Aspects of poverty : a review and assessment - Philippines (two volumes) (Vol. 2 of 2) : Main report with annexes and statistical appendix

Report No. 2984-PH Aspects of Poverty in The Phiilippines: A Review and Assessment FILE COPY (In Two Volumes) Volume II: Main Report with Annexes and Statistical Appendix December 1, 1980 Country Programs Department East Asia and Pacific Regional Office FOR OFFICIAL USE ONLY Document of the Wbrld Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otheiwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency unit = Pesos (P) US$1.00 = P 7.5 P 1.00 = US$0.135 FISCAL YEAR January 1 to December 31 ABBREVIATIONS AND ACRONYMS AMC - Agricultural Marketing Cooperative BCS - Bureau of Census and Statistics BHS - Barangay Health Station BLISS - Bagong Lipunan Sites and Services BM - Budget Ministry BPW - Bureau of Public Works CLT - Certificate of Land Transfer CORUM - Cotabato Rural Uplift Movement CPI - Consumer Price Index FIES - Family Income and Expenditure Survey FNNS - First Nationwide Nutrition Survey FNRI - Food and Nutrition Research Institute HYVs - High Yielding Varieties LBP - Land Bank of the Philippines LWUA - Local Water Utilities Administration MAR - Ministry of Agrarian Reform MASICAP - Medium and Small Industries Coordinated Action Program MEC - Ministry of Education and Culture MHS - Ministry of Human Settlements MMA - Metropolitan Manila Area MMINUTE - Metro Manila Infrastructure Utilities and Engineering Program MOA - Ministry of Agriculture MWSS - Metropolitan Waterworks and Sewerage System MPW - Ministry of Public Works MSSD - Ministry of Social Services and Development NA - National Accounts NACIDA - National Cottage Industries Development Authority NCP - Nutrition Center of the Philippines NCSO - National Census and Statistics Office NDS - National Demographic Survey NEDA - National Economic and Development Authority NFAC - National Food and Agricultural Council NHA - National Housing Authority NIA - National Irrigation Administration NMYC - National Manpower and Youth Council NNC - National Nutrition Council NWRC - National Water Resources Council  FOR OFFICIAL USE ONLY - ii - ONFE - Office of Nonformal Education OPT - Operation Timbang PCA - Philippine Coconut Authority PREPF - Population, Resources, Environment, and the Philippine Future PROGRESS - Program for Removing Sewerage from Streets RDA - Recommended Daily Allowance RHU - Rural Health Unit RWDC - Rural Waterworks Devqlopment Corporation SEAP - Self-Employment Assistance Program UPLB - University of the Philippines at Los Banos WD - Water District WHO - World Health Organization WSA - Water Supply Association ZIP - Zonal Improvement Program Thi dou ethsarsrce itiuinadmy eue yrcpet nyi h efrac  - iii - PREFACE 1. This report presents the findings of a mission which visited the Philippines in July/August 1979 to\ analyze the extent of poverty in the country and review relevant policies. A draft report was discussed with the Government in July 1980 in Manila. The Purpose of the Study 2. Alleviation of poverty in the Philippines has been a concern of the Philippine Government and the World Bank for some time. The Government has already instituted a number of programs and projects aimed at increasing the living standards of the poor and has commissioned special studies dealing with poverty and income distribution. The Five-Year Development Plan (1978-1982) of the Philippines includes poverty reduction as one of its major targets. The Bank has also become increasingly interested in the poverty issue in recent years and reports analyzing poverty have been completed for several countries, including those in the East Asia Region. Two previous reports on the Philippines by the ILO /1 and the World Bank /2 covered several social and distributional issues at great length but did not focus specifically on poverty alleviation. The Consultative Group for the Philippines, in its 1978 meeting, suggested that the World Bank should prepare a report on poverty and related issues for the information of the Group. This study has been prepared in pursuance of that decision and is meant to help the Government as well as other members of the Group in understanding the nature and extent of poverty in the Philippines so that appropriate measures can be devised to alleviate it. The Scope and Format of the Report 3. The scope of this report is fairly wide. It attempts to measure the extent of poverty and to explore its sectoral dimensions. Hlowever, systematic data at a sufficiently disaggregated level for this purpose were rarely available and, where available, were not easily reconcilable or susceptible to clear interpretation. As a result, some of the findings remain tentative, and this report cannot be more than a first step towards an increased understanding of poverty in the Philippines and the development of an integrated set of poverty alleviation policies. Further analytical work would be required in several areas, such as rainfed agriculture, urban development, and employment. But, the data base remains weak, and more definitive analysis can probably be undertaken only after the completion of the next Family Income and Expenditure Survey and Agriculture Census. /1 ILO, Sharing in Development, 1974. /2 World Bank, The Philippines: Priorities and Prospects for Development, 1976.  - iv - 4. This report consists of two volumes. Volume I presents an overview of the poverty situation and its policy implications. Volume II is the Main Report consisting of five chapters. The first chapter presents a profile of poverty by geographical area, occupation, and sector; it also traces time trends and discusses some measures:of inequality. Chapter 2 extends the discussion to other social indicators and basic needs. The third chapter deals with the employment aspects of poverty, including the extent of and reasons for the decline in real wages. Issues relating to rural poverty and agricultural development are examined in Chapter 4. The last chapter discusses the pattern of urban poverty and employment. Volume II also contains technical annexes and a statistical appendix. Composition of the Mission 5. The mission consisted of!Sarshar A. Khan (Chief), Dale M. Hill, Winston King, Dirk M. Leeuwrik, Deepak Lal, all Bank staff; Trung Ngo of FAO/CP; Jong-Goo Park of ADB; and Sudhir Anand, Simon Fass, and Allen Williams, consultants. Bruce Jones, Bank staff, assisted the mission in drafting the chapter on employment. The report also benefitted from the ongoing work on the Philippine Food, and Nutrition Plan with assistance from Harry Walters of the Bank staff.  ASPECTS OF POVERTY IN THE PHILIPPINES: A REVIEW AND ASSESSMENT (In Two Volumes) VOLUME II: THE MAIN REPORT (WITH ANNEXES AND STATISTICAL APPENDIX) Table of Contents Page No. GLOSSARY OF ABBREVIATIONS ........................... i PREFACE ..iii CHAPTER 1: THE EXTENT, PATTERN AND TRENDS OF POVERTY ................ 1 A. Income and Expenditure Data in the Philippines 1 - Limitations of Data ................................ 1 B. Measurement of Poverty .. 3 - The Concept of Poverty and Measurement Problems 3 - The Poverty Line. 4 C. The Poverty Profile ................................... 8 - Poverty Incidence by Location .................... 8 - Poverty Incidence by Size of Family ................ 10 - Poverty Incidence by Occupation . . 11 - Poverty Incidence by Sector/Subsector . . 14 D. Trends in Poverty, 1957-75 ......................... 14 E. Consumption Inequality .. 18 CHAPTER 2: BASIC NEEDS ............................................ 19 A. Food and Nutrition ....................... . 19 - Degree and Pattern of Malnutrition ............... 19 - Nutrition and Food Supply ...................... I ... 22 B. Health ....................... 24 - Health Indicators .. 25 - Health Services ................ 27 - Expenditure on Health Services ............. 29 C. Education .................... 30 - Literacy Rates .................... 32 - Educational Facilities and Quality of Education 32 - Educational Expenditures ............. ............... 34 - Program and Policy Implications . .35 D. Housing ........................ 36 E. Water Supply and Sanitation .......................... 39 - Institutional Aspects of Water Supply and Sanitation 43 F. Conclusions and Recommendations . .44 CHAPTER 3: EMPLOYMENT, WAGES, AND POVERTY . . . 46 A. Trends in Employment and Real Wages . .46 B. The Supply of Labor ..50 C. The Demand for Labor, Sectoral Employment, and Industrial Promotion Policies. 51 D. Devaluation, Inflation and the Terms of Trade . 55 E. Conclusions and Policy Implications . .56  - vi - Table of Contents (Continued) Page No. CHAPTER 4: ISSUES IN RURAL POVERTY ................................. 58 A. Profile of Rural Poverty ............................ 58 B. Correlates of Rural Poverty ......................... 59 - Agricultural Productivity and Poverty Level ....... 59 - Farm Size, Farm Tenure, and Access to Land ........ 62 - Intensity of Land Use and Crop Mix .... ............ 63 - Quality of Land ................................... 64 - Secondary Sources of Family Incomes .... ........... 64 C. Profile of Major Poverty Groups ..................... 67 - Rice and Corn Farmers ............................. 67 - Rice and Corn Laborers ............................ 69 - Coconut Farmers and Laborers ...................... 70 - Small Sugar Farmers and Sugar Plantation Laborers . 71 - Coastal Fishermen ................................. 72 - Farmers in Marginal Rainfed and Upland Areas ...... 73 D. Trends in Agricultural Growth and Poverty Incidence in the 1970s .............................. 75 - Population-Land Trends ............................ 76 - Factors Affecting Regional Poverty Incidence: 1970-75 ........................................... 76 - Distribution of Agricultural Growth by Subsector: 1970-75 ........................................... 78 - Agricultural Growth, Post-1975 Period .... ......... 82 E. Conclusions and Policy Implications .... ............. 83 - Reorientation of Support for Lowland Rice and Corn Farmers ...................................... 84 - Agricultural Development in Rainfed Areas ......... 85 - Increasing Underst'anding of Land Settlement ....... 86 - Agrarian Reform and Redistribution of Income ...... 86 - Increased Poverty Orientation in Other Agricultural Programs ............................. 86 - Area Development Programs ......................... 87 CHAPTER 5: ISSUES IN URBAN POVERTY ................ .. ............... 88 A. Urbanization Trends ................................. 88 B. The Profile of Urban Poverty ........................ 89 - The Pattern of Urban Poverty and Recent Changes ... 89 - Correlates of Urban Poverty ....................... 90 - Dynamics of Urban Poverty ......................... 91 C. Urban Employment .................................... 93 D. Government Programs for Assistance to the Poor ...... 95 E. Conclusions ......................................... 97 ANNEX 1-A: REVIEW OF THE INCOME AND, EXPENDITURE DATA .... ........... 99 - The Family Income and Expenditure Surveys ......... 100 - Justification for Using Expenditure Rather than Income Data .................................. 103  - vii - Table of Contents (Continued) Page No. - Definitions of Urban Areas Used in FIESs ........... 105 - 1971 and 1975 FIES .............................. ... 105 ANNEX 1-B: THE POVERTY LINE ..................... .................... 106 - Rural-Urban Price Differences ............ .......... 112 ANNEX 1-C: COMPARISON OF FIES AND NATIONAL ACCOUNTS BASED ON ESTIMATES OF PERSONAL INCOME AND EXPENDITURE . ............ 114 - References ............................................ 118 ANNEX 4-A: AGRICULTURAL PROGRAMS AFFECTING PRODUCTION AND RURAL INCOMES IN THE 1970s ...... .... . ........................... 121 - Foodgrain Self-Sufficiency Programs ................ 121 - Irrigation Development . .............................. 124 - Agrarian Reform Programs ........................... 128 - Cooperative Programs ................................. 132 - Land Settlement Programs .......... ................. 133 - Recent.Programs Benefitting Diversified Rainfed Agriculture ......................................... 135 - Programs Being Developed in the Coconut and Fisheries Sectors . .. ............ 137 ANNEX 5-A: TRENDS IN URBANIZATION IN THE PHILIPPINES .... ............ 139 - National Urbanization .............................. 139 -Regional Urbanization .............................. 142 -Growth of Urban Centers ............................ 145 STATISTICAL APPENDTX ................... 151 MAP OF THE PHILIPPINES - Classification of Provinces by Geographical Regions (As of 1975)  - viii - Tables in Text Table No. Page No. Chapter 1 1.1 Sensitivity of Poverty Incidence to the Poverty Line, 1971 .... 6 1.2 Poverty Incidence by Region and Rural/Urban Area, 1975 ........ 9 1.3 Poverty Incidence by Rural/Urban Area and Size of Family, 1975 12 1.4 Poverty Incidence by Occupation and Region, 1971 ............... 13 1.5 Poverty Incidence by Industrial Sector and Region, 1971 ....... 15 Chapter 2 2.1 Nutritional Intake and Sufficiency Level in the Philippines, 1978 ........................................................ 21 2.2 Distribution of-Malnourishment among Children .................. 21 2.3 -Infant Mortality and Percentage of Pre-Schoolers Affected by Second and Third Degree Malnutrition ........................ 24 2.4 Some Socioeconomic Indicators by Region, 1973 ................. 26 2.5 Distribution of Rural Health Units and Barangay Health Stations, 1974 ............. ; ............. 28 2.6 Poverty Incidence by Edu'cation of Household Head, 1971 ........ 31 2.7 Literacy Rates (1970) and Poverty Incidence, 1971 ............. 33 2.8 Urban and Rural Water Supply Coverage by Region, 1975 ......... 40 Chapter 3 3.1 Labor Force, Employment and Unemployment, 1957-79 ............. 47 3.2 Real Wage-Rate Index, 1957-78 .................................... 48 3.3 Employment Generation by Sector in the Philippines and Korea, 1963-76 ......................................... .............. 54 Chapter 4 4.1 Poverty Incidence and Agricultural Productivity Measures by Region, 1971 ............................ eso,rces ........... 60 4.2 Poverty Incidence, Productivity, and Land Resources by Region, 1971....................... I....................... 61 4.3 Gross Margins per ha by Crop Enterprise, 1976 ................. 65 4.4 Ranking of Crops on the Basis of Net Expected Return .......... 66 4.5 Growth in Poverty, Regional Output, and Population ............ 77 4.6 Average Annual Growth Rate for Value Added in Agriculture, Fisheries and Forestry (Constant 1972 Prices) ............... 79 4.7 Trends in Yields of Selected Crops ............................ 81 Chapter 5 5.1 Structure of Aggregate Urban Employment, August 1971 and August 1976, by Employment Class, Industry and Occupational Group .94  1. THE EXTENT, PATTERN, AND TRENDS OF POVERTY 1.01 This chapter analyzes the general dimensions of poverty in the Philippines. The first two sections discuss data problems and the method used to measure poverty for purposes of this report. The third section presents poverty incidence (percent of population falling below the poverty line) by urban/rural area, region, and family size for 1975. This profile is supple- mented by poverty incidences for various occupations and sectors, the necessary data for which were available only in the 1971 Family Income and Expenditure Survey (FIES). The fourth section discusses the trends in poverty and stresses that the FIES data for various years need to be interpreted carefully. The last section presents some inequality measures. Although longer term trends are analyzed in this and other chapters, the focus of the report is on the situation in the Philippines in the 1970s. A. Income and Expenditure Data in the Philippines /1 1.02 Several public and private agencies have in the past collected income and expenditure data on a nationwide basis in the Philippines. A review of the various surveys shows that the most comprehensive among them are the FIESs conducted by the National Census and Statistics Office (NCSO). These surveys have a national coverage, are based on a reasonable sample size and design, and have followed a fairly consistent methodology since they were initiated in 1957. Therefore, despite the weaknesses discussed below, these surveys have served as the primary data source for this report. Other major surveys suffered from more serious limitations in terms of area coverage, sample design, and comparability over time. Limitations of Data 1.03 Before discussing. the poverty profile and trends derived from the FIES data, it is necessary to note the limitations of these data. The main problems arise from variations in the time lags between the survey and the period covered, the types of data collected, the definitions and area coverage, and the questions asked. 1.04 First, the FIESs used the interview and recall method to collect data./2 Since some time elapses between the period covered by the survey /1 For a detailed review of the nature, coverage, and problems of the data, see Annex 1-A. /2 The households in the survey sample were asked questions regarding their incomes and expenditures. Since they do not usually keep records, they had to rely on their memory about what they earned or spent during the period covered by the survey, called the reference period. The record keeping method is intended to be used in.the 1980/81 FIES.  - 2 - and the reporting time, it is difficult for the respondents to recall their full income and expenditure. Both income and expenditure, therefore, tend to be underestimated. This difficulty increases with the length of the recall period. The problem is further compounded if the questionnaire is not sufficiently comprehensive or is not followed up by supplementary questions. 1.05 The FIESs in the Philippines have been conducted with varying time lags and, therefore, can be presumed to suffer from different degrees of recall lapse and underestimation of incomes and expenditures. The 1957 FIES was conducted in March 1957 and its reference period was March 1, 1956 to February 28, 1957. The 1961 FIES was conducted in April 1962 and the 1965 FIES in May 1966 but their reference periods were calendar years 1961 and 1965 respectively. The 1971 FIES covered the period from May 1970 to April 1971 and was conducted in May 1971. The income part of the 1975 FIES was conducted in December 1975, and the expenditure part in March 1976 (discussed in detail below). The recall lapse and resultant underestimation would, therefore, differ from survey to survey. 1.06 Second, the information collected by the FIESs has changed over time. Several variables were successively added after the 1957 FIES: family size in the 1961 FIES, area of residence and household type in the 1965 FIES, and some socio-economic/demographic variables (e.g., number of family members employed for pay or profit, age, education, occupation, and employ- ment sector of household head) in the 1971 FIES. In 1975, some economic variables (e.g., industrial affiliation of the household head) were dropped. 1.07 Third, some definitions and area groupings have changed over time. The definition of urban areas changed in the various FIESs, with the major change being inclusion of peripheral urbanized areas around urban centers starting in the 1971 FIES. The regional groupings of provinces have also changed, making it difficult to construct a systematic regional comparison over time. For 1971, however, this report has regrouped the provinces into regions to match the 1975 grouping./l 1.08 Fourth, the income data in the various FIESs were found to be more deficient than the expenditure data. The income questions asked have varied a great deal, being very detailed in some cases but less detailed in /1 The Philippines is divided into 13 regions. The division and regional names have changed periodically, the last change occurring in 1978. This report is, however, based on the regional division of the country used in reporting the results of the 1975 FIES. To allow intertemporal comparisons, the mission regrouped the provincial results of the 1971 FIES in accordance with the 1975 regional division. The map at the end of the report presents the 1975 regional division and names.  others./l On the other hand, the expenditure questions asked in the various surveys have remained virtually identical and were probed more extensively than the income questions. Information was sought on a very detailed list of several hundred items. In all the surveys, the reference period for expenditure on food, beverages, and tobacco was the week immediately preceding the interview. This minimized the recall lapse./2 The family expenditure estimates based on FIESs were also closer to those in the National Accounts (NA) than the income estimates (see Annex 1-C). 1.09 For the above reasons, the expenditure data have been used for estimating poverty incidences in this report. B. Measurement of Poverty The Concept of Poverty and Measurement Problems 1.10 At present, two concepts of poverty are commonly in use: relative poverty and absolute poverty. The former uses some measure of the relative shares in income or consumption of various sections of population, such as income share accruing to the lowest 20% or 40% of the population. This measure shows the inequality of distribution of income but does not say anything about the level of deprivation of the various sections of people. The term absolute poor, on the other hand, refers to a group that is unable to meet specified minimum needs, i.e., the group that falls below a poverty line. The term does not imply that all persons falling below the poverty (minimum needs) line are either equally or absolutely deprived. Within the "absolute" poor, those near the top of the group may be only on the margin of subsistence whereas those at the bottom may be seriously deprived. This report focuses on absolute poverty. /1 For example, the 1975 FIES, unlike some of the earlier surveys, did not ask specific questions on net income from production of articles for own use, or separate questions on commissions, tips, bonuses, etc. or on income earned from activities not operated by any member of the household, or on non-work income from land, poultry, etc. These and other variations not listed here would not only seriously affect the comparability of income surveys over time but would also cast doubt on a poverty profile based on at least the income data of the 1975 FIES. For details, see Annex 1-A. /2 This procedure creates some problems of its own. The week may not be representative of the whole year. To a great extent, the randomness and large size of the sample would ensure against serious biases on this account. All surveys were conducted between March and May. A quick check by the mission showed little bias at the national level among various surveys. There was, however, some evidence of biases across regions as prices and quantities consumed changed at different rates among them.  -4- 1.11 A systematic study of absolute poverty, however, faces a number of problems. First, to identify the poor we need to draw a poverty line based on some idea of minimum needs. Yet needs vary among individuals and over geographical areas even within a country. Second, most income and expendi- ture surveys have the family as their unit of observation and thus fail to capture significant 4ifferences in the composition of, and distribution of expenditures within, the families. Third, consumption requirements and prices of goods and services differ significantly between, and within, urban and rural sectors. Drawing one poverty line, or even separate urban-rural poverty lines, therefore, ignores these differences (of unknown magnitude) and may distort the measurement of poverty significantly. Finally, income and expenditure data are collected only every few years and may be affected by temporary factors (for examnple, changes in weather or external economic conditions). Measurement errors may also vary in size. Changes in relative price and consumption patterns introduce further complications into the assessment of time trends. 1.12 The net effect of these problems is to introduce considerable arbitrariness as well as uncertainty into the measurement of poverty incidence and even more so in the trends therein. These problems are illustrated by the fact that, depending upon the definition of minimum needs and the type of data used, the incidence of poverty estimated for the country by various researchers has ranged between 25% and 80%. Keeping in view these difficulties, we have tried to arrive at a reasonable and operationally useful poverty estimate. The Poverty Line / 1.13 The two main criteria used for drawing a poverty line are the consumption basket of the "representative poor" and the "least-cost" consumption basket necessary to meet specified minimum needs. In each case, the major component of the consumption basket is food. 1.14 The "representative" consumption basket has the advantage of using the actual consumption pattern of the poor, presumably reflecting their preferences. However, this criterion presents several problems. First, the poor are not known independently of the poverty line still to be estimated. Therefore, the consumption pattern of some relatively low income group is used./2 Second, this criterion ignores the potential use of cost minimizing efforts of households (voluntary or induced), such as using low cost foods with relatively high nutrition. Third, the actual consumption basket may contain items that society views as unnecessary or even unhealthy such as the use of alcohol and tobacco. /1 For further details, see Annex 1-B. /2 The World Bank guidelines suggest the use of the food basket corresponding to the 20th percentile in the national income distribution.  1.15 The "least-cost" consumption basket approach attempts to meet the minimum recommended dietary and other subsistence requirements at the lowest possible cost. Its advantage is that it determines the income necessary to meet the minimum needs of an individual (or a family). If the individual then chooses to spend the income.on unnecessary items or on food that is not cost-efficient (i.e., provides less nutrition per peso spent), this decision reflects his preferences but not forced deprivation. Nutritional inadequacy (and poverty) then results from his own choice. The problems with this approach are that it may ignore the prevailing tastes and that a cost- efficient choice may require nutritional education on a large scale. 1.16 Since both approaches have their advantages and disadvantages, we decided to draw a poverty line taking into account the components of both least-cost and actual consumption baskets. However, greater weight was given to the latter in order to reflect the actual preferences of the population. 1.17 A least-cost diet subject to nutritional constraints was estimated by two Filipino authors, Tan and Holazo, using linear programming./l They specified four other basic needs: medical care; housing, together with fuel, light, and water; clothing; and education- They computed poverty lines for a typical six-member family for various regions for 1971; the average poverty line approximated a family income of P 2,160, or P 360 per capita. We have used this as an estimate of the least-cost consumption basket. 1.18 To calculate a comparable poverty line based on actual consumption of the representative poor for 1971,/2 we relied on the 1971 FIES and the National Food and Agriculture Council (NFAC) survey of October-November 1970. The lowest income class for which the NFAC survey reported consumption data was P 400 or less per capita per year./3 The caloric content of the food component of this consumption basket amounted to 1,881 and was proportionately increased to meet the recommended (by the Philippine Food and Nutrition Research Institute) daily allowance of 2,016 calories for an average Filipino. The nonfood needs were then calculated from the actual expenditures /1 Tan, E.A. and Hlolazo, V., "Measuring Poverty Incidence in a Segmented Market," Institute of Economic Development and Research Discussion Paper 7814, School of Economics, University of the Philippines, Quezon City, December 1978. /2 The basic poverty line for the present report was constructed for 1971 (and then extrapolated for other years) because: (a) the 1971 FIES appeared to be the most reliable; and (b) 1971 not being an end-year of the period being studied, extrapolations from it would be less distorted. /3 Consumption of lower income classes was not reported separately. The per capita income level of P 400 corresponded to a percentile level between the 20th and 40th in the 1971 FIES, and is thus somewhat above the percentile level recommended by the Bank guideline.  - 6 - of the family income class in the 1971 FIES roughly corresponding to the P 400 per capita income group in the NFAC survey. Adding these food and nonfood consumption baskets resulted in a poverty line budget of P 560. 1.19 If a poverty line were to take into account elements of both "least cost" and "actual preferences," it would be somewhere between these two extremes of P 360 and P 560. We believed that greater weight should be given to the actual consumption pattern to take into account people's preferences but some allowance should be made for least-cost elements in the poverty budget. Consequently, a per capita poverty line of P 500 for 1971 was adopted which was about 10% below the "actual preference" consumption basket but 40% above the "least-cost" consumption basket. 1.20 Sensitivity of Poverty Incidence. Since the above poverty line, like others, is an approximation, it is useful to test the sensitivity of poverty incidence to different poverty lines (see Table 1.1)./l Not surprisingly, poverty incidence is quite sensitive to changes in the poverty line; an increase of the poverty line by P 50 raises poverty incidence by five to six percentage points. Although the poverty line used in this report is believed to be realistic,-Table 1.1 gives a rough idea of the size of error in results if the "true" poverty line is significantly above or below it. Table 1.1: SENSITIVITY OF POVERTY INCIDENCE TO THE POVERTY LINE, 1971 Percent Percentage % increase increase point increase in poverty Poverty in poverty in poverty Poverty line line incidence incidence incidence P 450 - 32.5 P 500 11.1 38.7 19.0 6.2 P 550 10.0 43.7 12.9 5.0 P 600 9.1 48.6 11.2 4.9 1.21 Urban/Rural Poverty Lines. Ideally, separate poverty lines should have been derived for different geographical areas based on local prices and consumption patterns. This was not, however, possible because price indices for individual regions were based on different and separately specified baskets of goods; without a common basket of goods they could not provide a reliable interregional comparison of price levels. Even separate price indices for urban and rural areas were not available; all prices are collected /1 Table I.5, Statistical Appendix, gives poverty incidences for various poverty lines for 1975.  in urban centers. Since there are known to be signiificant differences between urban and rural prices, it was necessary to have reasonable estimates of separate urban and rural prices. Following estimal:es by other researchers, the mission set the urban poverty line at a level one-third higher than the rural poverty line. This difference in the two poverty lines, which is assumed to cover differences in both price levels and consumption patterns between the two areas, appears to be of a reasonable magnitude. Treating the overall poverty line of P 500 for 1971 as the weighted average of urban and rural poverty lines, separate urban and rural poverty lines were then calculated. A similar calculation was made for 1975 after inflating the 1971 poverty line by the CPI./l Similar calculations were not made for the period before 1971 because (a) the definition of urban areas had changed significantly in the various FIESs, and (b) the focus of the report is on the 1970s. 1.22 Calculation of Poverty Incidences. Poverty incidence could be calculated for either individuals or families. However, from a policy point of view, it appeared to be more appropriate to construct the poverty profile in terms of family characteristics and the family head (assuming he is the primary earner) because this approach also captures the dependents of the income earner. Hence, poverty incidences by family are given in this report. 1.23 The FIES data show percentage of families of each size (from one member to 12 +) falling into 17 expenditure classes. These data are presented separately for such variables as location, occupation, industry, and education. To obtain poverty incidences, first the per capita expenditure level for various expenditure classes was calculated. Then the number of families falling below the per capita expenditure corresponding to the chosen poverty line were added up and expressed as a percentage of total number of families./2 1.24 The format for the poverty incidence tables given in the report is similar: there is a cross-classification of poverty by two variables. In each cell (which defines a population group by particular values of the two variables), the first figure gives the percentage incidence of poverty in the group and the second gives the number of poor fami]ies./3 This information allows us to focus on two distinct aspects of poverty. The poverty incidence figure allows us to locate "high risk" groups (who may, in fact, account for only a small proportion of population in poverty). The absolute number of poor families allows us to locate particular concentrations of poverty. /1 The 1975 poverty lines are: Urban = P 1,103; Rural = P 827. /2 For details, see Annex 1-B. /3 The tables in the Statistical Appendix also contain a third entry which provides the total number of families for each cell.  -8- C. The Poverty Profile 1.25 This section discusses variations in poverty incidence by region, urban/rural area, and family size based on the 1975 FIES, the most recent survey available. Data on some other variables such as industrial affiliation and occupation of the household head were not available as part of this survey. The 1971 FIES data were therefore relied upon for the analysis of these variables. Although these data are a decade old, and the pattern of poverty may have changed somewhat since then, they serve as a useful reference point and provide insights into some aspects of poverty./l 1.26 As noted in Section A, the 1975 expenditure survey was conducted about five months after the reference period and nonfood expenditures may have been underestimated to a greater degree in this survey than in previous years./2 Furthermore, the ratio of family expenditure estimated from the 1975 FIES to the one estimated from the national accounts data is 0.70 compared to 0.94 for 1971 (see Annex 1-C). Thus, the 1975 FIES appears to have underestimated family expenditures more than the 1971 FIES. This underestimation is probably largest at the higher income levels, but must also have occurred at lower income levels. Consequently, if the expenditures were underestimated by 5-10%, as seems probable, the "true" poverty incidence would be 2-4% lower than the tables show. Poverty Incidence by Location 1.27 Table 1.2 presents poverty incidences by region and rural/urban area for 1975. Poverty incidence for the whole country was 45%, with urban poverty at 40% and rural/poverty at 47.5%. Since nearly 70% of the populiation lives in rural areas and rural poverty incidence is higher than urban, rural areas accounted for nearly three-fourths of the total poor. 1.28 Poverty incidence varied'substantially across regions, from a low of 29% in Central Luzon to a high of 73% in Northern Mindanao. This variation was greater in rural areas than in urban areas. The regions with the highest poverty incidence were Northern Mindanao,/3 Central Visayas, /1 The 1971 FIES also contained data on expenditure distribution by level of education and employment status of the household head. Poverty incidences based on these data are discussed in Chapters 2 and 3 respectively. /2 Food expenditure data related to the week preceding the actual time of the survey and, therefore, did not suffer from this weakness. /3 Northern Mindanao showed an unusually high poverty incidence in 1975 with almost 30% increase over 1971 (see Statistical Appendix, Table I.3), for which the mission could not find a satisfactory explanation. Never- theless, there was a consensus during field discussions that Northern Mindanao is a high poverty area.  - 9 -- Table 1.2: POVERTY INCIDENCE BY REGION AND RURAL/URBAN AREA, 1975/a Region I Region II Region III Region IV Region IV-A Region V Region VI Cagayan Central Metropolitan Southern Western Ilocos Valley Luzon Manila Tagalog Bicol Visayas Rural Poverty Line P 827 Per Capita Urban Poverty Line P 1,103 Per Capita Rural 37.6 44.9 26.9 (0) 47.0 56.4 49.4 (173.0) (131.2) (128.7) (0) (301.0) (242.5) (255.4) Urban 42.9 50.5 33.9 30.9 43.1 50.4 43.4 (42.1) (19.2) (62.7) (237.8) (107.3) (43.9) (70.4) Total 38.5 45.6 28.9 30.9 45.9 55.4 48.0 (215.1) (150.4) (191.4) (237.8) (408.3) (286.4) (325.8) Region VII Region VIII Region IX Region X Region XI Region XII Central Eastern Western Northern Sout:hern Central All Regions Visayas Visayas Mindanao Mindanao Mindanao Mindanao Rural 62.6 56.0 46.0 75.3 41.2 28.4 47.5 (274.6) (202.8) (125.5) (228.3) (130.6) (72.0) (2,262.8) Urban 52.5 56.1 45.4 61.1 42.4 49.9 40.2 (81.8) (44.9) (18.6) (40.9) (49.2) (23.4) (842.1) Total 59.9 56.0 45.9 72.8 41.5 31.7 45.3 (356.4) (247.7) (144.1) (269.2) (179.8) (95.4) (3,104.9) /a In each cell, the first figure shows the (7) incidence of poverty and the second in parentheses shows the number of poor families (in thousands). Source: FIES, 1975, NCSO.  - 10 - Eastern Visayas, and Bicol, in that order. These regions-also showed relatively high poverty in both rural and urban areas. The highest concentrations of the poor were in the more populous regions of Southern Tagalog, Western Visayas, and Central Visayas, which together accounted for about 35% of the total poor. 1.29 While the 1975 poverty-incidence in the Philippines appears to be high, it was, in fact, only slightly above the average for middle income LDCs according to one Bank study using a uniform poverty line./l Another study, however, shows that the Philippines had a significantly higher poverty incidence than the other large developing countries of the East Asia Region except for Indonesia./2 It should be noted that these are only rough estimates because, given substantial variations in poverty thresholds and quality of data, making accurate comparisons across countries is even more difficult than measuring poverty in one country. Poverty Incidence by Size of Family 1.30 For the country as a whole, as well as for rural and urban areas separately, poverty incidence was below average up to a five-member family and above average beyond it (Table .1.3). More significantly, the larger was the family the higher was the incidence of poverty over practically the whole range of family size in both rural and urban areas. Families with six or more members constituted about 52% of the total families and accounted for 68% of the poor families and about 30% of the poor population./3 1.31 Since no allowance was made for the lower needs of children, poverty incidence in larger families, who tend to have a higher proportion of children, is probably overstated. The pattern noted may also reflect the life cycle effect of families experiencing reduced per capita income during reproductive years and before ~children have reached income-earning capacity. Nevertheless, the increase in poverty incidence for-most family sizes appears to be significant enough that the relationship would hold even if allowance could be made for the lower needs of children. These results lend support to the hypothesis that low income families tend to be large and large 11 Montek S. Ahluwalia, et al, Growth and Poverty in Developing Countries, World Bank Staff Working Paper No. 309 (Revised), World Bank, Washington, D.C., 1979. This study used a uniform poverty line of income per head accruing to the 46th percentile of, the population in India and adjusted the income data for differences in purchasing power among countries. /2 Willem Bussink, et al, Poverty and the Development of Human Resources: Regional Perspectives, World Bank Staff Working Paper No. 406, World Bank, Washington, D.C., 1980. /3 Poverty incidence by family size based on the 1971 FIES also gives similar results. See Statistical Appendix, Table 1.4.  - 11 - families tend to be poor because of their high dependency ratios and lower capacity for saving and investment. This suggests that poverty redressal would simultaneously require family planning, more employment opportunities, and decreased dependency in larger families. Poverty Incidence by Occupation /1 1.32 Poverty incidence showed surprisingly small variation within the rural and urban occupations (Table 1.4) in 1971./2 Apart from higher earning professions, poverty incidence in the major rural occupations varied between 59% for tenant farmers to 48% for farm owners. The largest concentration of poverty was in tenant farmers (24% of the total number of poor), followed by farm owners (20%), fishermen (7%), unspecified farmers (7%), and farm laborers (6%). 1.33 The somewhat high incidence of poverty among tenant farmers (59%) compared to farm laborers (52%), who are usually considered to be among the poorest, was a little surprising. Since the 1971 FEES reported a relatively low number of farm laborers, it is possible that some of them were included among tenants due to classification problems. It is also possible that their expenditures were more in cash and, therefore, have heen fully captured, while those of other (subsistence) farmers have not. Hence, the apparent difference in the poverty incidence of these groups may be partly due to measurement problems. 1.34 The regional distribution of poverty for the occupational groups is discussed in detail in the sectoral chapters of the report. It may, however, be noted here that extremely high incidences of poverty existed in certain occupations and regions. For example, poverty incidence was over 90% among tenant farmers and farm laborers in Central Visayas, over 80% among unspecified farmers in Western Mindanao, Central Visayas, and Eastern Visayas, over 80% among farm laborers in Bicol and Central Visayas, over 70% among farmer tenants in Eastern Visayas and Northern Iind.anao, and 87% among fishermen in Central Visayas. Since most of these high incidences pertained to the poorer regions, they again emphasize the regLonal character of poverty. These instances of extreme poverty probably need further investigation to again determine if special steps are needed to deal with them. /1 Based on the 1971 FIES. /2 Note that Tables 1.4 and 1.5 use a common poverty line of P 500 because expenditure distributions for occupation and industry were not available by urban and rural areas separately. With separate poverty lines, poverty incidence for rural occupations would probably drop by about five percentage points and that for urban occupations would increase.  - 12 - Table 1.3: POVERTY INCIDENCE BY SIZE OF FAMILY, 1975 Number of Persons in Family 1 2 3 4 5 6 7 Rural Poverty Line P 827 Per Capita Urban Poverty Line P 1,103 Per Capita Rural 11.7 15.6 26.8 36.3 40.6 55.3 60.1 (7.84) (43.37) (148.20) (237.40) (283.39) (380.46) (331.75) Urban 5.1 10.5 20.4 30.6 37.5 44.6 51.7 (1.78) (13.96) (48.35) (99.45) (123.37) (129.34) (120.46) Total 9.4 13.9 24.9 34.4 39.6 52.1 57.6 (9.62) (57.33) (196.55) (336.85) (406.76) (509.8) (452.21) Number of Persons in Family All 8 9 10 11 12 or more families Rural 63.2 66.2 65.2 69.7 65.5 47.5 (300.2) (229.05) (142.14) (82.25) (76.64) (2,262.8) Urban 54.6 61.8 62.7 64.1 61.0 40.2 (105.38) (77.25) (53.92) (34.61) (34.16) (842.1) Total 60.8 65.2 64.5 67.9 64.0 45.3 (405.58) (306.3) (196.06) (116.86) (110.8) (3,104.9) /a The first number in each cell gives poverty incidence (%) and the second number in parenthesis shows the number of poor families (in thousands). Source: FIES, 1975, NCSO.  Tqbl. 1.41 POVERSt INCIDINCIS 81 OCCUIPATION AND R8GIO 1971 /na Pove-t, Un. * 500 PAr C ptt, Maio- I Rbolon II 0,00n IV R l ItV.-A Roln V RZoV V !2 Vll lakjOt VlIl lo_fLI 1 on X Rn Xl on Xl 1.5.5 Cgaaap Cootral K.tr.5,olittn Southern W t1cr. Cantr.! l t 1.?. W ter. 80wrt tn S.uther, CVnMAl Ph llppin- Tlot.. VallUy tcon KNnll Tas lnI BLcol VlMyIS Vly.. 1V.47yes 9lnd.o. H10d6a.o HIMnde.p Mindaa.. Prof...Ion... * 22.0 2.8 1.2 8.2 9.0 5.7 6.5 7.2 * 33.3 1.4 2.9 5.6 .d,lol.trntlon (0.0) (3.3) (1.3) (10.) (0.6) (3.6) (2.4) (2.2) (1.3) (0.0) (6.0) (4.3) (0.8) (31.9) oud clnloul vorkor. /b -k1. oork.r. 26.0 * 12.4 0.8 27.9 19.7 27.9 34.0 35.3 17.6 39.1 I1.8 7.3 20.0 (5.9) (0.6) (6.6) (5.3) (14.0) (6.3) (13.4) (12.6) (11.3) (6.0) (4.3) (2.0) (2.2) (94.0) F.r.nr o.n.r. 34.0 53.0 9.2 ' 36.8 59.0 43.4 80.7 66.1 *0.0 57.5 02.6 13.0 48.4 (28.2) (39.7) (3.0) (0.0) (52.7) (44.6) (30.0) (48.8) (72.7) (29.4) (57.)) (65.2) (12.6) (301.0) Faner pant ocnon. 49.7 79.2 17.4 ' 37.7 57.8 51.8 92.8 84.2 ' 75.0 42.5 * 50.4 (15.45 (12.7) (4.9) (0.0) (11.3) (4.0) (13.0) (11.1) (11.8) (5.5) (3.0) (2.6) (1.0) (94.3) lam,? tnn.,t 53.9 69.6 27.4 ' 42.5 66.1 50.6 91.3 79.3 73.8 79.8 62.4 40.0 09.2 (64.1) (67.0) (30.9) (0.0) (50.1) (83.9) (54.0) (55.7) (55.6) (27.3) (48.75 (36.2) (11.3) (584.0) F.r" - not po- 54.6 80.3 15.O 38.0 73.9 37.0 83.8 03.0 85.4 72.5 03.0 37.1 57.6 tIOted .nd (13.1) (10.4) (3.8) (0.0) (3.6) (20.0) (3.)) (21.8) (28.2) (20.5) (18.1) (5.8) (17.8) (164.8) tuber gather.?. 7.?s lubonne. 25.0 66.7 33.9 ' 48.9 80.0 49.2 0.9 65.8 29.3 4 51.9 (2.0) (6.7) (8.8) (1.0) (23.)) (22.4) (43.3) (21.8) (7.2) (1.0) (6.0) (9.4) (2.0) (68.8) Flab rn.0n 67 4 * 31.0 I 49.4 55.6 *6.9 87.0 65.6 53.8 70.3 M 19.) 53.6 I.borer. and (16.2) (0.0) (7.4) (0.1) (23,2) (153.9) (19.15 (34.8) (23.5) (15.6) (146.) (4.1) (3.9) (173.2) rel.ted c.ost.d. Lo8an.rOd otbr ' ' a, 4 ' ' * . * 4 4 4 a 4 41 4 erect? ,ork, (0.0) (0.0) (0.0) (0.0) (3.7) (3.1) (2.0) (1.0) (3.0) (0.0) (2.5) (1.0) (0.5) (18.6) incr,..q ..r..0n ' ' ' ' ' ' ' ' ' ' ' ' 28.6 .nd n1.t.d (1.4) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (2.0) (0.0) (0.0) (1.0) (0.0) (0.0) (4.3) 'Tr*.port cod 42.2 4 20.4 12.5 37.1 28.0 22.3 52.6 48.5 30.0 38.3 15.4 2 28.7 coaunioatboa. (11.4) (2.0) (10.3) (10.5) (16.7) (4.2) (6.9) (22.6) (6.3) (3.3) (4.6i (5.0) (1.3) (103.5) corker. C,.fltn, prodot- 52.3 42.9 16.2 13.8 38i3 32.8 29.6 54.5 46.9 31.0 69.4 18.9 26.9 31.7 n.on vork.r e labretna (35.6) (11.6) (15.1) (25.8) (46.7) (18.4) (28.1) (44.9) (21.6) (9.6) .(5.0) (6.4) (7.8) (287.5) 0.Ooi?,4. apont 24.2 16.4 11.s 40.6 27.6 23.7 51.3 45). 15.2 53. . 37.3 27.7 Iad r -r.ttioo (4.6) (4.6) (3.4) (7.7) (13.0) (5.0) (6.4) (1504) (0.0) (2.0) (9.1) (1.8) (6.1) (81.2) Oc,p.tlon not 33.8 37.3 11. 9 9.3 27.2 38.9 33.1 51.8 44.0 55.0 42.6 25.3 12.8 28.8 rap .t.d or (20.3) (9,7) (9.)) (7.1) (21.2) (14.4) (19.5) (22.8) (13.2) (6.4) (9.4) (4.3) (2.7) (170.1) Isd q..t*ly d.ctrlbed All nIc.uptloM 41.7 54.7 17.6 9.1 35.1 48.8 38.3 64.7 59.7 38.1 58.2 39.7 20.4 38.7 (225.5) (166.9) (107.9) (61.3) (264.5) (242.1) (245.1) (355.3) (257.2) (120.8) (193.*): (143.6) (67.%) (2,456.0) A l. ..th -e11 the th lr.t flgr- choc th. (2) 1oidd.. ot p.o.arty nd tbv. soad ln pr.np h.esa lbov.. the n-btc at poor f4.1114s (Un thaoqam):. /b InClud.c t lc, n ..tia 4nd .aaag.rll (far. *nd not..) vorkar.. 040tc 0.pl0 Ola tfl insl for 'tatl.t..ally a .I.d Inf.r.ct.. Soo?oet 21dS. 1971. NSCSO  - 14 - Poverty-Incidence by Sector/Subsector 1.35 Table 1.5 presents the incidence of poverty by industrial sector/ subsector (of family head) and region.based on the 1971 FIES./1 Rice /2 and corn farming showed a consistently high poverty incidence across regions. By far the largest number of poor (48% of all poor families in the Philippines) were dependent on these two crops. 1.36 At the national level, poverty incidence showed relatively small variation among farming families dependent on different crops. There was, however, a large variation in poverty incidence in individual crops across regions. For example, the incidence was very high among rice and corn farmers in Eastern and Central Visayas but very low in Central Luzon and Central Mindanao. Similarly, poverty incidence in coconut farming, "other crops", and fishing was very high in Central and Eastern Visayas but very low in Central Mindanao. 1.37 Table 1.5 also shows that poverty incidence was generally low in manufacturing and services, the lowest being in government and community services. Hawking, peddling, and, domestic services /3 were the only nonagricultural categories with a relatively high poverty incidence, which nevertheless was lower than for most agricultural categories. D. Trends in Poverty, 1957-75 -1.38 ---Several problems arose in-estimating time trends--in-poverty incidence. The first problem related to the'use of appropriate price index for adjusting the poverty line to various FIES years. To adjust the poverty line over time, we should ideally use a price series based on the poverty basket of goods and services rather than on an average basket such as the CPI. However, the Philippines does not have a low income CPI series covering the whole country and all the FIES years. In fact, the only series which spans all the FIESs and covers the Philippines is the CPI with 1965 as the /1 It is reassuring that the data on occupation and industrial sector of the family head, which were gathered independently by means of quite separate questions, are consistent. For example, the occupation table shows 323,000 fishermen families with a poverty incidence of 53.6%, while the industry table shows a total of 329,000 families engaged in fishing, hunting, etc., with an identical poverty incidence. /2 Rice and "palay", which means paddy rice in Filipino, are used interchangeably in this report. /3 These categories are not separately shown in Table 1.5. Poverty incidence in hawking and pedddling was 43% and domestic services 48%. See Statistical' Appendix? Table I.2.  TWbA .0: ?OVfnTf IIICID80 9CC BY 280OT0AL 0ECrOR A6IM 020006 1971 LA Poverty Lin 500 Par Calte belayn 1 FlmJj XI ojan Xll PnItoa IV kA lbotn 1V0A bRaon V Rna10n VI ROnton 6(1 R4.Ooa VIII Mgtaln 02 Raton I RtOln ) MAjrL II1 Tota1 Cn4yan Co-trol brr-p-tce r Maeten. anto V-tr abn-r-. eob- nrrt.ara Sootlero C0nt.-l ?hiltpptro lloeo Vally L.aA.n HGta Isaa S- n DAiL VOLspy. Vlszyco VOeYe