A Dose of Reality for the Dismal Science
Op-ed in the Financial Times
© Financial Times
For a few years, advocates of rapid fiscal austerity have argued as though public debt is like a black hole—once it reaches a certain size, it collapses in on itself under its own weight and pulls the economy down with it. Crisis awaits the spendthrift. A 2010 academic paper by Carmen Reinhart and Kenneth Rogoff, two eminent economics professors, provided many pundits and politicians with the desired evidence for this instinctive view. They seemingly found that economic growth fell off sharply when national debts reached 90 percent of gross domestic product.
The claim that there was a clear tipping point for the ratio of government debt to GDP past which an economy's walls caved in never made any sense.
A new study that has attracted lots of attention, however, shows that no such sharp fall-offs occur. Put aside the details of Excel coding errors and statistical weights. In fact, forget that specific paper. The claim that there was a clear tipping point for the ratio of government debt to GDP past which an economy's walls caved in never made any sense. If such a critical level were to hold, there would have to be some equally abrupt causal mechanism by which the dire predictions for growth would have to come to pass —perhaps interest rate rises, a currency crisis or an increase in hoarding and saving resulting from feared future tax rises? Such an event would be clearly visible in the data among those countries that went past the debt event horizon of 90 percent. But it is not there.
A casual perusal of 20th century economic history, let alone more rigorous econometric analysis, turns up multiyear periods in the United Kingdom and United States following the second world war, and in Belgium, Italy, and Japan in the past 20 years, when public debt was greater than 90 percent of GDP but nothing much happened. Either stagnation in economies led to slowly rising debt levels, as in Italy or Japan of late, or growth returned and debt levels declined, as in the United Kingdom and United States in the 1950s. The latter two escaped the black hole of debt without an austerity rocket booster.
In other words, slow growth is at least as much the cause of high debt as high debt causes growth to slow—which if you stop to think about it, as some of us did before this week, makes more sense. The current UK economy is exhibit A for such a dynamic, with its slowing growth and mounting debt. On the other hand, public debt is sometimes incurred by spending on constructive things with a positive return, such as infrastructure and education. So it is common sense that slow growth is always bad for debt accumulation, but not all debt accumulation is bad for growth. Thus, the causality runs more dependably from growth to debt than vice versa.
Some would like to turn the debunking of the fiscal event horizon claim into a cautionary tale about macroeconomic policy advice in general. They would throw up their hands, saying macroeconomic analyses either inherently depend upon too little data to have reliable results, or inevitably will be selectively picked up by ideologues and opportunistic politicians to suit their purposes.
Perhaps both: There will always be some willing economist who can play with the data to provide a credible-seeming study to support any given politically influential point of view. This, however, is far too defeatist, if not craven, a conclusion to draw.
Yes, I have some scars from having had to fight policy battles over and over—against spurious claims about expansionary budget consolidations, low or zero fiscal multipliers and the imminent threat of inflation during recession. I have heard claims from the 1930s repeated in Japan in 1999, in Britain in 2010, and again in the euro area today.
Certainly, academics who oversell their conclusions to get their lines into a speech or to come up with a rule that can be named after them can become disproportionately influential. And the academic establishment does not help matters by encouraging the production of new and counterintuitive results irrespective of their validity in practice. It also disincentivizes the replication and empirical testing of previously published work. Yet, eventually, the truth will out—as it did this week.
Every time that the truth does emerge from the data, a few more researchers and officials find something lasting. This process of intellectual attrition took a long while to convince policymakers that trade protectionism is harmful, and that adherence to the gold standard is ill-advised. But we have now reached a state of public debate where only the obviously self-interested advocate the first, and only the obviously loony defend the second. Perhaps, as the mistakes of the current UK and euro area economic leaderships become more evident, discussed, and analyzed, similar progress will be made in limiting policy discussion to the more reasonable options.
It is a victory for common sense and good policy that the International Monetary Fund has publicly decided to reverse its past mistakes and come out clearly for sensible fiscal approaches—not least recognizing the impact on growth of cutting the deficit and that reducing public debt is a task that should primarily occur once countries are out of recession.
The moderate middle ground between panicked austerity and heedless spending is the right place for policy advisers to be. Such moderation will not suit many of those who seized upon the event-horizon hypothesis—that result's attraction was that it suggested a crisis was around the corner as a spur to consolidation.
But let us celebrate rather than mourn what this reevaluation of the evidence demonstrates, even though most already should have known it: Too much public debt has its costs for growth, but the extent of those costs depends on the reasons the debt accumulated and the trajectory of the economy. And it is not worth provoking a crisis to forestall a crisis that is unlikely to come, for black holes are rare and not of this Earth.