The good vote banks


Markets and democracy are arguably two of the most elegant and useful creations of humankind. Like twins, they are often found together, and naturally share some features. They are alike since both involve collectives of humans behaving strategically. A study of markets and how they succeed or fail to deliver the socially optimal outcome can illuminate how the workings of a real democracy with its real failures can be improved.

Economists do it with models, often very elegant ones. Beginning with models of ideal markets, they have identified what are called market failures that plague markets in the real world. They have discovered ways to address those failures so that real markets can be nudged to grind out results closer to that of ideal markets. Consider the work of Peter Diamond, one of the three winners of this year’s Nobel prize in economics, which includes the study of labour market imperfections and their consequences.

Mr Diamond’s insight, called the Diamond Paradox, involves the friction introduced by search costs in the functioning of labour markets. Workers incur the cost of searching for jobs, and firms incur the cost of recruiting workers. Add to that the matter of expectations, and the outcome can deviate from that of an ideal labour market. If the workers’ expectations are that firms will strenuously seek recruits, workers will expend effort in job seeking; and if firms anticipate workers will carefully examine job opportunities, firms will put effort into differentiating themselves to attract the most suitable workers. The outcome, or equilibrium, will be good for all.

Instead, if both parties’ expectations are that the other party will not bother much, then the outcome will be disheartened workers and uninterested firms leading to unemployment, or a bad equilibrium. The paradox is the existence of unemployed workers simultaneous with job vacancies.

Outcomes depend on the aggregation of individual expectations. If a sufficiently large number of individuals expect a certain outcome, then the outcome expected could indeed influence what actually happens. This can be understood through Mr Diamond’s fun little “coconut economy” model. It is set on an island far away where people consume only coconuts which they harvest from palm trees. However, due to a peculiar taboo, islanders are forbidden from eating the coconuts that they pluck themselves. Since they can only eat coconuts plucked by others, they must find another person with coconuts and trade.

It is costly to pick coconuts since it means climbing a palm tree. If an islander expects no one else to gather coconuts, then it will be pointless for her to incur the cost of picking coconuts since she will have no one to exchange them with. This will be a rational expectation if all others also have the same expectation, and the predictable outcome will be starvation all around. Contrariwise, if everyone believes that a sufficient number of others will also pick coconuts, then a vigorous coconut market will evolve with full tummies all around.

Moving from markets to democracy, substituting voters for workers, and political parties for firms, we can see an analogous mode of failure for a democracy. Like for workers in a labour market, the voters’ rational expectations about the usefulness of their vote on the aggregate can lead to either a good or a bad outcome.

Democracy is not just about voting but rather about informed choice. It is costly for voters to inform themselves about political parties. Besides there’s time and effort required to vote. If the expectation is that others will not be making the personally costly effort of making informed choices, then the individual voter will rationally conclude that it is not worth the cost of informing himself about which party best deserves his vote and then voting—because his vote would not count in the outcome he desires.

Political parties, in their turn, noting that voters are not bothering to inform themselves, or are disinclined to vote, will rationally not put in any effort in the costly endeavour of differentiating themselves to appeal to voters. The outcome will be disastrous: first, no political party puts in any effort in attracting informed voters; and second, a set of political parties that are hard to differentiate. The parties then don’t bother to address the concerns of voters and thus misgovern without fear of consequences. The desirable outcome would occur only if voters expended effort required for informed voting, and political parties responded appropriately to the voters’ efforts.

One mechanism to nudge democracy from the bad equilibrium to the good equilibrium readily comes to mind. That is, somehow change the expectation of the voter from one that says that his vote does not matter (which would be rational if he believes that others will not be voting) to one that says his vote matters (because others will also be voting.)

Our voter will vote if he is assured that sufficient numbers of like-minded voters will also vote. This can be achieved by creating a coalition of voters who ex ante commit to voting, and this coalition choosing the party or the candidate to vote for based on a set of values shared by the members of the coalition.

Photo: Al Jazeera

Let’s consider this in the context of urban Indian voters. It is generally known that they largely choose to not vote, believing that their votes don’t count. With sufficient numbers of them holding this view, the expectation is rational since it amounts to a self-fulfilling prophecy. Political parties, in turn, also rationally respond to this by not even bothering to seek the votes of this segment of voters, and after elections, ignoring their concerns. This further alienates the urban voters. In essence this is voluntary disenfranchisement of the urban middle-class voter which partially accounts for the election of undesirable people to political office.

The remedy for this could be the formation of an association of voters whose members will internally decide on specific candidates (“primaries” so to speak) based on how closely candidates match the principles of the association, following which all members will vote, and equally importantly, vote only for those chosen candidates. This allows the association to make a credible claim that its members’ votes matter on the aggregate—both to every individual member of the association and to political parties.

In other words, this association of urban educated voters is an artificial “vote bank,” much like the existing vote banks that are based on other demographic characteristics such as caste and religion, and which currently have a baleful influence on the political outcome. Based on the idea that “if you can’t beat them, join them,” it recognises that in the real world, the introduction of another vote bank (which would be unthinkable in an ideal world) may lead to improvements.

Democracy as an ideal works flawlessly in an ideal world. But like markets and their failures, in the real world democracy failures lead to seriously flawed results that have terrible consequences for hundreds of millions in a country like India.If we honestly confront the reality of democracy failures, we can figure out a way to address them urgently and seriously. The time for vote banks for good governance.

Atanu Dey is a fellow at the Takshashila Institution and blogs at




  1. Paranoid Android says:

    2 questions –

    1. Is this constitutionally allowed? As in, if I decide that the whole of my company and my apartment complex is going to vote based on how cool we think the candidates sunglasses are, and by that metric Karuna wins over Bala Saheb, are we allowed to aggregate our votes thus? So, in short, does the stated purpose matter for these banks?

    2. How would you aggregate the votes physically? Would you be allowed to proxy vote for 10,000 people? How about 10? Do you collect their voting cards and cast a vote? Without a logistical route, it is a good concept, but nothing more.

    Also, how about the “negative vote” or a “vote against” concept?

  2. parijat says:

    Atanu, this is so brilliantly clarifying I could cry. Thanks!

  3. […] It’s not easy, and we have to be innovative. See for instance, Atanu Dey’s interesting idea to form middle-class vote banks to induce good […]

  4. […] As long as the middle class confines itself to the role of a cheer leader, refuses to change its voting habits or rather cultivate the habit of voting, none of the reform we wish to see can happen. There is no incentive for our political class to deliver because their cash for campaigns comes from the rich; votes are bought from the poor while everybody in the middle is too busy cursing them. Hence it is important for our revolutions to include the rich and the poor, people of all castes and religion. It is important for us to create good vote banks. […]


Leave a Comment


You must be logged in to post a comment.