In defence of Reinhart and Rogoff
The public criticism Carmen Reinhart and Kenneth Rogoff have received is harsh. Even their detractors' figures show that higher debt is correlated with lower GDP growth
Academic economic papers rarely receive the sort of mass reception that brings coverage in the Guardian and the Telegraph so by the standards of its field, ‘Does High Public Debt Consistently Stifle Economic Growth? A Critique of Reinhart and Rogoff’, was something of a blockbuster.
The eponymous Carmen Reinhart and Kenneth Rogoff are economists who, in a 2010 paper, ‘Growth In a Time of Debt’, found that “whereas the link between growth and debt seems relatively weak at ‘normal’ debt levels, median growth rates for countries with public debt over roughly 90 percent of GDP are about one percent lower than otherwise; (mean) growth rates are several percent lower.”
These results, fleshed out to book length for the successful ‘This Time Is Different’, have been quoted by George Osborne, Paul Ryan, and Olli Rehn in support of their measures to get spiralling government debts under control.
Last week’s paper by Thomas Herndon, Michael Ash, and Robert Pollin of the University of Massachusetts Amherst claimed to have proved this wrong. They had recreated Reinhart and Rogoff’s results and found that the pair had reached their figure of a GDP ‘growth’ rate of -0.1 percent per annum for economies with government debt of over 90 percent of GDP thanks to “coding errors, selective exclusion of available data, and unconventional weighting of summary statistics.” Most embarrassingly, the ‘coding error’ was a proper schoolboy error; Reinhart and Rogoff missed some of the numbers out of their calculations.
In truth the idea that there was an Iron Law such that an economy would shrink as soon as it’s government debt hit 90 percent of GDP, the ‘strong form’ of Reinhart and Rogoff (pushed more by the political practitioners than them it ought to be said), was always iffy. It smacks of the sort of bogus causation derived from correlation which is the basis for much modern macroeconomics.
There are, for example, different types of debt. Advocates of higher spending often point to the 260 percent of GDP the British government owed in 1816, the 180 percent it owed in 1919, or the 220 percent it owed in 1945. This, they tell you, proves that Britain’s economy can bear an even greater burden of debt than the 70 percent of GDP it has doubled to in the last five years.
But you don’t have to be David Starkey to know that in 1816, 1919, and 1945 Britain had run up that debt to pay the cost of defeating a tyrant and as soon as that was done we stopped. It was an expense we had to meet and defray over time, the wartime borrowing was classic ‘consumption smoothing’.
To put it another way, when the British government started spending heavily in 1792, 1914, or 1939 there was a definite endgame for this spending: the restoration of the Bourbon monarchy, the defeat of the Kaiser, the overthrow of Hitler. The very moment those goals were accomplished spending would fall rapidly.
Our current level of government spending, by contrast, is not being undertaken to safeguard this country and its neighbours from conquest but to maintain a public sector and welfare state grown fat on borrowing and tax revenue from an unsustainable bubble in the absence of that bubble and those tax revenues. We are not smoothing consumption, we are sucking it out of tomorrow. And, unlike Pitt the Younger, Herbert Asquith, or Neville Chamberlain, present day advocates of higher spending cannot give an endgame for their proposed accrual of debt.
The point of this for evaluating Reinhart and Rogoff’s work is to note that one load of debt is not necessarily the same as another. There ought to be a little nuance to the picture, there are no magic numbers.
But even with that said it can still be argued that Reinhart and Rogoff have been hard done to this last week. They are, as they say in ‘This Time Is Different’, involved in the on-going process of growing their data set (which, rather unwisely, they have been quite proprietary about) and since 2010 they have revised their conclusions in the light of new data which Herndon, Ash, and Pollin had access to.
As Reinhart and Rogoff wrote in the Wall Street Journal, in a 2012 paper with Vincent Reinhart they found GDP growth rates of 2.4 percent for economies with government debt over 90 percent of GDP, pretty close to the 2.2 percent calculated by Herndon, Ash, and Pollin.
Indeed, and despite what some excitable commentators have proclaimed, Herndon, Ash, and Pollin have not found no correlation between high debt and low growth. They have found a weaker one than Reinhart and Rogoff in 2010 and about the same as they found in 2012, but they have still found one.
As page 21 of their paper states: if debt is below 30 percent GDP growth comes in at 4.2 percent, if debt is between 30 percent and 60 percent of GDP growth comes in at 3.1 percent, if debt is between 60 percent and 90 percent of GDP growth comes in at 3.2 percent, and if debt is over 90 percent of GDP growth comes in at 2.2 percent. Even on Herndon, Ash, and Pollin’s figures higher debt is correlated with lower GDP growth.
And there is perhaps a more profound point to note. Reinhart and Rogoff have fessed up to the coding error but the “selective exclusion of available data, and unconventional weighting of summary statistics” which Herndon, Ash, and Pollin accuse them of is, in fact, the very stuff of modern macroeconomics.
The ‘facts’ which dominate and guide our economic lives such as GDP, the CPI, or even unemployment, are not objectively given but are constructed using just such subjective methods, a prime example are the nonsensical unemployment figures of the United States. If this furore provokes a little scepticism so much the better, but it should spread much further than one paper.
John Phelan is a Contributing Editor for The Commentator and a Fellow at the Cobden Centre. He has also written for City AM and the Wall Street Journal Europe. He blogs at Manchester Liberal and Tweets @TheBoyPhelan
We are wholly dependent on the kindness of our readers for our continued work. We thank you in advance for any support you can offer.