David R. Henderson  

A Pareto-Optimal Move

Where Is the Political Flynn E... Bailout Watch...

Economists often talk about Pareto-optimal moves, that is, changes in policy that make some people better off without making anyone else worse off. But we have trouble coming up with any real examples. It's an easy exercise to show that eliminating restrictions on sugar imports into the United States would create greater dollar gains to U.S. consumers and foreign producers than the losses to the domestic producers and the few favored foreign producers who would lose the scarcity value of their quotas. But it's not a Pareto-optimal move.

An editorial, "Senior Liberation Act," in today's Wall Street Journal, though, discusses such a move or, more correctly, as close to such a move as we're ever likely to see. In 1993, the Clinton administration, without any change in legislation or even the standard change in regulations, started to disallow people who receive Social Security and who are 65 years old or older from leaving Medicare. Some seniors want out of Medicare. Letting them out would be Pareto-optimal. If the government forces them to stay in or lose Social Security benefits, the vast majority of the small percent who would have fled Medicare would stay in and the government would spend money on them. But letting them out would reduce the government's spending on Medicare with an increase in Social Security spending only for that tiny sliver of the population that wanted out of Medicare so badly that they were willing to forgo their Social Security benefits. The Journal writes:

If even 1% of Medicare-eligible retirees voluntarily opted out, Medicare expenditures would decrease by about $1.5 billion a year, and by some $3.5 billion a year by 2017.

Changing this government policy, which some Medicare recipients are suing to do, is not literally Pareto-optimal. If a substantial number of Medicare beneficiaries leave, then a small number of federal government employees might lose their jobs administering the program and would go to their next-best job, thus losing some producer surplus. But it's close to Pareto-optimal.

Comments and Sharing

COMMENTS (14 to date)
Marc writes:

Lose their jobs?

Prakhar Goel writes:

Please... When was the last time that government was able to think in such clear logical terms?

Likely scenario: There is some union/other interest group interested the government managing health care.

Richard writes:

Why isn't this an adverse selection problem, with the healthiest seniors opting out?

David writes:

Richard - it's not an adverse selection problem because they're not paying anything for Medicare. From the article: "even though they paid Medicare taxes throughout their income-earning years and though they are not asking for that money back."

So, sure, maybe the average cost per person goes up for Medicare, but the total cost would go down. Adverse selection only applies to insurance programs, not welfare programs.

Scott Wentland writes:

Some politicians would feel bad if people left Medicare...

David R. Henderson writes:

Bryan Caplan pointed out that the usual term is not a "Pareto-optimal move," even though I can find instances of that in the literature, but a Pareto improvement. Point accepted. Thanks, Bryan.

Rolo Tomasi writes:

I agree with Marc. The government will not fire people simply because they have nothing to do. I think we can safely say this is an example of a (actual not almost) Pareto improvement.

Dick king writes:

OK, David, I don't see it. Why would people withdraw from a free program? Surely there must be some downside to being in medicare, such as a copay or monthly fee or some legal requirement that any doctor that sees you must obey even if you are willing to pay cash?

If there's anything like that, then the taxpayers might lose when some people withdraw and there could be an adverse selection problem.


David W. writes:

dk - I looked up the Medicare fee schedule, and you're right, there is a charge - 96.40/mo (or higher if you have income above 85K/person). In addition, anyone with less than 10 years of Medicare taxes on record has to pay an extra 443/mo. I have to admit, I'm surprised, given the size of the taxes for Medicare, that they bother to collect anything from the beneficiaries.

Given the prices of medical services, I'd be very surprised if any significant portion of people who've paid their Medicare taxes incur expenses less than that, but it is possible.

I guess I'll have to oppose mandatory Medicare for other reasons.

David R. Henderson writes:

Dick King and David W.,
What David W. found on line was the premiums for Medicare Part B. As far as I know, there is no requirement to sign up for Medicare Part B. The requirement these people are opposing is to join Medicare Part A, for which premiums are zero.
David Henderson

Jayson Virissimo writes:

David, I'm glad you joined the Econlog team. Econlog is one of my favorite blogs and I think it'll be even better now that you're here. Also, I appreciate the work you did on the Concise Encyclopedia of Economics. Its a great reference work. The reading lists at the end of the articles are especially handy for autodidacts like me.

David R. Henderson writes:

Thanks, Jayson.
P.S. I'm still learning. I don't know if I'm supposed to respond to compliments, but what the heck.

quadrupole writes:

Why would someone opt out of medicare?

How about to get better care.

Increasingly, doctors are opting out of medicare. As in most systems, your top folks leave first... so if you want to see the best doctors and you are a medicare patient you are increasingly out of luck.

Why not pay yourself? Well... the process of a doctor who has opted out of medicare taking a medicare patient as a private contract patient is very complex and if mishandled, can wind up in large fines and prosecution of the doctor... so most simply opt not to treat medicare recipients under any circumstances.

Bob Murphy writes:


I second Bryan's complaint. If a student had put "Pareto-optimal move" when it was not a move to a Pareto optimal position, then I would have taken points off. :)

BTW if you want to have a cute trick question for your students, tell them to imagine s situation where there are enough cranky citizens such that any change in government policies will always make at least one person less happy. What does this imply about Pareto optimality? I bet at least 90% will get it wrong.

Comments for this entry have been closed
Return to top