Logic Deficit: Why Were Reinhart-Rogoff Ever Taken Seriously?
The controversy continues to simmer around the Reinhart-Rogoff (RR) paper and the now famous Excel spreadsheet error that led to claim that debt-to-GDP ratios above 90 percent led to sharply lower growth rates. The University of Massachusetts paper that exposed this mistake has led many people to reconsider their earlier acceptance of the Reinhart-Rogoff 90 percent debt cliff.
While that is a positive development, the re-examination should go a step deeper and ask why anyone ever took their argument seriously in the first place. It’s not just the arithmetic on debt-to-GDP ratios that tripped up RR; it was the basic logic of their argument.
If we accept the RR thesis, something bad happens to countries when their debt-to-GDP ratio exceeds 90s percent, which causes them to experience prolonged periods of slow growth. It is difficult to see how this could possibly be the case since debt is only one side of a country’s balance sheet, countries also have assets. For there to be any actual relationship between debt and growth it would seem that it would have to be debt, net of assets, and growth.
The RR story, where they purport to find a relationship between debt and growth, would be like finding a relationship between household debt and future income, without considering whether or not they own a home. In the RR approach to family finances, a family with $200,000 in debt who rented their apartment would be viewed as being in the same situation as a family with a $200,000 mortgage on a house that is worth $500,000.
While the second family would have $300,000 in assets after deducting their mortgage, like the first family it would also have $200,000 in debt. And since RR only concern themselves with debt, both families would count as facing equal debt burdens.
It would of course be absurd to imagine that these two families are in the same situation financially. And it would be equally absurd to imagine that countries with similar debt burdens but different amounts of assets are in the same position.
Suppose that the United States and Belgium both had debt-to-GDP ratios of 100 percent. This would put them in the same column in the RR table. However the United States has a vast amount of valuable land and an enormous coast line with fishing rights that it could in principle sell. It also has massive reserves of oil and gas, much of it on public land that could be sold.
It could even sell the rights to tax revenue. That should not sound far-fetched. The proposal to issue carbon permits in a cap and trade system essentially amounted to selling off (or giving away) the right to collect a carbon tax.
The government could issue permits to emit carbon over the next two decades and auction them off all at once. Depending on the levels set, such an auction could easily raise several trillion dollars, knocking 10-20 percentage points off the country’s debt to GDP ratio. If anyone believed that a too high debt to GDP ratio was a major drain on growth then it is hard to see why we would not want to go this route.
While Belgium surely has some assets to sell, and it could probably engage in a similar sale of a right to collect future taxes, relative to its GDP it almost certainly would have a far smaller stock of assets than the United States. (Belgium’s revenue is already more than 50 percent of GDP.) This means that the RR analysis would count the U.S. and Belgium as being in the same position, even though the United States effectively has a house to offset its liabilities, while Belgium does not.
There are good reasons not to sell off assets like land or mineral rights, but if anyone really believed in the RR cliff then such asset sales would pay enormous dividends. If we took RR’s uncorrected numbers, which implied a growth penalty of more than 2 percentage points to countries that crossed the 90 percent threshold, a $1 trillion sales of assets could lead to a growth dividend of more than $4 trillion in the first decade alone. This $4 trillion growth dividend is in excess to getting the $1 trillion market value of the assets themselves.
No part of this story makes any sense. No one, not even a Harvard economist, really believes that a country with no assets and a country with a huge amount of assets are in the same financial situation just because they face the same debt burden. Nor does anyone believe that there is the potential massive growth dividend available from selling off government assets that the RR analysis implied.
In short, RR’s analysis never was plausible, even if their spreadsheet had been correct. The fact that so many people in positions of authority were willing to embrace RR says a great deal about these people. We can and should discard RR’s outlandish claims about the impact of debt on growth. We should also be asking some serious questions about our elites and their willingness to accept economic nonsense just because it fits their agenda.
Dean Baker is the co-director of the Center for Economic and Policy Research (CEPR). He is the author of The End of Loser Liberalism: Making Markets Progressive. He also has a blog, "Beat the Press," where he discusses the media's coverage of economic issues.