LS polls: Why vote share won’t mean seats

All States’ surveys rely on vote share and translate them into seat projections.  Is it that precise?

Look at seats won and vote shares of two fronts in Kerala since 1977 (as delimitation of 20 constituencies started then). The vote share gap has varied from 10 to 2.5 percent. In 1977, the difference in vote share was 10 percent and the winning Congress-CPI-RSP-Muslim League-Kerala Congress- NDP-SRP front took all 20 seats, leaving none for the CPM-led front.

At the national level, Congress was routed, with the first non-Congress government, led by Janata Party, coming to power.
LS polls 1980 took place after a substantial realignment of political forces at the State-level. At the national level, Congress, led by Indira Gandhi, staged a comeback with 351 out of 544 seats in LS.

But in Kerala, LDF along with Congress, led by A. K. Antony, and Kerala Congress (Mani) group, notched 12 seats as against 8 by Congress (I)-led UDF. Vote share gap was 10.5 percent but the difference in seats was only four.

These two elections proved Kerala contrarian to the national trend but also sent a message that vote shares always do not translate into number of seats.
The counter-contrarian trend was visible in 1984, with Kerala voting in tandem with rest of India post-Indira assassination, which gave Rajiv Gandhi an unprecedented mandate on the crest of a sympathy wave. The vote share gap was 4.4 percent and the gap in seats won was 14 (UDF 17, LDF 3).

There was a quick reversal in 1989, when contrarian trend became visible as Kerala gave Congress-led front 17 out of 20 seats when Congress was voted out of power in national elections.

Elections 1991 saw Kerala being counter-contrarian again courtesy another sympathy wave, the vote share gap rose to 10.6 percent but the gap in seats won got reduced to 12.

From 1996 onwards, the election results have represented counter-contrarian trends as Congress as a national party did not get an overwhelming advantage in the LS elections of 1996, 1998 and 1999. In these three elections the vote share gap narrowed down to 2.5, 3.6 and 2.1 respectively whereas gap in seats won was 0, 2 and 2 respectively.  

In 2004, the infighting in UDF gave unprecedented advantage to LDF and in 2009 factionalism in CPM gave the same to the UDF.  The vote share gap was 5.6 in 2004 with gap in seats won was 17, whereas in 2009, vote share gap was 7.7 and the gap in seats won was 12.

Our electoral history tells us that arithmetic translation of vote share to seat projections can be fallacious A constituency-wise closer look and sixth sense together may give a clearer view.

But what can be stated clearly is that voter needs to be cautious of any survey despite its analytical rigour and statistical robustness.

A look at the statistics so far:

(The author is a researcher in issues of federal polity in India)