Poor in India: Region-wise estimates

Recently Professor YK Alagh wrote two opinion pieces on Posture-nomics and Models and Realities in The Indian Express where he referred to my work on poverty estimates. Some of these were not publicly available and are now put up in a working paper, Incidence of Poor and Poverty Risk in India across NSS Regions for Rural and Urban areas, 2004-05 and 2009-10.

Abstract

The paper provides an estimate of incidence of poor and poverty risk in India across NSS regions for 2004-05 and 2009-10 in rural and urban areas. It raises concern on increasing poverty risk and also incidence of poor in some regions. These are not necessarily among the relatively worse-off CABMOUJ (kab mouj, referring to Chhattisgarh, Assam, Bihar, Madhya Pradesh, Odisha, Uttar Pradesh and Jharkhand) states, but they also happen to be in some of the better performing states like Andhra Pradesh, Delhi, Gujarat, Haryana and Punjab.

Extracts

...

Our NSS region wise analysis shows that poverty risk being greater than unity has reduced from 33 regions in 2004-05 to 26 regions in 2009-10 for rural areas and has increased from 35 regions in 2004-05 to 36 regions in 2009-10 for urban areas. Further, from the 26 regions with poverty risk greater than unity in 2009-10 for rural areas, 18 regions indicated an increase in poverty risk when compared with 2004-05 and poverty risk also showed an increase in another 10 regions. However, from the 36 regions with poverty risk greater than unity in 2009-10 for urban areas, 20 regions indicated an increase in poverty risk when compared with 2004-05 and poverty risk also showed an increase in another 18 regions. Relatively speaking, this indicates a tendency towards convergence or greater concentration of poverty in some regions for rural areas while there is an increase in spread across regions for urban areas. This may also explain the effect on account of population shift or migration from urban to rural areas in our understanding of decomposing poverty change (Mishra 2014). 

The CABMOUJ states, the North-East states, other hilly and mountainous regions and some smaller entities in rural areas do give the impression of the increasing incidence of poor and poverty risk in largely rainfed and other vulnerable regions. That apart a matter of increasing concern is that there are regions in well-performing states where one not only observes an increase in poverty risk, but also an increase in the incidence of poor. They are both rural and urban areas of Plains Southern in Gujarat (the concerns raised by Alagh (2014a, b)), Northern Punjab and Central Plains in West Bengal, and urban areas of Coastal Andhra Pradesh, Delhi, Eastern Haryana, Himachal Pradesh, Mountainous and Jhelum Valley in Jammu & Kashmir, and Eastern Maharashtra among others.

Concluding Remarks

With reductions in poverty, the public policy focus should be among regions and sub-groups with relatively higher incidences and higher poverty risk. At the same time, it should also be a matter of concern that the incidence of poor is increasing among a sub-group of populations. At first glance, they seem to be among some CABMOUJ states, but a deeper probing indicates that this seems to be emerging in North-East states and also in some regions among the better-off states like Andhra Pradesh, Delhi, Gujarat, Haryana and Punjab. This should be an important concern for public policy.

Other recent related work

Mishra, S. (2014). Decomposing Poverty Change: Deciphering Change in Total Population and Beyond, Review of Income and Wealth, forthcoming.

Pathak DC, Mishra S. (2011). Poverty Estimates in India: Old and New Methods, 2004-05. Working Paper No. WP-2011-015. Indira Gandhi Institute of Development Research, Mumbai.

Pathak DC, Mishra S. (2013). Poverty in India and Its Decompositions: A Critical Appraisal of the New Method, in S. Mahendra Dev (ed.) India Development Report 2012-13, Oxford University Press, New Delhi, pp.209-223 and 414-417, see pre-print version.

The link to the current paper is Incidence of Poor and Poverty Risk in India across NSS Regions for Rural and Urban areas, 2004-05 and 2009-10.

It is all in the line, our poverty line

The poverty estimates for 2009-10 in India by the Planning Commission has pegged the per capita per day poverty lines at 28.65 rupees in urban areas and 22.43 rupees in rural areas. Comparable poverty lines for 2004-05 are 19.29 rupees and 14.89 rupees respectively (see a note on poverty estimates for 2004-05).

Compared to 2004-05, the poverty line of 2009-10 has increased by 48.5 per cent and 50.6 per cent in urban and rural areas respectively.  During the same period, the Consumer Price Index for Urban Non-manual Employees (CPI-NME) and Agricultural Labourers (CPI-AL) increased by 48.8 per cent and 54.9 per cent respectively, as indicated in Key Indicators of Household Consumer Expenditure in India 2009-10 (see Table T7, p.17). An independent calculation of the CPI for Industrial Workers (CPI-IW) indicates that between the financial years of 2004-05 and 2009-10 CPI-IW increased by 45.2 per cent. 

The 2004-05 estimates were updated to June 2001 in the Planning Commission's affidavit to the Supreme Court in October 2011 and pegged at 32 rupees and 26 rupees (rather, 32.16 rupees and 26.03 rupees) for urban and rural areas respectively. If 2009-10 estimates are used to update for June 2011 then an inflation accounted for by CPI-IW and CPI-AL indicates a poverty line of  32.25 rupees and 25.35 rupees for urban and rural areas respectively. Thus, the poverty line for rural, if not for urban, India indicates a decline when compared to the affidavit. 

These all may give the impression that we might be close at the urban level but miss the bus at the rural level.  However, a closer look at CPI-IW for food between the two financial years indicates an increase of 58.7 per cent. A relatively higher increase for food in CPI-AL is also expected.

The food price inflation in recent years is a known fact, but it has implications for the methodology used by the Planning Commission. At a given level of basket with food comprising a higher share, the poverty line should have increased more than the general price increase. If this has not happened then as per 2009-10 estimates from the consumption expenditure the price increase in non-food items was higher or something still remains unexplained - the calculation of the poverty line is a black box.

Further, at the actual expenditure there could be a shift in the consumption basket from food to non-food items and this would not be because of choice, but because of compulsions arising out of relative prices. The estimates given by the Planning Commission, who have already done the calculations, do not elaborate on this point. But, then they reiterate that it not about a point, it is a line.

This reminds me of a story. A well meaning ruler appointed 'accomplished' persons as advisers who articulated improving well-being through hard work, dedication and competition. The ruler was so convinced that people who were lazy, casual and non-performers (including the poor, the sick, and the old) were asked to improve, failing which they were banished from the kingdom. The latter was followed in letter and spirit by able and objective administrators who were not swayed by human emotions. Despite that, every time the ruler reviewed the situation there always existed a pool of poor and sick and some young people who looked old. To top it all they were losers and were the ones who could not make use of all the incentives given (in the form of subsidies and tax deductions among others) to succeed. No one could provide any answers.

The ruler was worried and would go into contemplative moods about what went wrong in banishing the lazy hares. Suddenly, one fine day (we will narrate the incident some other time) it chanced upon the ruler that competition by design has to have a winner and a looser. No matter what you do you cannot do away with the looser. Then how does the ruler showcase his achievements among the people and more so among the comity of nations. Reversing the policy of banishing 'lazy hares' or not incentivizing winners is not possible. Thus, the ruler after deliberation with thinkers agreed to take a middle position 'competition that is inclusive'. The change in the slogan was mere cosmetic and things continued as usual except for the fact that the non-performers were not banished. They were allowed to stay and some efforts were made to rehabilitate some of them and in some cases a policy of 'give them fish without teaching them how to catch' was also followed.

The combination of policies did not have the desired effect. Poverty did not get reduced. Good minds were put to work to arrive at a method of calculating poverty. The method did away with the baggage of calorie consumption instead of adding nutritional concerns. It was made robust enough for other minds to have difficulty in comprehending and their efforts in replicating it will only result in they loosing their minds (only Alexander can cut the Gordian knot). Thus, all eyes were on the method. And lo and behold, it showed results.  One has to concede that it is all in the line, our poverty line.

(Revised 21 March and 24 March 2012. Thanks to Mitali Chatterjee, Kinnary Desai and Smit Gade for comments).