Document(s) 4 de 16

It has already been mentioned in the last chapter that the Task Force of the Planning Commission on Employment Opportunities (July 2001) had in its Report (popularly known as the 'Ahluwalia Report') focused attention on the sharp decline in employment growth. The Committee rightly pointed out that while a part of the slowdown in employment growth was due to an increase in the rate of unemployment, much the more important part of the decline has to be ascribed to a slow down in the growth of the labor force. This is largely because of a fall in the participation rates (PRs) as measured by the NSS.

The Committee did not take a firm position on the reasons for the fall in participation rates, but the arguments presented included non-economic factors. e.g., the expected shift in the activity status of the younger age group towards education; increase in the share of the population aged 60 and over; a reversal or 'correction' of the increase in PRs of certain age groups recorded in the 1993–1994 NSS (p. 45). At the same time its assertion that the decline in the PRs in the prime age group is 'within the margin of sampling errors', while partly applicable to males, is, as we shall see, certainly not true for females.

A follow-up Report by a 'Special Group' of the Planning Commission (May 2002) firmly put its emphasis on a slowdown in the growth of demand for labor as the culprit on the observed trends in labor force and employment growth. Noting that the fall in employment growth was accompanied by a higher rate of GDP growth, the Committee concluded: 'It means that the capacity for job creation per unit of output went down about three times compared with that in the 1980s and the early 90s' (ibid., p. 336). It suggested that the nature of economic growth had become more capital-intensive, both due to structural changes, and the 'rightsizing' of labor use.

In this chapter we shall go into some detail into the reasons for the fall in participation rates which is the main cause of the observed decline in employment rate. Our major purpose would be to see if we can confirm in any conclusive way if this is the result more of demand-side developments (as implied by the 'Special Group' Report).

The theoretical perspective

The measured labor force at any point of time (and the volume of employment) is determined by the equation of the demand and supply functions of labor. If the observed magnitude changes over time it might be because of the shift of the demand or the supply functions or by a combination of the shifts in both. Thus if the labor force or employment growth falls over time we cannot conclude that it is because of a fall in the demand for labor. It might easily be caused more by a fall in the supply of labor. We cannot measure the contribution of supply and demand factors to the observed slowdown without a fully specified and estimated model. While this might be difficult to achieve we can at least infer the relative importance of the two sets of factors qualitatively by looking at trends in wage levels over the time periods covered. This can be illustrated by Figure 4.1.

The Figure 4.1 shows the supply-and-demand framework for wage determination in a dynamic setting. The x-axis measures participation rate while the y-axis measures the wage per worker. The participation rate in effect measures the flow of labor per unit of time out of the potential stock. The demand function is then downward sloping: the lower the wage level the more is the flow of labor sought by employers from the existing stock. The upward sloping supply function is propelled by the 'substitution effect' of changes in wage levels. A higher wage will see a higher rate of allocation of time to the labor market at the expense of other activities, like leisure, household work or education.

As we have seen the post-reform period in India saw a significant fall in the rate of growth of employment compared to the previous decade. It will be shown in the next section that this decline was the counterpart of a decline in the participation rate In the competitive labor-market framework presented in Figure 4.1, this must imply that, if this decline has been caused by a shift in the demand function with a relatively unchanged supply function, the wage per worker must fall in the second period (situation 2 in Figure 4.1). We have already seen in the last chapter that this has not been the case. The real wage has increased in the post-reform period. This outcome is only possible if the supply function has shifted upward. Such a shift is possible if the 'income effect' of higher welfare levels causes household members to supply less labor to the market for a given wage growth. An increasing growth rate of wages could indeed both cause and sustain such an increase in household income levels.

Figure 4.1 Wage-determination framework.

In the empirical work presented below we shall be investigating the possibility of such an income effect for different demographic groups and sectors.

It is of course possible that the wage determination in the Indian economy cannot be interpreted within the competitive labor-market framework in the post-reform years. The higher rate of wages might have been the result of institutional factors. In this case the rate of employment growth (and consequently participation) could fall in response to the higher wage growth, and the gap between the growth of labor supply and of labor demand would have opened up leading to a higher rate of unemployment (situation 3 in the Figure 4.1). In fact the second Report of the Planning Commission hinted at this possibility without actually analyzing it in detail. There are two objections to this alternative hypothesis. The rate of unemployment did increase in the post-reform years, but we shall see in the discussion below that the magnitude of this increase was not really large enough to account for the large fall in the growth rate of employment. Second, and more crucially, it is hard to believe that institutional determination of the higher growth rate of wages would be at all realistic in an economy in which the bulk of employment is in the 'unorganized' sector.

There is a third possibility which is suggested by the 'segmented labor market' model. Segmentation implies that labor markets in the different segments are subject to different sets of supply-and-demand factors. In this case falling demand in one segment causing reduced participation might coexist with increased demand in other segments putting an upward pressure on wages. If, then, the observed data refer to participation in the first segment and wages in the second, we might indeed have the apparently inconsistent picture referred to above with slackening demand coexisting with rising wages in the aggregate. Segmentation might be caused by many factors and depending on the focus of our analysis could refer to gender, formal-informal, caste and many others. In the context of the current issue the factor we should emphasize is the distinction between primary and secondary earners. In the methodology of the NSS the former is called Usual Principal Status (UPS) and the latter Subsidiary Status (SS). This distinction is important in the context of the issue being discussed, because as we shall see a substantial part of the fluctuation in participation has taken place in the market for SS workers, while the wage or earnings data which are reported are for the markets of UPS workers.

The SS labor market is dominated by females, the majority in self-employed status, who are generally part of agricultural households and divide their time between household activity and work on the family farm as required. The amount of time spent on the latter is not measured by the NSS which only records the number such workers during the period of the survey. During an upsurge in economic activity, as might have happened during the introduction of new technology in paddy in the early eighties, the increase in demand for labor might be partly met by an increase in workers of secondary status working for varying lengths of time. Over a period of time as the labor market settles down to the new level of activity if the technical change is of a lasting kind, a new deployment of labor would normally be worked out by farmers in which some of the secondary workers used would be replaced by UPS workers on more permanent basis. Since the time spent by the latter on economic work is normally much larger than that by secondary workers, the numbers recorded as employed by the total (including both UPS and USS workers) might actually fall although the number of UPS workers might have increased. Thus the total employment figure might give a misleading idea of the change in demand in the labor market. We have to look specifically to the numbers in UPS status to see how the employment change in relate the recorded change in wage levels.

We will follow this line of analysis in the empirical work reported below.

It might be objected that in the discussion above we might have overemphasized the importance of wage workers in the Indian labor markets. In so far there is a large presence of self-employed workers, an analysis focused on wage trends might be misleading.

How important is wage labor in the Indian labor market?

The share of wage workers in UPS (principal) employment is more than half–slightly less in rural areas. Some part of wage labor is, however, supplied by secondary workers. It is important to get an idea about what proportion of the UPSS employed (principal and subsidiary together) actually participate in the labor market as wage labor either as a first or a second job. This total gives us the proportion of the employed who actually respond to wage signals in the labor market directly. Looking at the (larger) UPSS labor force as a whole, it was seen that the proportion of wage workers goes up (relative to the UPS count) by 3.5 percent in rural areas but it marginally declines in the urban areas. Evidently in rural areas more of the total enter the labor market as wage workers in subsidiary employment. In the urban areas relatively more workers in secondary status are to be found in the non-wage employment segment.

Decline in labor-force growth and distribution of fall in LFPR among different demographic groups

The fall in the rate of growth of employment in the post-reform period can be shown to be basically due to the fall in the participation rate. It has already been pointed out in the last chapter that the rate of unemployment in the Indian labor market is low, either with or without the secondary workers. Even if the rates of unemployment might have gone up in the post-reform era a bit its overall impact on the employment growth is small.

Our discussion in this section, therefore, concentrates on the changes in the participation rates, and the possible reasons for these changes.

We first set out the change in the rates of growth of the total (UPSS) labor force between the different NSS rounds.

Table 4.1 suggests that there was substantial decline in the labor-force growth from the 1980s (38th to 50th round or 1983 to 1993–1994) to the 1990s (50th to 55th round or 1993–1994 to 1999–2000) from 1.71 percent to 1.23 percent. The decline in the labor force could be due to decline in the working-age population or due to the decline in the labor-force participation rate (LFPR). Keeping the LFPR of 1999–2000 the same as 1993–1994 we find that hypothetical labor-force growth would have been much higher at 2.16 percent–much higher that even in the 1980s. It clearly shows that it is the decline in the LFPR that is mainly responsible for slower growth of employment in the 1990s.

We examined separate graphs for the age-specific LFPRs for all the three rounds for rural male, rural female, urban male and urban female.

The overall impression from the four profiles of age-specific PRs is that the most important change in the post-reform period is the decline in female participation, particularly in the rural economy. As far as males are concerned, the profiles for both the rural and the urban areas showed marginal declines in LFPR in age groups 5–19 years and 59+ age groups in 1990s. The decline in male LFPR in 5–19 years was substantial in the 1980s but slowed down in the 1990s–in both rural and urban areas. By contrast, the female LFPR showed a decline in all age groups between 1993–1994 and 1999–2000. In the previous period between 1983 and 1993–1994 the decline in both rural and urban areas were sharper in 5–19 age groups, but there was no substantial changes in the older age groups. The distinctive change in the latest period is the marked decline in female LFPR in the working age group as well.

As a substantial proportion of females participate in the labor market in a subsidiary capacity (i.e., enter the labor market only for a part of the year) it is worthwhile to look at female LFPR separately for UPS (Principal) and SS (Subsidiary) categories. This is portrayed in Figures 4.2a to 4.2d.

Figure 4.2a Rural female UPS labor-force participation rate.

Figure 4.2b Urban female UPS labor-force participation rate.

As in the case of males, the UPS female LFPR in all areas showed marginal decline only in the 5–19 age group in the 1990s most of the decline having taken place in the earlier decade. In the urban areas a marginal decline was observed in other age groups as well. But the real dramatic changes seem to have taken place in the category of subsidiary female labor. The LFPR graphs in both rural and urban areas showed shift inwards in the 1990s–signifying a decline in LFPR in all age groups. This contrasts strongly with the movement in the previous period in the 1980s, the adult age groups of 25–49 years showing a substantial increase in LFPR in both rural and urban areas. The patterns and the nature of shift of the LFPR graphs suggest that there was an upsurge in female subsidiary labor demand in the period between 38th and 50th rounds and females belonging to the age group 25–49 were in the best position to respond to it. However, between the 50th and the 55th rounds there seems to have been a substantial decline in the demand for subsidiary workers and it is reflected in the inward movement of the entire LFPR curve for female subsidiaries (It is to be remembered that for the subsidiary labor force there is no unemployment, all are employed.) This is an important point–which is valid both for the rural and the urban sectors, but quantitatively more so for the rural.

Figure 4.2c Rural female subsidiary labor-force participation rate.

Figure 4.2d Urban female subsidiary labor-force participation rate.

Accounting for the decline in the labor force in the nineties

Three significant points emerged from the discussion in the last section: First, the decline in the 5–19 age group of the primary labor force (UPS); Second, a marginal decline in the 59+ UPS labor; last, a substantial decline in female subsidiary labor supply.

The question is what is the relative contribution of each of these segments to the total decline in labor supply in the 1990s? Table 4.2 throws light on this question. It presents the actual (A) and hypothetical (D) labor force, the latter on the assumption of no change in age-group specific LFPRs between the 50th and 55th rounds.¹

The results show that decline in LFPR actually contributed to a fall of 23 million in the labor force. The last column shows that female subsidiary labor force (of all ages) contributed 47 percent of the total decline. The 5–19 UPS LFPR contributed 30 percent and aged LFPR (59+) contributed another 10 percent. These three factors combined accounted for as much as 88 percent of the total hypothetical labor force decline in the 1990s. Clearly the first component–the withdrawal of subsidiary females from the labor force–leads the list in terms of the diagnosis of the observed fall in employment growth. But before coming to this topic, we discuss briefly the fall in LFPR in the 5–19 and 59+ age groups.

5–19 UPS labor-force participation rate

The distribution of all persons in this age group in different principal activities is given in Table 4.3.

One can clearly see that the increase in proportion of students is the main factor responsible for decline in the work participation rate in both the rural and the urban sectors. There is a difference between the two periods: between 1983 and 1993–1994 and the subsequent period of the 1990s. The earlier period experienced substantial jump in the category of students but the shift was relatively more from the category of 'nowhere children' (doing nothing). In the latter period, it was largely a shift from UPS worker and domestic work. However, the withdrawal from work in absolute term was higher in the 1980s compared with the 1990s. We conclude that it is either demand for education and/or better educational facilities that seem to be the prime reason for decline in labor-force participation in this age group. However, increase in the students' participation rate in this age group seems to be tapering off in the urban areas in the 1990s (Table 4.3).

We examined the relationship of the participation of this demographic group to the levels of the expenditure per capita of the households to which they belonged. The tabulation was done separately for the rural and the urban households. Participation was found to be negatively related to household expenditure levels in both sectors, but the negative relationship was stronger in the urban areas. The slope of the negative relationship clearly decreased over time, but was still strong in the 1999–2000 survey, particularly in the urban sector.

Aged 59+ labor-force participation rate (LFPR)

The post-liberalization years saw a reversal in participation trends in the rural sector for both males and females, but not in the urban areas. Participation rates fell in contrast to the positive trend in the previous decade. In the urban sector the relatively slow decline in participation already noticed in the previous decade continued at more or less the same pace. Overall the share of the post-retirement workers in the total withdrawal from the labor force in the 1993–2000 years is small.

In the 59+ age group it is generally argued that income effect predominates when withdrawal from labor force is observed over time. Only in the case of urban females do we see a consistent decline in LFPR over the three rounds and also for the top three quintiles. The trends are much more mixed for the other categories, though for both rural and urban males the decline in participation in the 1990s seems to have been strongest in the top two quintiles.

It should be emphasized that this result, although suggestive, cannot be conclusive about the income effect on the participation of seniors. This is because the earnings of the seniors staying on in market activity itself affect the expenditure level of their household.

Female subsidiary labor force

Change in the demand for subsidiary labor

As argued earlier, there was a sudden upsurge of female subsidiary labor demand in the 1980s followed by a contraction of labor demand in the 1990s. This fluctuation requires further probe. Tables 4.4a to 4.4b present the distribution of subsidiary female employment by major industries and occupations.

As we can see subsidiary female employment is concentrated in a few agricultural and allied activities, i.e., growing of cereals and animal husbandry. They contributed 80 percent of female subsidiary employment in rural areas and even a sizeable part of such employment in the urban labor market.

We, therefore, looked specifically at the growing of cereals sector in rural areas since it constituted the largest chunk of female subsidiary labor supply. This sector contributed 4.9 million out of six million of additional subsidiary female employment between the 38th and 50th rounds. At the same time it contributed 5.8 million out of 6.3 million of decline in female subsidiary employment between the 50th and 55th rounds.

However, analysis across states for this sector show that four states Karnataka, Madhya Pradesh, Maharashtra and West Bengal² played the major role in absorption of female subsidiary labor in this sector in the 1980s and subsequent decline in the 1990s. These states include the districts which experienced a spread of the green revolution in the 1980s. Thus this piece of evidence gives credence to the argument that the spread of labor absorbing green revolution technologies in the 1980s bumped up the demand for labor in these areas, which in the short run could only be met by increasing use of female secondary labor. Subsequently in the 1990s as rising labor costs led to the introduction of labor replacing technologies, the additional demand for female labor subsided. This process might have been also helped by hurdles faced by oilseed development program in the 1990s in parts of this region.

It should be noted that most of this adjustment took place, not in the wage-labor market, but among the self-employed workers. It can be seen from Table 4.4a that agricultural wage labor accounted for a small portion of total female subsidiary employment in the rural areas, and even less so in the urban sector. Further examination of the NSS data by employment status showed that casual wage labor accounted for only 16.9 percent of all female subsidiary workers (in the age group 15 and above) in the rural areas in the 55th round–down from 21.7 percent in the 50th. The corresponding percentages in the urban sector were 12.1 and 20.6. Much the more substantial share of such employment in both sectors was accounted for by the 'self-employed' and the 'helpers' categories.

Evidence on the 'income effect'

If there is a significant income effect affecting participation then we would expect the opportunity cost of leisure to increase at all household welfare (or income) levels–but it would presumably increase more at higher levels of welfare. Consider a supply function of labor-relating participation rate (of, say, prime-age females) to the household welfare level. At any point of time, for a given distribution of household incomes, we would expect this curve to turn down quite sharply as the effect of higher household welfare begins to overshadow the substitution effect. When at a later date average income of all household increases, the supply curve relating participation of this group to the household welfare level is pushed downwards. Thus there is less participation at all welfare levels–but the point at which there is a significant fall in the slope of the curve comes earlier in the lower part of the household welfare distribution. We have seen that a very important portion of the change in participation in the post-liberalization period is accounted for by the fall in the number of females of subsidiary status. We can try to see which household welfare groups have typically contributed to the withdrawal of labor in the female subsidiary status. The index for household welfare used is the mean per capita expenditure level. We use the groupings as provided by the NSS reports.

Table 4.5 Distribution of subsidiary employment across APCE groups for ages 5+

It is seen that in the rural sector the female subsidiary workers in the 55th round area coming much more from a lower expenditure group than those in the 50th round. Notice in particular that the P2/P1 ratio, as defined in Table 4.5, has fallen from a value of 1.66 to 1.04 in the post-liberalization years. The value of this ratio for rural males has also fallen, but not by as much.³ The evidence strongly suggests that the withdrawal of subsidiary workers–which was identified as a dominant feature of the change in the rural labor markets over this period–came increasingly from higher household expenditure groups. Interestingly enough, the trend in the urban sector is the exact reverse. The P2/P1 ratio increased substantially both for males and females, suggesting that the withdrawal observed for subsidiary workers in the urban labor markets came increasingly from lower expenditure groups. We conclude that the 'income effect' seems to have been a factor in the fall in participation of subsidiary workers, particularly females, in the rural areas, but that other factors (e.g., education or social connection) might have been more important in the falling participation rate in the urban economy.

Principal-status labor-force participation

Let us now see how the supply of labor in the UPS category behaves in contrast to the supply in the UPSS status discussed in the last section. The change in the growth rates of UPS labor by gender and sector are shown in Table 4.6.

Unlike in the case of the UPSS labor force no decline in the growth rate is observed in the 1990s compared with the 1980s. However, there is an important gender difference. The male labor-force growth fell in in the 1990s, whereas female labor-force growth increased by 50 percent. The last row in Table 4.6 shows the hypothetical growth rate which would have occurred if the PRs had remained at the same levels as in the 50th round. The significant point to note is that in the rural labor market although the actual growth rate for female principal workers between the 50th and the 55th rounds was below the 'derived' growth rate for this group, it had nevertheless increased compared with the previous period between the 38th and the 50th rounds. Further scrutiny about changes in age-specific participation rates for principal females shows that this increase is really due to an increase in PRs for the prime age groups 25–59 (see Figure 4.2a).

We did a D–A analysis for rural females in the 55th round by broad age groups, showing the difference between the derived figure of the labor force (on the assumption of PRs being unchanged from the 50th round) and the actual labor force reported. A negative figure indicates that the PR for the relevant age group has increased. It was seen that the D–A statistic for the 25–59 age group was minus two million compared to the positive 0.5 million for the total rural female workers of UPS status. Thus while for the rural principal females as a whole there was a net marginal decline in PRs, the PRs for the prime age groups had increased to a significant degree, relative to the trends in all other age–sex groups in the rural labor market for UPS workers.

Considering that there had been a substantial fall in the PRs for the rural females in secondary status in this period (see Figure 4.4a), we conclude that there was some shift of employment from subsidiary to principal status during this period. It is quite consistent with rationalization of the labor force where principal workers are preferred compared with the subsidiary labor force when growth is sustained over a period of time.

It is clear, therefore, that the observed decline in the labor force–and the attendant fall in employment–has not affected the principal labor market. On the contrary the evidence suggests a tightening of labor conditions in this market. Since the reported wage rates refer to this labor market, we would expect this tightening to be reflected in an increase in real wages. This is indeed what we see across the board for different classes of principal workers.

Evidence on wage trends

We have already seen in Chapter 3 that the growth rate of the manual casual wage per day shows a slight acceleration in the rural sector as a whole, for both males and females. More detailed statistics are given in Table 4.7 for the manual as well as the non-manual parts of the casual labor market. The wage rate accelerated substantially in the second period for both sexes in non-agriculture in rural areas and in the urban areas.

While all groups seem to have experienced an upward trend, the acceleration in the second period was stronger for non-manual workers and in the urban areas. The evidence suggests that the increase in demand for labor in the non-manual labor market supplemented the relative increase in demand for female principal workers, which seems to have been caused by the rationalization of labor deployment in the manual agricultural labor market.

Conclusion on withdrawal of labor

The observed fall in the supply of labor in the 55th round due to decrease in PRs has three major components: the 5–19 age group in UPS status (30 percent of the total decrease); the older 59+ age group in UPS (11 percent); and the females of working age group in subsidiary status (47 percent). It is the last which has been the subject of extended discussion and alternative explanations. Our analysis based on a reading of the historical record suggests that the key to an understanding of this phenomenon is the upsurge in the demand for labor in the early eighties due to the second wave of the green revolution in paddy cultivation and also in oilseed cultivation. This increase in demand was met in the short run by a lift in the participation of SS females as shown in the data for the 38th and 50th rounds. As the economy adjusted to the new level of labor demand in agriculture labor deployment was gradually changed with more use of female labor of a more regular kind. Thus we get a shift from SS females to UPS females between the 50th and the 55th rounds. Since the supply of effort by UPS workers is at a substantially higher level than for the SS workers, this led an overall decrease in female number of workers over all. While this restructuring of the female labor demand is the basic cause of the observed fall in PRs of SS females, we can also discern an element of the income effect leading to a withdrawal of female secondary workers from higher income groups. As explained the wage rates in agriculture are determined in the market for the more regular (UPS) workers. It seems that the increase in demand for such workers continued to keep ahead of the increase in supply due to natural growth of the working-age population, so that the rate of increase in wages increased in the nineties.

Appendix 1 concept of different types of labor force

Usual Principal and Subsidiary Status (UPSS) include persons in the labor force by both major and minor time criteria. In other words, it includes both principal and subsidiary status categories of persons in the labor force.

UPS labor force refers to the persons those who are included in the labor force by major time criterion.

5.0.15 Usual activity status: The usual activity status relates to the activity status of a person during the reference period of 365 days preceding the date of survey. The activity status on which a person spent relatively longer time (major time criterion) during the 365 days preceding the date of survey is considered the principal usual activity status of the person. To decide the principal usual activity of a person, he/she is first categorized as belonging to the labor force or not, during the reference period on the basis of major time criterion…. For the persons belonging to the labor force, the broad activity status of either 'working' or 'not working but seeking and/or available for work' is then ascertained again on the basis of the relatively longer time spent in the labor force during the 365 days preceding the date of survey.

(Instruction manual, 55th round, schedule 10, section 5.0.15.)

In this study, the Subsidiary Status (SS) labor force is defined as persons who are pursuing non-economic activities (out of labor force) by major time criterion (UPS) but belong to the labor force by minor time criterion. It excludes persons who are included in labor force by UPS to avoid double counting. Since unemployment status is determined by major time criterion, those belonging to labor force only on the basis of subsidiary status by default are all workers.

5.0.16 Subsidiary economic activity status: A person whose principal usual status is determined on the basis of the major time criterion may have pursued some economic activity for a relatively shorter time (minor time) during the reference period of 365 days preceding the date of survey …. It may be noted that engagement in work in subsidiary capacity may arise out of the two following situations:

1. a person may be engaged for a relatively longer period during the last 365 days in economic/non-economic activity and for a relatively shorter period in another economic activity and
   
   
2. a person may be pursuing one economic activity/non-economic activity almost through-out the year in the principal usual activity status and also simultaneously pursuing another economic activity for a relatively shorter period in a subsidiary capacity.

(Instruction manual, 55th round, schedule 10, section 5.0.16.)

In our concept of subsidiary status labor force we have only included the persons who are engaged in non-economic activity for a 'relatively longer period during the last 365 days' but engaged in economic activity for a 'relatively shorter period' or 'pursued non-economic activity almost throughout the year in principal usual activity status' but pursued another economic activity for 'relatively shorter period in a subsidiary capacity'. In this fashion we managed to get UPSS labor force = UPS labor force + SS labor force which is additive.

Appendix 2: estimating the absolute number of the labor force

NSS rounds are sample surveys. They do calculate India's population but these are generally underestimated and the level of underestimation is going up over the year. We can get corrected population estimates for these three mid-year dates by interpolating population figures from three decadal population census of India–1981, 1991 and 2001.

NSS differs from decadal census in terms of age-group distribution population. To adjust for the 38th, 50th and 55th rounds of NSS we have used the five-year age-group distribution of the 1981 Census, 1991 Census and National Health and Family Welfare Survey-II (NFHS) 1998 respectively. The 13 age groups that we have considered are 0–5, 5–10, 10–15, 15–20, 20–25, 25–30, 30–35, 35–40, 40–45, 45–50,50–55, 55–60 and 60 and above. NSS also differs from decadal census in terms of rate of urbanization rate and sex ratio. To get around this problem we used the Census-adjusted NSS mid-year population and age-group distribution separately for rural male (RM), rural female (RF), urban male (UM) and urban female (UF). Thus, we calculated population for these four sections of population for 13 age groups separately.

Labor-force participation rate (LFPR) for each of these four sections of population for all three rounds have been generated for all 12 age groups (for 0–5 age group LFPR is not calculated) from unit-level data. By multiplying LFPR for each of them with the respective population cohort gives us the labor force for each of this population cohort. By adding up the labor force of all age groups and dividing it by its respective population we could derive the LFPR at more aggregate levels. Our calculated LFPR at aggregate level marginally differs from published LFPR figures of NSS.

A similar procedure has been adopted for calculating the number of workers in Chapter 3 and elsewhere.

Document(s) 4 de 16