Bouie vs. Complexity

by Kevin D. Williamson

Regarding my piece “The Victim Presidency,” Jamelle Bouie writes: “It’s not a defense of Obama to say that this betrays an ignorance of the size and complexity of the federal bureaucracy.”

Complexity is of course something that I spend a great deal of time writing about. It is my main theme of late. If you’re in the market for a book about the problems that complexity presents for, say, the regulation of the health-insurance market, I’m your huckleberry. Interested in knowledge problems as they relate to the Affordable Care Act or structured finance? Happy to point the way.

Mr. Bouie insists that he is not simply trying to make an excuse for the president’s revealed incompetence in sundry matters, but of course that is precisely what he and other apologists for the administration are doing. If they were really interested in complexity as such, then they would bring it up on the front end of the policy debate, rather than on the back end.

I’ve seen this happen so many times that every other policy debate looks to me like an ancient rerun of Three’s Company: Do you think there’ll be a comic misunderstanding in this episode, too? It unfolds like this: Politicians on the Barack Obama model promise that they will muster their native intelligence and empirical evidence to bring order to, e.g., the health-care industry, through the judicious application of regulation. People like me tell them that the effects of such regulation are almost certainly going to be other than what was intended, because such markets are too complex to be understandable, predictable, or steerable, even in principle. Even if every bureaucrat who touches health care or the labor market has the brain of an Einstein and the soul of a St. Thomas Becket, it will not turn out the way it is intended. And then, when it doesn’t turn out as intended, Jamelle Bouie et al. protest that the toldya-so chorus “betrays an ignorance of the size and complexity of the federal bureaucracy.”

And they never even consider the question: If the federal bureaucracy is so vast and complex that its behavior cannot be adequately managed, how is it that the phenomena that the bureaucracies are tasked with managing—orders of magnitude more complex than the bureaucracies themselves—are supposed to be manageable? To consider the question with any intellectual rigor is to accept real, meaningful, epistemic limits on what government can do.

Consider quant-geek Paul Wilmott’s argument that all financial models are inherently inaccurate for the simple (but not really simple) reason that models have an effect on financial reality, and models that attempt to account for their own effects end up in an infinite recursive loop. For similar reasons, there is a pretty good argument that the net effects of expert-endorsed federal economic policy variations over the past century are indistinguishable from random variation, and that policy variations are a relatively minor factor in economic performance. Even assuming that we could, in theory, construct accurate and useful economic-policy models, there is the nagging question about whether politicians are inclined to do so: Back when he was a good economist rather than a fixed-gear rant-monkey, Paul Krugman did excellent work making mincemeat of orthodox ideas about national competitiveness and the fanciful policy prescriptions associated with them—ideas put into the service of agendas that are mainly about, in his words, “status and power,” establishing a phantom criterion that is “extremely useful as a political device.”

I would be endlessly pleased if Mr. Bouie and those of his bent suddenly developed a newfound appreciation for the problem of complexity. Perhaps next time around we could consider it  before we draw up the policy rather than after it has failed. Make excuses if you must, but learn from your mistakes.