Bryan Caplan  

Right Cut, Wrong Side: Obama Botches the Payroll Tax Holiday

Behaviorism Works Because It's... 54 Libertarians Tell Their Sto...
With perfectly flexible wages, it doesn't matter whether tax law says "employees pay" or "employers pay."  Tax incidence depends on supply and demand elasticity, not legislative intent. If wages are nominally rigid, however, the law matters.  If you cut a tax on employers, this reduces labor costs, increases the quantity of labor demanded, and reduces surplus labor.  If you cut a tax on employees, in contrast, this increases worker compensation, increases the quantity of labor supplied, and increases surplus labor.

In both cases, admittedly, a tax cut might directly increase demand and, with nominal wage rigidity, increase employment.  But when you cut taxes on employers, the incentive effect and the fiscal effect work in the same direction.  When you cut taxes on employees, the incentive effect and the fiscal effect work in opposite directions.

That's why Obama's proposed payroll tax holiday botches an idea of truly Singaporean cleverness.  Instead of giving the tax cut to employers, where it would do the maximum good, or splitting it evenly, where it would do intermediate good, he's giving all of it to employees, where it does the minimum good:
The payroll-tax reduction under discussion now would cut the 6.2% Social Security tax levied on a worker's wages to 4.2%. A worker making $40,000 a year would save $800, and some economists say that could help stimulate demand at a time when the economy remains relatively weak.

The employer's half of the tax--also 6.2%--wouldn't be affected under the White House proposal, and thus the cost of hiring new workers wouldn't be directly affected.
Of course, if Obama's goal is simply to salvage his fading popularity, he's right on target.

P.S. If you object that you want to put the money into the hands of people who will actually spend it, remember that cash flow is a good predictor of business spending.

HT: Alex Tabarrok

Comments and Sharing

COMMENTS (17 to date)
John Thacker writes:

Greg Mankiw makes this same point.

Eli writes:

It may salvage Obama's popularity, or it may harm it. Since with nominal rigidity, real wages go up, this tax cut will draw more people into the labor market, raising the measured unemployment rate. I'm not saying it will cause less employment, only that the rate reported in newspapers will be higher than it otherwise would be.

Dave writes:

How does your analysis change if your economic model is that profits are bad and employers are evil? Does the policy make more sense then?

Joey Donuts writes:

What does nominally rigid mean?

Does it mean wages workers put in their pocket are nominally rigid or tne nominally stated wage for the job is nominally rigid?

If its the first, why can't wages the employer pays (nominally stated wage for the job) go down with the tax reduction on workers salaries?

If its the second, why not tax the hell out of workers wages, it shouldn't make any difference to them.

I think you are assuming the second which seems to me silly.

Ben Hughes writes:

@Joey Where exactly did Bryan imply "it shouldn't make any difference to them [the workers]"?

First of all, Bryan is referring to *relative* taxation - given a fixed percentage of taxation - how should it be "allocated" between employee and employer. The phrase "tax the hell out of" implies you're changing the terms of the argument to debate the total taxation. That's a completely separate debate.

If you really do mean mean why not shift the entire tax burden on the employee, then: if your only stated goal is to decrease unemployment, this would probably be a good thing. But it also - precisely because of sticky wages - would *reduce* take-home wages across the board, by a fairly large amount. This might be seen as a way to break - by force - wage rigidity, but a large wage shock of this kind would presumably harm AD, which is exactly what Keynesians don't want, right?

I think the question comes down to whether you think - speaking only of the unemployment question - whether a very indirect "aggregate demand" increase will be more effective than a very direct "cost of labor" reduction, and I think that answer is most emphatically no.

Joey Donuts writes:

@ Ben Hughes

Did I read this incorrectly?

If you cut a tax on employers, this reduces labor costs, increases the quantity of labor demanded, and reduces surplus labor. If you cut a tax on employees, in contrast, this increases worker compensation, increases the quantity of labor supplied, and increases surplus labor.

It only increases labor compensation if the employer keeps the stated wage of the job constant. If the employer lowers the nominal stated wage to offset the tax decrease, The employee loses nothing. The employer lowers his labor costs. This is what I mean by the assumption that the stated wage is nominally rigid. Why wouldn't out of work employees offer up their services for lower nominal stated wages if take home nominal wages don't change. Thus providing incentives for employers to hire them.

It seems to me this possibility is ignored, which is what I find silly.

Dale Moses writes:


That is what we expect to happen in the long run; employers lower compensation until the burden is again distributed by the relative elasticities of supply and demand.

But in the short run these prices may be nominally rigid in the manner described. We can think of this as negotiation costs. If an employer has a contract with their employees for a monthly/weekly wage, that wage is likely to be stated in nominal dollars.

If the employer is going to lower it, he has to notify his employees at the very least (depends of course on the employment deal). If his employees are not happy about it then you have costs. If these costs are higher than the pay cut the employer will keep their nominal wage.

Rather than expecting a simple net present value computation and then a wage shock, the employer likely maximizes their profits by smoothing out the nominal wage decrease.

Anyhoo: whether or not that matters really depends on whether or not people are actually willing to work less if they get paid less. Since people tend to want to work 40 hours per week every week and tend to do so regardless of the wage(for various reasons, and especially for salaried employees) it may also be the case that there is no or almost no supply effect as discussed above.

If there is no(or little) supply effect for labor then all we see is aggregate demand stimulus and the value of the cut is based on the marginal propensity to consume of those people who receive it (assumed to be higher for lower income people, which the payroll tax hits more than other taxes).

On the other side, giving the money to employers assuming that nominal wage rigidity(I.E. employees are not able to demand higher wages due to the payroll tax cut) means that they will produce more. Normally we would say they are likely to use this windfall for both labor and capital. But for the sake of argument lets say that businesses have plenty of excess capital (yes, i know this is not the story that the tax cut folks are yelling from the heavens) and so hire mainly labor or spend it themselves.

Both of these are likely a wash, picking whether you want your chicken first or your egg first. If we give the money to workers they buy more and this drives demand for goods and services and causes industry to produce more in response to the increased demand. If we give the money to employers this lowers input costs and they produce more in response to the lower costs.

Hyena writes:

I really like Mankiw's explanation, which is very clear on why it would matter which side you do the cut on.

You're not going to get a Social Security cut on the employer side, however. It's not feasible because most people don't realize that there are employer shares to taxes.

MikeDC writes:

1. Almost all Americans are economically illiterate, or worse, irrational.
2. President Obama would rather make a faulty move that appears helpful than make a truly helpful move because
3. The particular economic illiteracy of his political base is that any "tax cut for business" must be bad.
4. This particular form of economic illiteracy has been fostered by President Obama and the Democrats for years, right up to, and nearly simultaneous with his angry bluster about the need to improve the economy in spite of the sanctimonious purity of those on the left.

In short, he's reaping exactly the seeds he's sewn. A truly beneficial move would be more unpopular because of his party's repeated promotion and appeal to unsound economic policy.

The less beneficial half-move will be more popular with the base, but it's still not going to generate much of value because the base is mad about caving in on other poorly thought out, populist economic promises broken. Any benefit will be lost in the wash.

The lesson here would be that a political party that actually attempted to understand economics and educate their particular mob accordingly, rather than further inflaming them, would be in a better position to help.

Brad Warbiany writes:

If this were a permanent change, or a 5, or even a 2 year change, I might agree. Over a longer term, an employer-side payroll tax cut will improve employment.

But at a single year, I'm not sure I buy it. You might have a very marginal improvement in hiring, but I think it's more likely to go to the bottom line in an environment where many businesses are thinking about survival rather than expansion.

The Democrats, IMHO, view this as a stimulus program, not an jobs program. They know wages are sticky, and that means that most workers will see a 2% boost to take-home pay for 2011. Even for high-income earners, you're talking about

I completely agree that putting the tax cut to the employer side of the ledger is more likely to improve employment, but I don't think that was the intent (regardless of what any politician might publicly say).

Ano writes:

There are three big problems with your arguments here.

Problem 1:
Your analysis assumes that the goal is to minimize "surplus labor" rather than something more directly appropriate as a measure of economic health, such as employment-to-population ratio or capacity utilization. If all we cared about was the amount of "surplus labor" we could just say "if you say you want a job but you don't have one, you go to jail." Boom! No "surplus labor!"

So when you say "If you cut a tax on employees, in contrast, this increases worker compensation, increases the quantity of labor supplied, and increases surplus labor," you are saying that discouraging workers from looking for work is good economic policy. I don't think you believe that.

Problem 2:
Your discussion of incentive effects on demand for labor doesn't seem very realistic as a large-scale job-creator. With 15% unemployment among workers with no HS diploma, do you really think that employers don't have the power to lower their pay offer by 2% and get workers? Really? Millions of unemployed people are sitting home because they're holding out for $10.20 per hour when they've been offered $10 per hour? The payroll tax reduction is a very expensive way to affect these incentives at the margin for 1 year. Your analysis works for improving the long-run employment-to-population ratio but seems really far-fetched as a way to reduce unemployment right now. The labor market is not exhibiting 9.6% unemployment and 17% underemployment because it is having a labor supply problem.

Problem 3:
Your point that "cash flow is a good predictor of business spending" does not jibe with reality right now. Businesses are profitable and are piling up cash rather than investing because they aren't seeing the demand to justify capacity expansion. Pointing out that during normal times cash flow leads to investment is a non sequitur in a time when businesses are putting cash in 0.5% return bank accounts rather than investing. How could adding more cash to that pile lead to a similar effect on aggregate demand than a payroll tax cut for workers?

Brian writes:

Brad has a good point that a one year cut will not likely have much effect on the stated wages offered for most mid to high level jobs, though I suspect stated wages can adjust downward much more quickly for lower wage jobs - to the point that fair living wage laws protect them.

Ano's point is also valid that businesses show no interest (or need) in investing or growing because consumer demand remains depressed. Until that changes, no amount of tax incentives to businesses will improve employment, regardless of how you package it.

In that light, adding liquidity on the consumer side in the form of a 2% increase in take-home pay could be the best way to stimulate consumer demand and thereby increase employment by providing business with a reason to grow. Earlier attempts to do this were not successful because consumers used them to pay down debt and increase savings. These actions are beneficial and in fact unavoidable during and after a recession, so the earlier stimuli were likely helpful from the standpoint of speeding up the process, however many signs now indicate that this personal economic re-trenching has progressed sufficiently to expect this new stimulus may well have the intended direct effect of increasing consumer spending. If that is the case, we should see great benefits.. but of course that still remains to be seen.

dlc writes:

The problem with the analysis is that it ignores the temporary nature of the proposed cut. If payroll taxes were reduced permanently the effects on labor demand would be obvious. But tell most business owners that they get a payroll tax "holiday" of one year, and they're far less likely to hire additional workers since they expect the long-term labor cost structure to be unchanged. More likely, the cash from temporarily cheaper labor would be stashed (status quo) or spent on capital, where another part of the proposal allows 100% expensing next year, further reducing income tax liability. An employment booster, it would not be.

Oh yeah. And the Republicans would never let a temporary tax cut on business owners expire, regardless of the deficit situation.

Lord writes:

I agree that in normal times your argument would be correct, but these are far from normal times. The Fed has still been running too tight monetary policy and as a result we are still in a conserve cash mode. Cutting the employer side will further increase profits but will not stimulate demand because the demand for their goods will not rise. This is what happens during depression, lower costs lead to less, not more investment as the demand for money exceeds the demand for goods. On the other hand, giving it to employees, particularly given its regressive nature, means it will be quite efficiently turned into demand that employers will respond to. The reason is it is a demand, not supply problem and demand for money most of all.

Larry writes:

In all my years of hiring, never once did a mandate come down to hire more people because there was a tax cut available. Hiring is a function of work, which generally has a direct correlation to sales. We would pocket the tax credit on any replacement employees we had to hire, but it never sparked hiring for its own sake.

Voodoo economics, GHW Bush was right. It's time to return to the 70% marginal tax for the wealthy then use that money to make sure everybody had a job that pays a living wage. It's when this country has worked best for the majority.

Steve Roth writes:

>P.S. If you object that you want to put the money into the hands of people who will actually spend it, remember that cash flow is a good predictor of business spending.

I've been waiting for you to haul out this canard again. You're essentially saying that if we give businesses enough money, they'll spend it all.

And when you said in a previous post that I'm too lazy to look up that savings get invested and those investments get spent, you were essentially saying that savings translates directly and immediately in business spending, so contributes immediately to AD. Cmon.

We know that money given to the poor impacts velocity much more than money given to the rich.

Bryan Caplan knows better than to (repeatedly) cite a single study demonstrating one (medium weak) effect as if it were definitive proof proof of a large theoretical macro construct of which that effect is one part.

Also on the validity of that one study: Propter hoc?

Banker Zero writes:

So your argument is that an employee who gets $800/yr somehow ends up working harder? They all of a sudden become more productive?

And apparently you think a company that is taxed less will somehow all of a sudden need more labor? Why? is there all of a sudden more work to be done?

Corporations are hoarding cash. Why wouldn't they simply hoard this extra cash? or why wouldn't they use it to pay more for "talent" (ie someone making way more than 40k/yr).

The purpose of the cut is that it will have a stimulative effect, and spur consumption. You seem to assume that the purpose was to create jobs. In a sense, that is the purpose, but on a temporary basis - added consumption will do more towards that end.

Businesses need to look farther ahead than 2 years to figure out their respective workforces. A temporary cut on their end of the payroll tax would do very little for the economy as a whole. For some reason you think that money, and not need will create demand for a workforce.

If more people are shopping, a store needs more cashiers, if more people are buying widgets, a factory needs more people to man the machines. If the amount of shoppers stay consistent, but the store gets a tax break... you say they'd need more labor. Your argument is flawed for obvious reasons. Reasons a middle school kid could point out.

Comments for this entry have been closed
Return to top