Reinhart, Rogoff, and How the Macroeconomic Sausage Is Made - Justin Fox - Harvard Business Review - 0 views
-
This is watching the sausage of macroeconomics being made. It's not appetizing. Seemingly small choices in how to handle the data deliver dramatically different results. And it's not hard to see why: The Reinhart-Rogoff data set, according to Herndon-Ash-Pollin's analysis, contained just 110 "country-years" of debt/GDP over 90%, and 63 of those come from just three countries: Belgium, Greece, and the UK. This is a problem inherent to macroeconomics. It's not like an experiment that one can run multiple times, or observations that can be compared across millions of individuals or even hundreds of corporations. In the words attributed to economist Paul Samuelson, "We have but one sample of history." And it's just not a very big sample.
-
So what to do about it? One response is to dig for more data, and Reinhart and Rogoff have been doing that, going back to 1800 to examine episodes of public debt overhangs. Another is to have different people crunch it in different ways, which is what Herndon-Ash-Pollin did, or assemble different data sets, as several other scholars have done.
-
But the biggest challenge may be how to present it.
- ...1 more annotation...