というエントリをサイモン・レンールイスが書いている(原題は「Could austerity’s impact be persistent」)。

If I could carry just one message into mediamacro to bring it more into line with macro theory, it is that nominal interest at their lower bound represent a policy failure. Unconventional monetary policy is a very unreliable substitute for interest rate changes and fiscal policy as a way of controlling the economy, and a temporary fiscal stimulus can reliably get interest rates off their lower bound. This was the big mistake that most advanced countries made in 2010, and painfully slow recoveries were the result. The UK is currently making the same mistake, which is why the macroeconomic impact of the Labour and LibDem programmes is so much better than the Conservatives’ continuing austerity.
In the textbook macroeconomic models, this policy mistake can have a large but temporary cost in terms of lost output and lower living standards. This is because in these basic models a short term lack of demand does not have an impact on supply. Output in the longer run is determined by the number of those wanting to work, the capital stock and technology, all three of which are assumed to be independent of short term demand shortages. However it looks increasingly like these textbook models can be wrong.
In a new study (pdf), Gustav Horn and colleagues at the IMK institute in Germany looked at how persistent the impact of negative fiscal shocks (higher taxes or lower spending) had been on output. Their analysis is a refinement of earlier studies by of Blanchard and Leigh, and more recently Fatas and Summers. They find that the impact of recent fiscal shocks have been persistent rather than temporary, at least so far.
Although this persistent impact is not part of textbook models, economists have explored effects of this kind (the collective name for which is ‘hysteresis’). There are many theories about why it could happen, such as theories of endogenous growth.


各年マイナス幅(%) GDP縮小幅(%)
2011 -1 -1 -1 760
2012 -1 -2 -3 2,280
2013 0 -2 -5 3,800
2014 -0.43 -2.43 -7.43 5,647
2015 -0.43 -2.86 -10.29 7,820
2016 -0.43 -3.29 -13.58 10,321
2017 -0.43 -3.72 -17.3 13,148
2018 -0.35 -4.07 -21.37 16,241
2019 -0.3 -4.37 -25.74 19,562
2020 -0.3 -4.67 -30.41 23,112



I do not have to argue that such permanent effects are certain to have occurred. The numbers are so large that all I need is to attach a non-negligible probability to this possibility. Once you do that it means we should avoid austerity at all costs. In 2010 austerity was justified by imagined bond market panics, but no one is suggesting that today. The only way to describe current Conservative policy is pre-Keynesian nonsense, and incredibly harmful nonsense at that. That was why I signed the letter.

*1:cf. ここ