About my call for better government, my friend Lynne Kiesling writes:
I do think Will is creating a false dichotomy in his fine-hair-splitting. “Norms of anti-corruption and civic responsibility” are not substitutes for institutions, like a constitution, that recognize the inducement to corruption that is inescapable when some subset of a population has legal power to determine outcomes. The point that I think Will is missing is that the incentive is inescapable, even if the actual corruption does not occur.
Put another way: institutions matter. Formal and informal institutions matter. Constitutions that define and limit the role of government and norms of civic virtue are institutional complements in creating relatively better government than we would have in the absence of these institutions. But the reason that we need the formal institutions, and particularly formal institutions that define the scope and limit of government power and action, is that civic virtue is often insufficient to deter elected representatives from following the lure of the ever-present corruption incentive.
I agree with just about everything Lynne says. Though I think that if she checks her North and Greif she'll find that norms count as institutions, in their broad sense of the term. And one of the deepest facts of the institutional world is that conscience is cheaper than police.
Anyway, here's something I said in the earlier comments thread, in response to Tim Lee:
[I]t's hard to make government work well. The public choice guys are right that its more about structure than public-spiritedeness. It is laughably naive and romantic to think that sufficient public-spiritedness will deliver good government. But it remains that we WANT good government, and public-spiritedness helps. Libertarians seem loathe to admit this, and I think it's a problem for us.
I think this is a disagreement of emphasis and strategy. Yes, civic virtue is insufficient. But it's also necessary, and necessary is a big deal. The best constitution in the world isn't worth a damn in a context of pervasively lousy norms. Incentive structure and institutional form is so important that political economists often feel that a breath spent affirming honesty and public-spiritedness is a breath wasted. After all, you could be telling folks just how important incentives are.
What I was objecting to in Steve and Mike's posts was not their indisuputably sound opinion that institutions matter, or the idea that it is not surprising when opportunities for corruption are seized. What I thought I detected, and what I was objecting to, was a sense of vindication in the view that people with political power are irremediably corrupt and cannot, must not, be trusted. Because this view is false. If there is going to be political power, we must trust people not to abuse it, and many people don't abuse it, for which we should be grateful. If we can remove incentives for abuse, and therefore lean less on frail conscience, then we should. But I think an indiscriminately scathing attitude toward politicians and political power (of which I have often been guilty) is harmful, both to the level of social trust that does in fact help determine the effectiveness of our suboptimal institutions, and also to the public credibility of libertarians, many of whom do have an especially rigorous grasp of what it would take to make our institutions work better.
Let me draw a parallel and see if it flies. I think the corporate form suffers from some thorny agency problems. The incentives of owners and managers are often poorly aligned. And I'm convinced some of the recent financial crisis is a consequence of corporate executives abusing the trust of their creditors and shareholders. However, I am unimpressed with arguments that this calls into question the legitimacy of capitalism, the corporation, or corporate executive power in precisely the same way I am unimpressed with the suggestion that Rod Blagojevich calls into question the legitimacy of democratic power.
We need better institutions. But there is no insitutional design, whether it be of a public system of democratic governance or a private system of corporate governance, that is so airtight in aligning the interests of principals and agents that conscience and trust are unneeded. One of the reasons many people are skeptical of “cynical” public choice-types is that the quest for incentive-compatible institutions can look like an attempt to squeeze all the trust out of the system. And it is indeed an attempt to rely less on trust. But the point is not to rely less on trust; the point is to make our institutions more likely to deliver what they promise. Emphasizing, truly, that the system can't possibly work without some level of virtue and trust is a good way to reassure skeptics that you haven't declared jihad on fellow feeling and are not out to wring the inefficiency from our institutions by wringing out the humanity.