Democracy and Distribution
A fascinating, if slightly tongue-in-cheek, political philosophy discussion could be beginning in the blogosphere. Josh Barro started it by asking "Is Iowa Necessary?" and faulting its governor Terry Branstad for supporting federal subsidies for wind power but not solar:
Branstad told Radio Iowa he agrees with Romney that money used to support Solyndra, the solar panel manufacturer, was "wasted" but that the wind tax credit is "a way different thing."Ultimately, Barro is asking what is to be done about politicians who abuse democratic processes to extract rent for their narrow constituencies. Matt Yglesias joined in, posting a link to this proposal by political theorist Xavier Marquez, who suggests randomly assigning legislators to new constituencies when they have to run for re-election. Marquez' idea would seek to remove the political incentive for rent-seeking on behalf of a particular local interest group -- politicians would have to reintroduce themselves to their new constituencies and justify their records at the end of each term -- while preserving democratic accountability structures.
What's so great about the wind credit? Well, according to Branstad, it has encouraged the construction of wind turbines all over Iowa, which means jobs for Iowans and rental income for Iowa farmers. If that sounds to you a lot like the arguments for subsidizing solar power -- and the arguments for every industrial subsidy ever -- you're not alone.
Of course, the really important difference between wind subsidies and solar subsidies is that Iowa is windy and not especially sunny. If the purpose of the federal government is to do nice things for Iowa, then obviously it should prioritize wind over solar.
Branstad's attitude isn't new. Iowa has long leveraged its status as the first caucus state to hold politicians hostage, demanding their support for good-for-Iowa, bad-for-the-country policies from farm subsidies to ethanol mandates and now the wind energy tax credit.
I find Marquez's proposal more than a little absurd -- I mean that in the best way possible -- but I think it begins the search for real solutions. (My main objection to his proposal is that it doesn't really eliminate the potential for rent-seeking, it merely shifts it in time and creates a credibility problem. If politicians could promise to seek rents on behalf of their new district during re-election, then nothing is solved. In fact, when representatives are moved through congressional redistricting, this is a common occurrence. My second objection is that I think it weakens accountability structures to have your performance judged by someone else than whoever installed you. Not only would constituents have limited time to learn the records of their new representatives, they would be less careful to elect representatives because "they wouldn't be theirs.")
There are broader questions to ask here, though: How do we design a political system which removes the incentives for, or the capacity to, extract rents -- without eliminating democratic accountability? How do we get politicians to represent their constituents without representing their exclusive interests?
The economist Mancur Olson thought about these questions in two very important books, The Logic of Collective Action (1965) and The Rise and Decline of Nations (1982). To give an absolutely unjust summary of these two works, the former puts forth a theory of how groups organize, while the latter seeks to apply the theory empirically on history. Olson's theory is that groups form on the basis of their common interests, and so the greatest incentives for organization are for those who could share the most concentrated gains, while there are weaker incentives for groups with diffuse gains. Over time, Olson thinks, these "distributional coalitions" are able to extract rent for their narrow interests at the expense of more diffuse groups -- the largest of which, of course, is the nation itself.
Branstad should make terribly good sense within Olson's vision. Iowa is a small state in terms of population, but it exercises outsize power not just because of its huge over-representation in the Senate, but because its citizens share a very concentrated set of interests -- ones which combine with other distributional coalitions in other agricultural states.
And yet, revoking statehood does not make sense. The ad absurdum extension of Barro's argument is that all states should lose their power to elect representatives because their sets of interests are too narrow, and that all electoral business for national legislatures should occur at the federal level. But that ignores the point that distributional coalitions can still assemble, and that arguably they can be stronger whenever they can select the closest group, which is easier among large populations. If New York City was not tied down to the State of New York, it would make sense for its representatives to form a distributional coalition with other urban centers and advance a much narrower set of interests. The construction of U.S. states, in other words, does force some (but perhaps not enough) diffusion of the interest groups.
There are a few obvious ways to limit rent-seeking:
(1) Take away the goodies. If the returns to rent-seeking are low, distributional coalitions will be much weaker. This would involve a major decrease in discretionary federal spending and economic involvement.
(2) Decrease the leverage of distributional coalitions in bargaining. If Iowa didn't have a de facto veto over American agricultural policy, or West Virginia over American energy policy, then agricultural and energy policies could be closer to optimal for the nation, instead of Iowa and West Virginia respectively. One way to reduce leverage comes from the "holdout problem": reducing the thresholds for legislative action will reduce the power of the marginal holdout to extract rent. Most obviously, that would mean ending the filibuster.
(3) Make the most diffuse interest group more concentrated. Federalism may be a potent solution to federal rent-seeking. Shifting power from the federal government to local governments will increase the concentration of interests of the most diffuse group. That will reduce the relative power of distributional coalitions -- if Iowa had to pay for Iowa's farm subsidies, instead of paying less than 2 percent of their cost, my bet is that Iowa would want a whole lot less farm subsidies.