It’s the alliances, stupid

The Congress engineered the right alliances in the right places and so won many more seats with fewer votes. The opposite was true of the BJP, argues Many reasons have been given in the media for the poll outcome — poor governance, neglect of the poor, neglect of rural areas. I believe that poor governance was key, because the swing, almost everywhere, was against the current or recently ousted incumbent state government. Misgovernance is so widespread that people vote against their respective state governments rather than for any national entities.

The Congress did very well, yet fared dreadfully in states where it ruled. It lost all but one seat in Kerala, all but two in Punjab, and all but one in Uttaranchal. Why so, when there was a pro-Congress wind? Because Congress rule (or shall we say misrule) led to a strong anti-incumbent vote. The NDA (and particularly the Telugu Desam Party) also suffered from anti-incumbency. Poor governance cuts across parties.

In the four states where governments were voted out in 2003 (Himachal Pradesh, Rajasthan, Madhya Pradesh, Chhattisgarh), the swing against the ousted party continued. Clearly, new governments have a honeymoon period of a year or two before they become the new hate objects.

But there was another important factor at work. What mattered critically was the management of alliances. Despite a revolutionary change in seats in Parliament, the changes in vote share were very small (see accompanying table). When the margin between victory and defeat is thin, you need the right alliances in the right places. The Congress forged better alliances, the NDA worse. And that made a huge difference.

The accompanying table is culled from the internet. The data are preliminary and subject to final revision, so do not treat them as accurate to the last decimal point. But they look good enough for coming to broad conclusions.

The most sensational finding is that the Congress party actually suffered a loss of 1.48% in vote share, from 28.3% in 1999 to 26.82% this time. So it was not the charisma of Rahul or appeal of Sonia that mattered. It was not even disgust with the NDA. Rather, the Congress engineered the right alliances in the right places, and so won many more seats with fewer votes.

The opposite was true of the BJP. Its own vote share fell by only 1.54%, roughly the same as for the Congress. Even the NDA vote as a whole fell by no more than 2.97%. The vote of parties constituting the Congress combine also declined by 1.26% compared with 1999. Yet the Congress combine managed a net gain of 66 seats, while the NDA suffered a loss of 89 seats. This cannot be explained by the difference in vote share. It is due above all to the quality of alliances. Good alliances did not give the Congress combine a huge swing in vote share. Rather, it gave the combine additional votes in constituencies where it mattered. And that made all the difference.

The most disastrous decision of the NDA was, of course, to switch partners in Tamil Nadu from the DMK combine to the AIADMK. But for this, the NDA would have won more seats than the Congress combine, and Atal Bihari Vajpaye might have made a bid to become prime minister for the third successive time!

Another BJP ally to fare poorly was the TDP in Andhra Pradesh. But the fall in TDP’s vote share was not dramatic (from 3.65% to 3.23%). However, the Congress combination included the TRS and Left parties. As an alliance, it proved far superior.

Overall the Congress gained 31 seats in the Lok Sabha and the BJP lost 42 seats. However, net gains or losses mask the phenomenal volatility in the fate of individual seats. The Congress lost almost two-thirds of the seats it had held in 1999 (65 out of 114). The BJP lost half the seats it had won last time (90 out of 180). But the Congress also gained 96 seats it had lost last time, and the BJP gained 48.

Swings of fortune are wildly volatile from the viewpoint of the individual candidate. The above data suggest that more than half the sitting legislators of the two main parties, whether they ended up on the winning or losing side, lost their seats. The data may exaggerate: some legislators may have shifted seats and won again. But overall, India has few safe seats for any but the top leaders (except of course in Marxist West Bengal). This fragility further underscores the need for the right alliances at the right time.

Uttar Pradesh and Bihar are exceptions to the general phenomenon of wild swings against incumbents. The fall-out of the Babri Masjid means that in these states Muslims stick to Yadav protectors, and the combination is stable. No ousting of incumbents here, despite misgovernance.

How do we improve governance? There are many ways, but they require a consensus across parties and that is difficult. All parties now host so many criminals and kickback experts that good governance will put hordes of legislators of all parties in jail. So, rather than improve governance, parties prefer to try and increase their share of the spoils of misgovernance. No way out of this shameful equilibrium is in sight.

What do you think?