Where Rogoff-Reinhart went wrong
In retrospect, last week’s debunking of one of the key conclusions of Kenneth Rogoff and Carmen Reinhart about government debt looks inevitable. The whole story, from the initial lavish praise for the Harvard professors to the current harsh criticism, is a sad reminder of the power of ideology in the angry debate over economic policy.
In 2011, the two eminent professors claimed to show a tipping point for government borrowing. If the debt amounted to more than 90 percent of GDP, the GDP growth rate was typically much slower than in more fiscally prudent countries. When Thomas Herndon, a mere graduate student at the University of Massachusetts, redid the maths this year, he also found a correlation between higher government debt and slower growth. But there was nothing remotely like a tipping point.
The new paper was a blow to the politicians who relied on the Rogoff-Reinhart 90 percent line to support fiscal “austerity” (smaller government budget deficits). But they were always foolish to trust a study which drew a universal conclusion from a small sample of countries in vastly different situations.
Insofar as the Rogoff-Reinhart research had any value, it merely restated something that should have been obvious anyway: unsatisfactory economic performance and excessive government borrowing generally go together.
The connection is social, not financial. In societies that can get things done, respond well to challenges, compromise when necessary and do not spend more than is affordable, the government is likely to be fiscally competent and the economy is likely to be effective. Conversely, if the society is deeply divided, the economy is probably enfeebled and there is a high chance of a political deficit – to stay in power governments then need to spend more than they take in taxes.
The statistical analysis supports this correlation, but it’s really common sense, like the relationship between obesity and bad eating habits. The social analysis should serve as a warning to the austerity crowd. A balanced government budget is not going to restore an ill economy to good health or unify a divided society. The political deficit and social divisions will just appear elsewhere, perhaps taking the form of greater political instability.
The austerity-promoters would also do well to admit that large quantities of government borrowing and spending can be helpful, for example during and after wars, natural catastrophes and recessions. And the choice to raise funds by borrowing rather than through increased taxes is at least as much political as economic.
After all, debt and taxes are interchangeable in terms of cash flow, as long as taxpayers hold all government debt. The two financing mechanisms can be integrated, so the higher taxes needed to pay for higher debt loads are exactly compensated by the interest income and principal repayments taxpayers receive from the government. Government borrowing does not necessarily “crowd out” other economic activity any more, or any less, than an equivalent quantity of tax revenue.
It’s clear that Rogoff-Reinhart has methodological and theoretical problems, but their opponents, the advocates of “stimulus” (larger deficits), should be careful about gloating.
Pro-deficit economists cannot counter Rogoff-Reinhart with a persuasive historical study of their own, because peacetime deficits have almost never been as high a share of GDP as they are now. Deficit-doves often cite the U.S. Great Depression, but even if government spending reversed the 27 percent decline in GDP between 1929 and 1933, the precedent is not clearly relevant to the recent 4 percent decline.
The pro-deficit camp does have a plausible theory. Government deficit spending can make up for activity unnecessarily lost through an external shock, for example the 2008 financial crisis. But the stimulus crowd should admit the theory’s limits. The more the government spends, the more likely it is to spend foolishly, especially when the government suffers from the sort of large political deficit common in easily shocked economies. Spending financed by borrowing, or by newly created money, is no less likely to be wasted than spending financed by taxes.
Why did the implausible 90 pecent Rogoff-Reinhart debt threshold ever gain credence? And why do stimulus defenders ignore the dangers of ever larger governments? Because the austerity-stimulus debate is ultimately a battle in the ideological war over the proper role of government in society. Both sides fervently believe they are right: governments need to be restrained or governments need to be let loose. Both find supporting evidence everywhere and contrary evidence nowhere.
No facts will decide this argument, but the current mix of policies is particularly hard to interpret. For austerity fans, deficits are too high to count as truly austere; for their opponents they are too low to count as genuine stimulus. There will certainly be enough evidence to show that the other side’s approach has failed. And it’s a safe bet that the next Thomas Herndon will find easy pickings.