On the proper allocation of the coronavirus vaccine


The Grumpy Economist is spot on—the best way to distribute the vaccine is via auction.

I say this because I actually care about those who desperately need the vaccine to survive. There are some very confused people out there who think that arguments for free-market solutions are heartless—but quite the opposite is true.

This is because, as with any other form of limited resource, these vaccines should go to those who need them the most. That is: they should go to those who, because of age or any other reason, will get the most value from their protection.

How can you argue with that? And luckily, modern science has a unit of measure for how much people value things: the dollar. It's not perfect, sure. But nothing else even comes close. Money is the unit of caring, after all, or put another way, it's the unit of account we use for Coasian bargaining. So let's take advantage, as we generally do, of the Pareto-optimal distribution of resources Coase promises us. 1

Sometimes people like to argue from externalities. That is almost always the wrong answer, and here especially so. First and foremost, externalities are not an exception to the Coase theorem. Yes, sometimes it can be hard to fully define property rights for externalities, which is a prerequisite for taking full advantage of Coase. But hard doesn't mean impossible: perhaps, after thinking deeply about the specifics for a while and not seeing a solution, it might not unreasonable to give up. But anyone who starts by throwing up their hands and declaring "externalities" in an attempt to avoid even thinking about property rights is being silly. That's not even an argument—it's an attempt at an excuse.

But none of that matters all that much here. The public benefit from increasing population-wide immunity is fairly constant: if it's one person rather than another who gets the vaccine first, it counts towards the same threshold of herd immunity. Even those who try to claim that the vaccine should be distributed in some other way will argue on the basis of who should be protected first (such as the elderly). It's true: the elderly should be protected first, and why? For their own benefit, in that they are the ones who suffer the highest personal risk to their lives and wellbeing from the coronavirus.

Many people have the very justifiable intuition that they want to help the poor. Sure! I'm in favor of that altruistic urge: give money to the poor. Have the government give money to the poor, even. But that's a separate issue. Don't try to mess with the one and only system that we we know actually works for distributing resources.

It's certainly true that many people across the board have gotten poorer as a direct and indirect result of covid. Indirect—in that many people have been unable to work for fear of covid. And direct—specifically, in that they're going to have to pony up for a vaccine if they want to be able to return to something resembling normalcy. 2

The government has taken the all-but-unprecedented step of just giving people money to help alleviate, even if imperfectly, those indirect costs of covid. Great! It can do the same for the more direct costs as well: either by giving people vouchers or whatnot that can be used to pay for vaccines, or, better yet, by giving people the value of those vouchers in cash.Giving the cash equivalent is just as good when it comes to making up for the cost of the vaccine, and is more efficient: both in the sense that there is no need to print and distribute paper vouchers, and also in the sense that it gives people more flexibility to do things like pay for the vaccine out of their own pocket even before the government manages to get its act together and get their money back later.

Insofar as you want the government to help people who are poor, that's the way to do it. It helps the poor no less (as it can be scaled up to provide however much money the government chooses to spend on vaccines), and it avoids interfering with getting the vaccine to the people who, again, need it the most.

But all that aside—even if, for some reason, the government wants to handle most of the direct vaccine distribution itself, for it to forbid the selling of vaccines is beyond the pale. Even if we assume that the government is surprisingly good at both logistics and at determining need (neither of which is remotely the case), it's an impossibly hard argument to claim that it's perfect at these. But to forbid others from even trying is in fact to make that claim. Will the government ever accidentally miss anyone? Will some bureaucratic mistake, or even just red tape, ever interfere with someone getting the vaccine they need in time? Even the most ardent supporters of central planning must concede that around the edges there will be imperfections.

Mistakes happen. If someone does indeed fall between the cracks, for you to say that they must have no recourse but to beg for bureaucratic reconsideration is senseless and cruel. It's the elevation of government bureaucratic to holy writ, the claim that the government is the be-all and end-all, that one deserves nothing but what the government, in its ineffable wisdom, chooses to give you.

Perhaps not everyone will necessarily have the resources to overtake such a bureaucratic slip-up by purchasing the vaccine on their own, especially if the government is expending its resources on the bureaucracy instead of giving out checks as suggested above. Nevertheless, for the government to put its foot down and forbid even the possibility of acting outside of the scope of of its processes is nothing but pure arrogance of the highest order.

Further reading:

  • Free Market Vaccines — the inspiration for this post. It also discusses several other issues, most notably the immense cost in human life of the delay in approving the vaccine. Strongly recommended.
  • Priority Care — about how prices help with the details of prioritization.
  • Covid 12/24: We’re F***ed, It’s Over — Zvi Mowshowitz discusses vaccine auctions in his latest in a long line of excellent posts about the state of covid.
  • A Vaccine Auction — about one particular way to structure a vaccine auction which would have some nice properties.


Sure, perhaps in practice we don't perfectly meet the preconditions of the Coase theorem, and so we're not quite at the Pareto-frontier. But even this approximation tends to work out pretty damn well, certainly better than central planning.


This cost exists as a real cost regardless of if you charge for it the way I'm proposing here or if you pay for it with tax money. Making the payment be indirect doesn't actually itself change the cost one way or another. It's still a real cost, even if you want to measure it in the lack of other government services that could have been provided with the same money. That's not to say it's not worth the price—it is entirely possible, even likely, that using that tax money to help defray the cost of the vaccine is the best way to use it.