When Thomas Piketty’s "Capital in the 21st Century" came out in 2013, it quickly became a favorite of the political left and neo-Keynesian economists as his findings fit the narrative of increasing income inequality. Paul Krugman said “ Mr. Piketty’s contribution is serious, discourse-changing scholarship in a way most best sellers aren’t. And conservatives were terrified.” Larry Summers hailed it as a “Nobel Prize-worthy contribution.”
There may be just one problem with Piketty’s earth-shattering revelation: it appears to be wrong.
First is was Acemoglu and Robinson, who by using a simple regression analysis, concluded “the main economic force emphasized in Piketty's book, the gap between the interest rate and the growth rate, does not appear to explain historical patterns of inequality (especially, the share of income accruing to the upper tail of the distribution).”
The most recent example comes from none other than the IMF’s Carlos Goes, where Piketty’s work is described as “ rich in data, [but] the book provides no formal empirical testing for its theoretical causal chain.”
The author continues “I find no empirical evidence that the dynamics move in the way Piketty suggests. In fact, for at least 75% of the countries examined, inequality responds negatively to r − g shocks, which is in line with previous single-equation estimates by Acemoglu and Robinson (2015).
The results also suggest that changes in the savings rate, which Piketty takes as relatively stable over time, are likely to offset most of the impact of r − g shocks on the capital share of the national income. Thus, it provides empirical evidence to the model developed by Krusell and A. Smith (2015), who say Piketty relies on a flawed theory of savings.
The conclusions are robust to alternative estimates of r − g and to the exclusion or inclusion of tax rates in the calculation of the real return on capital. Knowing if Piketty’s hypothesis is correct is of crucial importance. Such is the case because the policy solutions designed to counter increasing income inequality trends in advanced economies will need to tackle the underlying causes of inequality.
The results presented here show that observed increases in income inequality in advanced economies are largely uncorrelated with changes in r −g, which suggests that one needs to look for the causes of inequality (and potential solutions) elsewhere.
Perhaps before overthrowing the status quo, Piketty should focus on avoiding the cardinal sin of statistical analysis, and differentiate between correlation and causation. Then again, since the critique does come from the IMF, the agency which has been wrong about virtually everything in this decade; this time may not be any different. Of course, since the non-falsifiable art of economics is as far from science as an EBT recipients is from caring about the S&P all time highs, we may never know who is right in this particular debate.