Philippines - Recent trends in poverty, employment and wages

討＿が＿-，、：_,1,:.：数－＿玉’「’ CURRENCY s Average Exchange Rate Year (Pesos per US$) 1970 5.90 1971 6.43 1975 7.24 1980 7.51 1981 7.90 1982 8.54 1983 11.11 1984 16.70 FOR OFFICIAL USE ONLY THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES Table of Contents Page No. EXECUTIVE SUMMARY ........................................ v I. INTRODUCTION .............................................. 1 (i) The Background ......................... 1 (ii) Objectives of the Present Report ....... ........ 1 II. FAMILY INCOME AND EXPENDITURE DATA ...... ............... 2 A. Sources of Data ................. ............ 2 (i) Income/Expenditure Data Prior to 1976 ....... 2 (ii) Data Sources and Data Availability, 1976-1983 ...... 2 B. Limitations of the ISH ........................... 3 (i) The Income Concept in ISH ................. 3 (ii) Underreporting of Income ............................ 3 (iii) Differential Implications of Family Income and Family Expenditures ........................... 5 (iv) Ranking by Family Income, Family Expenditure and Per Capita Expenditure: Illustration of Quantitative Differences .............. 6 III. TRENDS IN POVERTY AND INEQUALITY ........................... 8 A. Trends in Poverty Incidence ..................... 8 (i) Poverty Trends, 1965-75 ...................... 8 (ii) Poverty Trends, 1971-83 ............................ 9 (iii) Poverty, 1979 and 1983: A Panel Study of 3,294 Households .................................... 12 (iv) The Poverty-Employment Linkage ..................... 14 B. Trends in Inequality ..... .......... o .... .. 16 (i) Family Income Inequality .................... .. 16 (ii) Rural-Urban Inequality ....... . ............... 18 This report is based on the findings of a reconnaissance mission which visited the Philippines in September 1984. The mission consisted of Bhanoji RAo (mission leader) and Sudhir Anand (consultant). The report was discussed with the Government in March 1985. Subsequently, in June 1985, written comments from the Government were received and these were taken into account in the preparation of the final report. Tis document has a restricted distribution and may be used by recipients only in the performance of omkofrcil duies. Its contents may not otherwise be diclsed wiftut Wod Bank authodntion. -11 - IV. TRENDS IN EMPLOYMENT AND WAGES ............................ 20 A. Employment Trends ................................... 20 (i) The Data Situation ...................... ............ 20 (ii) Employment and Unemployment ........................ 21 (iii) Sources of Employment Expansion ..................... 23 B. Trends in Real Wages .... 25 (i) The Data ................ ...........................25 (ii) The Trends, 1970-83 ............................... 25 (iii) Wages and Poverty Thresholds ................................. 31 V. RECOMMENDATIONS .......................................... 33 (i) Monitoring Poverty _-.................... ........ 33 (ii) Data Collection and Dissemination .................. 33 (iii) Statistical Organization o.................. 34 REFERENCES ....................... 35 ANNEXES: 1. Integrated Survey of Households ........ ............ 38 2. The 1975 Integrated Census of the Population and Its Economic Activities ....................... ........................ 51 3. The 1981 Census of Agriculture and Fisheries.................. 54 4. Sources of Employment and Wage Data... ................ ........ 60 5. Poverty Lines in Earlier Philippine Research................... 66 6. Poverty Lines Based on Rice, Other Food and Non-Food Expenditures ............... 69 7. Statistical Appendix ..................... ........ ..... 76 TABLES IN TEXT 1. Comparison of National Accounts and ISH Income Estimates, 1980-83............... ..... 4 2. Comparison of National Accounts and FIES Income and Consumption Expenditure Estimates, 1961-75........... 5 3. Poverty Estimates from Income and Expenditures Distributions... 7 4. Poverty Lines and Poverty Incidence, 1971-83...... .. ......... 10 5. Poverty Incidence by Main Source of Income........... .... 12 6. Shifts in Poverty Group 1979-83, Urban and Rural Areas......... 14 7. Quintile Income Shares 1971-83................................ 16 8. Population, Labor Force, Employment, Unemployment and Underemployment, 1971-83..................................... 22 9. Underemployment Among the Poor and Nonpoor by Occupation and Industry, Fourth Quarter, 1978......... ............ 26 10. Index of Average Daily Basic Wage Rates of Laborers in Industrial Establishments in Metro Manila 1970-80............ 28 11. Daily Wage Rates of Laborers in Agricultural Operations, 1970-82.. ....................... 29 - 111 - 12. ILlustrative Comparison of Poverty Threshold Earning Levels and Average Daily Wage Rates, 1971-83..................................... 32 TABLES IN ANNEXES ANNEX I 1. Summary Information on ISH Surveys - ISH Family Income Data.... 46 2. Status of ISH Tabulation...................... ............... 48 ANNEX 4 1. Sources of Data on Employment, 1956-83......................... 61 2. Employment Estimates from the Household Surveys and Establishment Surveys 1978 and 1980.......................... 62 3. Comparative statistics on Employment Status of Persons 15 Years Old and Over, Based on Weekly and Quarterly Reference Periods, Third Quarter, 1976............................... 63 ANNEX 6 1. Rural and Urban Population, 1980 and Retail Price of Rice, 1975, 1978 and 1983 by Region............ o .................. 71 2. Regional Consumer Price Indices 1975-83.... ................... 72 3. Price Indices for Deflating the 1983 Poverty Line to 1975 and 1980-82.... ...................... ................. 73 4. Poverty Lines, 1975-83........................................ 74 5. Comparison of Selected Poverty Lines..... .. ................. 75 STATISTICAL APPENDIX 1. Per Capita Income and Consumption, 1970-83..................... 77 2. Family Income Distribution Data, Third Quarter, 1980........... 78 3. Family income Distribution Data, Fourth Quarter, 1980.......... 79 4. Family Income Distribution Data, Third Quarter, 1981........... 80 5. Family Income Distribution Data, Fourth Quarter, 1981.......... 81 6. Family Income Distribution Data, Third Quarter, 1982........... 82 7. Family Income Distribution Data, Fourth Quarter, 1982.......... 83 8. Family Income Distribution Data, Third Quarter, 1983.......... 84 9. NCSO and World Bank Data on Metro Manila Income Distribution, First Quarter, 1983................................... ........85 10. World Bank Data on Metro Manila Household Expenditure Distribution. 86 11. World Bank Data on Metro Manila per Capital Expenditure Distribution .......6....9...... ....................... 87 12. Poverty Incidence Trends, 1965-75, Bank Poverty Report, and Abrera Study ............. .................................... 88 13. Regional Poverty Lines and Poverty Incidence, 1983............. 88 14. Poverty Incidence, 1979-83, Among a Panel of 3,294 Families.... 89 15. Percentage Distribution of Families with Specified Income Increase Between 1979 and 1983............................... 89 16. Poverty Incidence by Level of Education 1979 and 1983........ 90 17. Income Shares of Bottom 20% and Top 20% of Families in Selected Countries ...................... o .......... . .......... 90 18. Estimates of Average Annual Family Income, Rural and Urban..... 91 19. GDP Per Worker in Agriculture and Nonagriculture, 1975-83...... 92 - Iv - 20. Estimates of Ratios of Agricultural and Nonagricultural Incomes Per Worker, 1975-83....... .... . . ... . .. .. ...... *.... 92 21. Terms of Trade in Agriculture, 1969-82............... 93 22. Estimates of Number of Households, Population and Average Household Size, 1961-83. .. .. . .. . . ... . .. . ... ... . .. . .. o..o. . 94 23. Poverty Incidence Among the Employed Persons, 1978...*..... .... 95 24. Structure of Employment, 1965-82 .....o*........ 0 .*...... ...... 96 25. Employment Change by Sectors in Urban Areasoee.oe.asono...a. 97 26. Employment Change by Sectors in Rural Areaso........... .**.. 98 27. Direct Government Employment in Relation to Total Employment, 28. Employment Growth 1976-82 by Category of Workers ............... 99 29. Employment Change by Major Regionso........... ...s.99.....es 100 30. Unemployment Rates by Sex, Metro Manila and Other Areas, 31. Occupation by Three Principal Educational Levels, 1976 and 1982. 101 32. Average Daily Basic Wage Rates of Laborers in Industrial Establishments in Metro Manila, By Occupation, 1970-80 ....o.. 102 33. Nominal Daily Wage Rate Changes in Metro Manila, 1980-83 ....... 103 34. Legislated Minimum Wage and Reported Wage, 1971-1982 ...... o.... 104 35. Estimated Number of Households and Employed Persons by Rural and Urban, 1976 and 1982................ .............. 105 36. Projected Labor Force, 1980-2000 .................... 105 -v- THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES EXECUTIVE SUMMARY Introduction 1. Despite rapid economic growth during the 1970s, the Philippines has been noteworthy a country having a relatively high level of poverty. A 1980 Bank study noted that in 1975, 45% of the Philippine families were estimated to be Living in poverty. Real wages, furthermore, showed a persistent downtrend during the 1970s, despite increases in real per capita income. The report also pointed out that the distribution of income changed little during 1965-71. 2. The purpose of this review, which is the outcome of a reconnaissance mission which visited Manila in September 1984, is to make a quick survey of recent trends in poverty, employment and real wages, evaluate the relevant data and make recommendations to improve the data base. The Trends in Brief 3. This report points to both adverse and favorable trends affecting Living conditions of the people of the Philippines in the past ten to twelve years. One adverse trend was the decline in the real wage, which was also associated with an increase in underemployment, relatively high rate of growth of unpaid family workers in the rural areas and the failure of industry to absorb additions to the urban labor force. The other adverse trends include an increase in family income inequality and a widening of rural-urban income differentials. The major favorable trend was a reduction in the percentage of families living in poverty in both rural and urban areas. The apparent contradiction between the wage and poverty trends can in part be explained by the following observations. First, in rural and urban areas, even though real wages declined, families headed by wage-earners generally had lower levels of poverty incidence. Thus, a decline in real wages still left most of the families with total family incomes above the poverty line. Second, demo- graphic changes point to a reduction in average household size from 6 in 1971 to 5.6 in 1980, while the increase in female labor force participation has resulted in an increase in the average number of employed persons per house- hold from 1.8 in 1971 to about 2 in 1980. These two changes lowered the dependency burden from 3.3 in 1971 to 2.8 in 1980. Despite these observa- tions, the deficiencies in the data base, both in regard to wages and in respect of family incomes may also be at the root of the apparent contradictory trends in wages and poverty. 1/ Aspects of Poverty in the Philippines: A Review and Assessment, Report No. 2984-PH, December 1, 1980. -vi- The Data Situation 4. Since 1976, the National Census and Statistical Office collected data on family incomes as part of the Quarterly Integrated Surveys of House- holds (ISH). The earlier Family Income and Expenditure Surveys (FIES), obtained data on both incomes and consumer expenditures. Expenditure distri- butions are generally preferred to income distributions for poverty analysis. However, since ISH did not collect expenditure data, poverty analysis in the present report had to be based on income data. 5. A second Limitation of the ISH is that the data officially published and tabulated so far are mostly based on income per family, which does not adjust for family size differences. This limitation is partly surmounted in this report by obtaining a few special tabulations on the distributions of families by per capita income class. 6. Employment and wages data also have some limitations. The coverage and concepts of labor force and employment have changed in 1976 with the launching of the ISH. Data on rural wages cover only four crops - rice, corn, coconut and sugarcane. Urban wage data were limited to Metro Manila and the collection of the data on a comparable basis was discontinued in 1980. A. Major Findings Trends in Poverty Incidence 7. The analysis in this report is based on poverty lines that provide the minimum required food and nonfood expenditures. For the benchmark year 1983, separate poverty lines were computed for rural and urban areas and the thirteen regions of the country. Special poverty level food and nonfood weights were used to extrapolate the 1983 rural and urban poverty Lines to earlier years. 8. There was a decline in the incidence of poverty in the early 1980s compared to the early 70s. The percentage of the families below the poverty line declined from 57% to 47% (1971-1980) in the rural areas, from 35% to 28% in the urban areas and from 51% to 41% in the country as a whole. Furthermore, during 1980-83, the level of poverty did not change significantly despite worsening economic conditions and declining per capita incomes. The number of familes in poverty, however, increased from 2.5 million in 1971 to 2.8 million during 1981-83 in the rural areas, and from 3.3 million to 3.6 million nationally. In the urban areas, the number remained constant at about 0.8 million. In 1984, the poverty situation probably worsened because of the very high rate of inflation and the negative economic growth rate. 9. The highest rates of poverty incidence (between 50% and 51%) were found in Western and Central Visayas. Relatively high rates (around 40%) were also observed in Ilocos, Cagayan Valley, Bicol, Western Mindanao and Northern Mindanao. Rates of around 30% were found in Central Luzon, Southern Tagalog, Southern Mindanao and Central Mindanao. This finding is broadly similar to that of the Bank's 1980 study. - vi - 10. A little over half (54%) of the urban families derived their main livelihood from the non-agricultural sector by providing skilled and unskilled labor. Poverty incidence among these families was relatively Low at 16%. The next numerically large group comprises a fifth of the urban families deriving income predominantly from dividends, rents and pensions. This group also had a relatively low poverty incidence (18.5%). Among the families who were own- account workers in non-agricultural sectors (18% of the total), poverty incidence varied from 17% to 24%. Finally, farming, fishing and Livestock in the urban areas were characterized by relatively high rates of poverty incidence (between 45% and 50%); but only 8% of the total urban families depended on these activities in 1983. 11. In the rural areas, relatively high rates of poverty incidence occurred among the families dependent on corn, coconut, other crops (notably sugarcane) and fishing, and among agricultural labor families. Even among the families dependent on manufacturing and those deriving livelihood from dividends, rents and pensions, the poverty incidence rates were close to 40% to 50%. Non-agricultural labor was a source of livelihood with relatively low poverty incidence (26%). This group, however, accounted for only a quarter of the total rural families. 12. The characteristics of a group of 3,294 families in the 1979 ISH which was also covered in the 1983 ISH were studied. Poverty incidence among these families declined marginally from 43% in 1979 to 40% in 1983. Of the poor in 1979, about 40% were able to raise themselves from poverty by 1983, while 60% continued to remain poor. Meanwhile, of the nonpoor in 1979, 25% fell below the poverty line by 1983 while 75% continued to be nonpoor. Thus, while the overall poverty incidence changed only marginally between 1979 and 1983, there was substantial mobility between the poor and the nonpoor groups. Some 27% of the rural families and 23% of the urban families were poor in 1979 as well as in 1983. 13. Within the families referred to in the previous para, among those headed by persons with elementary education, regardless of urban or rural location, poverty incidence was a high 50% in 1983. Rural-urban differentials become greater, however, as one moves higher on the educational ladder. For instance, within rural and urban families headed by high school graduates, the respective rates of poverty incidence were 37% and 28%; and within the rural and urban families headed by college graduates, the respective rates were 23% and 14%. These rural-urban differentials as well as the fairly high absolute levels of poverty incidence may be in part due to the possible excess supply of high school and college graduates. 14. Data on the distribution of the employed population by earnings Levels were available for 1978. By suitably converting the per capita poverty line into an earnings threshold, it was found that 34% of the urban earners and 40% of the rural earners were poor (on the basis of earnings from work -- not counting pensions, rents, dividends and similar unearned income accruals). Data by sector of activity point out relatively high poverty incidence in all the services sectors in urban areas and in agriculture, manufacturing and distributive trade in rural areas. - viii - Trends in Inequality 15. During the period 1956-71, according to the available evidence, there was little change in the income distribution, and the bottom 40% of the families received a share of income of around 12%. Between 1971 and 1980-83, however, there was a slight worsening of income distribution. The share of the bottom 40% declined from 12% in 1971 to less than 10% during 1980-83. The shares of all the first four quintiles declined similarly, while the share of the top 20% increased from 53% in 1971 to 59% in 1980-83. 16. The Philippines compares unfavorably to certain other Asian countries such as Hong Kong, Japan, Korea and Taiwan where the share of the bottom 20% of the families is between 6% and 9% compared to 3.9% in the Philippines. Income distribution in the Philippines is very similar to Malaysia and more or less similar to some of the Latin American countries. 17. As in the case of family income inequalities, there was a deter- ioration in the rural-urban income differentials. The rural-urban per capita income ratio declined from 0.75 in 1975 to about 0.5 in the period 1980-83. Income per worker in agriculture was about half of the income per worker in nonagriculture in 1975. This deteriorated to a third in the early 1980s. This was brought about, not by a deterioration in agricultural productivity, which in fact increased during 1975-80, but by a deterioration in the agri- cultural terms of trade, by about 30% during 1975-82. This deterioration was partly the result of Government policies which tended to raise the prices of agricultural inputs and kept down output prices. It was also in part due to a relative deterioration in the prices of the country's export commodities. Domestic agricultural price policies are presently under review by the Government. Employment Trends 18. The average annual labor force growth rate during 1971-82 was a high 3.6%. This reflects both a rapid growth of the working age population, and a rising labor force participation rate especially in the case of the females. Between 1971 and 1982, labor force participation rose from less than 60% to 64%. The result has been a downward pressure on real wages and growing levels of underemployment. In the Philippines, as in many other developing countries, unemployment rates do not indicate the well-being of the popula- tion. In fact, low unemployment rates -- such as the 4.1% of the Last quarter of 1983 -- indicate that economic difficulties force many people to do some work and earn some income. The underemployment rate (percent of the labor force working less than full-time as well as working full-time, but wanting additional work), on the other hand, is a better indicator of the labor market situation. The rate was on the increase since 1978 and reached its peak, 30% of the labor force, in the last quarter of 1983, compared to 14% in 1978. 19. The share of agriculture in total employment declined from 60% in 1965 to 50% in 1971 and then averaged 52-53% during 1976-82. Manufacturing share remained around 19-20% since 1965. Services share increased from 22% in 1965 to about 30% in 1971 and stayed at that level in the 70s and the early 80s. Thereewas thus virtually no relative shift from agriculture to industry in the past one and a half decades. -ix - 20. Urban employment growth during 1976-82 (2.8% p.a.) slackened com- pared to the growth rates during 1971-76 (4.9Z p.a.) and 1965-70 (3.6% p.a.). Urban employment increase largely came from the services sector. This sector accounted for 93% of the total job increase during 1976-82, compared to 731 during 1971-76 and 641 during 1965-70. In general, industry failed to absorb the increase in the urban labor force. Recent economic difficulties especially reduced the tempo of employment expansion in Metro Manila. The region accounted for only 5Z of the employment expansion during 1976-82 as against its share of 161 in the population increase. 21. Rural employment expanded at an annual rate of 5.1% during 1976-82 as against 4.7% during 1971-76 and a mere 0.6% during 1965-70. The very high rural employment growth rate during 1976-82 was partly due to the expansion of rural non-agricultural activities during 1976-78 and partly due to the drift of labor from the urban areas to the rural areas during the economic recession of 1981-82. In the rural employment increase, the largest shares were from agriculture (94% during 1971-76 and 59% during 1976-82). During 1976-82, however, rural services accounted for 28% of job expansion. This dominance of agriculture and services in recent years was, of course, accompanied by large increases in underemployment. In 1982, over a fifth of the total work force was unpaid family workers, the large majority (88%) being found in the rural areas. This category of workers grew at an average annual rate of 9.6% during 1976-82, a clear indication of lack of regular jobs for wage and salary workers. 22. By occupation and industry, the evidence clearly points to relatively higher levels of underemployment among the poor than among the non- poor. A general expansion of work opportunities will thus help alleviate poverty. Trends in Real Wages 23. Real wages of skilled and unskilled workers exhibited a declining trend in Metro Manila during 1970-1980. During 1980-83 the trend continued except that workers in some of the skilled occupations appeared to have experienced modest growth in real wages. In agriculture, wage data were available in respect of workers in palay, corn, sugarcane and coconut. In general, real wages declined during 1970-74, increased during 1975-78 and then declined again. 24. The increase in the agricultural wage rate during 1975-78 could be due to various factors. Relatively high rates of agricultural CDP growth occurred during 1975-78. This period also saw a stable overall CDP growth rate, relatively high investment growth rates, the favorable effects of the commodity boom of the earlier years, and special rural infrastructure programs. The incidence of underemployment was low during 1977 and 1978, an indication of a tightening of the rural labor market assisted by the movement of labor from rural to urban areas. Finally, during 1975-78 inflation rates were low. All these factors helped to bring about an increase in the agricultural real wages during 1975-78. The economic environment deteriorated in subsequent years, slack developed in the labor demand, and inflation started to erode the real value of the nominal wage rates. 25. In the urban areas, employment expanded at 4.9% during 1971-76 and 2.8% during 1976-82 but real wages either declined or stagnated. The healthy overall economic growth environment of the first period and part of the second was counter-balanced by excessive labor supply (indicated by the 70 to 90% service sector share in employment growth) and the result was only a modest increase in the money wage. Thus when the inflation rate was high during 1971-74, real wages declined. They stagnated when the inflation rate was low during 1975-78. The deterioration in demand for Labor and the high inflation rates of the post-1978 period led to declining real wages. 26. Given the very high inflation rate and the negative rate of economic growth of 1984, real wages have probably dropped significantly in both rural and urban areas in the past year. Sustained economic growth and at most modest inflation over relatively long periods of time can provide a basis for growth in real wages. The process can be greatly aided by a slower labor force expansion. However, action on the population front today will produce benefits only some 15 to 20 years hence. For 1985-2000, however, the employment problem is of a serious magnitude, with annual additions to the labor force amounting to almost a million people. B. Recommendations Monitoring Poverty 27. At present, little direct attention is paid by the Government to monitoring poverty or the impact of development programs on poverty and employment. Because of the importance of this issue in future development, it is recommended that poverty monitoring should be undertaken on a formal and regular basis by the Covernment. To facilitate this, urgent attention is needed to streamline and improve data collection. Data Collection and Dissemination 28. Household Incomes and Expenditures: The practice of holding more than two quarterly ISH resulted in the accumulation of a backlog of unprocessed data. Both from the view point of quick data processing and to affect improvements in the survey operations, only one or at most two ISH per year should be conducted. Once every two or three years, the ISH should be expanded to collect family income and expenditure data. Tabulations should be prepared by per capita income and per capita expenditure classes in addition to tabulations by family income and family expenditure classes. 29. Data from the Agricultural Census should be processed quickly and MCSO should publish important national and regional tables on a comparable basis for 1971 and 1980. 30. Labor Force and Employment: Each year in the ISH, labor force and employment data must be collected as at present. Establishment surveys should concentrate on obtaining data on output and value added rather than on employment and earnings. The Ministry of Labor and Employment should have full access to ISH labor force and employment data and should discontinue its household surveys. BAEcon should conduct, once in two years if not annually, comprehensive socioeconomic surveys, rather than the present rice and corn surveys, to obtain detailed information on the employment and related economic conditions in the rural economy. 31. Wage Rates and Prices: The Ministry of Labor and Employment should collect quarterly data on wage races in non-agricultural sectors in key cities. The proposed BAEcon rural surveys would provide rural wage rate data with an expanded coverage. NCSO should prepare and publish rural and urban consumer price indices for the general population as well as low income groups. 32. Statistical Organization: The problems of data collection and dissemination are not limited to household expenditure, employment and wages. There is a need to review the entire statistical system to suggest ways to ensure the timely collection and dissemination of statistical information without duplication of efforts. Such a review should address all relevant aspects including staffing, salaries and budget. THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES I. INTRODUCTION (i) The Background The Bank's poverty report 1 on the Philippines was issued in 1980. It contained poverty incidence data based on a 1975 family income and expenditure survey (FIES) and detailed analysis of poverty based on a 1971 FIES. The present information on the poverty problem is thus quite dated, yet the problems of unemployment, poverty and living standards remain crucial. While no data of the FIES type have been collected during the past several years, some preliminary tabulations on household income distribution have been published for the period 1980-83 on the basis of "integrated surveys of households" launched in 1976. Published data from these surveys contain distributions of families by family income class for the different regions. These surveys do not provide data on family expenditures, and hence cannot yield results comparable with the FIES 1975 and FIES 1971. One can, however, try to evaluate the family income distributions from the surveys of 1980-83 and then make some inferences on recent trends in poverty. The Bank's 1980 poverty report, in the chapter on employment and wages, pointed out that the period 1957-78 was marked by a tendency for the real wages to decline. This was mostly attributed to the relatively high growth rate of Labor force (about 3% p.a.). The report also pointed out that in most sectors, value added per worker stagnated. Based on these two observations, it has drawn two implications: steps should be taken to slow labor supply growth and policies should be geared to increase employment and productivity. On the employment front, the report thus took an essential first step, namely, to signal the trends of very high labor force growth, stagnating productivity and declining real wages. These trends need to be updated, especially at the present time when the economy is going through a major crisis and the adjustment and recovery programs will have important implications to employment, productivity and real wages. (ii) Objectives of the Present Report Against the background described in the earlier section a reconnais- sance mission visited the Philippines in September 1984. This report which is the outcome of the mission has the following objectives: (a) evaluation of the currently available post-1975 data on employment, wages and income distribution, with respect to coverage and quality of data; 1/ World Bank, Aspects of Poverty in the Philippines (1980). -2- (b) a brief analysis of available data focusing on trends in employment/ unemployment, real wages, income distribution and numbers below the poverty line, pointing out significant changes, if any, in the situation since 1975; and (c) recommendations for improving the quality and coverage of the data. II. FAMILY INCOME AND EXPENDITURE DATA A. Sources of Data (i) Income/Expenditure Data Prior to 1976 In 1957, 1961, 1965, 1971 and 1975 the National Census and Statisti7al Office (NCSO) conducted family income and expenditure surveys (FIES).1 The survey of 1971 obtained data for the 12 month period of May 1970 to April 1971 and was conducted in May 1971. Income data for 1975 were available from the Integrated Census of the Population un ertaken by NCSO and expenditure data from the Survey on Family Expenditures.- The income data in 1975 were obtained in December 1975, the questions referred to annual income and the int?me questions were less definite and detailed in comparison to the 1971 FIES.- Expenditure data for 1975 were actually collected some five months after the income survey, thus with a relatively long recall period. (ii) Data Sources and Data Availability, 1976-83 After 1975, NCSO did not conduct a comprehensive family income and expenditure survey. Instead, starting in 1976 an integrated household survey (ISH) was held each year in selected quarters.- Income data from the surveys of 1976 through 1978 have not yet been cleaned; data for 1979 are being cleaned and data for 1980 through 1983 are cleaned but not yet processed (as of October 1984). At present, data are available on the aggregate distri- butions for 1980 through 1983. In addition, the Bank mission obtained income 2/ For a brief description of these and other related surveys, see Terasaki (May, 1977). 31 These two surveys correspond to the former FIES, but there are two dif- ferences compared to the earlier surveys: (1) field operations on income and expenditure were undertaken separately, and (2) expenditure survey used a sub-sample of the integrated census. 4/ See World Bank Poverty Report (1980). 51 1976 - 3rd Quarter, 1977 - 1st, 3rd and 4th quarters, 1978 - all four quarters, 1979 - all four quarters, 1980, 1981 and 1982: 3rd and 4th quarters, and 1983: 1st, 3rd and 4th quarters. -3- data on about 3,300 families for 1979 as well as 1983 thus providing a basis for finding out what happened to the income of each family during 1979-83. B. Limitations of the ISH (i) The Income Conce2t in ISH The ISH income concept is quite comprehensive. Total family income includes primary income (earnings) and receipts from other sources received by all family members during the reference quarter as participants in any econ- omic activity (either as laborer, proprietor, capitalist or entrepreneur) or as recipients of transfers, pensions, grants, etc. Primary income includes wages and salaries, commissions, tips, bonuses, family and clothing allow- ances, and net receipts/profits derived from the operation of family-oriented enterprises/activities and the practice of a profession or trade. Imputed rental values of owner-occupied houses, interests, royalties, and profits from sale of stocks and bonds, rentals, including landowners' share of agricultural products, pensions, supports, gifts, winnings, proceeds from insurance, etc., constitute the receipts from other sources. Also included in the estimated total family income are the receipts from family sustenance activities which are not considered as family-operated activities (see Annex I for more details). (ii) Underreporting of Income (a) ISH-National Accounts Comparison: The fact that the income concept is quite comprehensive does not automatically ensure a high degree of accuracy in reporting by respondents. There are hardly any igernal consistency checks that can be built into the various income questions.- The result is, more often than not, a high degree of income underreporting deliberately as well as because of recall lapse and absence of accounting and bookkeeping practice by most families. In addition, when income data is gathered as part of a large questionnaire covering various other socio-economic aspects, the income question may not receive adequate attention by the respondent as well as the enumerator. One method employed to find the degree of undercoverage in ISH income is the comparison of ISH and national accounts data. For the period 1980 through 1983, the comparison indicated that ISH income fell short of personal income from national accounts o the extent of half to one-third, about the same as the pre-1980 record.' In contrast to this high apparent 6/ For instance, when detailed consumer expenditure data are also collected, a check can be made by comparing the two data sets. The respondents can be asked to explain the sources of other revenue if expenditure is high relative to income. Where income is quite high relative to expenditure, questions can be raised on the use of savings. 7/ See also the works of Mangahas cited in the references. -4- undercoverage of income, the degree of undercoverage of consumption expendi- ture in the family expenditure surveys relative to personal consumpti in national accounts varied from 6% to 30% in the pre-1980 FIES surveys. These results are not very different from the situation in other developing conotries. The fact that a comparison between the survey estimates of family income and personal income from national accounts provides a basis to assess the degree of underestimation in the surveys does not mean that an across-the- board adjustment of survey data is desirable and can be defended, for, there is no way one can safely make the heroic assumption of an equal degree of under-reporting by one and all of the thousands of families in the survey.- In any case, while such an adjustment will affut the poverty incidence numbers based on an absolute poverty cut-off,-- it will not make any difference to the income distribution. Table 1: COMPARISON OF NATIONAL ACCOUNTS (NA) AND ISH INCOME ESTIMATES, 1980-1983 1980 1981 1982 1983 N.A. Personal Income (P Bln) 200.5 231.0 256.7 290.3 ISH family income (P Bin) a/ 108.5 138.4 169.9 191.6 ISH-NA Income Ratio 0.54 0.60 0.66 0.66 Sources: NEDA and NCSO a/ Based on 3rd and 4th quarter data. 8/ It is quite plausible that personal consumption is understated in the national accounts since in all likelihood private investment has been overstated as evidenced by the very high ICOR's in recent years. 9/ For an example of such adjustments, see Daly et al, Income Distribution, Food Consumption and Poverty, Staff Paper No. 7, Ministry of Agriculture of the Philippines, June 1981. 10/ An across-the-board adjustment of the incomes of all famiLies would reduce, for instance, the poverty incidence of 59% in 1971 (based on the poverty line in the 1980 Bank report and the family income distribution) to 35.2%. Similarly, the poverty incidence in 1975 will decrease from 64.4% in 1975 to 23.3%. The 1971-75 poverty trend stands reversed because of a relatively higher degree of undercoverage of income in the 1975 survey. -5- Table 2: COMPARISON OF NATIONAL ACCOUNTS (NA) AND FIES INCOME AND CONSUMPTION EXPENDITURE ESTIMATES, 1961-1975 1961 1965 1971 1975 Ratio of FEIS to NA Income 0.65 0.70 0.68 0.53 Ratio of FIES to NA Expenditure 0.71 0.85 0.94 0.70 Source: World Bank, Aspects of Poverty in the Philippines, Vol. II, p. 115. (b) A Comparison of NCSO and World Bank Income Distribution Data for Metro Manila for First Quarter 1983: Metro Manila income data for the first quarter are available not only from the NCSO, but also from a household survey conducted by the University of the Philippines researchers under the auspices of the Bank's urban research group. From the NCSO data, the estimated average family income was P 3,860 per month. The World Bank survey estimate was P 4,184, about 8% more than the NCSO estimate. Another implication of the two data sets also may be noted. Based on a Metro Manila pov y line of P 695 per month per family for the first quarter of 1983,- the NCSO distribution indicated that 7% of the families were below the poverty line, whereas the World Bank data indicated a 4% poverty incidence. The implication is an under reporting of incomes in the NCSO survey in the very low income groups. (For instance, below a family income of P 1,000/month, NCSO indicated 18% of the families as against 10.8% in the World Bank survey.) (iii) Differential Implications of Family Income and Family Expenditures The Bank report issued in 1980 adopted a poverty line of 912 per capita or P 5,147 per family for the year 1975. When the family poverty line was applied respectively to the family expenditure and income distributions, the percentages of families below the cut-off were 53.2% and 64.4%. This is because of the phenomenon that family income estimates were relatively more downward biased than expenditure estimates. Hence, to find the extent of poverty incidence, one prefers data on family expenditure distributions to family income distributions. The ideal data, however, are distributions of families by per capita expenditure classes. Such data are to a large extent, free from the bias in poverty incidence measures based on family expenditure distributions, where the bias arises due to the high degree of positive association between family expenditure and family size. Thus, when a per capita poverty line is defined and converted to a per family poverty line based on the average family size, and when the per family poverty line is applied to a family expenditure distribution to measure the incidence of poverty, the measure is biased on two 11/ For the methodology of this poverty line, see Annex 6 of this report. counts. First, there will be some families below the cut-off with family size less than the average and second, there will be some families above the cut- off with famiLy size larger than the average. As a result, some famili H below the cut-off and some above it, may be wrongly identified as poor - and the net effect may be upward or downward bias on overall poverty incidence. (iv) Ranking by Family Income, Family Expenditure and per Capita Expenditure: Illustration of Quantitative Differences (a) 1971-75 National Data: The data in Table 3 illustrate the degree of difference between the poverty incidence numbers when a given poverty line is applied to family expenditure and income distributions. 12/ Ranking by family (or household) expenditure and ranking by per capita expenditure will not identify the same families below and above the appropriate poverty cut-offs. For instance, in the South Asian data sets, covering the period 1968 to 1975, Visaria found that only 14 to 28% of the households were assigned to the same deciles when ranked by per- household and per-person criteria. Large families tend to have large total expenditures and smaller average expenditure and thus the particular households which get classified as poor will be almost entirely different between the two types of rankings. See Visaria, P., Poverty and Living Standards in Asia, LSMS Working Paper No. 2, World Bank, 1980. The disparities in the results of income distribution analysis from household income and per capita income rankings has been pointed out in Cautam Datta and Jacob Meerman, Household Income or Household Income Per Capita in Welfare Comparisons, W.B. Staff Working Papers, No. 378, World Bank, 1980. For example, the paper shows that the distribution of benefits from public education is very inegalitarian if household income is used as a basis for analysis. The reverse occurs when per capita income is used. In a recent paper, Lipton had reviewed the evidence and concluded that the ranking by outlay per household should be considered useless for poverty studies and that the ranking by outlay per person should be clearly preferred. See Michael Lipton, Demography and Poverty, World Bank Staff Working Papers, No. 623, World Bank, 1983. -7- Table 3: POVERTY ESTIMATES FROM INCOME AND EXPENDITURE DISTRIBUTIONS Percent below poverty threshold Difference Source and Poverty Line Year A B B-A World Bank Report a/ Threshold 1971 44.9 59.0 14.1 P 500/annum/capita in 1971 1975 53.2 64.4 11.2 Abrera study a/ Food Threshold 1965 34.0 48.3 14.3 P 520/annum7capita in 1971 1971 41.3 56.9 15.6 Abrera Study a/ Total 1965 70.1 75.5 5.4 Threshold: P 877/annum/capita in 1971 1971 69.5 77.7 8.2 al See Annex 5. A: Based on expenditure B: Based on Income The differences in the poverty incidence rates based on income and expenditure data tend to be relatively high in the middle of the income distribution, tapering off towards the end. There is no way one can derive an adjustment formula to proceed from one distribution to the other. The differences, how- ever, could be quite substantial and significant. Clearly, one cannot combine poverty measures from expenditure and income distributions and assess trends. One has to consider either of them only, but, with the full under- standing that expenditures tend to be better covered than incomes. (b) World Bank Data for Metro Manila for First Quarter 1983: The ready availability of the data set at the Bank helps to assess the differences in poverty incidence based on distributions of families by family income class, by family expenditure class and by per capita expenditure class. The poverty line per family in the First Quarter of 1983 for Metro Manila was P 695 per month (see Annex 6). World Bank data on the distribution of families by family income class indicate a poverty incidence of 4%. The family expen- diture distribution, on the other hand, indicates a poverty incidence level of 7.5%. Next, when the poverty line is taken on a per capita basis, (P 131), poverty incidence amounted to 10.6%, which is, of course, a relatively more realistic figure. (c) Concluding Observation: Comparative poverty incidence estimates obtained from family income and family expenditure distributions of NCSO indi- cated that the estimates from the income distributions were smaller than those from expenditure distributions. This will perhaps be the case as long as incomes are severely under reported relative to expenditures; but, even this cannot be taken for granted as evidenced in regard to Metro Manila. A more -8- important point, however, is the invalidity of estimates based on the ranking of families by family incomes and family expenditures. From the World Bank data on Metro Manila on the average family size by family (monthly) expenditure class it was found that from "less than P-167" up to P-2,500, the average family size increased very rapidly from 2 to about 7. From P 2,500 to "P 16,664 and over," it increases rather slowly from 7 to 8.3. The poor families would then appear to be those with the small average family sizes. On the other hand, when households are ranked by per capita expenditure, the household size declines as per capita expenditure increases and the poor families are those with the highest average household size, a result almost diametrically opposed to the one produced by a ranking of families by family expenditure. The conclusion is that distributions by per capita incomes and expenditures are the right data set for poverty studies. III. TRENDS IN POVERTY AND INEQUALITY Note on the Validity of the Results. As stated in Section II the data available for recent years (1980-83) are distributions of families by family income class (ISH data), and not by per capita income class. Similarly the data for 1975 and earlier years wer 3 ased on rankings of families either by family income or family expenditure.- The observations in Sections A(i) and A(ii) below are thus of limited significance in regard to poverty incidence data for any particular year. They could, however, be used as indicators of broad trends over time. In addition, sections A(ii), and A(iii) below provide an analysis based on per capita income distributions drawn from the ISH of 1983 (Third Quarter). A. Trends in Poverty Incidence (i) Poverty Trends, 1965-75 A study by Abreraw found an increase in poverty incidence from 34% to 48% from 1965 to 1971 when the poverty threshold was based only on food adequacy, but found poverty incidence to be about constant at a high 70% when a much higher threshold was considered. A high threshold naturally inflates poverty incidence. On the one hand, it looks odd to infer a much larger poverty incidence in the Philippines than in, say, India where about 40 to 50% of the people have been found to be in absolute poverty. On the other hand, as the vast majority (such as 70% or 80%) get classified as poor, such a measure is unlikely to be useful for policy decisions and actions with a sharp poverty focus. 13/ So far, NCSO has never published data by per capita expenditure or per capita income. However, the Bank poverty report (1980) obtained bivariate distributions of families by family expenditure and family size, thus overcoming the limitations of not having per capita expenditure distributions. 14/ See Abrera (1976). The Bank's 1980 poverty report used a poverty line which roughly identified the bottom 40% in 1971 as belonging to the poverty group. Between 1971 and 1975 quite apart from a slight increase in the overall and rural incidences of poverty there was one notable trend in the data in the report. It was the jump in the urban poverty incidence from 29% in 1971 to 40% in 1975. Such a major change in a relatively short period did not occur in the evidence presented by Abrera for 1965 and 1971. (ii) Poverty Trends 1971-83 (a) Note on the Poverty Line: For the purposes of the analysis in this report, relatively conservative poverty lines were used. In 1983 the annual poverty Lines were P 7,363 for a rural family and P 10,584 for an urban family. These amounts of purchasing power were just enough to provide the needed food and non-food items (See Annex 6). Special poverty level food and non-food weights were used to deflate the 1983 poverty lines and obtain the lines for earlier years. (b) Rural-Urban Poverty Trends: Several scholars have noted the inade- quacy of the incom o5 ncept and other limitations of the 1975 data compared to the data for 1971- Whether for this reason or not, one can see the conspicuous increase in the rural and urban poverty incidences in 1975 com- pared to 1971 (Table 4). If one ignores the 1975 data in Table 4, and consid- ers 1971 and early 1980s, a modest and plausible decline in rural and urban poverty incidence is noticed in the early 1980s compared to 1971. There was thus some trickle down of growth to the lower income groups. The 1980-83 "fluctuations" are not worthy of any special interpretation since sampling errors, differences in the degree of coverage etc., can cause such minor variations. There is yet another interesting way to pick a data set from Table 4. This is by looking at broad data comparability of the income from the sample surveys and the income from national accounts. The last column of the table provides the ratio of the income estimates from the two sources. In 1971 and 1982, personal income in the household surveys amounted to a little over two-thirds of the personal income from the national accounts. The com- parison of 1971 and 1982 confirms the reduction in rural and urban poverty incidences in the early 1980s compared to 1971. 15/ See, for instance, the Bank's Poverty Report (1980) for a concise summary of the differences. A major problem was that the number of households estimated by the 1975 FIES was somewhat low, 6.8 million as against 7.2 million covered by the 1975 integrated census. See also Annex 2 for limitations of income data. - 10 - Table 4: POVERTY LINES AND POVERTY INCIDENCE, 1971-1983 a/ Poverty lines Poverty incidence (per family) (% of familiez below Survey to Year b/ pesos/annum poverty line) c/ NA income Rural Urban Rural Urban Rural and ratio Urban 1971 2,341 2,989 57.4 35.1 50.7 0.68 1975 4,570 5,514 63.8 53.2 60.6 0.53 1980 5,537 7,758 46.7 28.5 40.8 0.55 1981 6,384 8,965 47.8 28.4 41.4 0.58 1982 6,892 9,790 42.2 24.6 36.6 0.67 1983 7,363 10,584 45.4 26.0 39.0 0.60 /a See Annex 6 for details of methodology relating to the poverty lines. 7b 1980-83 third quarter data. 7- Extreme poverty may be denoted with reference to an income level that is just enough to buy the food that can provide the recommended calories. This level of income equals the value of 16kg of rice per head per month (see Annex 6). The incidence of extreme poverty declined from 24.9% in 1971 to 20.1% in 1980 and 15.1% in 1983 in the rural areas and from 7.9% in 1971 to 6.1% in 1980 and 5.9% in 1983 in the urban areas. The number of poor families in the country as a whole increased from 3.3 million in 1971 to 3.6 million in 1980-83. In the rural areas, the number increased from 2.5 milion to 2.8 million. In the uban areas, the number remained constant at 0.8 million. In an earlier para, it was noted that distributions of families by per capita income class (in contrast to per family income class) were obtained for the third quarter of 1983. Based on these distributions, rural and urban poverty incidences were 43% and 20% respectively, as against 45% and 26% noted in Table 4. The relatively large differential in regard to the urban areas must be due to the relative predominance of young families, with a small family size who get classified as poor when poverty incidence is measured on the basis of the distribution of families by family income class. Thus, the "true" poverty incidences, especially in the urban areas, will be less than the rates indicated in Table 4. However, it is expected that the conclusion on the general decline in poverty incidence in the early 1980's compared to the early 1970's is still valid. (c) Regional Variation in Poverty in 1983: Poverty lines and the rates of poverty incidence for the thirteen administrative regions are given in Table 13 of Annex 7 based on the rankings of families by per capita income. Metro Manila had the least poverty incidence of 11.2% in 1983. The highest rates of poverty incidence (about 50%) were found in Western and Central - 11 - Visayas. Relatively high rates (around 40%) were also observed in 11ocos, Cagayan Valley, Bicol, Western Mindanao And Northern Mindanao. Central Luzon, Southern Tagalog, Southern Mindanao and Central Mindanao had rates around 30%. While these trends are broadly consistent with general perceptions, the incomes data could be subject to differential degrees of under-reporting and this could affect the regional comparison. (d) Poverty Incidence by Main Source of Livelihood, 1982 and 1983: Tabulations on the distribution of families by per capita income classes were specially prepared by NCSO for the rural and urban areas and by main source of livelihood for the third quarters of 1982 and 1983. From these distributions, poverty incidence rates were computed by main source of livelihood. The results are given in Table 5. In the urban areas, where a third of the Philippine families live, the incidence of poverty was 19.2% in 1982 and 19.7% in 1983. The distribution of families by main source of livelihood changed little in 1983 compared to 1982. A little over half (54%) of the urban families derived their main livelihood from the non-agricultural sector by providing skilled and unskilled Labor. Poverty incidence among these families was relatively low at 16%. The next numerically large group comprising a fifth of the urban families derived income predominantly from dividends, rents and pensions. This group also had a relatively low poverty incidence (18.5%). Among the families (18% of the total), who were own-account workers in non-agricultural sectors, poverty incidence varied from 17% to 24%. Farming, fishing and livestock in the urban areas were characterized by relatively high rates of poverty incidence (between 45% and 50%); but only 8% of the total families depended on these sectors. Among the rural areas, where two thirds of the country's families lived, significant annual shifts in the patterns of livelihood occur. For instance, for 9.9% of the country's total families (or 15% of the total rural families) in 1982, the main source of income was palay farming. In 1983, only 5.2% of the total families (or 7.7% of all rural families) belonged to that group. On the other hand, families deriving major part of income from coconut increased from 3.7% of total families in 1982 to 4.9% of total families in 1983. A number of factors such as yield and price changes by crop determine whether the main source of income of a family stays the same or changes over short periods of time. In view of this, it is not possible to identify a fixed set of families as palay or coconut families, since no such clear-cut set exists year after year. In spite of this limitation, some labels are quite useful in order to associate source of livelihood and level of poverty incidence. The data in Table 5 indicate that relatively high poverty inci- dence (50% to 60%) occurs among the families dependent on corn, coconut, other crops (notably sugarcane) and fishing and among agricultural labor families. Even among the families dependent on manufacturing and those deriving liveli- hood form dividends, rents and pensions, the poverty incidence rates were close to 40% to 50%. Non-agricultural labor was a source of livelihood with relatively low poverty incidence (26%). It does seem that sustained poverty alleviation in the rural areas requires major structural changes in the rural economy, bringing about shifts in sources of livelihood from agriculture to nonagriculture. -12- Table 5: POVERTY INCIDENCE BY MAIN SOURCE OF INCOME Third quarter Third quarter of 1982 of 1983 Poverty % of Poverty % of Region and main incidence total incidence total source of income (1) families (%) families Urban 19.2 32.9 19.7 33.0 Farming 43.1 2.1 45.7 1.9 Fishing, livestock, forestry 43.8 0.7 48.9 0.7 Manufacturing 17.5 1.0 23.8 0.9 Wholesale and retail trade 20.7 3.3 20.1 3.7 Other services 21.3 1.7 17.4 1.5 Dividends, rents and pensions 18.6 6.4 18.5 6.6 Non-agricultural labor 15.4 17.7 16.3 17.7 Rural 42.4 67.1 43.3 67.0 Farming - Palay 36.2 9.9 33.5 5.2 - Corn 58.8 6.3 62.5 5.8 - Coconut 56.0 3.7 41.0 4.9 - Other crops 53.4 4.1 52.6 3.8 Agricultural labor 53.8 5.9 56.5 7.1 Fishing 51.7 3.6 52.1 3.6 Livestock, poultry, forestry 41.0 2.0 46.8 2.5 Manufacturing 37.3 1.1 51.1 1.2 Wholesale and retail trade 34.2 3.1 35.7 3.5 Other services 31.0 1.4 28.9 1.3 Dividends, rents and pensions 48.7 11.6 50.0 12.1 Non-agricultural labor 24.2 14.4 25.8 16.0 (iii) Poverty, 1979 and 1983: A Panel Study of 3,294 Households (a) The Data: A household-by-household comparison of the filled-in questionnaires for the third quarters of 1979 and 1983 was undertaken by the NCSO staff and a total of 3,294 households, each with the same name and Location of the head of household in 1979 and 1983, were identified. For each household, household size, household income, head's occupation and head's education were analyzed for 1979 and 1983. The results are reported below. (b) Data are not a Representative Sample: The average household sizes in 1983 should be around 5.2 for rural and 6.3 for urban. The panel showed 5.6 for rural and 5.7 for urban. In the urban areas, relatively larger households were left out of the panel. The average family income in the third quarter of 1983 from the total sample was P 3,090 for rural and P 7,983 for urban. The panel estimates were P 3,224 for rural and P 5,796 for urban. Clearly, the urban data were least representative of the national sample, - 13 - biased downward both in relation to family size and average income and hence would over-estimate urban poverty. Despite this limitation, some insights can still be gained on the progress of the group of families over a period of four years. (c) Overall Poverty Incidence and Mobility from Poor to Non-Poor: Among the families under consideration, poverty incidence declined marginally from 45% in 1979 to 42% in 1983 in the rural areas, and from about 40% to 35% in the urban areas (Table 6). Further analysis of the poor and non-poor families revealed a very important feature. Among the urban poor in 1983, 65% were poor in 1979 also, but, as many as 35% were not poor in 1979. Similarly, in rural areas, among the poor in 1983, 64% were poor in 1979 also while the remainder of 36% were not poor in 1979. Some 27% of the rural families and 23% of the urban families were poor in 1979 as well as 1983. Nationally, poverty incidence declined from 43% in 1979 to 40% in 1983 among the families under discussion. Of the poor in 1979, about 40% were able to raise themselves from poverty by 1983, but, 60% continued to remain poor. At the same time, 25% of the nonpoor in 1979 fell below the poverty line by 1983 whereas 75% continued to be nonpoor. Thus, while the overall poverty incidence changed only marginally between 1979 and 1983, there was substantial mobility between the poor and the nonpoor group. Incomes of different families increased by different magnitudes during 1979-83. Among the urban and rural families respectively, 27% and 30% had zero or negative nominal income growth and a further 21% each had positive income growth below the inflation rate (58% from third quarter 1979 to third quarter 1983). Thus, a half of thl6Orban and a half of the rural families had experienced real income increases- Such volatility of income changes goes hand in hand with a high degree of mobility from poor to non-poor and non-poor to poor categories for a significant proportion of families. In such circum- stances, reliable poverty analysis requires data on distribution of families by per capita expenditures since consumption expenditure is a reasonable proxy for "permanent" income especially for the low income groups. Despite this valid observation, a hard-core of rural and urban poor, those who were poor in 1979 as well as 1983, could be identified from the per capita income distribu- tions and they made up for 27% of the families in the rural areas and 23% of the families in the urban areas. Further data on household characteristics are available at the NCSO and tabulation and analyses work on them would be useful for policy purposes. 16/The Development Academy of the Philippines collected data on socio-econ- omic status and perceptions from 500 heads of households in 1981 (March- May). Considering the past three years (that is, 1978-81), 42% felt that their Lives improved, while 31% responded that their lives deteriorated. See Mangahas (December 1981). These findings are comparable to the ones noted above. - 14 - Table 6: SHIFTS IN POVERTY GROUP, 1979-1983, URBAN AND RURAL AREAS - Poor Families Nonpoor Families Status in in 1983 in 1983 Total Region 1979 No ('000) Z No ('000) % No ('000) Z Urban Poor 237 22.7 177 16.9 414 39.6 Nonpoor 129 12.3 502 48.1 631 60.4 Total 366 35.0 679 65.0 1,045 100.0 Rural Poor 600 26.7 418 18.6 1,018 45.3 Nonpoor 338 15.0 893 39.7 1,231 54.7 Total 938 41.7 1,311 58.3 2,249 100.0 Source: Panel of 3,294 families from ISH, 3rd Quarter Surveys (Data processed at the World Bank). (d) Poverty Incidence by Family Head's Educational Level: Among the families headed by persons with elementary education, regardless of the loca- tion (urban or rural), poverty incidence was 52% in 1979 and 47%-49% in 1983. Urban-rural differentials begin to emerge as one moves up from high school to college and higher levels, with urban poverty incidence lower than rural poverty incidence by level of education. Among the families headed by high school and college qualified persons, poverty incidence rates respec- tively were 37% and 23% in 1983 in the rural areas and 28% and 14% in the urban areas. These are fairly high rates of poverty incidence and could be attributed in part to the possible excess supply of such graduates relative to needs. A related issue is the quality of education. (iv) The Poverty - Employment Linkage (a) A Note on the Data: The ISH tables provide data on earnin s of the wage and salary workers and the net income of own account workers.--- The data are provided by occupation, industry, education and number of days worked in the refernece quarter. Such data for 1976, 1977 and 1978 are available in the ISH Bulletins on Labor Force. Data for 1979-83 are not yet fully tabula- ted, but, when they become available, some insight can be obtained on the relative positions of different occupational groups, sectors and educational 171 Of the three categories of "employment status," the third, namely, unpaid family workers do not have any explicit earnings of their own. - 15 - categories. In this section, the linkages between poverty and emyhyment are analysed on the basis of the data for the fourth quarter of 1978.- (b) From per Family to per Earner Poverty Lines: The data on earnings are for 1978. Hence the per family poverty lines for the third quarter of 1983 in Annex 6 were first extrapolated backwards to 1978 fourth quarter, by using rural and urban price indices. To convert the per-family9yoverty lines to per-earner, an average of 2 earners per family was assumed- The earnings thresholds that signify poverty in the fourth quarter of 1978 were P 750 per urban and P 543 for rural. (c) Poverty among Earners by Urban and Rural: Among the urban erners, 34% were poor in 1978. The proportion was 40% in the rural areas. The urban poverty incidence among families would be less than the 34% indicated by pov- erty among the earners because incomes from interest, dividends, rents and transfers, which are not included in the earnings could be relatively more significant in the urban areas than in the rural areas and also possibly because of relatively higher under-reporting of earnings. Rural poverty inci- dence among families may be more than the incidence among earners because of the possible existence of a relatively large number of families living on meagre "other income" without participating in employment and therefore with- out any earning members. (d) Poverty among Earners by Occupation and Industry. In the urban areas, poverty incidence was 42% among the sales workers and 41% among the service workers. These two categories make up for a third of the urban earners. Another third of the earners were the occupational group of production workers, transport operators and laborers and 30% of this group had earnings below the threshold. In the rural areas, sales workers, service workers and agricultural workers respectively had poverty incidence rates of 45%, 49% and 43%. Data by sector of activity point out relatively high poverty incidence in the services sectors (distributive trade, community services and personal services) in urban areas and in agriculture, manufactur- ing and distributive trade in the rural areas. In fact, rural manufacturing had a poverty incidence of 52% and distributive trade had 45% as against the 42% in agriculture. Perhaps, in the contemporary rural setting, landlessness drives people into some petty manufacturing and trade activity with relatively high poverty incidence. The urban areas also attract the landless where since "petty manufacturing" can be very expensive to launch, the avenues open are distributive trade and personal services. 18/ NCSO, Integrated Survey of Households Bulletin: Labor Force, Third Quarter 1978, Fourth Quarter 1978, Manila, September 1979. 191 The average number of employed persons per family was about 2 during 1980-83. - 16 - B. Trends in Inequality (i) Family Income Inequality (a) The 1956-71 Trend. The income shares of the bottom 40% were 12.6% in 1956, 12.1% in 1961, 11.5% in 1965 and 12.0% in 1971. 20e overall Gini coefficients were respectively 0.48, 0.50, 0. J,and 0.49.- The data for the different years were not strictly comparable,- and hence one might at best state that there was no strong evidence to conclude that inequality narrowed or widened during 1956-71. (b) The 1971-83 Trend. Data on quintile shares for 1971, 1975 and 1980-83 are given in Table 7. The first impression one gets is that the data for 1975 are out of tune with the rest since the degree of income under- coverage was the highest and since the income shares of the bottom 20% and the bottom 40% were relatively very high. Thus one might wish to ignore the 1975 data in the inter-temporal comparison. Second, the data for the period 1980-83 could be treated as a composite, since the quintile shares mostly exhibit fluctuations over a stable trend. Table 7: QUINTILE INCOME SHARES, 1971-1983 a/ Shares in total income of families belonging to Ratio of ISH personal Year Bottom Bottom Bottom Bottom Top income to National 20% 40% 60% 80% 20% Accounts estimate 1971 3.7 12.0 25.6 47.1 52.9 0.68 [1975] 5.1 14.2 26.6 45.0 55.0 0.53 3rd Quarter 1980 2.9 9.9 22.0 42.0 58.0 0.55 3rd Quarter 1981 2.8 9.9 21.2 40.8 59.2 0.58 3rd Quarter 1982 3.0 9.4 20.9 40.8 59.2 0.67 3rd Quarter 1983 3.4 9.8 21.3 41.2 58.8 0.60 aI Computed by straightforward interpolation of the cumulative family income distributions. From first and second,' the third point emerges which is based on a comparison of the income shares of 1971 on the one hand and the 1980-83 shares on the other. The income shares of the bottom four quintiles declined and the share of the top 20% increased. One might also note that the relatively poor 20/ See Sta. Romana III (September 1976) 21/ See Terasaki (May 1977). - 17 - economic performance of the 1981-83 period22/ tended to depress the income shares of the middle and upper income groups and had the effect of inflating the shares of the lower income groups. If this is allowed for, the final con- clusion that emerges is a widening of income inequality in the past decade (1971-80) of fairly high rates of economic growth. It was widely acknowLedged that the periQ 1971-80 was especially characterized by rapid agricultural sector growth-- - at an average of about 5% a year - and the Leading sector in this growth was rice, the growth of which was assisted by irrigation, high yielding seeds and credit. The 24/ Government also undertook agrarian reforms in the rice and corn sectors.- It must now be concluded that all these had not in anyf helped to reduce the degree of inequality in the family income distribution.- (c) Growth, Poverty and Inequality. In retrospect, it might be said that the growth record of the 70s in the Philippines was of mixed welfare consequences. To the extent that poverty incidence declined during the period both in the rural areas and the urban areas, growth did trickle down to the lower income groups. However, the trickle down was quite inadequate to improve the income shares of the bottom quintiles of the population. It was so because rapid economic growth over a short period of time can hardly be accompanied by structural changes, productivity improvements and employment shifts that lead to sustained increases in real wages and enhanced share of labor income in national income which bring about a reduction in overall income inequality. (The estimated share of wages and salaries in family income in the Philippines declined from 44% in 1971 to 39% in 1982). (d) Philippines and Other East Asian Countries. Data on quintile shares are available for 1970 or 1971 for Hong Kong, Japan, Korea (Rep.), West Malaysia and Taiwan (1972). In all countries except West Malaysia, the income share of the bottom 20% of the families was between 5.6%-8.8Z. In West Malaysia it was 3.8%, about the same as in the Philippines (3.9%). The share of the top 20% varied from 37.2% in Taiwan to 49% in Hong Kong with Japan and Korea falling in between. Malaysia's top 20% had 56% share and in the 22/ See World Bank, The Philippines: An Agenda for Adjustment and Growth, Report No. 5258-PH, November 1984. GNP growth slackened to 3.4% in 1981 and less than 2% in 1982 and 1983 as against the average of 6% during the 1970s. 23/ See H.T. Oshima (1983). 24/ See World Bank, Aspects of Poverty (1980). As Mangahas (December 1981) pointed out, there was a little of every thing (e.g. land reform, family planning, dispersal of industry, food self-sufficiency, rural credit, price control etc.) that could have led to reduction in inequality and poverty, but the actual outcome was different in regard to inequality. 25/ As data from the 1981 Agricultural Census, now being processed, become available, a full assessment of inequalities within the agricultural sector and trends in inequality could be undertaken. - 18 - Philippines it was 54%. The Philippine and Malaysian distributions are closer to those of the Latin Amercian countries than East Asian countries. It is well known that land reform has played an important role in reducing income inequities in Japan, Korea, and Taiwan. It does not appear to have as yet been the case in the Philippines, despite the well publicized land reform program affecting rice and corn farms. (ii) Rural-Urban Inequality (a) Income and Productivity Differentials. In a recen 6 tudy, NEDA examined the changes in the rural-urban income differential- The rural- urban per capita income ratio was 0.75 in 1975, as obtained from the basic data from the 1975 FIES. For the period 1980-82, NEDA could not directly obtain rural and urban mean income estimates since the ISH tabulations were not available. Instead, NEDA used the NCSO average family incm estimates by region and classified some regions as rural and some as urban.- On this basis, it was found that the rural-urban per capita income ratio declined from 0.75 in 1975 to 0.63 in 1980 and 0.60 in 1982. NCSO had since brought out the ISH household income data by rural and urban. During 1980-83, rural income per head was no more than half of urban per capita income and there was no clear upward/downward tendency in the past four years. This ratio, how- ever, was much less than the 0.75 observed in 1975. The evidence presented here suggests a deterioration of rural-urban differentials in the 80's compared to 1975. Another way of assessing the rural-urban differential is by looking at the proxies of agricultural and non-agricultural incomes, that is, GDP originating in each of the sectors. The productivity differential at current prices deteriorated in the 80s compared to 1975. GDP per person at constant prices increased during 1975-80 in both sectors but stagnated since then in both sectors. Average productivity in agriculture was a constant proportion of the productivity in non-agriculture. The ratio was a low one-third; but, the ratio of sectoral productivities would not be exactly equal to the ratios of sectoral per worker incomes, since the ratio of o2all personal income to total CDP in 1975 was about 0.78 in the Philippines,- and it could be diffezent between the two sectors. Sectoral income-GDP ratios are not available. It may be assumg that in agriculture the ratio will be higher than the national average,- and equals to, say, 0.90. With an agriculture 26/ NEDA, Some Aspects of Rural-Urban Welfare Differential, (August 1984). 27/ The regions with rural population comprising over 70% of the total as rural areas. Regions III, IV, NCR and VII were thus classified as urban and the other nine regions were classified as rural. 28/ 0.78 in 1975 and 0.75, 0.76, 0.75 and 0.76 respectively in the four years 1980 through 1983. 29/ In general, depreciation and other deductions as percent of CDP are likely to be relatively lower in agriculture. - 19 - share in GDP of 0.29, the income-GDP ratio in non-agricultural sector for 1975 can be derived as (0.78-0.29 x 0.90) - 0.71 = 0.73. This leads to a ratio of agric 0 ural income per worker to non-agricultural income per worker of 0.46- On the basis of computations such as this, it was found that the agricultural income per worker at current prices which was almost half of the non-agricultural income per worker in 1975 deteriorated to a third by the early 1980s. It is thus clear that the agriculture to non-agriculture income differentials indeed widened and could have set the basis for the deterioration of the rural-urban income differentials noted earlier. This, however, happened even though the ratio of real (constant price) productivity remained constant. There are thus the following trends that need an explana- tion: (1) the real productivity improved in 1980, compared to 1975 in both sectors and remained stagnant during 1980-83; (2) the real productivity ratio remained constant throughout; and (3) the differentials in sectoral income per worker, and along with it rural-urban income disparity, deteriorated. Leaving aside the 1980-83 productivity stagnation, all other trends imply a deterioration in the agricultural terms of trade. (b) Terms of Trade in Agriculture. Since 1975, agricultural terms of trade deteriorated. The implicit price deflator for value added in agricul- ture, relative to the price index of agricultural inputs declined by 28.8% during 1975-1982. During the same period, the agriculture to non-agriculture relatives of implicit price deflators of gross value added declined by between 25% and 30%. Relative to rural consumer prices, the implicit price deflator for agriculture declined by 33%. Over prolonged periods of modern economic growth, as labor moves from agriculture to industry, as rural real wage rates increase, and as mechanization improves agricultural productivity, one would expect a secular decline in agricultural terms of trade relative to industry. However, the post-1975 reduction in the agricultural terms of trade in the Philippines was not part of any secular trend brought about by the conditions of modern economic growth. Agricultural productivity improved or stagnated, agriculture continued to absorb labor at a positive rate, and rural real wages declined or stagnated. These resulted largely due to Government policies that affected agricultural prices and incomes on the one hand and policies that tended to discourage labor-intensive manufacturing and construction on the other. Prices of the Philippine export commodities also deteriorated compared to import prices. As pointed out by the recent NEDA study 31/ on rural-urban welfare differential, the deterioration in the agricultural terms of trade could be attributed largely to government intervention policies which tended to raise prices of agricultural inputs but kept down the prices of agricultural outputs. Such intervention policies involved the imposition of price controls, tariffs and taxes, and subsidies, which taken together, were biased 30/ (0.37) x (0.90 t 0.73) where 0.37 is the productivity ratio. 31/ NEDA (August 1984). - 20 - against the agriculture sector and in favor of the manufacturing sector. These policies were generally aimed at promoting industrial production, raising government revenues, and protecting consumer welfare. The cost of these policies, however, seems to have been the relative deterioration of farmers' incomes and the hampering of the growth of agricultural production. What is more, the overvalued exchange rate, the relatively low real interest rates, high effective protection and the various other industrial incentives promoted capital intensive production (for domestic use mostly) in the non- agricultural sectors. Large public corporations with access to huge amounts of domestic and external credit exacerbated the process of inefficient technological choices. Labor often found its way into low productivity service occupations rather than into manufacturing and construction. In addition, labor moved from the villages to the towns and into the urban services sectors. IV. TRENDS IN EMPLOYMENT AND WAGES A. Employment Trends (i) The Data Situation (a) Household Surveys. Household surveys in the Philippines conducted at different times in each year since 1 provide labor force data on a continuous and fairly consistent basis Between the old household surveys of 1956-76 and the post-1976 integrated household surveys, there were two principal differences: the first was in the age cut-off of the population covered (10+ in the former surveys and 15+ in the ISH) and the second was the reference period (one week in the former surveys and one quarter in the ISH). The first change is justified on basis that most children in the 10-14 group are outside the labor force and there is no point in going below the legal minimum age of 15 for entry into the labor market. The consequences of the second change are a slight increase in the overall employment rate and a decline in the unemployment rate. Thus, for instance, for the third quarter of 1976, the former survey concepts and ISH concepts respectively provided labor force participation rates of 59.5% and 60.5%, emplGyment rates of 93.7% and 94.8% and unemployment rates of 6.3% and 5.2%. (b) Change in the Reference Period. The reference period change was quite an issue in the academic circles in the Philippines. Most academics contend that the ISH reference period of one quarter reduced the unemployment 32/ See Annex 4 for sources of data. 331 According to the 1975 Integrated Census of Population, of the 5.6 million persons aged 10-14, 0.5 million or about 9% had "gainful occupations." If such trends continued in recent years, the cut-off of 15± might have excluded between 2-3% of the labor force. - 21 - rate to a misleadingly low level.34 While this is true, data on unemployment are hardly useful as indicators of economic well-being of the people in the developing countries where problems of low productivity, lack of skills and lack of full-time work are the principal dimensions of the employment problem. ISH provide data on the number of days worked in the reference quarter, cross- tabulated by earnings, occupation, and education. Improvements in collecting these data by providing intensive training to the enumerators may be a better strategy rather than simply reverting to a one-week reference period. (ii) Employment and Unemployment (a) Labor Force Participation. The overall labor force participation rate has been on an increasing trend. For instance, in 1976-77, the rate was around 60%, whereas in 1982-83 it was around 64%. Much of the increase was due to changes in the age structure in the case of males and increases in age specific participation rates in the case of females. The average of the annual labor force growth rates during 1971-82 was 3.6%. (b) Unemployment and Underemployment. In a country like the Philippines with predominance of rural activities, petty trade, small manufacturing and various types of personal and community services, establishments and households where production takes place are largely informal in organization structure and personnel practices. Firms (and households) cannot afford to come in and go out of business frequently as in the industrialized countries. Unemployment rates, therefore, do not move in unision with the upswings and downswings in economic activity. In fact, if anything, low open unemployment rates indicate the need for relatively more and more people to do some work and help families tide over an economic crisis, as in the case of 1983 (when the lowest unemployment rate of 4.1% was recorded for the fourth quarter - see Table 8). The inadequacy of employment opportunities is better reflected in the lack of full-time work when in fact people like to do more work and when people have full-time work and yet want more work presumably to earn a little more income to meet their family and personal obligations. The ISH classify people who want additional work by the number of days they have worked in the 34/ See for instance the University of the Philippines Workshop Report (June 1984) which says "In 1976 the reference period was changed from the past week to the the past quarter. Moreover employment was expanded to include such activities as home gardening and raising poultry, as long as there was a harvest. These changes effectively increased the numerator of the employment rates. The base of the employment rate, which is the labor force, was also changed. Where before the labor force included the population 10 years old or over, who were either employed or looking for work, the cutoff age was changed to 15 years, thus reducing the size of the denominator. All these changes tend to reduce the unemployment rate figure. What is additionally vexing is that, since data based on the older definition are no longer being collected, researchers no longer have a basis for comparing employment performance before and after 1976." Table 8: POPULATION, LABOR FORCE, EMPLOYMENT, UNEMPLOYMENT AND UNDEREMPLOYMENT, 1971-83 1976 1977 1978 1979 1980 1981 1982 1983 1971/a 1976/a (Qr 3) (Average) (Average) (Average) (Qr 4) (Qr 4) (Qr 4) (Qr 4) Population 15 + ('000) 25,811 31,376 24,837 25,695 26,688 27,918 28,929 30,023 30,978 31,907 Labor force ('000) 12,911 16,244 15,018 15,328 16,502 17,536 18,543 19,005 19,698 20,521 Employed ('000) 12,246 15,427 14,238 14,547 15,652 16,802 17,746 18,017 18,614 19,671 Unemployed ('000) 666 818 780 781 850 734 797 988 1,084 850 Underemployed ('000) 1,862 1,634 3,628 2,897 2,324 n.a. 3,875 5,180 5,445 6,259 Visibly/b ('000) 807 829 2,178 1,588 1,236 n.a. 1,928 2,827 2,813 3,405 Invisibly/e ('000) 1,055 805 1,450 1,309 1,088 n.a. 1,947 2,353 2,632 2,854 L.F.P.R. (%) 50.0 51.8 60.5 59.7 61.8 62.8 64.1 63.3 63.6 64.3 As % of L.F. Underemployment (%) 14.4 10.1 24.2 18.9 14.1 n.a. 20.9 27.3 27.7 30.5 Visibly (%) 6.3 5.1 14.5 10.4 7.5 n.a. 10.4 14.9 14.3 16.6 Invisibly (%) 8.2 5.0 9.7 8.5 6.6 n.a. 10.5 12.4 13.4 13.9 Unemployment (%) 5.2 5.0 5.2 5.1 5.2 4.2 4.3 5.2 5.5 4.1 a Population aged 10+. b Working less than 40 hours a week (60 days/quarter) and wanting additional work. c 'Wbrking 40 hours or more (60 days/quarter) and wanting additional work. Source of Basic Data: NCSO Household Surveys. - 23 - reference quarter. Using a "full-time norm" of 60 days per quarter, those who want additional work (the underemployed) can be classified into the two categories of visibly underemployed (those who worked less than 60 days) and invisibly underemployed (those who worked 60 days or more). The highest underemployment rate was noted in the fourth quarter of 1983, the worst manifestation of the then emerging economic crisis. The rate, yever, was increasing since 1978, after a temporary drop in 1977 and 1978 (c) Employment Structure: Between 1965 and 1971, there was a reduction in the share of agriculture in total employment from about 60% in 1965 to 50% in 1971. During the years 1976-82, the sector share was around 52-53%. Manufacturing share changed very little since 1965 remaining around 19-20%. Services share increased from 22% in 1965 to about 30% in 1971 around which level it continued during 1976-82. There was thus virtually no relative shift from agriculture to industry in the past one and one half decades. (iii) Sources of Employment Expansion (a) Urban and Rural Rates of Employment Growth. Employment growth during three sub-periods was as follows: 1965-70: 3.6% p.a. in urban and 0.6% p.a. in rural, 1971-76: 4.9% p.a. in urban and 4.7% p.a. in rural; and 1976-82: 2.8% in urban and 5.1% in rural. While import substituting industrialization provided the impetus for urban employment growth during 1965-70, the relatively high overall rates of economic growth and particularly the high rate of agricultural sector growth provided the impetus to urban as well as rural employment growth during 1971-76. In the period 1976-82, however, employment grew at an apparently very high rate in rural areas compared to the urban areas which contradicts past experience. Two reasons may be advanced to explain this. First, as noted earlier, the rural (non- agricultural) sector activities were particularly at a peak level during 1976- 78. Second, the lack of momentum in urban labor absorption, especially after 1980, resulted in a swelling of the numbers in rural employment without much increase in full-time employment. (b) Sectoral Contribution to Employment Expansion. During 1965-70, in the urban employment expansion, 64% came from services. It was 73.5% during 1971-76 and 93.2% during 1976-82. Industry failed to absorb the growing urban labor force. Rural employment growth patterns differ between the three sub- periods. During 1965-70, 56.6% of the increase in total employment was from services, 30.5% from manufacturing and 24.5% from construction. Agricultural absorption, however, was negative, and as a result, overall rural employment growth was only 0.6% p.a. The period 1971-76 was characterized by relatively 35/ NEDA, in its recent study of rural-urban differentials pointed out that rural employment expanded at a higher rate during 1976-78 than urban employment contrary to the usual trend. "It was specifically between 1976-78 that the promotion of rural-based non-farm activities was given emphasis, and rural infrastructure, such as rural roads, bridges and rural electrification and water supply were intensified to bolster employment activities in the rural areas." (NEDA, August 1984), p. 41. - 24 - large rate of labor absorption in agriculture. Some 94Z of the total employment expansion was due to agricultural employment increase. During 1976-82, agriculture contributed 58.7% and services contributed 27.6% to rural employment expan n, accompanied, of course, by large increases in underemployment.- (c) Employment Expansion by Class of Worker. The increase in employment during 1976-82 can be further analyzed by class of worker. Of the increase in urban employment 29% were wage and salary workers, 41% own account workers and as much as 30% unpaid family workers. In the rural employment growth, 44% was made up of unpaid family workers and 34% own account workers. In 1982, some 22% of the work force (or a total of 4.1 million) was unpaid family workers (3.6 million in rural areas). The category increased by 77% in six years or an extremely high rate of 9.6% p.a., a clear indication of lack of regular jobs as wage and salary workers and lack of small business opportunities to be own account workers. It also signals the existence of serious underemployment. (d) Employment Expansion by Major Region. Metro Manila, Luzon, Visayas and Mindanao are considered here as the major regions. During 1976-82, Metro Manila accounted for 5% of the total employment increase, even though its contribution to national population increase was around 16%. Luzon and Mindanao shares in employment and population increase were about equal (35% and 34% in Luzon and 30% and 29% in Mindanao). Visayas accounted for about 30% of the employment increase, larger than the 21% share in the population increase. The relatively small share of Metro Manila in employment growth compared to other regions and compared to its own share in population growth would only imply relatively large open unemployment in Manila. This, of course, was validated by an 11% male unemployment rate and 16% female unemployment rate in 1982 in Metro Manila as against less than 2.7% male unemployment rate and 7.5% female unemployment rate in the rest of the Philippines. The relatively low unemployment rates in the regions outside Metro Manila to a large extent mask the high levels of labor underutilization. (e) Education and Occupation. Compared to most middle income countries, the Philippines historically had very high enrollment rates in secondary schools and higher educational institutions. For instance, in 1981, Philippines enrollment in higher educational institutions was 26%, which was surpassed y by four countries out of a group of 60 middle income countries.- Normally, education has a positive effect on income distribu- tion and poverty alleviation; however, if there is a mismatch between 36/ Most researchers (e.g. Tidalgo 1976 and 1982) on Philippine employment trends have noted the labor saving nature of technology adoption in agriculture, industry and construction and the resultant expansion of service sector employment (as a last resort). While the comment on technology adoption may be true, the labor absorption patterns differ between rural and urban areas. 37/ See World Development Report, 1984, pp. 266-267. - 25 - education and occupation and if the educational levels are not adequately used, such a result may not accrue. The percentage of college level work force occupying professional, technical and related occupations was slowly declining in the Philippines (during 1976-82 it declined from 34.5% to 31.5%) and this broadly supports the view that the highly educated work force has not been adequately utilized. The issue of appropriate utilization of educated manpower assumes even greater significance in the Philippines where higher education continues to expand. (f) Underemployment, Occupation, Sector and Poverty. The ISH data of the fourth quarter of 1978 are used to asses the differences in the levels of underemployment among the poor and non-poor within the high poverty occupa- tions and sectors. The data are given in Table 9. By occupation, the evidence clearly points to relatively higher levels of underemployment among the poor. A general expansion of work opportunities will thus help alleviate poverty. The same signal comes from the data by industrial sector. However, in the urban services sector, which has absorbed increasingly higher percent- ages of people in the past decade, poverty cannot be significantly alleviated by increasing the work opportunities, for, the percent underemployed was low and equal both among the poor and the non-poor in services. Poverty in this case was the result of Low-productivity and low rewards. This comment applies to some extent to the service sector in the rural areas also. Another import- ant aspect is the high and equal (60%) rate of underemployment in the rural agricultural and manufacturing sectors among the poor in contrast to the relatively low rate in the service sector poverty group. Household manufac- turing of some sort was the alternative to agriculture in the rural areas and sufficient full-time work did not exist in both sectors. Thus, poverty alle- viation in rural areas requires a major expansion of work opportunities both in agriculture (where the majority are now occupied and who presumably possess the skills for agriculture) and non-agricultural sectors. B. Trends in Real Wages (i) The Data A brief note on the sources of data and recent developments in regard to data collection is given in Annex 4. The following points summarize the situation. First, data on agricultural wages, obtained by the Bureau of Agricultural Economics were limited to labor force in rice, corn, coconut and sugarcane. The Bureau is continuing this activity. Second, the Central Bank, until 1980 collected wage rate data in Ketro Manila from selected establishments. This activity was abandoned when the Government assigned the responsibility to NCSO. Third, the responsibility was shifted to the NWC in 1983 and the first (preliminary) set of wage rates for March 1983 were available recently. Problems of comparability and limited coverage apply to both rural and urban wage series. (ii) The Trends, 1970-83 (a) Metro Manila, 1970-1980: The Central Bank series indicate a declin- ing trend in the real wage rate of the skilled and unskilled workers -26- Table 9: UNDEREMPLOYMENT /a AMONG THE POOR AND NONPOOR BY OCCUPATION AND INDUSTRY, FOURTH QUARTER, 1978 Urban Area Rural Area Among the Among the Among the Among the Occupation/Industry poor non-poor poor non-poor Occupations Sales workers 20.1 8.2 36.4 17.8 Service workers 12.3 7.6 24.3 5.6 Agricultural, forestry and fishery workers 57.0 31.2 61.8 33.1 Production workers, transport operators and laborers 42.0 15.0 56.2 17.6 Industries Agriculture, forestry, fishing 57.4 30.0 62.2 33.0 Manufacturing 34.0 8.3 57.9 14.7 Wholesale and retail trade 18.3 7.8 35.5 16.8 Community and personal services 17.4 17.7 32.9 26.7 /a Percent of persons working less than 60 days. Source of Basic Data: NCSO, ISH, 1978 4th Quarter. - 27 - (Table 10). In actual fact, the drop was highest between 1970-1974 (51% for skilled workers and 53% for unskilled workers). Then followed a period of near v7stancy of the real wage (1974-78) with further declines in 1979 and 1980, These trends were brought about by steady increses in nominal wages and different rates of inflation during different periods. The inflation rates in Metro Manila in 1971, 1972, 1973 and 1974 were 15%, 10%, 14% and 34% respectively. These are high rates compared to 1975 (8%), 1976 (6%), 1977 (8%) and 1978 (7%). The situation worsened again in 1979 (19%) and 1980 (19%). (b) Metro Manila, 1980-83: The Central Bank data for 1980 and the NWC data for 1983 (March) for a set of ten occupational categories indicate that for some groups real wages have declined in March 1983 compared to 1980. These groups were electricians, foremen, masons and common laborers. The other groups (carpenters, compositors, latheman, linotypists, painters and plumbers) had higher real wages in March 1983 compared to 1980. (c) Agricultural wages, 1970-1982: The BAEcon data on nominal wage rates in palay, corn, sugarcane and coconut are converted into real wage rates by applying the consumer price index for areas outside Metro Manila (Table 11). (The NCSO does not prepare a rural CPI). Real wage rates in Palay declined during 1970-1974, increased during 1975-1978 and declined again in 1979 and 1980, with little change in 1981 and 1982. Real wages in the corn sector declined marginally between 1970 and 1974, increased during 1975-77 and then declined until 1982. The 1975-78 increase and the subsequent decline are broadly true of wages in sugarcane and coconut sectors also. In general, the real wage rates in the early 80s were lower than in the early and especially mid-7Us. (d) A Tentative Explanation of the Pattern of Real Wage Movements: The decline in rural and urban real wages during 1970-74, increase in rural areas and constancy in urban areas during 1975-78 and the decline Aft-. that (up until 3 J82/83) needs some explanatory hypothesis if not a precise explana- tion. Before attempting an explanation, however, some of the limitations of the data need to be mentioned. First, problems and errors may be present in the series on nominal wages especially since the Government agencies col- lecting the data as well as the reporting establishments are 48r,cious of Government minimum wage guidelines issued from time to time.-' There is, however, no simple way of evaluating the accuracy of the wage da-s (except by commissioning an independent wages survey). Second, the index number problem can vitiate the real wages significantly. Assuming that data limitations have 38/ These trends were shared by a number of occupational groups (skill categories). See Table 32, Annex 7. 39/ It is useful to note that the situation to be explained is not one of a simple declining trend. 40/ Fortunately no statistical relationship has been found between the legislated and the reported wages series (see Table 34, Annex 7). -28- Table 10: INDEX OF AVERAGE DAILY BASIC WAGE RATES OF LABORERS IN INDUSTRIAL ESTABLISHMENTS IN METRO MANILA, 1970-1980 (1972 = 100) Nominal Real Period Skilled Unskilled Skilled Unskilled 1970 90.6 88.4 114.4 111.6 1971 95.3 94.4 105.1 104.1 1972 100.0 100.0 100.0 100.0 1973 105.3 102.6 92.4 90.0 1974 115.1 110.8 75.6 72.8 1975 119.7 120.1 72.7 72.9 1976 124.4 126.2 71.2 72.3 1977 137.5 132.9 72.9 70.4 1978 154.4 138.4 76.1 68.3 1979 170.1 145.8 70.8 60.7 1980 180.9 151.5 63.7 53.4 Source: Central Bank. -29- Table 11: DAILY WAGE RATES OF LABORERS IN AGRICULTURAL OPERATIONS, 1970-82 Consumer price Daily wage rates (pesos/day) index outside Palay Corn Sugarcane Coconut Manila Nominal Real Nominal Real Nominal Real Nominal Real 1972 = 100 1970 3.44 4.35 3.05 3.86 3.18 4.03 3.38 4.28 79.0 1971 3.85 4.24 3.44 3.78 3.32 3.65 4.21 4.63 90.9 1972 4.36 4.36 3.76 3.76 3.67 3.67 4.41 4.41 100.0 1973 4.46 3.91 4.15 3.64 5.12 4.49 4.73 4.15 114.0 1974 5.56 3.65 5.56 3.65 5.88 3.86 6.36 4.18 152.2 1975 6.59 4.00 6.41 3.89 7.78 4.72 7.33 4.45 164.6 1976 8.97 5.13 8.77 5.02 7.28 4.16 8.73 4.99 174.8 1977 9.86 5.23 9.55 5.06 10.96 5.81 9.22 4.89 188.6 1978 10.42 5.14 9.76 4.81 9.99 4.92 10.18 5.02 202.9 1979 10.71 4.44 10.38 4.31 11.00 4.56 10.62 4.40 241.1 1980 10.60 3.72 10.66 3.74 10.97 3.85 10.71 3.76 284.9 1981 11.91 3.69 10.80 3.35 11.88 3.68 13.45 4.17 322.5 1982 13.07 3.68 11.11 3.13 12.76 3.59 15.32 4.32 355.0 Source: BAECON. - 30 - not affected the noted real wage patterns, they can be explained by the trends in labor supply, labor demand, and the overall macro-economic environment. Throughout 1970-82, labor supply (in an ex ante or potential sense) was expanding rapidly (ex post average annual growth rate was 3.6%). In such circumstances, relatively sluggish demand for labor tends to pressure money wages down, though money wages may not in fact decline for reasons such as setting wages equal to or slightly higher than previous levels and the probable practice of adhering to the (increasing) trend indicated by the legislated minimum wage levels. Under this scenario, whether or not real wages increase will depend on the inflation rate, determined largely outside the labor market. If the demand for labor is buoyant, money wages will be under an upward pressure. Then also the real wage movement depends on the inflation rate except that a macroeconomic environment conducive to stimulat- ing labor demand is more likely to go hand in hand with a relatively modest rate of inflation. During 1971-76 and 1976-82, rural employment expanded at high rates (4.7% and 5.1% respectively), but real wages increased onLy for a short while during 1975-78. Relatively high rates of agricultural GDP growth accrued during 1975-78. This period also saw a stable overall GDP growth rate, rela- tively high investment growth rates, the favorable effects of the earlier commodity boom, and special rural infrastructure programs. The incidence of underemployment was low during 1977 and 1978, an indication of some tightening of the labor market aided by rural to urban migration. Finally, during 1975- 78 inflation rates were relatively low. The economic environment deteriorated in subsequent years, slack developed in labor demand, and inflation started to erode the real value of the nominal wage rates. The situation was exacerbated by pressures on rural labor supply due to movement of labor from urban to rural areas. In the urban areas, employment expanded at 4.9% during 1971-76 and 2.8% during 1976-82. The healthy overall economic growth environment of the first period and part of the second (1977-78) was counterbalanced by rural- urban migration and excess labor supply (indicated by the 70% to 90% service sector share in employment growth) and the result was only a modest increase in the money wage. The relatively high inflation rate during 1971-74 contri- buted to the real wage decline and the low inflation rate of the 1975-78 period helped to keep the real wage rate constant. The deterioration in the demand for labor and the high inflation rates of the post-1978 period led to declining real wages. Sustained growth and modest inflation over relatively long periods of time ensure the best conditions for growth in real wages via growth in labor demand (cutting down underemployment), and increases in productivity. The process can be greatly aided by a slower (rather than a faster) labor force expansion which requires action on the population front today to derive benefits some 15 to 20 years hence. - 31 - (iii) Wages and Poverty Thresholds Trends, 1971-83: In the previous sections, it has been noted that during 1971-83, poverty incidence has declined, but, underemployment has increased and real wages have generally declined. In order to "reconcile" (at least partly) the noted trends, data on the poverty thresholds and wage rates are brought together (Table 12) and their implications are noted below. In the rural areas, wage rates are shown for laborers in corn farming where historically, the rates have been lower than in rice, coconut and sugar farming. The corn labor wage rates were lower than the rural poverty threshold wage levels in both 1971 and 1975. The situation changed in the subsequent years and the wage rates were higher than the poverty thresholds. In the urban areas (Metro Manila), the wage rates for common laborers (the lowest paid occupational category) were higher than the poverty threshold throughout 1971-83. The situation of falling real wages could thus coexist with a situation of declining poverty incidence because the wage rates were higher than the poverty thresholds. In addition, and most important, the levels of poverty incidence were in fact very low among the wage earner families in the urban areas and nonagricultural labor families in the rural areas. This could be also due to the significance of non-wage incomes for some of the families. However, the data in Table 12 do point out one disturbing trend in both urban and rural areas. It is the rapidly narrowing gap between the wage rate and the poverty threshold in recent years. It is quite likely that this reduction reflects the pressure of the expanding labor force on the job market and the slackening of employment growth in the last couple of years, indicated by the recent increases in underemployment. Prospects, 1984 and Beyond: With the economy registering negative growth in 1984 and inflation running at a very high rate, a worsening of employment, wages and poverty was almost certain. The situation for the period 1985-87 also is not very bright. With labor force expanding by 2.1 million persons in-the three years and economic growth projected at only 3% p.a., unless steps are taken to minimize the adverse effects of the slow growth on the lower income groups, the poverty situation will continue to worsen The Government anticipates a 4.4% annual growth rate in agricul- ture -- which if realized will go a long way to prevent poverty escalation in the rural ares. As for the urban areas, with projected growth rates of 2.5% in industry and 2.3% in services, the situation is less comforting. Between 1985-95, the annual addition to the labor force averages 0.8 million. This increases to a million persons between 1995-2000. To address the employment problems of such magnitudes, relatively high rates of economic growth, fairly high rates of investments and economic policies favoring labor-intensive development have to be pursued. 41/ Updated Philippine Development Plan, 1984-87, NEDA, September 1984, p. 22. - 32 - Table 12: ILLUSTRATIVE COMPARISON OF POVERTY THRESHOLD EARNING LEVELS AND AVERAGE DAILY WAGE RATES, 1971-83 Monthly Poverty poverty threshold line per daily earning Average daily wage Wage to poverty family (P) level (P) /a rate (P) threshold ratio (Z) Rural Metro Manila laborers common Year Rural Urban Rural Urban in corn/b laborers Rural Urban 1971 1985 249 3.9 4.8 3.4 9.8 87.2 204.2 1975 381 460 7.7 8.8 6.4 12.5 83.1 142.0 1979 425 500 8.2 10.1 10.4 15.2 126.8 150.5 1980 461 545 8.9 11.0 10.7 15.8 120.2 143.6 1981 532 629 9.7 13.4 10.8 n.a. 111.3 n.a. 1982 574 688 10.5 14.7 11.1 n.a. 105.7 n.a. 1983 614 882 11.2 18.8 n.a. 19.4 n.a. 103.2 /a Computed by deflating the monthly poverty line by an assumed number of 26 working days and the average number of employed persons per family. The average number of employed persons was 1.9 and 2.1 respectively in 1976 and 1982 in rural ares and 2.0 and 1.8 in 1976 and 1982 respectively in the urban ars. For the period covered in the above table, the following averages were assumed: Rural: 1.9 for 1971 and 1975, 2.0 for 1979-80; and 2.1 for 1981-83. Urban: 2.0 for 1971 and 1975; 1.9 for 1979-80; and 1.8 for 1981-83. /b Wages are about the lowest in corn compared to rice, sugar and coconut. - 33 - As for the next century, since the new entrants to the labor force are yet to be born, the country can opt for promoting population control seriously and effectively now so as to reduce the intensity of the future labor force explosion. V. RECOMMENDATIONS (i) Monitoring Poverty Monitoring poverty is a serious matter and the Government cannot ignore this responsibility. The Government, however, was in favor of monitor- ing the progress of the bottom 30%. Since there is no way of identifying a fixed group of families or persons as the bottom 30%, exercises on monitoring their progress will become controversial and complicated. From the standpoint of long-run political and social stability, the most desirable course is for the Government to establish simple poverty lines (such as those suggested in Annex 6), monitor both poverty incidence and the poverty focus in its development programs and projects. To facilitate this, streamling and improving the data collection are recommended. (ii) Data Collection and Dissemination (a) Household Incomes and Expenditures: The practice of holding more than two quarterly ISH should be reviewed since data processing and the publi- cation of data can not be handled expeditiously when more than two quarters are covered in a year (see Annex 1). Collecting household income data in each and every quarterly ISH did not really serve any purpose. Instead, once in every two or three years, data on incomes, expenditures and savings should be obtained in detail by expanding the scope of the ISH for the concerned years. Tabulations should be provided by per capita income and per capita expenditure classes in addition to tabulations by family income or family expenditure classes. The 1981 Census of Agriculture and Fisheries is an important source of information on agricultural and fishery output. The data refer to 1980. The previous agricultural census was conducted in 1971. The two censuses together provide valuable data not only to assess the changes over a decade in rural levels of living, but also to find the sources of agricultural growth and its distributive benefits. The NCSO should process the 1981 census data quickly and publish important national and regional tables on a comparable basis for 1971 and 1980. (b) Labor Force and Employment: The present efforts to strengthen the ISH (see Annex 1) should be continued and fully implemented. Each year labor force and employment data must be collected as at present. However, an ISH should be conducted only once a year or atmost twice a year. Establishment surveys should concentrate on obtaining data on output and value added rather than on employment and earnings. The Ministry of Labor and Employment should have timely and full access to ISH labor force and employment data and should discontinue its household surveys. BAEcon should conduct, once every two years, comprehensive socio-economic surveys, rather than the present rice and - 34 - corn surveys, to obtain detailed information on the employment and related economic conditions in the rural economy. (c) Wages and Prices: The Ministry of Labor and Employment should collect quarterly data on wage rates in the non-agricultural sectors in key cities. The proposed BAEcon rural surveys would provide rural wage rate data with an expanded coverage. As for prices, the CPI's are produced for the regions and the country, but not by rural and urban classification. The price data routinely collected by NCSO are sufficient to prepare reasonable rural and urban price indices. Such indices should be prepared for the general population as well as for the low income groups such as agricultural laborers in the rural areas and the industrial workers in the urban areas. (iii) Statistical Organization The problems of data collection and dissemination are not Limited to household expenditure, employment and wages. In cases like national accounts, some weaknesses are known to exist. The weighting pattern for consumer price index is outdated. Trade data are said to be off from partner data. It is common knowledge that data on public enterprises are limited. Vital registra- tion is inadequate for estimating birth rates and death rates. The list can be long. The need is for a comprehensive review of the entire statistical system with a view to recommending measures to ensure the timely collection and dissemination of statistical information without duplication of efforts by different agenices. Such a review should go into all relevant aspects including staffing, salaries and budget. - 35 - REFERENCES 1. Abrera, Ma. A.S., "Philippine Poverty Thresholds" in M. Mangahas (ed)., Measuring Philippine Development, Manila: Development Academy of the Philippines, 1976. 2. Bautista, Romeo M., The Development of Labor Intensive Industry in the Philippines, (Unpublished) July 1979. 3. Bautista, R., Power J., et al, Industrial Promotion Policies in the Philippines, Manila: Philippine Institute for Development Studies, 1979. 4. Berry, A., "Income and Consumption Trends in the Philippines, 1950- 70", Review of Income and Wealth, September 1978. 5. Daly, Rex F. Merlinda D. Ingco and Clemencia H. Padrinao, Income Distribution, Food Consumption and Poverty, Staff Paper No. 7., Agricultural Policy Analysis Staff, Ministry of Agriculture, Quezon City, June 1981. 6. Dutta, G., and Meerman, J., Household Income or Housebold Income Per Capita in Welfare Comparisons, Staff Working Paper No. 378, World Bank, 1980. 7. ILO., Sharing in Development: A Programme of Employment, Equity and Growth for the Philippines, Geneva, 1974. 8. Jain, Shail, Size Distribution of Income: A Compilation of Data, World Bank, 1975. 9. Lal, Deepak, Wages and Employment in the Philippines, World Bank, October 1979. 10. Lipton, M. Demography and Poverty, Staff Working Paper No. 623, World Bank, 1983. 11. Mangahas, Mahar, Quizon, J.B. and Lim, Antonio, A Critique of NCSO 1975 Family Income and Expenditure Survey (Unpublished) 12. Mangahas, Mahar, "Income Inequality in the Philippines: A Decomposition Analysis," in Income Distribution, Employment and Economic Development in Southeast and East Asia, the Japan Economic Research Center, Tokyo, July 1975. 13. Mangahas, Mahar, "What happened to the Poor on the Way to the next Development Plan?" Development Academy of the Philippines, Manila, December 1981. 14. NEDA, Some Aspects of Rural-Urban Welfare Differential, Manila, August 1984. - 36 - 15. Oshima, Harry T., "Sector Sources of Philippine Postwar Economic Growth: The overall record in Comparative Perspective" Journal of Philippine Development, First Semester, 1983. pp 1-44. 16. Power, J.P. and Sicat, G.P., The Philippines-Industrialization and Trade Policy, Oxford University Press, 1971. 17. Rao, V.V. Bhanoji, Poverty in Indonesia, 1970-80: Trends, Associated Factors and Research Issues, World Bank, January 1984. 18. Republic of the Philippines, Five-Year Philippine Development Plan, 1978-1982, Manila, September 1977. 19. Republic of the Philippines, Five Year Philippine Development Plan, 1983-87, Manila, May 1982. 20. Republic of the Philippines, Updated Philippine Development Plan 1984-1987, Manila, September, 1984. 21. Sta. Romana III, Leonardo, Income Inequality in the Philippines: A Survey of Evidence, Working Paper, Income and Assets Distribution Research Project, Institute of Economic Research, Hitotsubashi University, Tokyo, September 1976. 22. Tan, E.A., and Holazo, V., Measuring Poverty Incidence in a Segmented Market, Discussion Paper No. 78-14, Quezon City: School of Economics, University of the Philippines, December 1978. 23. Terasaki, Yasuhiro, Overviews of Philippine Income Distribution Studies (I): Data Availability, Working Paper, Income and Assets Distribution Research Project, Institute of Economic Research, Hitotsubashi University, Tokyo, May 1977. 24. Tidalgo, R., "Labor Absorption in the Philippines, 1956-1973" Philippine Economic Journal Vol. 15, Nns. 1 & 2, 1976. 25. Tidalgo, Rosalinda, and Esguerra, Emmanuel F., Philippine Employment in the 1970s, Manila: Philippine Institute of Development Studies, July 1982. 26. University of the Philippines, An Analysis of the Philippine Economic Crisis, A workshop Report (by a panel of 10 economics professors), June 1984. 27. Visaria, P., Poverty and Living Standards in Asia, LSMS Working Paper No. 2, World Bank, 1980. 28. World Bank, The Philippines: Priorities and Prospects for Development, 1976. - 37 - 29. World Bank, Aspects of poverty in the Philippines: A Review and Assessment, Vol. II: Main Report, Report No. 2984-PH, December 1, 1980. 30. World Bank, World Development Report, 1984. 31. World Bank, The Philippines: An Agenda for Adjustment and Growth, Report No. 5258-PH, November 1984. -38- ANNEX 1 Page 1 THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES Integrated Survey of Households (i) Introduction In November 1976, the regular quarterly labor force survey conducted by the National Census and Statistics Office (NCSO) was 'integrated' with the semi-annual agricultural survey conducted by the Bureau of Agricultural Economics (BAEcon) to form a single survey called the Integrated Survey of Households (ISH). (The integration was enabled by Presidential Letter of Instruction No. 521 and NEDA Memorandum Circular No. 7-77.) The ISH was planned with technical assistance from NEDA's Inter-Agency Committee on Survey Design, which advised on sampling, the formulation of concepts, the definition of terms, and the questionnaire design. The ISH is a joint undertaking of various government agencies. The principal agencies involved are the National Economic and Development Authority (NEDA) which is the coordinating agency, the Bureau of Agricultural Economics(BAEcon) and the NSCO, which are the operating agencies. The ISH was envisaged as a continuing quarterly survey of non-institutional households. The ISH was designed to serve the following objectives: (a) to measure the level of employment, unemployment and underemployment in the country; (b) to gather data on employment generation, production and productivity of family-operated agricultural and non-agricultural activities; (c) to gather data on rice and corn production, household rice stock, and inventory of livestock and poultry; and (d) to gather data on both individual and family earnings and income from other sources. (ii) ISH Questionnaire Design The ISH questionnaire is structured into several parts: Part I - Data on Individua's Part II-A - Data on Own Family-Operated Activities Part II-B - Data on Family-Sustenance Activities not Considered in Part II-A Part III - Data on Crops, Livestock and Poultry. The data on individuals (Part I of the Questionnaire) covers demo- graphic characteristics (age, sex, marital status, highest grade completed, and usual occupation), and economic characteristics (various items pertaining 39q - ANNEX 1 Page 2 to employment and unemployment). In addition, receipts from dividends, pensions and transfer are also obtained in Part I. Data of own family-operated activities (Part II-A of the question- naire) refers to information on production and net receipts from various economic activities (fishing, manufacturing, etc.). Similar data are obtained in Part II-B on the family-sustinance activities not considered in Part II-A. Data on crops, livestock and poultry was obtained in the final part of the questionnaire for processing by BAEcon. (iii) Instructions for Collecting Employment Data in ISH For a few selected questions, an edited extract of the instructions for obtaining employment data from respondents to the ISH questionnaire is given below. It shows that the data collected (if reliable) can be extremely useful to study labor utilization, intensity of employment, unemployment (labor-time not at work but available for work), and visible underemploy- ment. The survey also allows the employment data to be linked with income Tearnings). Did the respondent have a job or business or unpaid work on family farm or business, even for only one hour at any time during the past quarter? Explain to the respondent the meaning of work; ask probing questions to ascertain existence of a job or business, or unpaid work on family farm or enterprise during the reference quarter. Work is any activity of a person done for pay, profit or without pay on a family farm or enterprise or an activity done by a farm operator or member of his family on another's farm on 'exchange labor' arrangement. (Work for pay, work for profit, work without pay on family farm or enterprise were defined.) Examples of activities which are not considered as work are: housekeeping in own home; building, repairing or painting own house; volunteer work; activities of students in civic action; and begging or gambling. What was his primary job or business during the past quarter? What other job or business did he have during the past quarter?: Primary job/business is any gainful activity of a person which is permanent and a full-time job, lasting for 1 year or longer or had lasted or was expected to last for one year or longer, regardless of whether he had a job/business or not during the reference quarter. If a person has two permanent jobs consider the one where he worked more hours as his primary job. If, however, these two jobs have equal hours of work, consider as primary job the one where he derives more income. Other jobs/businesses are any gainful activities of a person which may be permanent or temporary, full-time or not, aside from his primary job. No. of Full Days Worked: For the purposes of this survey a full days work is based on a person's normal prescribed working hours per day whether it be 4 hours, 5 hours, 10 hours, as long as he was paid a full day's pay. Normal Working Hours per Day: For the purposes of this survey, number of normal working hours per day refers to the usual or prescribed -40- ANNEX 1 Page 3 working hours of a person for a full day's work in his job/business. (iv) Instructions for Collecting Income Data in ISH Adapting the ISH Manual for Field Workers (revised August 1978), the instructions for obtaining income data from the respondents for a sample of items are given below. Wage/Salary During the Past Quarter: These items must be asked for all persons with entry for primary and secondary jobs except for the following: unpaid family worker on a family-operated activity, and operators or managers of a famiLy-operated activity who do not receive salaries or wages either in cash or in kind from the family business or enterprise. Cash salary or wage includes basic compensation received in cash, cash commissions, tips, bonuses and family and clothing allowances in cash. Salary and Wage in kind includes those compensations received by an employee in the form of goods such as rice, corn, fish or any other form of payment not in money, and rental value of housing quarters provided to an employee. The amount to be entered must be the total wage and salary earned (even if not received during the quarter) before deductions of contributions for insurance, retirement and medicare premiums, withholding tax and other contributions reflected in the payroll. Take note that total wage and salary include basic pay plus com- missions, tips and allowances as defined earlier. However, salesmen and brokers who worked on commission basis should report only the net amount received after deducting all expenses incidental to their jobs as salesmen or brokers. Receipts from non-economic activity during the past quarter: This particular section aims to collect data on receipts from non-economic activity as well as earnings of persons less than 10 years old, either in cash or in kind during the reference quarter. These receipts are dividends, interest and roylaties and profits for sale of stocks and bonds; rentals, including land- owner's share of agricultural products and rental value of owner-occupied house; and pension, retirement benefits, support, gifts, winnings, proceeds from insurance, etc. Data on Own Family-Operated Activities: There are five blocks comprising Part II-A of the Questionnaire. These are: (A) fishing; (B) manu- facturing and processing; (C) service-type activities and construction; (D) forestry and logging, hunting and trapping, mining and quarrying; and (E) farming, gardening and livestock and poultry raising. All five blocks ask for information on production, cost of operation, net receipts from the par- ticular activity, and labor inputs of family and non-family members during the reference quarter. Information on the extent of production and disposal of produce are also gathered. The instructions issued to the enumerators are illustrated below in respect of fishing activity. Block A - Fishing: There will be entries in Block A - Fishing, if any member of the family did some fishing or gathered live shells and seaweeds for sale or for consumption, on own account during the past quarter. This includes deep-sea and off-shore fishing and fishpond/pens and seaweed farm etc. Fishing enterprise managed and operated by cooperatives and corporations 41 -ANNEX 1 Page 4 are not considered family operated activities, and therefore they are to be excluded. "Managers," "encargados," "caretakers" and other workers having direct supervision in a fishing activity who are paid in cash and/or in kind as employees should not have reports in this block. A lessor or one who rents out fishpond, fishing boat or other fishing equipment as a regular enterprise, is not considered an operator of a fishing activity but rather operator of a rental business. The activity will be reported in Block C (Service type Activities and Construction.) Total catch, harvest or gathering (in kg): Enter the total catch, harvest or total quantity gathered by type of fishing used during the refer- ence quarter in kilograms. In the case of fingerlings and bangue or shrimp fry, report the number gathered. The report for the total catch, harvest or gathering must include the quantity consumed by the household, those sold or processed for sale, those given away or paid for debt, and those given as share to workers. (Conversion to Kilograms: If the report on total catch is given in units like "tiklis" (basket), "banyera" (tub), "kahon" (box), etc., ask the average weight in kilos that the unit contains. Multiply the weight given by the number of units or containers to obtain the figures in kilo- grams. When a respondent knows the total value of catch but he does not have any idea about the weight of the catch, simply divide the total value of catch by the average selling price per kilogram during the quarter, to estimate the weight in kilograms.) Total value of catch, harvests, or gathering (Pesos): Report the total value of catch, harvest or gathering for each type of fishing activity in pesos. "Value" includes the total amount of money received by the family from the sale of its catch, harvest or produce plus the imputed cost of the catch, harvest or produce, which was consumed by the family, paid for debt, given as share to workers, processed for sale and/or given away free. The imputed value should be based on the market price prevailing in the locality at the time of the catch, harvest or produce. Net receipts in cash and in kind: Total net receipts refer to excess of the total value of catch during the reference quarter over the cost of operation. To compute the total net receipts in cash or in kind, deduct all the expenses incurred in cash and the imputed value of expenses in kind from the total value of the catch. If operator had non-family member business partners, operator's share of net receipts (percent): Ask the respondent if the operator of the fishing activity had a business partner who is not a family member. If so, inquire about the percentage share of the operator from the net receipts both in cash and in the form of goods. It should be stressed here that if the operator had a partner who is not a family member, the report on family production, distribution of produce and cost of operation shall be those of the whole enterprise and not the percentage of the operator alone. Data on Family-Sustenance Activities: Part II-B deals with household activities not considered as family-operated activities. These are also sources of income which are not included in Part II-A. Some examples of these activities which could augment the family income are: raising one or two 42 ANNEX 1 Page 5 fowls, raising a few eggplants or tomatoes, fishing or gathering shells for a day's meal, etc. Goods produced from these activities are intended for home consumption or for sale if harvests are plenty. In some cases where a house- hold operates a sari-sari store and at the same time raises a pig or a few fowls for home use or for sale, this household will have entries in Parts I, II-A, and II-B. The members engaged in operating the store will have entries in Parts I and II-A while those involved in raising chickens or a pig will have entries in Parts I and II-B. Had the members raised 3 or more pigs or 30 fowls, there should have been another report in Part II-A under E - Farming, Gardening and Livestock and Poultry Raising. But if the members of a household engage in two or more related activities like fishing, which is the family-operated enterprise, and shell gathering which a household member casually does for food, the entry must be only in Parts I and II-A under A- Fishing. The proceeds from shell gathering will be reported in additon to the net receipts of the operator in fishing in Part II-A. (v) Survey Design and Sampling Frame A two-stage stratified sampling design was adopted with barangays (villages) and households as the primary and secondary sampling units, respec- tively. The basic frame from which the sample barangays were drawn is the 1975 Integrated Census of Population and its Economic Activities (ICPEA) (see Annex 2). The barangays were stratified on the basis of the economic activity If most of their working population as: (1) palay; (2) corn; (3) other agri- cultural crops, (4) fishing; (5) manufacturing; and (f) other economic activi- ties. However, if at least 20% of the working population was engaged in manu- facturing activities, the barangay was automatically classified under stratum (5), manufacturing. The sampling frame for the second stage of sampling was prepared by listing all households in the selected barangay. During the listing, the households were classified as either engaged in the stratum activity (substratum 1) or not engaged in the stratum activity (substratum 2). At least one member of the household should have been working as an operator or salary/wage earner in the activity for the household to have been considered in substratum 1; otherwise the household was classified in substratum 2. Starting with the third quarter of 1977 the barangays, which were earlier stratified according to the (six) economic activities of their populations, were further stratified into urban and rural areas. It was also decided that a fixed sampling fraction would be adopted for each stratum. Stratification was done within each province because provincial estimates were sought. The table below shows the siumary of the total samples selected for all provinces using a fixed overall sampling ratio in each stratum. ANNEX 1 Page 6 No. of No. of Sampling sample sample Stratum fraction barangays households Urban palay 1/050 211 1,688 Urban corn 1/050 80 640 Urban other agricultural crops 1/050 88 704 Urban fishing 1/050 65 520 Urban manufacturing 1/250 420 3,360 Urban other economic activities 1/250 650 5,200 Rural palay 1/300 1,037 6,222 Rural corn 1/300 543 3,258 Rural other agricultural crops 1/300 436 2,616 Rural fishing 1/100 321 1,926 Rural manufacturing 1/250 311 1,866 Rural other economic activities 1/250 543 3,258 Total 4,705 31,258 Due to greater variability generally observed in the urban areas compared with the rural areas, a sample of 8 households (4 households for engaged, 4 for not engaged) was fixed for urban barangays and 6 households (3 for engaged and 3 for not engaged) fixed for rural barangays. The sample barangays were selected from each stratum with probabiLity proportional to size using the 1975 ICPEA number of households as the measure of size. On the basis of the predetermined number of samples per substratum, the sample households were selected from the list of households within each selected barangay. Systematic random sampling was done independently in the selection of sample households for each substratum. Replacement of samples was not allowed although it was possible to have nonresponse cases due to a critical law and order situation, or inaccessibility of the selected sample barangays or sample households. The allocation of sample households to the twelve strata was first made at the national level, and using the resulting national allocation levels as controls, they were then allocated to the different provinces. The overall sampling ratio - the ratio of the number of households to the estimated total number of households - was held constant within each stratum at both the national and provincial levels. (vi) Concepts, Definitions and Classifications A barangay is the smallest political subdivision in the country, and is the basic geographic enumeration area for the ISH. A city or municipality is composed of several barangays. The following guidelines used in the 1975 Integrated Census of the Population and its Economic Activities (ICPEA) were adopted in classifying urban areas in ISH: -44- ANNEX 1 Page 7 (a) In their entirety all cities and municip lities having a population density of at Least 1,000 persons per km (b) Poblaciones or central districts of municipalities and cities which have a population density of at least 500 persons per km2 (c) Poblaciones or central districts (not included in (a) and Cb)) regardless of the population size, which have the following: (i) street pattern, i.e. network of streets in either parallel or right- angle orientation; (ii) At least six establishments (commercial, manufacturing, recreational and/or personal services); and (iii) At least three of the following: a. A town hall, church or chapel with religious services at least once a month; b. A public plaza, park or cemetery; c. A public place or building where trading activities are carried on at least once a week; d. A public building like a school, hospital, puericulture and health center or library. (d) Barangays having at least 1,000 inhabitants which meet the condi- tions set forth in (c) above, and where the occupation of the inhabitants is predominantly nonfarming or nonfishing. Rural areas were defined as those not falling under any of the above classifications. Employed persons include all those who, during the reference quarter, are 15 years old and over as of their last birthday and are reported as being either at work or with a job but not at work. Unemployed persons includes all those who, during the reference quarter, are 15 years old and over as of their last birthday and are reported to be wanting and looking for work. Also included are persons reported as wanting full-time or part-time work but not looking for work because of the belief that no work is available, or because of temporary illness, bad weather or other valid reasons. Persons not in the labor force are those who are without jobs and not wanting work, or wanting work but not looking for work for reasons other than those mentioned earlier. They include housewives, students, disabled or retired persons and seasonal workers who are not working and not looking for work during the survey period. The employment status of persons 15 years old and over is determined on the basis of answers to a series of interrelated questions. The data on occupation and industry relate to the job held during the reference quarter. Persons employed at two or more jobs are reported in the job at which they worked the greatest number of hours during the quarter. (vii) Data Processing Editing and coding of information are usually done in the 72 provin- cial offices of the NCSO, but a general review is carried out at the Central Office by the Household Surveys Division of Parts I and II of the schedule. BAEcon processes Part III (agricultural production) which is only canvassed every semester. 45 ANNEX 1 Page 8 Right after the survey, the questionnaires are edited by the provin- cial staff to resolve inconsistencies which need field verification. The next stage of quality control is the general review of questionnaires by the Household Surveys Division at Central Office, Manila. Each questionnaire is subjected to close scrutiny and to verification of entries and codes. For example, it often happens that received questionnaires have omissions, impossible entries, inconsistent entries, and entries of unreasonable magni- tude. Editing is meant to inspect and correct these errors. Any inconsis- tencies of entries which cannot be resolved are referred back to the field offices for correction. During keypunching, accuracy is ensured through complete verification of all keypunched information. The last stage of quality control is the machine-edit for completeness and consistency of records and entries. Error listings are prepared and the subject matter division makes the necessary correction on computer printouts. The tape file is then updated on the basis of the corrections. As soon as the file is clean some test tables are produced for final analysis and approval of the subject matter division. If found to be within expected standards, the final tables are prepared. A descriptive analysis of survey results is done by the staff of the Household Surveys Division. Advance results from manual tallies are published about two months after the quarterly survey is conducted. The final results are published much later, and currently there is quite a backlog of data to be published. (viii) Review of ISH Surveys Conducted to Date To date some 23 rounds of ISH have been undertaken, on which summary information is given in Table 1. Since the start of the ISH in November 1976, 10 of the 23 survey rounds conducted included Part III (the BAEcon question- naire). Of the 23 surveys, 6 (1st quarter 1977, 4th quarter 1977, 2nd quarter 1978, 4th quarter 1978, 1st quarter 1984 and 2nd quarter 1984) contained only Part I (Labor Force Characteristics and Income), of which all but the last two have been tabulated and published in final form. The processing for the remaining survey rounds is still to be completed, and problems have been encountered in making consistent Part I (Individual Data) and II (Data on Family-Operated Activities). However, some of these rounds do have final tabulations for Part I (3rd quarter 1976, 3rd quarter 1977, 1st quarter 1978, 3rd quarter 1978 and 4th quarter 1978). It should also be noted that only the 3rd quarter and 4th quarter survey rounds were conducted for each of the years 1980, 1981, and 1982. Part III of the questionnaires was processed separately by BAEcon and the plan to match the NCSO ISH files with the BAEcon files to produce common tabulations has not been implemented as yet. Monitoring and scheduling of the program for processing the quarters of 1979 and later years has been turned over to the subject matter divisions. Modifications and revisions have been incorporated, one of which is the introduction of the structural check program which screens the consistency of five major fields in order to make Parts I and II agree and thus simplify the machine-editing procedure. The Household Surveys Division has not decided whether Part II of the earlier survey rounds before 1979 are Table 1: SUMMARY INFORMATION ON ISH No of Parts of question- sample Year Quarter naire included households Nature of data Publication/date of publication 1976 3rd 1, II-A, III 30,000 Not tabulated 1977 1st I 15,000 Not tabulated 3rd I, I-A, III 31,258 Not tabulated 4th I 15,000 Not tabulated 1978 lt I, 11-A, III 31,258 Unweighted (preliminary) Special release 286/01-04-80 2nd I, 15,000 Unweighted (preliminary) Special release 251/03-12-79 3rd I, II-A, II-B, III 31,258 Unweighted (preliminary) Special release 286/01-04-80 4th I 15,000 Unweighted (preliminary) Special release 286/01-04-80 1979 1st I, II-A, II-B, III 31,258 Unweighted (preliminary) Special release 301/03-10-80 2nd I, II-A, II-B 15,000 Unweighted (preliminary) Special release 320/05-21-80 3rd I, II-A, II-B, III 31,258 Unweighted (preliminary) Special release 325/07-14-80 4th I, II-A, II-B 15,000 Unweighted (preliminary) Special release 336/08-15-80 1980 3rd I, II-A, II-B, III 31,258 Unweighted (preliminary) Special release 422/08-20-82 4th I, II-A, II-B 15,000 Unweighted (preliminary) Special release 409/05-13-82 1981 3rd I, II-A, II-B, III 31,258 Unweighted (preliminary) Special release 476/08-15-83 4th I, II-A, II-B 15,000 Unweighted (preliminary) Special release 431/11-12-82 1982 3rd I, II-A, II-B, III 31,258 Weighted (preliminary) Special release 476/08-15-83 4th I, II-A, II-B 15,000 Weighted (preliminary) Special release 478/08-22-83 1983 1st I, II-A, II-B, III 31,258 Weighted (preliminary) Special release (unnumbered)/11-02-83 3rd I, II-A, II-B 31,258 Weighted (preliminary) Special release 501/01-06-84 4th I, Il-A, II-B 15,000 Weighted (preliminary) Special release (unnumbered) 1984 let I 17,694 2nd I Metro Manila only -47- ANNEX 1 Page 10 to be processed. The current system does not seem to be applicable to the 1977 and 1978 series because of changes in the questionnaire format. The status of the ISH tabulation program for the period 1976-84 is shown in Table 2. The Household Surveys Division with the help of new graduate trainees is committed to complete the final results of all years since and including 1979. Strong recommendations have been made by some agencies to process only Part I so as to compile the high priority labor force tabulations. (ix) Problem Areas Identified in the ISH1 (a) Questionnaire Design. As noted, the ISH aims to obtain a compre- hensive overview of economic activities of both individual (Part I) and house- hold production units (Part II), including agricultural production. The analytical structure is patterned after the 1975 ICPEA. However, it appears that consistency between Parts I and II of the questionnaire has not been fully tested. Moreover, questionnaire changes have also been made periodi- cally, which has necessitated programming revisions and added to processing problems. Thus, data processing has been delayed because of the problems of making Parts I and II of the questionnaire consistent. The questionnaire has also been difficult to implement in the field because of its complexity. Interviewers find difficulties in making the information collected in Parts I and II consistent. (b) Sampling Design. The sampling design appears to be operationally too complex, with almost 9,410 raising factors: 2 (engaged or not engaged) for each of the 4,705 sample barangays. In contrast, the 1971 FIES used 4 raising factors - 2 for urban and 2 for rural PSUs. Moreover, survey operations are subject to possibly large biases due to difficulties experienced in updating the sample frame, especially for split barangays. The frame mapping and updating problem are of some concern. (c) Data Collection/Field Operations. Problems encountered in this area are related to both administrative and technical issues. The present travel allowance provided to interviewers appears inadequate and this seems to be contributing to nonresponse. The lack of transportation (jeeps, motorcycles and pumpboats) in the field is also a serious problem. Offshore/high cost interview areas including inaccessible and critical barangays should perhaps be treated separately as special enumeration areas; they should not be included in the regular survey operations. (d) Pre-Central Office Processing. Although the tabulation of data at the provincial and regional offices will help to reduce editing errors, there are still too many incomplete and erroneous questionnaire returns found in processing at Central Office (miscoding and omission of entries). More 1/ In March 1985, at thetime of discussing the report, the government representatives stated that some of the problems noted here were resolved and others were being resolved. Table 2: STATUS OF ISH TABULATIONS 1976 3rd qtr Part I, II & III (BAEcon Processing) Special Report No. 1 Employment data only 1977 1st qtr Part I 3rd qtr Part I, II & III (BAEcon Processing) Bulletin No. 48 Employment data only 4th qtr Part I Bulletin No. 48 1978 lit qtr Part I, II, III (BAEcon Processing) Bulletin No. 49 Employment data only 2nd qtr Part I 3rd qtr Part I, II, III (BAEcon Processing) Bulletin No. 50 Employment data only 4th qtr Part I Bulletin No. 50 1979 1st qtr Part I, II & III (BAEcon Processing) Handtally of incnme and employment (unweighted) 2nd qtr Part I, II 3rd qtr Part I, II & III (BAEcon Processing) 4th qtr Part I, II 1980 3rd qtr Part I, II & III (BAEcon Processing) Handtally of income and employment (weighted) 4th qtr Part I, II 1981 3rd qtr Part I, II & III (BAEcon Processing) Handtally of income and employment (weighted) 1982 3rd qtr Part I, II Handtally of income and employment (weighted) 4th qtr Part I, II Handtally of income and employment (weighted) 1983 3rd qtr Part I, II Handtally of income and overseas employment (weighted) 4th qtr Part I, II Handtally of income and overseas employment (weighted) 1984 3rd qtr Part I Handtally of income and overseas employment (weighted) 4th qtr Part I (Metro Manila only) Handtally of income and overseas employment (weighted) A -49- ANNEX 1 Page 12 questionnaire spot-checking should be encouraged at the provincial and regional offices to reduce data cleaning problems at the Central Office. (e) Central Office Processing. The data encoding process needs more systems and subject matter divisions' participation. Editing and imputation procedures need improvement. Historical data files from which data could be used to impute missing values would also be helpful. In this regard, use of shuttle questionnaire forms and control cards should be considered. Data processing for the ISH is part-manual (the preliminary statis- tics based on hand-tally summaries) and part-computerized (the statistics for final publication). The two are independent, and thus the data-processing design would seem to be inefficient. The timetable of processing activities has not been realistic. Since a quarter's load cannot be finished in one quarter, the previous set is still being processed, when the next quarter's data are received. This has led to a growing backlog. Experience shows that the processing for a quarter normally takes 6 months, so that in one year at most two surveys can be finished assuming no delays in cleaning the data. It appears that every quarterly survey is a system by itself and has to be maintained by the EDP group of systems analysts and programmers. Standardization is lacking. The system has lost its operational effectiveness with the addition of extra processing steps caused by systems changes within a quarter and further adjustments for successive quarters' requirements. With respect to data requirements, items are added or deleted from time to time, which affects the system programs. Acceptable procedures for machine-editing have not been established as yet, and the EDP group resorts to short-term remedies for problems encountered along the way, which leads to delays. Staffing is a highly critical problem due to personnel turn-over not only at EDP but also at the Household Surveys Division. Employees concerned with system development, computer programming, analysis of statistical reports, etc., are trained statisticians with two-year contracts. However, even before completing their required terms, they resign or transfer to the private sector or other agencies in response to more attractive job offers. (x) Ongoing Activities Toward Improving the ISH A number of activities are currently underway to improve the house- hold survey system. These include: (a) A set of proposed core and special modular questionnaires covering labor force data and a number of subject areas has been prepared. These have been implemented with effect from 1st quarter 1984, and consist of a core questionnaire (a shortened Part I) and special supplements or "riders." (b) A proposal has also been made to the Technical Committee on Survey Design to revise the current ISH sample design to make it simpler. A design using 4 or 5 independent subsamples has been proposed and is being implemented. This should alleviate some of the present NCSO bottlenecks in the area of ISH data collection and processing. -50- ANNEX 1 Page 13 (c) More regional tabulation activities are being encouraged to reduce the Central Office workload. (d) The University of the Philippines Statistical Center has a project to identify and recommend better editing and computation methods to alleviate the processing backlog. (xi) Generating Annual Income Statistics from the ISH The ISH presently gathers data on both individual and family incomes. It generates information on income levels and distribution pertain- ing to the reference quarter. Although initially intended to be conducted quarterly, the ISH has sometimes not been undertaken owing to limited resour- ces. The quarterly income data have not, however, been aggregated to generate annual income statistics. The later are more meaningful because of seasonal- ity of incomes, especially in agriculture. Since data on family incomes in ISH are collected and processed independently every quarter, a family may well belong to different income classes in different quarters (particularly a family in the agricultural sector). Thus, if the tabulated quarterly statis- tics were simply to be added up by income class, this would tend to create a bias in the annual income distribution. To generate an accurate picture of annual income distribution, the quarterly income of each family should be aggregated to get annual income. Another problem concerns the quality of the family income data. The ISH was intended mainly to generate statistics on labor force characteris- tics. Hence, data on family incomes were only of secondary interest. As a consequence, the questionnaire did not include sufficient detail to ensure accurate recall, particularly in respect of income (net receipts) from family - operated activities. There is, therefore, a need to examine further Part II (Family-Operated Activities) of the ISH questionnaire. Presently, net receipts during the reference period are obtained by deducting from the value of gross production all the costs incurred. Entries for the quarters have to be examined to check the consistency of the values and accuracy of the estima- tion procedures. Furthermore, the editing procedure has likewise to be reviewed to ensure consistent and valid income entries. A joint project has been proposed between PIDS, NCSO (Household Surveys Division) and NEDA (Statistical Coordination Office) to generate annual family income statistics for 1979, 1981 and 1983 at the national and regional levels. The tabulations will provide information on income distribu- tion as well as socio-economic characteristics of particular income groups. But it is important to note that the resulting ISH income statistics will not be fully comparable to previous FIES distributions. This is because of differences in sampling, questionnaire design, and processing and editing procedures. - 51 - ANNEX 2 Page 1 THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES The 1975 Integrated Census of the Population and its Economic Activities (ICPEA) (i) Introduction In 1975 the National Census and Statistics Office conducted the Integrated Census of the Population and its Economic Activities (ICPEA), which was the precursor of the subsequent ISHs. The ICPEA was a two-phase opera- tion. Phase I, conducted in May 1975, was designed to provide comprehensive statistics on the age and sex composition, geographic distribution, fertility, migration pattern, manpower, and other demographic, economic and social char- acteristics of the Filipino people. Phase II, undertaken in December 1975, gathered data on the economic characteristics of persons 10 years old and older in activities like farming/gardening, livestock and poultry raising, fishing, manufacturing and other household-operated activities for the period January 1 to December 31, 1975. While Phase - was a complete enumeration, Phase II was conducted on a sample basis, with a complete listing of house- holds in Phase I serving as a frame in selecting the sample households for enumeration in Phase II. This was the first time that a quinquennial census of the country was undertaken. Except for the 1960 and 1970 censuses, Philippine censuses had been undertaken at irregular intervals ranging from 9 to 21 years between censuses. As stated in Presidential Decree No. 650 which authorized the undertaking of the 1975 Integrated Census, there was a "demand for more recent census data on the Philippine population to fulfill data requirements for national, social and economic planning and the formulation of national and sectoral policies such as in manpower development and utilization, health, education, housing, and population control." (ii) General Concepts and Terms For the purposes of the ICPEA, clear definitions have been adopted for barangays, urban and rural. These have been broadly similar to the defi- nitions in the 1970 census and the integrated surveys of households (see Annex 1 for the definitions). (iii) Method of Enumeration and Sampling Design The usual procedure of enumerating the population by means of per- sonal interview was followed in Phase I. Every household within the enumer- ator's assigned barangay was visited and a responsible member of the house- hold, termed the respondent, was interviewed. Information on all items in the population schedule was obtained for each household member, except for those not falling within the age group required for a particular item. Data on technical/vocational skill, migration, fertility and usual occupation/ -52- ANNEX 2 Page 2 activity were collected on a 100% basis (unlike the 1970 Census when such data. were obtained on a 5% sample). A listing of all household heads in all barangays throughout the country was undertaken in Phase I and used as a sampling frame for the 1975 ICPEA Phase II. From the listing of households in each barangay, a 10% sample in all urban barangays and a 5% sample in all rural barangays were selected for personal interview. Each barangay was treated independently. Using a random start for every barangay, sample households within each barangay were selected systematically. A total of 454,844 sample households were finally selected and interviewed. The ICPEA Phase II Questionnaire is in the form of an eight-page booklet. Apart from the household identification block (page 1), it has two main parts: Part I - Data for Individual Household Members. (Part I-A: Personal Characteristics, Educational Attainment and Technical Skill; and Part I-B: Economic Activities of Persons 10 years old and over.) Part II - Data on Farming or Truck/Home Gardening, Livestock and Poultry Raising, Fishing, Manufacturing/Processing and Other Economic Activities Part I-A asks questions on the names of household members, their relationship to the household head, their sex, age, marital status, highest grade completed or degree obtained, and vocational or technical skills. Part I-B asks questions on the economic activities/enterprises during 1975 of each household member 10 years old and over, and collects data on class of worker, number of hours worked (averag-t number of hours per week and number of weeks) and net receipts or yearly salary/wage (in cash and kind) in each activity (farming/gardening, livestock/poultry, fishing, manufacturing/processing, other activities). Part I-B(2) also asks for the occupation in activities not operated by any member of the household, e.g., wage and salary workers (lawy- ers, carpenters, etc.), students, housekeepers, pensioners, domestic helpers, etc. Part II is subdivided into five sections: A, B, C, D, and El efer- ing to the different enterprises operated by members of the household Each section, except E, calls for data on production and disposal of produce, cost of operation and other expenses, equipment and facilities used in 1975, and 1/ A: Farming or truck/home gardening. B: Livestock and poultry. C: Fishing. D: Manufacturing and processing. E. Other economic activities. 53 - ANNEX 2 Page 3 net income. Section D also asks for the number of non-household members who are engaged on a contract or consignment basis. (iv) The Income Data of ICPEA Phase II The 1975 ICPEA Phase II was conducted to gather data on economic activities which are broadly classified into own-household operated activi- ties. The detailed questions on these activities would have helped respon- dents to recall income from different sources, and thus helped achieve a better coverage of income data. However, some questions on inheritance and rental from agriculture and owner-occupied houses - income items not normally considered as "activities" but as income-earning "assets" and imputable to higher income groups - were not included. Furthermore, there was no specific question on the imputed rental value of owner-occupied housing in ICPEA Phase II (unlike the ISHs). A single questin in Part I-B simply asked for "Pension, Gifts, Winnings, Dividends, etc." Total Family Income in the 1975 ICPEA Phase II (and hence also in the 1975 FIES) is defined as aggregate income in cash or in kind of family members from work, and from sources other than work. This includes net receipts from activities not operated by the household, such as wages and salaries and receipts from pensions, gifts, winnings, dividends, etc. Items omitted from the design of the survey include: the rental value of owner- occupied house; borrowings of all kinds; interest from bank deposits; employ- ment benefits like hospitalization; scholarship; capital gains due to changes in the value of fixed assets; increase in stocks; and money received from sale of personal assets. By excluding such imputed income, the 1975 ICPEA Phase II reflects lower mean incomes than it would have done if these items were also investigated with the economic activities of the households. Also, forest activities such as firewood-gathering, and other income-in-kind items that are classed as family-sustenance activities, may have been under-reported owing to the survey's primary concern with economic activities yielding money incomes. Given that the focus of the 1975 survey was on the economic activ- ities of households, a better income coverage may be expected of the household operated activities, such as household manufacturing and family agriculture, than of family-sustenance activities. -54- ANNEX 3 Page 1 THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES The 1981 Census of Agriculture and Fisheries (i) The Census Objectives The 1981 Census of Agriculture and Fisheries (CAF) is a comprehen- sive collection and compilation of statistics from the agricultural and fish- eries sectors of the Philippines. The specific objectives of the Census were: (a) To determine the structure and characteristics of farms such as their size, tenure, land use, area planted to crops and production, number of livestock, poultry, etc.; (b) To determine the number and distribution of household enterprises engaged in fishing, and to gather information on the operation of these household enterprises; (c) To provide a benchmark for improving the reliability of current survey estimates; (d) To provide a sampling frame for other statistical undertakings; and (e) To provide basic data for national and local development planning. The 1981 CAF was conducted in two phases. Phase I involved the listing of households and household-operated activities in all barangays in the country, and the gathering of selected information on agricultural opera- tions such as the area of farms and arable lands, area under permanent meadows and pastures, and inventory of livestock and poultry. The listing of house- holds from which the farms were identified served as the frame for the second phase of the Census. Phase II involved the enumeration of sample farms and sample fishing operators to gather more detailed information. (ii) General Concepts and Definations Agricultural operation covers the following: (a) the growing of crops such as palay, corn, vegetables, fruits, nuts, and other agricultural products; (b) the raising of livestock such as carabaos, cattle, horses, hogs, etc., and the production of livestock products such as fresh milk; (c) the raising of poultry such as chickens, ducks, geese, etc., and the production of poultry products such as eggs; and (d) other agricultural activities such as cultivation of ornamental plants, trees, and bushes for sale; mushroom, earth- worm, silkworm, and honeybee culture. An operator/self-employed in agricultural operation is a person who takes the technical and administrative responsibility of managing a farm or a holding. He is responsible for making day-to-day decisions in operating the farm or holding, including the management and supervision of hired labor. The 55 - ANNEX 3 Page 2 farm operator may work on the Land alone or with members of his household, or he may not work on the land but may employ others to work on the land. He may or may not be the owner of the land. A farm or a holding is any parcel or parcels of land used wholly or partly for agricultural operations and operated as one technical unit by one person alone or with others regardless of title, legal form, size and loca- tion. A farm or a holding operated as one technical unit is one where the parcels of land are operated under a single management and where the financial resources for the operation come from the same management. Specifically, a farm is defined as a holding which satisfies any of the following condi- tions: (a) at least 1,000 square meters are devoted to crops; (b) at least 10 hectares are under permanent meadows and/or pastures; (c) at least 10 heads of large animls are raised, such as carabao, cattle, horse, regardless of age and kind; (d) at least 20 heads of small animals are raised such as hogs, goat, sheep, regardless of age and kind; (e) at least 100 heads of poultry are raised, regardless of age and kind; (f) at least 50 heads of rabbits are raised; and (g) a combination of land areas and animals are raised which amount to at least 10 agricultural units. One agricultural unit is equivalent to one-tenth of the cut-off in each of the aforementioned cate- gories. A parcel of the holding or farm is any piece of land entirely sur- rounded by water, road, river, canal, etc., which is not contiguous to any other part of the holding or farm. Physical area refers to the actual measured area of the land, reporting the area only once regardless of the number of times it has been planted during the reference period. Effective area is the area of land used for agricultural operations, reporting the area as many times as it has been planted and harvested with the same or different crops during the reference period. (iii) Sampling Design The sampling frame was based on the result of the listing. The farms were identified from the listing and were classified into two strata, Large and small farms. A large farm is one which satisfies any of the follow- ing conditions: (a) area devoted to crops is at least 4 hectares; (b) area under permanent meadows/pastures is at least 20 hectares; (c) at least 10 Large animals are raised; (d) at least 50 small animals are raised; (e) at Least 500 poultry are raised; and (f) at least 50 rabbits are raised. Farms which do not satisfy any of these conditions are considered as small farms. (The parcel or parcels of land that comprise a farm do not necessarily have to be located in the same municipality, province or region. For farms whose parcels are located in different municipalities, one questionnaire is alloted for the information for parcels in the same municipality. Thus, the number of questionnaires for a farm is equal to the number of municipalities in which the farm's parcels are located.) All large farms were enumerated in the Census of Agriculture, but only 10% of the small farms. Selection of samples by stratum was done at .the municipality level with the farm as the sampling unit. -56- ANNEX 3 Page 3 Phase II of the Census of Fisheries involved the enumeration of sample fishing operators. From the Phase I listing which was used as the sampling frame, a stratified list of all fishing operators or self-employed in fishing was made. Fishing operations were classified into three strata: municipal or household fishing, commercial fishing, and aquaculture. Munici- pal fishing covers fishing operations without the use of a boat, or with the use of a boat of 3 tons gross or less. It is also called household or suste- nance fishing. Commercial fishing covers fishing operations using a boat or boats of more than 3 tons gross. Aquaculture refers to the culturing of aquatic products such as fish, oysters, seaweeds and other aquatic resources in sea, lakes, or rivers. In Phase II, a 10% sample of municipal fishing operators was taken, but a 100% coverage of operators in commercial fishing and aquaculture. Selection of the sample by stratum was done at the level of municipality with the fishing operation as the sampling unit. (.v) The Agriculture Questionnaire (CAF Form 3) The Agriculture Questionnaire (CAF Form 3) is constructed by farm and by municipality/city where the farm/parcels are located. If a farm is located in two or more municipalties/cities, the operator is interviewed separately and a CAF Form 3 is accomplished for the parcels located in each municipality/city. The Agriculture Questionnaire has the following main sections: Section I - Land Utilization (size of panels, land use in 1980, pastureland) Section II - Size and Tenure of Farm Section III - Crops (interplanted, succession, temporary. permanent, household manufactured sugarcane and coconut products) Section IV - Livestock, Poultry and Other Domesticated Animals Section V - Agricultural Power and Ceneral Transport Facilities and Farm Equipment (source of power, means of transport and farm equipment and facilities) Section VI - Irrigation (sources and methods) Section VII - Fertilizers and Soil Dressing, Other Agricultural Chemicals and High-Yielding Varieties of Palay Section VIII - Other Products (e.g., firewood, coconut shells, nipa) Section IX - Association of Agriculture with Other Industries Setion X - Paid Workers in Agriculture Section XI - Farm Population Section XII - Source of Income (all categories of income are covered) ~7 ANNEX 3 Page 4 (v) Municipal and Commercial Fishing Questionnaire (CAF Form 4) Municipal fishing covers fishing operations using boat(s) of 3 tons gross or less or raft(s) or without boat. Data on municipal fishing is based on sample fishing households where at least one member did some fishing between January 1 and December 31, 1980. Commercial fishing covers fishing operation using boat(s) of more than 3 tons gross anytime from January 1 to December 31, 1980. This includes single proprietorship or household-operated fishing whether offshore or deep see. Likewise covered are managed fishi-ig enterprises operated by corporations, cooperatives and institutions. The Municipal and Commercial Fishing Questionnaire has the following main divi- sions: Part I - General Identification Portion Part II - Section I - Data on Individuals (Fishing houshold members, their characteristics, net receipts, salaries, wages, etc.). Part III - Section II - Data on Fishing Operations (Form of organization, employment, fishing gear used in 1980, fishing crafts used-in 1980, operating cost, catch and gathering in 1980, catch by fishing ground and landing area, problems encountered and information on credit programs). (vi) Questionnaire on Prices of Agriculture and Fishery Products (CAF Form 5) Item 1 - Price of Agriculture and Fishery Products (Name of Product, Local Unit, Price per Unit, Weight in Kilograms) Item 2 - Fish Landing Facilities (e.g., Type of Wharf/Fish Landing Area). (vii) Summary Results from Phase I of the 1981 Census of Agriculture Phase I of the 1981 Census of Agriculture identified 3,438,552 farms in the country in 1980. this represents an increase of 46.04% over the figure registered in the 1971 Census of Agriculture. The increase in the number of farms is evident not only at the national level. Lower level aggregates indi- cate the same direction. In 1980, the 9.03 million hectares covered by farms comprised 30.11% of the country's total land area. Region IV accounted for the bulk of the total farm area followed by Region XI and Region V. On the other hand, NCR had the least followed by Region I. Compared with 1971, the country's total farm area in 1980 increased by 540,619 hectares or 6.36%. The expansion in total farm area did not keep pace with the 46.04% growth in the number of farms from 1971 to 1980, and thus led to a decrease in average farm size from 3.61 hectares in 1971 to 2.63 hectares in 1980. At the national level, compared with 1971 the registered arable land area in 1980 fell proportionately and absolutely despite the expansion in total farm area. The positive contribution by the regions in Mindanao were -58- ANN 3 Page 5 offset by reductions in farm area in regions in Luzon. The increase in the area under permanent crops from 1971 to 1980 was responsible for the increase in the total farm area. During the same period, a decrease was observed in the area under permanent meadows and pastures. (viii) CAF 1981 Tabulation Plan for Agric,Ature Statistical tables for agriculture are grouped under four classifi- cations, namely, consistency tables, preliminary tables, final tables by residence of operator, and final tables by Location of farm parcels. The consistency tables have been prepared mainly to assess the validity and correctness of the data that are being processed and tabulted. There will be two sets of consistency tables. One set will be based on the unedited files and the other set will be based on the edited files. Since the machine processing of the questionnaires is done by pro- vince, tabulations by location of farm parcels cannot be completed straight away. The preliminary tables, therefore, are designed to determine the con- tribution of farm parcels located outside the residence of the operator prior to the preparation of the final tables. This will indicate the difference expected between the final tables by residence of operator and by location of farm parcels. Statistical tables by residence of the farm operator will include all the data from the farm of an operator regardless of the location of farm parcels. The data will be accounted to the residence of the farm operator. Only data from farms located in the same province and municipality will be included in the tabulations by location of farm parcels. Hence, data from farm parcels of a farm that are located outside the residence of an operator will be included in the province and municipality where the parcels are located. One of the major requirements of statistical users relates to data that are necessary for assessing poverty. Among the identifed correlates of poverty are small farm size, limited access to land or high tenancy rate, low physical productivity, low-value crop mix, lack of non-farm employment, remoteness of area, and lack of infrastructure. Given the objective of monitoring the welfare of the poor, there is a need to measure these elements and identify the affected population. Useful information can thus be provided by tabulating farm characteristics according to residence of operator. There is also a need to define land-use patterns, cropping intensi- ties, crop mix, available technology, available inputs, and levels of agricultural productivity. Monitoring these by location will help provide information towards the improvement of land use, marketing, etc. The tables by residence of operator will consist mainly of cross- tabulations of the various farm-items with farm size. The tables by residence of operator are to be printed by province but, to reduce printing cost, sum- maries at a lower level of geographical classification will be stored on tape. For tabulations by location of farm/parcel, the definition of farm 59 ANNEX 3 Page 6 adopted in the Census will be reformulated to include only parcels of agricul- tural land that are located in the same municipality and operated as one techno-economic unit. Since processing of the questionnaires will be done by province, final tables by location of farm/parcel cannot be produced straight away. Preliminary tables which consist only of data from farms/parceLs loca- ted within a province will be released first, and data from farms/parcels located in provinces outside the residence of the operator will be tabulated after the processing of data from the individual provinces has been completed. -60 - ANNEX 4 Page 1 THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES Sources of Employment and Wages Data A. Employment Data (i) Household and Establishment Surveys A number of sources provide employment data. The most continuous and fairly consistent sources are the series of household surveys conducted since 1956 by the NCSO. (See Table 1 for a listing of the sources.) There were also establishment surveys running parallel to the household surveys. In the years 1956-71, the establishment surveys covered the manufacturing sector only. Since 1973, establishment surveys covered almost all the sectors of economic activity. In 1975 and 1978, censuses of establishments were also held. Table 2 shows the employment estimates from the establishment surveys and the household surveys for 1978 and 1980. In respect of agriculture, construction, transport and personal and private services, the census of establishments estimates of employment are very low compared to household survey estimates. This is so for more sectors when the 1980 household survey employment estimates are compared to establishment survey estimates. Clearly, establishment surveys cannot and are not meant to provide employment data. -61- ANNEX 4 Page 2 Table 1: SOURCES OF DATA ON EMPLOYMENT, 1956-83 A. Household Surveys 1956-64 Philippine Statistical Survey of Households (May and October) 1965-73, except 1970 Bureau of Census and Statistics Survey of Households (May and October until 1969 and four times a year during 1971-73) 1974-75 National Sample Survey of Households (2 or 4 selected months) 1976-83 Integrated Survey of Households (generally 2 or three quarters). In 1979, all four quarters were surveyed. B. Establishment Surveys 1956-71 Annual Surveys of Manufacturing 1973-74 } 1976-77 } Annual Survey of Establishments 1979-83 } C. Economic Censuses 1960, 1971, 1981 Censuses of Agriculture 1961, 1967, 1972 Industry and Services 1975, 1978 Censuses of Establishments -62- ANNEX 4 Page 3 Table 2: EMPLOYMENT ESTIMATES FROM THE HOUSEHOLD SURVEYS AND ESTABLISHMENT SURVEYS 1978 AND 1980 Employment (in thousands) Percent of total 1978 1980 employment covered by Househld. Househld Estab. 1978 1980 survey Estab. survey survey estab. estab. Sector of Activity (4th qr) Survey (4th qr) (Average) survey survey Agriculture, Forestry and Fishing 8,702 392 8,894 261 4.5 2.9 Mining and quarrying 67 59 130 57 88.1 43.8 Manufacturing 1,916 1,465 1,850 1,177 76.5 63.6 Construction 480 80 600 110 16.7 18.3 Utilities 55 - 55 36 - 65.5 Transport, communications 681 261 725 215 38.3 29.7 Finance and insurance 308 178 308 116 57.8 37.7 Wholesale and retail trade 1,745 964 1,799 NA 55.2 NA Personal and private services 2,660 559 2,791 693 21.0 24.9 Source: NCSO, ISH 1978, ISH 1980, Census of Establishments 1978, and Establishment Survey, 1980. (ii) The Integrated Household Surveys These surveys were conducted starting from the third quarter of 1976. They replaced the earlier labor force surveys. Compared to the labor force surveys, there were two changes: These relate to the reference period and the age cut-off. Both are explained as follows: "In a developing country like the Philippines where the economic activities of many persons are so varied and unstable, a 'past week' as a reference period does not present a very vivid picture. It is the contention of planners that the 'past week' as 1/ See Annex 1 for a detailed discussion on the questionnaire, sampling design, concepts, etc. -63 - ANNEX 4 Page 4 a reference period is suited to industrialized countries and not for develop- ing countries. It is believed that the 'past quarter' as a reference period in the country's survey will afford the users a more realistic indication of the nation's economic and social wellbeing. On the other hand, the concept on labor force has been sharpened to include only persons who are 15 years old and over instead of covering persons who are 10 years old and over. Such2 modification was made to conform with the Labor Code of the Philippines."- Past-week and Past-quarter Implications: To evaluate the implications of moving from the old concept of "past week" to the new concept of "past quarter" as a reference period, a comparative table using both reference periods was prepared from the 1976 third quarter survey (see Table 3). Table 3: COMPARATIVE STATISTICS ON EMPLOYMENT STATUS OF PERSONS 15 YEARS OLD AND OVER BASED ON WEEKLY AND QUARTERLY REFERENCE PERIODS, THIRD QUARTER: 1976 Past week Past quarter Number Percent of Number Percent of Employment status ('000) total ('000) total Household population 15 years old and over 24,837 24,837 In the labor force 14,776 59.5 15,018 60.5 Employed 13,841 93.7 14,238 94.8 Unemployed 935 6.3 780 5.2 Not in the labor force 10,061 40.5 9,820 39.5 It can be gleaned from the table that with the "past quarter" as a reference period, the number of persons in the labor force and the level of employment are higher than the levels recorded with the "past week" as a reference period. The increase which is quite substantial can be partly explained by persons who were in unstable, irregular, seasonal, or short-time jobs which could not be accounted for in a very short reference period as the "1past week". (iii) Other Surveys of Employment Data MOLE-HIM: The Ministry of Labor and Employment (MOLE) started in June 1984 the first survey of households in Metro Manila. The survey is known 2/ See ISH, 1976 -64- ANNEX 4 Page 5 as the Household Information Monitor (HIM) and after the pilot in June, was repeated in August. The principal purpose of HIM is to generate quick information on employment and unemployment. The reference period is a week (Sunday to Saturday). The population surveyed is the household population. Labor force includes all those who are 15 years and older who have worked in the reference week or looking for work. Underemployed persons are those wanting additional work. In the first survey of June 1984 for Metro Manila, the total population (15+) covered was only 716 persons. The August survey covered 758 persons. These are very small sample sizes. The surveys found 14.4% unemployment rate in June and 4.2% in August. MOLE: Monthly Report on Employment and Labor Turnover: The reports are based on short questionnaires sent for each month to several establish- ments to find out the accessions and separations of employment. B. Wage Data (i) Recent History of Data Collection Prior to 1980, the National Census and Statistical Office, the Central Bank, and the Wages Commission, were involved in the collection of data on compensation, earnings and wages, especiaiiy in the non-agricultural sector (BAEcon collects the agricultural wage data). Statistics on average earnings and compensation are produced by the NCSO/NEDA and gathered through the Integrated Quarterly Survey of Establish- ments (IQSE); the Annual Survey of Establishments (ASE); the Census of Estab- lishments (CE); the Integrated Survey of Households (ISH); and Survey of Key Enterprises in Manufacturing (SKEM). Wage rate data, on the other hand, is found to be inadequate, if not totally lacking, in the statistics produced by NCSO. The only significant series that draws the trend and levels of monthly wage rates for selected occupations was the Central Bank's Survey of Laborers in Industrial Establish- ments in Manila and Suburbs which was discontinued in 1980 and the Survey of Base Compensation, Personnel Practices and Fringe Benefits conducted by the then Wage Comission, now the National Wages Council, which was also stopped in 1980. The Presidential Letter of Instruction 1082 of 1980 entrusted the HCSO (and NEDA) with the responsibility to immediately develop a new series of indicators on production, employment, earnings and wage rates. The NCSO surveys of establishments could help gather all the needed data except the wage rates. To correct this deficiency, NEDA in May 1983 issued a circular allocating the functions and responsibilities for the collection, processing and dissemination of wage rates to the National Wages Council. NWC launched the quarterly occupational wage survey in June 1984. In the near future, therefore, one could hope to get a consistent data set on the non agricultural sector wages from the NWC. For agricultural wages, as in the past, BAEcon series will continue to be the single source. 65 ANNEX 4 Page 6 (ii) National Wages Council's Wage Surveys Survey of Base Compensation, Personnel Practices and Fringe Benefits: This is conducted within the Metro Manila area. Data obtained include average monthly compensation (overall, by employment size, and for unionized and non-unionized firms), summary of personnel practices and fringe benefits, and occupational wage data (ranges by industry). Forty-three surveys have been conducted covering 27 industries excluding the sectors of agriculture, forestry, fishing, utilities and commerce. The surveys were held during August 1974 up to 1980 and then the survey was suspended when the former Wage Commission was abolished. Occupational, Employment and Wages Survey: Conducted jointly by the NWC and the Labor Statistics Service (LSS) in March 1983 for the first time, the aim is to obtain data by region and by industry. Data are being processed. Occupational Wage Survey: This is a nationwide quarterly survey and covers 35 non agricultural industries and 1,742 establishments. The first survey started on August 20, 1984, and covered the second quarter of 1984. Industry-Occupational Survey on Employment and Wages This was conducted once in 1977 and was not yet repeated again. The survey was spon- sored jointly by the NWC, LSS and a number of other agencies. (iii) Wage Data 1970-80: Central Bank Series The average daily basic wage rates of laborers in industrial establishments in Metro Manila are derived from data furnished by 102 estab- lishments covering automobile dealers, brokerages, lumber yards, furniture manufacturers, printing and publishing, transportation, and construction firms, cigar and cigarette factories, painting contractors and machine shops. These rates represent the daily basic average rates of all classes of workers in each group. Data are available up to and including 1980. (iv) BAEcon Agricultural Wage Data The data are obtained from surveys of farm labor engaged in palay, corn, coconut, sugarcane and vegetables. Recent (e.g., 1984) surveys involved about 9,000 farmers nationally, drawn from a sample of 1,600 barangays out of a larger number covered in the Rice-Corn Surveys. Selection of the barangays is purposive, to ensure coverage of the specified crops. Information on labor use, alternative jobs of farm labor, use of family labor, etc., were collected on a short questionnaire, but tabulations were available only on wages. Daily wage rate data were obtained for laborers receiving and not receiving meals and snacks. -66 - ANNEX 5 Page 1 THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES Poverty Lines in Earlier Philippine Research (i) Poverty Threshold Based on the Rice-Wage Formula (1948 and 1971) In 1948, the then Bureau of Census and Statistics (BCS) in cooperation with the Department of Social Welfare (DSW) developed the Rice- Wage Formula, intended to measure at any time or place the minimum earnings that a low-wage earner should receive to be ab to clothe, feed, shelter and educate a family of 5 with decency and health.- The estimated rice equivalent of an expected threshold wage was 145 gantas. In 1971 the family size assumption changed from 5 to 6. From that year on, the evaluated rice- wage was 165 gantas of rice per month. A poverty threshold can be found by multiplying 165 with the price of rice. One limitation of this exercise was, the estimate did not take into account any nutritional norms in regard to the food consumption data employed as part of the consumption basket. (ii) Asia Research Organization - Gallup International (1971) 2/ The Asia Research Organization, for Gallup International conducted a survey of urban areas and elicited the response to the following question: "How much income per week do you think a family of four (husband, wife, 2 children) would need for health and comfort?" The majority of over 40% respondents perceived the need to be between P 50-P 100. The mid-point of this range being P 75, converting the family size to 6, the annual urban threshold is P 5,850 per family of six. (iii) FNRC - Food Threshold (1973) The Food and Nutrition Research Center formulated a breakfast, lunch, snack and dinner menu that would be both tasty and adequate in recommended levels of nutrients. For Metro Manila, the basket was priced and for May 1974, the food budget for a reference family of 6 was F18.12 per day. (iv) ABRERA Food and Total Thresholds (1961-74) 31 On the basis of the FNRC food threshold for Metro Manila for 1974, using the food price indices over time and space, Abrera derived the three 1/ See Felisa R. Barreto, An Analytical Study of the Rice-Wage Formula, BCS Technical Paper, May, 1973, quoted in Abrera (1976). 2/ See Abrera (1976). 3/ Abrera (1976). -67- ANNEX 5 Page 2 regional annual food thresholds for the different years. These are given below. Year Manila Other urban Rural 1961 1,789 1,432 1,253 1965 2,476 1,981 1,733 1971 4,284 3,428 3,000 1974 6,633 5,306 4,643 The food-thresholds for the bench-mark year was converted to a total threshold based on a food expenditure to total expenditure ratio of 0.6. The "all items" consumer price indices were used to obtain the total threshold over time for the three regions. The thresholds are given below. Year Manila Other urban Rural 1961 3,590 2,872 2,513 1965 4,497 3,597 3,147 1971 7,203 5,762 5,000 1974 10,550 8,844 7,738 (v) Tan-Holazo Least Cost Threshold (1971) 4/ I E.A. Tan and V. Holazo of the University of the Philippines in a paper brought out in December 1978 employed linear programming to determine the least cost consumption basket that would provide the recommended daily per capita levels of calories (2106 k. cals), proteins (50 gms) and other nutrients as well as other basic needs. The result was a series of regional poverty lines with an implied average per capita poverty threshold of P 360 per year. (vi) The World Bank Poverty Line (1971) The Bank methodology up until recently was to find the food expendi- ture and the food basket of the 20th percentile of the income/expenditure dis- tribution, compute the calories obtained by that group, and use the 4/ E.A. Tan and V. Holazo (1978). -68 - ANNEX 5 Page 3 expenditure-calories relationship to estimate the food expenditure needed to obtain a norm level of calories. Based on the food to non-food ratio of the 20th percentile, the food expenditure estimate ia blown up to an overall poverty threshold. This procedure gave an estimated per capita poverty line of P 560 per year.5 In the Bank's report on poverty in the Philippines, a poverty line of P 500 was adopted. (vii) DAP Metro Manila Poverty Threshold (March-May 1981) The Development Academy of the Philippines has made a study of social well-being in Metro Manila. For this purpose, 500 household heads were interviewed during March-May 1981. The poverty threshold income is an indica- tor of the cost of living as perceived by the different household heads. The median threshold was between P 1,300-P 1,500. The level of P 1,400 was the mid-point of this range. (The average value of the responses was P 1,920). (viii) National Wages Council Necessities Line for Metro Manila (1981) The necessities line was P48.10 per day for a family of six. Detailed explanation is not available. The necessities line (obviously) is based on a basket of goods. It is almost certain that a nutritional norm has been taken into account. (ix) Ministry of Social Welfare Cut-off to Determine Needy Families (1983) The cut-off was P5,400/family/annum in 1983. The basis for the cut- off is not known. About 22.6% of the Filipino families are found to be needy in 1983. The families in need numbered 2 million and the needy population numbered 12 million in 1983. 5/ See World Bank, Aspects of Poverty in the Philippines (Volume II), 1981, p. 110. 6/ Source: Mangahas (December 1981) pp 14-15. ANNEX 6 Page 1 THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES Poverty Lines Based on Rice, Other Food Items and Non-Food Expenditures There are basically three principal sources of calories in the diet of most Filipinos and other peoples of South East Asia. They are rice, fish and meat. Rice and fish are relatively even more important sources of calories than meat. It can be shown by direct comparison of expenditure per calorie that rice is the cheapest source of calories. Consumption of 16 kg of rice per month will provide a person with an intake of about 2100 k. cal per day on the average, and the level of rice consumption is al? sufficient to obtain the recommended level of protein (about 50 gms/day).- The value of 16 kg of rice per month in 1983 was P 50.56 in urban areas and P 49.76 in rural areas (see prices in Table 1 at the end of this Annex). From the 1982 Food and Nutrition Survey, it was found that the compo- sition of totaL food expenditure for the country as a whole was 40% for cereaLs, 30% for meat, fish and poultry and 30% for milk, fruits, vegetables, sugar, fats and oils etc. If meat, fish and poultry are ignored, cereals constitute about 60% of total food expenditure. From the respective rural and urban rice expenditures of P 49.76 and P 50.56 and the cereal share in total food expenditure, the computed minimum required per capita monthly food expend Pures are P 82.93 and P 84.27 respectively for rural and urban areas.- The next step in the construction of poverty lines is the blowing up of the food threshold to a total threshold. After eliminating beverages, tobacco, taxes, gifts, special purchases and miscellaneous goods, the pattern of consumption according to the 1975 FIES indicated that food expenditure amounted to 62.3% of total expenditure in the rural areas and 51.3% in the urban areas. Clearly, since these ratios are averages across all income groups, in the relatively lover income groups, the percent spent on food would be substanpally higher. A plausible estimate would be 70% for rural and 60% for urban.- Based on these, the poverty threshold levels of per capita 1/ The computations were explained in detail in Poverty in Indonesia, 1970- 1980 Trends, Associated Characteristics and Research Issues, World Bank January 1984. 2/ Note that these levels of food expenditures might in fact provide slightly more calories and proteins than the levels recommended. However, some of the excess might be lost in processing and storage. This and other similar issues could be looked into as and when an official poverty line might be adopted. 3/ Recent evidence from the World Bank Metro Manila Household Survey, 1983 indicates that food expenditure in total expenditure in Metro Manila amounted to 70% in the first two deciles of the population and 50% in the last two deciles. ANNEC 6 -70 - Page 2 monthly expenditures are P 118 for rural and P 140 for urban (fractions of pesos ignored). For extrapolating the 1983 poverty lines to the earlier years, appropriate price indices we requiced. The general consumer price index is not suitable for the purpose.- For instance, the general consumer price index more than doubled between 1975 and 1983. But, the rice price increased by 64% in the rural areas and 58% in the urban areas during that period. In the poverty line computed above, food has a heavy weight at 70% in rural areas and 60% in urban areas. Accordingly, the price deflators for the rural and urban poverty lines were obtained by taking a weighted average of rice price indexes and general consumer price indexes, with weights (0.7, 0.3 and 0.6, 0.4) equal to the food and non-food proportions in the poverty-line. The poverty-line price deflators are shown in Table 3. Poverty lines per capita per month for 1983 and the per family per annum poverty line for 1971-1983 are given in Table 4. The poverty Lines obtained in this annex and selected thresholds from earlier research are provided in Table 5 to facilitate comparison. Some of the earlier thresholds are much higher than those of this paper. 4/ The consequence of using the CPI may be illustrated either by taking 1983 as the benchmark or by taking 1975 as the benchmark. In the former case the 1983 benchmark poverty line gets relatively more heavily deflated than warranted by the poverty line consumption basket. The 1975 poverty lines will amount to P 3,339 (rural) and P 4,730 (urban), substantially lower than the estimates in Table 4 below. If 1975 is used as the bench mark, then, the CPI will over-inflate the 1983 poverty lines which will then be P 11,747 and P 15,359 for rural and urban areas respectively and much higher than the poverty lines in Table 4. -71- ANNEX 6 Page 3 Table 1: RURAL AND URBAN POPULATION, 1980 AND RETAIL PRICE OF RICE, 1975, 1978 AND 1983 BY REGION Region Population (000) Price of rice per kilo 1980 1975 1978 1983 National Capital Region Urban 5,926 1.90 2.10 3.10 Ilocos Urban 842 NA NA NA Rural 2,699 1.94 2.11 3.13 Cagayan Valley Urban 344 NA NA NA Rural 1,872 1.86 2.09 3.13 Central Luzon Urban 2,009 1.93 2.14 3.28 Rural 2,794 1.97 2.14 3.01 Southern Tagalog Urban 2,258 2.02 2.16 3.20 Rural 3,861 1.90 2.09 3.25 Bicol Urban 746 NA NA NA Rural 2,731 1.95 2.13 3.05 Western Visayas Urban 1,283 NA NA NA Rural 3,243 1.92 2.07 2.97 Central Visayas Urban 1,212 NA NA NA Rural 2,575 1.95 2.29 3.12 Eastern Visayas Urban 610 NA NA NA Rural 2,190 2.13 2.12 3.14 Western Mindanao Urban 431 NA NA NA Rural 2,097 2.16 2.16 3.18 Northern MIndanao Urban 733 NA NA NA Rural 2,026 1.94 2.14 3.19 Southern Mindanao Urban 1,124 NA NA NA Rural 2,223 1.97 2.14 3.05 Central Mindanao Urban 427 NA NA NA Rural 1,844 2.02 2.11 3.07 Urban Ia 17,945 1.93 2.12 3.16 Rural Ia 30,155 1.97 2.13 3.11 /a Prices are weighted averages of regional prices with population weights. In respect of the rural areas, the market price may be somewhat higher than the true cost to the consumers since they include those who consume their own produce. Source: NCSO - 72 - ANNEX 6 Page 4 Table 2: REGIONAL CONSUMER PRICE INDICES, 1975-1983 (1978 = 100) Region 1975 1979 3Qr 1980 3Qr 1981 3Qr 1982 3Qr 1983 3Qr Metro Manila 82.3 124.4 144.3 162.3 179.4 194.8 Ilocos 76.5 119.9 138.3 159.8 180.5 196.7 Cagayan Valley 79.9 118.6 141.4 162.3 173.0 186.8 Central Luzon 75.1 119.2 138.0 153.0 169.0 187.0 Southern Tagalog 75.2 117.8 136.3 154.4 174.1 187.8 Bicol 74.3 126.2 146.3 163.5 177.4 190.1 Western Visayas 80.9 121.8 144.5 159.9 175.4 190.8 Central Visayas 81.9 121.2 142.8 160.7 186.8 209.0 Eastern Visayas 77.4 125.3 141.9 160.3 172.1 180.6 Western Mindanao 77.1 113.9 132.7 150.0 165.6 181.2 Northern Mindanao 74.4 123.7 149.8 169.5 189.0 204.7 Southern Mindanao 75.1 121.8 141.8 161.5 174.1 188.5 Central Mindanao 73.2 121.2 139.5 160.2 169.7 185.3 Philippines 77.5 121.3 141.2 159.4 176.0 191.3 Urban 80.5 122.9 142.6 160.1 176.6 192.6 Rural 75.8 120.4 140.4 159.0 175.6 190.4 Source: Regional indices are from NCSO. Urban indices are weighted averages of Metro Manila, Cagayan Valley and Central Luzon, the weights being the urban populations of the three regions. Rural indices are derived as residuals from the Philippines indices and urban indices. -73- ANNEX 6 Page 5 Table 3: PRICE INDICES FOR DEFLATING THE 1983 POVERTY LINE Type of Price Index Rural Urban Food Price Index (3rd Qr 1983 = 100) 3rd Qr 1982 94.2 93.0 3rd Qr 1981 88.1 85.8 3rd Qr 1980 75.9 72.8 3rd Qr 1979 71.7 69.7 1975 61.1 63.3 1971 31.3 32.4 Consumer Price Index (3rd Qr 1983 = 100) 3rd Qr 1982 92.2 91.7 3rd Qr 1981 83.5 83.1 3rd Qr 1980 73.7 74.0 3rd Qr 1979 63.2 63.8 1975 39.8 41.8 1971 22.1 23.1 Deflator for Poverty Line (3rd Qr 1983 = 100) 3rd Qr 1982 93.6 92.5 3rd Qr 1981 86.7 84.7 3rd Qr 1980 75.2 73.3 3rd Qr 1979 69.2 67.3 1975 54.7 54.7 1971 28.5 28.7 Source: See Table 1 for example of rice prices and Table 2 for regional consumer price indices. The methodology for obtaining the deflator was explained on page 2 of this Annex. ANNEX 6 Page 6 Table 4: POVERTY LINES, 1975-1983 Rural Urban Per capita monthly poverty line, 1983 (3rd qr) P 118 P 140 Average Family Size (1980 Census) 5.2 6.3 Poverty Line per family per annum, 1983 (3rd qr) P 7,363 P 10,584 Poverty Line /a - 1982 (3rd qr) (Pesos) 6,892 9,790 - 1981 (3rd qr) (Pesos) 6,384 8,965 - 1980 (3rd qr) (Pesos) 5,537 7,758 - 1979 (3rd qr) (Pesos) 5,095 7,123 - 1975/b (Pesos) 4,570 5,514 - 1971/c (Pesos) 2,341 2,989 al Special price deflators for the poverty level consumption basket were used. See Table 3. b/ Average family sizes were 5.9 (rural) and 6.0 (urban) as in the basic data source. c/ Average family sizes were 5.8 (rural) and 6.2 (urban) as in the basic data source. 75 ANNEX 6 Page 7 Table 5: COMPARISON OF SELECTED POVERTY LINES/a Present Paper Rice equivalent/month 227 kg/Family of 6 - Rural 266 kg/Family of 6 - Urban Rice Wage Formula (NCSO, 1948, 1971) Per Month 165 gantas (approximately 430 kg) Annual/Family (in Pesos) Present Paper - Thresholds (for 1971) Rural: 2,341 Urban: 2,989 Abrera - DAP Study (for 1971) Rural: 5,000 Urban: 5,700-7,200 Tan-Holazo Least Cost Basic Needs Threshold (for 1971) 2,160 World Baak Poverty Report (for 1971) 3,000 Annual/Family (in Pesos) Present paper -- Thresholds (for 1981) Urban: 8,q65 DAP - Metro Manila Threshold (March-May, 19I) Netro Manila: 16,800 National Wages Council Necessities Line (1981) Metro Manila: 17,556 /a See Annex 5 for details of other poverty lines. - 76 - THE PHILIPPINES: RECENT TRENDS IN POVERTY, EMPLOYMENT AND WAGES STATISTICAL APPENDIX Table 1: PER CAPITA INCOME AND CONSUMPTION, 1970-1983 Annual percentage change Personal Consumer Real per capita Per capita Personal consumption price Consumption Per capita consumption Year income expenditure Population index Income expenditure income expenditure (P m1n) (P mn) ('000) 1970 33,106 29,552 36,800 0.4501 1,999 1,784 1975 89,727 76,165 42,200 0.7754 2,742 2,328 6.53 5.47 1976 106,211 87,120 43,350 0.8498 2,883 2,365 5.14 1.60 1977 121,869 102,626 44,540 0.9298 2,943 2,478 2.07 4.79 1978 139,094 118,846 45,760 1.0000 3,040 2,597 3.29 4.80 1979 164,243 146,577 47,000 1.1755 2,973 2,653 -2.20 2.15 1980 200,474 178,119 48,300 1.3898 2,986 2,653 0.46 0.02 1981 231,020 206,942 49,500 1.5706 2,972 2,662 -0.50 0.32 1982 256,726 234,486 50,800 1.7319 2,918 2,665 -1.80 0.13 1983 290,323 268,188 52,000 1.9053 2,930 2,707 0.42 1.56 -di - 78 - ANNEX 7 Page 2 Table 2: FAMILY INCOME DISTRIBUTION DATA, THIRD QUARTER, 1980 Rural Urban Urban & Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class Z of Z of Z of Z of Z of Z of (in pesos) families income families income families income Less than 500 16.6 2.1 3.2 0.18 12.2 1.1 500 to 999 36.3 8.8 10.6 1.27 27.9 4.8 1,000 to 1,999 63.4 26.9 29.6 6.77 52.4 16.1 2,000 to 2,999 78.4 43.5 45.9 14.4 67.7 27.9 3,000 to 3,999. 86.7 56.8 59.7 23.5 77.8 38.9 4,000 to 4,999 91.5 66.6 68.5 31.1 83.9 47.5 5,000 to 7,499 96.4 80.3 82.2 47.1 91.7 62.5 7,500 to 9,999 98.2 87.6 89.3 58.8 95.3 72.2 10,000 to 14,999 99.2 93.4 94.& 70.5 97.7 81.2 15,000 to 19,999 99.5 95.7 97.0 78.6 98.7 86.7 20,000 to 24,999 99.7 97.5 98.5 85.6 99.3 91.2 25,000 to 29,999 99.7 97.5 98.5 85.6 99.3 91.2 30,000 to 39,999 99.7 98.2 992. 89.6 99.6 93.7 40,000 to 49,999 99.8 98.5 99.4 92.1 99.7 95.2 50,000 to over 100.0 100.0 100.0 100.0 100.0 100.0 Total /a (5,840) (12.67) (2,838) (14.78) (8,677) (27.45) /a Households in thousands and income in billions of pesos. -79- ANNEX 7 Page 3 Table 3: FAMILY INCOME DISTRIBUTION DATA, FOURTH QUARTER, 1980 Rural Urban Urban & Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class Z of Z of Z of Z of Z of Z of (in pesos) families income families income families income Less than 500 13.1 1.7 3.0 0.2 9.8 0.9 500 to 999 32.2 8.1 9.4 1.2 24.8 4.6 1,000 to 1,999 61.1 27.3 27.9 7.0 50.3 16.8 2,000 to 2,999 77.6 45.7 44.6 15.5 66.9 30.1 3,000 to 3,999 86.5 59.7 59.4 26.1 77.7 42.3 4,000 to 4,999 91.3 69.4 68.4 34.3 83.8 51.2 5,000 to 7,499 96.8 84.5 83.4 52.9 92.4 68.1 7,500 to 9,999 98.4 90.5 90.5 65.4 95.B 77.6 10,000 to 14,999 99.6 96.8 95.5 77.4 98.2 86.8 15,000 to 19,999 99.7 98.0 97.6 84.9 99.0 91.3 20,000 to 24,999 99.8 98.7 98.5 88.8 99.3 93.7 25,000 to 29,999 99.8 99.1 98.7 90.2 99.4 94.5 30,000 to 39,999 99.8 99.1 99.2 93.7. 99.6 96.4 40,000 to 49,999 99.8 99.2 99.4 94.7 99.7 97.0 50,000 to over 100.0 100.0 100.0 100.0 100.0 100.0 Total /a (5,874) (12.93) (2,857) (13.87) (8,731) (26.79) /a Households '.n thousands and income in billions of pesos. -80- ANNEX 7 Page 4 Table 4: FAMILY INCOME DISTRIBUTION DATA, THIRD QUARTER, 1981 Rural Urban Urban & Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class Z of Z of Z of Z of Z of Z of (in pesos) families income families income families income Less than 500 13.0 1.5 3.3 0.2 9.8 0.8 500 to 999 31.8 6.8 9.1 0.9 23.8 3.6 1,000 to 1,999 58.6 22.5 24.5 4.6 47.3 12.8 2,000 to 2,999 74.0 37.2 40.7 11.1 62.9 23.1 3,000 to 3,999 83.0 49.2 53.6 18.3 73.2 32.5 4,000 to 4,999 88.0 58.0 63.1 25.2 79.6 40.2 5,000 to 7,499 94.8 74.0 78.1 39.9 89.1 55.5 7,500 to 9,999 97.4 82.9 86.1 51.1 93.5 65.6 10,000 to 14,999 99.0 90.3 92.9 64.5 96.8 76.2 15,000 to 19,999 99.5 93.4 95.7 72.4 98.1 81.8 20,000 to 24,999 99.7 94.9 97.4 78.6 98.8 85.9 25,000 to 29,999 99.8 96.4 98.2 82.0 99.1 88.4 30,000 to 39,999 99.9 97.3 99.2 87.3 99.5 91.7 40,000 to 49,999 99.9 97.9 99.3 88.2 99.5 92.5 50,000 to over 100.0 100.0 100.0 100.0 100.0 100.0 Total /a (5,978) (15.28) (2,916) (17.95) (8,894) (33.22) la Households in thousands and income in billions of pesos. -81- ANNEX 7 Page 5 Table 5: FAMILY INCOME DISTRIBUTION DATA, FOURTH QUARTER, 1981 Rural Urban Urban & Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class Z of % of X of Z of Z of % of (in pesos) families income families income families income Less than 500 11.0 1.21 1.9 0.1 8.0 0.6 500 to 999 28.4 5.9 6.9 0.7 21.3 3.0 1,000 to 1,999 55.3 20.4 18.3 3.2 43.2 11.0 2,000 to 2,999 71.6 35.1 32.7 8.5 58.9 20.5 3,000 to 3,999 80.9 46.8 45.1 14.8 69.3 29.3 4,000 to 4,999 86.6 56.1 55.2 21.4 76.4 37.1 5,000 to 7,499 94.5 73.3 72.7 37.2 87.4 53.4 7,500 to 9,999 96.7 80.4 81.9 48.9 91.9 63.1 10,000 to 14,999 98.9 89.9 91.4 65.R 96.5 76.6 15,000 to 19,999 99.4 94.6 95.4 75.9 98.2 83.7 20,000 to 24,999 99.6 95.3 97.7 81.4 98.9 87.3 25,000 to 29,999 99.8 96.5 98.6 83.9 99.1 89.1 30,000 to 39,999 99.8 96.5 98.6 88.7 99.5 92.2 40,000 to 49,999 99.8 96.7 99.0 91.2 99.6 93.6 50,000 to over 100.0 100.0 100.0 100.0 100.0 100.0 Total /a 8.948 36.00 la Households in thousands and income in billions of pesos. -82- ANNEX 7 Page 6 Table 6: FAMILY INCOME DISTRIBUTION DATA, THIRD QUARTER, 1982 Rural Urban Urban & Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class % of % of Z of % of Z of Z of (in pesos) families income families income families income Less than 500 7.8 0.74 1.78 0.07 5.83 0.38 500 to 999 23.2 4.35 5.4 0.42 17.3 2.2 1,000 to 1,999 49.5 16.0 18.4 2.9 39.2 9.0 2,000 to 2,999 66.1 28.7 32.3 7.32 54.9 17.2 3,000 to 3,999 76.4 39.8 44.4 12.6 65.8 25.2 4,000 to 4,999 82.9 48.8 53.3 17.6 73.1 31.9 5,000 to 7,499 91.9 65.6 69.4 30.2 84.4 46.4 7,500 to 9,999 95.6 75.2 78.5 40.3 89.9 56.3 10,000 to 14,999 98.4 85.7 88.3 55.4 95.0 69.2 15,000 to 19,999 99.1 89.8 93.0 65.6 97.0 76.6 20,000 to 24,999 99.4 91.6 95.4 72.3 98.0 81.1 25,000 to 29,999 99.6 93.2 96.5 76.1 98.5 83.9 30,000 to 39,999 99.7 94.4 98.3 83.8 99.2 88.6 40,000 to 49,999 99.7 95.0 98.7 86.1 99.3 90.1 50,000 to over 100.0 99.7 100.0 100.0 100.0 100.0 Total /a 6.161 19.72 2,995 23.50 9,111 43.22 /a Households in thousands and income in billions of pesos. -83- ANNEX 7 Page 7 Table 77 FAMILY INCOM4E DISTRIBUTION DATA, FOURTH QUARTER, 1982 Rural Urban Urban 6 Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class Z of 2 of l of Z of Z of Z of (in pesos) families Income families income families income Less than 500 4.5 0.4 0.8 0.0 3.3 0.2 500 to 999 19.5 3.9 4.8 0.4 14.7 2.0 1,000 to 1,999 46.6 15.7 14.8 2.3 36.1 8.6 2,000 to 2,999 63.5 28.5 27.5 6.4 51.6 16.5 3,000 to 3,999 74.9 40.2 40.9 12.2 63.6 25.1 4,000 to 4,999 81.7 49.4 50.6 17.7 71.3 32.3 5,000 to 7,499 91.5 67.0 68.5 31.7 83.8 48.0 7,500 to 9,999 95.6 77.5 79.0 43.3 90.0 59.1 10,000 to 14,999 98.4 87.6 88.9 58.5 95.1 71.9 15,000 to 19,999 99.3 92.3 93.4 68.4 97.2 79.4 20,000 to 24,999 99.4 93.3 95.8 75.2 98.1 83.5 25,000 to 29,999 99.5 94.1 97.3 80.3 98.6 86.6 30,000 to 39,999 99.7 95.5 98.4 84.8 99.1 89.7 40,000 to 49,999 99.8 96.9 98.8 87.4 99.3 91.8 50,000 to over 100.0 100.0 100.0 100.0 100.0 100.0 Total /a 9A6E 41.72 /a Households in thousands and income in billions of pesos. -84- ANNEX 7 Page 8 Table 8: FAMILY INCOME DISTRIBUTION DATA, THIRD QUARTER, 1983 Rural Urban Urban & Rural Quarterly Cumulative Cumulative Cumulative Cumulative Cumulative Cumulative Income class Z of Z of % of Z of Z of X of (in pesos) families income families income families income Less than 500 6.6 0.6 1.1 0.0 4.8 0.3 500 to 999 21.9 4.3 4.8 0.4 16.3 2.3 1,000 to 1,999 49.8 17.5 16.8 2.6 39.0 9.3 2,000 to 2,999 66.6 30.8 31.1 7.0 54.9 17.6 3,000 to 3,999 77.1 42.8 43.3 12.3 66.1 25.8 4,000 to 4,999 83.7 52.3 52.5 17.5 73.6 32.9 5,000 to 7,499 92.5 69.5 70.0 31.0 85.2 48.0 7,500 to 9,999 96.0 79.1 79.3 41.0 90.7 57.8 10,000 to 14,999 98.3 88.0 89.0 55.6 95.4 70.0 15,000 to 19,999 99.0 91.8 93.7 65.8 97.4 77.3 20,000 to 24,999 99.3 94.4 95.7 71.3 98.3 81.6 25,000 to 29,999 99.4 95.2 97.0 75.7 98.8 84.4 30,000 t-t 39,999 99.4 96.4 98.1 80.5 59.3 87.6 40,000 to 49,999 - 99.4 97.2 98.8 84.3 99.5 90.1 50,000 to over 100.0 100.0 100.0 100.0 100.0 100.0 Total /a 6,254 19.33 3,074 24.54 9,328 43.87 /a Households in thousands and income in billions of pesos. -85- ANNEX 7 Page 9 Table 9: NCSO AND WORLD BANK DATA ON METRO MANILA INCOME DISTRIBUTION, FIRST QUARTER 1983 NCSO data for WB data for WB data 1st quarter January 1983 average Monthly income Z of % of No. of % of income % of (Qty income r 3) families income households households January 1983 income Less than 167 0.3 /a 7 0.4 64 /a 168- 333 0.5 /a 5 0.3 282 /a 334- 667 5.3 0.8 49 2.9 554 0.4 668- 999 11.9 3.3 121 7.2 851 1.5 1,000- 1,333 12.6 4.5 182 10.8 1,142 2.9 1,334- 1,667 9.0 4.3 173 10.2 1,497 3.7 1,668- 2,499 19.1 12.2 311 18.4 2,029 8.9 2,500- 3,333 13.7 12.8 258 15.3 2,844 10.4 3,334- 4,999 13.7 17.8 225 13.3 4,049 12.9 5,000- 6,667 6.1 11.0 131 7.8- 5,692 10.6 6,667- 8,333 2.9 8.0 73 4.3 7,536 7.8 8,334- 9,999 1.5 4.2 39 2.3 9,150 5.0 10,000-13,333 1.9 6.8 38 2.3 11,072 6.0 13,334-16,666 0.7 3.5 27 1.6 14,974 5.7 16,667 and over 0.8 10.7 49 2.9 34,900 24.2 Total 100.0 100.0 1,688 100.0 4,1E4 100.0 /a Less than 1%. Sources: NCSO and World Bank. -86- ANNEX 7 Page 10 Table 10: WORLD BANK DATA ON METRO MANILA HOUSEHOLD EXPENDITURE DISTRIBUTION Monthly household No. of Average expenditure class households Percent of household (pesos) in sample households size Less than 167 1 0.06 2.0 168- 333 3 0.19 3.0 334- 667 100 6.20 4.3 668- 999 203 12.58 5.3 1,000- 1,333 307 19.02 6.0 1,334- 1,667 261 16.17 6.8 1,668- 2,499 333 20.63 6.9 2,500- 3,333 174 10.78 7.1 3,334- 4,999 122 7.56 7.5 5,000- 6,667 58 3.59 7.7 6,667- 8,333 23 1.43 7.7 8,334- 9,999 9 0.56 9.1 10,000-13,333 13 0.81 8.9 13,334-16,666 3 0.19 6.3 16,667 and over 4 0.25 8.3 - 87 - ANNEX 7 Page 11 Table 11: WORLD BANK DATA ON METRO MANILA PER CAPTTA EXPENDITURE DISTRIBUTION Monthly Average Average Average Percent of per capita No. of house- household household expenditure expenditure house- hold expenditure expenditure spent on food class holds /a size (Total) (Food) (PSF) Less than 25 - 25- 49 - - - - - 50- 99 53 8.8 705.40 477.58 67.7 100- 149 189 8.0 992.74 642.15 64.7 150- 199 242 7.2 1,231.19 773.41 62.8 200- 249 229 6.6 1,449.85 857.25 59.1 250- 299 193 6.0 1,602.45 895.45 55.9 300- 349 159 5.9 1,895.08 1,031.43 54.4 350- 399 97 6.0 2,230.97 1,145.36 51.3 400- 449 125 5.8 2,495.25 1,233.28 49.4 450- 499 } 500- 549 95 5.4 2,910.16 1,345.14 46.2 550- 599 } 600- 699 66 5.6 3,568.00 1,579.17 44.2 700- 799 45 5.6 4,179.73 1,659.18 39.7 800- 899 37 6.0 5,050.89 1,919.46 38.0 900- 999 10 5.1 4,791.70 1,515.00 31.6 1,000-1,999 51 5.4 7,189.45 2,322.94 32.3 2,000-2,999 5 4.8 11,712.80 2,834.00 24.2 3000 and over 6 3.8 19,677.50 2,916.67 14.8 ./a Total in sample was 1,688. Data in this table available for 1,605. -88 - ANNFX 7 Page 12 Table 12: POVERTY INCIDENCE TRENDS, 1965-1975 BANK POVERTY REPORT, AND ARRERA STUDY Source and poverty Percent of families below the poverty threshold threshold (1971) 1965 1971 1975 p.a. per capita Rural Urban Total Rural Urban Total Rural Urban Total World Bank Report (P 500) n.a. n.a. 37.2 45.9 29.0 40.5 47.5 40.2 45.3 Abrera study Food threshold: (P 520) 38.6 23.9 34.0 48.0 24.7 48.0 n.a. n.a. n.a. Total threshold: (P 877) 76.1 56.8 70.1 76.1 51.7 69.5 n.a. n.a. n.a. Sources: See World Bank (1980) and Abrera (19761. Table 13: REGIONAL POVERTY LINES AND POVERTY INCIDENCR, 1983 /a Value (P ) Poverty of 16 kg of Food to non- line Poverty rice and other food ratio at per capita inc- Region food /b poverty line /c per month dence /0 (in pesos) (in pesos) (7) Metro Manila R2.7 0.59 14n 11.2 Ilocos 83.5 0.74 113 40.3 Cagayan Valley 83.5 0.74 113 43.1 Central Luzon 80.3 0.69 116 27.4 Southern Tagalog 86.7 0.6R 12R 11.3 licol 81.3 0.74 110 42.7 Western Visayas 79.2 0.68 116 50.0 Central Visayas 83.2 0.68 122 48.1 Eastern Visayas 83.7 0.77 109 33.0 Western Mindanao 84.8 0.66 128 40.1 Northern Mindanao 85.1 0.75 113 38.6 Southern Mindanao 81.3 0.66 123 33.3 Central Mindanao 81.9 0.74 111 28.4 /a See Annex 6 for methodological details. Tb Regional rice prices were used. Expenditure on cereals as a proportion of total food expenditure was assumed to be equal to 0.6 in all the regions. /c The 1975 FIES food ratios were raised by a factor of 1.2 (which is the ratio of 0.68 to 0.57 where 0.68 is the proportion spent on food at the poverty line and n.57 is the overall proportion spent on food, both for the Philippines as whole). /d Based on distribution of families hv per capita income class. Refers to third quarter. 89- ANNEX 7 Page 13 Table 14: POVERTY INCIDENCE, 1979-83, AMONG A PANEL OF 3,294 FAMILIES Poverty incidence (percentage of families with per capita incomes below the per capita poverty line Area 1979 - 3rd quarter 1983 - 3rd quarter Rural 45.3 41.7 Urban 39.6 35.0 Source: Basic Data on households was supplied by NCSO. Data were processed at the Bank. Table 15: PERCENTAGE DISTRIBUTION OF FAMILIES WITH SPECIFIED INCOME INCREASE BETWEEN 1979 AND 1983 Range of per capita Percent of families income increase Urban Rural Zero or negative 27.3 29.9 Positive up to 58% 20.9 21.1 Over 58% 51.8 49.0 Total 100.0 100.0 Note: The inflation rate for the 1979-83 peso was 58%. Source: See Table 14. -90- ANNEX 7 Page 14 Table 16: POVERTY INCIDENCE BY LEVEL OF EDUCATION, 1979 AND 1983 Educational level of Urban Rural the household head 1979 1983 1979 1983 Elementary (Grade 1-7) 52.5 49.3 51.8 46.8 High school (Grade 1-4) 32.6 28.1 37.0 36.7 College (1 to 5 years) 22.1 14.0 25.8 22.7 Bachelor or Higher Degrees 11.6 4.7 3.7 5.4 All levels 39.6 35.0 45.3 41.6 Source: See Table 14. Table 17: INCOME SHARES OF BOTTOM 20% AND TOP 20% OF FAMILIES IN SELECTED COUNTRIES Income shares of Bottom 20 Top 20 Hong Kong (1971) 5.6 49.0 Japan (1971) 8.8 37.6 Republic of Korea (1971) 7.2 43.4 Malaysia (West) (1970) 3.8 56.2 Philippines (1971) 3.9 54.0 Taiwan (1972) 8.8 37.2 Brazil (1970) 3.0 61.5 Nxico (1969) 4.2 63.2 Source: Shail Jain (1975). - 91 - ANNEX 7 Page 15 Table 18: ESTIMATES OF AVERAGE ANNUAL FAMILY INCOME, RURAL AND URBAN Rural/urban ratio Urban Rural Family Per capita Year (P) (P) income income /a 1975 6,789 5,139 0.757 0.750/b 1980 23,991 8,689 0.362 0.446 1981 24,617 10,221 0.415 0.512 1982 31,391 12,895 0.411 0.507 1983 31,929 12,366 0.387 0.477 /a In 1975, the urban and rural household size were respectively 6.03 and 5.88. In 1980, the urban household size was 6.35 and the rural household size was 5.15. These were used to get the ratio of per capita incomes for 1980-1983. /b This changes to 0.776 if the FIES household sizes are replaced by those from the Integrated Census of Population (census estimates are 6.13 and 6.18 for urban and rural respectively). Table 19: GDP PER WORKER IN AGRICULTURE AND NONAGRICULTURE, 1975-1983 GDP at current GDP at 1972 GDP per worker at current GDP per worker at constant prices (P bln) Prices (P bln) Employment ('000) prices prices Agri- Nonagri- Agri- Nonagri- Agri- Nonagri- Agri- Nonagri- Ag./Non- Agri- Nonagri- Ag./Non- Year culture culture culture culture culture culture culture culture Ag ratio culture culture Ag ratio 1975 33.0 81.6 18.2 50.2 7,190 6,605 4,590 12,354 0.37 2,531 7,600 0.33 1980 61.8 204.3 23.7 68.8 8,793 7,596 7,028 26,896 0.26 3,120 9,057 0.34 1981 69.4 235.9 24.6 71.6 9,459 8,075 7,337 29,214 0.25 3,046 8,367 0.34 1982 76.2 264.2 25.4 73.7 9,086 8,428 8,387 31,348 0.27 3,014 8,745 0.34 1983 82.1 298.7 24.8 75.3 9,139 9,115 8,983 32,770 0.27 2,721 8,261 0.33 Table 20: ESTIMATES OF RATIOS OF AGRICULTURAL AND NONAGRICULTURAL INCOMES PER WORKER, 1975-1983 Share of agriculture Assumed Derived Agricultural Agricultural income Overall income to in total GDP income-GDP income-GDP nonagricultural per worker as a ratio GDP ratio at At current At constant ratio in ratio in non- productivity of nonagricultural Year current prices prices prices agriculture agriculture ratio income per worker --- at current prices ---- 1975 0.78 0.29 0.27 0.90 0.73 0.37 0.46 1980 0.75 0.23 0.26 0.90 0.71 0.26 0.33 1981 0.76 0.23 0.26 0.90 0.72 0.25 0.31 1982 0.75 0.22 0.26 0.90 0.71 0.27 0.34 1983 0.76 0.22 0.25 0.90 0.72 0.27 0.34 poll ANNEK 7 93 - Page 17 Table 21: TERKS OF TRADE IN AGRICULTURE: 1969-1982 I II III IV Year Index % change Index Z change Index % change Index Z change 1969 113 99 95 91970 112 (0.9) 97 (2.0) 96 1.1 1971 101 (9.8) 105 8.2 103 7.3 110 1972 100 (1.0) 100 (4.8) 100 (2.9) 100 (9.1) 1973 111 11.0 109 9.0 113 13.0 106 6.0 1974 106 (4.5) 103 (5.5) 110 (2.7) 128 20.8 1975 104 (1.9) 105 1.9 110 0.0 133 3.9 1976 96 (7.7) 102 (2.9) 103 (6.4) 132 (0.8) 1977 92 (4.2) 103 1.0 100 (2.9) 128 (3.0) 1978 88 (4.3) 100 (2.9) 97 (3.0) 125 (2.3) 1979 88 0.0 95 (5.0) 93 (4.1) 120 (4.0) 1980 74 (15.9) 81 (14.7) 80 (14.0) 96 (20.0) 1981 70 (5.4) 80 (1.2) 78 (2.5) 92 (4.2) 1982 74 5.7 79 (1.2) 76 (2.6) 89 (3.3) Definitions: - IPIN Agri GVA (excl. Forestry and Fishery) x 100 Price index of agricultural inputs - IPIN Agri GVA (excl. Forestry and Fishery) xr 100 IPIN industry GVA II - IPIN Agri GVA 100 IPIN Others (Industry GVA and Services GVA)1 - IPIN Agri GVA (excl. Forestry and Fishery) r 100 Consumer Price Index Rural IPIN - Implicit Price Index GVA = Gross Value Added Source: NEDA, August 1984. -94 - ANNEX 7 Page 18 Table 22: ESTIMATES OF NUMBER OF HOUSEHOLDS, POPULATION AND AVERAGE HOUSEHOLD SIZE, 1961-1983 Average Employed No. of Mid-year household Employed population per Year households population size population household ('000) ('000) ('000)/a 1961 4,427 28,239 6.38 - 1965 5,132 31,881 6.21 - - 1971 6,347 37,862 5.97 11,680 1.84 1980 8,677 48,317 5.57 16,657 1.92 1981 8,894 49,526 5.57 17,534 1.97 1982 9,111 50,740 5.57 17,514 1.92 1983 9,328 51,956 5.57 18,254 1.96 /a For August (1971-75) or 3rd Quarter (1980-83). Sources of Data: No. of households is from NCSO surveys (FIES for 1961-1971 and ISH, 3rd Quarter, for 1980-83). Mid-year population estimates are from the 1983 Philippine Statistical Year- book. Employment data are from Labor Force Surveys and ISH. Table 23: POVERTY INCIDENCE AMONG THE EMPLOYED PERSONS, 1978 Percent of persons Percent of persons below poverty level below poverty level earnings /a earnings /a Occupation Urban Rural Industry Urban Rural Professional, technical and related workers 5.9 6.3 Agriculture, forestry, fishing 54.2/d 42.4/e Administrative, executive and Mining and Quarrying 25.0 12.0 managerial workers 4.4 16.7 Manufacturing 26.6 51.4/e Clerical and related workers 12,4 13.8 Utilities 6.3 6.2 Sales workers 42.2/b 44.9/c Construction 35.8/d 21.5 Service workers 40.57- 48.77W Wholesale and Retail Trade 42.27d 45.2/e Agricultural, forestry and fishery workers 54.97w 42.57W Transport and Commnications 19.2 22.7 Production workers, transport operators Finance and Insurance 7.6 13.5 and laborers 30.1/b 39.7/c Community and Personal services 35.0/d 27.7 All Categories 33.8 40.2 33.8 40.2 * /a Poverty lines per family for the fourth quarter 1978 were P 4,345 for rural and p 6,004 for urban. On the basis.of 2 earners per family, the quarterly earning thresholds were P 543 for rural and P 740 for urban. /b In the total employed persons, the shares of these groups were: sales, 15.8%; service, 17.1%; agriculture, 10.2%; production workers, 31.0%. c In the total employed persons, 8.3%, 20.2%, 62.3% and 17.9% were in the categories of sales, service, agriculture and production workers. /di Sectoral shares in total employment were 10.52 (agriculture), 5.0% (construction), 15.8% (wholesale and retail trade) and 35.1% (services). /e Sectoral shares in total employment were 62.7% (agriculture), 10.11 (manufacturing) and 8.2% (wholesale and retail trade). Note: Employed persons here exclude the unpaid family workers. I-.0% - 96 - ANNEX 7 Page 20 Table 24: STRUCTURE OF EMPLOYHENT, 1965-82 Sector of activity 1965 1971 1976 1978 1980 1982 Agriculture Agriculture, forestry, hunting and fishing 59.9 49.7 52.8 52.3 51.8 52.1 Industry Mining, manufacturing, construction utilities, transport and communications 18.0 20.5 18.9 19.2 19.6 18.2 Services Commerce, government, domestic, personal and community services 22.1 29.8 28.3 28.3 28.6 29.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 Source of Basic Data: NCSO .97 - ANNEX 7 Page 21 Table 25: EMPLOYMENT CHANGE BY SECTORS IN URBAN AREAS Employment in thousands Mining, Month, year Total Agriculture manufacturing Construction Services May 1965 3,016 507 568 137 1,804 May 1970 3,598 542 661 219 2,176 Change 1965-70 582 35 93 82 372 (%) 100.0 6.0 16.0 14.1 63.9 August 1971 3,756 349 659 201 2,547 August 1976 4,779 456 795 229 3,299 Change, 1971-76 1,023 107 136 28 752 (%) 100.0 10.5 13.3 2.7 73.5 3rd Quarter 1976 4,548 701 - 1,528/a - 2,319 4th Quarter 1983 5,396 670 1,617 3,109 Change, 1976-83 848 -31 89 790 (%) 100.0 -3.7 10.5 93.2 /a Included are manufacturing, mining, construction, utilities and transport and communications. Source of Basic Data: NCSO ANNEX 7 -98- Page 22 Table 26: EMPLOYMENT CHANGE BY SECTORS IN RURAL AREAS Employment in thousands Mining, Month, year Total Agriculture manufacturing Construction Services May 1965 7,527 5,545 673 162 1,147 May 1970 7,760 5,518 744 219 1,279 Change, 1965-70 233 -27 71 57 132 (Z) 100.0 -11.6 30.5 24.5 56.6 August 1971 8,472 5,617 835 243 1,777 August 1976 10,648 7,670 941 262 1,775 Change, 1971-76 2,176 2,053 106 19 -2 (%) 106.0 94.3 4.9 0.9 -0.1 3rd Quarter 1976 9,690 6,957 1,280/a 1,453 4th Quarter 1982 13,218 9,026 1,765/a 2,427 Change, 1976-82 3,528 2,069 485 974 (Z) 100.0 58.7 13.7 27.6 /a Manufacturing, mining, construction and transport and communications. S99 - ANNEX 7 Page 23 Table 27: DIRECT GOVERNMENT EMPLOYMENT IN RELATION TO TOTAL EMPLOYMENT, 1976 AND 1982 Direct gov't Total employement employment Gov't employ. as ('000) ('000) % of total 1982 1976 1982 1976 1982 1976 Urban 5,396 4,548 842 743 15.6 16.3 Agriculture 670 701 16 23 2.4 3.3 Industry 1,617 1,528 106 92 6.6 6.0 Services 3,109 2,319 720 628 23.2 27.1 Rural 13,218 9,690 802 562 6.1 5.8 Agriculture 9,026 6,957 30 56 0.3 0.8 Industry 1,765 1,280 86 59 4.9 4.6 Services 2,427 1,453 686 447 28.3 30.8 Table 28: EMPLOYMENT GROWTH 1976-82 BY CATEGORY OF WORKERS Employment Change 1976-82 1982 1976 ('000) (%) Urban 5,396 4,548 848 100.0 Wage and salary workers 3,486 3,238 248 29.2 Own account workers 1,373 1,027 346 40.8 Unpaid family workers 538 284 254 30.0 Rural 13,218 9,690 3,528 100.0 Wage and salary workers 3,998 3,171 827 23.4 Own account workers 5,601 4,385 1,216 34.5 Unpaid family workers 3,620 2,065 1,555 44.1 Source of Basic Data: NCSO. ANNEX 7 Page 24 Table 29: EMPLOYMENT CHANGE BY MAJOR REGIONS Employment Change ('000) 1976-82 Percentage distribu- No. tion of 1975-80 Major region 1982 1976 ('000) (Z) population increase National Capital Area 1,999 1,773 226 5.2 15.9 Luzon 6,363 4,835 1,528 34.9 34.0 Visayas 5,827 4,537 1,290 29.5 20.9 Mindanao 4,425 3,093 1,332 30.4 29.2 Philippines 18,614 14,238 4,376 100.0 100.0 Source of Basic Data: NCSO Table 30: UNEMPLOYMENT RATES BY SEX, METRO MANILA AND OTHER AREAS, 1976-82 Unemployment as Z of labor force Male Female 1982 1980 1978 1976 1982 1980 1978 1976 Metro Manila 10.8 9.1 5.8 10.5 15.8 11.1 9.2 14.4 Other areas 2.7 2.5 1.9 2.4 7.5 6.9 6.3 7.7 Philippines 3.6 3.2 2.4 3.4 8.6 7.5 6.7 8.8 -101 - ANNEX 7 Page 25 Table 31: OCCUPATION BY THREE PRINCIPAL EDUCATIONAL LEVELS, 1976 AND 1982 1976 1982 Education and occupation (%) (%) Elementary Professional, technical and related 0.2 0.3 Administrative, executive, managerial 0.4 0.2 Clerical and related 0.3 0.4 All others 99.1 99.1 Percent of ntElementary" in Total Employment 54.3 51.6 High School Professional, technical and related 1.0 0.8 Administrative, executive, managerial 1.2 0.8 Clerical and related 3.5 3.2 All others 94.3 95.2 Percent of "High School" in Total Employment 21.2 24.9 College Professional, technical and related 34.5 31.5 Administrative, executive, managerial 4.1 4.5 Clerical and related 18.8 21.0 All others 42.6 43.0 Percent of "College" in Total Employment 17.2 17.4 Note: Each educational category includes graduates and non-graduates. Source of Basic Data: NCSO. Table 321 AVERAGR DAILY BASIC WAGE RAT&S OF LABORERS IN INDUSTRIAL ESTABLISKMENtS IN NETRO MANILA, BY OCCUPATION, 1970-1980 Cigaretta and Blacksmath sotlermen Carpentera . Drivere cigar elkera Capositorg Electriclane Foremca Period Nominal Real Nominal Ream Nominal Reai Noninal Rai Noinal Reai N-aHnal 0eal Nominal =Rai Nominal Rear (1) (2) (3) (4) (5) (6) (7) (8) (9) <10) (11) (12) (13) (14) (15) (16) 1970 9.41 11.88 12.99 16.40 9.97 12.59 9.81 12.39 7.89 9.96 10.03 12.66 13.31 16.81 16.56 20.91 1971 9.89 10.90 13.21 14.56 10.30 11.36 11.08 12.22 8.57 9.45 10.67 11.76 13.31 14.67 17.50 19.29 1972 11.26 11.26 13.45 13.45 10.66 10.66 11.51 11.51 8.81 8.81 12.48 12.48 13.39 13.39 18.34 18.34 1973 12.35 10.83 13.96 12.25 10.97 9.62 12.58 11.04 9.01 7.90 13.26 11.63 13.52 11.86 19.66 17.23 1974 14.73 9.68 14.63 9.61 11.46 7.53 13.48 8.86 9.49 6.24 13.57 8.92 14.92 9.80 22.24 14.61 1975 15.51 9.42 15.34 9.32 11.74 7.13 14.14 8.39 9.64 5.86 13.80 8.38 15.80 9.60 23.62 14.34 1976 16.03 9.17 15.67 8.97 12.35 7.07 14.51 8.31 10.36 5.93 13.96 7.99 16.23 9.29 25.13 14.38 1977 18.49 9.80 16.93 8.98 15.37 8.15 17.00 9.01 11.52 6.11 14.92 7.91 16.82 9.92 26.43 14.01 1978 21.24 10.46 19.10 9.41 17.55 8.65 18.83 9.29 11.65 3.75 17.48 8.61 19.98 9.84 27.84 13.72 1979 24.30 10.12 20.96 8,73 19.26 8.00 19.78 8.24 11.94 4.98 18.96 7.90 22.69 9.44 30.98 12.90 c 1980 27.50 9.68 22.06 7.77 19.90 7.02 21.88 7.70 12.77 4.30 19.16 6.75 24.01 8.41 - 35.56 12.50 Lathemen Lntypiats Tingeitha NMons Nachanics Painters P1umbere Common laborera Period Noinal vaa 1omNi1? URel Nomånar ai Nominal Reåi Wulnal aal Nominal Rai Nominal Kemi Nouinal Real (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) 1970 10.02 12.65 14.08 17.78 10.08 12.73 10.42 13.16 12.39 15.64 10.99 13.88 11.61 14.66 9.21 11.63 1971 10.87 11.98 14.70 16.21 10.61 11.70 10.53 11.61 13.43 14.81 12.07 13.31 11.59 12.78 9.84 10.83 1972 11.26 11.26 16.15 16.15 11.04 11.04 10.71 10.71 14.00 14.00 12.34 12.34 11.71 11.71 10.42 10.42 1973 12.45 10.92 16.64 14.60 11.81 10.36 10.70 9.39 14.44 12.67 13.54 11.88 11.71 10.66 10.69 9.38 1974 14.14 9.29 16.96 11.14 14.52 9.54 10.75 7.06 16.32 10.72 15.24 10.01 12.76 8.38 11.54 7.58 1973 14.51 8.82 17.04 10.35 15,06 9.15 11.16 6.78 17.44 10.59 15.86 9.63 13.31 8.08 12.51 7.60 1976 15.38 8.81 17.05 9.76 15.78 9.03 11.83 6.77 18.23 10.43 16.66 9.53 13.45 7.70 13.16 7.53 1977 15.97 8.47 18.01 9.55 17.33 9.19 14.76 7.83 20.51 10.87 17.20 9.11 15.92 8.44 13.84 7.34 1978 16.94 8.36 19.86 9.79 20.08 9.89 16.77 8.27 23.90 11.78 19.64 9.68 17.90 8.83 14.42 7.12 1979 18.97 7.90 22.43 9.34 22.93 9.55 18.57 7.73 25.86 10.76 21.28 8.86 19.22 8.01 15.19 6.32 1980 21.27 7.49 22.65 7.98 23.24 8.19 19.12 6.74 28.43 10.01 22.46 7.91 18.36 6.47 15.78 5.56 ANNEX 7 103. - Page 27 Table 33: NOMINAL DAILY WAGE RATE CHANGES IN MEO MANILA, 1980-83 Wate rate (P/day) 1980 1q83 Karch Percentage increase Occupation CR data NWC data 1983 March over 1980 Carpenter 19.90 28.84 44.9 Compositor 19.16 34.91 82.2 Electrician 24.01 29.18 21.5 Foreman 35.56 41.43 16.5 Latheman 21.27 37.55 76.5 Linotypist 22.65 36.86 62.7 Mason 19.12 22.47 17.5 Painter 22.46 33.44 48.9 Plumber 18.36 27.35 49.0 Common laborer 15.78 19.39/a 22.9 Memo: Consumer price index (1978 = 100) 141.5 182.3 28.8 /a The average monthly minimum salary over ten sectors was P 504. The daily rate was based on 26 working days in a month. 10- ANNEX 7 - 104- -AP4 Page 28 Table 34: LEGISLATED MINIMUM WAGE AND RFPORTF.D WAGE, 1971-1982 Metro Manila wage rates (P/day) Agricultural wage rates (P/day) Legislated Reported actual Legislated Legislated Reported wage Legislated highest for unskilled to reported lowest in to reported Year minimum workers ratio minimrum palay sector ratio 1971 8.00 9.84 0.81 4.75 3.R5 1.23 1972 8.00 10.42 0.77 4.75 4.36 1.09 1973 8.00 10.6q 0.75 4.75 4.46 1.07 1974 9.98 11.54 0.86 5.34 5.56 0.96 1975 10.65 12.51 0.85 5.74 6.5q 0.R7 1976 12.81 13.16 0.97 7.09 R.q7 0.79 '1977 15.18 13.84 1.10 9.47 9.86 0.96 1978 16.27 14.42 1.13 10.55 10.42 1.01 1979 21.19 15.19 1.53 13.51 10.71 1.26 1980 29.85 15.7R 1.R9 17.28 10.60 1.63 1981 31.82 n.a. n.a. 18.53 11.91 1.56 1982 31.82 n.a. n.a. 19.65 13.07 1.50 Source: NUC ANNEX 7 - 105.- Page 29 Table 35: ESTIMATED NUMBER OF HOUSPHOLDS AND EMPLOYED PERSONS BY RURAL AND URBAN, 1976 and 1982 Employed Persons Households ('000) Employed Persons Per Fousehold Year, Quarter Rural Urban Rural Urban Rural Urban 1976 III 5,11R 2,300 9,690 4,548 1 89 1.98 1982 IV 6,197 2,967 13,218 5,396 2.13 1.82 Source: NCSO, ISH Table 36: PROJECTED LABOR FORCE, 1980-2000 Year Males ('000) Females ('000) Total ('000) 1980 11,576 5,982 17,558 1985 13,669 7,568 21,237 1990 15,861 9,338 25,199 1995 18,216 11,381 29,597 2000 20,680 13,646 34,326 Source: NCSO IBRD 17234 1å CLASSIFICATIONOP PROV1NCE 120 ir UNE e= BY GEOGRAPHICAL REGIONS N( NATIONALCAPTAL REGION 11."DS'" WESTERN VISAYAS A on A~ PHILIPPINES Abra -Capir lieur Antle, 20 Maumin al CENTRAL VISA VAS - A.rr PanlIii Ngme Orianual CAGAYAN VALLEY .gobeg 9ns . Skallar CYl EASTERNVISAYAS pancmal Lndors KlCiipavaa Norihen Sima?r ~ Smnr- -- -- Reg nal U ndanm Nu~uVisa Lfe1w ^porr, .OQino ' SoudNm Leyte Laag Irnational Bunderi j CENTRAL LUZON WESTERN NINDANAO Nuava Eclij Zea~ga del Norge Dir~ Tr~c Zembanga del Sur Zamblai Bailan Pamng lulu gtilaen Tawitiel Batman ~ NORTHLERN MIN DANA lagen fUTftERN TAGA JDG .Suriadal a Norge Auroa CAn~gin Ausa di Nort. ---. RigalMmiOriental Boguma 0 50 m00 ir0 2m0 250 30 C ~ M ~i Oa O ccdid nu K ILO M ETERS S 9 1!! I.. .. . . . . . Bukdn MILES de AgundalSur 0 50 100 50 200 EASTERN MINDANAO Mnda Orini -SurigndelSur Tari~ Mkndara Oakdenam Davis Orianimi - abnuan ombian Oavn Angeles Pa~ Davoo del Sur - BCOL St Comto 6 - Camrin . Nom aNTRAL MINDANAO LMANILA CaLan~du Sur L~;malNrt - Lanco de SMr t Notc~ta, Pa bilacp - Sutmn Kuderat L Colapen Leg90pi SM Allen ras o Puerto cu~ Priftcesa Or CHINA * VJ†SllS † Bco ~O nasan 12r 12r