A Flawed Landmark
A much cited paper contains some errors, but it doesn’t matter as much as you’d think.


When Reinhart and Rogoff took an even more expansive view than they did their original paper, and looked at data since the 18th century, they found numbers similar to the ones that their critics found with just post–World War II data. Other studies have also yielded similar results. So yes, there is a correlation between high debt and significantly slower growth, but there isn’t a “trigger” (a non-linear correlation) that deflates economic growth when government debt reaches a certain point.

But now that you understand the controversy and what the evidence actually says, here’s the bad news: It doesn’t matter that much.

For one, it never made that much sense to cite the Reinhart-Rogoff study as evidence that debt has crippling effects on economic growth, because the paper sets out a correlation, not a causation.

As to which way the causality works, it seems more likely that slow growth causes high debt than vice versa: It drives up social-welfare spending, depresses tax revenues, and makes the denominator of the debt-to-GDP ratio grow more slowly.

Debt and deficits, of course, do present an obstacle to economic growth. Almost every economist agrees that excessive borrowing eventually slows growth and crowds out private investment in a seriously harmful way — something that will be a problem on our current fiscal trajectory. High debt also limits the ability of a central bank to ease policy in a recession, which is probably even more important than the fiscal constraints it imposes. But the relationship will never be as clear as the Reinhart-Rogoff paper made it look, and there is no compelling empirical argument that the U.S. is nearing a growth cliff because of its debt, as the paper originally was taken to suggest.

So should those who supported austerity on the basis of Reinhart and Rogoff’s correlation be embarrassed? No more than liberal economic commentators who rail against austerity purely based on basically the opposite correlation. They claim that fiscal contractions have been devastating to European economies merely because there’s a correlation between austerity and slower growth — shoving aside the possibility that the worst-off European economies have had to implement the most contractionary fiscal policies.

It’s enough, really, to force both sides to remember the capricious nature of the social sciences — something many conservatives were rather good at until Kenneth Rogoff and Carmen Reinhart wrote a paper with a very convenient conclusion.

This shouldn’t be too wrenching, since the political battle lines were never drawn by the data anyway. American conservatives who want dramatic short-term deficit reduction really just want a much smaller government; if they actually do fear a debt crisis that cripples growth, presumably they will still believe it even when one piece of their empirical case has been weakened. Liberals who oppose almost any spending reduction because they believe government spending is justly and efficiently allocated will not win the argument because they found an Excel error. Most Europeans who favor austerity do so because they object to subsidizing their suine southern neighbors or oppose greater fiscal integration of the euro zone.

That said, could the weakening of a prominent piece of evidence — even if it never should have been all that useful — shift the battle away from debt hawks? At the margins, it might matter a bit. Stanford economics professor John Taylor wrote over the weekend suggesting that it already has: “A participant in last Friday’s G20 meetings told me that the [Reinhart-Rogoff] error was a factor in the decision to omit specific deficit or debt-to-GDP targets” in the meeting’s conclusions.

But what does Taylor himself still think? “The main arguments now for controlling the growth of spending and gradually bringing the U.S. federal budget into balance overpower any one study, right or wrong.”

– Patrick Brennan is a William F. Buckley Fellow at the National Review Institute.