Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

Labour Force Participation: Employment Patterns

Respondents, irrespective of age, were considered to be members of the labour force if they worked in any capacity either on or off their own farms. Generally, thiscategory included all family members except young children, those few with disabilities, and those who did not have to work (e.g. an older, landed mother-in-law). The predominant work pattern was recorded for all members (table 19). For example, if a householder worked on his or her own farm for five days and as an agricultural labourer for two days, the individual work pattern was noted as "own farm." However, if there was further fragmentation, the "combination code" was used to denote a person who had no clear-cut pattern. Employment on the family farm was the major occupation, and accounted for nearly half the total work time. Employment as an agricultural labourer accounted for almost one-quarter of the work time.

TABLE 20. Percentage of households with various numbers of working members

TABLE 21. Reasons for not working: individual responses by day

a. Includes people who might have spent Part of the day In market-town transaction business.
b. Death, politics, visitors.

Working Members

Approximately 180 of the 350 respondents (about 50 per cent) made up the labour force; on a weekly basis between 44 per cent and 80 per cent of the respondents worked. As determined from the aggregate household data, nearly half of the households had two or three working members. There were no working members in only 2.5 per cent of the households, and in no households at any time was there only one working member. These data are summarized in table 20.

Labour Force Participation: Days Worked by Round

Total days of labour-force participation by each round or week were calculated (fig. 4). Peak and slack periods are quite evident and relate to the intensity of agricultural activities as well as to cultural and social events. Weeks 2 and 3, when employment was very low, were deemed auspicious times for marriage in the village. This period was followed by a time of full employment during harvest. Employment peaked again during the monsoon season (weeks 15 to 26) as paddy was transplanted and then weeded. A period of relative inactivity (weeks 27 to 30) preceded harvest. The sharp dip at week 40 reflects the major festival, Sankranti. When these same data are examined by household, that is, person days worked by household for each week, peaks are evident for each seven-day period, so that, if a person worked, he or she generally worked for a full week.

Reasons for Not Working

The reasons for not working on any day are presented in table 21. Voluntary unemployment and absence from the village, when people were visiting relatives, were the most frequently cited reasons. Other major reasons included domestic activities, such as child care, involuntary unemployment (when an individual sought work but did not get any), health, and marriages or festivals. The percentage of those people who did not work because of illness ranged from less than 1 per cent to slightly more than 5 per cent. Illness affected this group quite randomly; at only one time during the study was the rate of illness uniformly low-March and April, i.e. the summer season. During the remainder of the year, a high rate of illness in one week (4 per cent) was often followed by a low rate the next (1 per cent or less) with no discernible pattern. The percentage of those involuntarily unemployed ranged from 0 to 15 per cent. A six-week period (weeks 46 to 51) in February and March was the only consistently high period of this type of unemployment: the figures at this time ranged from 7.7 per cent to 14.1 per cent. Otherwise, rates were extremely low (mean 3.3 per cent).

Total days worked per week per round

TABLE 22. Wage rates for agricultural labour, cash or in-kind, per day

a. On a contract basis only
b. Lower value during slack season

Wage Rates for Agricultural Labourers

Rates of pay for agricultural workers were averaged by category of employment (table 22). Little difference was found between seasons except for weeding and shelling of groundnuts: slack-season daily wages were 3 rupees while peak-season wages were 5 rupees per day (US$1.00= 1 2.5 rupees).

Wages were paid in kind (as paddy) during a four-week period in the summer season and for a longer period during the kharif and post-kharif season, which lasted from mid-August to the end of December (weeks 17 to 38). The value of such in-kind payment equalled or exceeded cash-wage rates. Sorghum was paid only twice and in small amounts (9 and 39 kg).

Crop Production

Paddy (rice) is the principal crop grown in Dokur, owing to the availability of irrigation (now about 50 per cent of all cropped land). Groundnuts are also grown in large quantities as a cash crop. Sorghum (jowar), pigeon pea, and finger millet (rag) constitute the remaining crops, which are grown in smaller quantities and are thought of as subsistence crops.

Although yields of rabi paddy are higher, the largest single harvest of the season is kharif paddy, because of the acreage cultivated. Less land is used for groundnut cultivation and fewer farmers are growing the crop each year owing to the presence of a widespread groundnut virus. The remaining crops are grown principally as dryland crops.

During the study period any fertilizers or pesticides used by farmers invariably went towards improving either of the two major crops. Cropping activities fall into three major seasons (table 23); the summer season crop is generally small and restricted to those farmers who have adequate irrigation.

The yields of the major crops of ail participant farms indicate the magnitude of the harvest. When examined along with the labour-force graph, the yield data give a picture of the seasonal aspects of crop production and labour activity. From weeks 19 to 37 (mid-August to the end of December), labourers received wages in kind (as paddy) as well as in cash. These wages increased markedly during the November-December harvest.

The range of yields for each farm was wide because of variations in acreage under cultivation, access to irrigation, and other farm inputs. Concerning the latter, smaller farmers were clearly at a disadvantage, since their families were forced to work as labourers to supplement their incomes, and this often meant that they had to delay sowing, transplanting, or harvesting their own plots until the optimal time had passed.

TABLE 23. Crop production schedule, by week

TABLE 24. Health classification of study respondents on physical examination.

a. Only 90 per cent of the study respondents underwent physical examination. The percentages refer to those who were examined.

Physical Health Examinations

The physical examinations were of limited usefulness for comparative analysis because the number of examinations per child was highly variable.

Some summary data are, however, useful indicators of health problems. Of the 111 members whose stool samples were collected (32 per cent of the respondents), half (49.5) per cent) were free of parasites. This sample was, however, biased towards children. The remaining half harboured amoebic cysts, Giardia, Ascaris, hookworm, tapeworm, and liver fluke. The majority of those infected had two or more parasites.

Approximately 10 per cent of the respondents (all male) were never examined, principally because of shyness. Table 24 shows the results for the remaining 90 per cent.

Fewer than 10 per cent of the study members were afflicted with an ailment that was serious, e.g. otitis media, hypertension, pneumonia, asthma, tuberculosis, and urinary tract infection. The most frequently encountered minor conditions were wounds, arthritis, colds, and cataracts.

Assets

Assets consisted principally of land which was valued at between 10,000 and 20,000 rupees per acre for wet land and between 3,000 and 6,000 rupees for dry land, including pasture land. Other major assets included houses, outbuildings, gold and silver jewellery, farm equipment, and animals.

Mean assets for the 40 families were valued at 113,600 rupees (S.D. + 166,070) or approximately $9,500 per family. Assets ranged from 3,100 rupees for a landless family to 945,000 rupees for a large family with 32 acres of wet land.

Relationships between Food Availability and Energy Needs

Correlations between weekly food availability and energy requirements could not be carried out as had been planned for two unforeseen reasons. First, it was impossible to obtain reliable information on stocks at the beginning of the study. Second, wages were often paid weeks after work was performed, and the complex system of borrowing and repayment that occurred during these periods was beyond our ability to assess accurately. However, family energy needs were assessed, as well as the actual availability of food grains, and taken together these assessments form the basis for the determination of levels of adequacy.

These family energy needs, calculated as a cereal equivalent, in kilograms, were estimated as follows. The recommended cereal allowances (RCA) for various ages were accepted as an indicator of family needs [8]; 10 per cent of the RCA was added to this value to compensate for the typical Dokur village diet, which is almost wholly cereals. The RCA was also adjusted if a family was away temporarily, but was not adjusted for daily fluctuations, such as those resulting from casually absent members or guests. The adjusted figure was then multiplied by 365, which pro" vided the approximate yearly total for kilograms of cereal required. To determine if this cereal requirement was equal to the total potential cereal availability, the latter was estimated for each family for the year as the sum of:

1. The cereal equivalent in kilograms of the monetary value of the groundnuts produced for the year.
2. The cereal equivalent in kilograms from total wages received for the year.
3. The kilograms of cereals produced and/or received as in-kind payment of wages.

Of the 40 study families, 11 (27.5 per cent) did not have enough food to cover basic needs, as determined by the method just described. However, eight of these families managed to meet at least 80 per cent of their energy needs, using supplementary means to do so.

Relationships between Study Groups with Respect to Anthropometric Indices, Income, Illness, and Energy Intakes

As mentioned earlier, the study sample was divided into two groups based upon female labour-force participation. Although there were women from nearly all categories of farm size, the majority of them were from the smaller farms or, in two cases, from landless families. Non parametric t-tests indicated no difference between the two groups in weight for height. A significant difference was noted between the two groups in three areas: total income, percentage of days ill, and the percentage of the energy requirement met. Total wage income was significantly higher for group 1, i.e. women who worked on their own farms (t-value = 5.41, P < 0.001); the same was true with respect to the adequacy of energy intake (t-value = 4.91, P < 0.001) and illness (t-value = 5.94, P < 0.001).

Credit and Debt

Total annual income from all sources was calculated for each household using the weekly transaction data. In 18 of 40 households (45 per cent the family was in debt. An analysis of weekly income and expenditure for each family revealed that all were in debt some of the time, although there was no seasonal trend. Instead, there was a pattern of large sums being received as back wages or from calling in money owed to the family, which in turn would be used within one to two weeks to pay off debts called in by someone else.

There was no correlation between the percentage of the family energy intake met and income (from all sources), when energy intake and the prior week's income were correlated. An influential factor must certainly have been the pattern of credit and debt characteristic of the study households.

The weekly transaction data also provided information on the families who failed to have adequate harvests or income from wages. They had all been able to borrow either cooked food, raw rice, or paddy. Paddy was borrowed in small quantities, without interest, or by the bag, with interest. Cooked food from a caste-fellow or a large farmer-who clearly had stocks and for whom the family worked-was always available. Caste-fellows and neighbours also provided small quantities of rice or paddy. When a full bag of paddy was borrowed, it was always with interest. in the case of a small farmer, repayment was sometimes based on the borrower's own cropping pattern. If he harvested only one crop a year, he was not required to return the paddy until after his next harvest.

The willingness to loan crops or food has not always been a part of village life, but has developed gradually. Nearly six years ago the farmers in Dokur became eligible for easier credit with the State Bank of India. At the same time the Land Mortgage Bank expanded its services. During this period a considerable number of all farmers took out loans to increase irrigation. In 1976 only 30 per cent of the cultivated land was irrigated as opposed to 50 per cent in 1982 [3]. A significant proportion of the newly irrigated land has been planted with rabi groundnut, a cash crop that pays from two to five times more per kilo than paddy. Therefore, farmers have had little difficulty making their loan repayments.

A study of the village labour market, including labour force participation data, wage rates, and land transaction records, revealed that small farmers have gradually sold their land to larger farmers and shepherds, who have an excess of cash due to an increase in meat and wool prices, and have then joined the labour pool. Data on wage rates and labour-market participation from 1976 to 1979 indicate that the sharp increases in wages over the intervening years as well as vastly increased opportunities for employment have made small-scale farming less attractive [24]. An additional incentive for many small farmers to sell their land and join the labour ranks has been a sharp increase in land prices since the late 1970s.

Thus, larger landowners have rapidly expanded their resources and their debts and have become labour-dependent because farming is not mechanized. Their labour dependency has led them to ensure that labourers do not fall short of food. For this reason, expected seasonal variations in household food intake did not occur to any significant extent.

Statistical Analyses: Determinants of Energy Intake

Non-parametric correlations (Spearman-R) were carried out prior to running stepwise regressions. It was thought that the correlations would provide useful indicators for the regression analyses.

Non-parametric Correlations

Correlations were carried out between percentage of family energy requirement met and household variables using the weekly household data, except where specified:

• PCWF = percentage of time worked by females 12 years or older.
• PCILL = percentage of days ill/total member days.
• NINC = net income.
• WINCO = income from wages in rupees.
• NWK = number of working members.
• WP = principal work pattern (on/off farm)- dummy variable.
• NMEM = number of household members.
• WETLAND = acres of wet land farmed.
• UPLAND = total land cultivated, acres, last kharif.
• TOTASS = total assets, in rupees.
• H4WK = harvest during previous four weeks- dummy variable.
• PCKCAL = percentage of family energy requirement.

The PCKCAL variable was created by dividing the total family energy consumption by the total requirement specified by Gopalan [8].

The correlations were done only for the first 23 weeks of data. Significant correlations are shown in table 25.

TABLE 25. Result of Spearman Rank Order Correlations test of percentage of average family energy requirement met and household variables

a. Significance level = 0.05.

TABLE 26. Results Of multiple regression analysis of percentage of average family energy requirement met and household variables

Regression Analysis

Stepwise regression analysis was carried out separately for each of the 51 weeks for which dietary recalls were available. The dependent variable was PCKCAL, the percentage of the family energy requirement met, as described previously, and the independent variables were the eleven variables tested for non-parametric correlations.

The variables listed in table 26 were significant with the tolerance level for the regressions set at 0.05.

That an increase in female labour-force participation

(PCWF) during the slack season (weeks 1 to 13) increased family energy intake was not surprising. There was little work except for women at this time of the year, which covered the period of the dry season, when there is a small summer harvest, to the early rainy season, when women begin to transplant paddy. The R-squares for these equations ranged from 0.13 to 0.25.

Having harvested a crop within the four weeks (H4WK) previous to the dietary recall was significant just four times (weeks 24 to 33) and accounted for 11 to 18 per cent of the variation. These weeks corresponded to the early months of the harvest season.

The illness variable (PCILL) was significant three times and was negative: as the percentage of time that householders were ill increased, family energy intake decreased. The three consecutive weeks when the illness variable was significant (21 to 33) fell within the period of high illness rates for both adults and children (figs. 3 and 4). R-squares ranged from 0.14 to 0.27.

No variables were significant (at the 5 per cent level) for the 24-week period. In those weeks where there was a significant variable (even once), the R-squares ranged from 0.11 to 0.27.

continue

Contents - Previous - Next