Sunday, November 8, 2020

Covid cycles

Theory: (From An SIR model with behavior)

Fact: (from Scott Gottlieb via Marginal Revolution)


I do not mean to toot my horn, as many other graphs from the model did not look like that. This particular graph did, and really offers a sad interpretation of what's going on. In the model that produced the graph, people and policymakers react to the current death rate in deciding how much risk to take by going out. 

It is entirely individually rational for people to go out and party when very few around them are infected. Sadly, that means the disease collectively ramps up. Then it is individually rational for people to cut back, and the disease slows down. Cycles can result.  

Public policy is supposed to get on top of these cycles, by stamping out disease when it is low, the same way you keep taking antibiotics even when you feel better. It is the policy that has failed rational expectations here, not people. (No, that does not mean lockdown business and print money so we all can stay home and order stuff that comes by magic from Amazon. Ambitious testing would have done the trick. Or at least containing the summer's wave of super spreading parties.) 

Sherwin Rosen, Kevin Murphy and José Scheinkman have a beautiful JPE paper Cattle Cycles describing a related phenomenon. But our governments are supposed to be smarter than cows. 



16 comments:

  1. What should be the policy when a large chunk of people think its just another flu?

    ReplyDelete
    Replies
    1. Isn't "people think its just another flu" a result of government policy?
      --E5

      Delete
    2. Anecdotally, as soon as Las Vegas casino opened up, there were a herd of people outside who poured in like it the Superbowl. Drive by SF on a pleasant day and you find crowds packed tightly together in the parks, on the streets, and the sidewalks. I shouldn't generalize but I suspect people just don't care.

      Delete
  2. I've long wondered how the government response (at all levels, Democrat or Republican) would differ had it not been an election year.

    ReplyDelete
  3. Until the end, I thought you might be referring to the governments of Sweden and Japan. Oh, well....

    ReplyDelete
  4. I have a different theory. Covid is snaking through the economy and gets congested at points where its path crosses itself, and that is why we have cycles. It is exactly like federal spending, federal spending gets congested and caused a regime change every eight years, generally. In fact, the math that connects path search and differential equations over the Z plane was discovered by a research at Microsoft some twenty years ago, then he died in a mountain climbing accident.

    Toot your horn, but see if my post gets past the mark.

    ReplyDelete
    Replies
    1. I had a variant of this thought as well. The hypothesis is that R drops as superspreaders get infected and recover, choking off further exponential spread. But then patterns of behavior change and there is a new group of superspreaders and you have another wave. (Simple example: in warm weather the receptionist at the pool can become a superspreader, which changes to the receptionist at the health club in winter.)

      Overlaid on this phenomenon is the virus moving to new geographic locations that have not been hit hard before. These have been somewhat isolated previously, but once infections reach these locations they find a whole new group of spread patterns through which the virus runs.

      There are probably other aspects of the way that the spread works that are not well understood, but that could certainly produce this pattern.

      Perhaps much more testing or banning parties would have curtailed the spread more, but I am not so sure.

      Delete
  5. "But our governments are supposed to be smarter than cows."

    Bwhahaha! You're definitely an idealist Dr. Cochrane. Cows seem to be smarter at the present moment.

    Institutions matter just like North said. He was at Hoover, too.

    ReplyDelete
  6. "But our governments are supposed to be smarter than cows." I think the state governments are acting fairly rationally here. Regarding news of covid cases, direction of change is much more salient than magnitude of change, so the increase in approval rates from a large drop in cases is not much larger than the increase from a medium-sized drop in cases. Thus states relax restrictions when cases are falling, which boosts approval rates further. My alternative theory is that if states get the # of cases really low, then an influx of cases from surrounding states could bring news of huge percentage spikes, which is bad for approval rates. I haven't looked at the data, so I'm just speculating.

    ReplyDelete
  7. Confirmed cases are weakly correlated with deaths because early testing was largely confined to sick people in hospitals and now it includes precautionary testing of asymptomatic young workers and students. The positivity rate of testing in NYC was over 50% during the first peak but only 1.8% now that many neighborhoods are full of recovered people with some resistance ("immunity"). New virus hot spots are least likely in places that have already been hit hard. The April wave was concentraged in the Northeast which therefore was relatively untouched by second and third waves. The second July wave hit the South, which had largely escaped the April crisis). New outbreaks have largely been concnetrated in the Midweast and sparsely-populated Plains states, which escaped April and July waves. As odds of reinfection fall because of fewer susceptible targets, the virus reproduction rate will fall below 1 so that new cases will no longer replace those recovered and dead. A vaccine will accelerate this same process and, since it must be rationed for months, will narrow the scope of new infections to the young and healthy who will later be at very low risk of reinfection (exactly as if they had been vaccinated).

    ReplyDelete
  8. "Public policy is supposed to get on top of these cycles" .... Oh my goodness! Cochrane is a closet Keynesian! A sheep in wolf's clothing!
    --E5

    ReplyDelete
    Replies
    1. No such luck E5. Competent public health is not the same thing as Keynesian stimulus. Libertarian does not mean anarchist -- property rights, rule of law, public goods, simple externalities and public health are appropriate government functions. --BB

      Delete
    2. From your list of 'appropriate government functions', your revealed philosophy falls into the category 'classical liberalism', and not 'libertarianism'.

      See "https://en.wikipedia.org/wiki/Right-libertarianism" for an interesting discussion in contrast and comparison between the various 'schools' of libertarianism and between 'libertarianism' and 'classical liberalism' or 'neo-liberalism'. 'Libertarianism', and 'libertarian' differ from 'classical liberalism' and 'neo-liberalism' in significant aspects. For example, F. Hayek and M. Friedman, according to the article, referred to their economic philosophies as 'classical liberalism' (now referred to as 'neo-liberalism').

      The term 'liberal' has become so strongly associated with a particular political party in the States that there is an understandable desire to avoid using the term and variations on the term to describe political and economic philosophies that differ markedly from those associated with that particular political party.

      'Libertarian' is a term that differs sufficiently in sound and context from 'liberal' to serve as a differentiator, but it is incorrect as a label for what is, philosophically, classical liberalism.

      Delete
    3. I thought that public health was probably an appropriate government function at the start of the pandemic, but after seeing the poor job done here, I have some doubts. If government is to have a dominate public health role it should (i) demonstrate institutional competence and (ii) be dealing with a well known problem with a well known solution.

      In the case of COVID institutional competence is questionable (e.g. too many regulatory burdens) but potentially fixable.

      But COVID has turned out not to be a well known problem with a well known solution and this has made the government role especially problematic. There is disagreement even now about how COVID spreads, what the best NPIs might be, how dangerous its is, who might have immunity. In the face of uncertainty heterogeneous responses of actors in the marketplace usually are a much better means of information discovery than government directed mandates or R&D. In many of its actions, government run public health has choked off that information discovery with inflexible regulations and lockdowns. Most of these have probably been done in good faith, but they have fallen prey to many problems identified by public choice economists.

      At a minimum these problems mean that government run public health should have taken a much lighter hand and been much more modest in its claims of knowledge.

      Delete
  9. You are really missing the central point: Lockdowns are a dispute for access to public space. You are simply siding with the people that want to use force to prevent one group from using this space (healthy people) so that the other group can use it (less healthy people). Do you have scientific evidence that this has any impact on the rate of transmission? Not even that... But you want more use of force.

    I honestly find it very disturbing that even "libertarians" supports this type of tyranny.

    ReplyDelete
  10. if youre looking at national aggregate data and seeing cycles youre wrong. its just single peaks occurring at different times in different geolographical areas, which looks like a cycle in aggregate.

    ReplyDelete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.