Daily Archives: September 30, 2013

Chris Giles

On Thursday evening, I put out the following tweet after reading Martin Wolf’s trenchant column on austerity.


It was a throwaway remark relating to an academic paper by Òscar Jordà of the Federal Reserve Bank of San Francisco and University of California, Davis and professor Alan Taylor, also of UC, Davis. I will confess that I had not read the paper when I tweeted, but had read this summary on Vox.eu.

The paper has gained celebrity status among economists and commentators who like to criticise the UK government’s fiscal policy. Martin Wolf has quoted it repeatedly, as has Jonathan Portes, director of the National Institute of Economic and Social Research and many others including professor Simon Wren Lewis of Oxford University.

Quite properly, Prof Taylor emailed me on Friday to ask why I was so “deeply unpersuaded” and asked me to account for my words. This has sparked a lively email exchange between us, which I think has informed both sides and certainly forced me to read the full paper closely. The emails are private, but I think the paper has gained such traction that my concerns deserve to be written down. Some of them were prompted by this excellent blog by David Smith, economics editor of the Sunday Times.

In a nutshell, while I think the new estimator in the paper is clever, its application to the UK as an example is wrong. The upshot is that the headline result — that three fifths of the UK’s growth disappointment is caused by austerity — is not supported by the evidence in the paper.

I am not claiming a “Reinhart and Rogoff” moment, and nor is David Smith to my knowledge, but it does serve to illustrate that understanding your data is as important as getting the right econometric estimator and that I am surprised that so many UK academics did not spot the causes for concern. If they did, they chose not to highlight them.

A separate point is that when academics have been using the paper as a weapon to bash the Treasury, they have not looked at the size of the effect Jordà and Taylor are claiming. They claim austerity from 2010 to 2013 has left the level of gross domestic product 3 per cent lower than it would have been had there been no consolidation at all. This is not a large number and very close to the view that the Office for Budget Responsibility would get using its modest fiscal multiplier. It estimated output was already 1.4 per cent below the level of “no austerity” in 2011-12 after which there were two more years of deficit reduction pain implemented.

So, the result, even if true, does not prove deficit reduction to be misguided. But that is for another time, the issue here is what is wrong with the claim that three-fifths of weakness is caused by austerity and that austerity has hit ouptut by 3 per cent. The short answer is “a lot”. Read more