|
|
5. The Rural Labor Market in Indonesia
Rosamund NaylorAn understanding of the rural labor market is central to an analysis of the effects of past and prospective rice policies in Indonesia. Labor remains the most important input in rice production, and rice traditionally has been the largest source of rural employment on Java. But the relative importance of rice has diminished rapidly in the past two decades, and the effects of rice policy on the rural labor market are much different today than in earlier periods. Just how different they are depends on the efficiency with which the rural labor market functions.
The extent of integration between rice and other rural labor markets, and between rural and urban labor markets, influences the changes in employment and wage levels that would result from changes in rice policy. If the labor market for rice is isolated from the rest of the economy, changes in profitability of rice production can be expected to have a direct impact on employment and wages in the rice sector. The large size of the rice sector in the Indonesian economy ensures that such changes would have major significance for aggregate employment and rural incomes. If, on the other hand, the labor market is well integrated, changes in rice employment can be accommodated with relative ease by other parts of the economy; farmers would have difficulty reducing real wages paid in rice production for fear of losing workers to other pursuits.
The efficiency of rural labor markets in Indonesia is an issue of considerable controversy. Studies undertaken at different times create such contrasting scenarios that it is difficult for the reader to believe that the various authors are considering the same country. Analyses also have been plagued by a lack of reliable data so that no consensus has emerged about such fundamental phenomena as the direction and magnitude of change in the level of rural wages. This chapter reassesses the structure of rural labor markets with attention focused on two principal issues-competitiveness in the market for unskilled labor and real wage trends.
The first section of the chapter presents evidence from field surveys conducted in i987-88 on off-farm job opportunities and relative wages (off farm versus on-farm) for unskilled rural workers. The goal of this work is to investigate whether off-farm jobs are widespread, accessible, and well paid relative to those in rice production. Attention also is given to the
selection of employment during the course of the year to understand the role of the rice economy in providing full employment for unskilled laborers. The next section examines recent changes in labor-hiring institutions and work practices in rice production to determine how cultural arrangements evolve in relation to demand and supply conditions in the rural and urban labor markets. The third section provides an analysis of real wage trends in rice production on Java between 1976 and 1988. These trends are evaluated in the context of the observed changes in employment opportunities.
Alternative Views of the Rural Labor Market
Before the introduction of Green Revolution technologies in the late 196os, the rice sector on Java was characterized as a subsistence sector, capable of absorbing an expanding labor force but not of stimulating rural economic growth (Geertz 1963). Significant gains in rice production and land productivity resulted from increasingly intensive cultivation practices, but at best these changes maintained per capita income levels: "The ever driven wet-rice village found the means by which to divide its growing economic pie into a greater number of traditionally fixed pieces and so to hold an enormous population on the land at a comparatively very homogeneous, if grim, level of living" (Geertz 1963, p. 100).
Geertz's equally dismal outlook for the future was premised on a continuous growth of the rural population (and the rural labor force) within Java's confined productive land area and on little expansion of off-farm employment. Until 1970, few productive employment opportunities were available to unskilled workers in the nonagricultural sectors of the rural economy.
Labor market surveys conducted in the 1970s, reported by White (1976, 1979) and Hart (Hart and Sisler 1978; Hart 1986a, 1986b), found that nonagricultural job opportunities for unskilled workers in rural areas were limited in scope and earning potential and that employment in rice production did not rise appreciably with higher rice output. These conditions led to monopsonistic behavior on the part of landowners and increasing income polarization between landholders and landless laborers. Field surveys from the 1970s also supported the existence of a segmented labor market, evidenced by widely different wage rates within the same village for a given age-sex group and season (Lluch and Mazumdar 1985). Hart presented evidence that this divergence in wage rates was attributable to preferential hiring arrangements within the rice sector based on income and social status (Hart and Sisler 1978; Hart 1986b).
Empirical investigations in the ig8os depict a markedly different pattern. Workers earn comparable wages in agricultural and nonagricultural activities. In many areas, labor moves easily between sectors and regions of the economy in response to relative wage differences and the seasonal availability of work opportunities. Research conducted in the mid-1980s by Collier et al. (1988) and Manning (1986a, 1986b, 1988) concluded that an expanding range of employment opportunities, brought about by high rates of economic growth in both rural and urban locations, improved the bargaining position of unskilled workers and helped to distribute income gains more equitably throughout the economy. These analysts found the labor market for unskilled workers on Java increasingly well integrated. The extent to which this transition in rural labor markets has occurred on and off Java remains a source of debate among researchers in Indonesia.
Recent Empirical Evidence on the Rural Labor Market
Field survey evidence presented in this section is used to describe the income-earning activities available to unskilled workers. This evidence helps to establish whether the process of marginalization-the movement of unskilled labor from the agricultural sector into low-paying and low-productivity jobs in the nonagricultural sector is prevalent in rural Java or whether, instead, unskilled laborers are able to work in nonfarm jobs at wages comparable to those earned in agriculture. These results, in turn, show the extent to which landowners are able to exert monopsony power over landless laborers. In other Asian countries, the emergence of a wide range of equally remunerative employment opportunities for unskilled laborers in nonagricultural activities has provided convincing evidence of
1The rural labor market survey was part of a larger survey of rice production technology, crop substitution, and rural labor markets associated with rice cultivation conducted by Heytens and Naylor in 1987-88. The survey of rural labor markets consisted of interviews with rice farmers, farm laborers, owners of rural industries, and unskilled workers in rural industries. The numbers of respondents in the sample were as follows:
On and off the farm. bSee note 2.
Because the survey sites were chosen to emphasize dominant areas of rice production, they are biased slightly toward areas of strong economic growth. Few remote upland regions with poor access to large cities or towns were included in the survey.
a transition from monopsonistic to competitive labor market conditions (Bardhan 1984; Squire 1981).
The survey results confirm previous findings that most rural laborers on Java pursue multiple income-earning activities, within a given day, within a season, and over the course of a year (Collier et al. 1982b, 1988; White 1986; Manning 1986a, 1986b). Regions vary in employment opportuni-ties, and the survey results identify several distinct patterns in labor market behavior. Both institutions and workers' behavior are remarkably flexible in accommodating the pursuit of full employment. But in general, labor mobility is greatest between the urban informal and rural labor markets and between the agricultural labor market and the off-farm rural labor market (Rucker 1985). Construction activity in the cities and villages has been one of the most significant forces attracting male laborers from the agricultural sector since the early ig8os (Manning 1986b, 1988).
Job opportunities for landless laborers in rice production in any given season depend on the proportion of sawah planted in paddy. In the rainy season (November to February), when virtually all of the sawah is in paddy, laborers work in agriculture for two months on average. The number of labor days employed in rice production is higher for workers who harvest with contract teams (tebasan)1 and for those in areas where irrigation schedules stagger the planting and harvesting cycles for farms throughout the season. In the dry season, employment in rice production varies considerably by region. The diversity of production environments in the survey sites thus provides a broad basis for evaluating labor market mobility between sectors and regions on Java.
Three crops of rice are grown on one-third of the sawah in Klaten (Central Java), and laborers work for one to two months on average during each of the dry-season plantings (March to June and July to October). In Majalengka (West Java), paddy is planted twice a year in the irrigated sawah and once a year in the rainfed sawah, but rice is rarely triple cropped because of insufficient water. Laborers are able to find em-ployment in the rice sector for roughly two months in each of the rice seasons. Many look for work outside of the agricultural sector in the second dry season. Rice is triple cropped on only 5 to 1o percent of the sawah in Kediri (East Java). Total days worked in agriculture per laborer fall from fifty to one hundred days in the rainy season to twenty-five to forty days in the second dry season, when only a small portion of the sawah is in paddy.
The substitution of palawija crops for paddy in the dry season reduces the demand for labor in the agricultural sector. Labor inputs for soybeans and corn are only 50 to 75 percent of those for rice in the dry season (Table 5.1). In Kediri, where corn is grown extensively in the sawah during both dry-season plantings, laborers are able to work in corn production for fifteen to thirty days per planting. Soybeans and vegetables are the pre-dominant palawija crops in the second dry season in Majalengka. La-borers remaining in the village work for twenty to forty days per season in the production of these crops. Sugar and tobacco are labor-intensive crops, but less so than paddy. In Klaten, government sugar and tobacco programs permit male laborers to work for three to five months per year in sugar production and male and female laborers to work for up to seven months per year in the tobacco fields.
Interviews with farm laborers and workers involved in various off farm occupations indicate that they engage in multiple-earning activities dur-ing the course of a year or within a season rather than within a given day. Casual and self-employed activities such as informal trading, driving becaks (rickshaws), and making or carrying bricks are the main off-farm activities pursued within a given season. Employment of longer duration, mostly work on infrastructure or construction projects, is more prevalent on a seasonal basis (i.e., during the dry season).
Employment Patterns in Kabupaten Kediri (East Java)
Interviews with unskilled workers in a wide range of rural activities show that labor supply is allocated on the basis of relative wages (including wage equivalents of meals), job duration, physical exertion, perceived risk, and family life-style. Table 5.2 reports a sample hierarchy of employment opportunities for unskilled labor based on survey interviews in Kediri in 1987. At the top of the hierarchy are jobs outside of the agricultural sector which offer significantly higher pay and benefits on a semi-permanent basis. For example, road work on a piecework basis pays an average of RP 3,000 to 4,000 per day for a five- to six-month period, compared to the average wage for males in agriculture of Rp i,goo per day (meals included) for the same number of hours (eight hours per day). These off-farm activities generally require strenuous physical exertion or some training on the job. Because barriers to entry into this class of activities are rare for unskilled workers, these jobs contribute to higher wage levels in rural areas.
Below these high-paying activities are jobs that pay approximately the same wage as agriculture on a daily basis but offer long duration of several months or even years. Within this group, employment preference between agriculture and other jobs is mixed. For example, a construction worker, who could earn Rp 1,500 per day on a six-month basis, preferred agricultural work because he received meals in addition to the same daily wage. Third in the hierarchy are jobs that offer a higher daily wage than
agriculture but are not guaranteed on a daily basis. Although respondents in these activities clearly preferred nonagricultural work because of the greater physical exertion required for farm labor, they continued to work on their own land and for other farmers in all seasons on a part-time basis.
The fourth class of job opportunities is the most representative for landless laborers on Java throughout the year. This class of self-employed workers includes becak drivers and informal traders. Earnings are comparable to those for nonfarm activities and agricultural work, and labor moves smoothly between these sectors within each season. These off farm activities are not marginal in the sense of having inferior quality or low income-earning potential, and they are often preferred to agricultural work. On an hourly basis, wages are roughly the same as those in agriculture. Within this category, the opportunity cost of agricultural labor generally equals the wage rate. The constant flow of workers between the farm and nonfarm sectors for this category of jobs is evidence of a high degree of integration within the regional labor market.
The final category of employment opportunities consists of jobs that are clearly inferior to agricultural labor in both wage and quality of work but offer a means to employ family members with few work alternatives and low opportunity costs to the family. This category is made up of formal sector jobs that can absorb labor on a full-time or seasonal basis. Although these jobs offer permanent wages, working conditions are poor, hours are long, and meals are not included. Most of the laborers in this group are young women who often return to their villages to take part in rice harvests.
Employment Patterns in Kabupaten Majalengka (West Java)
The pattern of off-farm employment opportunities varies between regions. Fewer off-farm employment opportunities exist for unskilled labor in Majalengka than in Kediri. As in much of West Java, widespread migration to the major cities of Jakarta and Bandung occurs on a seasonal basis (Manning i986b). Information and transportation between rural and urban areas is good, and housing generally is provided by employers, often construction firms, in the cities. Earnings are higher in the cities than in the villages. The migrants are mostly male, causing the role of women in the rural labor force to be more pronounced than in East Java (Sjahrir 1990).
Village workers tend to migrate to the cities in groups. In some cases, 70 percent of the unskilled laborers routinely leave the village together to work on a single construction project. In other villages, virtually no seasonal migration to the urban areas occurs. Information about work opportunities in urban areas generally flows by word-of-mouth, although occasionally a person from a project in the urban area is sent to the rural areas to recruit workers. Migrants leave the village either for an extended period (three to five months) or for several shorter periods (one to two months) during the year.
Although emigration is a prominent part of rural employment patterns in Majalengka, opportunities also exist locally (Table 5.3). Tile factories are the principal nonagricultural activity in the region. The larger tile factories operate throughout the year and hire two hundred daily workers (men, women, and children) on average. To secure an adequate supply of daily laborers, the large factories send trucks to the villages to pick up those who wish to work that day and deliver them back to the villages after work. The small factories hire forty daily workers on average, who come on their own from nearby villages. Entry into the industry is unrestricted. Wages in the tile factories are similar to those in agriculture. Employment in local construction also is prevalent during the dry season, as are a variety of informal jobs such as making bricks, digging and carrying sand from the rivers for construction projects, and collecting firewood.
Employment Patterns in Kabupaten Klaten (Central Java)
Klaten is the most densely populated of the survey sites. Supplies of labor are ample, and the likelihood of an "involuted" labor market rife with market failures appears greater than in the other sites. Instead, a very numerous and diverse set of demands for unskilled labor has evolved within the Klaten labor market. The nonfarm sector in this region is dominated by a wide range of small-scale industries such as food processing, textiles, batik, furniture and plywood manufacturing, and ironworks. Table 5.4 shows the number of industries and employees by size classification which were in operation in 1986. Most of the industries are house-hold industries, and demand for unskilled labor by this group is much higher than demand for labor from the larger industrial groups. Most household industries are not mechanized; inputs consist primarily of labor and raw materials.
Several farm laborers interviewed in Klaten work in construction for one to four months during the dry season either within their villages or in the nearby cities of Solo or Yogyakarta. Many others work in small rural industries on a part-time basis when they are unable to find work in the sawah. Interviews with unskilled laborers and owners of industries confirm a high degree of labor mobility between the agricultural and non-agricultural sectors of the rural economy. Many of the unskilled employees of the industries return to the sawah to participate in the paddy harvests. Others work in a variety of tasks in the cultivation of paddy as well as in industry on a part-time basis. All of the laborers interviewed had
knowledge of a range of employment opportunities and usually allocated their time to the most remunerative activity available. Table 5.5 lists the industries surveyed in Klaten and the wages paid for each activity.
Nonagricultural wages for unskilled laborers in factories are often lower than agricultural wages. Factory work is attractive because employment is full time or continuous for many months, the work is less physically
difficult, and the value of the product is not high (e.g., bread factory, tobacco factory, and woor [ingredients for clove cigarettes] manufacturing). Wages are the same or higher than agricultural wages for piecework labor and for physically difficult work, especially work on construction projects. For seasonal work in infrastructure projects, labor is taken directly from the agricultural sector, and contractors base their wage rates on local agricultural wages. Wages for unskilled laborers in the larger factories, such as the government textile firm and the publishing firm, are similar to those in agriculture, but benefits are provided and the work is often full time. For becak driving, the hours worked per day are long, but the intensity of work effort is less than that demanded of agricultural labor. For the most part, therefore, wage differentials are easily explained by factors other than market failures-duration of employment, difficulty of work, and the value of output.
The Rural Labor Market in Kabupatens Pinrang and Sidrap (South Sulawesi)
Employment activities differ greatly between rural areas on and off Java. South Sulawesi was chosen as an off Java survey site to provide a contrast with the rural labor market on Java. The results of the survey in South Sulawesi cannot be generalized for all areas outside of Java, but they do indicate broad employment patterns, characteristic of parts of Indonesia that have a much higher ratio of land to labor than does Java.
More than half of the sawah in Pinrang and Sidrap is now double cropped in rice. Large public sector investments in technical irrigation systems during the 1970s transformed a majority of the sawah in the region from rainfed to irrigated rice land. Agroclimatic conditions and labor constraints prevent farmers from planting paddy three times a year. Palawija crops are rarely grown in the sawah; low yields and the lack of adequate processing and distribution facilities in the area cause these crops to be much less profitable than rice. Rice varieties have a long growing cycle (IR 42 seed variety with 135 days to maturity), and there is usually a fallow period of a month or more between crops.
Labor inputs in paddy production in Pinrang-Sidrap are significantly lower than on Java. This difference arises because of the relative scarcity (and higher cost) of hired labor in South Sulawesi. Most families in Pinrang and Sidrap have access to land through purchase or sharecropping arrangements. The number of landless laborers available for hire is relatively low, and women of Buginese (an ethnic group native to South Sulawesi) origin do not participate in rice production apart from harvesting. The shortage of labor available to the agricultural sector in the region has led to the widespread use of family labor and gotong-royong
(mutual unpaid labor exchange between friends and neighbors) for most tasks in rice cultivation.
The use of hired labor for planting and harvesting has increased since the mid-ig8os. Most of the hired laborers are Javanese migrants who live in Polmas, a kabupaten neighboring Pinrang and Sidrap. Javanese women, as well as men, hire themselves out for planting, whereas preharvest workers of South Sulawesi origin, both hired and gotong-royong, are all male. The availability of migrant workers originating from Java has been essential to break the labor constraint during peak periods of demand in many areas. The proportion of migrant workers employed in rice production varies enormously by village depending on location and accessibility.
Virtually all of the male farmers interviewed in Pinrang-Sidrap work exclusively on the farm-in the sawah for most of the year when paddy is grown and in gardens during the fallow seasons. Although most women do not work outside of the household except during paddy harvests, a variety of home industries-primarily sewing and producing sutras (traditional dress) for local and regional markets-contributes to household income. Evidence of multiple earning activities on and off the farm, characteristic of the labor market on Java, is clearly lacking.
Many of the off-farm activities employing hired labor in Pinrang-Sidrap are closely connected to the rice sector. Examples include irrigation expansion and improvement projects, the manufacturing of bags for harvested and milled rice, retail stores for tractors, seeds, and other inputs, and private rice mills. Many of these activities hire local (mostly male) labor, when paddy is not planted, for wages equal to or higher than those earned in rice production. Table 5.6 shows a sample of employment opportunities and wages outside of rice production.
The majority of off-farm employment opportunities are related to the rice sector either directly through production linkages or indirectly through expenditures of rice earnings. For example, home building and home improvements, activities that employ many men from both the villages and the larger towns in the fallow season, are largely dependent on farm incomes generated from the previous rice harvests. Home industry employment for women, such as sewing, is also tied to local expenditures after the harvests.
Employment opportunities unrelated to the rice sector are less significant. A small percentage of the population in each village goes on merantau (work outside of the region for long periods of time, e.g., in Kalimantan or Malaysia). The proportion of the population on merantau has diminished since technical irrigation became widespread in the region and increased employment and income opportunities in rice production became available.
Comparison of Rural Labor Market Conditions on Java and in South Sulawesi
Recent field survey evidence on employment and wages in the rural labor market on Java does not support the hypothesis that there is a labor surplus economy in which job opportunities are limited and wages remain at a minimum subsistence level (Geertz 1963). Evidence from the 198os depicts a largely competitive rural labor market in which unskilled laborers move between the agricultural and nonagricultural sectors of the economy on a seasonal basis to find employment at comparable wages.
The wide range of employment opportunities available to unskilled workers provides evidence that the ability of landowners to exert monopsonistic power over landless laborers is much weaker than that observed by Hart in the 1970s (ig86a, ig86b) and has disappeared in many areas.
Agricultural labor is neither at the top or the bottom of the hierarchy of potential employment opportunities available to unskilled laborers on Java. The notion of landless laborers being pushed out of agricultural jobs into marginal activities with lower hourly wages is not supported by recent empirical evidence. A more plausible explanation is that unskilled labor is being pulled into the rapidly growing nonagricultural sector. Some agricultural workers have left the sector for full-time work, but many have increased part-time employment while maintaining their presence in agriculture. This growth has greatly diminished underemployment during seasons of slack agricultural demand. A large proportion of the laborers interviewed prefer to remain in agricultural work when they have adequate opportunities and can earn returns comparable to those in off-farm work.
The opportunity cost for unskilled laborers in rice production appears to be about equal to the wage rate paid in the agricultural sector. During the slack agricultural seasons, earnings from construction, becak driving, selling fruits and vegetables in the market, or employment in a variety of other formal and informal activities are on par with the agricultural wage rate. Labor demand in the rice sector remains important for rural income distribution and employment on a seasonal basis. But broader-based growth in rural incomes, which creates the demand for labor-intensive rural construction and services, and the expansion of employment opportunities in urban areas have become essential elements of the demand for unskilled labor in the rural economy.
Generalizations about labor market behavior on Java are not applicable to many regions off Java. In the South Sulawesi survey site, for example, the rural labor force is more dependent on the rice sector as a source of income and employment than are the areas surveyed on Java. Fewer off-farm employment opportunities exist in rural areas in South Sulawesi, and greater access to cultivable land for most villagers in Pinrang and Sidrap mitigates the need to pursue multiple-earning activities within each season. Although unskilled workers do migrate between islands in Indonesia, the rural labor markets between regions on and off Java are not yet completely integrated.
Institutional Changes in Labor-Hiring Practices
Adjustments in traditional employment practices in rice production offer further evidence of increased competitiveness within the rural labor
market. The changing role of cultural practices in determining employment and wage levels for unskilled workers within the rice sector is examined in this section. The premise is that labor-hiring institutions evolve in response to long-term shifts in labor demand and supply. In particular, cultural practices that serve to divide work opportunities among the largest possible number of laborers give way to piecework, cash payments, and contract hiring of labor teams when labor becomes scarce.
Labor-Hiring Institutions
Wide regional diversity in labor-hiring and payment arrangements exists within the Indonesian rice economy, both on and off Java. Existing labor-hiring arrangements include gotong-royong, bawon, ceblokan and kedokan, and tebasan. Although the specific terms of these arrangements may differ among regions, or even within villages, their structure is largely consistent throughout the rice economy.
Gotong-royong is the mutual exchange of unpaid labor between friends and neighbors within and outside of agricultural production. A reciprocity agreement exists in the institution of gotong-royong, but apart from meals no wages are paid for work performed. Within the rice sector, friends and neighbors work together on cultivation tasks and rotate between their farms until the job is completed for each farm. Gotong-royong is used most commonly for labor-intensive preharvest tasks such as planting.
Under the bawon system, anyone who wishes to harvest may do so (at least within a given region or village). Workers come to the farmers' fields on the day of the harvest, help with the harvesting, and divide among themselves a standard share of their harvest as a payment in kind. This share, usually between one-tenth and one-fifth of the amount harvested, differs between regions according to local scarcity of labor. By adapting to local labor market conditions, the bawon system ensures that returns per worker are lowest in regions of most ample labor supply. Incomes are thus spread among the largest possible number of workers in each region.
Ceblokan refers to the hiring arrangement for planting paddy in which a female laborer receives no payment for the task of planting, but she has exclusive access to the harvest and often earns a greater share of the harvest than under the traditional open-harvest system. The kedokan system is similar to the ceblokan system but includes additional tasks such as weeding. Although husbands and other family members are permitted to help the women with the harvest, the full wage share of the harvest is paid to the woman who performed the preharvest tasks. To ensure job security, female laborers are willing to defer receipt of wage income until after the harvest.
The tebasan system is a method of harvesting in which the farmer sells for cash the standing crop of paddy to a labor contractor (penebas) before the harvest. The penebas hires his own workers, often a group of laborers from his village who travel throughout the region harvesting for the same contractor.
Changes in Labor-Hiring Institutions
The prominence of traditional labor-hiring practices is influenced by local labor demand and supply conditions. But evidence from field surveys reveals considerable changes in hiring practices for both harvest and preharvest tasks in rice cultivation. For example, hiring patterns in Majalengka, where the system of ceblokan for planting has been used for many years, are beginning to change in response to regional labor shortages. In most villages surveyed, the farmers still use the ceblokan system; if there is abundant labor, the women are willing to plant and wait for the returns from the harvests. In areas of scarce labor, however, wages are paid in addition to the one-sixth to one-fifth share of the harvests for ceblokan workers. In the village of Kertaninagin (Kecamatan Kertajati), farmers began paying wages in 1985; 8o percent of the farmers paid wages in the first year, and now all farmers pay wages.
The ceblokan and kedokan systems of hiring labor for preharvest tasks still are used in some areas of Kediri, although cash wages have largely replaced these institutions for planting and weeding. Evidence of the decline in the use of the ceblokan and kedokan systems in Majalengka and Kediri is consistent with the empirical results of micro-studies in other areas of Java (Mintoro et al. 1984). In areas of seasonal labor shortages, caused by men migrating to cities or men and women working in rural off-farm activities, the institutions of ceblokan or kedokan have been replaced by wages. In villages with less access to off-farm work, ceblokan and kedokan continue to be used.
The choice of harvesting method and the share paid to harvesting workers similarly reflect local conditions in the labor market. In many parts of Java, the use of contract teams of labor for harvesting under the tebasan system has replaced the traditional bawon system of harvesting. In most cases, the tebasan system for harvesting is used to break a seasonal cash constraint of farmers (e.g., for education expenses) or to facilitate the organization and management of harvesting labor. Because of the growth in employment opportunities, the advantages of the bawon system in distributing income among the largest numbers of workers have become less important. The harvest share received by the workers on a tebasan team generally is lower than that paid to labor under the bawon system, but the work is more likely to be full time and thus provides a larger total income to the worker.
The use of contract labor arrangements with wages paid on a piecework basis also has become prevalent for certain preharvest tasks. In most of the villages surveyed in Klaten and Kediri, labor for planting is hired on a contract basis to ensure the farmer of adequate labor supplies at peak periods. The contract cost for hiring planting labor is generally higher than the cost of paying daily wages to individual workers. The piecework wage received by planting workers under the contract system allows farmers to pay laborers according to their productivity.
The evolution of labor-hiring practices in rice production in Pinrang and Sidrap has differed from that in the survey areas on Java. Many of the labor institutions on Java grew out of labor surplus conditions, whereas for decades labor availability for rice production in Pinrang-Sidrap has been scarce relative to the demand for labor services. Family labor is used for most preharvest tasks in rice production in response to the unavailability or high cost of hired labor. Gotong-royong traditionally has been used for the labor-intensive task of planting to ensure labor during peak periods of demand.
Throughout the region, the institution of gotong-royong for planting is being replaced by hired labor. The employment of hired labor for planting began in the early ig8os in most villages, and it is still widely used. The surveyed farmers who use hired labor report that they changed their practice because of the unavailability of gotong-royong labor for planting on a timely basis. Farmers who crop technically irrigated sawah are particularly dependent on timely production practices because of their irrigation schedules. The tebasan system for harvesting has not been introduced in Pinrang-Sidrap. Harvesting throughout the region continues to be done by the bawon system, and the harvesters receive one-tenth of the harvest.
Implications of Institutional Change
Increased rice production, rural economic growth, and the expansion of off-farm employment opportunities for unskilled workers have led to widespread changes in traditional hiring and payment practices in all of the survey regions. On Java, the use of institutional arrangements designed to spread returns to labor among a large number of workers is declining as demand for labor from outside the rice sector rises. Market wages have become a more prominent means of payment and labor allocation. The asymmetric adjustment in wages that normally occurs in the process of economic development-inflexibility of wages to move below a subsistence level under labor surplus conditions but flexible upward movement in wages in response to competitive conditions-has been reinforced by changes in labor-hiring practices in the Indonesian rice economy.
New institutional practices governing the organization of labor in the rice sector give increasing attention to emerging labor scarcities. Contractlabor groups have replaced daily wage labor in many regions, particularly for the labor-intensive tasks of planting, weeding, and harvesting. The use of contract labor reduces the possibility of labor shortages at peak seasons and enhances the efficiency of performance of each task. Groups are paid on a piecework basis according to the amount of land cultivated or the amount of paddy harvested. These adjustments in labor-hiring arrangements in response to rises in workers' marginal productivity are consistent with the development of an increasingly competitive labor market in rural Indonesia.
Wage Trends in the Rice Sector
The reduced prominence of labor surplus conditions in rural Java is reflected in the pattern of real wages within the rice sector as well as in changes in the use of traditional employment practices. This section examines secondary data on real wages for individual tasks in rice production from 1976 to 1988.3 The analysis is confined to the main rice-growing provinces on Java (West, Central, and East Java), where unskilled labor moves relatively freely between the farm and nonfarm sectors of the rural economy and hired labor is used extensively in the production of rice.
The Choice of Deflators
Widespread disagreement exists in the literature and among current researchers in Indonesia about the path of real wages in the rice sector. The controversy rests on two issues-the choice of deflator applied to the nominal wages, and the region and time period of analysis. Mazumdar and Sawit (1986) and Naylor (1989) have shown that a wide range of real wage trends can be calculated from available price data in Indonesia, even if a single series of nominal wages is used for all calculations. The results vary from a stagnant trend in real wages between 1976 and 1988, when a comprehensive index of household expenditures is used to deflate nominal wages, to a 50 to loo percent rise in real wages during the same period, when a simple rice price deflator is applied to nominal wages. This inconsistency in results makes it impossible accurately to assess real income growth and changes in the structure of the rural labor market based on credible real wage trends.
3Data on nominal wage rates have been collected regularly by the Central Bureau of Statistics (BPS) since 1976 in provinces on Java and since 198o in ten other provinces. The BPS survey includes 18o farmers in West Java, 261 farmers in Central Java, 261 farmers in East Java, and 36 farmers in Yogyakarta. The number of farmers in the other provinces is much smaller. For further reference on these surveys, see Korns (1988) and Godfrey (1987). The analysis in this section focuses on hoeing (male) and weeding (female) wages from the BPS survey. These wages are paid largely in cash rather than in kind.
Income and price changes in the rural economy lead to systematic changes in consumption bundles over time, which bias estimates of real wage trends based on any deflator with fixed expenditure weights.4 Although rice constitutes a major expenditure item of rural households (between 2o and 2.5 percent for low- and medium-income households), the choice of the rice price as a deflator does not incorporate the full range of household expenditures such as clothing, housing (purchase or building materials), and transportation. As incomes rise and a smaller proportion of household income is spent on food items, the use of a broader index of household expenditures becomes essential .
Errors in weighting and calculating the various components of a comprehensive household expenditure index are more prevalent than in constructing a simple index of rice prices. Close examination of Indonesia's rural consumer price index by province reveals large distortions within the relatively minor expenditure category of chilies.6 Removal of the misleading component of the index yields a plausible representation of real wages .7
The results obtained by using a revised index of consumption expenditures (excluding chilies) as a deflator for wages in rice production are shown in Figures 5.1-5.6.8 Data for hoeing are used to represent move-
4The Laspeyres index and the Paasche index are commonly used price indexes in economic welfare analysis. The Laspeyres index weights prices by quantities in the base year, whereas the Paasche index weights prices by quantities in the final year. Accordingly, the Laspeyres index overestimates a rise in the cost of the original bundle and the Paasche index underestimates an increase in cost. The consumer price index paid by farmers that is used in this analysis is calculated by the Laspeyres weighting method. As a result, real wage trends based on the index tend to be understated. Because the results err on the conservative side, a calculated increase in real wages would indicate a definite improvement in the welfare of agricultural laborers.
SThis trend is based on Engel's Law, which states that the income elasticity for food in the aggregate is less than one and approaches zero as incomes rise.
6The rural consumer price index used in this analysis is the household consumption component of the farmers' terms of trade index, 1976 = ioo (BPS). This index includes the costs of food, housing, clothing, and other goods and services, weighted by average per capita expenditures for rural households in each province. The consumer price data in the farmers' terms of trade index are often used as a deflator in the evaluation of real wages in rice production since both the wage and price series are collected simultaneously from the same sample.
7Extraordinary growth in chili prices between 1981 and 1988 caused the food component of the consumer price index to rise disproportionately. Further investigation of the reporting process at the village level is necessary before recorded chili prices can be used with confidence as part of a deflator of rural household expenditures. The BPS released a revised index of the farmers' terms of trade in mid-1989, which has a new base of 1983 = 100. Because of substantial changes in all expenditure categories and reweighting among the components of the index, the original and revised price series cannot be spliced to form a consistent deflator from 1976 to 1988. For further detail on the index problem, see Naylor (1990)
8The revised index consists of the index of prices paid by farmers in the BPS farmers' terms of trade series without the chili component.
ments in male wages; those for weeding represent female wages. Trends in real wages rise throughout the first half of the ig8os for both male and female tasks in the main rice-producing provinces on Java. Growth in real wages during the early ig8os is consistent with several reinforcing influences in the Indonesian economy: rising demand for labor in off-farm activities stimulated by revenues earned from rice production; increased government spending in rural areas leading to the expansion of employment opportunities; and the economy-wide effects of the oil boom on the urban demand for labor (Manning 1986a, 1986b; Godfrey 1987; Collier et al. 1988). Although the increase in real wages does not show the cause of growing labor demand, wage movements are consistent with a tightening of the rural labor market and an absence of surplus labor.
Region and Time Period of Analysis
Figures 5.1-5.6 also illustrate the regional disparity in real wages between 1976 and 1988. Real wage growth was rapid in East Java, slower in Central Java, and almost stagnant in West Java. There is no consistent difference between the trends in male and female wages across regions. Wage levels for both male and female tasks remained higher in West Java than in East and Central Java throughout the reporting period (Table 5.7). Relatively high wage levels in West Java are the result of earlier rapid growth associated with expansion of the Jakarta economy. Nevertheless, the data show some convergence in wage levels. The fastest growth in wages during 1976-88 occurred in East and Central Java, the provinces with the lowest initial wage rates. A converging pattern of agricultural wages across regions reflects broad-based rural economic growth consistent with the expansion in the rice economy throughout Java during the early 198os.
Despite their rapid growth in the ig8os, wages in rice production in Central Java remain substantially below those paid in East and West Java. The larger supply of labor in Central Java compared with that of other provinces on Java is probably the principal explanation for the relatively low wage levels. In 1985, Central Java had a reported population density of almost 8oo people per square kilometer, whereas population densities in East and West Java were roughly 650 people per square kilometer (BPS 1987). Because transportation within and between provinces on Java is well developed, wage differentials most likely reflect the psychic and resource costs of migration.
Two recent disquieting turns of events in the rural labor market are apparent from the aggregate data. Between 1985 and 1987, real wages for unskilled workers in the rice sector remained stagnant and in some years
declined in East and Central Java (Table 5.7). In addition, the converging pattern in real wage trends across provinces that was evident during the early ig8os was reversed in recent years. Real wage growth in West Java, the province with the highest wage levels, exceeded that in East and Central Java between 1985 and 1987. Wage levels in Central Java fell further below those in East and West Java during the same period. Stagnant real wage rates in East and Central Java seem to indicate that improvements in labor welfare, which were pronounced during the first half of the ig8os, have not been sustained since 1985. Slower growth in labor demand in both urban and rural locations directly reflects the temporary downturn in economic activity associated with the oilled recession in Indonesia.
More time and evidence are needed before definite conclusions regarding welfare changes can be made, however. A surge in economic growth, led by marked improvements in manufacturing output and nonoil exports, has begun as Indonesia enters the REPELITA V period (the fifth five-year plan, 1989-93). Real wages in rice production in East and Central Java increased in 1988 and could rise substantially during the next five years, especially if labor demand expands in nonagricultural sectors and in the production of agricultural commodities such as horticultural crops and livestock products, which are labor-intensive and have high income elasticities of demand.
Conclusion
Field survey results on employment opportunities for unskilled workers, adjustments in labor-hiring practices in the rice sector, and rising real wages in agricultural and nonagricultural activities provide evidence of a broadly competitive and well-integrated rural labor market on Java in the ig8os. Employment in the agricultural sector remains an important source of income for landless laborers on a seasonal basis. But nonfarm activities play an increasingly dominant role in contributing to rural household employment and income. Wages in rice cultivation are roughly comparable to wages for unskilled labor in seasonal jobs in rural non-agricultural activities. Unskilled workers face a wide range of income-earning alternatives and allocate their time to various activities depending on relative returns.
In response to increased competition in the labor market, many farmers on Java have discarded the use of traditional labor-hiring practices, previously designed to accommodate large surpluses of labor, and instead pay cash wages to ensure adequate and timely availability of labor. Contract arrangements are used extensively in rice production, especially in planting. The evolution of payment and employment practices reflects changing demand and supply in the labor market. Labor surplus conditions, defined by an unlimited supply of labor available at a minimum subsistence wage, no longer exist on Java. Wages for agricultural workers are determined endogenously by demand and supply in the rural labor market, and farmers pay competitive wages to secure labor on a timely basis.
Appendix 5.1. Field Survey Data on Wages in Rice Production
Appendix 5.1 contains data on local wages by task in rice production for the survey areas on Java. The appendix is designed to supplement the data provided in the text for readers interested in the level and diversity of wage rates at the local level. The first section of data includes wage rates (Rp per day and Rp per hour) for hoeing (male) and weeding (mostly female) tasks in rice production by village for each of the survey regions on Java. The regions are presented in the following order: Kabupaten Klaten in Central Java (Tables A5.1-5.4); Kabupaten Majalengka in West Java (Tables A5.5-5.8); and Kabupaten Kediri in East Java (Tables A5.9-5.12).
The data on hoeing and weeding wages by region are followed by village-level data on planting wages under contract arrangements for Kabupaten Klaten in Central Java (Tables A5.13-5.14) and Kabupaten Kediri in East Java (Table A5.15). Contract planting arrangements were not common in 1988 in Kabupaten Majalengka in West Java. The wage data for planting provide a cost comparison between the contract and noncontract systems of hiring labor for planting. The final table (A5.16) illustrates the wide variation in harvest shares paid to labor in the bawon and tebasan systems between villages in Kabupaten Klaten (Central Java).
Tebasan is a contract arrangement by which a farmer sells a standing crop of paddy to a labor contractor (penebas), who hires a team of workers to harvest the sawah.
|
|