Oxford University Press's
Academic Insights for the Thinking World

The Perverse Effects of Transparency?

Adrian Vermeule, author of Mechanisms of Democracy and co-author with Eric Posner of Terror in the Balance: Security, Liberty, and the Courts, is a Professor of Law at Harvard Law School. (The article below draws upon material in Chapter 6 of Mechanisms of Democracy: Institutional Design Writ Small, and upon Elizabeth Garrett and Adrian Vermeule, “Transparency and the U.S. Budget,” forthcoming in Fiscal Challenges (Garrett, Graddy and Jackson, eds.) (Cambridge University Press, forthcoming October 2007)).

The new-ish Democratic Congress has instituted several reforms in the direction of greater transparency in the lawmaking and budget-setting process. Prominent among these is a rule requiring that “earmarks” — roughly, appropriations targeted towards specific projects in particular House districts – be collected into a single list and, crucially, that the legislators who push for an earmark must attach their names to it. Recently, the New York Times reported that the new rules have greatly increased the number of earmarks (Edmund L. Andrews and Robert Pear, “With New Rules, Congress Boasts 9780195333466.jpgof Pet Projects,” New York Times, August 5, 2007). The distribution of earmarks among legislators has also changed: “Many lawmakers say the increased openness has put the [chairs of appropriations subcommittees] in an awkward position. Because everyone can see who is receiving what, rank-and-file members are clamoring for their districts to obtain a bigger share of the goodies. Similarly, constituents in home districts are becoming bolder as the earmarking process becomes less mysterious.”

The article does not mention interest groups as such, but the “constituents” it describes are not John and Jane Voter; they include military contractors, state and local universities and museums, and the Coast Guard. Transparency always has a dual effect: it allows organized good-government groups and voters at large to monitor what their representatives are doing, and also allows organized interest groups who feed off the federal budget to ensure that the representatives whose election campaigns they support are bringing home the bacon. Opacity or secrecy, by contrast, makes it harder for voters to monitor their legislator-agents, but also ensures that legislators cannot make credible commitments to interest groups. Where earmarks are opaque, the legislator who swears to the pork producers, or the teachers’ union, that she has steered money in their direction lacks credibility; the interest groups know that that she knows that her assurances cannot be verified. So one general effect of transparency is to bind legislators more tightly to the wishes of the median voter in their districts, another is to bind them more tightly to the wishes of special interests who may or may not support the same policies as the median voter. Which of these causal effects will dominate, and whether the result is good or bad (on net), is unpredictable and doubtless varies across contexts. A well-known theory of John Ferejohn’s explains self-imposed legislative transparency by drawing upon principal-agent theory; legislators are agents who impose greater transparency on themselves in order to induce principals to elect them rather than opposing candidates, and to award them more authority. By reducing monitoring costs, transparency reduces the risk that legislators will stray from what the principals want. Again, however, the “principal” can be either the median voter or any third party who benefits from monitoring legislators’ behavior. To be sure, it is possible that before the recent reforms interest groups could penetrate the opacity of the earmarking system at lower cost than voters at large, in which case transparency might have leveled the playing field; but this is a factual claim, and the article suggests to the contrary that interest groups have benefited as much or more.

Whether the recent increase in earmarking is indeed a genuinely perverse effect, from the social point of view, depends upon the optimal level of earmarking, about which no one has a very good theory, and upon the likely alternatives. In the absence of a given earmark, would there be no appropriation at all, or a general appropriation, or just an earmark for some other legislator’s pet project? But the increase in earmarking is clearly an unanticipated effect, and perverse insofar as pro-transparency reformers expected or hoped that sunshine would disinfect the appropriations process by reducing earmarks. Transparency is a game with not two players – legislator and voters – but three – legislators, voters, and interest groups. The stock arguments for transparency, such as the claim that sunshine disinfects, are thus hopelessly simplistic.

Recent Comments

There are currently no comments.