Why I’m worried about the US

28 March 2020

The US, and in particularly the New York City area (my home town, where part of my family and many friends live), are showing the potential to be the worst-impacted locations in the world.

What do the data say about where the US is now and what’s likely to happen in the near future? And why does the US appear uniquely unable to respond to the challenge? (Spoiler: it’s not mainly about partisan politics.)

(Most charts below, except where stated otherwise, are from the stupendously outstanding Our World in Data. If you’d like to support their important work, you can donate here; I just did.)

Many cases, growing quickly …

The US has now surpassed 100,000 cases (28 March figure; chart below goes through 27 March):

The US is now adding more cases daily than any other country. This is the rolling three-day average as daily cases reports can be lumpy based on lags in testing and reporting:

And, using a chart from this highly recommended New York Times article, we see that the growth rate of total new cases remains particularly high in New York city:

The same article has a novel way of comparing severity in different locations that I like a lot, though it requires some thinking. This chart compares two measures: the growth rate in total confirmed cases (using a trailing 7-day measure to smooth out lumps), and the confirmed cases per thousand people, which is a measure of the degree to which COVID-19 is penetrating into the local population (“attack rate”).

There are a few things to think about when looking at this chart.

Any location–a town, a city, a country, a cruise ship, a nursing home, a family–which sees an outbreak of COVID-19 will show very rapid growth in the beginning. But we should obviously worry much more about a location growing cases from a base of many cases than a base of few cases. New York City alone had more than 26,000 cases as of 27 March, around 5% of the world’s total: a 40% growth rate in NYC is much more concerning than, say, a 100% growth rate in a small town.

What this chart shows is that New York is uniquely bad in combining three factors: a large population (20 million), a high attack rate (>2%), and a continuing high rate of growth.

And another New York Times article–actually a feature helpfully updated daily–shows US cities on the same “doubling time” chart that I’ve republished several times from Our World in Data. (This chart is of deaths, not cumulative cases; note that it’s plotted on a logarithmic scale so a straight line means consistent exponential growth.)

We see here that New York is on a particularly brutal curve, doubling deaths roughly every two days.

… and not slowing down

I’ve recently discussed the importance of doubling times and whether they are increasing or decreasing. Unfortunately, the doubling time in the US has remained stubbornly high for three weeks now. (This is an update of the doubling time chart from a prior post.)

Repeating a point from that earlier post, with updated figures:

If the US were to continue to see the number of confirmed cases double ever 3 days, the number of cases would increase 813-fold in 30 days. A 813-fold increase in the number of confirmed cases would take us from 85,991 cases on 27 March to more than 69 million cases on 25 April!

This is just a statement about how the math works, not a prediction. For many reasons, I don’t think that will happen. To name a few:

  • The growth rate in reported cases in the US is likely greater than the growth rate in actual cases, because the rate of testing is increasing. Over time, the two growth rates should converge.
  • Control measures are increasing and will slow the growth rate.
  • Even if nothing were done, as we get into the millions and tens of millions of cases, some degree of herd immunity kicks in. The effective reproductive number decreases as there are the proportion of infected (and therefore immune) people increases.

What’s the special challenge in the US?

Looking at different responses around the world, it’s become clear that a major pandemic exposes unique strengths and weaknesses of different cultures and different systems of government, at least in terms of their ability to respond to a pandemic.

Simplifying enormously, I think about four factors which seem particularly relevant to a society’s ability to respond to the crisis effectively:

  1. Quality and inclusiveness of health care system.
  2. Transparency.
  3. Degree of central government control.
  4. Individualism versus collectivism.

It goes without saying that better health care systems, which often but not always correlate with wealth, are better able to respond. But the inclusiveness of the health care system is important in a severe epidemic in at least one way: the extent to which all citizens believe and trust that they will be looked after by the system, regardless of their personal resources, may correlate with their willingness to temporarily forgo income.

Transparency has been a clear benefit in the response in many countries, with figures published daily, and extraordinary collaboration among scientists and epidemiologists to find solutions. There’s a strong argument that lack of transparency was at the root of the failure to control the epidemic earlier in China.

Considering government control as well as culture, I’ve been struck by many accounts of how China has exercised extraordinary innovation to battle COVID-19, but also used highly invasive approaches ranging from using big data to monitor and direct individual citizens, to removing infected individuals from their families. This video by a British man living in Wuhan is a particularly interesting overview of the steps the Chinese are taking on the ground today. I don’t know about you, but it’s really hard for me to imagine most Western European countries, let alone the US, tolerating this degree of invasion into the private sphere.

It’s been striking how a number of countries with roots in Confucianism, including South Korea, Singapore, and Taiwan, have been able to execute extreme lockdowns with reportedly high compliance and high effectiveness.

Looking at Europe, it’s similarly striking the degree to which in countries including France, Germany, and Switzerland have been able to implement stringent control measures. Some countries, like France, have a degree of central power that simply doesn’t exist in the US; others, like Switzerland and Germany, may have significant power within regional governments, but enjoy a high degree of trust in the competence of the government and the professionalised bureaucracy.

As relatively communitarian societies with strong memories of the World Wars, many European countries have not only accepted the sacrifice required to implement control measures, but have also managed their response on a national level. For example, in France, our local departement in the South West has no deaths and only a handful of cases. Alsace, on the other hand, has been badly hit. So the country is aggressively sending doctors and resources to Alsace, and the military is moving patients from overwhelmed hospitals in Alsace to other locations.

In comparison, the US is at the extreme of both a culture rooted in individualism and individual rights, and a highly decentralised government where many powers reside with states, cities, and other local governments. Moreover, the US must be at an all-time low in terms of faith in government.

What makes me especially pessimistic about the ability of the US to respond is the combination of extreme individualism, the relative weakness of the federal government, and the low faith in government. Watching Governor Cuomo’s daily briefings has hammed home the degree to which each city and state is choosing its own response, and to which different locales are competing for scarce resources.

A depressing outlook

At the end of the day, not every country, state, city, or region will respond effectively. Many countries in Europe are taking the difficult medicine of lockdown for long enough to get control of the epidemic, and will then be able to gradually reopen the economy while implementing aggressive track/trace/quarantine against any new outbreaks.

Imagine that one country in Europe, Covidia, either can’t or won’t implement the measures to get the epidemic under control, and the virus spreads through a significant proportion of the population. Other countries in Europe will close their borders with that country, and maintain strict quarantine rules for anyone entering the country from another. Covidia is in bad shape, but many countries are able gradually to return to normality over time.

In the US, it seems to me that we have the worst of all possible worlds. Some cities and states are moving to strict lockdown, accounting for around half of the population. But half of the population is not under lockdown.

Until every location with even a few cases implements strict control measures, COVID-19 will continue to spread. And unlike in other regions of the world, if one state or city is badly hit, it’s not obvious that there’s a way for its neighbors to close the borders.

So the US is effectively choosing intense economic pain without the benefit of the ability to come out of lockdown in 6-8 weeks. The result will be one of two scenarios, both bad.

In one, the US will need to continue to implement increasingly strict and broad control measures, extending the period of economic impact far longer than it needed to be–with catastrophic results for employment and the economy as a whole. But at least the virus will come under control over time.

In the second, the US will be politically unwilling, or unable, to do so, and COVID-19 will continue to spread rapidly. The result will be a much greater attack rate into the population; health care systems being overwhelmed around the country; more sickness and death than would otherwise have been the case; and and extended period in which economic uncertainty persists, perhaps with see-sawing between tighter and looser controls.

Do you have a more optimistic story about how things will evolve in the US? Please tell me; I’ll share the best arguments in the next few days.

2 thoughts on “Why I’m worried about the US

  1. Hey, Chris! I hope all is well. Well written post; thanks for writing and sharing.

    At least the data I’ve looked at a national level shows that the rate of infection (doubling rate) appears to be slowing, counter to the argument you presented. Your own data appears to support that; the US was accelerating for a bit dropping from 9 days to 3 days, but recently inching its way back to 4 days. That all said, I still concur that at more local levels we will see growth rates that exceed the national average, and as a consequence COVID-19 will “burn through” these communities decimating them. NYC, as you note, is basically on fire at the moment with doubling every two days.

    That aside, I concur with your view that in the US, we will not see the degree of containment we’ve seen in other countries whether Europe of Asia. As noted, we’ve taken stringent steps in only parts of the US, while other parts are lax or non-existent. What I suspect we will see in the US for the next 12-18 months is NYC played out over and over again. Large population centers will be impacted as containment comes too slowly, and counter-measures thus prove ineffectual. My guess is we’d see 2-3 cities like NYC before we awake to the fact that this will roll through the US, and likely not till we get to 4-5 cities before we have changed course in any material way that looks substantively different than our current response. As you note, this in-between state is the worst of worlds as you effectively push the US into the recession without nearly the results you’d want if the entire US had taken this uniformly more seriously.

    Note, the post you linked to (importance of doubling times, https://wordpress.com/block-editor/post/subjectiveprobability.com/202) ask me to sign into WordPress and start a post.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: