In the coming months and years, there’s going to be a lot of debate about how countries responded to COVID-19 – and how they should have responded. What impact did lockdowns have? Were they more or less important than travel restrictions? Did citizens experience lockdown fatigue? Should vaccines have been approved more quickly? Was enough weight given to individual rights? Did some populations benefit from pre-existing immunity?
I don’t intend to provide an answer to any of these questions here. What I do intend to provide is a brief descriptive analysis of rich countries’ responses to the COVID-19 pandemic. For this purpose, I downloaded two datasets: Our World in Data’s COVID-19 database, and the Oxford Blavatnik School’s COVID-19 Government Response Tracker.
Although data are available for over 150 countries, I restricted the sample to those with a GDP per capita of more than $25,000 (using the measure given in Our World in Data’s database). It is of interest to focus on countries with relatively high levels of income, since one would assume that they have the greatest capacity to deal with novel viruses like SARS-CoV-2. (Countries with populations of less than 100,000 were also excluded.) I obtained cumulative values for each variable of interest up to 15 December because many countries had missing data on government response measures after this date.
Another reason for imposing an income cut-off is to exclude countries where reported numbers of COVID-19 deaths may substantially understate actual numbers. At the present time, we have to assume that large numbers of countries are not accurately reporting their COVID-19 outcomes. This is not necessarily because of corruption; it could simply be due to lack of testing infrastructure.
Here are some examples. A recent study found that more than 16,000 people may have died of COVID-19 in Khartoum (the capital of Sudan), whereas the official figure is 477. Indonesia has reported only 83 COVID-19 deaths per million people (as of 3 January), yet – according to the Financial Times – Jakarta has seen 71% excess mortality this year, which is more than some cities in the “hardest hit” countries. South Africa (one of the most economically advanced countries in Africa) has reported an order of magnitude more COVID-19 deaths than almost every other country in Sub-Saharan Africa. This may be partly due to South Africa’s higher average age, although it seems unlikely that this can explain the entire disparity.
Looking at countries with GDP per capita > $25,000, there are only 15 that reported less than 100 COVID-19 deaths per million people (as of 15 December 2020). These countries could be said to have successfully “contained” the virus. They are (in ascending order of per-capita deaths): Taiwan; Singapore; New Zealand; Brunei; South Korea; Malaysia; Japan; Australia; UAE; Norway; Iceland; Qatar; Finland; Trinidad and Tobago; Cyprus. Thirteen of the countries fall easily into three categories: East Asian countries (Taiwan, Singapore, South Korea and Japan); ultra-wealthy countries with authoritarian governments (Brunei, UAE and Qatar); and geographically peripheral Western countries (New Zealand, Australia; Norway, Iceland, Finland and Cyprus).
The two others are Malaysia and Trinidad and Tobago. These are both at the lower end of the GDP per capita scale for the sample, which may be indicative of underreporting. In addition, Trinidad and Tobago is an island with a population of only 1.4 million, so will have faced fewer challenges in containing the virus than most other countries. (Note that Taiwan has a missing value for GDP per capita in Our World in Data’s database, but it was included because its GDP per capita is known to be substantially higher than $25,000.)
Now let us consider which policies the fifteen most successful countries implemented. As to lockdowns, six out of fifteen had zero days of mandatory stay-at-home orders during the period in question: Taiwan; Brunei; Norway; Iceland; Qatar; Finland. Australia had the most days at 162; all the others had less than 120. The fact that these numbers are quite low is presumably due to the fact that, as mentioned, most or all of the fifteen successfully contained the virus. The association between COVID-19 deaths per million people and days of mandatory stay-at-home orders was positive but only borderline significant (r = .32, p = 0.03, n = 46). I assume this reflects chance and/or reverse causality.
Only two of the fifteen had zero days of mandatory workplace closures (for some sectors) during the period in question: Taiwan and Brunei. The others had between 22 days (in the case of Japan) and 274 days (in the case of Trinidad and Tobago). Somewhat surprisingly, the association between COVID-19 deaths per million people and days of mandatory workplace closures was positive and significant (r = .40, p = 0.006, n = 47). Once again, however, I assume this mostly reflects reverse causality: i.e., countries that managed to contain the virus were able to keep workplaces open.
As to travel restrictions, four out of fifteen had zero days of total border closures during the period in question: Singapore; South Korea; Japan; Iceland. The others had between 59 days (in the case of Finland) and 271 days (in the case of New Zealand). Note that the UK, France, Italy, Sweden and the US all had zero days of total border closures. Interestingly, there was a moderate negative association between COVID-19 deaths per million and days of total border closures (r = –.37, p = 0.01, n = 47). The scatterplot for this relationship is shown below. It indicates that no country with many days of total border closures had a high COVID-19 death rate.
Among rich countries, those that managed to contain the virus up to 15 December were – with two exceptions – either East Asian countries, ultra-rich countries with authoritarian governments, or geographically peripheral Western countries. These “successful” countries were not characterised by longer-than-average lockdowns (most likely because they managed to keep case numbers low, and so had less need for lockdowns). Several of them did, however, implement total border closures for prolonged periods of 2020 (notably Brunei, Australia, New Zealand, and Trinidad and Tobago).
It makes sense that geographically peripheral Western countries were successful in containing the virus: their first cases probably arrived later, and they have had fewer international travellers throughout the pandemic. One can also see why ultra-rich countries with authoritarian governments had less trouble containing the virus, especially given that the three in question – Brunei, UAE and Qatar – are again somewhat geographically peripheral. (I am assuming, of course, that they have been reporting their COVID-19 outcomes accurately.)
The “puzzle”, then, is why East Asian countries were so successful. It may come down to some combination of: greater state capacity; higher levels of compliance; less individualistic behaviour; more pre-existing immunity; and lower levels of vitamin D deficiency. Japan, incidentally, may be the most puzzling of all. It is a very large, highly connected country. Yet it had zero days of mandatory stay-at-home orders, only 22 days of mandatory workplace closures (for some sectors), and zero days of total border closures.
Please note that this is only meant to be a very rough and preliminary analysis. More sophisticated methods will be needed to determine the precise impact of lockdowns, travel restrictions and other policies. In addition, this analysis does not take into account the effects of the “new variants”, which could change the picture substantially going forward.
Image: Frederic Edwin Church, Coast Scene, 1852
Thanks for reading. If you found this newsletter useful, please share it with your friends. And please consider subscribing if you haven’t done so already.