Reinhart-Rogoff: a structure of knowledge goes down

An important structure of knowledge went down last week. The Reinhart-Rogoff paper “Growth in a Time of Debt,” which claimed to show that real GDP growth is weak when government debt hits a “threshold” of 90% of GDP, was taken down by a grad student doing what sounds like a normal seminar assignment. The assignment: replicate a paper. The student chose to replicate the Reinhart-Rogoff paper; amazingly was given access to the data; tried to replicate it; found obvious errors and questionable decisions; and now he, Thomas Herndon, is an intellectual star.

“Above 90 percent, median growth rates fall by one percent, and average growth falls considerably more,” Reinhart and Rogoff had concluded. After Herndon, this finding appears dubious.

Others have documented the extent to which this paper — and this finding — enjoyed a prominent and established position in the knowledge hierarchy, especially for an economic paper that had never even been peer-reviewed (see Konczal, Dube, O’Brien, Lowrey, Bright. And for something more in Herndon’s own words, see here and here). For example, Krugman wrote:

The intellectual edifice of austerity economics rests largely on two academic papers that were seized on by policy makers, without ever having been properly vetted, because they said what the Very Serious People wanted to hear. One was …. Reinhart/Rogoff on the negative effects of debt on growth. Very quickly, everyone “knew” that terrible things happen when debt passes 90 percent of GDP.

Taking off from Krugman, the fall of this knowledge structure is important because the movement to solve the US debt problem appears to be the same movement that has long wanted to shrink and shrink and shrink the US government until it’s broke and domestically impotent. The Reinhart-Rogoff paper represented a structure of knowledge this movement used in the struggle to further this recycled interest. The paper didn’t create the interest, or make the interest more likely to succeed. But it helped give the interest an empirical right to exist, a sense of legitimacy. Otherwise, the idea to radically cut government spending during a time of little or slow growth might simply be known as economic suicide.

Not surprisingly, I admire Martin Wolf’s take for its clarity and respect. I especially like how Wolf uses Reinhart and Rogoff’s earlier book — This Time is Different — to argue against the authors’ later paper. Here is a longish excerpt from Wolf in the Financial Times (no copyright infringement intended, just appreciate the work being published here):

The recent critique by Thomas Herndon, Michael Ash and Robert Pollin of the University of Massachusetts at Amherst makes three specific charges against the conclusions of professors Reinhart and Rogoff: a simple coding error; data omissions; and strange aggregation procedures. After correction, they argue, average annual growth since 1945 in advanced countries with debt above 90 per cent of GDP is 2.2 per cent. This contrasts with 4.2 per cent when debt is below 30 per cent, 3.1 per cent when debt stands between 30 per cent and 60 per cent, and 3.2 per cent if debt is between 60 per cent and 90 per cent. In their response, professors Reinhart and Rogoff accept the coding errors, but reject the critique of aggregation. I agree with the critics for reasons given by Gavyn Davies. The argument that data covering a long period of high debt should count for more than data covering a short one is persuasive.

Nevertheless, their work and that of others supports the proposition that slower growth is associated with higher debt. But an association is definitely not a cause. Slow growth could cause high debt, a hypothesis supported by Arindrajit Dube, also at Amherst. Consider Japan: is its high debt a cause of its slow growth or a consequence? My answer would be: the latter. Again, did high debt cause today’s low UK growth? No. Before the crisis, UK net public debt was close to its lowest ratio to GDP in the past 300 years. The UK’s rising debt is a result of slow growth or, more precisely, of the cause of that low growth – a huge financial crisis.

Indeed, in their masterpiece, This Time is Different, professors Reinhart and Rogoff explained how soaring private debt can lead to financial crises that generate deep recessions, weak recoveries and rising public debt. This work is seminal. Its conclusion is clearly that rising public debt is the consequence of the low growth, itself explained by the crisis. This is not to rule out two-way causality. But the impulse goes from private financial excesses to crisis, slow growth and high public debt, not the other way round. Just ask the Irish or Spanish about their experience.

It follows that, in assessing the consequences of debt for growth, one must ask why the debt rose in the first place. Were wars being financed? Was there fiscal profligacy in boom times, which is almost certain to lower growth? Was the spending on high-quality public assets, conducive to growth. Finally, did the rise in public debt follow a private sector financial bust?

Different causes of high debt will have distinct results. Again, the reasons why deficits are high and debt rising will affect the costs of austerity. Usually, one can ignore the macroeconomic consequences of fiscal austerity: either private spending will be robust or monetary policy will be effective. But, after a financial crisis, a huge excess of desired private savings is likely to emerge, even when interest rates are very close to zero.
In that situation, immediate fiscal austerity will be counterproductive. It will drive the economy into a deep recession, while achieving only a limited reduction in deficits and debt.

Moreover, as the International Monetary Fund’s Global Financial Stability Report also notes, extreme monetary stimulus, in these circumstances, creates substantial dangers of its own. Yet nobody who believes in maintaining fiscal support for the economy in these specific (and rare) circumstances thinks that “fiscal stimulus is always right”, as Anders Aslund of the Peterson Institute for International Economics, suggests. Far from it. Stimulus is merely not always wrong, as “austerians” seem to believe.

This is why I was – and remain – concerned about the intellectual influence in favour of austerity exercised by Reinhart and Rogoff, whom I greatly respect. The issue here is not even the direction of causality, but rather the costs of trying to avoid high public debt in the aftermath of a financial crisis. In its latest World Economic Outlook, the IMF notes that direct fiscal support for recovery has been exceptionally weak. Not surprisingly, the recovery itself has also been feeble. One of the reasons for this weak support for crisis-hit economies has been concern about the high level of public debt. Professors Reinhart and Rogoff’s paper justified that concern. True, countries in the eurozone that cannot borrow must tighten. But their partners could either support continued spending or offset their actions with their own policies. Others with room for manoeuvre, such as the US and even the UK, could – and should – have taken a different course. Because they did not, recovery has been even weaker and so the long-run costs of the recession far greater than was necessary. This was a huge blunder. It is still not too late to reconsider.

[bold added]

This entry was posted in an actually thriving labor market, conservative movement, debt, economic recovery, macro-economics, the great contraction, the great contraction 2007-2012. Bookmark the permalink.

One Response to Reinhart-Rogoff: a structure of knowledge goes down

  1. Pingback: A second post on Reinhart-Rogoff | price of data

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s