Wednesday, April 16, 2008

Vaccination: Compulsion vs. Incentives

Thomas Pogge defends mandatory vaccination policies on the grounds that this is "the only way to overcome the collective action problem." It is not:
Let's distinguish two forms of regulation, reflecting the statist vs. Hayekian distinction. One option is to make the undesirable activity illegal. The alternative is to make it costly. More generally, we can regulate activities either by using the blunt instrument of the law, or else by the more subtle manipulation of market forces. I think the latter will often be preferable.

Let's suppose that $1000 is more than enough to counterbalance the public health costs of a non-vaccinated person. In that case, what possible reason could we have for denying people the right to opt out of getting vaccinated if they care so much that they're willing to pay this cost?


  1. Possible Reason #1: Some people who want to opt out are probably mistaken about the risks of vaccination (such as the autism myth), and letting them pay to opt out may not be a benefit to them. This is an especially strong reason when mistaken parents are putting their children (along with other people's children) at risk by not vaccinating them.

    Possible Reason #2: If the government mandates vaccinations, it may be signaling that vaccination is something good to do. If the government requires a $1000 fee for non-vaccination, it may be signaling that non-vaccination is desirable, if you can afford it. This is a bad thing to signal, for various reasons. For instance, it may make people think more selfishly when considering vaccination (not taking the benefits to others into account), it may convince people that the personal benefits of non-vaccination are greater than they are, and it may make people who can't easily afford non-vaccination upset. Also, if the public argument for vaccination is "this is something that each member of the community can do to benefit the whole community," people may be less inclined to believe this argument if they see that rich people can buy their way out of vaccinations, since this seems, on its face, to be inconsistent with the public justification of vaccinations. See here for discussion of related issues with putting prices on things.

    Possible Reason #3: The spread of contagious disease is a local, non-linear phenomenon. The marginal social cost of an additional non-vaccination is relatively low in neighborhoods where almost everyone is vaccinated, but much higher in neighborhoods where the vaccination rate is near or below the threshold where the spread of disease becomes much more likely. Since there will probably have to be a single nationwide price for non-vaccination, despite the variability between neighborhoods in wealth and vaccination rates, the price probably can't be very well calibrated to marginal cost. This means that the 'fee for non-vaccination' set-up will be less efficient than a real market, with smaller net benefits, which weakens the case for using fees instead of mandates (although, on its own, I guess this is not a reason for mandates instead of fees).

  2. How about: option 2 unjustly advantages rich people. Is your motivation for objecting to Pogge with this alternative based on a defence of freedom to choose (vs. paternalism)? If so, then why restrict this freedom only to bankers, lawyers, people without children, etc.?

  3. Barry -- I don't think that's any objection to the present proposal. Maybe it's an objection to having a market economy in the first place, or at least one where some people have extremely limited purchasing power. We should institute a universal basic income to increase the real freedom of the poor. But limiting everyone's options equally is certainly not the answer (cf. the 'leveling down' objection to egalitarianism, which I consider decisive).

    Note that this can be made equivalent to raising taxes across the board, and simply paying a $1000 bonus to everyone who agrees to get vaccinated. I assume you don't think this "unjustly advantages rich people."

    (That may be a better way to frame the policy, actually, in light of Blar's very interesting point #2.)

  4. Hi Richard,

    I do agree that distributive inequalities, insofar as they cannot be disentangled from the allocation of goods, will be part of an objection to the market. But I wasn't relying on that.

    I also think that egalitarianism can withstand the levelling down objection, though this is perhaps not the place to give my reasons. However I wasn't relying on that either.

    Nor do I think a universal basic income is relevant to my objection.

    My point was that the current question is an applied one. In deciding it we need to take how things are as given. That means taking for granted an unjust distribution of wealth, in my opinion. Given that, if we think it is important to be free to choose X then we should not restrict access to X to rich folk, for doing so will restrict it unjustly.

  5. Barry, what do you think of the proposal to give a $1000 bonus to those who agree to get vaccinated?

  6. Sorry - I meant to say. I don't think this is any different. The net cost to poor people is the same. Rich people will in fact be sacrificing much less to relinquish their bonus than poor people. Any given rich person may not know to the nearest $1000 how much cash is in their current account. But it might be equivalent to a poor person's month's wages.

  7. What is the "cost" to poor people of offering them the option to gain $1000?

    As far as I can see, my proposal does not hurt them at all. (It's a Pareto improvement: some benefits, no harms.) Your grounds for objecting seems to be that it benefits the rich even more. Hence my invocation of 'levelling down'. For the sake of "equality", you would restrict the freedom of many, and increase the freedom of no-one. If that doesn't strike you as wrongheaded, I'm not sure what else I can say.

  8. the problem is that once you introduce the option of a reward for doing X, the people who need the reward more will be more inclined to do X. in fact the $1000 is worth much more to a poor person than a rich person. for the poor person it may be completely irrational to opt out of the vaccination. the rational poor person will hence be less free than the rational rich person on this proposal.

    there are two reasons why the $1000 offer is not a pareto improvement. for one thing the money is not coming from nowhere. it is state money that could presumably be spent on public services or invested into the economy. if the government is responsible, then the $1000 would accrue to most people more or less indirectly anyway. so its not simply that in one case they are forced to get the vaccination and in another they are given $1000 to get it. for in the first case they were effectively given public services instead of cash.

    the second reason is better: in the move from, say, spending the set of $1000's on public services and enforcing vaccination to giving the money to those who get vaccinated, there is an additional status inequality. rich people have the freedom to rationally opt-out of vaccination. very poor people have not. if even one poor person suffers a loss of well-being on account of this asymmetry that would suffice to deny pareto. but i reckon there would be systematic loss of welfare.

  9. "the rational poor person will hence be less free than the rational rich person on this proposal."

    Again, the problem is entirely in the background fact that poor people are already less free than rich people. My proposal does not make the poor any less free than they already are. It does make the non-poor more free. But that's a good thing. (It's a good thing for anyone to be more free, so long as it doesn't make anyone else less free.) You keep looking at the question through the distorting zero-sum lens of class warfare and relative benefits. The way to tell whether the poor are harmed or benefited by my proposal is not to compare them to the rich under my proposal, it is to compare them to the poor under the status quo!

    "for one thing the money is not coming from nowhere."

    I was assuming a tax hike on the rich to pay for it, thus building in a redistributive element (a very mild UBI, if you like). But in fact the Pareto aspect is clearer in the original version (i.e. replace strict compulsion with a $1000 opt-out option). All we've done is to add an option; nobody has had any options removed. So it's certainly a pareto improvement in freedom.

    Your second reason - "status inequality" - is more interesting. But I give it zero normative weight. It's morally repugnant to restrict some people's freedoms merely because of others' envy/resentment. Down that road lies Harrison Bergeron.

    (Pareto improvements become impossible if you allow such jealous preferences to count, so that a benefit for one person is ipso facto a "harm" to others.)

    In sum: I grant that radical egalitarians may oppose my suggested policy since they think that increasing the freedom of the rich is an intrinsically bad thing. But I don't think we should accept such an illiberal view (to put it mildly).

  10. I'm so sick of the leveling down objection. There are a million answers to it, but the one that I'm inclined to raise here in Barry's defense is a bullet-biting one: there are some situations where it is OK to level down.

    Obviously, we ought not to level down in terms of goods that matter -- no Harrison Bergsonish weights on the dancers. But is it really the objectionable kind of "leveling down" to refuse to give the rich a stupid little privilege (paying to get out of vaccination) in the name of equality?

    Also, do you support allowing rich people who are drafted into the military to pay poor people to take their places? After all, that's a pareto improvement in freedom... and a pretty good reductio, I think.

  11. "do you support allowing rich people who are drafted into the military to pay poor people to take their places?"

    Sure, why not? (Well, there are confounding political effects in this case, if the rich are the ones who get to decide whether to implement a draft, and they can more easily escape it, this creates a sort of political moral hazard. So I'd be cautious for that reason. But the mere fact of "inequality" is no moral problem. It shouldn't bother any reasonable person from behind the veil of ignorance. Again: jealousy is unreasonable and repugnant.)

  12. Isn't the objection to egalitarianism based on the repugnance of jealousy just an example of the genetic fallacy? Whether egalitarian policies would reduce jealousy, or whether those who support egalitarian policies are motivated by jealousy, is irrelevant to the normative standing of egalitarianism.

    I suppose a charitable way of understanding the jealousy objection is to suppose that itm is directed against arguments in defense of egalitarianism that include premises like "inequality makes people jealous" and "making people jealous is bad," in which case the jealousy objection amounts to a claim that people are responsible for their own morally culpable emotions. But since no argument for egalitarianism sounds like that, well...

    (Also, a little pithier: privilege is at least as repugnant as jealousy.)

  13. Huh? Maybe I'm missing something. I thought that levelling-down egalitarianism just is an expression of jealousy writ large, i.e. a preference that others not be better off than you are. (Maybe this is a loose sense of 'jealousy', since it needn't be tied to any particular emotional feel.)

    It just strikes me as completely insane to think that there's anything intrinsically bad about some people being privileged. What we should want, of course, is for everyone to be so privileged. But better to have some people benefitted than none. Again, the veil of ignorance makes this clear -- there's no way levelling-down policies could pass this test.

  14. That is a pretty loose sense of jealousy, and I doubt that the (alleged, debatable) repugnance of the strict sense carries over.

    In particular, it's loose in two ways, not just one. It's loose in the way you identified (it's not attached to a particular emotional state), but it's also loose in a more important way: it's not attached to any particular individual-level disposition. When we talk about jealousy, we usually think we're talking about a person who is worse off having a disposition to take stuff away from the person who is better off. But when we talk about egalitarianism, even "leveling-down egalitarianism," we're talking about a collective disposition to express our equality by not structuring our relationship such that it has unnecessary privileges.

    So imagine that we're behind the veil of ignorance. We have a choice between two worlds. In one world, nobody is allowed to wear purple hats. (Suppose purple hats are very expensive.) In the second world, only people who were descendants of William the Conqueror are allowed to wear purple hats. To choose the first world would be an example of "leveling-down egalitarianism," but we can imagine that the representatives in the original position would do so, for the reason that choosing the second world would give offensive inherited privileges in exchange for no real benefit to anyone. The di minimis "pareto improvement in freedom" isn't worth the collective cost in creating social distinctions.

  15. "choosing the second world would give offensive inherited privileges in exchange for no real benefit to anyone."

    I can understand it being "offensive" if one is unjustly prevented from wearing purple hats oneself. I can't imagine why any reasonable person would care that Willy's descendants get to wear purple hats. It's perverse to resent that others receive a benefit. Each person should be intrinsically pleased (or at least indifferent) that Willy's descendants have this extra freedom. The only legitimate grounds for concern would be the possibility of instrumental badness down the line, e.g. if the purple-hat wearers decided they were superior and so began to oppress others, making them worse off in absolute, and not merely relative, terms.

  16. You're missing the expressive features of the situation. By denying the purple hat to all, rather than giving it just to the bluebloods, we (collectively, bluebloods and redbloods alike) are expressing respect for one another as equal citizens by refusing to create unnecessary status distinctions. And creating the status distinction would be offensive just because it would express a lack of respect.

  17. I agree that one shouldn't express disrespect. But we should increase freedom. Hence, to avoid unnecessary conflict here, it would be most inadvisable for us to adopt conventions of social meaning according to which increasing freedoms for some was understood as expressing disrespect for others. If such conventions are already present, we should do what we can to undermine and change them. (I personally don't perceive any necessary 'disrespect' here, so it must be possible to overcome this pernicious attitude.)

    On the other hand, I suppose one might argue that in the majority of cases there will be no good reason to restrict access to only a subclass. So a general aversion to inequalities of access, though strictly unwarranted, may nonetheless prove globally optimal in bringing about positive social change. I'm not sure whether that's true, but it'd be an interesting line of argument.

  18. Your first paragraph makes lots of sense, but to take it seriously, we'd have to be really careful not to allocate goods on the basis of wealth. In our history, wealth distinctions have traditionally been really strongly associated with disrespectful treatment by the wealthy toward the less wealthy. ("Let them eat cake.") Giving more privileges to the wealthy just reinforces that practice.

    Perhaps, then, we could allocate the limited-supply good of non-vaccination by, e.g., determining how many people (per area, etc.) could non-vaccinate without threatening public health, then holding a lottery, and then permitting the market to reallocate in some fashion that isn't wealth dependent. (Maybe people could trade non-vaccination for some other kind of social privilege that was allocated similarly.)

  19. Or, more simply, redistribute wealth in the first place.

    Though in the meantime, I think very few people (Barry aside) would interpret my alternative framing -- i.e. a $1000 "public health bonus" for getting vaccinated -- as expressing disrespect for the poor. So that could be happily implemented even now, right?

  20. Hang on a minute.

    Paul has done a great job of arguing against the view that privileging a minority may have unfortunate consequences. More can be said; but there remain two more pressing questions: one, whether we can enable freedom of access without privileging anyone, and two, if we are going to privilege some folk, who should it be?

    On the first, unless I missed it somewhere, some form of income-assessed charge would assuage some of my concern.

    On the second, even if wealth had been distributed fairly, there may be better subsections of the population to privilege than the rich folk. However wealth is not fairly distributed, alas. So why not privilege those who have been otherwise underprivileged, or worse, treated unjustly by an imperfect system? (The logistics could surely be arranged with taxes.)

  21. generally paying people to do things makes them feel better about it (particularly when it is a social good) and fining them makes them feel worse about it.

    For example speeding became less "ok" in NZ after people started getting fines for it.

    It seems to me unlikely that vaccination would be a special case.

    However I don't like the idea all that much because money must come from somewhere and it would create clusters if unvaccinated people etc etc

  22. "So why not privilege those who have been otherwise underprivileged, or worse, treated unjustly by an imperfect system?"

    How? Show me a freedom-increasing proposal and I'll support it in an instant, just as I did the freedom-increasing proposal discussed in the main post.

    Now, I made the most obvious freedom-increasing proposal relating to this particular issue. I think it's very obviously an improvement on the status quo. But I grant that there may be even better conceivable improvements, e.g. combining this with wealth redistribution (UBI). [Though, as suggested in my first comment, it seems a bit off-topic to bring up those broader issues here.] I assume that whatever more complicated alternative policy you have in mind ("The logistics could surely be arranged with taxes") will just be some narrower and less efficient (but similarly politically unrealistic) version of this. No?

  23. Hi RIchard,

    Let me see if I have your argument straight, and then try again to get you to see my perspective.

    We want to allow people to be able to opt-out of vaccinations but only at significant cost. You proposed $1000 for everyone who got the vaccination.

    I opposed this for two similar reasons. Firstly the real value of $1000 is far more for poor people than rich. Secondly the real value of $1000 is sufficiently low as to be inconsiderable for some rich people and out of the question for some poor people. This would not be a problem in a society in which wealth was distributed fairly. But since wealth is not distributed fairly in our society, this is a problem. (Notice that this is the only broader claim I am making, and it has the status of an obvious social fact.)

    In effect, your solution to the problem is to privilege rich people. Then we had a discussion about pareto and so on. That is interesting, but my argument doesnt turn on it.

    My claim is this: if indeed there is no solution to the problem which does not involve privileging some section of society, I don't see why we should privilege rich people. My weak claims were that this was arbitrary and that other people are in greater need of social privileged. My strong claim is that this is horribly unjust.

    If indeed there is no solution that doesnt involve privileging some section of society, why not privilege poor folk, or public sector workers, or single parents, or old people. Doing this will be easy: we simply say: vaccination is compulsory for everyone apart from public sector workers. If you baulk at this that is only because your proposal is less forthright in its privileging.

    But I suggested a solution without privilege: we have an income-sensitive amount of money which is returned. According to this scheme it would cost poor folk less to opt-out than rich folk. This is perfectly consistent with society-wide norms of distribution.

  24. Note that part of my policy is that the opt-out cost is meant to be sufficient to compensate for the public health costs for society to tolerate a non-vaccinated (pandemic-prone) individual. To arbitrarily "privilege" some group in the way you suggest would screw up these incentives. It's not as though I'm saying any rich person can opt out of vaccination for free. Though market-based solutions invariably have the effect of "privileging" the rich in some sense, there's nothing arbitrary about it. (Nor is it a formal constraint built into the policy -- the limitations are not as clear-cut as you imply. Lower-middle class people may choose to make other sacrifices in order to afford this cost, if it is very important to them.) There's a cost involved, and this creates efficiency (within limits; cf. Blar's objection #3).

    Also, I dispute your characterization of the situation, i.e. the claim that it is essentially "a problem" to benefit the rich. On your view, every new technological advance, every improvement in elite education, every expensive new medical technique, all these additions to the market must be seen as "problems" (at least to some degree). This is an obscenely illiberal view, and I couldn't disagree with it more strongly. Hence all the 'pareto' stuff. You view does turn on this, because without it -- i.e., from a liberal rather than radical egalitarian perspective -- your objection is a non-starter; there is no "problem" here for you to object to.

    Anyway, since you're so puzzled by why we should want to "privilege the rich" (i.e. implement a market-based solution), let me just list some of the advantages over your proposals:

    * It is non-arbitrary. Only those who are able to pay the public health costs of their opting out will be able to do so. This is manifestly fair. It's a classic externality problem (cf. pollution, etc.). If you are going to impose a cost on others by becoming pandemic prone, you ought to pay this cost yourself, rather than externalizing the cost and expecting others to bear it.

    * This is the core problem with your "income sensitive" proposal: it effectively subsidizes poor people to make bad public health decisions, since they don't pay the full cost of their choices. That's simply bad policy. Give poor people money if you want -- then they can use it for whatever (hopefully good) choices they want. But subsidizing undesirable choices is a terrible idea. Policy-makers need to pay more attention to the incentives created by their policies.

    * Your "free opt-out for public sector workers" proposal is, of course, even worse in this respect.

    * It is also less efficient in the sense that it is not a good method for distributing the limited "good" of avoiding vaccination to (all and only) those who desire it most. The absence of any pricing mechanism means that any old Joe worker might avoid vaccination simply because they can -- with disastrous public health consequences. Further, Paul the private sector worker has no possible means of securing the good no matter how desperately he hates vaccination. (Maybe he has strict religious objections; maybe he's terrified of needles; maybe he has a health condition which makes vaccinations risky for him. Whatever his motivation, a market-based solution lets him act on it. Bureaucratic policies, by contrast, won't be able to take into account every such eventuality.) Even if Paul is poor, he has the opportunity under my policy to pay to opt-out, if he cares enough to make the requisite sacrifices (work more, spend less on other items, get a loan, etc.).

    * This point in my favour is enhanced significantly by the addition of a UBI, which I think makes for the ideal policy combination.

    Here's a quick dilemma: are we after an ideal policy, or a politically realistic one? If the latter, your anti-market proposals are not realistic (in addition to being less desirable for the reasons I've pointed out). If the former, they are very obviously less ideal than combining market policies with wealth redistribution (UBI). So, either way, there's no point in advocating illiberal, anti-market solutions.

  25. (Which is all to say, I can "see your perspective"; I just strongly disagree with it!)

  26. As Richard notes in the current system every thing that involves money is in a sense a denial of freedom to the poor. If we don't like it that much we are probably communists or something similar.

    regardless, on a practical level I think it is a major policy mistake to try and solve every policy issue (including this one) in every decision.

    there are, of course, some policy instruments that are more effective than others at achieving aims (such as increasing freedom of the poor). If you spread out that effort amongst less efficient mechanisms you will just achieve the same thing at a much greater cost to all your other goals. The UBI is an example of a fairly efficient method.

    > we have an income-sensitive amount of money which is returned.

    We already have a tool for getting money off the rich (tax) and for giving to the poor (social welfare) there is no need to reinvent a whole department of government to do the same things all over again.

    Also poor people are the key demographic for requiring vaccinations. they are the community that would probably initiate the outbreak and would suffer most from it. If Richards equilibrium price is $1000 the one in your system would probably be $995 and add almost no freedom to them (otherwise it would expose the poor community to the virus outbreak). On the other hand for warren buffet I presume you would want it to be negative 5 billion or so* which would deny you the ability to use that incentive to make him do other things besides 'get vaccinated'.

    *just imagine the corruption incentives!


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)