The science behind political coalitions
Alliance-building depends largely on size, winnability and the parties’ ideological cohesion
“Politics is the art of the possible.” This phrase is often invoked by political spokespersons while rationalising the latest coalition their party has cobbled together. Over the years, Indian democracy has thrown up many examples of natural — and not-so-natural — coalitions.
In the recent parliamentary election, three main coalitions, National Democratic Alliance (NDA), United Progressive Alliance (UPA) and Mahagathbandhan (grand alliance), are trying their luck. All three have their quirks. A couple of years ago, few would have predicted an alliance between the Samajwadi Party (SP) and the Bahujan Samaj Party (BSP). Recently, a more plausible coalition between the Aam Aadmi Party (AAP) and the Congress has fallen apart after hectic negotiations.
What prompts different parties to come together? Which coalitions are more likely to emerge than others? Is it about ideological cohesion or winnability or a combination of the two? Is it possible to say anything concrete about these issues at all?
Political scientists and economists have wrestled with these questions for long. In the Indian political context, coalitions can take broadly three forms: pre-election coalition of informal voting blocks (think of various caste equations analysts talk about); pre-election coalition of parties for contesting elections; and post-election coalition of parties for government formation. Of the three, government formation is the most structured and analytically more tractable situation. Consequently, it can often be predicted with surprising accuracy, though insights carry over to other cases as well.
Nobel Laureate economist Robert Aumann is well known for presenting many interesting ideas in this field. Specifically, Aumann has been predicting which coalition will get formed in the Israeli parliament with some success. Even with a few parties, the number of coalitions which can be formed is large. So how can a successful coalition be predicted in advance?
Roughly speaking, Aumann’s idea is that political parties enter the bargaining for coalition with certain specific goals in mind. They wish to maximise the power of the coalition they are a part of. At the same time, they also wish to maximise their own power within the coalition. It is the fine balance between these two goals that determines their coalitional choices. As it turns out, this simple idea has some surprising consequences, and it can rule out many potential coalitions.
Aumann’s analysis implies that a successful coalition should not be too small, otherwise it won’t be powerful and won’t achieve much. This much is intuitive. But a successful coalition cannot be too large either, otherwise its members will lose control over the affairs of their coalition, and they will have negligible impact on the outcomes. Consequently, their bargaining power within the coalition will be negligible, and they will have less to gain by staying in the coalition. At some point, they will leave, bringing down the effective size of the coalition.
Think of the coalition between the SP and the BSP in Uttar Pradesh. Why did they separate two and half decades ago and why have they joined hands now? Political commentators may say it was due to the difference in temperament of political personalities and so on, but a look at the voting shares of both the parties over the years suggests a different conclusion. Between the 1990s and 2014, both parties commanded large vote shares. Their combined strength would have swept the poll, but at the same time would have made the coalition too large and unwieldy. Such a coalition would have brought SP/BSP to power, but almost certainly couldn’t have accommodated the political ambitions of their senior and middle level leadership.
Following the emergence of Narendra Modi, there was a tectonic shift in terms of voting loyalties in Uttar Pradesh. Both the SP and the BSP lost their vote shares in a big way. Now, they were in the situation where a coalition became a tantalising possibility. Consequently, both set aside their political differences and the alliance was formed.
A similar reasoning applies to the government formation in Karnataka. When the result of the recent assembly elections was announced, there was a hung assembly. The numbers were such that either the BJP or the Congress could have formed an alliance with JD(S). But a BJP/JD(S) alliance would have been unmanageably large, possibly leading to internal disaffection. A Congres/JD(S) alliance was smaller, more cohesive and made more political sense.
Insightful as Aumann’s analysis is, two additional factors are relevant in the context of coalition formation in India. One is ideology and the second is the forward looking behaviour of the parties. Although a coalition between parties from different ends of the political spectrum is not unheard of (think of the BJP-PDP experiment as a recent example), such coalitions are politically costly and difficult to justify to the core supporters. As such, ideologically incompatible coalitions are formed less frequently.
Finally, political parties tend to think about the future as well. If the terms of proposed arrangement are such that they will allow one party to grow at the expense of another, a political party will think twice before entering in such an arrangement no matter how important winning the current election is. The Congress’ reluctance to enter into an arrangement with AAP (which is demanding seats in Haryana and Punjab) and with BSP (which was reportedly demanding seats in Madhya Pradesh) is grounded in such considerations.
To conclude, coalition building is not merely the art of the possible. It is backed by science as well.
Avinash M Tripathi is associate research fellow (economics) at Takshashila Institution
The views expressed are personal