They offer better opportunities for trial-and-error learning

From Jeffrey Friedman, RIP by Ilya Somin.  The subheading is Friedman was a notable - and underrated - political theorist.

While Jeff was generally considered a libertarian, or at least libertarian-adjacent, he largely rejected conventional libertarian  political thought. He was equally skeptical of standard approaches to democratic theory. Like libertarian (and some non-libertarian) critics of modern democratic government, Jeff argued that widespread voter ignorance was a serious danger. But, unlike them, he claimed that most such ignorance was inadvertent "radical" ignorance, rather than individually rational behavior (as believed by most economists and other social scientists). He also argued that similar radical ignorance afflicted expert decision-making on public policy, and conventional social scientific analysis. He thus concluded that markets generally produce better results than government, not because of superior incentives, but because they offer better opportunities for trial-and-error learning.

An interesting point.  I am inclined towards the incentives and disincentives aspects of markets (both for individuals and enterprises) but the point about superior beta-testing of ideas is a valid one.