Behind the `shining India'

On a Guwahati roadside, waiting for the food their parents would bring. Seen against the widespread hunger, the high economic growth loses much of its sheen. - RITU RAJ KONWAR

THERE is a strong taboo against marring an auspicious occasion with the churlish remark. And the dawn of the new year has been, by all accounts, an exceptionally bright conjuncture for the Indian economy - an astonishingly high aggregate growth rate in the first half of fiscal year 2003-04, foreign exchange reserves bulging to a figure of $100 billion, and stockmarkets cresting new waves not on the backs of new companies of uncertain parentage and prospects, but on the basis of seemingly solid investor confidence in established Indian corporate giants.

In the midst of the welter of good news, a stubborn fact has most inconveniently intruded, though without drawing the kind of attention it may deserve. A recent report of the Food and Agricultural Organisation (FAO), the State of Food Insecurity in the World 2003, indicates that over a fifth of India's population still suffers from chronic hunger. Far removed from the rhetoric of a "shining India" where the "feel-good factor" is pervasive, India figures in the FAO's categorisation a little worse off than Honduras and a little better off than Mali. Indeed, India is one among 17 countries where the number of the undernourished decreased in the first half of the 1990s, before increasing in the second half to almost completely offset the gains of the earlier part of the decade. Tracking the incidence of hunger in India over three reference periods - 1990-92, 1995-97 and 1999-2001 - the FAO plots an initial decline from 214.5 million to 194.7 million, before a near total reversal of all gains pushed up the number of the undernourished to 213.7 million.

These figures make an eloquent point, simply because they pertain to a period prior to the debilitating drought of 2002, which caused a sharp drop in agricultural output and incomes, compelling the government to step in with direct livelihood support programmes that by all accounts fell far short of requirement. Outlays in employment and other forms of livelihood support were increased marginally in fiscal year 2002-03. But their impact has been marginal. Physical achievements in terms of man-days of employment generated, dwelling units established and habitations covered were not significantly different from earlier years.

Seen in this light, the news on the growth front loses much of its sheen. A 7.4 per cent growth in agriculture and allied sectors in the first half of 2003-04, coming on the back of a 3.5 per cent decline in the earlier year, just about restores levels of output and income to where they would be in a course uninterrupted by adversities of the weather. It cannot yet be deemed to have made a significant dent on the problem of poverty and undernourishment. The trend set since the mid-1990s, of an increasing incidence of hunger, cannot yet be considered reversed. This puts the FAO estimates firmly on a collision course with official figures on the incidence of poverty in India.

Government figures on poverty are based on sample surveys conducted at discrete points of time. It is now the official orthodoxy that between 1993-94 and 1999-2000, when the 50th and 55th rounds of the National Sample Survey (NSS) went into consumer expenditure levels and patterns across the country, the number of people living in poverty declined by the order of 60 million. In absolute numbers, the number of the poor in the latter period has been put at 260 million. This is considerably in excess of the FAO estimate of the incidence of undernourishment, when a degree of congruence should be expected. The NSS estimates on poverty are based on nutritional intake directly estimated from surveys, while the FAO arrives at its figures by considering the total food budget of a country - production and imports less exports - and adjusting for the degree of inequality in incomes and access to food.

In normal circumstances, the direct survey should be expected to yield the more reliable figures, were it not for a change of definition introduced in the 55th round of the NSS. The "recall period" canvassed for the households and individuals was changed from that used in the 50th round: where the standard practice had been to estimate consumption across the basket for a uniform 30-day reference period, the 55th round took both a one-week and a 30-day reference for food and a year-long period for all other items. In dealing with this incompatibility between the 50th and 55th rounds, the Planning Commission chose to use the 30-day reference period but observed that the "estimates may not be strictly comparable to earlier estimates of poverty".

Economists suggest that the use of a multiplicity of reference periods could have engendered confusion among the respondents and "contaminated" the data. Experiments with the seven-day recall period in earlier rounds of the NSS have shown that it tends to increase consumption of food items by about a third in comparison to the 30-day recall. Similarly, the one-year recall for items other than food yields figures for consumption across income groups that are much less unequal than those arising from the 30-day recall. Despite the use of the standard 30-day reference for poverty estimation, there have been suggestions that the concurrent canvassing of the two recall periods could have vitiated the data at source.

Some efforts to correct this error by re-examining the primary data have arrived at the finding that the decline in poverty was not as great as revealed in the survey. Economists at the World Bank and Delhi University have concluded rather, that the number of the poor could have decreased by half the magnitude indicated: 30 million as opposed to 60 million. Another effort by the economist Abhijit Sen from Jawaharlal Nehru University, Delhi, has suggested that even if the incidence of poverty did decline in the 1990s, it was only a relative decline and no greater in magnitude than in the period 1987 to 1994. The absolute number of the poor however, could not have fallen between 1993-94 and 1999-2000.

If the direction of change, rather than the magnitudes involved were to be considered, then this finding establishes a broad congruence with the FAO findings. The basic proposition that emerges is that the decade of liberalisation through the 1990s may have had an initial favourable impact in alleviating poverty. But this was reversed in quick time and initial gains eroded by the end of the decade. The years following could not, realistically, have had much positive impact. Far from pressing ahead with the reforms agenda in the belief that it has had a uniformly beneficial impact, this is clearly reason to pause and rethink priorities.

A LITTLE consideration would show how the official estimates on poverty in the 1990s have stood in splendid isolation from all other economic parameters. Indeed, to restore an element of logical consistency between the different sets of official figures, it is necessary to either reject the claim of a decline in poverty over the 1990s, or condemn as irrelevant most of the economic statistics collected in the country. For instance, the growth rate in agriculture has fallen below the rate of growth of population through the 1990s, while the population dependent on agriculture for their livelihood has remained almost constant in proportionate terms.

All surveys of employment that have been carried out indicate a rapid erosion of opportunities, a decline in self-employment and the growing "casualisation" of labour. And all through the 1990s, the per capita availability of cereals has been on a steady downward course: from 435 grams per day in 1990 to 385 in 2001. Availability of pulses has also gone down rapidly from 41 grams per day to 29 grams in the same period of time.

These parameters are crucial, since the FAO has found that successful strategies to cope with poverty and undernourishment have invariably involved a high rate of growth in agriculture, increasing per capita food availability and significant official spending on public health. None of these conditions has been met in India through the 1990s.

The official response to the reality of declining availability of cereals and pulses, has been to claim that a process of dietary diversification is under way, with growing prosperity and livelihood security propelling a shift towards fruits, vegetables and animal products. To bolster this claim, a new series of the National Accounts Statistics (NAS) was introduced in 1999, with the base year pegged at 1993-94, purportedly to correct - among other things - the long-running tendency for official statistics to under-state the production and consumption of fruits and vegetables.

These claims have been roundly rejected in most informed circles. As the economist Utsa Patnaik has argued: "It is a grave mistake to think that the absorption of grain per head falls with rising income: on the contrary, all empirical evidence shows that the absorption of foodgrains for all purposes, always goes up with rising per head income under normal conditions (this includes both the direct use plus the indirect consumption through conversion of grain as feed into animal products)."

With the more politically articulate constituencies being primed for early general elections with a flood of good news, it may be necessary now to hold up the actual record of 12 years of liberalisation to the light and examine its impact across the country. In particular, it may be time for the overdue recognition that the neglect of a sector that accounts for the livelihoods of 60 per cent of the population has consequences.

Even with the new series of national accounts, it is evident that investment in agriculture has fallen sharply in the 1990s. While private investment has stagnated as a proportion of gross domestic product, public investment has collapsed. This is an outcome of fiscal adjustment and the rapid erosion of the government revenues that is not evident to the beneficiaries of the massive tax concessions all through the 1990s.