Fuzzy logic, part two

The finance minister8217;s reply to the general debate on the Budget was as pathetic as the performance of the economy. The key question be...

The finance minister8217;s reply to the general debate on the Budget was as pathetic as the performance of the economy. The key question before the nation a decade after reforms has to do with the performance of the economy. The finance minister was honest to admit that the momentum has gone out of growth, but asked why the terrible outcome of the last two years alone should be considered; why not the far better record of his first two years? It seemed to have escaped his attention that it was the Budget for 2002-03 that was under discussion and attention must necessarily focus on present failure rather than past achievement.

And, in any case, what past achievement? In both 1998-99 and 1999-2000, which were his better years, growth was far below the trajectory of 7-8 per cent recorded in the mid-8217;90s. Moreover, Sinha8217;s first year recorded plus 6 per cent growth only because Chidambaram8217;s 8216;8216;miracle8217;8217; budget had depressed growth to below 5 per cent in 1997-98 8212; after reaching a peak of plus 8 per cent the previous year, bettered only once earlier: Rajiv Gandhi8217;s spectacular plus 10 per cent in 1988-89, the only time the economy has grown in double digit figures.

What is one to make of a finance minister who actually praises himself by comparing his growth rate in 2000-2002 8212; the slowest since the Fifth Plan ended more than 20 years ago 8212; to growth rates between the First and Fourth Plans, and in doing so fails to mention that the so-called 8216;8216;Hindu8217;8217; rate of growth in the first quarter century of Independence was five times higher than average annual growth in the pre-Independence, first half of the 20th century?

An immediate consequence of the unprecedented acceleration of growth in the eighties was a sharp fall in poverty, reducing the percentage of those below the poverty line to 25 per cent, as per the National Accounting System NAS. Since the NAS percentage seemed too rosy for policy purposes and too complimentary to his predecessors8217; management of the economy, V.P. Singh8217;s Planning Commission decided that for estimating poverty, it was not the NAS figures but the figures given by the National Sample Survey NSS which would be taken as correct. Thus, since the last decade, while all other statistics are furnished by the NAS, the poverty estimates alone are taken from the NSS. This gave V.P. Singh and his cohorts the immense satisfaction of telling the country that poverty levels were not a quarter but actually more than a third of the population 8212; some 36 per cent.

Then came reforms 8212; and unfortunately for the reforms lobby which cuts across political parties the first quinquennial NSS after reforms showed poverty to be stagnating or even perhaps increasing, notwithstanding the initial fillip which reforms had given to growth. That trend was confirmed by each successive annual survey. This contradicted the NAS data 8212; which was hardly surprising since NAS had shown poverty to be declining in the eighties, and it was to cover up this awkward political fact that V.P. Singh and his successors resorted to the ruse of an altogether separate survey on consumption patterns of the poor to discredit the Gandhis, mother and son.

When the ruse backfired on the reforms lobby with NSS survey after NSS survey showing that poverty levels were stagnating year after year, Statistics Minister Arun Shourie put his fertile head together with Praveen Bisaria, head of the NSS Organisation, and came up with the idea of changing the basis on which the NSS quizzes the poor about their consumption patterns. The basis of questioning was expanded to a 30-day recall 8212; that is, the poor were asked what they ate 30 days ago instead of being asked what they spent on food seven days ago. Whether a poor illiterate is more accurately able to recall what his family ate a month ago or a week ago is the kind of arcane argument I prefer to leave to statisticians. The point is that even if the 30-day recall is more compatible with other NAS statistics than what is thrown up by remembering the previous week, the latest quinquennial survey is not comparable to the previous quinquennial or annual surveys since the methodology of the estimates has been changed. Oranges cannot be compared to apples. Moreover, poverty is not an independent variable. If all indices of growth are in decline 8212; GDP, agriculture, industry, employment 8212; how can poverty be declining, and that too at the fastest rate ever?

I, therefore, asked the statistics minister whether the methodology of the latest survey had been applied to the previous survey; or, alternatively, the methodology of the previous survey had been applied to the latest; and if neither were feasible, what was the Planning Commission8217;s estimate of the degree to which the latest statistics had been rendered non-comparable to the previous survey. On all three queries, I drew a negative response from the minister: no, the one methodology had not been applied to the other; and, no, the Planning Commission neither knew nor intended to find out the extent to which the earlier results were non-comparable with the latest data.

Story continues below this ad

I brought this to the attention of the finance minister in the course of the general debate on the Budget. His reply was to indignantly point out that Bisaria had died and it amounted to defaming the distinguished dead to question statistical compatibilities! His particular grouse was that I had described this as 8216;8216;fudging the figures8217;8217;. What would you call it? It is jugglery of this order which is discrediting the independence of the NSSO and making our country, once a model for statistical accuracy in the developing world, a laughing stock among professional statisticians and economists. I quoted a number of respected commentators from the World Bank to The Economic and Political Weekly 8212; to underline the damage being done by twisting statistics to suit particular ideological or political ends.

Some may think these unprocessed poverty statistics serve the ruling party8217;s ends. But the truth will out. Clearly, no one in Delhi, Uttaranchal, Punjab, Uttar Pradesh or Manipur 8212; indeed, no one anywhere outside the fanatical mobs mobilised by the Sangh Parivar in Gujarat and elsewhere, and possibly not even them 8212; believes for a moment the drivel on poverty statistics being purveyed by this government of Marie Antoinettes.