The Telegraph
 
 
IN TODAY'S PAPER
CITY NEWSLINES
 
 
ARCHIVES
Since 1st March, 1999
 
THE TELEGRAPH
 
 
Email This Page
THE POVERTY OF DATA
- Junk in the name of official data for the Northeast

For this column, let me begin with a table that I find very odd. This table shows poverty or head count ratios, that is, the percentage of population below the poverty line. And it is for the Northeast, with Sikkim included in my definition of the Northeast. Calculation of poverty ratios requires data and such data usually come from expenditure surveys conducted by the National Sample Survey. The NSS sometimes has thin or small samples. With small samples, data are regarded as not terribly reliable. Therefore, when poverty ratios are cited, they come from large sample NSS data. The problem is that the NSS does not have large samples every year. Typically, large samples occur with a frequency of five years. We had large samples in 1983-84, 1987-88, 1993-94 and 1999-2000. Using data from large samples, the planning commission now calculates poverty ratios for various states. These are separately calculated for rural and urban areas and then spliced together to obtain a combined figure.

The first panel in the graphic is now self-explanatory. This consists of poverty ratios computed by the planning commission for the Northeast, for the years 1983, 1993-94 and 1999-2000. I have reproduced the combined poverty ratios, since the separate rural/urban figures are irrelevant for purposes of the point I want to make. And these figures are from the planning commissionís national human development report 2001, published in 2002. Barring Mizoram, you notice very high levels of poverty in the region. While all-India poverty ratios have dropped by ten percentage points (from 36 per cent to 26 per cent) from 1993-94 to 1999-2000, the drop in the Northeast (even in Mizoram) has been less than proportionate. If these figures are accurate, together with undivided Bihar, undivided Madhya Pradesh, undivided Uttar Pradesh and Orissa, we need to seriously worry about poverty levels in the Northeast.

The figures in the second panel are what the planning commission expects poverty ratios to be in the Northeast in 2006-07. These are projections given in the tenth five year plan (2002-07) document. Notice that not only are the poverty declines from 1993-94 to 1999-2000 not expected to be replicated between 1999-2000 and 2006-07 (except perhaps in Sikkim), poverty ratios are actually expected to increase in Manipur and Mizoram. I find that exceedingly odd. This is stranger still because the tenth five year plan has statewise targets of gross state domestic product growth. The annual average growth rate in Arunachal Pradesh is expected to be 8 per cent, in Assam 6.2 per cent, in Manipur 6.5 per cent, in Meghalaya 6.3 per cent, in Mizoram 5.3 per cent, in Nagaland 5.6 per cent, in Sikkim 7.9 per cent and in Tripura 7.3 per cent. Letís ignore whether these growth rates are reasonable or not and they are also separately broken down into primary, secondary and tertiary sector components. The tenth five year overall target of 8 per cent annual gross domestic product is itself unreasonable. But the GSDP growth and poverty ratios are both targets. Are they mutually consistent, although the answer does depend somewhat on the composition of growth' I think not. With that kind of growth, how can you have such marginal declines in poverty, and in two instances, even an increase' Doesnít make sense and something is clearly wrong.

Perhaps a clue can be found in what the planning commission does with the poverty figures for the Northeast. Let me quote from the national human development report. ďPoverty Ratio of Assam is used for Sikkim, Arunachal Pradesh, Meghalaya, Mizoram, Manipur, Nagaland and Tripura.Ē Except for Assam, we donít even have poverty ratios for the other northeastern states, because the planning commission doesnít believe the NSS samples are large enough for those regions. Assamís poverty figures are being superimposed on the other states. This is misleading, to say the least. Why arenít poverty figures for the other states exactly the same as Assamís figures then' I donít know the exact methodology, but I believe something like the following goes on. For exact details, you will have to ask the planning commission. We have separate rural and urban poverty figures for Assam and we have a breakup of population in the other states according to the rural/urban division. Assamís rural poverty figures are then applied to that particular stateís rural population and ditto for the urban segment.

Because the rural/urban share varies from state to state, you will therefore have a different poverty ratio for the state from the Assam figure. No doubt the details are somewhat more esoteric, because there is a business of what population data to use (1991 or 2001 Census) and a question of matching the 77 NSS circles, but that in essence, is what goes on. Hence, if you look at the graphic again, you will find that with the exception of Mizoram, all the other poverty ratios are clustered around the Assam figure. We might as well junk the poverty figures for the other northeastern states and look at the Assam figure alone. So much for planning.

Life becomes even more bizarre when you consider what some independent researchers (outside the planning commission) have done. Laveesh Bhandari and Amaresh Dubeyís work is a case in point. They have used NSS data, but their own methodology, to work out poverty figures for the northeast. This is the third panel, based on NSS data for 1999-2000. The two sets of 1999-2000 figures are remarkable in their deviations, for every state other than Assam. Meghalaya, Mizoram and Nagaland particularly stand out. If these figures are correct, we are indeed talking about significant poverty in Assam and poverty of slightly different dimensions in Arunachal Pradesh, Manipur, Sikkim and Tripura. We donít need to worry about Meghalaya, Mizoram and Nagaland. Accordingly, the tenth planís figures for 2006-07 will also go for a six.

The problem probably is that our mindsets havenít changed. We tend to think of the Northeast as a homogeneous entity and ascribe all of Assamís characteristics to that region. I donít know who is right, the planning commission or Bhandari and Dubey. More accurately, I know that what the planning commission does is wrong. But I donít necessarily know that Bhandari and Dubey are right. What is obvious is that we have junk in the name of official data for the Northeast. Certainly for poverty and perhaps in other instances as well. To be able to plan, you need reliable data. Otherwise, it is garbage in and garbage out.

Top
Email This Page