Trend of rural employment and wages in Pakistan.

Chaudhry, M. Ghaffar ; Chaudhry, Ghulam Mustafa

INTRODUCTION

Trends in rural employment and wages are important in a number of ways. For example, a steady growth of job opportunities is a precondition for productive employment of labour force under rapid growth of population. Rising real wages of the working class would be essential for incessant improvements in the standard of living of the masses. Lack of sufficient employment opportunities in rural areas together with the consequent stagnating (even declining) wages may be a potential cause of mass movement of rural labour to urban areas and attendant formidable economy-wide problems. Similarly, aggregate growth rates of employment and wages in contrast with those in productive sectors have an important bearing on trends in income distribution and poverty. Rapid growth of population, predominance of rural sector and a general lack of studies on rural labour market conditions in less developed countries, including Pakistan, call for a study such as the present one, which explores the trends of rural employment and wages in Pakistan.

The paper carries four sections. Section 1 surveys the present state of the rural labour market. Section 2 reports trends in rural employment and discusses the various factors underlying those trends. Employment situation being the basic determinant, wage trends, especially those in agriculture, are highlighted in Section 3. Section 4 summarises the findings of the study and in their light makes some policy recommendations.

1. STATE OF RURAL LABOUR MARKET

Rural labour market is the largest segment of the labour market in Pakistan. According to available statistics, it accounted for 73 percent of the total labour force and for more than 70 percent of total employment in 1989-90 [Government of Pakistan (1990)]. Within the rural labour market, agricultural sector employs more than two-thirds of the rural labour force and slightly more than half of the total labour force. In this large market as a whole, both employment and wages are determined by the free interplay of market-forces. However, since the market is highly segmented, variations occur in the employment and wages in its various segments. Villages, being physically separated, represent independent segments and each of them has its own distinct market conditions. In some villages, landlords command a monopsonistic position for employment of rural labour. In others perfect competition may prevail, and in still others shortages of labour may be the fundamental problem. Apart from market conditions, variation in employment and wages in villages may also result from variations in available skills and the resource endowments of those villages. Although a comprehensive study of employment and wages must address all these issues, we, for want of the necessary data and time, are only concerned here with the average trends in employment and wages. This is justified by the fact that increased mobility of experienced rural labour in Pakistan has had the effect of considerably weakening the above-mentioned forces so that one finds it more appropriate to study trends of average rural employment and wages in the rural market as a whole.

2. TREND OF RURAL EMPLOYMENT

Labour force surveys by the Government of Pakistan are one of the major sources of information on the general level of employment as well as rural employment. To give some idea of the trends involved, time-series data on employment have been reported in Table 1.

These data show that employment in rural areas was about 98 percent of the total rural labour force during the 1960s and early 1970s. The rates however fell to below 97 percent in the late Seventies and the Eighties. These rates correspond to the unemployment rates of under 2.0 percent in the former period against 3.0 percent in the latter period. In contrast with the adverse movements in employment rates, under employment seems to have witnessed considerable improvement. While the underemployed represented nearly 18 percent of the labour force until the mid-Seventies, the percentage varied between 11 and 13 in the latter part of the Eighties. A number of fairly well-established explanations could be cited for the intertemporal trend in the rates of employment and underemployment [Chaudhry (1981)]. They include, among others, such significant factors as limitations of the labour-force approach to measurement of employment, rural- labour exodus, slackness of urban labour-market, and job opportunities in agriculture relative to supply of labour.

The employment estimates reported in Table 1 are unusually high and are of necessity, the result of the labour-force approach to the problem in a less developed country. Firstly, there is the problem of discouraged workers. It has been argued that unavailability of job opportunities tend to force many workers out of the labour force [Bardhan (1978) and Shah and Sathar (n.d.)]. Thus to a certain extent labour supply depends on its own demand and leads to fairly stable and high rates of employment. The same applies to services of housewives, who, during peak demand period would participate actively in farm work but would be totally confined to household work during slack- demand periods. Secondly, the labour-force approach to employment makes little sense in a society where more than 80 percent of the labour force consists of self-employed or unpaid family-workers. The problem is compounded by the informal, short-term and seasonal nature of rural jobs and by the joint responsibility of household members for supply of labour.

While employment has tended to be over-stated because of the above-mentioned factors, intertemporal changes in employment, though dampened, result from other three factors. The opening up of international markets and expansion of government-job opportunities were typical feature of the early Seventies and continued to ensure low rates of open unemployment in the rural areas. According to available information, an estimated 1.8 million workers, mainly from rural areas, were offered jobs in the international market during the decade of the Seventies [Gilani, Khan and Iqbal (1981)]. The over-saturation of government departments in the late Seventies and the return of Pakistani workers from the Middle East during the Eighties resulted in rising unemployment-rates in the two sub-periods. As rural areas continue to act as a reservoir for the unemployed, over-saturated urban markets and high rates of unemployment, especially among the educated youth, have been factors in the rising rural unemployment. More than any of these factors, however, changes in the demand for and supply of labour in agriculture have been responsible for corresponding changes in rural employment.

In recent years, here has been a growing consensus in literature about a positive relationship between agricultural output and demand for labour. Mellor (1988), for example, shows that elasticity of employment in agriculture with respect to the growth of agricultural output is likely to be close to 0.6 in a traditional economy. The deteriorating rural employment situation in the 1980s, in part therefore, should be attributed to slower growth of agricultural output [Government of Pakistan (1990)]. The elasticity of employment, however, is likely to vary with changes in the labour intensity of the production process, cropland increases and changes in cropping patterns. Precise estimates of employment growth, therefore, need to be made on the basis of trends in these variables. The results of our exercise conducted on this basis are reported in Table 2.

Time-series data on physical input of labour per crop-acre are lacking. However, the existing data indicate that physical input of labour went up at the rate of nearly 0.70 percent per annum in Pakistan's agriculture [Chaudhry (1982)] during the 1960s and early 1970s. We have assumed continuation of the constancy of this rate for the rest of the period. The data reported in Chaudhry and Chaudhry (1992) clearly show that there has been a consistent shift in cropping pattern towards more labour-intensive crops in Pakistan's agriculture. Although its significance has declined with the passage of time, it remains a dominant source of added employment in agriculture throughout the period under consideration. Similarly, increases in cropland amounted to more than one percent in the 1960s and 1970s but fell to under one percent in the early Eighties with negative growth rate of 0.26 percent per annum in the subsequent period. Overall, the growth of job opportunities in agriculture progressively declined over the period under consideration. It amounted to nearly 3.0 percent in the decade of the Sixties, dwindled to about 2.5 percent during the 1970s, and fell further to 2.0 percent during early Eighties. By contrast, the growth rate of job opportunities in the late Eighties was only 0.65 percent per annum.

Two factors explain the falling growth rate of job opportunities in the crop production sub-sector. The growth of job opportunities has been considerably in excess of the long-term growth rate of economically active population in agriculture over the period from 1959-60 to 1989-90. In contrast, economically active population in agriculture grew at the rate of 1.92 percent per annum between 1964-65 and 1989-90 [Government of Pakistan (1990)] causing, general, scarcities of labour. Owing to these scarcities, availability of labour increasingly constrained cropland and labour intensive cropping pattern in the late 1980s. The process was hastened by the falling and bare-minimum profitability of agriculture in the wake of elimination of input subsidies and the consequent rapid increases in the prices of key agricultural inputs relative to those of agricultural commodities. The adverse effect of this policy on employment is reflected in declining reliance on Green Revolution technologies which had been responsible for promotion of multiple-cropping [Brown (1970); International Labour Organization (1977) and Mohammad (1970)], raising farm productivities [Gill (1973) and Mohammad (1970)], shifting of the cropping pattern towards labour-intensive cash crops [Mohammad (1970)], direct addition, except in the case of tractors, to labour input per acre [Brown (1970) and Kaneda and Ghaffar (1970)] and the creation of new jobs within and outside agriculture [Chaudhry (1982)].

3. GROWTH OF RURAL WAGES

The intertemporal trend of rural wages is shaped by the same forces as shape the trend of rural employment. Since demand and supply forces determine wages, they are better reflectors of the changing labour market conditions than employment trends. Table 3 helps in appraising the real situation by reporting available wage data for casual workers from 1959-60 to 1984-85.

Several interesting conclusions follow from the data in Table 3. Nominal daily-wage rates in 1959-60 were indeed low, not exceeding Rs 1.36 per day. They took 10 years to double but a more than redoubling of these rates took place in just five years, rising from Rs 3.00 in 1969-70 to 7.53 in 1974-75. Although wage rates between 1974-75 and 1979-80 did not rise quite as rapidly as during the previous five years, a three-fold increase seems to have occurred during the period from 1979-80 to 1984-85. In contrast with these increases in nominal wages, increases in real wages were considerably dampened by rising prices of consumer goods and agricultural commodities. Although trends are the same, consumer-price increases were more pronounced than the increases in agricultural commodity prices in the entire period under consideration as well as in its sub- periods. Deflated by the two price indices, the resultant real- wage rates witnessed a near doubling over the period from 1959-60 to 1974-75. They, however, went up by more than 140 percent in the following decade. Whether one looks at nominal or at real wage rates, there has been considerable variation in the annual growth rates over time. Average annual growth rates of rural wages were about 4.7 percent during the early Sixties but had gone up to 5.07 percent and 6.65 percent between 1964-65 and 1969-70. By contrast, the decade of the Seventies saw falling growth rates followed by a large increase in the early Eighties. While the figures for 1984-85 are provisional, the increase in growth rates of wages is manifest from the growth rates for 1979- 80 to 1982-83.

Many implications of the above discussion may be of interest. The consistent increases in real wages in agriculture suggest that Pakistan may no longer be considered a labour-surplus economy. This becomes clear when we compare the growth rates of Pakistan's wages with those of the basically labour-surplus economies of India and Bangladesh. Against the negligible real- wage increases in India and the falling real wages in Bangladesh between 1960-61 and 1979-80, the real wages of agricultural workers in Pakistan witnessed continuous and substantial increases during the same period [Khan and Lee (1984)]. In fact many a studies indicate that Pakistan's agriculture might be facing labour shortages, especially during the peak-demand periods rather than surpluses of labour [Chaudhry (1982)]; Guisinger (1978) and Jose (n. d.)].

When we compare the growth rates of real wages with those of the value added by agriculture, it becomes obvious that the growth of real wages has been in excess of the value added by agriculture during most of sub-periods, with the exception of the years 1974-75 to 1979-80. This implies that income distribution in agriculture deteriorated only between 1974-75 and 1979-80 in contrast with the substantial improvements during the rest of the period. The same follows from the considerable narrowing of the rural-urban wage differences. For example, daily wage rates for agriculture were only 50 percent of urban wage rates in the mid-Seventies but stood on a par with them in 1984-85 [Government of Pakistan (1990)]. It is more than a mere coincidence that the above conclusion is quite consistent with the long-term trends of rural income distribution as reported by the various Household Income and Expenditure Surveys.

Before concluding this section, two limitations of the analysis of wage trends may, however, be noted. The analysis has been based on wage rates for casual workers and ignores wage rates for permanent workers owing to the lack of consistent time-series data. Studies that have compared wage rates of the two groups indicate that wage rate increases of casual workers have been more pronounced than those for permanent workers [Guisinger (1978)]. This fact in combination with the falling use of permanent labour is liable to slightly overstate the growth rate of wages over time. Again it is feared that conclusions regarding income distribution based on these wage rates may be erroneous if a casual worker works only a few days a year. Such a fear, however, seems to be baseless, as rising wages reflect a rising demand for labour and would normally be accompanied by an increase in the number of work days [Guisinger (1978)]. Since labour incomes are a function of employment and wage rates, the positive growth rates of job opportunities in agriculture as discussed in Section 2 would counterbalance any overstatement of growth rates of wages and thus reinforce our conclusions.

4. CONCLUSIONS AND POLICY RECOMMENDATIONS

The purpose of the present paper was to investigate the trends of employment and wages in the rural areas of Pakistan. It was found that the levels of rural employment have been quite high despite the rising unemployment and falling underemployment during most of the Eighties. While the high levels of employment were attributable to limitations of the labour-force approach as a measure of employment in a less developed country, trends of employment, unemployment and underemployment were the result of the exodus of rural labour, and slackness of urban labour-market and job opportunities in agriculture. While the Green Revolution induced a strong demand for labour in the rural areas in the Sixties, it was complemented by a rapid expansion of the public-sector employment in the early Seventies and increased migration of Pakistani labour to the Middle East during most of 1970s. By contrast, the oversaturated urban sector, return migration of workers from abroad, growing unemployment among the educated and the falling profitability of agriculture contributed to growing unemployment rates in the rural areas in the 1980s. However, growth of demand for labour in agriculture, except in 1989-90, remained well above the increases in labour supply throughout the period from 1959-60 to 1984-85 and may have resulted in growing scarcities of labour. This is supported by rapid increases of real wages in Pakistan's agriculture. Although growth rates of agricultural wages fell sharply between the 1974-75 and 1979-80, they again began to increase at an accelerating pace in the 1980s. The trends of both employment and wages in comparison with the changes in the value added by agriculture suggest an improvement in income distribution during most of the period under consideration, except between 1974-75 and 1979-80.

One of the most probable reasons for deteriorating rural- employment situation in the Eighties has to do with the adverse agricultural price policy and its consequent effect on job opportunities in agriculture. Obviously the low and falling agricultural prices relative to those of inputs have tended to limit the margins of intensive land cultivation and the growth of job opportunities fell far short of the growth of labour supplies during the 1984-85 to 1989-90 period. Unless corrected, such a price policy is likely to have devastating consequences for rural employment and wages in the future. As a corrective action, stability of reasonable profits must be guaranteed by appropriate manipulation of commodity and input prices. While elimination of input subsidies would be commensurable to rising input prices, its ill effects should be countered by raising commodity prices to world levels. Apart from these steps, an improvement in rural employment can be brought about through infrastructural development and rural industrialisation, and all-out efforts must be made to improve the current situation. A long-term solution to the unemployment problem would be curtailment of labour supplies to rural areas by effective population-control measures, rapid industrialisation and a deliberate policy of labour export.

Comments on "Trend of Rural Employment and Wages in Pakistan"

Adequate considerations on evolving a mechanism ensuring a continuous generation of productive employment opportunities, unfortunately, has been missing in Pakistan's development planning and policy formulations. Employment promotion, despite being mentioned as one of the important objectives of the planning exercises, in fact has been seldom translated into concrete action plans. The sectoral priorities, allocation of financial resources and different fiscal and monetary measures, announced from time to time, have rarely considered the implications for employment. In fact, a number of such measures have been found to be at variance with employment.

Unfortunately, the academicians and scholars, also did not address employment related issues to the extent the need was felt. This has led to an insufficient empirical work in this area. The situation for rural areas being the worst. Against this background any work on employment especially that concentrating on rural area, should be seen as a useful addition and fully used. The attempts of the authors, therefore, to analyse the state of the rural labour market, trends in rural employment and growth of rural wages, are commendable.

A careful look at this interesting piece of work, however, indicates that somewhere something is either missing or inadequately covered. The fact that the whole exercise is overwhelmingly concentrated on the crop sub-sector of rural agriculture does not help but to conclude that injustice has been done to the title. Then again some of the discussions and conclusions emerging thereof, at least to me, appear to be improper. Further elaborations of these points appear under (i) general comments and (ii) specific comments.

GENERAL COMMENTS

It is not at all understandable as to why the authors have limited their analysis of wage trends in rural areas to the agricultural sector and there too to the crop sub-sector. Moreover, in an essentially informal environment i.e. self- employment and unpaid family helpers implying disproportionately smaller wage-income earners, why the incomes of the employed have been excluded from the analysis. The argument of the authors about the nonavailability of such data seems to be based on inadequate information on some of the important sources such as: (i) Housing, Income and Expenditure Surveys, (ii) Surveys of Small Household and Manufacturing Industries, and (iii) Social Indicators of Pakistan. Use of these data sources would have certainly enhanced the usefulness of the analysis.

The limitations of the labour force approach in capturing the employment scene is mentioned without suggesting alternatives and, importantly, even without adequately discussing them. The argument that people tend to withdraw from the labour market due to bleak job opportunities the discouraged workers phenomenon does not seem to be applicable to the rural areas. The information in the Labour Force Survey (LFS) on that part of population which is neither in school nor in the labour force could have been analysed to substantiate the discouraged worker argument. The authors also appear to be unaware of the revised LFS questionnaire which addresses captures the extent of such workers. Similarly, the authors appear to be making no difference between job opportunities and work/employment opportunities. While the former essentially is wage employment, the latter can be wage employment, self-employment and unpaid employment as family members.

Great emphasis has been given to older references in support of a situation/observation pertaining to a latter period. For instance (i) explaining the factors responsible for the inter-temporal trend of employment and unemployment up to the present time, a reference of 1981 Chaudhry (1981) has been given, (ii) for the phenomenon of discouraged workers, supporting reference appears to be of 1978 Bardhan (1978), (iii) falling agricultural prices and higher costs of the present time are linked and explained with the earlier work, mostly of the 70s, etc.

SPECIFIC COMMENTS

The discussion on the extent of under-employment, defining one as underemployed if found working less than 35 hours a week, could have been more meaningful, had the under-employed were distributed in different groups of weekly hours. Use of income criteria, based on the use of Housing, Income and Expenditure Surveys, would have been even more meaningful.

Explanations of inter-temporal changes during the 70s and 80s, do not reflect the realities. This particularly refers to the argument that the capacity of the government to provide employment affected the rural employment scene. Similarly, the linking of rising educated unemployment, that too without figures and sources, to that of rural unemployment does not seem to be adequately substantiated.

It would have been more useful for the reader, if the methodology used for the construction of Table 2 was briefly mentioned instead of merely indicating sources of the table. The assumption that the rate of increase of physical labour input in agriculture in the present time would be similar to those of the 60s and 70s, probably may not be realistic, especially in the wake of the considerable degree of mechanisation in farm operations.

It is difficult to understand as to how labour scarcities in agriculture in the 80s have been arrived at. And then how they are linked with the falling base minimum profitability in this sector due to elimination of subsidies, increasing prices of inputs unmatched with a rise in output prices.

Based on these observations, I would propose to the authors that either they should enlarge their analysis by extending it to the whole rural areas by making use of the available additional data sources, or in case this is not acceptable to them, then they should better concentrate only on the crop sub-sector.

In the end, I shall be failing in my responsibilities, if I do not congratulate the authors for venturing into a field where little empirical work has been done. The comments, I must confess, have essentially been provoked by their excellent work.

Sabur Ghayur

Friedrich Ebert Foundation, Islamabad.

REFERENCES

Bardhan, P. K. (1978) On Measuring Rural Unemployment. Journal of Development Studies 14:3.

Chaudhry, M. Ghaffar (1981) Rural Employment in Pakistan: Magnitude and Some Relevant Strategies. Islamabad: Pakistan Institute of Development Economics. (Research Report Series No. 131.)

REFERENCES

Bardhan, P. K. (1978) On Measuring Rural Unemployment Journal of Development Studies 14:3.

Brown, Lester R. (1970) Seeds of Change: The Green Revolution and Development in the 1970s. New York: Praeger Publishers.

Chaudhry, M. Ghaffar (1981) Rural Employment in Pakistan: Magnitude and Some Relevant Strategies. Islamabad: Pakistan Institute of Development Economics. (Research Report Series No. 131.)

Chaudhry, M. Ghaffar (1982) Green Revolution and Redistribution of Rural Incomes: Pakistan's Experience. The Pakistan Development Review 21:3.

Gilani, Ijaz, M. Fahim Khan and Munawar Iqbal (1981) Labour Migration from Pakistan to the Middle East and its Impact on the Domestic Economy. Islamabad: Pakistan Institute of Development Economics. (Research Report Series Nos. 126-128.)

Gill, Amjad A. (1973) Pakistan Agricultural Development and Trade. Washington, D.C.: Economic Research Service.

Guisinger, Stephen (1978) Long-term Trends in Income Distribution in Pakistan. World Development 6: 1271-1280.

International Labour Organization (1977) Employment Strategy-Pakistan Project: Findings and Recommendations. Geneva.

International Labour Organization (1990) Yearbook of Labour Statistics 1990 and for Other Relevant Years. Geneva: ILO.

Jose, A. V. (n. d.) Agricultural Wages in India. New Delhi: Asian Employment Programme.

Khan, A. R., and Eddy Lee (1984) Introduction. In Azizur Rehman Khan and Eddy Lee (eds) Poverty in Rural Asia. Bangkok: International Labour Organization, Asian Employment Programme.

Kaneda, H., and M. Ghaffar Chaudhry (1970) Output Effects of Tubewells on the Agriculture of Punjab. The Pakistan Development Review 10:1.

Mellor, John W. (1988) Lectures on Agricultural Growth and Employment. Islamabad: Pakistan Institute of Development Economics. (Lectures in Development Economics Series No. 7.)

Mohammad, Ghulam (1965) Private Tubewell Development and Cropping Pattern in West Pakistan. The Pakistan Development Review 5:1.

Pakistan, Government of (1977) Survey Report on Cost of Production of Important Crops in Pakistan. Islamabad: Ministry of Food, Agriculture, Cooperatives, Under-developed Areas and Land Reforms, Agriculture Wing, Planning Unit.

Pakistan, Government of (1983) Labour Force Survey 1978-79. Karachi: Statistics Division. And Similar Surveys for 1982-83, 1984-85 and 1985-86.

Pakistan, Government of (1990) Economic Survey 1989-90. Islamabad: Finance Division, Economic Adviser's Wing.

Pakistan, Government of (1991) Agricultural Statistics of Pakistan 1989-90. Islamabad: Ministry of Food, Agriculture and Cooperatives, Food and Agriculture Division, Economic Wing and Similar Issues for Various Years.

Shah, Makhdoom Ali, and Zeba A. Sathar (n. d.) Quantification of Unemployment and Underemployment in Pakistan: Problems of Conceptualization and Measurement. Employment Planning and Basic Needs in Pakistan. Islamabad: Pakistan Manpower Institute and Bangkok: Asian Employment Programme.

M. Ghaffar Chaudhry and Ghulam Mustafa Chaudhry are, respectively, Joint Director and Staff Economist at the Pakistan Institute of Development Economies, Islamabad.

Table 1
Rates of Employment, Unemployment and Underemployment
in Pakistan for Selected Years since 1966-67

 Percentage of Rural Labour Force that is:

 Underemployed
 (i.e. Working
 Less than 35
Years Employed Unemployed Hours per Week)

1966-67 98.48 1.52 15.44
1967-68 98.65 1.35 15.40
1968-69 98.25 1.75 20.79
1969-70 98.26 1.74 18.94
1970-71 98.26 1.74 15.75
1971-72 98.64 1.34 18.87
1974-75 98.67 1.33 11.02
1978-79 97.00 3.00 15.14
1982-83 96.71 3.29 14.85
1984-85 97.05 2.95 11.32
1985-86 96.89 3.11 11.68
1986-87 97.50 2.50 12.36
1987-88 97.41 2.59 13.02

Source: [Chaudhry (1981) and Government of Pakistan (1983, 1990).]

Table 2
Annual Growth Rates of Job Opportunities in the Crop
Production Sub-Sector since 1959-60

 Percentage Increase in Job Opportunities
 per Annum due to

 Improved
 Labour-Input Cropland Cropping Total
Time Period Increase Expansion Pattern Increase

1959-60 - 1964-65 0.70 1.36 0.94 3.00
1964-65 - 1969-70 0.70 1.13 0.90 2.73
1969-70 - 1974-75 0.70 1.58 0.33 2.61
1974-75 - 1979-80 0.70 1.51 0.12 2.33
1979-80 - 1984-85 0.70 0.79 0.49 1.98
1984-85 - 1989-90 0.70 -0.26 0.21 0.65

Source: [Chaudhry (1982, 1977, 1991) and Government of Pakistan
(1977, 1991).]

Table 3
Trends in Agricultural Wages from 1959-60 to 1984-85

 Nominal
 Wage Consumer Implicit
 Rate (Rs Price Deflator
Years per Day) Index Agriculture

1959-60 1.36 100.00 100.00
1964-65 1.92 112.36 112.53
1969-70 3.00 137.03 126.96
1974-75 7.53 274.02 256.49
1979-80 13.30 435.22 392.80
1982-83 25.00 568.67 514.34
1984-85 42.50 644.67 585.34

 Real Wage Rate Annual Growth Rates
 Deflated by Over Previous Period

Years CPI IDA CPI IDA

1959-60 1.36 1.36 -- --
1964-65 1.71 1.71 4.69 4.69
1969-70 2.19 2.36 5.07 6.65
1974-75 2.75 2.94 4.66 4.49
1979-80 3.06 3.39 2.16 2.89
1982-83 4.10 4.86 10.24 7.47
1984-85 6.59 7.26 16.58 16.45

Source: [International Labour Organization (1990).]