Special articles
Recent Trends in Poverty and Inequality: Some Preliminary Results
Preliminary estimates from the published reports of the 61st round of the National Sample Survey suggest that while poverty did reduce during 1993-2005, the annual rate of reduction in this period was lower than in the 1970s and 1980s. More importantly, the bulk of this decline occurred in 1999-2005, with little or no reduction in poverty in 1993-2000, confirming the earlier consensus that the 1990s were indeed the lost decade for poverty reduction. Although the analysis is not conclusive, the fall in the relative price of food and the regional pattern of changes in employment and wages appear to underlie these trends. These results need to be explored in greater detail as and when the unit level data for 2004-05 become available. The paper also flags certain issues related to the poverty line which need to be settled once and for all.
HIMANSHU
S
poverty reduction during 1999-2005, so much so that this accounts for the overwhelming part of the total reduction during 1993-2005 and implies very little poverty reduction during 1993-2000. While this vindicates the conclusion reached earlier by Sen-Himanshu (2004) that the 1990s were largely a lost decade for poverty reduction, the surprise is that poverty appears to have reduced quite significantly after the late 1990s despite very slow agricultural growth and the now officially acknowledged increase in rural distress. This throws open many issues which can only be investigated fully after the 61st round unit level data are made public.
The paper is structured as follows. In the first section we report the trends in poverty and inequality based on comparable uniform reference period for major states and all India. This is done for the 38th, 43rd, 50th and 61st rounds. The issue of comparability of 55th round results and some tentative calculation based on EUSs are discussed in Section II. In Section III additional evidence on well-being is brought in to supplement the discussion on poverty and inequality in the recent period. Section IV puts these together in the broader context of poverty decline and its associated factors. The conclusion flags some of the important issues that need clarification in view of the trends emerging from a preliminary reading of the results of 61st round.
Poverty and Inequality in 2004-05, as Compared to 1993-94 and Earlier
The 61st round of canvassed information on consumption expenditure based on a 30-day uniform recall period for all items and also by the 365-day recall period for five less frequently consumed items – education, institutional medical care, durables, clothing and footwear. The recall periods used are identical to those used in the 50th round. However, the 61st round consumption expenditure survey schedules are not comparable to the 55th round for two reasons. First, no information was collected for food items based on seven-day recall as was done in the 55th in 1993-94 to 27 per cent in 1999-2000, the 61st round places round, and, second, unlike the 55th round which asked only this at nearly 29 per cent in 2004-05 – i e, an 8 percentage point 365-day questions for the five infrequently purchased items, the decline over 11 years rather than the 10 percentage point decline 61st round asked both the 30- and 365-day questions. over six years as claimed earlier. Similarly, urban poverty is now
Since results based on a uniform reference period from the 61st placed at nearly 26 per cent in 2004-05 against 33 per cent in round are fully comparable to the 50th and previous thick rounds 1993-94, a 7 percentage point decline over 11 years compared based on uniform reference period (URP), this is the first com-to the 9 percentage point decline over 6 years claimed earlier parison made here. Tables 1 and 2 present the poverty headcount from official 55th round data. ratio, poverty gap (PG), squared poverty gap (SPG) and Gini Moreover, matters are somewhat worse when we move to more ratios (all in percentages) for rural and urban areas separately sensitive measures of poverty such as the poverty gap and squared for the major states.2 It is clear from Tables 1 and 2 that poverty poverty gap which are indicators of severity and depth of poverty, has declined between 1993-94 and 2004-05 in rural areas of all respectively. On these two measures, Chhattisgarh and Orissa states as well as in most urban areas except Orissa and Uttaranchal. do worse in rural areas in 2004-05 than in 1993-94. In both these Nonetheless, the main conclusion must be that the 61st round states, poverty gap and squared poverty gap increased in results confirm that actual poverty reduction after 1993-94 has 2004-05 compared to 1993-94. Similarly in urban areas, PG and been much less than the earlier official assessment using the non-SPG measures have increased in Orissa, Chhattisgarh and Haryana. comparable 55th round. As against the earlier official claim For Chhattisgarh and Orissa, the more sensitive measures of that the rural poverty ratio had declined from 37 per cent poverty have worsened in both rural and urban areas.
Table 1: Comparable Estimates of Poverty and InequalityTable 2: Comparable Estimates of Poverty and Inequality(URP, Official Poverty Lines) (URP, Official Poverty Lines)
(Per cent) (Per cent)
Headcount Ratio Poverty Gap Headcount Ratio Poverty Gap Rural 1983 1987-1993-2004-1983 1987-1993-2004-Urban 1983 1987-1993-2004-1983 1987-1993-200488 9405 889405 88 9405 889405
Andhra Pradesh 26.8 21.0 15.9 10.8 5.86 4.35 2.9 2.0 Andhra Pradesh 41.2 41.1 38.8 27.1 10.9 10.6 9.3 6.1 Assam 44.6 39.4 45.2 21.7 8.75 7.45 8.3 3.5 Assam 25.9 11.3 7.9 3.7 5.6 1.5 0.9 0.5 Jharkhand 65.5 52.8 62.3 42.9 22.00 13.56 16.2 8.9 Jharkhand 40.5 34.6 26.5 20.7 10.9 7.8 5.2 4.7 Bihar 64.7 54.2 56.6 42.2 19.54 12.74 14.2 8.3 Bihar 61.6 63.8 40.7 38.1 18.5 16.6 9.7 9.3 Gujarat 28.9 28.3 22.2 19.4 5.64 5.44 4.1 3.4 Gujarat 41.9 38.5 28.3 14.2 9.7 8.2 6.2 2.5 Haryana 21.9 15.3 28.3 13.6 4.28 3.62 5.6 2.2 Haryana 26.4 18.4 16.5 15.6 5.8 3.6 3.0 3.2 Himachal Pradesh 17.0 16.7 30.4 10.9 3.58 2.63 5.6 1.5 Himachal Pradesh 11.0 7.2 9.3 5.0 2.8 0.7 1.2 1.0 Karnataka 36.3 32.6 30.1 20.0 9.73 7.88 6.3 2.7 Karnataka 43.6 49.2 39.9 33.3 13.3 14.1 11.4 8.9 Kerala 39.6 29.3 25.4 13.2 9.98 6.30 5.6 2.8 Kerala 48.0 38.7 24.3 20.6 14.7 10.0 5.5 4.7 Chhattisgarh 50.6 46.7 44.4 42.0 12.49 10.38 8.6 9.4 Chhattisgarh 50.7 36.0 44.2 40.7 14.5 9.8 11.5 12.9 Madhya Pradesh 49.0 40.1 39.2 35.8 13.95 10.64 9.8 7.8 Madhya Pradesh 56.1 50.0 49.0 42.3 16.1 14.5 13.9 12.4 Maharashtra 45.9 40.9 37.9 30.0 11.95 9.56 9.3 6.4 Maharashtra 41.1 40.5 35.0 32.8 12.1 12.4 10.2 9.2 Orissa 68.5 58.7 49.8 46.9 22.72 16.30 12.0 12.1 Orissa 54.0 42.6 40.6 43.7 16.7 11.1 11.4 14.1 Punjab 14.3 12.8 11.7 10.0 3.03 1.97 1.9 1.3 Punjab 22.9 13.7 10.9 5.0 5.9 2.3 1.7 0.6 Rajasthan 35.0 33.3 26.4 19.0 9.65 8.64 5.2 2.9 Rajasthan 41.2 37.9 31.0 28.5 11.5 9.6 7.0 6.2 Tamil Nadu 54.8 46.3 32.9 22.7 17.39 12.65 7.3 3.7 Tamil Nadu 51.9 40.2 39.9 24.1 15.4 11.5 10.2 5.3 Uttaranchal 25.2 13.2 24.8 14.9 4.00 1.99 4.4 1.9 Uttaranchal 22.4 20.4 12.7 17.0 5.9 4.2 3.2 3.0 Uttar Pradesh 47.8 43.3 43.1 33.9 12.70 10.25 10.6 6.7 Uttar Pradesh 52.7 46.4 36.1 30.7 15.1 12.7 9.3 7.2 West Bengal 63.6 48.8 41.2 28.5 21.06 11.58 8.3 5.4 West Bengal 33.5 33.7 22.9 15.4 8.5 7.4 4.5 2.6 All India 46.5 39.0 37.2 28.7 12.36 9.29 8.5 5.8 All India 43.6 38.7 32.6 25.9 11.4 10.2 8.0 6.2
Squared Poverty Gap Gini Squared Poverty Gap Gini Rural 1983 1987-1993-2004-1983 1987-1993-2004-Urban 1983 1987-1993-2004-1983 1987-1993-20048894 05 889405 88 9405 889405
Andhra Pradesh 2.00 1.41 0.87 0.65 29.7 30.9 29.0 29.4 Andhra Pradesh 4.1 3.9 3.2 1.9 33.2 36.1 32.3 37.6 Assam 2.63 2.04 2.21 0.90 20.0 23.0 17.9 19.9 Assam 1.7 0.3 0.2 0.1 26.1 31.0 29.0 32.1 Jharkhand 9.8 5.03 5.59 2.55 27.2 26.6 23.4 22.7 Jharkhand 4.2 2.6 1.6 1.5 30.9 32.1 32.5 35.5 Bihar 7.86 4.32 4.9 2.30 25.9 25.2 22.2 20.7 Bihar 7.1 5.9 3.4 3.0 28.5 26.6 28.2 33.3 Gujarat 1.69 1.59 1.16 0.91 26.8 26.1 24.0 27.3 Gujarat 3.6 2.6 2.0 0.7 28.5 27.8 29.1 31.0 Haryana 1.37 1.30 1.75 0.61 28.5 29.2 31.4 34.0 Haryana 1.9 1.1 0.9 1.0 34.8 28.7 28.4 36.5 Himachal Pradesh 1.16 0.71 1.62 0.35 27.1 28.4 31.1 Himachal Pradesh 1.1 0.1 0.3 0.3 35.8 29.2 46.2 32.6 Karnataka 3.69 2.80 2.01 0.63 30.8 29.7 27.0 26.5 Karnataka 5.5 5.7 4.4 3.1 34.2 34.0 31.9 36.8 Kerala 3.62 2.05 1.85 0.98 32.0 32.1 30.1 38.3 Kerala 6.2 3.9 1.9 1.6 38.9 36.9 34.3 41.0 Chhattisgarh 4.47 3.36 2.47 3.43 24.4 24.5 21.7 29.8 Chhattisgarh 5.6 3.6 4.1 5.4 32.2 32.1 30.6 44.0 Madhya Pradesh 5.54 3.97 3.58 2.31 31.5 30.6 30.0 26.8 Madhya Pradesh 6.2 5.6 5.3 4.8 29.8 33.3 33.6 39.7 Maharashtra 4.3 3.21 3.35 1.99 29.1 31.2 30.7 31.2 Maharashtra 4.9 5.2 4.2 3.5 34.6 34.8 35.7 37.8 Orissa 10.17 6.24 4.07 4.24 27.0 26.9 24.6 28.5 Orissa 7.1 4.2 4.3 5.8 29.0 31.0 30.7 35.4 Punjab 1.06 0.51 0.48 0.26 29.2 29.7 28.1 29.5 Punjab 2.3 0.6 0.4 0.1 33.9 28.8 28.1 40.3 Rajasthan 3.81 3.40 1.56 0.72 34.7 31.5 26.5 25.1 Rajasthan 4.7 3.4 2.2 1.9 33.9 34.6 29.3 37.2 Tamil Nadu 7.52 4.80 2.50 0.96 36.7 33.0 31.2 32.2 Tamil Nadu 6.3 4.6 3.9 1.6 35.1 35.8 34.8 36.1 Uttaranchal 1.04 0.46 1.08 0.42 29.2 28.3 24.4 28.5 Uttaranchal 2.0 1.2 0.9 0.7 30.5 35.1 27.5 32.9 Uttar Pradesh 4.7 3.4 3.64 1.93 28.9 28.5 28.3 29.0 Uttar Pradesh 5.9 4.7 3.4 2.3 31.5 33.5 32.6 36.9 West Bengal 9.46 3.99 2.45 1.42 30.0 25.8 25.4 27.4 West Bengal 3.2 2.4 1.4 0.6 33.5 34.6 33.9 38.3 All India 4.87 3.23 2.84 1.76 30.4 29.9 28.6 30.5 All India 4.4 3.8 2.9 2.0 33.9 35.0 34.4 37.6
Source: 2004-05 estimates are calculated from grouped data from NSSO Source: 2004-05 estimates are calculated from grouped data from NSSO
Report 508. Estimates for 1983, 1987-88 and 1993-94 are calculated Report 508. Estimates for 1983, 1987-88 and 1993-94 are calculated
from the unit level data respectively. from the unit level data respectively.
As far as inequality measured by Gini ratios is concerned, this has worsened between 1993-94 and 2004-05 in rural areas of all states except Bihar, Jharkhand, Karnataka, Madhya Pradesh and Rajasthan. This reverses the trend seen in the previous decade (1983 to1993-94) when inequalities came down in all the states except for Haryana and Maharashtra. In urban areas, inequality increased between 1993-94 and 2004-05 in all states as well as at the all-India level. Unlike all-India rural, where the Gini ratio in 2004-05 was about the same as in 1983, the urban all-India Gini was significantly higher in 2004-05 than in all previous NSS thick sample rounds.
Further, although poverty did reduce between 1993-94 and 2004-05, the pace of poverty reduction was lower than in the previous two decades. Tables 3 and 4 give the annual change in poverty and inequality measures for 1983 to1987-88, 1983 to1993-94, 1993-94 to 2004-05, 1987-88 to 2004-05 and 1983 to 2004-05 for rural and urban areas. These show a setback in poverty reduction after the 1980s. The pace of reduction of the all-India rural headcount ratio, which had averaged 1.1 percentage points per annum (ppa) during 1973 to 19833 and was 1.6 ppa in 1983-88, fell sharply to only 0.3 ppa during 1987-93 before recovering to 0.7 ppa during 1993-2005. Corresponding figures for all-India urban are 0.7 ppa during 1973-83, 1.1 ppa during 1983-88, 1.0 ppa during 1987-93 and only 0.6 ppa during 1993-2005. Taking both rural and urban together, the pace of poverty reduction fell from 1.1 percentage point per annum during 1973-88 to only 0.6 percentage point per annum during 1987-2005 and 0.7 percentage point per annum during 1993-2005. The 61st round should therefore set at rest all claims that economic reforms have led to faster poverty reduction.4
The annual rate of reduction in headcount poverty during 1993-2005 was lower in rural areas of 11 states and urban areas of 15 (out of 19) compared to 1983-1994; and the position is worse for the more sensitive indicators or on comparisons of 1983-88 with 1987-05. However, it is noticeable that poverty reduction was more during 1993-2005 than in 1983-1994 in rural areas of Assam, Bihar, Jharkhand and Uttar Pradesh, which are all states with higher than average poverty and where post-1993 GDP growth was less than the all-India average. Of course, all these poor states show worse urban poverty reduction performance during 1993-2005 than during 1983-94, but it remains the case that spatial pattern of poverty reduction does not fit easily with what is usually assumed regarding spatial patterns of GDP growth.5 Even otherwise, more states have seen a downturn in urban rather than rural performance in the post-reform period compared to the pre-reform period. Reasons for this disjunction
Table 3: Annual Percentage Change in Poverty and Inequality (URP, Official Poverty Line)
(Per cent)
| Headcount Ratio | Poverty Gap | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Rural | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 |
| Andhra Pradesh | -1.29 | -1.04 | -0.47 | -0.60 | -0.75 | -0.34 | -0.28 | -0.08 | -0.14 | -0.18 |
| Assam | -1.14 | 0.06 | -2.13 | -1.04 | -1.06 | -0.29 | -0.05 | -0.44 | -0.23 | -0.25 |
| Jharkhand | -2.82 | -0.31 | -1.76 | -0.58 | -1.05 | -1.88 | -0.56 | -0.66 | -0.28 | -0.61 |
| Bihar | -2.33 | -0.77 | -1.30 | -0.70 | -1.04 | -1.51 | -0.51 | -0.53 | -0.26 | -0.52 |
| Gujarat | -0.12 | -0.64 | -0.25 | -0.53 | -0.44 | -0.05 | -0.15 | -0.06 | -0.12 | -0.11 |
| Haryana | -1.46 | 0.60 | -1.33 | -0.10 | -0.39 | -0.15 | 0.13 | -0.31 | -0.08 | -0.10 |
| Himachal Pradesh | -0.07 | 1.27 | -1.77 | -0.34 | -0.28 | -0.21 | 0.19 | -0.37 | -0.07 | -0.10 |
| Karnataka | -0.81 | -0.59 | -0.92 | -0.75 | -0.76 | -0.41 | -0.33 | -0.32 | -0.30 | -0.32 |
| Kerala | -2.31 | -1.36 | -1.11 | -0.94 | -1.23 | -0.82 | -0.42 | -0.26 | -0.21 | -0.34 |
| Chhattisgarh Madhya Pradesh | -0.86 -1.98 | -0.58 -0.94 | -0.22 -0.31 | -0.28 -0.26 | -0.40 -0.62 | -0.47 -0.74 | -0.37 -0.39 | 0.07 -0.19 | -0.06 -0.17 | -0.14 -0.29 |
| Maharashtra | -1.10 | -0.76 | -0.72 | -0.64 | -0.74 | -0.53 | -0.25 | -0.27 | -0.19 | -0.26 |
| Orissa | -2.18 | -1.78 | -0.27 | -0.69 | -1.00 | -1.43 | -1.02 | 0.01 | -0.25 | -0.49 |
| Punjab | -0.32 | -0.24 | -0.16 | -0.17 | -0.20 | -0.23 | -0.11 | -0.06 | -0.04 | -0.08 |
| Rajasthan | -0.39 | -0.82 | -0.68 | -0.84 | -0.75 | -0.22 | -0.42 | -0.21 | -0.34 | -0.32 |
| Tamil Nadu | -1.88 | -2.08 | -0.93 | -1.39 | -1.49 | -1.05 | -0.96 | -0.33 | -0.53 | -0.64 |
| Uttaranchal | -2.67 | -0.03 | -0.90 | 0.10 | -0.48 | -0.45 | 0.03 | -0.22 | 0.00 | -0.10 |
| Uttar Pradesh | -1.01 | -0.45 | -0.84 | -0.55 | -0.65 | -0.54 | -0.20 | -0.36 | -0.21 | -0.28 |
| West Bengal All India | -3.30 -1.66 | -2.14 -0.88 | -1.15 -0.77 | -1.19 -0.61 | -1.63 -0.83 | -2.11 -0.68 | -1.22 -0.37 | -0.27 -0.25 | -0.37 -0.21 | -0.73 -0.31 |
| Squared Poverty Gap | Gini | |||||||||
| Rural | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 |
| Andhra Pradesh | -0.13 | -0.11 | -0.02 | -0.04 | -0.06 | 0.27 | -0.07 | 0.04 | -0.09 | -0.01 |
| Assam | -0.13 | -0.04 | -0.12 | -0.07 | -0.08 | 0.65 | -0.20 | 0.19 | -0.18 | 0.00 |
| Jharkhand | -1.06 | -0.40 | -0.28 | -0.15 | -0.34 | -0.14 | -0.37 | -0.07 | -0.23 | -0.21 |
| Bihar | -0.86 | -0.31 | -0.46 | -0.26 | -0.39 | -0.16 | -0.36 | -0.13 | -0.27 | -0.24 |
| Gujarat | -0.02 | -0.05 | -0.02 | -0.04 | -0.04 | -0.14 | -0.27 | 0.30 | 0.07 | 0.02 |
| Haryana | -0.02 | 0.04 | -0.10 | -0.04 | -0.04 | 0.16 | 0.28 | 0.24 | 0.29 | 0.26 |
| Himachal Pradesh | -0.10 | 0.04 | -0.12 | -0.02 | -0.04 | 6.02 | 2.71 | 0.24 | 0.24 | 1.45 |
| Karnataka | -0.20 | -0.16 | -0.13 | -0.13 | -0.14 | -0.26 | -0.37 | -0.04 | -0.18 | -0.20 |
| Kerala | -0.35 | -0.17 | -0.08 | -0.06 | -0.12 | 0.03 | -0.19 | 0.75 | 0.36 | 0.29 |
| Chhattisgarh Madhya Pradesh | -0.25 -0.35 | -0.19 -0.19 | 0.09 -0.12 | 0.00 -0.10 | -0.05 -0.15 | 0.03 -0.20 | -0.25 -0.14 | 0.74 -0.29 | 0.31 -0.22 | 0.26 -0.22 |
| Maharashtra | -0.24 | -0.09 | -0.12 | -0.07 | -0.11 | 0.45 | 0.15 | 0.05 | 0.00 | 0.10 |
| Orissa | -0.87 | -0.58 | 0.02 | -0.12 | -0.28 | -0.03 | -0.23 | 0.36 | 0.10 | 0.07 |
| Punjab | -0.12 | -0.06 | -0.02 | -0.01 | -0.04 | 0.11 | -0.10 | 0.12 | -0.01 | 0.01 |
| Rajasthan | -0.09 | -0.21 | -0.08 | -0.16 | -0.14 | -0.71 | -0.78 | -0.13 | -0.38 | -0.45 |
| Tamil Nadu | -0.61 | -0.48 | -0.14 | -0.23 | -0.30 | -0.83 | -0.52 | 0.09 | -0.04 | -0.21 |
| Uttaranchal | -0.13 | 0.00 | -0.06 | 0.00 | -0.03 | -0.18 | -0.45 | 0.37 | 0.01 | -0.03 |
| Uttar Pradesh | -0.29 | -0.10 | -0.16 | -0.09 | -0.13 | -0.08 | -0.06 | 0.07 | 0.03 | 0.01 |
| West Bengal All India | -1.22 -0.36 | -0.67 -0.19 | -0.09 -0.10 | -0.15 -0.09 | -0.37 -0.14 | -0.94 -0.11 | -0.44 -0.17 | 0.17 0.17 | 0.09 0.03 | -0.12 0.00 |
| Source: Calculated from Table 1. | ||||||||||
| Economic and Political Weekly | February 10, 2007 | 499 |
will require further investigation, as also the extent to which urbanisation of poverty may have been driven by migration of rural poor to urban areas. However, it is clear that the regional divergence stories on growth and poverty reduction that had emerged after the 55th round are now shown to be too simplistic.6 In the longer time frame of 1983 to 2004-05, there are some clear trends visible at the state level. The states which have done well consistently by this comparison in the last two decades are West Bengal, Tamil Nadu and Kerala in rural areas and Tamil Nadu, Kerala and Gujarat in urban areas. But even in these states, poverty reduced at a slower pace post-reforms. At the other end, longer period poverty reduction has been low in some rich states: Punjab, Haryana and Gujarat for rural and Maharashtra, Karnataka and Haryana for urban. Here too, the performance has not improved post-reforms.
Comparison with 55th Round
The previous section compared 61st round URP results with comparable results from rounds 50 and earlier, ignoring the 55th round (1999-2000) altogether. This is because there are no comparable URP estimates available from the 55th round. However, some comparison across these rounds is necessary since there is generally greater interest on what has happened more recently and also because a handle on this would clarify on the intensely debated matter of the extent of poverty reduction between the 50th and 55th rounds.
One problem in this context is that NSS Report 508 does not give the distribution of consumer expenditure by the mixed reference period (MRP), which is the first step necessary for comparisons with the 55th round. Nonetheless, since the report does give the MPCE by MRP against each size class of the distribution on URP basis, Mahendra Dev and Ravi (2007) have estimated MRP poverty counts from Tables 6R and 6U of this report assuming that the ranking of individuals by their MPCE on MRP basis is the same as their ranking by MPCE on URP basis. This assumes that MRP MPCE is strictly monotonic with URP MPCE.
This is obviously an approximation, since data from 43rd and 50th rounds show that although the MPCEs by URP and MRP are strongly correlated across households, the relationship is not
Table 4: Annual Percentage Change in Poverty and Inequality (URP, Official Poverty Line)
(Per cent)
| Headcount Ratio | Poverty Gap | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Urban | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 |
| Andhra Pradesh | -0.03 | -0.23 | -1.07 | -0.82 | -0.66 | -0.07 | -0.16 | -0.29 | -0.27 | -0.23 |
| Assam | -3.25 | -1.72 | -0.39 | -0.45 | -1.04 | -0.90 | -0.44 | -0.04 | -0.06 | -0.24 |
| Jharkhand | -1.29 | -1.33 | -0.53 | -0.82 | -0.92 | -0.69 | -0.54 | -0.05 | -0.18 | -0.29 |
| Bihar | 0.50 | -1.99 | -0.24 | -1.52 | -1.09 | -0.42 | -0.83 | -0.03 | -0.43 | -0.42 |
| Gujarat | -0.75 | -1.30 | -1.28 | -1.43 | -1.29 | -0.34 | -0.34 | -0.34 | -0.34 | -0.34 |
| Haryana | -1.76 | -0.94 | -0.08 | -0.17 | -0.50 | -0.49 | -0.26 | 0.01 | -0.02 | -0.12 |
| Himachal Pradesh | -0.84 | -0.17 | -0.38 | -0.13 | -0.28 | -0.45 | -0.14 | -0.03 | 0.01 | -0.08 |
| Karnataka | 1.24 | -0.35 | -0.60 | -0.94 | -0.48 | 0.17 | -0.19 | -0.22 | -0.30 | -0.20 |
| Kerala | -2.08 | -2.26 | -0.34 | -1.06 | -1.28 | -1.04 | -0.87 | -0.07 | -0.31 | -0.46 |
| Chhattisgarh | -3.27 | -0.61 | -0.32 | 0.28 | -0.47 | -1.05 | -0.28 | 0.13 | 0.19 | -0.07 |
| Madhya Pradesh | -1.36 | -0.67 | -0.61 | -0.45 | -0.64 | -0.35 | -0.21 | -0.14 | -0.12 | -0.17 |
| Maharashtra | -0.12 | -0.58 | -0.20 | -0.45 | -0.38 | 0.07 | -0.18 | -0.09 | -0.19 | -0.13 |
| Orissa | -2.54 | -1.27 | 0.28 | 0.07 | -0.48 | -1.25 | -0.51 | 0.25 | 0.18 | -0.12 |
| Punjab | -2.04 | -1.14 | -0.54 | -0.51 | -0.83 | -0.80 | -0.40 | -0.10 | -0.10 | -0.25 |
| Rajasthan | -0.74 | -0.97 | -0.23 | -0.55 | -0.59 | -0.42 | -0.43 | -0.08 | -0.20 | -0.25 |
| Tamil Nadu | -2.59 | -1.14 | -1.43 | -0.95 | -1.29 | -0.88 | -0.50 | -0.44 | -0.36 | -0.47 |
| Uttaranchal | -0.46 | -0.93 | 0.39 | -0.20 | -0.25 | -0.36 | -0.25 | -0.03 | -0.07 | -0.13 |
| Uttar Pradesh | -1.41 | -1.59 | -0.49 | -0.93 | -1.03 | -0.52 | -0.55 | -0.19 | -0.32 | -0.36 |
| West Bengal | 0.04 | -1.01 | -0.69 | -1.08 | -0.84 | -0.23 | -0.37 | -0.18 | -0.29 | -0.27 |
| All India | -1.09 | -1.05 | -0.61 | -0.75 | -0.82 | -0.26 | -0.32 | -0.17 | -0.24 | -0.24 |
| Squared Poverty Gap | Gini | |||||||||
| Urban | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 | 1983-88 | 1983-94 | 1993-05 | 1987-05 | 1983-05 |
| Andhra Pradesh | -0.05 | -0.08 | -0.12 | -0.12 | -0.10 | 0.63 | -0.09 | 0.48 | 0.09 | 0.20 |
| Assam | -0.30 | -0.14 | -0.01 | -0.01 | -0.07 | 1.10 | 0.28 | 0.28 | 0.06 | 0.28 |
| Jharkhand | -0.35 | -0.25 | -0.01 | -0.07 | -0.13 | 0.26 | 0.15 | 0.27 | 0.20 | 0.21 |
| Bihar | -0.28 | -0.36 | -0.03 | -0.17 | -0.19 | -0.41 | -0.03 | 0.46 | 0.39 | 0.22 |
| Gujarat | -0.23 | -0.15 | -0.12 | -0.11 | -0.14 | -0.15 | 0.06 | 0.17 | 0.19 | 0.12 |
| Haryana | -0.17 | -0.09 | 0.01 | -0.01 | -0.04 | -1.34 | -0.61 | 0.74 | 0.46 | 0.08 |
| Himachal Pradesh | -0.21 | -0.08 | 0.01 | 0.01 | -0.03 | -1.47 | 0.99 | -1.24 | 0.20 | -0.15 |
| Karnataka | 0.03 | -0.11 | -0.11 | -0.15 | -0.11 | -0.05 | -0.22 | 0.45 | 0.17 | 0.12 |
| Kerala | -0.51 | -0.41 | -0.03 | -0.13 | -0.21 | -0.45 | -0.43 | 0.60 | 0.24 | 0.10 |
| Chhattisgarh | -0.44 | -0.15 | 0.12 | 0.10 | -0.01 | -0.02 | -0.15 | 1.22 | 0.70 | 0.55 |
| Madhya Pradesh | -0.13 | -0.09 | -0.05 | -0.05 | -0.07 | 0.77 | 0.37 | 0.55 | 0.38 | 0.46 |
| Maharashtra | 0.07 | -0.07 | -0.06 | -0.10 | -0.07 | 0.04 | 0.11 | 0.18 | 0.18 | 0.15 |
| Orissa | -0.65 | -0.27 | 0.14 | 0.10 | -0.06 | 0.45 | 0.17 | 0.42 | 0.26 | 0.30 |
| Punjab | -0.38 | -0.18 | -0.03 | -0.03 | -0.10 | -1.14 | -0.55 | 1.11 | 0.68 | 0.30 |
| Rajasthan | -0.27 | -0.23 | -0.03 | -0.09 | -0.13 | 0.16 | -0.43 | 0.71 | 0.15 | 0.15 |
| Tamil Nadu | -0.38 | -0.23 | -0.21 | -0.17 | -0.22 | 0.16 | -0.03 | 0.11 | 0.01 | 0.04 |
| Uttaranchal | -0.18 | -0.11 | -0.02 | -0.03 | -0.06 | 1.01 | -0.28 | 0.49 | -0.13 | 0.11 |
| Uttar Pradesh | -0.27 | -0.24 | -0.10 | -0.14 | -0.17 | 0.44 | 0.10 | 0.39 | 0.20 | 0.25 |
| West Bengal | -0.17 | -0.17 | -0.08 | -0.11 | -0.12 | 0.24 | 0.04 | 0.41 | 0.22 | 0.23 |
| All India | -0.14 | -0.14 | -0.08 | -0.11 | -0.11 | 0.24 | 0.05 | 0.29 | 0.15 | 0.17 |
| Source: Calculated from Table 2. | ||||||||||
| 500 | Economic and Political Weekly | February 10, 2007 |
strictly monotonic. Many households in the bottom 40 per cent of population by MRP show much lower (even zero) consumption of low frequency items by URP while some others show very high relative consumption of these items. As a result, the distribution on URP basis is more skewed than on MRP basis. Nonetheless, our analysis suggests that although this could run into trouble on inequality measures and the more sensitive poverty indicators, this is likely to be approximately correct on the headcount ratio. We have, however, not repeated this method for two reasons. First, because the Dev-Ravi approximation will in any case have to be revised when NSSO releases the MRP distribution from the 61st round Consumer Expenditure Survey (CES), and till then their approximations suffice. Second and more importantly, even when the NSSO releases the MRP distribution from the 61st round CES, poverty counts from this will not be comparable to the 55th round. This is because contamination from the seven-day food questions in the 55th round CES is likely to have biased downward its estimates of poverty so that comparisons of MRP poverty estimates from the 55th and 61st round CES are in any case likely to underestimate the poverty decline between the 55th and 61st rounds.
We have therefore adopted a more direct approach to gauge poverty change between 1999-2000 and 2004-05 by taking recourse to the EUS of the 55th and 61st rounds. As in the 55th round, the 61st round EUS had an abridged schedule for consumption expenditure. The size distribution of population based on this was released in the 61st round report no 515 on employment and unemployment. Since recall periods used were the same in both the rounds, and questionnaires almost identical (although slightly more disaggregated in the 61st than in the 55th round), poverty numbers from the EUS of 55th round and 61st round are likely to be much more comparable than comparisons of 55th and 61st round data from the respective consumer expenditure surveys. It should be noted that since EUS consumption schedules use MRP but are abridged, these are expected to report higher poverty than MRP estimates from the CES. In the 55th round, the MRP count from EUS exceeded corresponding CES counts by 7 and 5.3 percentage points in rural and urban all-India. But since part of this was due to contamination from 7-day food questions in the 55th round CES, the corresponding differences are expected to be smaller in the 61st round. Poverty measures from the EUS are given in Table 5 for rural and urban areas separately.7
The poverty reduction obtained from EUS 55th and 61st rounds are 9.1 per cent in rural areas and 3.9 per cent in urban areas at the all-India level. These estimates of poverty reduction between 1999-2000 and 2004-05 are surprisingly high compared to the poverty reduction by the URP CES estimates between 199394 and 2004-05 reported in the previous section, especially in Assam, Bihar, Madhya Pradesh, Orissa and West Bengal.8 All-India rural poverty reduction from EUS during 1999-2005 turns out higher than the reduction by CES URP during 1993-2005, and even the lower urban poverty reduction from EUS is about 75 per cent of urban poverty reduction found from CES URP for the longer period. Since poverty estimates on URP basis for 61st round are comparable with 50th round poverty estimates, poverty reduction implied by these URP estimates between these rounds is free from any controversy and fully comparable. Along with these, if EUS estimates from 55th and 61st rounds are considered fully comparable, the implication is of very little poverty reduction during 1993-2000 followed by large reduction during 1999-2005.9
This is a surprising result in view of the fact that the latter period was one of lower GDP growth, especially in agriculture. However, it should be noted that Dev-Ravi also report higher poverty reduction rates during 1999-2005 (1.13 and 0.73 percentage points per annum in rural and urban India) than during 1993-2000 (0.68 and 0.70 respectively), and their state data also show this improvement to be largely limited to states with higher poverty. Further, since they do not adjust for the seven-day issue, actual reductions are likely to have been larger in 1999-2005 and lower in 1993-2000 than what they report. In this context it may be noted that EUS poverty counts from 61st round are 3 and 4 percentage points higher in rural and urban areas than Dev-Ravi’s MRP estimates from 61st round CES. These differences are much less than EUS-CES differences in 55th round MRP counts, which were 7 and 5.3 percentage points in rural and urban areas.
This matter of EUS-CES differences in 55th round MRP poverty counts was the basic difference between the Sen-Himanshu (SH) and Sundaram-Tendulkar (ST) results for the 55th round. Unlike ST, who assumed no contamination from seven-day food questions to 30-day estimates in 55th round and attributed the entire EUS-CES difference to the abridged EUS schedule, SH had attempted to decompose the 55th round EUS-CES difference between EUS abridgement and CES contamination. The SH estimates of the abridgement effect (i e, difference between actual 55th round EUS-CES difference and their assessment of the extent by which CES poverty was increased because of contamination by seven-day questions) ranged between 2 and 4 percentage points in both rural and urban areas. Since the 61st round EUS-CES difference appears to be much closer to the SH rather than ST interpretation of 55th round EUS abridgement effect, this is a priori evidence that ST were wrong and SH closer to the truth on 55th round poverty. However, this is a matter on which differences are likely to persist and, in any case, exact EUS-CES MRP differences will only be available when NSSO releases the distributions of MPCE by MRP. Available URP poverty estimates from the 61st round are not enough to gauge the required magnitudes fully since past evidence (as well as DevRavi’s 61st round approximations) suggests that poverty decline by MRP is higher than by URP.
Nonetheless, it is reasonable to conclude that 61st round data not only shows much less poverty reduction than earlier assessed from the non-comparable 55th round, most of the poverty
Table 5: Poverty Headcount Ratio from EUS
Rural Urban Annual Change
in Poverty 1999-2004-05 1999-2004-05 Rural Urban 2000 2000
| Andhra Pradesh | 13.8 | 8.5 | 33.6 | 23.7 | -1.1 | -2.0 |
|---|---|---|---|---|---|---|
| Assam | 48.3 | 20.3 | 10.5 | 2.9 | -5.6 | -1.5 |
| Bihar | 51.9 | 35.9 | 33.1 | 29.2 | -3.2 | -0.8 |
| Gujarat | 18.5 | 13.2 | 19.6 | 11.5 | -1.1 | -1.6 |
| Haryana | 14.2 | 11.9 | 15.7 | 13.5 | -0.5 | -0.4 |
| Karnataka | 30.1 | 18.6 | 31.8 | 29.5 | -2.3 | -0.5 |
| Kerala | 17.6 | 11.5 | 27.7 | 19.0 | -1.2 | -1.7 |
| Madhya Pradesh | 43.9 | 30.0 | 45.6 | 41.4 | -2.8 | -0.8 |
| Maharashtra | 32.9 | 23.9 | 3 3 | 33.9 | -1.8 | 0.2 |
| Orissa | 54.3 | 42.9 | 48.7 | 46.0 | -2.3 | -0.5 |
| Punjab | 11.5 | 9.4 | 9.6 | 5.0 | -0.4 | -0.9 |
| Rajasthan | 17.5 | 16.1 | 27.9 | 26.8 | -0.3 | -0.2 |
| Tamil Nadu | 31.5 | 18.9 | 22.2 | 21.4 | -2.5 | -0.2 |
| Uttar Pradesh | 35.7 | 28.9 | 36.4 | 29.4 | -1.4 | -1.4 |
| West Bengal | 43.8 | 29.4 | 17.9 | 12.1 | -2.9 | -1.2 |
| All India | 34.0 | 24.9 | 28.9 | 25.0 | -1.8 | -0.8 |
Source: Calculated from grouped data from NSS Reports 448 and 515.
reduction found above between 1993-94 and 2004-05 appears to have occurred in the period after 1999-2000. This conclusion is also consistent with available thin sample data. The mutually comparable URP estimates from 1989-90 (45th round) to 1998 (54th round) had shown no decline at all in rural poverty, while mutually comparable MRP estimates from 2001-02 (57th round) to 2004 (60th round) had shown rural poverty declining by about 7 percentage points. The rather strange upshot from the various available NSS datasets used together with official price deflators is therefore that while the period from 1989-90 to 1998 saw insignificant rural poverty reduction despite fairly high growth in measured agricultural GDP, the period after 1999-2000 is associated with significant rural poverty reduction despite stagnant agricultural growth. In conjunction with the fact that the 61st round results reported above show that the annual rate of urban poverty reduction nearly halved in the period 1993-2005 as compared to 1983-1994, although GDP growth accelerated, there is a general indication of disassociation between GDP growth and poverty reduction in the post-reform period.
Before closing this section, we present estimates of average MPCE on URP and MRP from the 61st round statewise since these are available in the published report. Tables 6 and 7 present the estimates of MPCE by URP as well as MRP for 50th, 55th and 61st rounds along with compound annual growth rate (CAGR) of MPCE. As expected, the growth rate of MPCE by MRP between 50th and 61st rounds is slightly higher in nominal terms than growth rate of MPCE by URP. But on MRP comparison, the rate of growth of nominal MPCE is almost half during 1999-2000 and 2004-05 compared to the rate of growth in nominal MPCE between 50th and 55th rounds. This is true for both rural as well as urban areas. However, in real terms, the rate of growth of MPCE appears to be better in the recent period compared to the period between the 50th and 55th rounds. This happens because the rate of consumer price inflation decelerated sharply between the 55th and 61st rounds. In rural areas, the rate of inflation in consumer prices based on CPIAL decelerated to only 2 per cent per annum between 1999-2000 and 2004-05, compared to more than 8 per cent per annum during 1993-94
Table 6: Nominal and Real MPCE in Rural India
| Nominal MPCE (in Rs) | Real MPCE (in Rs) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Rural | 1993-94 | 1999-2000 | 2004-05 | 1993-94 | 1999-2000 | 2004-05 | ||||||
| URP | MRP | Unadj | S-H | URP | MRP | URP | MRP | Unadj | S-H | URP | MRP | |
| Andhra Pradesh | 289 | 285 | 453 | 444 | 586 | 604 | 289 | 285 | 280 | 274 | 322 | 332 |
| Assam | 258 | 272 | 426 | 417 | 543 | 577 | 258 | 272 | 270 | 264 | 320 | 340 |
| Jharkhand | 215 | 220 | 375 | 367 | 425 | 439 | 215 | 220 | 244 | 239 | 256 | 264 |
| Bihar | 219 | 229 | 387 | 380 | 417 | 445 | 219 | 229 | 252 | 247 | 251 | 268 |
| Gujarat | 303 | 319 | 551 | 538 | 596 | 645 | 303 | 319 | 348 | 340 | 334 | 361 |
| Haryana | 385 | 381 | 714 | 699 | 863 | 905 | 385 | 381 | 470 | 460 | 493 | 517 |
| Himachal Pradesh | 351 | 362 | 684 | 670 | 798 | 836 | 351 | 362 | 450 | 441 | 456 | 478 |
| Karnataka | 269 | 278 | 500 | 488 | 508 | 543 | 269 | 278 | 301 | 293 | 285 | 304 |
| Kerala | 390 | 392 | 766 | 748 | 1013 | 1031 | 390 | 392 | 492 | 481 | 582 | 593 |
| Chhattisgarh | 227 | 238 | 366 | 360 | 425 | 445 | 227 | 238 | 227 | 223 | 250 | 262 |
| Madhya Pradesh | 262 | 267 | 415 | 407 | 439 | 461 | 262 | 267 | 258 | 253 | 259 | 271 |
| Maharashtra | 273 | 278 | 497 | 486 | 568 | 597 | 273 | 278 | 298 | 292 | 296 | 311 |
| Orissa | 220 | 225 | 373 | 366 | 399 | 422 | 220 | 225 | 224 | 220 | 237 | 250 |
| Punjab | 433 | 440 | 742 | 726 | 847 | 905 | 433 | 440 | 485 | 474 | 489 | 523 |
| Rajasthan | 322 | 329 | 549 | 537 | 591 | 598 | 322 | 329 | 355 | 348 | 343 | 347 |
| Tamil Nadu | 294 | 296 | 514 | 501 | 602 | 602 | 294 | 296 | 319 | 311 | 326 | 325 |
| Uttaranchal | 306 | 318 | 558 | 546 | 647 | 649 | 306 | 318 | 361 | 353 | 370 | 371 |
| Uttar Pradesh | 272 | 275 | 463 | 454 | 533 | 539 | 272 | 275 | 300 | 293 | 305 | 309 |
| West Bengal | 279 | 286 | 454 | 446 | 562 | 576 | 279 | 286 | 290 | 284 | 326 | 334 |
| All India | 281 | 287 | 486 | 476 | 559 | 579 | 281 | 287 | 306 | 300 | 318 | 330 |
| CAGR of Nominal MPCE (in per cent) | CAGR of Real MPCE (in per cent) | |||||||||||
| URP | MRP | URP | MRP | |||||||||
| 61-50 | 55O-50 | 55SH-50 | 61-55O | 61-55SH | 61-50 | 61-50 | 55O-50 55SH-50 | 61-55O 61-55SH 61-50 | ||||
| Andhra Pradesh | 6.64 | 8.07 | 7.68 | 5.89 | 6.36 | 7.08 | 1.01 | -0.26 | -0.63 | 3.48 | 3.94 | 1.42 |
| Assam | 7.00 | 7.76 | 7.40 | 6.25 | 6.68 | 7.07 | 1.98 | -0.14 | -0.47 | 4.74 | 5.16 | 2.05 |
| Jharkhand | 6.42 | 9.32 | 8.96 | 3.23 | 3.63 | 6.51 | 1.62 | 1.75 | 1.41 | 1.65 | 2.05 | 1.70 |
| Bihar | 6.01 | 9.15 | 8.79 | 2.82 | 3.22 | 6.22 | 1.23 | 1.59 | 1.26 | 1.25 | 1.64 | 1.43 |
| Gujarat | 6.33 | 9.57 | 9.14 | 3.19 | 3.68 | 6.62 | 0.87 | 1.50 | 1.10 | 0.70 | 1.18 | 1.14 |
| Haryana | 7.61 | 11.06 | 10.66 | 4.86 | 5.31 | 8.19 | 2.28 | 3.57 | 3.20 | 1.95 | 2.39 | 2.83 |
| Himachal Pradesh | 7.76 | 11.22 | 10.83 | 4.08 | 4.52 | 7.91 | 2.42 | 3.72 | 3.35 | 1.19 | 1.62 | 2.56 |
| Karnataka | 5.95 | 10.23 | 9.79 | 1.68 | 2.17 | 6.26 | 0.50 | 1.29 | 0.88 | 0.21 | 0.69 | 0.80 |
| Kerala | 9.06 | 11.79 | 11.37 | 6.13 | 6.61 | 9.18 | 3.70 | 3.85 | 3.46 | 3.79 | 4.26 | 3.82 |
| Chhattisgarh | 5.86 | 7.49 | 7.16 | 3.96 | 4.34 | 5.87 | 0.88 | -0.72 | -1.02 | 2.87 | 3.24 | 0.89 |
| Madhya Pradesh | 4.80 | 7.62 | 7.24 | 2.11 | 2.54 | 5.08 | -0.13 | -0.60 | -0.95 | 1.03 | 1.46 | 0.14 |
| Maharashtra | 6.90 | 10.13 | 9.73 | 3.74 | 4.19 | 7.18 | 0.75 | 1.14 | 0.77 | 0.85 | 1.29 | 1.01 |
| Orissa | 5.57 | 8.79 | 8.47 | 2.50 | 2.88 | 5.89 | 0.67 | -0.07 | -0.37 | 2.26 | 2.63 | 0.98 |
| Punjab | 6.29 | 9.12 | 8.71 | 4.05 | 4.52 | 6.78 | 1.12 | 1.65 | 1.27 | 1.52 | 1.98 | 1.59 |
| Rajasthan | 5.66 | 8.89 | 8.50 | 1.74 | 2.17 | 5.58 | 0.56 | 1.28 | 0.92 | -0.47 | -0.04 | 0.48 |
| Tamil Nadu | 6.75 | 9.64 | 9.16 | 3.21 | 3.75 | 6.67 | 0.95 | 1.28 | 0.84 | 0.38 | 0.91 | 0.87 |
| Uttaranchal | 7.04 | 9.84 | 9.43 | 3.06 | 3.51 | 6.70 | 1.74 | 2.15 | 1.78 | 0.55 | 1.00 | 1.42 |
| Uttar Pradesh | 6.29 | 9.06 | 8.68 | 3.10 | 3.52 | 6.31 | 1.03 | 1.43 | 1.08 | 0.60 | 1.01 | 1.05 |
| West Bengal | 6.58 | 7.99 | 7.66 | 4.85 | 5.24 | 6.55 | 1.43 | 0.19 | -0.12 | 2.89 | 3.27 | 1.41 |
| All India | 6.43 | 9.16 | 8.78 | 3.58 | 4.02 | 6.59 | 1.12 | 1.08 | 0.72 | 1.49 | 1.93 | 1.27 |
Notes: CAGR, Compound Annual Growth Rate; S-H are food adjusted MPCE estimates based on Sen and Himanshu (2004); Unadj:-Unadjusted; Deflator used is CPIAL; 55O: Official 55th round estimates; 55SH: Sen-Himanshu estimates for 55th round.
(per ’000 births)
and 1999-2000. In urban areas, CPIIW decelerated to 3.4 per cent per annum from 8.4 per cent per annum. Clearly, lower inflation after 1999-2000 drives much of the results regarding poverty reduction in the post-55th round period.
Indicators of Well-Being from Other Sources
The evidence so far suggests that the 1990s, whether defined as 1993-2000 or as 1987-2000, was indeed a period when poverty reduction suffered a setback. On the other hand, the period 1999-2005 appears to have seen significant poverty reduction. However, poverty or more specifically income poverty is only one indicator of well-being. It is important to supplement this information with other sources of well-being of population. So long as these indicators of well-being are related to poverty levels or even broadly to income levels, an improvement in these indicators should also imply a reduction in poverty and a general improvement in well-being.
An important aspect of well-being is access to employment.
On this, NSS results show an increase in unemployment throughout the 1990s, mirroring the set-back in poverty reduction. Table 8 summarises some of the results on growth and employment.
Table 8: Growth and Employment – Some Indicators
(Growth rates in per cent per annum)
1993-94 to 1999-2000 to 1993-94 to 1999-2000 2004-05 2004-05
Agricultural self-employment -0.53 2.89 1.01 Agricultural wage employment 1.06 -3.18 -0.89 Total agricultural employment 0.03 0.85 0.40 Agricultural GDP 2.88 1.76 2.37 Non-agricultural self-employment 2.34 5.72 3.86 Non-agricultural wage employment 2.68 3.79 3.18 Rural non-agricultural employment 2.26 5.27 3.52 Urban non-agricultural employment 3.13 4.08 3.46 Secondary sector employment 2.91 4.64 3.70 Tertiary sector employment 2.27 4.67 3.35 Total non-agricultural employment 2.53 4.66 3.49 Non-agricultural GDP 8.11 7.22 7.71 Total employment 1.02 2.85 1.85
Table 7: Nominal and Real MPCE in Urban India
| Nominal MPCE (in Rs) | Real MPCE (in Rs) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Urban | 1993-94 | 1999-2000 | 2004-05 | 1993-94 | 1999-2000 | 2004-05 | ||||||
| URP | MRP | Unadj | S-H | URP | MRP | URP | MRP | Unadj | S-H | URP | MRP | |
| Andhra Pradesh | 409 | 413 | 773 | 757 | 1019 | 1091 | 409 | 413 | 470 | 461 | 522 | 559 |
| Assam | 459 | 478 | 814 | 796 | 1058 | 1130 | 459 | 478 | 503 | 492 | 593 | 633 |
| Jharkhand | 410 | 422 | 670 | 655 | 985 | 1017 | 410 | 422 | 421 | 411 | 507 | 523 |
| Bihar | 313 | 334 | 565 | 553 | 696 | 730 | 313 | 334 | 355 | 347 | 358 | 375 |
| Gujarat | 454 | 477 | 892 | 872 | 1115 | 1206 | 454 | 477 | 548 | 536 | 601 | 650 |
| Haryana | 474 | 494 | 912 | 893 | 1142 | 1183 | 474 | 494 | 560 | 549 | 584 | 605 |
| Himachal Pradesh | 747 | 618 | 1243 | 1215 | 1390 | 1422 | 747 | 618 | 763 | 746 | 711 | 727 |
| Karnataka | 423 | 439 | 911 | 890 | 1033 | 1138 | 423 | 439 | 539 | 527 | 522 | 575 |
| Kerala | 494 | 486 | 932 | 909 | 1291 | 1354 | 494 | 486 | 548 | 535 | 648 | 679 |
| Chhattisgarh | 410 | 412 | 675 | 661 | 990 | 963 | 410 | 412 | 445 | 436 | 546 | 531 |
| Madhya Pradesh | 408 | 412 | 697 | 683 | 904 | 893 | 408 | 412 | 459 | 450 | 498 | 493 |
| Maharashtra | 530 | 543 | 973 | 952 | 1148 | 1228 | 530 | 543 | 593 | 580 | 567 | 607 |
| Orissa | 403 | 411 | 618 | 605 | 757 | 790 | 403 | 411 | 390 | 381 | 419 | 437 |
| Punjab | 511 | 530 | 899 | 878 | 1326 | 1306 | 511 | 530 | 587 | 574 | 739 | 728 |
| Rajasthan | 425 | 438 | 796 | 778 | 964 | 945 | 425 | 438 | 481 | 470 | 512 | 502 |
| Tamil Nadu | 438 | 450 | 971 | 949 | 1080 | 1166 | 438 | 450 | 598 | 585 | 575 | 621 |
| Uttaranchal | 510 | 508 | 836 | 818 | 978 | 1028 | 510 | 508 | 519 | 508 | 524 | 550 |
| Uttar Pradesh | 384 | 393 | 682 | 667 | 857 | 880 | 384 | 393 | 424 | 414 | 459 | 471 |
| West Bengal | 474 | 488 | 866 | 847 | 1124 | 1159 | 474 | 488 | 504 | 493 | 600 | 619 |
| All India | 458 | 468 | 855 | 836 | 1052 | 1105 | 458 | 468 | 526 | 515 | 548 | 575 |
| CAGR of Nominal MPCE (in per cent) | CAGR of Real MPCE (in per cent) | |||||||||||
| URP | MRP | URP | MRP | |||||||||
| 61-50 | 55O-50 | 55SH-50 | 61-55O | 61-55SH | 61-50 | 61-50 | 55O-50 55SH-50 | 61-55O 61-55SH 61-50 | ||||
| Andhra Pradesh | 8.66 | 11.00 | 10.61 | 7.13 | 7.59 | 9.23 | 2.25 | 2.17 | 1.81 | 3.52 | 3.96 | 2.78 |
| Assam | 7.90 | 9.26 | 8.85 | 6.78 | 7.25 | 8.12 | 2.37 | 0.82 | 0.45 | 4.73 | 5.20 | 2.58 |
| Jharkhand | 8.30 | 8.02 | 7.62 | 8.70 | 9.20 | 8.33 | 1.95 | -0.04 | -0.41 | 4.46 | 4.93 | 1.98 |
| Bihar | 7.55 | 9.19 | 8.79 | 5.23 | 5.69 | 7.37 | 1.24 | 1.04 | 0.67 | 1.12 | 1.57 | 1.08 |
| Gujarat | 8.51 | 10.98 | 10.57 | 6.22 | 6.70 | 8.79 | 2.58 | 2.33 | 1.95 | 3.46 | 3.93 | 2.85 |
| Haryana | 8.33 | 10.77 | 10.39 | 5.35 | 5.79 | 8.27 | 1.92 | 2.13 | 1.78 | 1.56 | 1.98 | 1.87 |
| Himachal Pradesh | 5.81 | 12.34 | 11.91 | 2.73 | 3.21 | 7.87 | -0.45 | 3.57 | 3.18 | -0.96 | -0.51 | 1.49 |
| Karnataka | 8.45 | 12.95 | 12.52 | 4.56 | 5.04 | 9.05 | 1.93 | 3.50 | 3.11 | 1.29 | 1.76 | 2.49 |
| Kerala | 9.13 | 11.46 | 10.99 | 7.74 | 8.29 | 9.76 | 2.50 | 2.02 | 1.59 | 4.38 | 4.91 | 3.09 |
| Chhattisgarh | 8.34 | 8.57 | 8.20 | 7.36 | 7.80 | 8.02 | 2.63 | 1.27 | 0.93 | 3.61 | 4.04 | 2.33 |
| Madhya Pradesh | 7.51 | 9.18 | 8.81 | 5.08 | 5.51 | 7.30 | 1.85 | 1.84 | 1.49 | 1.41 | 1.82 | 1.64 |
| Maharashtra | 7.29 | 10.20 | 9.80 | 4.77 | 5.23 | 7.70 | 0.62 | 1.47 | 1.10 | 0.46 | 0.90 | 1.01 |
| Orissa | 5.91 | 7.03 | 6.64 | 5.02 | 5.48 | 6.11 | 0.36 | -0.89 | -1.25 | 2.30 | 2.75 | 0.55 |
| Punjab | 9.06 | 9.20 | 8.79 | 7.77 | 8.25 | 8.55 | 3.41 | 1.72 | 1.34 | 4.38 | 4.86 | 2.92 |
| Rajasthan | 7.74 | 10.46 | 10.05 | 3.49 | 3.95 | 7.23 | 1.71 | 1.56 | 1.19 | 0.85 | 1.30 | 1.24 |
| Tamil Nadu | 8.54 | 13.70 | 13.26 | 3.73 | 4.22 | 9.05 | 2.50 | 4.89 | 4.48 | 0.74 | 1.21 | 2.98 |
| Uttaranchal | 6.10 | 8.65 | 8.26 | 4.21 | 4.67 | 6.61 | 0.24 | 0.36 | 0.00 | 1.16 | 1.60 | 0.72 |
| Uttar Pradesh | 7.58 | 9.64 | 9.23 | 5.21 | 5.69 | 7.60 | 1.64 | 1.27 | 0.89 | 2.12 | 2.58 | 1.66 |
| West Bengal | 8.16 | 10.05 | 9.64 | 5.99 | 6.47 | 8.19 | 2.16 | 0.55 | 0.18 | 4.19 | 4.66 | 2.19 |
| All India | 7.86 | 10.55 | 10.14 | 5.26 | 5.73 | 8.11 | 1.64 | 1.97 | 1.59 | 1.77 | 2.22 | 1.88 |
Notes: CAGR, Compound Annual Growth Rate; S-H are food adjusted MPCE estimates based on Sen and Himanshu (2004); Unadj:-Unadjusted; Deflator used is CPIAL; 55O: Official 55th round estimates; 55SH: Sen-Himanshu estimates for 55th round.
Table 9 presents growth rates of unemployment days, non-farm workers and wages of rural male casual workers for major states from the NSS employment-unemployment surveys.10 For the working population as a whole, daily status unemployment increased from 6.1 per cent in 1993-94 to 7.3 per cent in 19992000 to 8.3 per cent in 2004-05, and was higher in 2004-05 than 1993-94 in all states except West Bengal, Assam and Gujarat. More importantly, daily status unemployment among agricultural labour households (who are the poorest) increased from 9.5 per cent in 1993-94 to 12.3 per cent in 1999-2000 and further to
15.3 per cent in 2004-05. Yet, importantly, the surveys also suggest that the growth rate of employment, at 2.85 per cent per annum, outpaced population growth rate between 1999-2000 and 2004-05, after having been much less than population growth, at only 1.02 per cent per annum, between 1993-94 and 1999-2000. Closer inspection of the data suggests that most of this increase in employment during 1999-2005 was in the form of selfemployed with days of wage labour actually declining. But, nonetheless, one possible reason why poverty may have declined more in 1999-2005 period is that worker-population ratios were increasing during this period after a fairly significant decline in the previous period. Along with this, some other structural indicators from the employment-unemployment surveys also provide possible reasons for faster poverty reduction during 1999-2005 compared to 1993-2000. For example, the share of casual labourers in general (and agricultural wage workers in particular) actually declined during 1999-2005 after a sharp increase during 1993-2000 and also rural non-agricultural employment registered a sharp increase from 1999-2000 onwards, following almost complete stagnancy earlier in the 1990s.
However, matters are complicated by the fact that this surge in employment growth after 1999 was accompanied not by higher growth in wage rates but by their stagnation. Tables 10 and 11 summarise all-India trends in wages at constant 1999-00 prices. From these tables, the trend is not only of a clear deceleration in real wages of casual workers, there is an even greater deceleration in wages of regular workers in both rural and urban areas. That is, real wages decelerated for all workers significantly during 1999-2000 to 2004-05 compared to wage growth between 1993-94 and 1999-2000. And this was true for rural and urban, agriculture and non-agriculture, male and female and at all levels of education. If wages are taken to be the main indicator of wellbeing and poverty, then poverty reduction should have been less between 1999-2000 and 2004-05 compared to the previous period, a trend contradictory to the poverty estimates reported above.
The only way of reconciling such contradictory trends on wages and poverty, both from the same EUS data, is to concentrate on the poorest workers and on the regional picture. In the previous literature on poverty, it is not wages in general but wages of casual workers, particularly of agricultural labourers, that have been found to correlate very well with poverty indices.11 Given this, a re-look at Table 9 shows that, although casual real wages decelerated between 1999-2000 and 2004-05 for all-India, these increased at more than 3 per cent per annum in Assam, Bihar, Madhya Pradesh and Orissa, and in fact accelerated in most of these poor states. Also, it is evident from Tables 10 and 11 that the deceleration in real wages was least for casual workers in agriculture and independent evidence on this also shows better performance in states with higher poverty.12
Wages are an important indicator of well-being, but since these are also income indicators, it is useful to look at some of the non-income indicators. On this, the preliminary results are available from the National Family Health Survey (NFHS)-3 of 2005-06.13 The previous two surveys in this regard were NFHS-1 (199293) and NFHS-2 (1998-99). That is, these three surveys are roughly co-terminus with the NSS thick round surveys. The preliminary results available from the NFHS are given in Table 12. Of the various indicators, we have taken those which reflect the well-being of children under six. Since they constitute the most marginalised group, an improvement in well-being of children can be taken as improvement in well-being of the average population. On the basis of the three important indicators of infant mortality rate, the percentage of children who are underweight and have been immunised, we have also created an “Achievement of Babies and Children (ABC) index”.14
If these are any indication of well-being of children, it would be difficult to claim that the period between NFHS-2 and NFHS-3 saw more rapid improvement on these indicators compared to the period between NFHS-1 and NFHS-2. While the ABC index increased by 6.7 points (18.4 per cent) between NFHS-1 and NFHS-2, the increase in the later period was only 5.7 points (13.1 per cent). Further, some other indicators point towards deterioration in conditions of children between NFHS-2 and NFHS-3. For example, the percentage of children under age 3 who are categorised
Table 9: Growth Rate of Unemployment, Non-farm Workersand Wages by States
(in per cent per annum)
Unemployment Non-farm Worker Rural Male Real
Wage 1993-1999-1993-1999-1993-19992000 2005 2000 2005 2000 2005
| Andhra Pradesh | 4.9 | 8.2 | 0.7 | 7.1 | 4.3 | 1.5 |
|---|---|---|---|---|---|---|
| Assam | 0.4 | 0.8 | 9.9 | -0.1 | 2.4 | 3.5 |
| Bihar | 4.6 | 1.5 | 5.1 | 7.7 | 5.1 | 3.4 |
| GujaratHaryana Karnataka | -0.8 -5.2 -0.3 | -1.4 12.2 13.3 | 0.6 1.8 -1.4 | 4.5 8.6 3.8 | 3.4 2.8 3.8 | 1.1 0.8 0.8 |
| Kerala | 9.6 | 6.2 | 4.3 | 3.1 | 5.5 | 3.6 |
| Madhya PradeshMaharashtra | 7.5 8.0 | 13.7 10.9 | 3.7 0.0 | 7.8 5.5 | 0.9 3.1 | 3.4 -0.5 |
| Orissa | 1.1 | 10.9 | 2.0 | 11.1 | 1.7 | 5.9 |
| Punjab RajasthanTamil Nadu | 7.6 17.8 0.6 | 25.7 12.8 2.3 | 4.4 1.7 -0.3 | 6.7 7.1 1.1 | 0.2 3.4 6.5 | 0.0 0.6 0.0 |
| Uttar Pradesh | 3.2 | 4.3 | 3.7 | 7.4 | 2.9 | 1.5 |
| West Bengal | 13.0 | -6.5 | -1.0 | 4.3 | 2.8 | 0.2 |
(in per cent per annum)
| 1993-94 to 1999-2000 Agriculture Non-agriculture | 1999-2000 to 2004-05 Agriculture Non-agriculture | |||
|---|---|---|---|---|
| Male Female P er s on s | 2.80 2.95 2.78 | 3.67 5.13 4.19 | 1.38 1.04 1.31 | 0.67 1.51 0.76 |
Table 11: Growth Rate of Real Wages of Regular Workersby Education Status
(in per cent per annum)
Rural Urban 1993-94 to 1999-2000 to 1993-94 to 1999-2000 to 1999-2000 2004-05 1999-2000 2004-05
Not literate 6.18 -1.67 2.63 -1.00 Primary 3.88 -0.57 3.42 -2.20 Secondary 4.33 -0.72 4.37 -1.74 Graduates 6.04 2.00 5.27 1.91 All 5.38 0.56 5.01 0.21
as wasted, increased in 18 of the 22 states for which data is available,15with the all-India incidence rising from 14 per cent to 17 per cent. Children suffering from anaemia increased from 74 per cent to 79 per cent16 and similar was the case of women suffering from anaemia, with the overall percentage of women suffering from anaemia increased from 50 per cent in NFHS-2 to 51 per cent in NFHS-3.17
However, there are some NFHS indicators which show improvement between NFHS-2 and NFHS-3. For example, the incidence of stunting among children shows a fairly sharp fall from 46 per cent to 37 per cent, and this occurred in all states except Karnataka. Also, there was a decline in the percentage of women reporting low body mass index (BMI) in all states except Assam and Madhya Pradesh. Most importantly, just as with wage data, matters appear to have been better in some of the poorer states with Chhattisgarh, Rajasthan, Madhya Pradesh and Uttar Pradesh among the very few cases where ABC improvement was more between NFHS-2 and NFHS-3 than between NFHS-1 and NFHS-2. All these states, along with Orissa and West Bengal, also show better than all-India improvement in the ABC index between NFHS-2 and NFHS-3. The situation on student attendance in educational institutions is similar. The NSS shows attendance rates of 6-14 children up from 68 per cent in 1993-94 to 72 per cent in 1999-2000 to 82 per cent in 2004-05. This implies a better rate of improvement after 1999-2000.
Nonetheless, the results on poverty above sit awkwardly with the undoubted fact that an agrarian crisis (involving both large numbers of farmer suicides and starvation deaths in different parts of country) unfolded itself particularly in the period after 1997. The NSS estimates of per capita cereal consumption show this to be lower in 2004-05 than in 1993-94 for all states in both rural and urban areas. There is however no clear indication that the pace of decline was much different in 1999-2005 as compared to 1993-2000. Indeed, some poor areas such as rural Assam and urban Uttar Pradesh recorded an increase in the later period after declining in the earlier. Also, since these comparisons are likely to have been affected by 55th round contamination, it is possible that overall the decline in cereals consumption was actually greater in 1993-2000 than in 1999-2005. This would be in line with trends in the relative price of cereals which rose sharply in the first period and fell sharply in the second, although such price effects appear dominated by a secular shift of the cereal Engels curves throughout the 1990s.
Why Does Poverty Decline Fasterduring 1999-2005?
The previous section outlined some of the developments which would have crucial bearing on explaining poverty reduction. As far as poverty reduction is concerned, the only consistent story based on mutually comparable surveys and recall periods is:
(1) that poverty reduced during 1993-2005, but the annual rate of decline was lower than in the previous decade of 1983-1994 and (2) that it was also low compared to the annual decline between 1999-2005, implying that bulk of the decline between 1993-2005 happened in the last five years of the period concerned, with very little poverty reduction achieved between 1993-2000. Moreover, the picture emerging from the EUS of poverty reduction during 1999-2005 also shows that most of the poverty reduction at national level is driven by sharp poverty reduction in the states which had more than national average poverty till 1999-2000.18 The obvious question is what caused this high poverty reduction during 1999-2005 and whether this is real or just a statistical artefact. This becomes even more complex if juxtaposed against evidence emerging from other indicators of well-being, such as nutrition status. That is, if poverty did reduce much faster during 1999-2005 when agricultural output and wages were growing slower, why did many of these non-income indicators not improve even more in the period 1999-2005 compared to 1993-2000?
One possible answer to this puzzle is actually as old as studies on determinants of poverty. This concerns the relative price of food and cereals in particular. The impact of relative price as a variable has figured prominently in the initial debate on rural poverty in India, both independently of agricultural production and in relative significance with it.19 This was again reiterated by Sen (1996) while explaining poverty
Table 12: Child Indicators from NFHS
NFHS-1 NFHS-2 NFHS-3 (1992-93) (1998-99) (2005-06)
Infant mortality rate (per ’000 births) 77.3 67.3 55.5 Per cent children underweight
(age < 3 years) 51.1 46.7 43.3 Children fully immunised (in per cent) 38.3 44.2 46.0 ABC index 36.6 43.3 49.0
Source: Calculated as weighted average of 22 states for which data is available.
Table 13: Growth Rate of Consumer Prices by Groups
(in per cent per annum)
1993-94 to 1999-2000 to FGR/NFGR 1999-2000 2004-05 CPIAL Food Non-Total Food Non-Total 1993-1999food food 2000 05
Andhra Pradesh 8.80 7.21 8.38 1.93 3.37 2.30 1.22 0.57 Assam 8.10 7.57 7.98 0.40 4.14 1.26 1.07 0.10 Bihar 7.39 7.70 7.47 0.71 4.00 1.54 0.96 0.18 Gujarat 8.25 6.87 7.92 1.61 5.14 2.46 1.20 0.31 Jammu and
Kashmir 9.10 7.24 8.66 0.54 4.59 1.51 1.26 0.12 Karnataka 9.58 6.84 8.85 0.26 4.79 1.47 1.40 0.05 Kerala 6.64 10.59 7.66 3.16 -0.15 2.27 0.63-20.52 Madhya Pradesh 8.50 7.32 8.25 0.36 3.74 1.08 1.16 0.10 Maharashtra 9.52 6.76 8.87 2.13 5.40 2.88 1.41 0.39 Orissa 9.13 7.93 8.85 -0.87 3.71 0.25 1.15 -0.24 Punjab 7.11 7.86 7.29 2.40 3.19 2.60 0.91 0.75 Rajasthan 7.78 6.62 7.50 1.26 5.11 2.23 1.18 0.25 Tamil Nadu 7.57 9.92 8.24 2.20 4.29 2.84 0.76 0.51 Uttar Pradesh 7.96 5.96 7.51 1.42 5.22 2.28 1.33 0.27 West Bengal 7.99 7.30 7.80 0.81 4.64 1.87 1.09 0.17 All India 8.15 7.57 8.01 1.30 4.22 2.03 1.08 0.31 CPIIW Andhra Pradesh 8.48 8.32 8.64 2.64 7.27 3.49 1.02 0.36 Assam 8.59 8.34 8.37 1.00 4.28 1.95 1.03 0.23 Bihar 7.96 7.28 8.06 2.08 10.33 4.06 1.09 0.20 Gujarat 8.18 8.41 8.45 2.12 5.48 2.67 0.97 0.39 Haryana 7.98 9.62 8.46 2.60 7.18 3.73 0.83 0.36 Jammu and
Kashmir 9.28 10.58 9.72 5.19 7.83 5.68 0.88 0.66 Karnataka 9.11 8.18 9.12 2.20 7.76 3.22 1.11 0.28 Kerala 8.98 9.37 9.25 2.63 6.22 3.22 0.96 0.42 Madhya Pradesh 7.08 6.83 7.21 3.01 6.31 3.62 1.04 0.48 Maharashtra 8.33 9.74 8.60 2.89 8.34 4.29 0.85 0.35 Orissa 7.58 7.99 8.00 2.22 5.49 2.66 0.95 0.40 Punjab 7.16 7.86 7.35 2.29 6.41 3.24 0.91 0.36 Rajasthan 8.80 8.30 8.76 1.96 5.06 2.62 1.06 0.39 Tamil Nadu 8.20 8.44 8.41 1.81 7.45 2.97 0.97 0.24 Uttar Pradesh 8.16 7.66 8.26 2.48 5.68 3.02 1.07 0.44 West Bengal 9.35 9.24 9.44 1.27 3.76 1.73 1.01 0.34 All India 8.22 8.47 8.41 2.53 6.88 3.43 0.97 0.37
Note: FGR: Food growth rate; NFGR: Non-food growth rate.
reduction in the 1980s, particularly between 1983 and 1987-88, the later being a drought year.20
Table 13 gives the growth rates in consumer prices for food and non-food for rural and urban areas respectively.21 The last two columns of the table give the ratio of growth rates of the food and non-food groups. Between 1993-94 and 1999-2000, food prices were increasing faster than non-food prices in rural all-India by 8 per cent and this was the case in all states except Bihar, Tamil Nadu, Kerala and Punjab. After 1999-2000, the rate of increase in food prices was only 31 per cent of the growth rate of non-food prices. While this was true for all states in rural areas, what is important is the fact that this ratio was the lowest in Assam, Bihar, Karnataka and Madhya Pradesh, and food prices actually declined in Orissa. These states are also where the bulk of the rural poor is located and it is important to note that these are precisely the states that show the sharpest reduction in poverty during 1999-2005, as presented in Table 5. In fact, the fall in relative price of food is also responsible for some of these states showing higher real wage rate growth during 1999-2005. From this, one possible story emerges: that the high food price increase between 1993-94 and 1999-2000 nullified much of the poverty reduction that could be expected from the improvements in wages and agricultural productivity that did occur during the period; and that, conversely, because food price growth decelerated sharply between 1999-2000 and 2004-05, there was rapid poverty reduction despite lower growth of wage rates and agricultural output.
This story can be buttressed by some aspects of the regional pattern of poverty reduction found above and relating them to patterns in employment growth which did improve sharply between 1999-2000 and 2004-05. For example, although most nonpoverty indicators do not indicate any marked improvement in this period compared to 1993-2000, it does appear that the few success stories were also in states which have traditionally had higher than national average poverty with states like Bihar (including Jharkhand), Assam, Orissa, Chhattisgarh and West Bengal not only doing better on wage rate growth but also on various other indicators. Their performance on total SDP growth and agricultural SDP growth is also better in this period compared to the previous period.22In employment, Orissa shows the highest non-farm diversification followed by Haryana, Bihar and Madhya Pradesh. On unemployment, which is generally increasing, West Bengal shows an absolute decline followed by Gujarat and Assam. Bihar shows unemployment rising by only 1.4 per cent during 1999-05 compared to 5.1 per cent during 1993-2000. Since most of these states are also states with higher than national poverty, an improvement in these states also drives down national poverty at a faster rate. These regional patterns could have reinforced the effect of a lower relative food price.
However, although plausible, the above explanations cannot very convincingly explain large rural poverty reduction in a period of undoubted agrarian distress, unless the necessary links are further elucidated. It is not possible to do this until the unit level data from 61st round are released. Till then, such explanations must remain tentative and it is also not possible to reject an alternative interpretation: that the rural poverty decline estimated above for 1999-2005 is simply a statistical artefact driven by artificially low increases in the poverty lines because of low food prices. This interpretation is likely to be offered by those who have consistently argued against the current method of poverty calculations and their argument is likely to get strengthened if, as is likely, the nutritional intake data yet to be released from the 61st round turns out to show lower real food expenditure and lower nutrient intake during 1999-2005 despite the sharp fall in real food price.
Conclusion
The following emerge as unambiguous conclusions from the previous analysis: First, on a mutually comparable basis, URP poverty from CES has come down in all states between 1993-94 and 2004-05, but the annual poverty reduction is slower than in the previous decade of 1983 to 1993-94. Indeed, poverty reduction after the early 1990s has been slower than in the 1970s or 1980s. Second, again on a mutually comparable basis from EUS data, poverty seems to have reduced sharply between 1999-2000 and 2004-05. Third, although further analysis is required with MRP data from the 61st round CES when this becomes available, the two conclusions above imply that there was actually very little poverty reduction between 1993-94 and 1999-2000, so that the Sen-Himanshu estimates for this period were closest to the truth in the debate on this matter.
The paper has also put together some external evidence on improvements in non-income indicators of well-being between 1993-94 and 2004-05. On this, the evidence is mixed, but at first sight it appears that the period after 1999-2000 does not show any major improvements. However, there are a few sparkling stars, especially among the traditional high poverty incidence states in the eastern part of the country. It also appears that some of these high poverty states did better not only on these nonincome indicators but also on employment and income indicators. A possible optimistic story therefore is that certain developments have occurred in these high poverty states after 1999 that have not only caused them to show improvement but also, given their high weight in all-India poverty measures, driven down the all-India poverty estimates. This could still be consistent with agrarian distress, since most indicators of these, notably suicides, are concentrated in the relatively richer states and among farmers who may not be poor given the rock-bottom definition of the poor by the official poverty lines.23 Interestingly, some of these relatively rich states not only saw a worsening of income and employment indicators, their performance on non-income indicators is also a cause for concern. In any case, the 61st round throws open once again the whole issue of a possible growing disjunction between growth, poverty reduction and well-being. This is a matter that needs to be examined in detail.
However, very few policy conclusions are possible from the 61st round at this stage. The only message that emerges loud and clear is that the variable which matters most for measured poverty reduction is the relative price of food. This is partly because of the way poverty is currently measured, but it remains undoubtedly true that since most of the poor are net purchasers of food, changes in relative price of food have an immediate impact on their well-being. This message is particularly important in the current context when food prices are again sky-rocketing after a period of very low food price inflation and when the government appears to be at best indifferent to the fact that nutritional indicators, particularly for women and children, had decelerated in most states even during the period when food prices were low. Policy makers need to address urgently the current weaknesses of the Targeted Public Distribution System and the Integrated Child Development Services. The only other policy matter that can be highlighted from the discussion above is
is CPIAL; 55O: Official 55th round estimates; 55SH: Sen-Himanshu estimates for 55th round.
regarding the high growth in number of rural non-farm workers during 1999-05. Although this growth may well have contributed positively to poverty reduction by relieving pressure on a slowly growing agriculture, it is clear that most of this was at low wages and/or productivity and some, especially self-employment among females, could even have been distress driven. As in the case of food prices, the positive aspects of this may have dominated during 1999-2005 but this need not endure if, in the current obsession with GDP growth, it is simply assumed that higher GDP will somehow provide the required solutions.
Finally, with food consumption declining despite a fall in the relative price of food, there can be valid doubts about the way we currently measure poverty. First, with the share of food in total expenditure declining, the present weighting diagram for poverty lines is clearly unrealistic and may be exaggerating the recent poverty decline.24 Second, the interstate and rural-urban variations in existing poverty lines are well known to be faulty and it is high time that these are corrected. Thirdly, it is confusing that our present method forces us to arrive at very different poverty estimates for the same population by different survey instruments (such as MRP, URP, CES or EUS). Surely, what is varying is the way we measure the expenditure of the same individual and not whether the individual is poor or not. The correct approach would be to assign different poverty lines to these different instruments so that they show a similar incidence of poverty. An expert group has reportedly been set up on poverty measures. Its views on these matters are awaited eagerly.
EPWEmail: himanshu@gmail.com
Appendix
Official poverty lines for 1999-2000 were updated for each state using commodity-wise CPIAL for rural areas and CPIIW for urban areas following the Expert Group Method (1993). The statewise rural and urban poverty lines for 2004-05 used in this paper are:
State Specific Poverty Lines for 2004-05
(Rs/month)
Rural Urban
| Andhra Pradesh | 292.95 | 544.30 |
|---|---|---|
| Assam | 387.64 | 378.38 |
| Bihar | 356.36 | 461.40 |
| Gujarat | 353.93 | 540.80 |
| Haryana Himachal Pradesh | 414.76 394.20 | 504.20 504.20 |
| Jammu and Kashmir | 391.26 | 504.20 |
| Karnataka | 324.17 | 603.50 |
| Kerala | 429.07 | 562.90 |
| Madhya Pradesh | 324.48 | 569.00 |
| Maharashtra | 362.25 | 664.50 |
| Orissa | 325.65 | 544.00 |
| Punjab | 410.38 | 456.10 |
| Rajasthan Tamil Nadu | 374.57 351.86 | 531.10 551.70 |
| Uttar Pradesh | 369.76 | 487.10 |
| West Bengal All-India | 382.82 358.03 | 446.10 540.40 |
Notes
[I would like to thank Abhijit Sen for helpful discussion on many issues covered in this paper. Thanks are also due to Mahendra Dev and C Ravi for sharing their paper (published in this issue) and for useful comments. Usual disclaimers apply.]
1 This paper is based on 61st round data from the recently released NSS report Nos 508 and 515. The poverty lines for 2004-05 (see Appendix) have been calculated based on methodology suggested by the Expert Group on Poverty Estimation (1993) using commodity groupwise consumer price indices for agricultural labourers for rural areas and consumer price indices for industrial workers for urban areas.
2 Since three states, namely, Bihar, Madhya Pradesh and Uttar Pradesh,were bifurcated in 2001, the published reports from the 61st round are available for the bifurcated states separately. To make these comparable with previous estimates, these measures for Jharkhand, Chhattisgarh and Uttaranchal were calculated separately from unit records for the 38th, 43rd and 50th rounds. Moreover, poverty estimates for the bifurcated Bihar, Madhya Pradesh and Uttar Pradesh were also calculated. Bihar,Madhya Pradesh and Uttar Pradesh in this section refer to the present geographical regions represented by these states. However, these refer to the un-bifurcated states in later sections of this paper.
3 The pace of poverty reduction during 1973-74 to 1983 is from estimates of the expert group [GoI 1993].
4 The analysis here is on the basis of simple per annum changes in thepoverty ratio over survey years. However, some analysts prefer to normalise this change by the initial poverty level. Expressed as a percentage of the initial poverty rate, all-India annual reduction was 2.0 during 1973-88, 1983-94 and 1993-05. This measure has therefore not changed over the years, although a lower rate of 1.7 is obtained for 1987-2005.
5 However, matters need to be investigated further on this. Available SDPestimates at state level also indicate that many states with high poverty rates, e g, Assam, Bihar, Chhattisgarh and West Bengal seem to be doing reasonably better between 1999-2000 and 2004-05 compared to their performance between 1993-94 and 1999-2000.
6 Claims that poverty was growing more in states already more poor and that per capita MPCE was growing less where this was already less weremade, e g, by Deaton-Dreze (2002). However, Sen-Himanshu (2004) had shown that many of these claims were results of MRP-URP differences which were further exaggerated by the adjustments made to deal with 55th round incomparability. Also, Sen-Himanshu had noted that urban poverty was apparently growing more, although the growth of urban MPCE was much higher than that of rural MPCE.
7 In this table, Bihar, Madhya Pradesh and Uttar Pradesh refer to the geographical area of the erstwhile undivided states. 8 These five states have the highest difference between the per annum change in Table 5 and corresponding change during 1993-2005 in Table 3.
9 If the 61st round differences between EUS and CES-URP are taken as valid for the 55th round, rural poverty did not decline between rounds50 and 55, and urban poverty declined by only 2.8 percentage points.
10 Unemployment growth rates are growth rate of unemployment person days. Non-farm worker growth rate is by usual status.
11 The correlation between agricultural wages and poverty indices at the state level is around 0.9 for 43rd, 50th and 55th rounds [Deaton and Dreze 2002].
12 NSS has not yet reported statewise agricultural wages but somedetails are available from Wage Rates in Rural India (Labour bureau). This too shows all-India real wages actually declining for both males and females between 1999-2000 and 2004-05. But at the state level, this shows high real wage growth in Bihar (including Jharkhand), Orissa, Assam and West Bengal.
13 NFHS-3 results reported here are taken from the fact-sheet for variousstates available at www.nfhsindia.org. Key results for 22 states are available on the web site.
14 The idea for ABC index is taken from the FOCUS report (2006). Here it is the simple average of percentage of children fully immunised, per cent not underweight and the infant survival rate, all from NFHS.
15 The only states reporting a reduction in wasting were Assam, Chhattisgarh,Himachal Pradesh, Karnataka, Maharashtra and Orissa.
16 The states where the percentage increased are Andhra Pradesh, Assam, Gujarat, Karnataka, Kerala, Madhya Pradesh, Orissa, Tamil Nadu and Uttar Pradesh.
17 The states where the percentage of women with anaemia increased are Andhra Pradesh, Delhi, Gujarat, Haryana, Karnataka, Kerala, MadhyaPradesh, Rajasthan, Uttar Pradesh and West Bengal.
18 These states are Assam, Bihar, West Bengal, Madhya Pradesh and Orissa in descending order of annual percentage point poverty reduction during 1999-2005. These five states together accounted for a little over 50 per cent of all poor in rural India.
19 Saith (1981) demonstrated that while rural poverty and fluctuations in agricultural production were inversely related, fluctuations in consumer prices aggravated rural poverty. Dharam Narain in his unfinished work also argued that poverty was related to higher food prices. This specification was challenged because his use of nominal food prices as an explanatory variable was refuted by most economists who said that what really matters are relative prices, and if absolute prices need to be incorporated this should be doneby considering the rate of inflation rather than price level. As Sen (1985) argues, prices contributed to rural poverty because prices received failed
References
to catch up with prices paid as consumers of food. In other words it is
relative prices of food which are usually linked to rural poverty. Srinivasan(1985) demonstrated that the number of poor is a function of agriculturaloutput and the current prices of agricultural commodities. Gaiha (1989)demonstrated that while rural poverty and agricultural production wereinversely related, the effect of agricultural production in some cases wasweak or absent. The effect of price fluctuation on the other hand wasconsistently strong and often decisive. More specifically unanticipatedinflation in an index of consumer prices aggravated rural poverty.
20 Sen argues that despite the weakening of the link with agriculture, rural areaswere the beneficiaries of both increased agricultural prices as well ascheaper food prices. The price of agricultural goods in the 1980s were risingfaster than the general price level, reversing the earlier trend of movementof terms of trade against agriculture. But the fact that this increase inagricultural prices did not have unbearable inflationary implications waspartly because of the weakening of the link between agriculture and nonagriculture and partly because of the intervention of the state throughthe public distribution system. As a result, even though agricultural pricesas a whole increased faster than the general price levels, cereals pricesincreased slower so that it was possible for real wages to rise withoutincreasing product wages correspondingly [Sen 1996; Sen and Ghosh1993]. Coupled with increased income from increased agricultural prices, italso meant that the farming community remained insulated from the instability of agricultural production even during years of monsoon failure.
21 All-India commodity weights have been used to arrive at non-food growthfigures. Use of state level commodity weights will not change these resultssignificantly.
22 Among the states where per capita SDP growth increased during1999-05 compared to 1993-2000, the highest increase was observed inthe case of Assam, Bihar, Chhattisgarh and Orissa.
23 Incidentally, some of these high profile states such as Punjab, Maharashtra,Karnataka and Andhra Pradesh have also seen large farmer suicides andagrarian distress in general.
24 The current weighting diagram for poverty lines gives 81 per cent weightto food in rural areas and 75 per cent in urban areas, whereas in2004-05 the actual weight of food in the MPCE class around the povertyline is around 58 per cent in rural and 52 per cent in urban areas.
Deaton, Angus and Jean Dreze (2002): ‘Poverty and Inequality in India: A Re-examination’, Economic and Political Weekly, Vol 37, No 36, September 7.
Dev, Mahendra and C Ravi (2007): ‘Poverty and Inequality for All India and Major States: 1983 to 2004-05’, CESS working paper, processed.
Gaiha, Raghav (1989): ‘Poverty, Agricultural Production and Prices Fluctuations in Rural India: A Reformulation’, Cambridge Journal ofEconomics, 13(2), pp 333-52
Focus on Children Under Six (FOCUS) Report (2006): Citizen’s Initiative for the Rights of Children Under Six, available at www.righttofoodindia.org
GoI (1993): ‘Report of the Expert Group on Estimation of Proportion andNumber of Poor’, Perspective Planning Division, Planning Commission, New Delhi.
Saith, A (1981): ‘Production, Prices and Poverty in Rural India’, Journal of Development Studies, 17, January.
Sen, Abhijit (1996): ‘Economic Reforms, Employment and Poverty: Trends and Options’, Economic and Political Weekly, Special No, September.
Sen, Abhijit and Jayati Ghosh (1993): ‘Trends in Rural Employment and the Poverty-Employment Linkage’, ILO-ARTEP Working Paper, International Labour Organisation, New Delhi.
Sen, Abhijit and Himanshu (2004a): ‘Poverty and Inequality in India-I’,Economic and Political Weekly, Vol 39, No 38, September 18.
– (2004b): ‘Poverty and Inequality in India – Growing Disparities in the1990s’, Economic and Political Weekly, Vol 39, No 38, September 25.
Sundaram, K and S D Tendulkar (2003): ‘Poverty in India in the 1990s: Revised Estimates’, Economic and Political Weekly, Vol 38, No 46, November 15.
Sen, A K (1985): ‘Dharam Narain on Poverty: Concepts and Broader Issues’ in J Mellor and G Desai (eds), Agricultural Change and Rural Poverty:Variations on a Theme by Dharam Narain, Johns Hopkins University Press, Baltimore, MD, USA.
Srinivasan, T N (1985): ‘Agricultural Production, Relative Prices, Entitlements and Poverty’ in J Mellor and G Desai (eds), Agricultural Change andRural Poverty: Variations on a Theme by Dharam Narain, Johns Hopkins University Press, Baltimore, MD, USA.
NEW RELEASE
SAMEEKSHA TRUST BOOKS
Inclusive Growth K N Raj on Economic Development
Essays from The Economic Weekly and Economic and Political Weekly
Edited by ASHOKA MODY
The essays in the book reflect Professor K N Raj’s abiding interest in economic growth as a fundamental mechanism for lifting the poor and disadvantaged out of poverty. He has also been concerned that the political bargaining process may end up undermining growth and not provide support to those who were excluded from access to economic opportunities. These essays, many of them classics and all published in Economic Weekly and Economic and Political Weekly, are drawn together in this volume both for their commentary on the last half century of economic development and for their contemporary relevance for understanding the political economy of development in India and elsewhere.
Pp viii + 338 ISBN 81-250-3045-X 2006 Rs 350
Available from
Orient Longman Ltd
Mumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur Lucknow Patna Chandigarh Hyderabad
Contact: info@orientlongman.com
