THE verdict in Uttar Pradesh surprised the pollsters and political analysts, and it continues to surprise the psephologists even after the final figures are in. There is a paradox in the verdict: the BJP and its allies have secured 31.2 per cent of the votes, some nine percentage points down from last year but only a fraction below what they secured in the 1991 and 1996 parliamentary elections. Yet the BJP won only 29 seats (32, if the allies' seats are added) compared to between 50 and 60 seats it has secured in the past. For the Samajwadi Party, the result is even more paradoxical: it lost nearly five percentage points in terms of its vote share since the 1998 elections, but its seats tally has gone up from 20 to 26. The BSP recorded only a marginal increase in its vote share, from 20.9 to 22.1 per cent, but its seats tally rose from four to 14. The surprising element about the Congress(I) is that normally under the Indian electoral system one would not expect a party securing 17.3 per cent of the votes (including its allies) to win any seats. However, the Congress(I) romped home with 12 seats (including the tally of its ally, the Rashtriya Lok Dal). The first-past-the-post system that normally rewards the biggest party and punishes the rest does not seem to have worked that way in U.P. The party with the highest vote share, the BJP, won only 30 seats; the second and third highest-vote-share parties, the S.P. and the BSP, made major gains in terms of seats. In fact, even the fourth and lowest seat-share-party, the Congress(I), won a significant number of seats.
Conventionally, such a paradox would be explained by the Index of Opposition Unity (IOU). It was assumed that an increase in the IOU against the BJP would explain the sharp decline in its seats tally. But nothing of that sort took place in U.P. this time. There were no Opposition alliances, barring a minor seat adjustment between the Congress(I) and the RLD. In fact, the State-wide IOU came down from 53 per cent to 49 per cent between the 1998 and 1999 elections. It is not true to say that the BJP won big victories but lost narrowly. Its average margin of victory is smaller than its average margin of defeat. It is also not the case that the BJP accidentally lost a large number of seats by a very small margin. If anything, luck favoured the BJP: it won 12 seats by less than 10 per cent of the votes, while the S.P. and the BSP won only six seats each in this category. That is why U.P. remains a puzzle even after the verdict is fully known.
Part of the answer to this puzzle lies in the two other figures given in the table showing the vore share. This time two things worked against the BJP. In the past the BJP won over 50 seats with a fairly low vote share because its gap vis-a-vis the next most popular party, the Janata Dal or the S.P., used to be more than 10 percentage points. This time, like in 1996, the gap narrowed to only 7 percentage points. But unlike in the 1996 election, there was much greater resort to tactical voting at the constituency level. We find that this time the extent of constituency-level consolidation of votes against the BJP has been at its highest, more than 8 percentage points higher than last time. To sum up: the BJP's votes fell, and even though its main opponent also lost votes, the gap between the two narrowed; the gap was much lower at the constituency level where it actually matters, thanks to a higher level of polarisation of votes against the BJP at that level.
The other part of the answer lies in the region-wise break-up of the verdict. What looks like a difference of seven percentage points between the BJP and its main rival at the State level translated into a very different picture at the regional level. In the two small regions of Uttarakhand and Bundelkhand it trailed behind the Congress(I) and the BSP respectively in the matter of vote share. It lost all the four seats in Bundelkhand, while it won three out of four seats in Uttarakhand with narrow margins. In western U.P. the BJP's vote share was barely two percentage points over that of the Congress(I)-RLD combine, which won four seats there. In Poorvanchal, the area bordering Bihar, it trailed behind the S.P. by two percentage points. The BJP's vote share fell most sharply in this region while it could contain its seat loss to only four. It was ahead of the S.P. in Central and the Central East, but only by 1.5 and 2.5 percentage points respectively. Despite this slender lead the BJP lost as many as nine seats in the Central region where Sakshi Maharaj worked to defeat it.
As a result of all these factors, U.P. witnessed much closer triangular and quadrangular fights at the constituency level than the State-wise averages might indicate. Take for example the Ghatampur result: the BSP won this seat by securing 1,56,582 votes followed by the S.P. with 1,56,477 and then the BJP with 1,55,987 votes. In all, the results in 22 seats hinged on less than 10 per cent of the votes. As many as 50 seats changed hands in this election. Of the BJP's victories, 10 were in constituencies it had lost last time. Despite its overall gains, the S.P. lost 10 of the 20 seats it had won last time.
AN interesting thought experiment: what would have happened if the Assembly elections in U.P. were held simultaneously? This time the Election Commission has made Assembly segment-wise results available in record time and we can actually answer this question. The BJP would have performed much worse than in the Lok Sabha polls. It led in only 117 segments, and needed its allies' strength to surpass the S.P.'s figure of 128. The BSP would have improved its position and so would have the Congress(I). But the Congress(I) still has a lot of ground to cover before it can convert its extraordinary feat of making an 11 percentage point gain in the vote share into an increase in the number of seats. Unless the political alignments change, trends point to a hung Assembly in U.P. where every party is way off the majority mark.
Between the 1998 and 1999 elections most parties managed to retain about 80 per cent of their votes, with the notable exception of the S.P., which lost as much as 22 per cent of its votes to the Congress(I).
IN Uttar Pradesh, 49 per cent of the respondents agreed with the statement that one should vote the same way as the rest of the members of one's community do, which figure is a remarkable 14 percentage points higher than the national average. With the exception of the Congress(I), each party enjoyed almost unrivalled support from a particular Jati. The BJP collected 77 per cent of the Brahmin vote and 68 per cent of the Rajput vote. However, as it was also particularly popular with the other upper castes, these figures should not be treated as anomalous Jati occurrences but as part of a wider caste-based support base. However, in the cases of the BSP and the S.P. specific Jatis show voting behaviour markedly different from the other castes in their "set". Close to 80 per cent of the Yadavas voted for the S.P., which was over 50 percentage points more than the level of support that the remaining OBCs gave the party. Similarly, 74 per cent of the Jatavas voted for the BSP, compared to just 39 per cent of the other Scheduled Castes.
The Congress(I)'s expansion seems to have come from the lower castes. In 1998, its only major segment of support came from among non-Jatav S.Cs, of whom 42 per cent voted for it. Although the Congress(I) managed to retain support among the non-Jatav S.Cs even in 1999, its support increased noticeably also among the Jatavas and the other OBCs (non-Yadav), by 8 and 13 percentage points respectively. The Muslim vote is evenly split between the Congress(I) and the S.P., but the CSDS survey does not show any significant support for the BSP among Muslims.
The percentage of people who had made up their minds on whom to vote for a few days before the election was higher in U.P. at 44 per cent than the nationwide average of 36 per cent. The proportion of people who made up their minds before the campaign started was also significantly lower here, by 15 percentage points, than nationally. Dalits, Muslims and non-Yadav OBCs were by far the most likely groups to make up their mind late. This finding supports the hypothesis that these groups played the most important role in resorting to tactical voting against the BJP.
THE tables relating to class and education both serve to reinforce the fact that the BJP is very much the party of the elite. The very rich are more than three times as likely to vote the BJP than the very poor, and graduates and above are more than twice as likely to vote for the party than illiterates. However, the lower sections of society are far more evenly dispersed between the different parties.
The 1999 result in U.P. thus does not portend any early end to the political churning process; further volatility seems likely before any clear alignments can emerge. If anything, the revival of the Congress(I) and the unforeseen rise in the fortunes of the S.P. makes for a nearly unique situation of four-party system in U.P., a result that sharply diverges from the general trend in most other States towards a bipolar pattern.
