The US government spends billions of dollars a year subsidizing American farms, providing massive benefits for some farmers and dispersing the costs among millions of taxpayers. Once all the costs and benefits of lobbying and paying for the subsidies are tallied up, it turns out that they make the country worse off. One tangible result seems to be that these subsidies increase the prevalence of certain sorts of unhealthy foods, like soybean oil and corn syrup, in our diets.

So why do we have them?

It’s because the farmers who receive thousands — and sometimes millions — of dollars in subsidies have a huge incentive to lobby for the subsidies, while individual taxpayers have little incentive to actively oppose the subsidies, which only cost them a few dollars each. Politicians favoring the subsidies win the support of “Big Ag” without losing the support of anyone else.

Agricultural subsidies might not seem like a big deal in themselves, but the problem of concentrated benefits and dispersed costs applies to countless other programs, causing economic inefficiencies to pile up quickly. Public choice theory can help us understand why.

What is public choice theory?

Public choice theory is, roughly, the economic analysis of politics. Anthony Downs, a forefather of public choice, worried that economists had a tendency to “treat government as a machine” designed to maximize social welfare rather than as an institution run by flesh-and-blood human beings. Downs thought this traditional approach was flawed since “there is little point in advising governments to [maximize social welfare], or forming recommendations of action based on the supposition that they might, unless there is some reason to believe that they will.” We should therefore study the incentives facing real-world politicians, bureaucrats, and voters to see whether a given government program is likely to promote the common good.

One lesson of public choice theory is that real-world government programs often fail to promote the common good because of the problem of concentrated benefits and dispersed costs. That is, the benefits of many programs accrue to a few people, while the costs are spread out among the rest of us. Agricultural subsidies are a case in point.

It’s sometimes claimed that public choice worries about government failure assume that political actors are selfish. Politicians care only about enriching their campaign coffers, and voters care too little about the common good to hold them accountable at the ballot box. Indeed, this view has sometimes been fostered by public choice theorists themselves. For instance, Downs claims that rational behavior is “directed primarily toward selfish ends.”

However, it’s a mistake to think that the public choice theory of government failure rests on the assumption that political actors are largely selfish.

Even when most people are public-spirited, governments can fail.

To see why, consider Stanley Kelly’s remark from his introduction to Downs’s book, An Economic Theory of Democracy: “Just as firms that do not engage in the rational pursuit of profit are apt to cease to be firms, so politicians who do not pursue votes in a rational manner are apt to cease to be politicians.” Democracies select for politicians that win votes, regardless of whether that’s the politicians’ explicit aim. The trouble is, the problem of concentrated benefits and dispersed costs implies that the strategy that gets a politician the most votes will often be bad for the public.

Think back to farm subsidies. A typical voter won’t track politicians’ positions on all of the issues and vote them out for supporting bad policies. Why not? Because a single vote won’t change the outcome of the election. Casting your one vote for the best candidate for the country won’t result in the election of that candidate. For this reason, citizens may remain “rationally ignorant” of policy, not because they’re selfish, but because voting isn’t an effective means of serving the common good. So a candidate’s support for farm subsidies is unlikely to cost them the support of even altruistic voters because most of them won’t even know that the candidate supports the subsidies.

Farmers, by contrast, are more likely than typical voters to be attentive to a politician’s stance on farm subsidies because it directly affects their livelihood. And they may support farm subsidies not on selfish grounds, but because they sincerely think that the subsidies are good for the country. Perhaps they believe that “government subsidies help keep farming profitable and stable, allowing for the commercial finance of modern agriculture, the development of products and technologies that help farmers produce more food at a lower cost, and the preservation of production resources in case of future need,” as retired agricultural economist and Auburn University associate professor W. Robert Goodman argues in the Wall Street Journal.

The key point is this: endorsing farm subsidies will probably win a candidate some votes and lose them none. Thus, endorsing the subsidies is a vote-maximizing strategy. All else equal, then, a candidate that endorses the farm subsidies (perhaps with perfectly good intentions) will defeat a candidate that does not. In this way, even democracies populated by well-meaning people will select for policies that concentrate their benefits and disperse their costs, often to the detriment of the public good.