A new paper (pdf) by Nick Bloom and colleagues shows that there is "substantial dispersion" in the quality of management across US manufacturing establishments, which confirms earlier research. This raises a paradox.

The very fact that many poorly-run firms stay in business shows that market forces do not strongly and swiftly select against inefficiency.

Bloom's is not the only evidence we have here. Alex Coad's survey of corporate growth around the world found that firm growth is a "fundamentally random process" which is largely unrelated to past productivity or financial performance. Geroski and Gregg found that it was almost impossible to predict from past performance which firms got into difficulty in the 1990s recession. "Market selection criteria may be rather myopic" they concluded. And Bjorn-Christopher Witte has shown that market forces can sometimes actually select in favour of stupidity; they favoured irrational trend-followers in the tech bubble and risk-seeking banks in the mid-00s, for example.

All this suggests that markets do not select in favour of efficient firms not against poorly-run ones as rigorously as free market fanatics would have us believe.

Which brings me to the paradox. Despite all this, it seems clear that freeish markets are responsible for large chunks of productivity growth (pdf) - though the best natural experiment we have on this (the division of Korea) tells us that lots of emphasis (pdf) should be placed upon the "ish".

How, then, is it possible that market selection works so imperfectly and yet markets are key to prosperity?

There are (at least) two possibilities. One is that central planning of whole economies or of companies is so terribly inefficient (beyond a certain point) that even imperfect markets are a big improvement.

The other is that markets are not merely a technology but also a culture. As Gregory Clark and Deirdre McCloskey have argued, markets encourage - over time - the cultivation of "bourgeois virtues" such as hard work and prudence which themselves encourage growth. This helps explain why shock therapies of immediate transitions to market economies have had indifferent results.