So here's the plan. We're going to deleverage, re-regulate, clean up the mortgaged stables of Wall Street. We're going to reorganize, reform, revive the spirit of regulation. We're going to make banks work for us, particularly now that we own them. We're going to hammer down executive comp, cleanse credit default swaps through a clearinghouse, rethink accounting without accountants. Never never never never again. Only a small niggling question: How do we protect ourselves against, well, ourselves? This is where a careful, obsessive search for culpability is actually quite useful. If you believe, like most pols, that the responsibility for much of what occurred falls squarely on Wall Street, then the way ahead is straightforward. If, however, you believe that we've gotten here through close complicity between Wall Street, Main Street and Pennsylvania Avenue -- all those streets! -- then we face a more difficult task. I'm going to dwell on the latter, because if it's the former, well, Barney Frank and Christopher Dodd have already laid out sufficient plans. But if it's the latter, we're wrestling with a much more difficult set of problems peculiarly characteristic of our democracy.

Seven decades ago, economic collapse forced a similar realism on the American government. Faced with a shredded economy, we essentially erected a fourth arm of the state (fifth if you count the media) in the form of independent regulatory agencies such as the Securities and Exchange Commission. We had experimented with administrative government before -- the Federal Trade Commission, the Federal Reserve -- but with the New Deal, agencies proliferated and assumed new powers. In the late '30s, the rationale for this accretion of government power was clearly outlined in "The Administrative Process" by James Landis, a disciple of Louis Brandeis, an architect of the SEC and a dean of Harvard Law School. Landis' classic text represented a kind of high-water mark in the belief that independent agencies were necessary because of the advancing complexity of institutions and markets that required a) technical experts who understood them and b) professionals who could deal with issues rationally, objectively, scientifically.

We were more innocent then; even Landis, who remained a believer, criticized the evolution of these agencies in a report for President John F. Kennedy more than two decades later. Today the complexity of the financial system far outstrips what existed in 1960 and the capacities of our agencies today. But while there is a general consensus that some form of re-regulation is necessary, we are no longer as innocent as we were in 1937. Beyond inadequate budgets, or institutional confusion, the chronic weakness of an administrative state fits under the term "capture," a phenomenon critics have long worried about. In one sense "capture" refers to the tendency of a regulatory agency to identify more closely with the industry itself and less with any notion of common good -- or, in a more insidious form, to conflate industry and common good. The psychological underpinnings of the capture process are deep-set: Experts who fill these agencies come from and return to industries they supervise. They identify with those companies, even if they never sniff a bribe or a cushy job. Expertise requires intimacy; reject the latter, lose the former. Their very expertise creates a narrowness of vision. And one solution to insularity -- the regular replenishment of professionals over time -- can produce patronage and mediocrity, particularly in good times.

But the "capture" phenomenon goes deeper. Perhaps the most lethal form of capture is a mental state we identify with bubbles. The key to Landis' administrative process is mastery of what he thought of as scientific principles; that is, "problems" have rational "solutions." No matter that we have a more ambivalent relationship with "reality" today than that worldview suggests. But as time passes, clarity gets undercut by ambiguity and replaced by bureaucratic process. The most insidious capture is one not by a specific industry but by political consensus; majority rule is capture writ large. And this is the crux of our challenge. How, in a democracy, can independent regulators resist the popular will? Should they? After all, democratic theory, like efficient market theory, argues for the wisdom of crowds. Our experience with regulation is of agencies that, for reasons good and bad, lost their clarity of vision and mission and bent themselves to popular ideologies -- shedding any belief in navigation by objective measure (the media suffers from this, too, in a different way) opting instead to play off interests. So the question to ask as we tackle reform and re-regulation chronically recurs: How do we save ourselves from ourselves? Alas, there are no good answers.