Daniel Jordan Smith has written a very perceptive review of my book. There is little to argue about, but some clarifying remarks may be useful. Smith raises two key issues on which we might disagree—but perhaps the disagreement is more in emphasis than it might appear. These issues are the relationship between petty and grand corruption and the need for a greater historical perspective.
Yes, people do not like petty corruption. Petty corruption and grand corruption go together: Across 62 countries in the Global Corruption Barometer, where petty corruption is high, so is high corruption (r2 = .917). And where pickpocketing is rampant, so is corruption (pp. 81–84). But that does not mean that people see the two as equally problematic.
In many postcommunist countries, people I have spoken with view paying off the police as preferable to spending a day in court or going to the police department to pay the fine. People do not link petty corruption with inequality, and thus, petty corruption does not lead to less trust in fellow citizens. People blame themselves for corruption, as Smith argues. But they also exonerate themselves: We have no choice but to tolerate corruption? They resent far more the great wealth that political and business elites amass, and this leads to less trust, both in other people and in government.
I wish I could have done more tracing of the historical roots of corruption. I looked for ways to do this but found little I could do. I did link contemporary levels of corruption in the American states with historical data from 1880 on teacher pay, worker professionalism, and the Scandinavian population, as well as the 1924 vote for the reformist Progressive candidate for president, Robert LaFollette (pp. 229–31). Since then, I have come upon limited cross-national data from Tatu Vanhanen and Francois Bourginon on literacy rates in 1858 and economic inequality in 1910. Across 37 countries (mostly Western), the 1858 literacy rates explain 70% of the variance in 2008 corruption rates; among 25 countries, the literacy rates and the 1910 Gini (provided by David Samuels) together explain 83% of the variance in 2008 corruption. Emmanuel Saez gathered historical data on inequality in the United States. Claudia Goldin provided her time series estimates (with Edward Glaeser) on corruption from press reports. Inequality explains 45% of the variance in inequality over 59 years (1916–74). The inequality trap does have a historical basis. I wish I had these data before the book was published.
Finally, Smith asks if the decline of trust may reflect the turning away from face-to-face communications. In The Moral Foundations of Trust (2002), I show that group membership neither leads to nor stems from generalized trust. Inequality is the culprit over time in the United States, across the American states, and across countries without a legacy of communism.
The differences between us are minor, and I thank David Jordan Smith for raising these issues.
