[Before beginning, let me emphasize that this post is not discussing the use of government policy to provide income assistance for the unemployed. My focus is on the use of fiscal policy to boost aggregate demand.]
During the current epidemic, we are seeing an avalanche of articles advocating the use of fiscal stimulus. At times the media/twitter/blogosphere gives the impression that only a fool would fail to see the rationale for giving every American $1000. In fact, the evidence favoring fiscal stimulus is almost non-existent. I’ve discussed these issues before, but there is obviously a need to remind people of why fiscal stimulus is unlikely to work. Like Sisyphus, I have to keep making the same points, over and over again.
As recently as 2007, the consensus among macroeconomists was that monetary policy should be used to control inflation, and that fiscal stimulus was ineffective for all sorts of reasons. That consensus changed after 2007, but there isn’t any new evidence that would justify a change.
We’ve had 4 major fiscal policy shifts since 2007, and all four completely failed to have the impact predicted by Keynesian economists. Let’s examine each one of them.
1. In early 2008, the economy had fallen into a very mild recession. During the spring, President Bush got Congress to agree to a program of sending $300 checks to each person (excluding people earning less than $3000, and also excluding those who were upper middle class or rich.) Very soon after the checks were sent out, the economy fell off a cliff.
Now it’s true that the rebates probably boosted consumption and GDP slightly in 2008:Q2, but the Fed responded by tightening policy and hence by the second half of 2008 any beneficial impact was more than neutralized by monetary offset.
2. In early 2009, Congress passed President Obama’s $800 billion ARRA program. It was claimed that this would prevent unemployment from rising above roughly 8%, and that unemployment would rise to 9% without the program. Actually unemployment rose to 10%, with the program.
Obviously I cannot prove that the program had no impact, ceteris paribus;Â but does anyone seriously believe that the Fed would not have been more aggressive in the absence of the ARRA program?
3. At the very end of 2012, Congress sharply tightened fiscal policy. The deficit fell from $1050 billion in calendar 2012 to $550 billion in calendar (not fiscal) 2013. Roughly 350 Keynesian economists signed a letter suggesting that the fiscal austerity would slow the economy, and risked pushing us into recession. In fact, both NGDP and RGDP growth increased, and by 2013:Q4, the 12-month growth rate was considerably higher than in 2012:Q4, right before the austerity.
The explanation is simple. The Fed saw the fiscal austerity coming and rolled out some new monetary stimulus programs, including QE3 and a more aggressive form of forward guidance. The Fed successfully offset the fiscal austerity.
4. Between 2015 and 2019, the fiscal deficit exploded from $442 billion to $984 billion. This is not supposed to happen during a period when unemployment is falling to low levels. Keynesian models predict that all this unneeded fiscal stimulus during a period of low unemployment would push the inflation rate above 2%, and thus the Fed used “Phillips curve models” as a rationale to boost interest rates and prevent a rise in inflation. But the inflation never came, as fiscal policy doesn’t determine the rate of inflation, monetary policy does. Inflation actually undershot the 2% target.
And this leads to an important point that people miss. If fiscal policy doesn’t determine inflation, then it does not have a significant impact on aggregate demand (NGDP.) That doesn’t mean fiscal actions cannot affect RGDP; they can. But they do so by boosting aggregate supply, not demand.
If the fiscal authorities give every American $1000 it will be like throwing a stone in the ocean. It doesn’t create any new money, it just moves money around. (It doesn’t boost the supply side either.) The fiscal authorities borrow $1000 from John and give the $1000 to Jack. The Fed determines the rate of inflation through monetary policy, the creation of new money.
Just to be clear, I’m not saying it is impossible that fiscal policy could have some impact. You can conceive of a case where fiscal actions cause the central bank to adopt a different monetary policy, cause it to produce a different inflation rate. Central banks are often incompetent.
But the last 12 years gives us little reason for hope. We’ve done four major fiscal operations, and each one had an impact that was quite different from what was expected by Keynesian policymakers.
It’s time to focus on monetary stimulus, which is an order of magnitude more powerful than fiscal, is far less costly, and can be deployed much more quickly. Start by switching to level targeting (of prices, or better yet NGDP), and then adopt a “whatever it takes” approach to asset purchases. That’s the only way to meaningfully boost aggregate demand.
Speaking of Sisyphus, here’s Titian’s magnificent version:
READER COMMENTS
Dylan
Mar 21 2020 at 3:06pm
Scott,
Did you see the piece by Yellen and Bernanke in the FT on March 18th? I’d put a link to it, but that tends to get comments sent to purgatory for a bit. Was hoping to see your take on their proposals.
Scott Sumner
Mar 21 2020 at 5:56pm
I did a post on it.
Dylan
Mar 21 2020 at 6:37pm
That’s right. I forget sometimes that you have another blog. Off to Money Illusion to get better informed. Thanks.
Iskander
Mar 21 2020 at 3:45pm
How does fiscal policy affect NGDP except through the velocity of money? It could if it causes interest rates to rise but aside from that does it do anything?
Scott Sumner
Mar 21 2020 at 5:57pm
Perhaps it will also somehow encourage the Fed to print more money. I’m skeptical, but it’s possible.
Matthias Görgens
Mar 22 2020 at 5:25am
Fiscal policy can do all kinds of things. But the monetary policy ‘moves last’, ie can offset any fiscal policy.
If your monetary policy isn’t some variant of inflation, price level or ngdp targeting, fiscal policy can have an effect.
Eg during the 1930s Japan’s monetary policy was to keep government debt at 1% interest rates. That meant that monetary policy amplified fiscal policy instead of offsetting it: as the government went more into debt, the central bank bought more of its debt in exchange for new money.
Iskander
Mar 22 2020 at 4:51pm
I agree, I think you could get the same “fiscal policy” effects with a central bank pegging an exchange rate and a government borrowing to pay for large deficits. I guess talking about the effects of fiscal policy on NGDP/AD is almost meaningless without specifying the monetary regime.
Market Fiscalist
Mar 21 2020 at 3:52pm
If the aim of monetary policy is to keep the size of the money supply at the optimal level then what real difference does it make if the new money supply is adjusted  by buying and selling bonds or just by giving money away and taxing it back ?
Of course in normal times buying bonds is almost certainly the best option but in abnormal times when the monetary authority declines to use its full armory (or if the ‘it won’t work at the zero-lower bounds’ monetary policy skeptics are right) mailing checks may be the best option available.
Scott Sumner
Mar 21 2020 at 5:58pm
“Taxing it back” is very costly. We don’t know how the Fed would respond to fiscal stimulus, so it’s foolish to claim with confidence that it will boost inflation.
Market Fiscalist
Mar 21 2020 at 7:16pm
If the fed controlled the fiscal initiative as outlined here:
https://www.mercatus.org/publications/monetary-policy/covid-19-pandemic-direct-cash-transfers-and-federal-reserve
that would not be an issue
Scott Sumner
Mar 21 2020 at 11:12pm
What won’t be an issue?
Market Fiscalist
Mar 22 2020 at 10:31am
I was responding to your statement that ‘We don’t know how the Fed would respond to fiscal stimulus, so it’s foolish to claim with confidence that it [helicopter money] will boost inflation.’.
Actually I’d be generally curious about your response to the Beckworth article.
Scott Sumner
Mar 22 2020 at 12:46pm
How the Fed responds is always an issue. I can’t understand why you say it won’t be an issue.
Which part of the Beckworth article do you want me to respond to?
Market Fiscalist
Mar 22 2020 at 1:17pm
‘Step 3: The Fed Gets a Standing Fiscal Facility’
starts with ‘The final part of the operating framework would establish a standing fiscal facility for the Fed to use when doing helicopter drops. ‘.
I was figuring you would disagree with Beckworth on this part.
Mark Z
Mar 21 2020 at 3:52pm
I think this is a clear case of people (including some Keynesian economists) using Keynesian terminology to rationalize a policy they actually favor for completely different reasons. Greg Mankiw at least is fairly straightforward about the case for giving away money:
“Fiscal policymakers should focus not on aggregate demand but on social insurance. Financial planners tell people to have six months of living expenses in an emergency fund. Sadly, many people do not. Considering the difficulty of identifying the truly needy and the problems inherent in trying to do so, sending every American a $1000 check asap would be a good start.”
Basically, the actual argument isn’t really about fiscal stimulus, but retroactive social insurance. Though I disagree that a universal giveaway is a good idea. I stand to get a big check I don’t need and all you’d have to do is look at my tax returns to know I don’t need it. I’m not going to send it back of course.
Scott Sumner
Mar 21 2020 at 6:00pm
Even if unemployment rises to 20%, then 80% will still have jobs. Why give them $1000? Why not give the unemployed $5000?
robc
Mar 21 2020 at 6:15pm
Moral hazard?
Scott Sumner
Mar 21 2020 at 11:13pm
OK, maybe less than $5000. But there’s no reason to give it to the employed.
Shyam Vasudevan
Mar 21 2020 at 5:19pm
I think the argument for universal $1000 checks is to help people meet nominal debt contracts, not for stimulus. In the absence of a central bank that does whatever it takes to keep spending on its growth path, this will likely be necessary to prevent a steep contraction.
Scott Sumner
Mar 21 2020 at 11:13pm
Why do people who are still working need help repaying debts?
Shyam Vasudevan
Mar 22 2020 at 3:10pm
Making it universal reduces administrative burden and makes the policy seem more equitable. We don’t means-test aid after natural disasters.
Mark Z
Mar 22 2020 at 4:40pm
Sure we do: it’s targeted at the region affected by the disaster. We don’t give everyone in the country aid for an earthquake in California. Is it really that costly to target aid to people who are unemployed? This seems doubtful, since so many entitlements already are at least partly dependent on employment status.
I don’t think the equitability argument is very convincing either. Does anyone who is greatly concerned about equality really think giving Bill Gates the same $1000 check that you’re giving a homeless unemployed person contributes to equity? Usually, equity concerns cut in the opposite direction, toward making distribution more progressive.
Gordon
Mar 21 2020 at 5:46pm
Scott, I know in the past you’ve argued that interest rates are not the monetary transmission mechanism and that it’s asset prices and the hot potato effect. But given the high level of fear and panic, maybe those are not going to be as effective in the near term as the fiscal channel would be if the fiscal efforts have the full support of the Federal Reserve and won’t be undermined by the Fed.
Scott Sumner
Mar 21 2020 at 6:01pm
People made that argument in 2008, 2009 and 2012; it didn’t pan out. Fiscal stimulus simply won’t have much effect. Better to focus on humanitarian assistance.
Gordon
Mar 21 2020 at 7:03pm
I would agree that fiscal policy is a very poor replacement for monetary policy in helping to boost or stabilize NGDP and monetary policy trumps fiscal policy. I’m just wondering though if the Fed boosts the monetary base by X dollars by buying securities and the recipients of those dollars in turn purchase new securities issued by the government to fund direct payments to households and boosting unemployment benefits, do you consider that fiscal stimulus that has little effect on NGDP?
Scott Sumner
Mar 21 2020 at 11:15pm
Yes, the fiscal part of that policy has little impact.
Gordon
Mar 22 2020 at 5:12pm
That may need to be a little clearer in your writing. As Lars Christensen has written in the past, a large part of the effectiveness in monetary policy is expectations. Some of your writing could be interpreted to mean that the “hot potato” created by an increase of the monetary base by the Federal Reserve is “cooled” if the potato is in short order loaned to the federal government. It can also be seen as an implicit rejection of David Beckworth’s proposal to have the Federal Reserve partner with the US Treasury if conditions arise that will cause a severe collapse in aggregate demand.
Sven 2020
Mar 21 2020 at 6:13pm
Isn’t the real confusion precisely distinction between fiscal and monetary operations?
What is the distinction anyway? It cannot be distincted by the question “who is doing it”. It cannot be the distinction between “does it have distributional effects or not”, since monetary operations obviously have some distributional effect (it moves asset prices).
It can be distincted by the question “will the spending be paid for by (immediate or future) explicit taxes or will it be paid for by inflation tax”.
If that later distinction is used, then giving everyone 1000€ and paying for it by not paying for it (i.e. monetarize it and paying by inflation tax) is pure monetary policy.
Why should
Don Geddis
Mar 22 2020 at 4:54pm
The distinction is very easy to make. Monetary policy changes the quantity of money (monetary base). Fiscal policy does not.
“Pure” monetary policy (in the abstract ideal) has no direct economic effects except money supply changes. You are wrong to claim that monetary policy necessarily has “distribution” effects. Changes to the money supply change velocity, and aggregate demand. All the other changes depend on the current structure of the economy, and are never required effects of monetary policy. Not asset price rises, and not inflation. Those may happen, depending on what state the economy is in. But monetary policy does not necessarily cause those things to happen.
Fiscal policy causes changes in the economy only by redistribution. It cannot cause changes in aggregate demand (because of monetary offset) or the money supply. And fiscal stimulus must be paid back by future taxes.
Thomas Sewell
Mar 21 2020 at 8:48pm
The goal of sending everyone $1000, etc… is to encourage people to vote for the politicians who are seen to be responsible for sending the money.
The economic justifications are similar to those for spending for green energy or defense procurement. Enough plausibility to provide cover, but don’t actually make sense if you look too closely. Sure, there are true believers on both sides, but the pandemic is just enough cover for enough people who don’t look beneath the covers to closely.
Sadly, if one group of politicians is seen to do it, but the other side isn’t, then guess which side loses votes in the election and thus power?
Matthias Görgens
Mar 22 2020 at 5:31am
Scott, the Fed moved ever so slightly in the direction of free banking when they abolished reserve requirements. That should allow private banks to create more money. How much of an influence do you think that has?
Ideally, most of monetary policy should be left to the private sector like that. George Selgin showed how a fixed amount of based money in a free banking system leads to a variable amount of private money creation, varied just right to stabilize total nominal spending.
Abolishing reserve requirements was a good first step. But I guess they still have deposit insurance, capital requirements, implicit bailout guarantees etc on the books, so that first step might not have had much of an impact.
Scott Sumner
Mar 22 2020 at 12:47pm
Abolishing RRs won’t have much effect, although I support the move.
TMC
Mar 23 2020 at 12:16pm
A little OT, but what are the chances they put back the RR when this is all over?
Thaomas
Mar 22 2020 at 6:01pm
Although I’ve wondered for years why the Fed failed to maintain enough growth in aggregate demand (a very reasonable interpretation of its Congressional mandate) or even their own supposed target of 2% in CPE inflation, I have never hear the argument that they were even in part offsetting the “stimulus.”
Mike
Mar 23 2020 at 5:24pm
Scott – you’ve had me convinced were we talking about bank balance sheets, but my concern is around business transactions and the uncertainty faced right now in the face of indefinite constrained operations and/or shutdown. The vast majority of business transactions are on some variant of terms with most being 15 or 30 day. I don’t see how level targeting and adjusting the monetary base hits the place where we are likely to see the biggest drop in monetary velocity.
What we really need is to get everyone to agree that this is a temporary blip so that we don’t lose businesses that are otherwise viable due to a cash flow crunch caused by the temporary shutdowns. The loss of output during this time will not be good, but it pales in comparison to what would happen if the cascade effect could amount to.
Comments are closed.