Sept 20, 1999

Cover Story

Columns

Newsnotes

From the
Editor in Chief

Editorials

Eyecatchers

Voices

Business

Defence

Cinema

Books

Offtrack

Bodyline

Centrestage

Issue Contents

Judging the outcome isn't all that hazardous if past results are are monitored. The campaign provides the chemistry, the rest is organised number-crunching. Deputy Editor Swapan Dasgupta reads meanings in the jumble of statistics.

Is there already a Winner?

Triangular and quadrangular contests make psephology a high-risk venture. The 1998 election made life a little easier by focusing on three broad formations. This year, for the first time since 1977, it has been made even more simple. The battle is between the Congress-led alliance and the BJP-led alliance. Both sides have added and discarded some of their pre-1998 allies.

The starting point, therefore, is not the 1998 results but those results modified to take into account subsequent alliances, splits and mergers. Thus, the BJP alliance starts not with its 1998 vote share of 37 per cent but with a tally of 41.4 per cent. The Congress-led alliance kicks off with 31.8 per cent, 2.4 per cent more than actually secured in 1998.

In terms of seats, the effects are more dramatic. The consolidation of the anti-Congress vote in Andhra Pradesh, Karnataka and Bihar has proved advantageous for the BJP alliance, as has Sharad Pawar's breaking away from the Congress in Maharashtra. The merger of the RJP with the Congress has brought in the anti-BJP votes under one roof in Gujarat, and the Akali Dal split has also helped the Congress. Assuming that Pawar's NCP takes away one-fourth of the Congress vote in Maharashtra, G.S. Tohra takes away 15 per cent of the BJP-Akali Dal vote in Punjab, half the JD vote in Karnataka goes to the BJP-JD(U) but transfers in entirety in Bihar and Orissa and the turnaround of the DMK and AIADMK in Tamil Nadu follows their respective 1998 tally, the BJP alliance starts this election with an advantage.

Does low Turnout matter?
For parties, the real trick is to ensure that supporters are sufficiently motivated to come out and actually vote. This is easier said than done, though grassroots organisation helps. With frequent elections and restrictions on campaigning, the carnival atmosphere associated with earlier elections has disappeared. No wonder this election didn't witness a high turnout in the first round. Depending on who abstained, this can affect the results. If the abstention rate is higher among the middle classes, the BJP is inevitably the loser. Likewise, if the poor stay away, the Congress and parties like the BSP are handicapped. There is no pattern but surveys show that the poor take their vote more seriously than the better-off. They seem to be less distracted by long weekends and cricket matches. But don't jump to conclusions: the wave elections of 1971 and 1980 were accompanied by low turnout.

Can the BJP cross 272?
For the BJP, every election is accompanied by new expectations. In 1998, the two common questions were: Can the BJP cross the Vindhyas? Can it break the 20 per cent barrier? The party succeeded on both counts. But this time the question, "Can the BJP secure a majority on its own?", may have a negative answer. Opinion polls give the BJP alliance a 4 per cent swing. The party is contesting some 330 seats, against the 388 it fought in 1998. In an ideal situation for the party, after taking into account the alliances and splits, approximately 243 seats can be described as potential targets if the swing is uniform. Unfortunately, India rarely votes as one. The party could gain from the Congress split in Maharashtra, from the JD's incremental votes in Karnataka and Bihar and the TDP alliance in Andhra Pradesh but could lose out from the opposition consolidation in Gujarat and the Akali split in Punjab. The Uttar Pradesh results will depend not so much on how many votes it loses, but who gains. The realistic question for the BJP now is: Can it cross 200? For a party that won just two seats 15 years ago, that's not too bad.

Can the Congress recover?
Conventional wisdom decrees that 1998 was the worst year for the Congress and that it can't do any worse. Judging by atmospherics that's certainly true, though some opinion polls say otherwise. Actually, it's all a question of alliance arithmetic. The Congress has hesitantly entered into seat-sharing deals with the AIADMK in Tamil Nadu and the RJD in Bihar but its heart isn't into coalitions. In psephological terms, this means that the Congress has looked on passively while the BJP has systematically worked on improving the Index of Opposition Unity (IOU) -- a yardstick for judging the extent of anti-Congress consolidation. In the three crucial states of Karnataka, Andhra Pradesh and Bihar -- states which gave a total of 52 seats to its alliance -- a rising IOU has put the Congress alliance at an initial disadvantage. Statistics suggest that a 1 per cent change in the IOU changes the margin by 0.64 per cent. Moreover, it requires an approximately 3.23 per cent IOU change to offset a 1 per cent swing.

Taking into account all re-alignments, this means that the Congress begins the election with margin erosions of 15 per cent in Andhra Pradesh and 7 per cent in Karnataka and Bihar. In Maharashtra, if the NCP eats into 25 per cent of its vote, the BJP-Shiv Sena tally could go up to 35 from 10.

To survive in this election, the Congress doesn't need a gentle swing in its favour, it needs a tidal wave against the BJP.

What about Uttar Pradesh
Once again, Uttar Pradesh is The great imponderable of this election. Opinion polls are unanimous that the Congress will benefit from a significant swing in its favour. But the party starts with a ridiculously low base of 7 per cent which means even a huge swing of 10 per cent benefits it nominally. The associated question is: who loses the most? The greater the BJP loss, the more the gains for the BSP. Likewise, the more the SP slips, the more the BJP gains. To optimise the BJP's loss of seats, the Congress gains will have to be limited to specific regions and constituencies. For example, even if there is a 25 per cent uniform swing to the Congress -- the same gain registered by the BJP in the Ram wave of 1991 -- the party will not gain more than 14 seats. In other words, to achieve a repeat of 1993, when the BJP was outmanoeuvred by the SP-BSP alliance, there will have to be a lot of tactical voting by all communities. The mere shift in Muslim votes will not affect the BJP's fortunes materially, although it could cripple the sp. Unstructured anti-incumbency could lead to the Congress securing a better percentage and gaining in confidence. The BJP would end up losing votes but gaining seats. Such are the vagaries of quadrangular contests in a state that defies the realignment game.

How do SCs, STs, Muslims vote?
Nearly all opinion and exit polls have shown that the Congress is the party most favoured by the Scheduled Castes, Scheduled Tribes and Muslims. The BJP, while it has support among tribal communities in south Bihar, Madhya Pradesh and Orissa, is clearly at a disadvantage among these groups. According to the 1998 India Today-CSDS exit poll, 28 per cent of Scheduled Castes and 39 per cent of Muslims voted for the Congress and its allies; the BJP alliances secured 16 per cent and 7 per cent votes of these communities.

Despite this obvious imbalance, the voting pattern in sc and ST seats and Muslim-dominated constituencies indicates a good showing by the BJP. This mismatch is easily explained. Reserved constituencies are not necessarily dominated by Dalits and tribals. In fact, the number of these communities may actually be below 30 per cent. Interestingly BSP supremo Kanshi Ram never contests from reserved constituencies despite Dalits making up the core of his party. In Muslim-dominated seats, particularly those with a record of communal tension, the BJP has a loyal Hindu vote. If Muslim votes get divided, the BJP candidate often wins in triangular contests. In 1998, for example, the BJP's Mukhtar Abbas Naqvi won from Rampur (Uttar Pradesh) -- where the Muslim population is 42 per cent -- with only 32.8 per cent votes. If there is a Muslim swing to the Congress this year in Uttar Pradesh from the SP, this result may be repeated elsewhere.

Is the BJP an Urban phenomenon?
Nearly one in every four Indian lives in a city or a town. Yet, there are not more than 46 seats that can be classified as truly urban constituencies. There are other seats that have large urban components mingling with rural segments. The delimitation of constituencies hasn't taken place for 22 years and is long overdue. When the new boundaries are drawn, it is likely that the number of purely urban constituencies will register a sharp increase and, maybe, touch 100.

Ironically, such a process won't be entirely to the BJP's advantage. Compared to other parties, the BJP has undoubtedly more supporters in urban clusters than its rivals. This bias comes into full play in mixed constituencies. In Uttar Pradesh, Bihar, Madhya Pradesh, Karnataka and Haryana, BJP candidates normally lead by huge margins in urban segments and trail in the rural areas. But their urban margins are often so large that the rural losses are offset. In purely urban constituencies, this would lead to considerable "wasted" votes.

However, elections can't be won unless parties have both rural and urban support. Rather than describe Hindu nationalism as an urban phenomenon, it is more accurate to pin it down as an upsurge of the middle classes, both urban and rural. The latest India Today-Insight (ORG-MARG) poll shows among rural voters, 40.5 per cent prefer the BJP alliance and 31.5 per cent the Congress alliance. Among urbanites, the corresponding figures are 47 per cent and 32.5 per cent.

Is the Red Fort Impregnable?
A volatile electorate has both accepted and rejected all the parties in the states. West Bengal is the only exception. Here the Left Front has prevailed since 1977. Its worst performance was in 1984 when the Congress won 16 of the 42 seats. In 1998, the Left was surprised by the performance of the BJP-Trinamool Congress (TC) combine in Calcutta and its adjoining urban cluster. However, the BJP-TC can at best remain an irritant unless there is a consolidation of anti-Left votes. In 1998, the Congress, despite losing its deposit in 26 constituencies, secured 15 per cent of the votes. To defeat the Left, the BJP-TC would need to get 8 per cent votes from the Congress and secure a 4 per cent anti-Left swing.

Are Opinion Polls credible?
It has become customary for leaders and parties to believe and disbelieve opinion polls according to convenience. But are opinion polls worth taking seriously? The answer is mixed. In a country of India's size and complexity, it is impossible to be dead right. More so because there are imponderables like turnout and late swings. Also, while polls can broadly gauge the trends, the conversion of votes into seats is a difficult operation.

In assessing polls (we assume they are genuine), certain broad rules are worth remembering. First, polls are more accurate in reading the big picture. Be extremely wary of constituency polls and more kindly disposed towards overall national tallies. Even exit polls can go wrong in giving a correct state tally. Second, treat polls as indicative. They can tell you of the direction of the swing but they cannot always gauge its magnitude. Finally, beware of the loose use of terms. Psephology is the study of election data and is a branch of politics. Polling and forecasting are complex statistical exercises. In India, some charlatans have obliterated the distinction.

TRIVIA
> Atal Bihari Vajpayee is the only person to have won from four different states -- Uttar Pradesh, Madhya Pradesh, Delhi and Gujarat.
> Vajpayee enjoys the longest parliamentary vintage. He first won in 1957.
>The only seat never won by the Congress is Manjeri in Kerala's Malappuram district. It has been represented by the Muslim League since 1952.
> P.V. Narasimha Rao holds the record for the highest victory margin. He won the Nandyal by-election in 1991 by 5,80,297 votes.
> In 1989, K. Ramakrishna of the Congress won the Anakapalli seat in Andhra Pradesh with a nine-vote margin. In 1996, Satyajit Rao Gaekwad (Congress) prevailed in Baroda (Gujarat) by 17 votes.
> If re-elected, Kagapathi Pradhani of the Congress would have won from Nowrangpur (Orissa) 10 times in succession.
> If re-elected, it will be Indrajit Gupta's 11th term. He entered the Lok Sabha in 1960 and has lost only once since.
> The first general election began on October 25, 1951 and concluded on February 21, 1952 -- a total of 119 days. The second general election spanned 105 days, the third 110. In 1967, the exercise was over in 13 days and in 1980, four.
> The most resounding mandate was won by the Congress in 1971 in Himachal Pradesh (four seats). It secured 75.8 per cent of the votes. In the same election, it also secured 70.9 per cent of the votes in Karnataka (27 seats). In 1977, the Janata Party secured 70.4 per cent of the vote in Haryana (10 seats).
> Electorate in 1951: 17 crore. In 1999: 70 crore

Back | Next