Corralling Capture

Article excerpt

Regulatory capture is an idea at the center of virtually any discussion of the appropriate balance between Congress and administrative agencies. It has echoed throughout the proceedings of this conference. But it is also much-abused. So I want to use this Essay to make a plea for more rigor in how we think and talk about the idea of regulatory capture. Toward that end, consider three fairly simple framing points.

First, when we think and talk about capture, definitions matter. There is a standard 20,000-foot definition of "capture": a process by which policy is directed away from the public interest and toward the interests of a regulated industry. (1) Fair enough. But note the obvious problem here: The standard definition requires a conception of what the public interest is in the first place. Put another way, arguments about capture necessarily turn on a difficult counterfactual inquiry about what public-interested regulation would look like in capture's absence. (2) This is no mean feat, as virtually any policy can be framed in public interest terms.

Conceptual slipperiness has led to some recent efforts among legal scholars and political scientists to specify more clearly the problem of capture and define key terms. Some have now taken to talking about two different kinds of capture: materialist and non-materialist. (3) The materialist version of capture is the classic account. (4) Concentrated and diffuse costs and benefits create asymmetric stakes among interest groups. (5) Collective action problems make it harder for some groups to organize than others. (6) Here, capture is a structural problem that allows some interests to systematically win out over others.

Compare this classic conception to the newer, non-materialist theories of capture--sometimes referred to in the scholarly literature as "cognitive" or "cultural" capture. (7) This idea emphasizes interest-group capture of the administrative process through the creeping colonization of ideas. Thus, an industry can somehow convince regulators to think like it. (8) Consider financial deregulation, which some observers think has contributed substantially to our recent economic woes by lowering capital requirements, allowing banks to get into the securities business, and the like. (9) These changes, the argument goes, came about at least in part because the financial industry convinced regulators that what was good for Wall Street was good for America. (10)

The problem with the classic, materialist version of capture is surprisingly simple. If you look really hard at the political science and related literatures, it is difficult to find any good, solid empirical evidence that materialist capture exists at all. (11) Diffuse interests often do quite well. (12) Concentrated interests often lose, sometimes spectacularly. (13)

For its part, the non-materialist account of capture does not fare much better. Indeed, non-materialist capture begins to resemble, upon further examination, the marketplace of ideas. Some ideas win out; some do not. This helps underscore the more general problem, noted previously, that virtually any policy position can be framed as furthering the public interest. (14) It also suggests that, when deploying the concept of capture, it is quite easy to lose focus to the point where we are not talking about much at all.

Other efforts to define terms are more helpful. Legal scholars and political scientists now also talk about "strong" versus "weak" forms of capture. (15) Strong capture is the idea that interest-group rent-seeking is so pervasive and so socially costly that we might be better off without any regulation at all. (16) Weak theories of capture, by contrast, maintain that the regulation that gets kicked out the back end of the administrative process is less publically interested than it should be but is still on balance social-welfare-enhancing. (17)

The distinction is an important one, for it helps us to understand what remedies might be indicated upon a "capture" diagnosis. …

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.