ELECTION ANALYSIS

The Centre for the Study of Developing Societies (CSDS), New Delhi, has been undertaking post-poll surveys of a sample of 16,000 respondents since the general elections of 1996. According to the CSDS, the surveys are considered by political analysts as a n ideal source of data on voting behaviour. Unlike in the case of a pre-poll survey, there is no possibility here of the voter-respondent eventually changing his or her mind or not voting at all. And, unlike in the case of an exit poll, these surveys off er an opportunity to do a random selection of respondents and ask them a wide range of questions, the CSDS points out. Investigators of the CSDS have gone back to the same respondents after the 1996, 1998 and 1999 elections. According to the CSDS, this technique is considered the most precise and reliable way of measuring change over time, for it eliminates the possibili ty of change in two readings occurring simply as a result of differences in sample selection. The CSDS surveys have followed a rigorous technique of selecting a strictly random sample from the electoral rolls. As a result, the sample is known to match al l the known attributes of the Indian electorate within a margin of 1 per cent, the CSDS says. The sample involves 108 Lok Sabha constituencies spread over 20 States and Union Territories: it consists of 216 Assembly segments and 432 polling stations. The total sample size is 16,000, but since the survey does not make any substitution it normally achieves a target of between 9,000 and 10,000 respondents. In the current elections, three of the target constituencies of the CSDS sample will go to the polls only on October 28. The survey has hence been conducted in 105 Lok Sabha constituencies, covering 210 Assembly constituencies and 419 polling booths. The achieved sample is thus 9,069. The CSDS survey has been supported in part by Frontline.

Interpreting the mandate

YOGENDRA YADAV
SANJAY KUMAR

IF elections are only about who gets the right numbers to form the government, there is no doubt that the Bharatiya Janata Party and its numerous allies have emerged winners in the 13th Lok Sabha elections. Unlike on the previous two occasions, there is little uncertainty this time about who will form the government. For the first time in the last decade a pre-electoral alliance has crossed the majority mark. And it is not a wafer-thin majority. Once elections are held to the remaining six seats, the BJ P can hope for a situation in which no single ally of it will be able to bring down the government, not even the Telugu Desam Party (TDP). The new government will not be susceptible to blackmail by small allies. No wonder there is palpable relief in thos e circles within and outside the country for whom government stability is the predominant concern. So much so that no one stops to notice the fact that this is the fifth election in a row to have thrown up a 'hung parliament'.

But if election results are to be seen as anything more than the score in a numbers game in Parliament, if the larger political signs of a popular mandate are to be read, then the signals sent by the latest verdict are far from obvious. The winning coali tion has increased its parliamentary strength by 50 seats. But that is mainly because the BJP acquired new allies, rather than having won greater mass support. If the current score of the BJP and its allies is compared to what the current allies had in t he previous Lok Sabha, they have gained very little. In the previous Lok Sabha, they had 265 seats based on 42.3 per cent of the popular vote. The final picture for the new Lok Sabha shows that the current alliance has 296 seats based on 40.8 per cent of the popular vote. In other words, they have gained 34 seats and have lost 1.5 percentage points of the popular vote. The figures will surely dampen the enthusiasm of BJP strategists once the celebrations are over.

The BJP on its own has not been able to increase its tally in the Lok Sabha. It might, once the six remaining constituencies also go to the polls. But the fact remains that there is no substantial gain for the party at the end of a dream campaign. In ter ms of vote share too, the BJP has lost almost two percentage points, from 25.6 per cent last time to 23.7 per cent this time. It is true that this loss can be accounted for by the fact that the BJP contested 50 seats fewer than last time. But it cannot b e denied that the BJP's almost unstoppable upward march has come to a halt. However, there is a silver lining for the BJP: it has for the first time made inroads into new territories, thus taking another step towards becoming an all-India party. The BJP' s victories in Goa, Assam and the Andaman Islands may not contribute very much to its kitty, but they are crucial to its national ambitions. Elections 1999 have thus completed one phase of the BJP's rise to power. In this period the party underwent a geo graphical, social and political expansion. It has now reached a plateau of electoral success and is now subject to the normal cycle of success and failure. While the BJP's ascent to power at the Centre illustrates one face of this process, its defeat in Uttar Pradesh shows the other face.

The only real victors of this election seem to be some of the allies of the BJP, such as the TDP, the Biju Janata Dal(BJD) and the Trinamul Congress, which have added to their votes and seats. The TDP on its own has 3.7 per cent of the national vote shar e, besides 29 seats. It is too early to say whether the gains of its allies would necessarily strengthen the BJP. It is an important point, since the BJP's share of the combined seats of the ruling alliance has declined from 73 per cent at the time of th e constitution of the 12th Lok Sabha to 61 per cent at the beginning of the 13th Lok Sabha. The decline in its share of seats is bound to lead to a reduction in its clout within the reconstituted National Democratic Alliance (NDA).

FOR the Congress(I) too, the results are ambiguous. After successively losing its vote share in every round of elections since 1984, the Congress(I) seems to have staged something of a recovery. Its national vote share now stands at 28.5 per cent, an imp rovement of 2.7 percentage points over the last elections. If its allies are included, the gain is nearly five percentage points. The Congress(I) is back in power in Karnataka and Arunachal Pradesh and may well share power in Maharashtra. The recovery in Uttar Pradesh must be the best news the Congress(I) has had in quite a while. Much of the party's rise in vote share nationally is owing to the upswing in U.P.

Notwithstanding these consolations, the Congress(I) must confront the harsh reality that with a tally of 112, its contingent in the 13th Lok Sabha is the smallest ever. It is the third successive round of elections, and the second in which Sonia Gandhi c ampaigned, that the party has lost decisively. This defeat must be particularly hard to take, for less than a year ago it seemed that the 13th Lok Sabha was the Congress(I)'s for the asking. The BJP had lost three crucial Assembly elections - in Delhi, Rajasthan and Madhya Pradesh - and nationwide opinion polls had predicted swing of between six and seven percentage points against it. If that potential swing could not be translated into actual electoral victory, the Congress(I) has no one else to blame . The defeat in this round of elections has long-term implications for Sonia Gandhi's leadership, the party organisation and the very survival of the party in its present form.

THE Left Front has lost six seats and added a tiny fraction to its votes. Given its organisational problems in Kerala and the slowly rising dissatisfaction in West Bengal, its seems to have got away lightly. Even the slightest swing of votes away from it would have resulted in lots of red faces.

The Bahujan Samaj Party (BSP) looks like an obvious winner with a three-fold increase in its number of seats. While 14 parliamentary seats will no doubt give it a visibility that it did not have earlier, the fact that the BSP is shrinking into an only-U. P. phenomenon cannot be overlooked. Its nationwide vote share dropped by 0.4 percentage point, about 10 per cent of what it had last time, though there has been a 1.2 percentage-point gain in U.P. It lost its only seat outside U.P. and dropped votes in t he crucial State of Madhya Pradesh where it once hoped to become the third major player. Even in U.P., its gain in seats is owing to a very small swing of votes and tactical voting, a temporary phenomenon.

The same can be said of the Samajwadi Party (S.P.). Its remarkable achievement in this round of elections was to defy its obituaries once again. But its gain in seats has come despite a loss of 4.7 percentage points in its vote share in U.P. and one-fift h of its all-India votes. Like the BSP, it has disappeared from States other than U.P.

Sharad Pawar's Nationalist Congress Party (NCP) has proved to be only a Maharashtra phenomenon. While he showed that he can ensure the defeat of the official Congress, Sharad Pawar was unable to dislodge the Congress(I) the way Mamata Banerjee did in Wes t Bengal. With 22 per cent of the votes in Maharashtra, he trails the official Congress(I) by 12 percentage points.

In other words, while this round of elections has produced what looks like a coherent majority, much of it is owing to the vagaries of the electoral system. The attainment of a majority in Parliament does not imply a polarisation of voters in favour of t he winning combine. Nor does it indicate any effective bipolarisation of the party's political space. The seats occupied by the two leading alliances have not increased compared to the 1998 elections.

The third space, occupied by the various non-Congress, non-BJP formations, has not shrunk in any significant way. What has declined, of course, is the vision and organisational capacity of those wanting to create a Third Front in national politics. At th e State level, though, there appears to be a movement towards bipolar competition in Karnataka and Bihar. At the same time, Maharasthra and Tamil Nadu moved away from the two-party competition to triangular contests.

All this might create the impression that this verdict produced no difference. A quick look at the seat tally of the parties is likely to reinforce an impression of everything having come back to square one. In fact, nothing could be farther from the tr uth: such an impression is created because one looks only at the national aggregates. If one looks instead at the State level or below, one would discover evidence of enormous churning having taken place. Most of the States witnessed major political shif ts and reversals in this round of elections. So much so that one has to look for exceptions such as Gujarat, Kerala, West Bengal and Orissa, which did not see any major shifts in political preferences. Elsewhere it has been an election full of upsets. Si nce these upsets did not go in the same direction, they tended to cancel out one another, thus creating the impression of there being no change. An analysis of seat retention figures shows that of the 537 seats for which results are available, only 281 r eturned the same party that won in 1998. This ratio has been very low throughout the 1990s: 264 between 1991 and 1996 and 263 between 1996 and 1998. The figure will go down further for constituencies that returned the same member. Clearly, one is looking at the rejection of the incumbent at the ground level.

A detailed analysis of how seats changed hands between 1998 and 1999 shows that the BJP and its allies have been able to hold on to 157 of the 245 seats they held last time in these 537 seats. The Congress(I) and its allies retained only 53 of 161 seats. In other words, the BJP and its allies have retained two-thirds of the seats they won, while the Congress(I) holds only one-third. These figures indicate that the Indian elections continue to produce greater upheavals than elections in most other democr acies. Even parties on the upswing tend to lose a lot of seats. The BSP could retain only Mayawati's seat while the S.P. lost 10 of the 20 seats it won last time. As many as 50 of the 85 seats in U.P. changed hands this time.

An analysis of the verdict based on the urban or rural nature of the constituencies does not throw up any such clear pattern. The gains for the BJP and its allies have mainly come from rural constituencies this time, perhaps because the BJP had already r eached the saturation point in the urban constituencies. It is a comment on the expansion of the BJP that with its allies the party controls the majority of rural parliamentary constituencies as well. The Congress(I) gained votes both in rural and urban seats, but it incurred net losses in terms of seats in rural and semi-rural constituencies.

A comparison of the final picture of seats and vote share between reserved and general seats shows some interesting shifts. The Congress(I) and its allies, which made gains overall in terms of votes, have acutally lost about three percentage points in co nstituencies reserved for the Scheduled Tribes (S.T.). Their tally came down from 18 to eight in these constituencies. Since these seats are invariably in pockets where there is a concentration of the S.T. population, the decline indicates an erosion of S.T. support to the Congress(I). The BJP and its allies picked up nine of these seats and gained more than four percentage points in vote share. For the first time, the BJP and its allies won the majority of the seats reserved for the S.Ts. It is equally significant that the BJP alliance increased its vote share by 6.1 percentage points and seat share by 14 seats in the case of constituencies reserved for the Scheduled Castes (S.C.). This may not directly indicate any shift of S.C. voters towards the BJ P, but it does point to a social expansion of the BJP's base alongwith that of its allies. The Congress(I) got far fewer votes in the constituencies reserved for the S.Cs than in general constituencies or constituencies reserved for the S.Ts. Interesting ly, nine of the 14 BSP MPs were elected from non-reserved seats. Perhaps this is the first indication of a limited acceptance of the BSP among the sarva samaj that the party has begun speaking of.

DID the fact that voting was done in five phases influence the results? An analysis of the final results confirms the popular impression that it did. If the number of seats won by the BJP and its allies across different phases is compared, no clear patte rn emerges. They did better in the third and the fourth phases than in the second. But a pattern begins to emerge when the numbers of seats in terms of gains or losses are compared to the seats won in the same set of constituencies in 1998. The BJP and i ts allies gained by 26 and 21 seats in the first and second phases respectively. In the third phase there was a gain of only six seats, while there was a loss of three and six seats respectively in the fourth and fifth phases. An analysis of the phase-wi se swing confirms the picture of a steady decline of BJP gains in every subsequent phase of polling. However, the opposite is not quite true for the Congress(I) and its allies. They lost seats in the first two phases but did not quite gain in the last th ree. The gains went to others. In terms of vote, the gains were more or less evenly spread across all the phases, except the last one in which the Congress(I) and its allies did badly.

This clear evidence of the effect of the timing of the polls on the outcome shows that something of a Kargil effect cast its shadow over the first two phases. As the nation moved further into polling, that effect wore thin, and it was back to normal elec tions in their local settings. This finding raises the question whether the final outcome would have been different if the elections were held a little later, or for that matter a little earlier. It also questions the wisdom of staggering elections over such a long period.

THE evidence about the phase-wise voting also provides us a clue to understanding the big picture of the electoral verdict. It points to the possibility of a nationwide effect of the Kargil conflict and its aftermath on voting. The findings need to be co rroborated by survey analysis, but it seems that the BJP staged a recovery after its decline towards the end of last year. The manner of the dissolution of Parliament and the inability of the Opposition to form an alternative government, the Sharad Pawar -led split in the Congress(I), and the Kargil conflict made for conditions in which the BJP won back some of the votes it had lost. Yet all these factors put together did not create anything like a national wave of the kind the BJP hoped for. It just cre ated a favourable climate in which the BJP's claim to power was considered by the people. But the claim was finally settled on the terrain of State politics, the basic unit of political choice-making in this decade.

This combination of State and national-level trends meant that the BJP could get a verdict against incumbent non-BJP regimes, particularly in States such as Orissa and Bihar which has had unpopular governments in power for a long time. West Bengal and Ke rala, which have Left-led governments, and small States such as Arunachal Pradesh and Sikkim were the only exceptions.

The logic of incumbency, or rather poor governance, worked against the BJP and its allies too, as in U.P., Punjab and Karnataka, but only when the regime was very unpopular. In the rest of the States, a slightly favourable national climate and a redrawin g of alliances enabled the BJP to escape severe punishment. Its alliances in Tamil Nadu, Haryana and Andhra Pradesh helped it reap rich electoral harvest. By splitting the Congress(I), the NCP made the task of the BJP-Shiv Sena easy in Maharashtra.

In this sense the 1999 elections do not represent a fundamental break from the new kind of electoral system inaugurated in 1989. This phase of electoral politics put an end to the wave elections of the 1970s and 1980s. Elections are no longer decided sol ely on any national issue or personality. State-level politics has emerged as the primary arena of poltical choice-making. The fundamental question at the beginning of this round of elections was whether the 1999 polls would retain this character. One po ssibility for the BJP was to turn this into a national election in the sense in which the results of 1971 or 1984 was nationwide verdict. The other path was to accept the nature of political choice in the 1990s and work out a unique solution to the situa tion in each State. The BJP pursued both paths. The verdict makes it clear that the BJP did not succeed in changing the ground rules of poltical choice in the 1990s. It has won the game, at least for the time being, but by following rules that will not f avour it in the long run.