Why employees got the payroll tax cut

By Felix Salmon
December 7, 2010
Greg Mankiw discusses the economics of the payroll tax cut, and raises the question of whether it might have been better used to cut employers' share of payroll taxes, rather than employees' share:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Greg Mankiw discusses the economics of the payroll tax cut, and raises the question of whether it might have been better used to cut employers’ share of payroll taxes, rather than employees’ share:

This payroll tax cut goes entirely to the worker. This increases work incentives, but the main motivation is probably to increase take-home pay, consumer spending, and aggregate demand…

An alternative would have been to reduce the employer’s share of the payroll tax, at least to some degree. Given a sticky wage, this policy would have reduced the cost of hiring and, to the extent labor demand curves slope downward, increased employment. It would also have increased business cash-flow and, to the extent that firms are cash-constrained, increased business investment.

After raising the question, Mankiw stops short of answering it, leaving the possibility open that maybe Congress and the White House came to the wrong decision for political reasons. But of course they didn’t: this is a no-brainer.

Michael Cooper had a great article in the NYT in October about the last payroll tax cut, a/k/a “the Tax Cut Nobody Heard Of.” The idea is that the less noticeable a tax cut is, the more effective it is:

The tax cut was, by design, hard to notice. Faced with evidence that people were more likely to save than spend the tax rebate checks they received during the Bush administration, the Obama administration decided to take a different tack: it arranged for less tax money to be withheld from people’s paychecks.

They reasoned that people would be more likely to spend a small, recurring extra bit of money that they might not even notice, and that the quicker the money was spent, the faster it would cycle through the economy.

Economists are still measuring how stimulative the tax cut was. But the hard-to-notice part has succeeded wildly. In a recent interview, President Obama said that structuring the tax cuts so that a little more money showed up regularly in people’s paychecks “was the right thing to do economically, but politically it meant that nobody knew that they were getting a tax cut.”

So we know that people spend payroll tax cuts rather than saving them, which is exactly what we want any fiscal stimulus to do. Businesses, by contrast, are another thing entirely: they’re sitting on large and growing piles of cash, and showing little if any propensity to spend any excess cash which falls into their laps.

I certainly can’t imagine that a 1-year cut in payroll taxes would incentivize any employer to hire more people. A 2% cut in payroll taxes on a worker earning $30,000 a year amounts to a one-off payment of $600: very nice for the worker, but not remotely enough to get the employer to hire someone else. After all, payroll taxes will go back up to their normal levels in 2012.

And while business investment might increase “to the extent that firms are cash-constrained,” as we’ve already seen, they’re not really cash-constrained at all. Certainly large companies aren’t.

So really there’s no conceivable reason to give this tax cut to employers rather than employees. And if Mankiw weren’t being disingenuous*, he might make that a bit more obvious.

*Update: Apologies to Mankiw, I’ve now changed my mind, and no longer think he was being disingenuous. (But I’m still a fan of employee-side tax cuts.)

Comments
12 comments so far

Sahm, Shapiro, and Slemrod’s paper “Check in the Mail or More in the Paycheck:Does the Effectiveness of Fiscal Stimulus Depend on How It Is Delivered?” which is mentioned (but not by name) in the NY Times article finds just the opposite, evidence that that the propensity to consume and therefore the multiplier is much lower than lump sum tax cuts. This is not an effective method of stimulus if what you want to do is get consumers to spend more.

If consumers are likely to simply save money delivered in this way and leave it in low interest banking accounts then this is highly unlikely to have a large stimulating impact.

On the other hand, by lowering the marginal cost of labor it may well stimulate employment, which will be politically desirable regardless of the stimulating effect. The question is what is likely to have a larger impact on employment. As Mankiw points, in the frictionless case it won’t matter. In the presence of frictions (like minimum wages) we have reason to suspect the form this takes is not as likely to lead to higher employment.

Posted by OneEyedMan | Report as abusive

If an employer with 50 employees saves $600 per employee, they can hire another employee with those savings, increasing productive capability, correct? And by naive extrapolation, this would lead to a 2% reduction in the unemployment rate.

Posted by caveatBettor | Report as abusive

@OneEyedMan, I saw that paper, was not impressed: it wasn’t based on anything empirical, but rather just on surveys.

Posted by FelixSalmon | Report as abusive

@caveatBettor, the point is that those savings disappear in year 2, so unless you’re intending to fire that employee in year 2, you shouldn’t hire them this year.

Posted by FelixSalmon | Report as abusive

Is the temporary nature of the cut going to lead to a short term response regardless?

If there was a material increase in ag demand due to these cuts:

a)Wouldn’t the employees hired to produce for the demand just be fired when the demand receded in a year?

b)Would businesses simply force current workers to account for current demand if they know its temporary?

I suppose there are extra costs associated with firing an employee (morale, severance) which need to be taken into account when making a hiring decision.

It seems odd that the employment break gets the temporary treatment, but the employee break gets interpreted as a sustained increase in ag demand.

Perhaps its psychological – you can’t readily determine if an ag demand increase is due to the economy improving or temp tax cuts, so you make the hire regardless and live with it. You’re less likely to hire someone if you’re 100% sure the incentive will expire in a year.

Posted by djiddish98 | Report as abusive

@Felix Do you know that essentially all economic data is gathered by survey? I agree that paper has problems, but the fact that it is a survey isn’t one of them.

On the other hand, why do you believe that the delivery method should matter? To my knowledge this is the first test of that claim with any data. Since there isn’t any actual evidence that a hidden delivery method drives higher spending, I take the SSS paper as weak evidence that the “stealth money drop” method works no better and probably worse than lump sum transfers that are heavily advertised.

Posted by OneEyedMan | Report as abusive

I know how you think being sometimes wrong is necessary to being sometimes interesting, but I’m willing to believe you actually believe that this was the right way to do this cut. While it might seem hypocritical, I do think it’s likely you’re being disingenuous in suggesting that Mankiw is being disingenuous; it’s hard for me to imagine that you’re so ensconced in some sort of closed-minded echo chamber that you really don’t even believe that there are people who aren’t wrong with you (whether you would put it that way or not). The FoxNews/CNBC world notwithstanding, accusing people who (pretty clearly genuinely) disagree with you of being disingenuous is absolutely not necessary to be interesting.

Posted by dWj | Report as abusive

But this is the ideal thin end of the wedge. The payroll tax is a highly regressive tax. Considering that wages are moving to less than half of the national income it makes more sense to give a tax cut to those who are propping up aggregate demand. Hopefully it will be renewed for the next year and expanded. The ideal target is to get a total payroll tax holiday (for employees and employers) for a couple of years until the economy rights itself. The government can credit the Social Security accounts until then. It is worth a hit to the deficit. It is a better way to run a stimulus plan by giving money to the people most likely to spend it rather than giving the banks money which they will refuse to lend.

Posted by rahumkc | Report as abusive

I notice that another economist is being disingenuous. Bryan Caplan tries to explain why cutting on the employer side is better. He rests his whole argument on lowering the cost of labor.

http://econlog.econlib.org/archives/2010  /12/right_cut_wrong.html

Posted by libfree | Report as abusive

Mankiw is the very picture of disingenuousness, to the point where when he opens his mouth, you can pretty much assume it’s a lie that he knows is a lie. Indeed, it’s hard to find instances where this -isn’t- true.

Posted by EconomistDuNord | Report as abusive

@Felix: As long as the firm’s employee turnover rate is greater than 1%, this will not be a problem. Let’s assume the firm’s employees are born on the same date, but one each year, from 1931 through 1981. As long as some of those sixty somethings retire in two years, this still looks like a free call to me.

Posted by caveatBettor | Report as abusive

The employees got the Payroll Tax cut – because the “Payroll Tax” is actually what you pay to your Social Security account; so – if you pay in less – you will eventually receive less. Everybody is on to this except the workers. It used to be called FICA, remember?
Eventually you will get so many “Payroll Tax” cuts that you won’t have a Social Security Account at all and that is the object of the game.
It is the same with those Health Insurance and Communting expenses you can use to reduce your “Payroll Tax”.
All those fine economists speculating on the result – are either uninformed or, much worse, in favor of the game.

Posted by 2ndLook | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/