Political competition between the Left and the Congress in the forthcoming elections in West Bengal and Kerala is leading the Manmohan Singh-Sonia Gandhi government to stop reform dead in its tracks.

This is because more economic reform could lead either to lose.

But the Left stands to lose more. If that happens, it may pull down the United Progressive Alliance government. The government has thus decided to put self interest before national interest.

So the question arises, does political competition lead to as good an outcome as economic competition? The question is all the more relevant because in recent years there has also been a persistent comparison between India and China.

The latter's extraordinary economic performance has been attributed to its government not having to cope with political competition.

So suggestions are made from time to time, in a suitably low voice, that political competition is bad for growth. Going by the amazing behaviour of the Left, this does appear true. But is it really?

In a recent paper*, Timothy Besley, Torsten Persson and Daniel Sturm argue that it is not. They say, instead, that "political competition has quantitatively important effects on state income growth and state policies."

Like all good economists seeking a stairway to heaven, they have constructed a model. It seeks to "explain why political competition may enhance economic performance and uses the United States as a testing ground for the model's implications."

The model seeks to show how political competition affects policy and economic growth via the "quality of politicians".

It divides society into two groups: those who live off the land and those who live off things other than land.

Since the model is about the US, the chief policymaker is an elected governor. In India, that would translate into the Cabinet.

Vote banks exist and electoral advantage accrues from a surplus of committed voters. These voters are committed because their party adopts non-pliable stances.

In the US, these stances pertain to non-economic issues such as race or religion.

In India, the counterparts are Hindutva for the BJP and the protection of the PSUs for the Left. The Congress has no committed voters because it has no non-pliable stances.

This advantage, say the authors, reduces the incentive to appeal to "swing voters".

Now comes the zinger. The authors have plotted "the log of income per capita in each of the Southern states relative to the entire US against political competition in the state relative to the entire US, again using averages for each decade from 1930 to 2000."

What they get is rule of enormous implications. "The regression line has a slope of unity, suggesting that each percentage point of (relative) political competition is associated with a percentage point of (relative) income."

They say that this is not a mere coincidence but the result of a causal mechanism. In the absence of political competition, parties get supported on the basis of how far they can protect existing vested interests (the Indian communists again).

This support gives that party "a large advantage, blunting the responsiveness to voters over economic issues."

What happens then holds the key, as we can see from the Indian experience, not just in the case of the Left but the whole range of political parties.

"This allows narrow economic interests, antithetical to growth, to capture the political process�spilling over into the selection of low-quality candidates who are more susceptible to influence by special interests."

Criminals in politics are, I suppose, just an extreme case.

The model goes on to make predictions about growth and governance as they emanate from the degree of political competition.

The bottom line is that more political competition is as good for growth as more commercial competition.

We can ask why, in spite of so much political competition, the Indian growth rate is so low.

The answer is that, first, only the official growth rate is low; second, when political competition was less between 1951 and 1991, it was even lower; and third, perhaps most importantly, it would be higher if the size of committed voters relative to total vote share was lower.

*Political Competition and Economic Performance: Theory and Evidence from the United States, NBER Working Paper No. 11484, July 2005.