 Taxes are, of course, the other main avenue for reducing the government deficit. Their contribution to deficit reduction depends on how  - 14 - much they are raised. This paper will focus more on efforts to cut expenditures since concern about the negative effects of adjustment on the poor usually focuses on efforts to reduce government expenditures (cf. Cornia, et al, 1987). For a comprehensive treatment of the potential role of tax increases to reduce Peru's fiscal deficit, see World Bank (1988). TABLE 2: Selected Potential Sources for Reducing Peru's Fiscal Deficit A. Revenue Forgone in 1987 Due to Below Market Prices Charged by Public Enterprises Item Public Enterprise Forgone Revenue (Z GDP) Gasoline PETROPERU 1.1 Diesel Fuel PETROPERU 1.6 Electricity ELECTROPERU 0.5 Rice ECASA 0.3 Fertilizer ENCI 0.4 B. Government Payroll 1986 1987 Central Government (Z of GDP) 3.9 3.9 Non-Financial Public Enterprises (% of CDP) 2.3 2.1 C. Government Spending on Social Services 1984 1985 Ministry of Health (Z of GDP) -- 0.6 Ministry of Education (% of GDP) 2.6 -- Source: World Bank, 1988, and Suarez-Berenguela, 1987.  - 15 - Reduction of the external deficit is more complicated, but usually involves a major devaluation, which generally raises the prices of tradable goods. The likely effects of devaluation on the poor are discussed at various points below, but for a more technical discussion of the options available for reducing the external deficit see World Bank (1988). Negotiations are currently underway between Peru and its creditors on ways to fulfill past obligations and increase the flow of future funding. It is very likely that some kind of structural adjustment program, with or without the participation of international financial institutions, will have to be adopted. The remainder of this paper seeks to address the issue of how Peru's poor would be affected by typical structural adjustment measures and how the poor could be protected during the period of adjustment. The Poor in Peru in 1985-86 Identification of the poor using household survey data requires that households be ranked according to some criterion. In this paper the welfare levels of households are measured by adjusted per capita consumption levels, with children receiving smaller weights than adults when household size is determined.7- The PLSS consumption data capture a wide range of implicit expenditures, such as imputed rents, consumption of own products on the part 7/ The weights for children in the adjusted per capita consumption measure are 0.2, 0.3 and 0.5 for children aged 0-6, 7-12 and 13-17 years, respectively.  - 16 - of farm households, and the use value of durables, making this measure of welfare valid even for comparisons between urban and rural households.8/ The impact of adjustment programs on the poor will depend in the first instance on how badly off these households are before any adjustment program begins. We must, therefore, begin our analysis with an assessment of just how poor are low income households in Peru. Basic data on per capita consumption, both adjusted (smaller weights for children) and unadjusted, and per capita income are given in Table 3. This and subsequent tables use two poverty lines to define the poor, based on the poorest 10% and the poorest 30% of the population, respectively. In adjusted per capita terms, consumption for the poorest 10% of Peruvians is about 20% of consumption for the average Peruvian, while the poorest 30% consume about one third of the national average. Should consumption levels decline further because of an adjustment program, concern over the plight of Peru's poor would seem well justified. TABLE 3: Expenditure Levels of the Poor and Non-Poor: Peru 1985-86 (Intis per month at June 1985 prices) Per Capita Expenditures Adjusted Unadjusted Poorest 10% 115.3 76.9 Poorest 30% 189.0 126.1 Wealthiest 70% 740.5 514.4 All Peruvians 574.9 397.8 8/ See Glewwe (1987) for details on the calculation of household expenditures and a detailed argument for using adjusted per capita consumption expenditures to measure household welfare.  - 17 - The incidence of poverty in Peru varies considerably across geographic areas. To isolate this variation, we divide Peru into its three principal geographic areas, the coastal strip along the Pacific Ocean (Costa) the mountainous region that runs from North to South (Sierra), and the Amazon basin which lies to the east of the mountains (Selva). We further divide each of those areas into its urban and rural components. Table 4 demonstrates a crucial point regarding the potential effects of adjustment programs on the poor: even in Peru most of the poor are found in rural areas. Although less than half (44%) of all Peruvians live in rural areas, 70% of the poorest 30% and 83% of the poorest 10% of the population are found there.9/ This means that, as was the case in the C6te d'Ivoire, the effect of structural adjustment on Peru's poor is for the most part a matter of its effect on rural areas. This suggests that policies to safeguard the poor should be feasible in rural areas. TABLE 4: Location of the Poor by Region: Peru 1985-86 Poorest 10% Poorest 30% Wealthiest 70% All Lima 2.4 7.8 34.9 26.8 Costa Urban 5.0 10.8 17.1 15.2 Sierra Urban 7.5 9.1 11.8 11.0 Selva Urban 1.8 2.2 3.3 3.0 Costa Rural 9.4 9.8 6.1 7.2 Sierra Rural 62.7 50.2 22.2 30.6 Selva Rural 11.2 10.1 4.7 6.3 9/ These figures, as well as those in Table 3, differ slightly from those given in Glewwe (1987). This is due to a small error concerning expenditures that was detected and corrected after that paper was written.  - 18 - Yet it is possible, perhaps even likely, that those poor who do live in urban areas are more negatively effected by structural adjustment programs than the rural poor. This will be discussed below. Potential Effects of Recommended Policies on the Poor As discussed above, government policies can affect households in three ways: (1) by changing incomes; (2) by changing market prices of consumption goods; and (3) by changing the level or quality of government services. In this section a typical set of structural adjustment policies is examined in the Peruvian context to assess its likely effects on the poor. We start with policies that affect employment incomes. 1. Incomes from employment To understand how the poor will be affected by structural adjustment programs, one must examine their sources of income. Two important avenues by which structural adjustment programs immediately affect incomes of poor households are changes in government employment and changes in prices of agricultural outputs and inputs. In order to reduce government deficits, it is often recommended that some workers be dropped from government payrolls in an effort to reduce government spending. In the case of Peru a 10% decrease in the government payroll could reduce the deficit by 0.6% of GDP (cf. Table 2). Table 5 examines heads of household categorized by occupation and type of employer. The table supports two important conclusions: first, most poor Peruvians live in households in which the head is self-employed. In contrast, the number of household heads who work for the government is very small. Even if all heads of poor households who work for the government or  - 19 - for parastatal businesses lose their jobs only about 4% of the poorest 30% of the population and about 1% of the poorest 10% would be affected.l