One way to look at the issue of markets vs. government is to compare the relative strengths of entrepreneurs operating in the market with that of entrepreneurs operating through government. I think that this issue can be addressed along several dimensions:
1. Ability to resolve Coasian bargaining problems
Consider an entrepreneur wishing to solve an urban transportation problem. If the solution involves reconfiguring a lot of land, by building a new highway or rail system, then a market entrepreneur is likely to face a huge Coasian bargaining problem in trying to get all affected parties to come to terms that allow the project to be built. A policy entrepreneur, backed by the coercive power of government, can implement the solution by fiat.
I think that a lot of pro-government, anti-market bias comes from implicitly assuming that the policy entrepreneur knows everything about a problem. This is a troublesome implicit assumption, for a number of reasons. First, some problems are simply too complex to be fully understood. In James Manzi’s terminology, causal density is just too high. You cannot isolate the causes of phenomena, and so you cannot reliably predict the consequences of new actions.
Once we accept that knowledge is going to be imperfect, then the issue becomes comparing what a market entrepreneur is likely to know with what a policy entrepreneur is likely to know. As Hayek pointed out, the market entrepreneur works within a price system that coordinates local knowledge. In contrast, the policy entrepreneur relies on centralized knowledge. For opening, operating, and closing restaurants, local knowledge is likely to be more robust. For regulating nuclear power, centralized knowledge probably is more reliable.
3. Incentive to innovate
Large organizations are not well suited to innovation. The problem is that for an individual operating within a large organization, the risks and the rewards are both too small. If your project does really well, you get at most a trivial personal reward. If it costs a lot of money and flops, you personally suffer very little. The type of project that is optimal for a middle manager to launch is one that has a high probability of a small upside, even if it has a nearly unlimited downside. To prevent such projects from being launched, organizations set up procedures that make it difficult to undertake new projects.
(Note that a typical venture capitalist prefers a project with a nearly unlimited upside even if it has a high probability of a small downside. And self-funded entrepreneurs prefer projects where both the probability and magnitude of the potential upside are high enough to offset the downside risks.)
4. Incentive to evolve
Government has no reliable mechanism for discarding bad programs. The market has the profit-and-loss system. Thus, the market is better suited to evolution.
To me, the evolution issue is particularly important. Only a policy entrepreneur could undertake the DC Metro subway system. But once that works out badly, it does not fall by the wayside as would a market entrepreneur’s failed project.