And, of course, matters are much worse considering only those crops whose data is firm. Although current estimates of foodgrain production for 2003-04 are a whopping 19% higher than in 2002-03, this is only 4% higher than in 1998-99; and the overall Index of Agricultural Production (IAP) has increased less than 3% between 1998-99 and 2003-04. The IAP per capita of rural population is thus actually down 6% from 1998-99 while per capita foodgrains output is down 5%.

Moreover, this is not all. The feature that distinguishes the last five years, and worries farmers most, is that farm prices have drifted relatively lower despite low output growth. With agricultural GDP at current prices shown 27.3% higher in 2003-04 than in 1998-99 against 10.4% increase at constant prices, agricultural prices increased 15.4%. But during the same period, non-agricultural GDP is estimated to have increased 39.2% in constant prices and 69.1% in current prices, implying 21.5% increase in non-agricultural prices. The GDP estimates imply at least 5% terms of trade loss for agriculture over 1998-2004.

Again, matters might be worse since official terms of trade indices from the Ministry of Agriculture had already shown 5% decline during 1998-2002, when terms of trade from GDP estimates had shown only 2% decline. But in any case, taking terms of trade into account, the purchasing power of agricultural incomes has actually declined in per capita terms during 1998-2004 even by the optimistic NAS figures; and this decline is by more than 10% if one combines official Indices of Agricultural Production and Terms of Trade. Thus, despite some diversification, overall agricultural outcomes have been poor, both on output and prices.

As far as output is concerned, this is because rates of growth of yields per acre have declined very sharply for almost all major crops. Considering all the crops covered in the IAP, yield growth had averaged 2.5% per annum during the 1980s, dropped to around 1.5% by 1998-99, and has averaged only about 0.5% during the last five years. Underlying this are some long-term trends, e.g. the plateau reached by ‘green revolution' technology and inadequate public investment since the 1980s. But matters have worsened recently on these and other fronts, such as extension and cooperative credit, because the Fifth Pay Commission bankrupted state governments.

Of course, some states have done better than others on output growth and the private sector has begun extension activities for some crops in some regions. But, with very few winners and many losers, agricultural production performance is not an aspect in which India is shining. However, the political fallout is not entirely clear since, with most states opposition ruled, the blame-game will be about Centre-state responsibility and response. And the debate is likely to be on whose incumbency hurt more, not what is to be done in the future.

On the price situation, however, the NDA is much more vulnerable. After all, it was the Centre that caved in at WTO and dismantled quotas prematurely, exposing Indian agriculture to the international price volatility that many farmers blame for their woes. And, although attractive support prices were announced, farmers in most regions did not actually get any price support. Moreover, contrary to recommendations of some committees to extend support operations, there are plans to halt support purchases altogether and link support prices to insurance – a misguided alternative already being piloted in some districts where Rabi arrivals will coincide with elections.

Turning to rural poverty, can this have reduced given the poor agricultural outcomes? Also, given the near vacuum on policy initiatives on agricultural production and price instability, what are the ideas to deal with this? The important aspect that needs noting in this context is that rural incomes are not all from agriculture and that fall in crop prices, although this hurts farmers, can improve conditions for those who buy food.

Chart >> Click to Enlarge

The ‘shining' campaign has put particular emphasis on the achievement on roads and thus of rural connectivity. And available NSS data does show fairly massive rebound (by over 40%) in rural non-agricultural employment, particularly construction, transport and trade, during 1997 to 2001-02, after a collapse during 1990-91 to 1997. Moreover, available data on wages and prices show that the consumer price index for rural labour increased less than other price indices after 1998-99 and that real wage rates have increased. Further, rural non-food consumption is increasing whether one goes by NSS or independent data. Thus, there are clear signs of rural dynamism if one goes beyond agriculture.

However, paralleling the poor agricultural performance, there is also evidence of stagnation in agricultural employment while the number of rural workers dependent on wage employment has increased very sharply (over 45% since 1991). NSS data in fact show a doubling of current unemployment among usual rural workers over the period 1997 to 2001-02. Furthermore, although the ‘shining' campaign has correctly identified the fairly large antyodaya grain supplies as a major achievement of the NDA, overall per capita cereals consumption (whether measured by availability or from NSS consumption estimates) has declined.

This coexistence of some dynamism in rural non-agriculture alongside declining agriculture makes it difficult to identify exactly what is happening to overall rural well-being. This is compounded because the only reliable source of information on this, the NSS consumption expenditure surveys, have become non-comparable. Pressure from ‘reformers', stung by NSS results that showed increased rural poverty from 1990-91 to 1998, caused a change in the nature of the questions asked in these surveys.

Some experimental surveys had shown that asking 365 day, rather than 30 day, questions on items such as clothing and durable goods throws up improved distribution and that asking 7 day questions on food returns 30% higher food consumption than 30 day questions. So, instead of the uniform 30 day recall used previously, the 1999-00 NSS questionnaire was changed to only 365 day for clothing and so on and both 7 and 30 day questions for food. Not surprisingly, this led to much lower measured rural poverty, by almost 50 million.

Since then much has been written on comparability of subsequent NSS data, and it is sufficient to note that it is now agreed that rural poverty did not decrease by anything like 50 million and that, although the proportion of poor is likely to have declined somewhat, the number of poor may actually have gone up. More importantly, it is now agreed that, properly interpreted, NSS data from 1993-94 onward show very large increase in inequalities (see chart) – across states, between rural and urban, and within urban areas. Moreover, although evidence on inequalities within rural areas is less clear, the trend towards lesser inequality that had begun in the mid-1970s with rural development programmes and extended public distribution has clearly halted.

But, consistent with the divergent indications on agriculture and rural non-agriculture discussed above, the most interesting indication from NSS data from 1993-94 to 2001-02 is that although cultivators and agricultural labourers have done badly, and the self-employed in non-agriculture have also not done well, those employed for wages or salary in non-agriculture have done much better. In fact, salaried employment, proximity to urban growth, and the ability to migrate emerge as main determinants, not only of the ability to avoid poverty but also as sources of relative affluence. Although cultivators still dominate among the richest 20% in rural areas, the proportion among the rural rich of affluent non-cultivators with urban connections has increased rapidly.

This should interest those attempting to guess political outcomes. Although caste and religion continue to dominate networks of patronage, the combination of poor agricultural outcomes and rapid urban growth in recent years has shifted the rural balance from traditional elites towards those who can offer urban access. It is here that the BJP, traditionally much weaker in rural than urban areas, can be the biggest gainer. Rural India is definitely not ‘shining'. But access to the shine that exists, which the ‘shining' campaign has made even more apparent, is now through networks where that party is stronger. The moot question is whether this shift in balance merely affects how people tell outsiders how they will vote or whether this is so deep that it will actually show up in how they finally vote?

On this, NDA strategists might like to mull over the following about rural India. It remains overwhelmingly agricultural, is about to go into elections in a year after severe drought, and is being bombarded with celebrations of ‘shining' without any solution being offered for what is definitely a longer-run crisis. There have in the past been only four years before the current one when national income has grown more than 8%: 1967-68, 1975-76, 1988-89 and 1996-97. These have all followed a year of drought and all except 1975-76, when Indira Gandhi declared an Emergency, were election years. The ruling party suffered losses every time.