The best kind of fiscal stimulus
If we aren’t going to have monetary stimulus, what would be the best (or least bad) type of fiscal stimulus? In my view the best would be cutting the employer share of the payroll tax, as that seems to be the only way to reduce real wage costs. I view sticky nominal wages as one of the major causes of unemployment during recessions. The traditional Keynesian argument against wage cuts is that it merely leads to more deflation, without reducing real wages. I recall Gauti Eggertsson discussing this problem and Paul Krugman citing his argument. But I don’t buy the argument. The Fed controls the price level.
You might ask why if the Fed determines the price level, they don’t just go ahead and raise the price level, instead of letting the fiscal authorities run up trillions in budget deficits (stimulus plus cyclical deficits). I can’t answer that; you’ll have to ask the inflation hawks at the Fed. But for whatever reason, they’ve decided we have enough NGDP, and if Congress wants more they’ll have to do it the hard (and inefficient) way.
At least Bernanke made it pretty clear that he won’t allow further disinflation. He hinted that there is a sort of ‘Bernanke put’ around 1% core inflation, roughly where we are now. If this is the case, then any nominal wage cuts become real wage cuts. For very complicated reasons it is hard to directly cut the nominal wages of workers. But you can cut the nominal labor costs to companies very easily—just reduce payroll taxes.
Now that the Keynesian fiscal stimulus has not worked, the Obama team is looking at new ideas. They are rediscovering the merits of sticky-wage theories of the business cycle, which I am one of the few economists to still adhere to (most now assume sticky-price versions of Keynesianism.)
The Washington Post reports that Obama is considering a $300,000,000,000 payroll tax cut. No details are provided, but I’d guess that would finance a 2% cut for a couple years. That would mean a 2% cut in wage costs, which should cause the recovery to speed up. Of course one always must think about the monetary policy counterfactual. The hawks may become horrified by the increased debt and tighten further. But I think Bernanke can hold the line at 1% core inflation. If so, it may be our best shot, even if it is far less efficient than monetary stimulus.
Here’s the WaPo:
With less than two months until the November elections, the White House is seriously weighing a package of business tax breaks – potentially worth hundreds of billions of dollars – to spur hiring and combat Republican charges that Democratic tax policies hurt small businesses, according to people with knowledge of the deliberations.
. . .
Permanently extending the research credit would cost roughly $100 billion over the next decade, tax analysts said. And depending on its form and duration, a payroll tax holiday could cost more than $300 billion. While significantly less than last year’s stimulus package, both ideas would be far more dramatic than anything the White House has so far acknowledged considering.
HT: Tyler Cowen
Tags: Payroll tax, Taxation
2. September 2010 at 18:13
I agree that this is probably the *second* best kind of fiscal stimulus, but it doesn’t come close to the first best: preventing the anti-stimulus that arises from state budget cuts. Even if you don’t buy full Ricardian equivalence (I don’t), there is surely some offsetting effect from savings when stimulus is in the form of tax relief, and given rigidities in the labor market and elsewhere, it’s much easier to prevent perfectly good jobs from disappearing than to create them from scratch.
Granted, it sounds as if you think that the main effect of a payroll tax cut will be the change in incentives, which state aid cannot provide. But this change in incentives seems very small relative to the massive losses from lower payroll tax rates that it will entail, and it’s extremely difficult for me to imagine that the marginal dollar will have more effect by slightly decreasing the costs of labor (and mostly serving as a massive transfer to all the inframarginal cases) than by directly saving jobs from unnecessary elimination.
2. September 2010 at 19:06
Yah, this is great – suddenly the unions who own GM deserve a 2% pay raise, and some dying companies survive a bit longer. Boss, can I get that raise now????
Cut corporate taxes by 30% across the board – and make them permanent ($100B per year)… that way only the profitable companies get the juice.
Now you have incentivized productivity gains… and given us a strong competitive leg up against other countries. We are the United States of America, if we don’t have the smallest corporate tax rates in the world – we have gone horribly astray.
2. September 2010 at 19:11
Repeating the comment I placed in A moderate…post:
Would this be helpful in implementing the program?
http://macromarketmusings.blogspot.com/2010/09/right-kind-of-helicopter-drop.html
2. September 2010 at 19:21
Morgan,
My own research on the relationship on tax structure and growth argues strongly in favor of low to no corporate taxes (over the long run). Should we follow the path of Estonia in the short run? (Down 30% in GDP.) Is this problem not simply stated enough? In my opinion you DON’T GET IT!
2. September 2010 at 22:02
Mark,
Estonia is a small young country off to a good start – some top notch programmers there… but still a GDP of $19B, like a million people, and a negative population growth… cannot occur in any graspable language in a a sentence like…
“should we follow the path of Estonia?”
wtf are you talking about?
We are the world history’s leading capitalist country. Period. Our country? the one apparently you “live” in – chooses capitalism over democracy… every god damn time.
Our deepest cultural instincts, proven by the longest standing Constitution, and defined almost entirely by negative rights – is a living breathing systemic boot on the throat of the majority of greedy citizens protected by the majority of likely voters.
This country should have nothing but a progressive consumption tax – also note, after Scott mistakenly called payroll taxes a consumption tax, he dropped it.
Also note, Scott is unable to point to virtually ANY “sticky wages” not associated with public employees – I assume this is because he works in education. Meanwhile something like 55% of private sector workers have sen wage cuts, been told to work longer hours for the same pay, etc.
The the UAW membership votes to massively cut new employee wages the WHOLE concept of sticky is a giant lie…. a lie told by statists and neanderthals.
2. September 2010 at 22:51
This is going to be a great test case for my political theory that Republicans at this point will oppose anything Obama does. At the beginning of the stimulus debate, when pressed for what they, the Republicans would do, they proposed having a payroll tax holiday or at least cutting the payroll tax. So, Obama will meet them halfway and I am going to guess he is going to be waiting at the halfway point alone as usual.
2. September 2010 at 22:57
By the way, Morgan’s response to the proposal is a great microcosm of the usual Republican response.
I often hear Morgan attack people like me for suggesting that unfunded tax cuts are just as irresponsible as unfunded spending. Morgan rips me because (and I am paraphrasing) “tax cuts is simply returning to people what is theirs.”
So, now Obama proposes a tax cut and what does Morgan have to say?
“Yah, this is great – suddenly the unions who own GM deserve a 2% pay raise, and some dying companies survive a bit longer.”
You realize that using your logic a 30% corporate tax cut can also be construed as a give away to the unions who own GM.
The mental gymnastics that conservatives will undertake never ceases to amaze me.
3. September 2010 at 01:22
Why should the cut be only in the ’employer share’? If I remember my Price Theory correctly the effect would be the same either way, and cutting the ’employee share’ would presumably be more popular.
3. September 2010 at 01:43
@Roman,
All tax cuts are not created equal, for the exact good reason I explained. We do not want tax cuts that aid the unproductive.
GM: you fool, GM is not profitable, not like countless green and growing companies. Why do you not give some thought to my real point? Your policy actually gives more money to the companies with inflated wage balance sheets (the losers), I want to help the ones who have pared wages to the bone.
Tell you what, you guarantee that if I support this tax cut, you’ll support the good kind the new Republican congress passes and hands to Obama to rubber stamp – and I’m in.
This is the same thing Scott and you refuse to do on targeting NGDP – if you think that it is such a good policy, then make sure the new printed money gets handed to the current savers, the companies and individuals with the best balances sheets… use the new money to encourage Darwinism, to encourage creative destruction – reward the winners.
It already a politically complicated policy, if you want it to be successful, do you best to make the Tea Party happy.
I look forward to your new proposals.
3. September 2010 at 04:31
Matt, You said;
“Even if you don’t buy full Ricardian equivalence (I don’t), there is surely some offsetting effect from savings when stimulus is in the form of tax relief, and given rigidities in the labor market and elsewhere, it’s much easier to prevent perfectly good jobs from disappearing than to create them from scratch.”
I agree that it would boost AD, but that’s not the goal. The goal is to lower labor costs. And the goal is to prevent perfectly good jobs from disappearing. Each month millions of new workers become unemployed, and million more find jobs. If it can slow down the rate of workers losing jobs, that would help a lot. I agree that propping up S&L governments would have been better than what they actually did.
Morgan, You said;
“Yah, this is great – suddenly the unions who own GM deserve a 2% pay raise, and some dying companies survive a bit longer. Boss, can I get that raise now????”
The cut would apply to the employer share, and be short run only.
Marcus, I left a comment over at David’s blog. It would be great if any fiscal stimulus was accompanied by monetary stimulus, but of course I’d prefer monetary stimulus without the fiscal stimulus. I’m still worried about the national debt.
Morgan, I don’t work in the public sector. Our wages were cut 5.5% relative to the trend over two years, but even that wasn’t enough to prevent recession when NGDP fell 9% relative to trend.
Liberal Roman, Yes, it will be interesting to see how the Republican’s react.
David; You said;
“Why should the cut be only in the ’employer share’? If I remember my Price Theory correctly the effect would be the same either way, and cutting the ’employee share’ would presumably be more popular.”
This is true in the long run, when wages are flexible. But when wages are flexible, we should be out of the recession. In the short run nominal wages are sticky, so if you cut the employer share, the workers do not get any of the tax cut.
3. September 2010 at 05:02
Hey Scott,
You write: “For very complicated reasons it is hard to directly cut the nominal wages of workers.” Could you point me to things to read that discuss the “complicated reasons”?
3. September 2010 at 05:51
Off-topic, but when I saw this:
http://www.marginalrevolution.com/marginalrevolution/2010/09/the-small-schools-myth.html
I thought: I wonder if the whole “smaller countries have better governments” idea has the same cause?
3. September 2010 at 06:37
Scott,
I agree that the primary purpose of this policy would be to lower labor costs rather than boost AD. It’s just hard for me to imagine a world where employment is so elastic with respect to labor costs that this is as effective as just directly spending the money to prevent layoffs at the state and local level. Maybe the combination of lower labor costs and AD justifies it (and certainly it’s better than most of the alternatives), but it doesn’t seem like a first-best way to improve the employment picture.
3. September 2010 at 06:49
It makes Economics 101 sense, that when unemployment is high, reducing the cost of employment will increase hiring. There are many ways to achieve this — cutting payroll taxes is one way. But do you get the same “bang for the buck” cutting payroll taxes as you would hiring people directly via infrastructure spending? Also, if the payroll tax cut is only temporary, then forward-looking businesses will not increase permanent employment, they will pocket the windfall. This is a problem in general with temporary tax cuts, no matter who the recipient, that temporary tax cuts will often be devoted to savings or paying down debt. But temporary government spending increases aggregate demand directly.
And I agree with the first comment above that the most important stimulus right now would be for the federal government to help state & local governments avoid job cuts — avoid the “anti-stimulus” of the 50 states cutting back. A small portion of a payroll tax cut would go to state and local governments, but, again, you won’t get the same “bang for the buck” as if the entire sum went directly to state and local governments.
Targeted stimulus is probably a better solution — target the money toward recipients who will definitely spend it, thereby increasing the velocity of money, which is just as good as (if not better than) increasing the supply of money.
3. September 2010 at 07:18
Scott,
What is your view on a Kurzarbeit-style program? Wouldn’t that be more efficient on a per-dollar basis than cutting the payroll tax?
Hiring and training is expensive. If a decline in demand is cyclical, I’d imagine a firm would be open to cutting hours and having the government pick up, say, half of the difference in the wage bill.
If all firms who participated cut hours enough so their wages paid per ‘short work’ employee fell by $15,000, then it would only cost the government $7,500 to maintain that worker. Putting 5 million people on this program would only cost about $40 billion/yr assuming some administration costs. You’d also have a lower unemployment insurance bill which would offset much of the cost of this program.
3. September 2010 at 07:34
Scott, you are facing the wrong way.
There are 4.5M+ non-farm new hires every month. These are by and large made by GREEN AND GROWING companies.
Why in the world are you concerning yourself with stemming job losses from the RIPE AND ROTTING companies?
With any kind of tax policy you want ALL the juice to go to the green and growing guys, the profitable ones. Improve their balance sheet, so they can get more aggressive killing off the weak.
Think of your Sticky Wage hang up as a problem with the turn-over rate not being high enough.
Even if total lost jobs goes up by 1M, if we get 1.2M new hires in the same month – that’s a HUGE WIN.
Good tax policy will increase job turn-over. More new hires, even with more new losses.
—-
Let me make sure you get this – the purpose of tax policy is to encourage Creative Destruction. Long term progressive consumption taxes serve this purpose. Long term cuts in corporate taxes serve this purpose.
Short term “give the biggest benefits to the companies paying their employees the most” tax boosts – not so much.
3. September 2010 at 08:41
Nonfarm business unit labor cost has been falling at about a -3% rate over the past year-and-a-half, effectively cutting labor cost sharply. Moreover,labor share of business cost
has fallen to 57.6% of costs from 64% in 2001.
So what would cutting payroll tax give firms that they do not already have?
3. September 2010 at 10:31
Scott Sumner:
This is true in the long run, when wages are flexible. But when wages are flexible, we should be out of the recession.
Thanks for the explanation. I thought I might be overlooking an assumption.
3. September 2010 at 16:23
The U.S. Chamber of Commerce is lukewarm on the idea of a payroll tax holiday. http://www.cspan.org/Watch/Media/2010/09/02/HP/A/37782/US+Chamber+of+Commerce+Annual+Labor+Day+Briefing.aspx
They used some of the arguments I did in an earlier comment, that it would pull SOME employment into the present that would have happened in the future, but when the incentive is gone so are the jobs if everything were to remain the same. They pointed out other concerns about how the credit would be given because not all employers would qualify. They also mentioned additional stresses recently applied to the supply side that would impact the effectiveness of it, and likely prospects for recovery as a whole.
We cannot rape the supply side of the economy while encouraging additional public consumption and expect a robust recovery, payroll tax hoilday or not. In itself, I agree that a payroll tax holiday is probably better than nothing, but taking other new regulatory action into account, we might be lucky to end up with a break even situation if it helps at all.
3. September 2010 at 17:41
Ken, Most of the discussion revolves around two issues: Long term labor contracts, and money illusion.
Workers often sign one two or even three year contracts. The term money illusion refers to the situation where workers feel happier if they get a 3% raise in a year with 5% inflation, as compared to a 1% pay cut in a year with stable prices. In the latter case they actually do better in real terms.
Another problem is the public good aspect of wage flexibility. Suppose auto workers accept a pay cut but other workers don’t. The auto workers are still screwed, as auto sales will fall during a recession, and if other workers don’t take pay cuts when inflation falls, there often will be a recession.
There is an article by Akerlof that discusses wage rigidity, it is mentioned somewhere in the comment section of the next post.
Blackadder, It could, but among developed countries in Europe and East Asia, I think it also holds on average. But it would be worth looking at.
Matt, Let me put it this way. I believe that all the unemployment recently caused by falling NGDP, was caused by the fact that nominal wages didn’t immediate fall 8% below trend when NGDP fell 8% below trend. Since the trend is about 4% growth in wages, I predict that if wages had fallen 4%, and if the Fed held the fall in NGDP to 3% (or 8% below trend) unemployment should not have risen. I may be wrong, but I can’t imagine any other explanation for the huge rise in unemployment.
Doug, You raise a good point, which is why the tax cut should last as long as the recession is expected to last. Once it is over, we won’t need tax cuts to insure hiring.
Justin, I’m not an expert, but it sounds good. I gather it worked in Germany.
Morgan, I answered that in the next post.
Spencer, I’m pretty sure that data is distorted by the business cycle. It’s the flip side of the data showing huge productivity growth. Almost no one believes that productivity growth is real. I presume companies laid off their least productive workers, or maybe they are working the remaining workers to the bone.
But I’ll keep an open mind, I am certainly not enthusiastic about payroll tax cuts–I prefer monetary stimulus.
Bonnie, I assumed all firms would qualify. If not, I am more likely to oppose it. I also assumed it was just the employer portion that was cut. That’s how you lower employment costs. Today I read somewhere they were thinking of also cutting the employee share, which makes absolutely no sense.
But you do want to move some work from the future to the present—that’s the whole idea.
3. September 2010 at 23:40
[…] some kind of payroll tax cut. The discussion so far has been mixed with some folks like Scott Sumner, Tyler Cowen, and Arnold Kling endorsing it while others like Megan Mcardle and Mark Thoma […]
4. September 2010 at 05:33
Scott,
Cutting the employee side makes perfect sense. You assume that savings are equally distributed across every income, when almost assuredly they are not.
See this post “Savings rate has recovered if you ignore the bottom 99%” at Yves’ place for more.
http://economistsview.typepad.com/economistsview/2009/08/the-savings-rate-has-recoveredif-you-ignore-the-bottom-99.html
I think most people have ignored how spending and saving are distributed in our country, much to the detriment of the economy.
Any extra pay for the lower 50% will almost assuredly get spent immediately, which is where the payroll tax holiday hits
If you’re trying to shift the AD curve, increased consumer spending is what you want. Businesses making more money is good, yes. But simply having every customer spend more is hugely important too – and I would say more important.
I’ll say that increased corporate profits for our largest companies will not cause them to hire for a short term, small tax cut. We are already looking at near record corporate profits – why haven’t they hired already if this was the primary lever to pull?
4. September 2010 at 11:18
Mr. E, I agree that the employee side would have more effect on AD. But I just don’t see any fiscal stimulus having enough impact on AD to make much difference. Instead, I want to cut the employer side because it reduces the marginal cost of production. This should increase the demand for labor
4. September 2010 at 13:57
“Another problem is the public good aspect of wage flexibility. Suppose auto workers accept a pay cut but other workers don’t. The auto workers are still screwed, as auto sales will fall during a recession, and if other workers don’t take pay cuts when inflation falls, there often will be a recession.”
Scott, I thought you considered yourself a free-market guy?
Why would you care if auto-workers take pay cuts, but bike mechanics start making more (more high-end bikes being sold and ridden)?
You get my question?
5. September 2010 at 08:13
Morgan, You asked;
“You get my question?”
No, I don’t. I don’t care what bike mechanics earn, or anyone else.
8. April 2012 at 15:14
[…] bank fiscal levers it can use as part of its overall policy regime. Some post-Keynesians and market monetarists seem to like the idea of using payroll taxes as a fiscal lever (albeit with different rationales). […]