Some pretty interesting information has come to light on the research performed by Reinhart and Rogoff on the relationship between growth and government debt levels. Briefly, R&R’s 2010 paper,titled Growth in a Time of Debt, examined the relationship of public-sector debt to economic growth. Their analysis showed average growth sharply lower above a public-sector debt level at 90% of GDP. (U. S. public debt is now at 106% of GDP.)
It turns out that the two authors made a spreadsheet error in their calculations. Instead of debt above 90% being associated with slight economic contraction (growth rate of -0.1%), the data support a reduced, but still positive, growth rate of +2.2%. A pretty significant difference. (Note that they reported a median growth rate of +1.9%. We often catch our own analytical oversights when medians deviate significantly from averages.)
The issue of accuracy in a three-year-old academic paper is important, in large part because their analysis has been used by conservative politicians and commentators to argue against additional borrowing, and for more aggressive cutbacks in spending to achieve budget balance.
Yesterday, Reinhart and Rogoff made a sort of non-apology apology in The New York Times. Krugman et al have been all over the authors, arguing that the pro-austerity argument has now been entirely exploded.
It may be worth noting that Rogoff was the Chief Economist for the International Monetary Fund from 2000 to 2003. The IMF’s principal job is to help countries rescue their economies when public finances blow up. What causes such blowups? Pretty reliably, large public-sector deficits that trigger unaffordable borrowing costs. Once you can no longer service your debts, you end up in the not-so-sheltering embrace of the IMF, which requires budget-balancing measures as one of the conditions for providing emergency funding.
We know that water runs downhill, and we know that enough water risks drowning you. Reinhart and Rogoff suggested that we have quite an accurate measure of when the water level reaches our necks. The recalculated data suggest the debt danger point is less precisely quantifiable than economists would wish; it is much more a continuum than a precise inflection point. That does not change the historical observation that over-spending and over-borrowing, carried far enough, usually leads to be unpleasant consequences.
My concern remains that debt crises arrive without warning, even more so now that the Reinhard/Rogoff 90% tripwire has sharply-diminished predictive credibility. Once a debt crisis hits, as it has surely done for dozens of nations over the last several hundred years, you are left without attractive policy options, and with no serious alternative to the most brutal austerity.