Thread by @t0nyyates

Vaccines touch on a lot of economic policy questions. You could probably teach most of an econ course through the lens of vaccine policy.

For a start, public funding for free vaccines marks itself out from laissez faire. We could leave the private sector to do the research, manufacturing; and leave it up to the market to decide the price, and up to people whether they have it.

We decided not to, because we thought the market would produce vaccines too slowly, in too small quantities; + we made them free to boost take up which would otherwise factor in only personal health risk and not the wider externality of risks to others.

These were high speed and elemental decisions responding to an anticipation of disastrous market failure.

The procurement decisions invovled government as investor. Not wanting to put all eggs in one basket. Ensuring sufficient and efficient scale.

Those benefits had to be reaped weighing up the opportunities afforded by exploiting an international pool of expertise, vs the risks involved in cooperating with other states who could in extremis seize control of facilities or supplies.

Also, sadly, given the imperfections in our political economy [cue digression into further sub-module of said uni course], one fears that also weighed was the kudos of Buying and Winning British.

Procurement forced the parties on either side to sign contracts and to figure out strategies given those contracts. [Enter contract theory! Mechanism design!]

This activity reflected the history of politicians being bad at stuff, having the wrong incentives, and being corrupt; and it reflected the agency problems partering with a private supplier that is Too Needed To Fail.

The private party has to have enough of an incentive to do the job assigned, yet for this to happen in the face of what everyone knows, which is that the government's political and moral liability will make it a guarantour.

[Digression into comparable case study in financial regulation, deposit insurance, too big to fail.]

The 'nationalism' in vaccine policy of course expresses itself in manufacturing, distribution, donation decisions. As we are witnessing.

Cooperation is good if everyone sticks to the agreement. Manufacturing is also a portfolio decision. Get a big flood at your one and only factory [sound familiar] and you might be in trouble. If you are part of a large distributed cooperative network, much less of a problem.

But, agreeing to cooperate and finding out later that the other country all along intended to renege and seize vaccine facilities and output could be worse than not cooperating.

The decision of what to do with the doses you have revisits the nationalism/cooperation debate. A cooperative risk sharing agreement would direct vaccines to where most lives will be saved, which would be where the virus is most prevalent and the most vulnerable people are.

But if you hand over doses when others need them, will they reciprocate if the trajectory of the virus turns the tables on you?

It should be obvious that in the vaccine version of the econ/politics undergrad this is 'time for some game theory'!

Eg... can this be thought of as a one shot game? If you manage to cheat, is that a win you get to keep forever? Perhaps not, if there are further mutations to worry about. Or other points of leverage countries have over you.

In which case you might worry about your reputation as a cooperator/cheater. Or, given the way politics affects incentives, perhaps you can't afford to worry about that because you have an overriding need for the apparent win by the election, before the next round of the 'game'.

Whether you are a bunch of cheating or cooperating countries, how should you treat the non producing countries also afflicted with the virus. Sell it to them at the 'market' price, whatever that is. Or donate it to them. And under what terms?

What are the implications for your own selfish interests of laissez faire regarding the virus in non vaccine producing countries? Might we expect producers to behave 'ethically'? What leverage might the non producers have?

How should the producers' cooperative / cheating group behave towards covid denying countries? Can any kind of 'contract' be devised to make them do what is right for their own populations and the rest of us?

@toxvaerd1 and I are going to write about these things in an @EconObservatory piece hopefully. Which I reveal without implicating him in any of the botched bits of the above thread.

And I haven’t mentioned the ethics, economics and statistics that bear on the vaccine trialing process....

Nor how uncertainty colours each of the decision steps. Both the easy to calculate kind that you would find in a dice rolling game and the kinds harder to calculate because of the novelty of the situation.

Economists made a living pilfering the relevant mathematics and applying it to their own problems. And it pops up in vaccines everywhere.

It’s why you have the eggs and basket problem; why cooperation is hard to sustain (will the other country renege?); why lockdown easing is complicated ( digression into a course on control theory stolen from engineering, modified for ‘controlling ‘ humans).

Many aspects of the vaccination process manifest the explicitly intergenerational nature of the covid crisis. Trialling, prioritization, passporting, and just taking it.

Mortality rates are high for the old, but very small for the young . So vaccination is self-protection for the old, but largely a selfless act for the young to suppress the virus so that it does not kill the old. [bar hypothetical risk of mutations that later affect the young?].

Trialling for the old is tricky; could it be ethical to randomly consign some old to the certain lottery of death by covid, and others to get the uncertain amount of protection; not to mention subject their weak constitutions to unknown side effect risk?

Trialling for the young is tricky for different reasons. With what confidence can you extrapolate to the old [we bumped up against that one didn't we]? What is in it for the young? If they have bad reactions, they may have a long future ruined by them.

Delivery was similarly intergenerational. With a scarce initial supply, how to allocate them optimally to maximize life saved and minimize economic damage. We went for jabbing the vulnerable, rather than those who are most socially integrated [the young].

Then there's the issue of passporting + more broadly whether vaccines should be voluntary or compulsary. We go back to the externality inherent in individual behaviour. Which is large for most people for whom the mortality risk is very low [though many of those have old family]

It's nice not to have to brutalize ourselves and others by physically coercing people to have jabs. But it's also nice to have economic activity [including covid risky hospitality] freedoms restored too.

Logically in the middle is 'passporting' discriminating in allowing certain activities [hospitality/travel/work?] according to whether you have had a jab or not. This could help expand the economy quicker, benefiting the young, but it also discriminates against the young.

[that is, initially, and given we chose to vaccinate the old first].

Trialling also requires international cooperation. [Read antithesis of nationalism]. Sharing, pooling of data, resources and expertise. Rather is best done like that.

Best in the sense that you can get more expertise more quickly; and also more precision in your estimates of efficacy more quickly, by using more people.

Also there is the oddity that if one country seriously fucks up with the vaccine, it presents an opportunity for the rest to exploit high prevalence with vaccine tests. Provided you can get them to agree.

Alternatives being waiting - weighing the benefits of waiting [deaths avoided by jabbing us with something useless or harmful] with the costs [deaths incurred while the virus rips through us] - or conducting challenge trials and the sharpened ethical calculation that that entails

Anyway that's enough. Someone earlier replied 'you could do this [teach a course entirely through...] using climate change, and I agree. I hope that that course already exists.

Actually there's a sub thread to be added on beliefs, an important part of the vaccine process; and in econ generally we spend a lot of time studying how games or the macro economy play out as beliefs evolve, and corner examples like rational expectations [crudely true beliefs].

For example, when playing the various stages of the vaccine game [research, manufacture, distribute, donate] you have to decide whether to cooperate or be nationalistic, or promise keep or cheat later on.

To do that you have to figure out what the other country's strategy will be [form a belief about it] which means forming a view about their objectives, and constraints, and their beliefs [about your beliefs about their....].

Another example: once you get the results back about efficacy, you have to decide how to communicate them, and figure out what you want people to believe about them and how they will arrive at that belief. [Do they understand stats? If not, what do you do about it?]

Another: in trying to figure out how the economy will respond to different vaccination and lockdown release strategies, you want to know what people will come to believe about the health risks remaining and how they will respond to that...

[I forgot to make the connection with the EU wide and member state regulatory authorities approach to the reports of clotting problems with the AZ vaccine; a part of which is the calculation about how beliefs about safety respond].

Another beliefs example. Separating equilibria. If there are substantial numbers of unvaccinated, someone might make the calculation that the unvaccinated are risk-loving, and you are more likely to encounter the risk loving in risky places like night clubs.

This will deter the risk averse from going [even if they are vaccinated, because efficacy is not 100%] and make the expectation self-fulfilling. There are more compelling examples of this in the pre-vaccination phase.

[But you can see the problem surfacing again as worries about waning vaccinated immunity arise, or immunity in the face of new variants.]

That really is it. Should have been a blog post, sorry. And don't forget look out for the piece with @toxvaerd1 which will cover some of this territory.