Tom Means, a fellow UCLA Ph.D. in economics and a retired economics professor at San Jose State University, made a point to me the other day that seems obvious: California is too diverse a state for the same regulations to apply everywhere.
I take as given that social distancing makes sense. My wife and I started practicing it on or about March 10. In fact, one of my strong regrets (and I think “regrets” is too mild) is that the mayor of San Francisco and the governments of 7 urban counties in Northern California didn’t give social distancing a chance. I had gotten emails from Starbuck’s and from various restaurants about their plans to limit the number of people inside so that people would be at least 6 feet from each other. That was in place for a few days before the SF mayor and the urban county governments made their move. Do I know that it would have worked? No. But do I know that it would have had much less of a destructive effect on people’s economic lives? Yes. So I would have liked to see how much compliance there would have been before these governments took the next step.
But they did and so did the governments of the urban areas in southern California. Together, both northern and southern California urban counties probably account for 90 percent of California’s population. It’s not hard to understand, therefore, how a politician, Governor Newsom, looking at numbers, decided to put all of California under the shelter-in-place order. Government officials usually think that way: they look at the 90 percent, especially when it includes them, and don’t think hard about the 10 percent.
Tom Means made that point to me and passed on a note that he received from a friend of his. I’ve edited it slightly for grammar. Here it is:
Professor,
Its amazing how many people argue for a lockdown across states or even a country. If you look at the number of cases by county in California, you will see that the largest and probably the denser(?) counties have 90% of the cases. I live in Eldorado County which is a large county ranging from El Dorado Hills (where I live in a 55+ community just next to Sacramento county) and all the way up to South Lake Tahoe, a popular tourist attraction. As of last week, the county had 2 cases, like many other counties. Yesterday it increased to 3 cases, a whopping 50% increase. Nevertheless, many in our city are calling for complete isolation. They cite forecast and simulations of data, exponential growth formulas, that may apply to cities like NY, but clearly don’t apply to most of the counties and cities in California When I ask on our blogs what explains this discrepancy, they cannot answer. When I point out that maybe this dispersion in cases suggests that from a benefit/cost perspective we shouldn’t impose a one size fits all rule, they gasp in fear.
I would add that it’s much easier to socially distance when you live rurally and still go about much of your daily life. I think governments should let them.
The vast majority of people are afraid. I’m one of them. But we shouldn’t let our fear stop us from thinking on the margin.
This site shows a breakdown of cases and deaths by California county.
READER COMMENTS
Ian Fellows
Mar 30 2020 at 1:23am
It used to be that everyone was an armchair economist because everyone has a stake in the economy. Now the tables have turned and everyone is an armchair epidemiologist, especially the economists.
john hare
Mar 30 2020 at 4:52am
I live out in the woods and work construction with a small crew. For us, social distancing is very easy. When I’m sweating in the Florida summer, people have been social distancing from me for decades. Town with two (2) red lights is a mile away. Three other towns are about ten miles each in three directions with populations in the low to upper tens of thousands. For us, social distancing is a matter of slightly increasing the normal distancing and avoiding handshakes and gatherings. I have a hard time picturing this working in cities where elevators and buses are in common use.
Mark Z
Mar 30 2020 at 5:35am
I think for a lot of businesses that depend on a large number of customers per hour it wouldn’t be worth it to operate under these conditions anyway. For example, a restaurant in a big city (that has to pay big city wages) likely won’t bother to stay open if it can only have a few customers at once. For stores though I think it may be doable; there may be ways they can spread their customers out across time and lower density without sacrificing as much business. But for most bars and restaurants I think social distancing may be almost as bad for business as lockdown.
By the way, unfortunately, I think you’re going to win your bet with Charlie Hooper, and by a good margin at that. >1,000 fatalities in New York (state) already, mostly in the last week, and still a few weeks away from peaking. If it’s half as bad in the rest of the country I expect the US will reach 100k deaths by the end of summer.
Michael
Mar 30 2020 at 6:52am
A couple of points of skepticism.
Because of the exponential growth of the number of coronavirus infections, and because major metropolitan areas were likely infected first, I would not want to factor the low number of rural cases as compared to urban ones too heavily into my analysis.
If a shelter in place order was going to be enacted, a statewide order may make more sense than ordering shutdown in the major metropolitan areas only. A less-than-statewide order would likely induce some degree of exodus from cities into less populous areas, likely bringing the virus along with them.
That said, I do think rural areas, if they are good about adhering to best practices, should be capable of weathering the storm more easily than more populous areas. But in a way, rural populations are also at greater risk. Former (Obama Admin) director of CMS has noted that the biggest risk factor for dying during this pandemic is likely to be “need for a ventilator”, and I’m sure that people living in less population areas have reduced access to ventilators. I’d expect lower incidence rates but also lower survival rates among those affected in less populated areas.
JFA
Mar 30 2020 at 8:01am
Regarding the one size fits all policy, if speed is of the essence (as many people believe in the current moment) then taking the time to figure out what policy is best that takes into account all the differences in demographic and other factors of particular geographic areas might be too slow. And if there were to be differential social distancing guidelines for each locality, there would probably be a lot of rent-seeking to adjust those rules to exclude “my town” from more restrictive guidelines.
But those are separate considerations from whether “shelter in place” is the right policy.
Jon Murphy
Mar 30 2020 at 9:04am
I get your point, but allow me to push back on it a little:
A quarantine in a case like COVID is probably about as close to a true collective action problem as we will come in real life (unlike many things that are called collective action problems, such as many types of externalities, but are not). The quarantine is only effective if everybody does it. If people break quarantine, they risk being both carriers or getting sick. Furthermore, since this disease doesn’t show symptoms for up to 2 weeks after infection, people who are carriers may not even know and unwillingly spread the disease. Of course, all this is known to you.
But, when faced with a true collective action problem, the whole point of the solution is that there must be a one-size-fits-all solution. Individuals may be harmed on net, but the overall outcome is Kaldor-Hicks superior (or can be made Pareto-superior through compensation). The coordination is needed because it is in each individual’s interest to defect.
So, I think, if a government is going to do quarantine, it has to be one-size-fits-all given the nature of quarantine.
Jon Murphy
Mar 30 2020 at 9:06am
Point of clarification: I write on “quarantine” above and the original post is on social distancing. Lest it appear I am doing a bait and switch, I see social distancing and quarantining as two sides of the same coin. True, they are not identical, but for the purposes of my point, they are sufficiently similar.
Christophe Biocca
Mar 30 2020 at 9:40am
If that was true we wouldn’t have any exempted services, but there are quite a few. Even in Wuhan (which was the strictest lockdown seen so far), people would be allowed to go grocery-shopping every 3 days. A slightly more realistic model is that each reduction in overlapping human presences anywhere slightly reduces R0, and there’s enormous shared benefit in getting R0 strictly below 1, with milder benefit to further reductions. Which sounds awfully like an externality.
Now in practice taxing people based on the number of distinct individuals they interact/share surfaces with over 2 weeks (or more accurately, those individuals that weren’t themselves interacting directly) is unviable , so we fall back to regulatory methods as a second-best solution. But:
To the extent that there are exemptions (and we need them if only to avoid people starving), the current approach lets politically connected firms get exemptions (realtors are now essential services, Ontario’s government-owned liquor and cannabis stores are still open over here, etc).
Given that R0 isn’t an inherent property of the virus (hence why interventions to reduce it are considered at all) and that, combined with decreasing returns for R0 below 1, means regions with a lower baseline R0 would benefit less from interventions, so allowing them to adjust individually makes sense.
Thaomas
Apr 2 2020 at 8:31am
Actually, it’s not clear to me that the externalities are wide spread enough not to leave regulation at the local level, Henderson’s point, I think.
Concerning what those local regulations should be, it seems better to regulate proximity/congestion/infection-able contact than mandating everyone stay at home as a means of slowing the spread, although I recognize that “lock-down” might be a kind of second-best policy in some places, given that it wold take time and data to come up with an optimal policy.
Alan Goldhammer
Mar 30 2020 at 9:24am
David’s comment is valid up to a point. As some have noted, hospitals in rural areas are ill equipped to deal with more than a small number of patients who present with serious clinical symptoms. If hospitals are overwhelmed with SARS-CoV-2 cases, they will be unable to assist patients who are in need of assistance for other serious conditions. IMO, this is one of the key reasons for carrying this out. I suspect that most who follow this blog are practicing social distancing but think about what might happen if you slip in your home and break an ankle (this happened to my mother in law in our home 12 years ago). You call the EMT but there is no where to go because hospitals cannot take you.
As part of my daily research email, I’ve been looking at disease modeling. There is a nice pre-print looking at South Korea, where the outbreak began in church group and rapidly spread outward. They used metro traffic data from two large cities to look at social distancing and found a good correlation to control of new infections. Reports on other geographical locales are also being disseminated.
To quote former Secretary of Defense Rumsfeld, “But there are also unknown unknowns—the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones.” This is the crux of the problem and it all goes back to epidemiology. We have not done nearly enough testing to make any kind of prediction. When the pandemic wanes, we may find that the current control efforts were too stringent but nobody can state with certainty that is the case today.
There are going to be a large number of lessons learned as a result of the many missteps this year and missed opportunities in past years.
Alan Goldhammer
Mar 30 2020 at 12:54pm
For those interested, it appears the US task force is using THIS model for planning purposes. It’s being updated by Dr. Murray who is at University of Washington. The daily mortality rate is expected to peak in mid-April. There is also a link to the assumptions that went into the model. This is not inconsistent with other international models that I have seen.
David Seltzer
Mar 30 2020 at 6:26pm
Per John Cochrane: “The reproduction rate is the average reproduction rate. But not everyone is average. Every interesting distribution has a fat tail. In this lies a great danger and a great opportunity. Suppose there are 100 people with a 0.5 reproduction rate, and 1 super-spreader with a 100 replication rate. The average reproduction rate is 1.5. Clearly, locking everyone down is wildly inefficient. It’s much more important to find the 1 super-spreader and lock him or her down, or change the business or behavior that’s causing the super-spreading.” Makes sense. When talking about distributions, the fallback seems to assume a symmetric Gaussian distribution. What does the empirical distribution look like in terms of its moments?
Michael
Mar 30 2020 at 7:34pm
Interesting! I feel like that is something like what step 2 will look like. (By “step 2”, I mean the way that we reopen after social distancing and lockdowns).
Comments are closed.