The Eighteenth-Century Hockey Stick

Gregory Clark's book, A Farewell to Alms, described on the jacket by the New York Times as "the next blockbuster in economics," makes three startling assertions. The first is that living standards for the average human being showed virtually no increase between hunter-gatherer times and the year 1800, and turned sharply upwards thereafter, leading to a historical growth curve that looks like a hockey stick.¹ The second is that the enormous rate of economic growth since then is due to the spread of certain bourgeois virtues like hard work, savings, and reliability. Third, Clark attacks the currently fashionable notion among development economists that the growth of property rights and rule of law were the necessary conditions for economic takeoff.

Clark, an economic historian at the University of California, Davis, has filled his volume with charts, tables, and statistics on a wide range of economic data, including unusual items like daily calorie intakes of hunter-gatherers and implicit interest rates for societies without a money economy. However, despite this wealth of data, Clark's first thesis on flat living standards through 1800 is less remarkable than it first seems, while the evidence provided for the second thesis on the origins of the middle class and third on the unimportance of institutions is extremely limited.

The assertion that there was no increase in average living standards from hunter-gatherer times to 1800 is not technically correct. Clark uses the word "average" not to denote per capita income, as most economists would, but rather the income of the average lower class person, which admittedly in most agricultural societies meant upwards of 80–90 percent of the whole population. The fact that the average poor person in Europe in the year 1800 was poorer than their ancestors in 1300, or not much better off than poor people in Paleolithic times, simply reflects the inequitable income distribution of most agricultural societies. But per capita income was increasing steadily all this time, measured by the achievements we usually associate with the progress of civilization like the Great Pyramid or the Palace of Versailles.

The reason why the income of the poor moved upward so quickly after 1800, Clark argues, is that virtually all earlier societies were caught in a Malthusian trap. Any increase in incomes would lead to an increase in birth rates; since the marginal productivity of each additional worker was zero, more people had to divide the same amount of resources and incomes would fall. The only way to increase incomes was to increase death rates. Clark does in fact show that wages in England increased after the plague years beginning in the 15^th century. But the biggest Malthusian trap of all was China, with its huge population and periodic famines, that induced writers like Malthus and Ricardo to come to their gloomy conclusions that there was no way to raise the living standards of the poor.

Clark fails to demonstrate that the world was in fact caught in a Malthusian trap before 1800. The logic of the Malthusian model is ironclad, but is only as good as its going-in assumptions. For the trap to apply, all available land must be in use, and there must be a linear relationship between incomes and birth rates, with peasants deciding to have more children rather than buying a new roof or plough after a good harvest. Clark's own evidence suggests that as late as the 1700s when Malthus was writing, neither China nor Japan were caught in Malthusian traps. China was still a frontier society with unsettled land, and therefore could sustain population growth without falling living standards. Japan's population could grow substantially before the Meiji Restoration without impoverishment because it achieved remarkable increases in agricultural productivity. Clark has data showing a correlation between income and fertility for some groups, but also cites cases where societies could deliberately restrict fertility, meaning that the Malthusian logic would not necessarily apply to them. And so on.

Projecting...