Òscar Jordà of the San Francisco Fed and Alan M. Taylor of the University of California-Davis have a doozy of a working paper out on the macroeconomic effects of austerity. The chart above has the money stat: According to their numbers, the U.K.'s experiment with austerity starting in 2010 led to a 3 percent reduction in growth. If true, that's a big, big indictment of the Cameron government's policies.
Of course, you should always take macroeconomic studies that try to sort out the effects of government taxes or spending with a grain or two of salt. It's an area where experiments are impossible, and all the plausible alternatives fall short in some way. You can make conclusions based on a large number of studies on the same topic, but reading too much into any one paper is, as the Reinhart-Rogoff saga taught us, foolhardy.
All that being said, the Jordà/Taylor paper is notable. Generally, empirical papers on this subject take one of two approaches. The first, known as vector auto-regression (VAR) or structural vector auto-regression (SVAR), uses statistical techniques to infer when a shock to fiscal policy (e.g. a big cut in spending or increase in taxes) has occurred.
The second, known as the "narrative" approach and pioneered by UC Berkeley's David and Christina Romer, identifies shocks through news coverage and policy announcements. That avoids the possibility that a VAR approach will pick up shocks that aren't there, or fail to notice ones clearly announced by the government. On the other hand, it could mistake empty statements by government officials that don't correspond to spending patterns for real changes in policy, or miss minor unannounced policy changes. The two approaches tend to give very different answers, with VAR suggesting that austerity is less harmful to an economy than narrative approaches suggest.
Jordà and Taylor introduce a new method altogether, known as local projections / inverse-probability-weighted regression adjustment (LP-IPWRA), which has been used in medical research and applied microeconomics but not so much in macroeconomics to date. The approach is pretty complicated and way above my pay grade in any case. But it's designed to avoid the pitfalls of both narrative and VAR approaches by matching countries with similar economic characteristics, identifying which nations pursued austerity and which didn't. Then, when estimating the effects of austerity, they can control according to the nations' relative likelihood of pursuing austerity, given their economic situations.
They find that not only is austerity during downturns quite contractionary, but it even hurts growth during booms. "Our results underscore that austerity tends to be painful, but that timing matters," Jordà and Taylor write. "The least painful ﬁscal consolidations, from a growth and hence budgetary perspective, will tend to be those launched from a position of strength, that is, in the boom not the slump."
They then apply the findings to the specific case of the U.K., producing the chart you see above. But it could be even worse than that. "This caveat is the zero lower bound (ZLB) of monetary policy: the U.K. out-of-sample counterfactual corresponds to a liquidity trap environment, but the in-sample data overwhelmingly do not," the authors write. "In that case, the true residuals in 2010–13 could be much smaller than above, and the effects of austerity (i.e., the ﬁscal multipliers) even bigger, big enough to possibly explain most or all of the growth shortfall after 2010."
In plain English, what they're saying is that when central banks' interest rates have gone as low as they can go, fiscal policy is much more powerful than it is usually is. In this case, the Bank of England could not readily offset the impact of fiscal tightening by cutting interest rates, which means that they did more damage to growth. The countries Jordà and Taylor included in their sample mostly weren't in that situation, meaning they could be lowballing the damage that the austerity policy did to Britain.
Again, this paper is not definitive. And the newness of its methodology, while exciting, also opens the door for error. But if its results hold, the implications for the Cameron/Osborne government are extremely damning.