Monday Smackdown: George Borjas P-Hacking His Way Along...

An interesting paper saying that Glick and Rose's findings are not robust. I am generally pro-customs unions. I was taught when I was knee-high to a grasshopper that the Zollverein was a big deal. And I have always been impressed by the scale of cross-state trade in the U.S., which dwarfs cross-nation trade within Europe. But I may have to rethink—and I believe I certainly have to revise up my beliefs about how precisely these effects can be estimated: Douglas L. Campbell and Aleksandr Chentsov: Breaking Badly: The Currency Union Effect on Trade: "A key policy question is how much currency unions (CUs) affect trade...

...Glick and Rose (2016) confirmed that currency unions increase trade on average by 100%, and that the Euro has increased trade by a still-large 50%.... We find that the apparent large impact of CUs on trade is driven by other major geopolitical events correlated with CU switches, including communist takeovers, decolonization, warfare, ethnic cleansing episodes, the fall of the Berlin Wall and the whole history of European integration. We find that moving from robust standard errors to multi-way clustered errors alone reduces the t-score of the Euro impact by 75%. Looking at individual CUs, we find that in no cases does the time series evidence support a large trade effect, and that the effect breaks particularly badly once we find suitable control groups. Overall, we find that intuitive controls and omitting the CU switches coterminous with war and missing data render the trade impact of the Euro and all CUs together statistically insignificant...


#shouldread

Comments