In a recent paper in this journal, McCloskey (2015) repeats her argument that institutional rules, as economists typically mean the term, are not responsible for the dramatic increases in prosperity experienced in many countries over the past several centuries. The increase in income per person ‘was by a factor of a hundred, which cannot be explained by routine movements to an efficient equilibrium’. In other words, she argues, the improved incentives associated with traditional enhancements in institutions, such as the expansion of private property rights and improved freedom to trade, are not enough to explain a rise in incomes of the magnitude we have witnessed. Instead of the formal rules, McCloskey's alternative explanation is that changes in informal norms and ethical views related to the dignity of commercial activity are the real cause of the great prosperity of the last several centuries.
It is not my purpose to challenge her thesis directly in this comment. I am in fact deeply sympathetic with the general notion that culture matters, and may matter more even than formal institutions. Instead, the purpose here is to challenge a specific empirical claim she used to bolster her argument. The claim is worth quoting explicitly (McCloskey, 2015),
Bettering institutions do not explain the bulk of modern levels of income. New Zealand, for example, is honestly and efficiently governed. Italy is not. . .In 2009 in the Economic Freedom Rankings, New Zealand ranked first in its legal system and fifth from the top in its freedom from regulation. Italy in its legal system ranked sixty-third, just above Iran, and ninety-fourth in its freedom from regulation, just above the Dominican Republic. Italy, as any sentient Italian can tell you, has terrible public institutions.
Yet in real GDP per person New Zealand and Italy in 2010, were nearly identical, at $88.20 and $86.80 a day. . .
The beautiful thing about basing an argument on just two data points, assuming you get to pick, is that you can always tell the story you want and you always get a perfect R2. The problem for McCloskey is that we have data on the institutional rules for not two countries, but for over a hundred!
Figure 1 shows a scatter plot and accompanying OLS regression line for all 149 countries for which we have available data. The horizontal axis is the legal structure and property rights rating for the year 2010, measured on a 0–10 scale, taken from the Economic Freedom of the World index produced by Gwartney et al. (2014). The vertical axis is logged GDP per capita for 2010, measured in ppp US dollars, obtained from the World Bank's World Development Indicators online database.
Figure 1. Income and institutions.
In actual fact, current institutional measures, such as the same legal structure ratings McCloskey cites, do explain the bulk of current differences among countries’ income levels. The R 2 in this regression is not perfect, but 57% should satisfy most people's definition of ‘bulk’. For reference, the data points for McCloskey's exemplar nations, Italy and New Zealand, are indicated in Figure 1.
To be sure, McCloskey could respond that informal norms and ethical standards are more important than formal institutions or that norms and ethical standards are the driving force behind institutional change. These would be valid criticisms to consider. The bottom line, however, is that her claim that modern measures of formal institutional quality do not even correlate with current levels of income is simply incorrect.
In a recent paper in this journal, McCloskey (2015) repeats her argument that institutional rules, as economists typically mean the term, are not responsible for the dramatic increases in prosperity experienced in many countries over the past several centuries. The increase in income per person ‘was by a factor of a hundred, which cannot be explained by routine movements to an efficient equilibrium’. In other words, she argues, the improved incentives associated with traditional enhancements in institutions, such as the expansion of private property rights and improved freedom to trade, are not enough to explain a rise in incomes of the magnitude we have witnessed. Instead of the formal rules, McCloskey's alternative explanation is that changes in informal norms and ethical views related to the dignity of commercial activity are the real cause of the great prosperity of the last several centuries.
It is not my purpose to challenge her thesis directly in this comment. I am in fact deeply sympathetic with the general notion that culture matters, and may matter more even than formal institutions. Instead, the purpose here is to challenge a specific empirical claim she used to bolster her argument. The claim is worth quoting explicitly (McCloskey, 2015),
Bettering institutions do not explain the bulk of modern levels of income. New Zealand, for example, is honestly and efficiently governed. Italy is not. . .In 2009 in the Economic Freedom Rankings, New Zealand ranked first in its legal system and fifth from the top in its freedom from regulation. Italy in its legal system ranked sixty-third, just above Iran, and ninety-fourth in its freedom from regulation, just above the Dominican Republic. Italy, as any sentient Italian can tell you, has terrible public institutions.
Yet in real GDP per person New Zealand and Italy in 2010, were nearly identical, at $88.20 and $86.80 a day. . .
The beautiful thing about basing an argument on just two data points, assuming you get to pick, is that you can always tell the story you want and you always get a perfect R2. The problem for McCloskey is that we have data on the institutional rules for not two countries, but for over a hundred!
Figure 1 shows a scatter plot and accompanying OLS regression line for all 149 countries for which we have available data. The horizontal axis is the legal structure and property rights rating for the year 2010, measured on a 0–10 scale, taken from the Economic Freedom of the World index produced by Gwartney et al. (2014). The vertical axis is logged GDP per capita for 2010, measured in ppp US dollars, obtained from the World Bank's World Development Indicators online database.
Figure 1. Income and institutions.
In actual fact, current institutional measures, such as the same legal structure ratings McCloskey cites, do explain the bulk of current differences among countries’ income levels. The R 2 in this regression is not perfect, but 57% should satisfy most people's definition of ‘bulk’. For reference, the data points for McCloskey's exemplar nations, Italy and New Zealand, are indicated in Figure 1.
To be sure, McCloskey could respond that informal norms and ethical standards are more important than formal institutions or that norms and ethical standards are the driving force behind institutional change. These would be valid criticisms to consider. The bottom line, however, is that her claim that modern measures of formal institutional quality do not even correlate with current levels of income is simply incorrect.