Read on Twitter
*** What is automatic about the automatic stabilizers?
The discussion on the US extending the extra $600 in unemployment benefits has been couched in terms of automatic stabilizers. But if Congress needs to do something about it, how is this automatic? https://on.wsj.com/33qIhdj
The discussion on the US extending the extra $600 in unemployment benefits has been couched in terms of automatic stabilizers. But if Congress needs to do something about it, how is this automatic? https://on.wsj.com/33qIhdj
The CARES Act was a discretionary policy; why is repeating it now automatic? The Act by law expires now; isn’t extending it the opposite of automatic?
Like many other non-boring economic concepts, “automatic stabilizers” is used to mean different things https://bit.ly/2EIYi42
Like many other non-boring economic concepts, “automatic stabilizers” is used to mean different things https://bit.ly/2EIYi42
A strict definition of an automatic stabilizer is “The fiscal stabilizers are the rules in law that make fiscal revenues and outlays relative to total income change with the business cycle” The rules make eligibility or size depend on individual conditions
http://personal.lse.ac.uk/reisr/papers/16-autostab.pdf
http://personal.lse.ac.uk/reisr/papers/16-autostab.pdf
When the aggregate macroeconomy changes, individual conditions automatically change, and this affects fiscal revenues and outlays. When national unemployment rises, more people qualify for benefits. When national incomes rise, more people move up to higher tax brackets.
These stabilizers minimize policy discretion and errors. Ex ante, can set generosity or progressivity to balance macro goals with the their effects on social insurance and incentives to work. Without fine tuning, kicking in every time there is a recession
http://personal.lse.ac.uk/reisr/papers/20-optstab.pdf
http://personal.lse.ac.uk/reisr/papers/20-optstab.pdf
A less-automatic stabilizer would be “tying supplementary unemployment insurance and other support programs to the national unemployment rate or state unemployment rates.” Strictly speaking, this is a policy rule put into law to apply across recessions. https://brook.gs/33pKyWa
In theory, the rules would improve over the strict automatic stabilizers. One argument for a UI rule is that when unemployment is high, the disincentive effects are weaker (fewer jobs to find) and the macro benefits are higher (more idle resources).
http://econ.lse.ac.uk/staff/clandais/cgi-bin/Articles/aejoptimalUI_1.pdf
http://econ.lse.ac.uk/staff/clandais/cgi-bin/Articles/aejoptimalUI_1.pdf
Another argument is that knowing that the extra generosity of the safety net will be there in a recession, households will not save so much in fear of losing their job at the outset of recessions. These savings tend to amplify the downturns.
http://www.homepages.ucl.ac.uk/~uctpvst/HANK&SAM_analytical.pdf
http://www.homepages.ucl.ac.uk/~uctpvst/HANK&SAM_analytical.pdf
But, rule design is a subtle business: for instance, raising benefits at the onset of a recession but then soon cutting them, while still in a recession, may accelerate the recovery because it induces firms to offer jobs and workers to look for them. https://www.sciencedirect.com/science/article/abs/pii/S0304393214001664
A practical difficulty is that you have to tie the policy to a measure of the state of the economy. There are hard debates on how to measure how much slack there is in the labor market at any point in time. Generalized versions of Goodhart's law hit you. https://www.dropbox.com/s/b3z0khh1nte7x91/CEGS.pdf?dl=0
Of course, if policymakers get to change the measure every recession, then there is nothing automatic left. The best may be simple measures that are transparent. (But there is no perfect answer, simple measures can be thrown off by changes in trends)
https://www.hamiltonproject.org/assets/files/Sahm_web_20190506.pdf
https://www.hamiltonproject.org/assets/files/Sahm_web_20190506.pdf
A less-less-automatic stabilizer is to follow a rule, but not put it into law. A nice quote comes from Bernanke again: “Monetary policy should be systematic, not automatic.” The arguments he makes could be just as well applied to fiscal policy. https://www.brookings.edu/blog/ben-bernanke/2015/04/28/the-taylor-rule-a-benchmark-for-monetary-policy/
In rational-expectations economic models, there is only a slim distinction between a rule in the law, and a systematic benchmark to which policymakers usually stick to. But the distinction matters. Strategy may be a better word than rule (or automatic).
https://web.stanford.edu/~johntayl/2015_pdfs/House-HFSC-MPT-testimony-July-22-2015.pdf
https://web.stanford.edu/~johntayl/2015_pdfs/House-HFSC-MPT-testimony-July-22-2015.pdf
The US has been doing this for decades: every recession, the duration of unemployment benefits is extended. It is easy to predict this will happen again, and it stabilizes. But, every time, this is done differently, more or less. That limits its effect.
https://cpb-us-w2.wpmucdn.com/voices.uchicago.edu/dist/1/801/files/2019/06/ganong_noel_ui.pdf
https://cpb-us-w2.wpmucdn.com/voices.uchicago.edu/dist/1/801/files/2019/06/ganong_noel_ui.pdf
Finally are the less-less-less automatic stabilizers. Much of what I read in the press is whether to give an extra $600 to the unemployed for the next few months. Given current circumstances, this may have a large multiplier through household spending. https://institute.jpmorganchase.com/institute/research/labor-markets/unemployment-insurance-covid19-pandemic
If you think that aggregate demand is very depressed, and that interest rates are pegged by monetary policy, then the multiplier can be very large. Its stimulus benefits may exceed the disincentive of the high replacement rates, and justify the fiscal cost
https://sites.google.com/site/rohankekre/research/unemployment-insurance-in-macroeconomic-stabilization
https://sites.google.com/site/rohankekre/research/unemployment-insurance-in-macroeconomic-stabilization
But, this is pretty far from automatic. In fact, once you talk about multipliers, you are talking about the output response to a one-off policy *shock*. By definition these are not automatic. They are stimulus to the system.
https://econweb.ucsd.edu/~vramey/research/Ramey_Fiscal_JEP.pdf
https://econweb.ucsd.edu/~vramey/research/Ramey_Fiscal_JEP.pdf
Much of the current debate on (less-less-less) stabilizers is about a discretionary stimulus. It just happens to work through UI, which when untouched is automatic.
*That is fine*
Maybe discretionary stimulus is a good idea. Maybe it morphs into future automatic stabilizers.
*That is fine*
Maybe discretionary stimulus is a good idea. Maybe it morphs into future automatic stabilizers.
Recap:
1) an automatic stabilizer is in law and depends on household variables, which then vary with the business cycle
2) a policy rule is in the law and depends on macro variables
3) a policy strategy is not in the law, but in usual behavior
4) a discretionary policy is a shock
1) an automatic stabilizer is in law and depends on household variables, which then vary with the business cycle
2) a policy rule is in the law and depends on macro variables
3) a policy strategy is not in the law, but in usual behavior
4) a discretionary policy is a shock
Finally, a disclaimer: matters of definitions and the usage of terms, surely have some people who disagree with my categories above.
More important is to recognize that which of these 4 kinds it is will matter *a lot* for the effects of the policies. No matter what you call them
More important is to recognize that which of these 4 kinds it is will matter *a lot* for the effects of the policies. No matter what you call them
And for a fun ending to illustrate how everyone says they love automatic stabilizers, but then use the term to mean different things, you can read two giants of economics:
Friedman: https://miltonfriedman.hoover.org/friedman_images/Collections/2016c21/AEA-AER_06_01_1948.pdf
Solow: https://econpapers.repec.org/article/oupoxford/v_3a21_3ay_3a2005_3ai_3a4_3ap_3a509-514.htm
Friedman: https://miltonfriedman.hoover.org/friedman_images/Collections/2016c21/AEA-AER_06_01_1948.pdf
Solow: https://econpapers.repec.org/article/oupoxford/v_3a21_3ay_3a2005_3ai_3a4_3ap_3a509-514.htm
