The book challenges the modern orthodoxy of economics - that people are essentially the same everywhere, and with the right set of institutions, growth is inevitable - in three ways. First by showing that there were societies like medieval England where the institutional structure provided every incentive for growth, yet there was no growth. Second by pointing out that by objective measures the institutions of many highly successful modern economies, such as in Scandinavia, provide much poorer incentives to individuals than those of very poor economies. And lastly by showing that in the long run economic institutions that would prevent growth tend to get replaced endogenously by ones that are pro-growth.
Let me take a stab at defending the conventional wisdom.
1. I think that the time pattern of economic growth--nothing discernible for centuries, then a relatively rapid and dramatic increase--is a problem for just about any theory. How many of the usual explanatory variables are going to show such a pattern? Institutions are unlikely to remain constant for years, and then change (remember that for institutional economists, institutions are not just the formal characteristics of government; they are the habits and conventions that are deeply embedded in the society at large). But human qualities--the characteristics that are debatably cultural or genetic--are at least as unlikely as institutions to follow a path of stability interrupted by rapid evolution.
This reminds me of the Solow paradox. Remember that in 1987, Solow said that we see computers everywhere but in the productivity statistics. Well, 15 years later, productivity was soaring, and computers were on that phenomenon like white on rice. Looking back, you could calculate that total computing power in the economy in 1987 was infinitesimal compared with what came later. Once computers became powerful enough and pervasive enough, visible productivity gains soon followed.
Maybe the same story applies to the Industrial Revolution. The fast-growing sector, in this case textiles, may have been growing at a high rate all along. But starting from a low base, even if you double every twenty years, you don't affect the economy. Once the fast-growing sector reaches 3 or 4 percent of GDP, its rapid growth rate starts to have a visible effect on the overall growth. After a few more doubling times for the fast-growing sector, it becomes obvious that this sector is driving overall productivity.
2. A high standard of living in Scandinavia is not necessarily a refutation of institutional economics. Although their institutions are more anti-growth than ours in the dimensions of personal tax rates and labor market regulation, there may be other relevant dimensions (corporate profit taxes?) where our institutions are more anti-growth than theirs. The net differences may not be large.
3. The argument that bad institutions get replaced in the long run is not horribly devastating to institutional economics. One would hope that bad institutions do get replaced, although reading The Bottom Billion or Let My People Come is a reminder of how long the long run can be. In the less-than-long run, bad institutions are so entrenched that immigration is the fastest path to upward mobility.
For discussion: one element of the contemporary U.S. economy is a high participation rate for women in the labor force relative to, say, 1920. Does this reflect
(a) changes in the genetic or cultural makeup of women
(b) changes in technology
(c) changes in institutions (families, colleges and universities, employers, and government regulations)
I would not want to put a weight of 100 percent on (a) and 0 on the other factors. Similarly, I would not want to put a weight of 100 percent on Clark's theory that changes in the genetic or cultural makeup of men produced the Industrial Revolution.