David R. Henderson  


UBI: Reply to Dolan... UBI: Ed Dolan Responds...


Pardon the grammar. The correct grammar is: TINSTAAFL: There is no such thing as a free lunch.

Don Boudreaux has an excellent post this morning in which he takes on a post at the American Enterprise Institute about mandated paid parental leave. Here's the relevant statement from the AEI piece, written by Christopher Ruhm of the University of Virginia:

There is widespread public support for paid parental leave, including majorities in both major political parties, and both major party Presidential candidates in 2016 advocated for some type of leave policy.

A check of the link to the claim about widespread public support leads to this AEI article. The article states:
Support for the concept is bipartisan, with 83% of Democrats and 71% of Republicans in favor of this policy.

I tried to find the original poll question, but this is the best I could do. It doesn't state the particular question asked.

Don points out the big problem with Ruhm's claim, writing:
If there truly were widespread support for paid parental leave given its costs, employers would have every incentive to supply such leave to more workers. The fact that such leave is not supplied to more workers is the best available evidence that support for such leave in fact is not as widespread as Mr. Ruhm assumes it to be. The poll of the market where each respondent acts with something material at stake - in contrast to opinion polls (upon which Mr. Ruhm relies) where each respondent talks with nothing material at stake - reveals that most of the workers who Mr. Ruhm wishes to force to take a larger share of their compensation in the form of paid leave in fact prefer compensation packages featuring no paid leave but with higher take-home pay to packages featuring paid leave but with lower take-home pay.

I agree with Don that the market is the most reliable pollster. Employers, because they try to maximize their profits, respond to the demands of employees. If paid parental leave costs the employer, say, $2,000, and the employee values it at, say, $2,100, they can make a mutually beneficial deal. But if the employee values it at, say, $1,900, the deal won't happen. And it's good that it won't happen because such a deal would make the employer and employee both worse off. I wrote about this here.

But there's more to say. There's strong evidence from polling data that when people are asked if they want something nice, they will say yes. But when they are asked if they want something nice but they will have to pay something for it, the numbers tank.

Here's what Professor Ruhm's colleague Steven Rhoads wrote in 1985 in The Economist's View of the World, which ranks as one of my 3 favorites economics books by a non-economist:

Some years ago a similar poll found overwhelming support for increased spending on social programs. Seventy percent of respondents wanted more spent on the elderly. Sixty percent favored increases both for the needy and for education. And 54 percent wanted more spent on hospitals and medical care. But when the same people were asked if more should be spent even if more taxes were required, those favorable disposed dropped to 34, 26, 41, and 25 percent, respectively. Making clear who will pay the taxes can also have a dramatic effect. One poll that found 50 percent support for the use of tax monies to supplement the cost of operating bus services found only 27 percent support a few months later when the words "personal income tax monies" were used instead of "tax monies." We, the public, seem quite willing, if given half a chance, to believe that there is such a thing as a free lunch.

All italics in both quotes are in the original.

Now you might argue that this is about taxes whereas the original post was about mandates. But that makes my point, and Don's, stronger, not weaker. The likelihood that a mandate on a worker's employer will be paid for by the worker is actually higher than the likelihood that an expenditure program will be paid for by the particular person polled.

If the people polled were employees, and if they really understood who pays for the mandated benefits, many of them would answer: "Butt out, buddy. Keep your life-arranging hands off my wages."

Comments and Sharing

COMMENTS (14 to date)
AlanG writes:

David - I think the debate is really about what is a social good and once that is defined, who should pay for it. You are quite correct TINSTAAFL (had to cut and past so I wouldn't get this wrong). Even if we (the societal 'we') decide something is a social good the full impact is often not observed for a considerable time afterwards. A good example is Medicare which everyone around in early 1960s thought was a good idea. However, what was a modest program has since ballooned and we all know what the trustee's report states. I happen to like Medicare, not because I'm taking advantage of it but that it does help a lot of people who could not otherwise afford their healthcare.

My former employer had (and still has) parental leave for six weeks following child birth. I happen to think that's a good thing even though our kids were born before I started working there so it was not a benefit I was going to be able to take advantage of. Whether my salary was impacted because of this never concerned me.

We all end up paying for this things one way or another but IMO we are a stronger society because of this. There are just some things I can't be libertarian about.

quadrupole writes:

The real interesting way to phrase this would be "What percentage pay cut would you personally be willing to take in exchange for paid paternal leave?"

That would be telling.

pyroseed13 writes:

But isn't Ruhm arguing for a government program that would distribute these payments? In other words, the costs wouldn't necessarily fall on the employer, unless it was funded like unemployment insurance currently is.

Hazel Meade writes:

The problem is that most people assume that the paid leave is "free". They think it does not affect their salary. All other things being equal, if you were offered free paid leave or no free paid leave, of course any rational person is going to take the free paid leave.

All sorts of questions about government benefits or mandates are phrased this way, in which the pollster phrases the question in a way that allows the unstated assumption that there is no cost involved to prevail. It's a clever way for someone who wants to advance a socialist agenda to get support?

Would I like a free sandwich? Sure. Would I like a free sandwich if it means I have to pay into a monthly sandwich pool from which sandwiches are distributed according to need? Maybe not.

Thaomas writes:

But I've noticed that when it comes to policies people DO expect free lunches, or at least will not eat unless lunch is free. Take minimum wages. Folks point out that transferring income to low paid workers will cost some jobs (most folks think not very many, but some). Lunch is not free, but some people are willing to accept the cost.

Don Boudreaux writes:


You write, about minimum-wage policy, that you notice that "some people are willing to accept the cost" of lost jobs. Do you mean the people who actually lose, or who never get, jobs because of the minimum wage are among those who are "willing to accept the cost"? If so, I must disagree.

My observation from following this debate for decades is that, above all, the vast majority of minimum-wage proponents deny that any trade-off exists for low-skilled workers. But even for those relatively few minimum-wage proponents who admit that the minimum wage reduces job prospects for the very group of workers that it is ostensibly designed to enrich, it is not the minimum-wage proponents themselves who are "willing to accept the cost." Instead, these proponents are willing to have others accept the cost.

mariorossi writes:

Isn't there a question of imperfect information generating a market failure in this situation?

Any single employee who requested paid paternal leave is unlikely to receive it in a negotiation with an employer. The employer knows that the employee has much better understanding of the likelihood of the employee actually using the benefit. So this kind of benefit can only be offered as part of a large group negotiation, where there is less information imbalance (the employer can value the likelihood of a large population using the benefit almost as well as the individuals). This is only possible if the employer is very large.

You could mitigate such information imbalance if you could commit to work for the same employer for a long enough time, but that's not an enforcable contract and you couldn't control for effort anyway. I don't think you'd want to make such contracts legal to be honest.

Even if an employee prefers a different package, he/she has no way to credibily comunicate his own information to the employer so no trade can happen. Since people mostly control when they have children, the information asymmetry seems very significant.

This seems like a significant problem to me.

Obviously parental leave is no free lunch: employees pay for it one way or the other. It's basically a form of borrowing: you get a benefit when mostly young which you pay back later on...

You could argue there might be better mechanisms to provide the support, of course: maybe grants or loans to new parents directly from the government. These already exist to be honest...

Uday writes:

I think if the government were to pass a paid leave scheme, they should first abolish the corporate income and payroll taxes

Thaomas writes:


I am referring to the discussion about policy. Some people favor a minimum wage (at least if the EITC cannot be made substantially more generous); other do not. A higher minimum wage transfers income to low income workers at the cost of some other low income workers. There is a trade off; there is no free lunch = income transfer with no loss of employment.

Paid maternity leave is just a sort of in kind wage which will lead to some decrease in the nominal wage. So policy makers face the same kind of trade off.

David Seltzer writes:

I was a graduate student at Chicago when Milton Friedman taught. A number of econ students stenciled TANSTAAFL on sweat shirts and wore them to one of his lectures. Amused, he gave the little chuckle for which he was known. Another Chicago student was inspired to stencil, "In practice that's great, but how does it work in theory".

Jon Murphy writes:


So policy makers face the same kind of trade off.

Again, to Don's point, it is not the policymakers who face this trade-off. Rather, they force this trade-off upon others. They may observe it exists, but that is not the same as facing the trade-off.

Tionico writes:

the original, as I remember it from way back when it first came out, was "There Ain't No Such Thing As A Free Lunch"

Thaomas writes:

This is the kind of information that we need to decide on mandated family leave. It's not enough to postulate, probably correctly that there is SOME trade-off (lunch is not totally free) but the cost of the lunch compared to its benefits.

[I hope the link works]


Jon Murphy writes:


It's not enough to postulate, probably correctly

It's not a postulation. It's a statement of fact. The mere fact we live in a world with scarcity means there are always trade-offs.

What's more, doing a simple cost-benefit analysis is not enough. One must apply it correctly. The point of Don and my comments is that cost-benefit needs to be done, but the people who are affected must make the choices. Otherwise, the cost-benefit analysis is quite meaningless (all this assumes the analyst correctly captures all aspects of the exchange. It is highly unlikely this is the case since, on paper, a mandatory maternal leave policy may make sense it is unlikely that every worker feels the same way as the "average" person or that it will benefit/cost each worker the same in their subjective valuations).

Comments for this entry have been closed
Return to top