What We Can Learn from Coronavirus Per Capita Death Rates

Army Sergeant Marlenny Medin asks a patient about his medical history at the Javits New York Medical Station, which supports local hospitals during the coronavirus outbreak in New York City, April 8, 2020. (Specialist Nathan Hammack/US Army/Reuters)
The main takeaway from the massive disparities in death rates so far is their regional patterns.

Six months into the global coronavirus pandemic, and more than two months into the widespread lockdowns of American public life, we have lots and lots of data. But what have we learned? The hard data we have today remains flawed, and it can tell us a lot more about the “where” of the pandemic than the “why.” But just knowing the “where” is a start. One of the chief lessons we can take is the difficulty of comparing very different regions of the world. America isn’t South Korea or New Zealand, and never could have been. Another is quite how uniquely bad the situation in the New York metro area has been.

For today, let’s focus on a single metric: deaths per capita. It is the ultimate bottom line: How many of your people have you lost? It is still not a perfect measurement: It relies on accurate reporting (something not all governments can be trusted to do), it still has a time lag, and there remain varying standards in how different locales decide which deaths are “COVID-related.” But unlike infection rates, the numbers are not as sensitive to the availability of testing. And death rates are not only more available than hospitalization rates, they also measure the ultimate success or failure of hospitalization.


I’ll use the data from the COVID-19 Dashboard by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU), which is as good a source for comparable data as any. The numbers are pulled from the Johns Hopkins site as of the morning of May 20, 2020. U.S. data is available on a state-by-state basis, which allows us to compare variations within the United States to national data elsewhere. I’ve color-coded countries by broad geographic regions, which requires lumping together some places — for example, I included Australia and New Zealand with East Asia, despite major differences in their societies, on the basis of geography. Also, 14 countries with populations of a million or more are listed as having zero COVID-19 deaths; presumably, most or all of these are just not reporting data or are refusing to admit they have cases: Vietnam, Mozambique, North Korea, Cambodia, Rwanda, Laos, Papua New Guinea, Eritrea, Central African Republic, Turkmenistan, Mongolia, Namibia, Lesotho, and Timor-Leste. Regional numbers for Italy are via Statista.

The World in Regions

Before analyzing individual countries and states, it is worth considering the data on a cumulative basis, by regions of the world:

The divergence is dramatic — and illustrates the different baselines for judging the responses of different national and local governments. The outbreak in North America and Europe — the most advanced and prosperous parts of the world, all situated in the Northern Hemisphere where the outbreak came at the tail end of winter, overwhelmingly in temperate zones — has been four times the global average. At the opposite end of the scale are East Asia, South Asia, and Africa, all with a fraction of the cases. Government and social responses may have a greater role in East Asia, a region that contains a lot of densely packed, cosmopolitan countries with advanced economies. As I’ve noted previously of New Zealand (which, like Australia, is in the Southern Hemisphere, where the weather was still warm in February), there are a number of potential factors in play, from geographic isolation to low population density. That’s not true of Japan or South Korea.

It is, in any event, hard to explain the scale of variation between East Asia and Europe or South America solely on the basis of different government policies. What is more debatable is how much relates to demographic conditions (age, genetic factors, rates of obesity, and other risk factors) and how much to social practices (acceptance of public mask-wearing is more culturally ingrained by now in many Asian societies.)

The Worst

Here’s the 33 jurisdictions (nations or U.S. states) with the highest per-capita death rates:

As you can see from the above, the top of the global list is dominated by European countries and American states. The United States, with the largest raw number of reported deaths in the world, drops to eleventh when ranked on a per capita basis, behind ten European countries. Two of the top three are tiny microstates, but Italy, Spain, France, and Britain have all been hit worse than the United States. So has Sweden, with its much-vaunted experiment in developing herd immunity by keeping the country open. Ecuador is the lone representative of the Southern Hemisphere on the worst list.

New York also stands out on its own, well ahead of the rest of the world, with its neighbors New Jersey, Connecticut, and Massachusetts all above any nation of more than 100,000 people. In fact, outside of tiny San Marino, New York’s per capita death rate is nearly double that of any nation on earth. The national U.S. death rate drops from 280 per million to 167 per million outside of the Tri-State area. That’s still a high rate on a global basis, a bit higher than Canada’s, but it’s also lower than the regional average for Europe. The Northeast is really in its own category from the rest of the country, and is an enormous driver of the overall national numbers.

There’s a natural temptation to focus narrowly on the red/blue divide, given that the top of the list of states is heavily dominated by deep-blue states, their governments long run at many levels by Democrats. It’s fair enough to assess with hindsight what kinds of governments and reactions led to worse results, but even aside from my distaste for reducing tragic deaths to a red/blue divide while people are still dying, party and ideology are only one piece of what happened. The deep-blue Pacific coast, for the most part, came out better than the Northeast did.

Bad leadership, however, was clearly a major accelerant to why the New York metro area has suffered the worst outbreak on the planet. Infighting between Governor Cuomo and Mayor de Blasio, Cuomo’s catastrophic policy of sending infected patients back to nursing homes (with a death toll the state has been covering up), de Blasio’s long record of statements telling New Yorkers to keep going about their business, and the insistence on keeping the subways open all made the New York area a unique vector for disease even atop its unusual population density and dependence on closely packed mass transit.

Here’s the next-highest list:

American states are still grouped around European countries, with even Germany — one of the European countries that has done relatively well — above Florida, Arizona, California, and Wisconsin. Iran is the worst-hit country outside of Europe and the Americas. That undoubtedly is due in part to the fact that, like Ecuador and Peru, Iran’s topography is more mountainous, elevated, and temperate than that of its neighbors.

Here’s the third list, including the last few jurisdictions above the global average:

The American states below the world average include the southern border (Texas’s death rate is comparable to that of neighboring Mexico), the West Coast (Oregon) and some states in the interior that are (with the exception of Utah) comparatively less connected to global travel and migration.

Now, the countries with the fewest reported deaths per capita:

Somehow, I doubt that many thinkpieces will be written on how we should all try to emulate Ethiopia. Undoubtedly, many of the poorest countries are also underreporting, not on purpose, but because people there are likelier to die at home, and the health-care system is less able to devote resources to classifying deaths. The presence of Thailand and Taiwan, however, is interesting. Thailand was one of the world’s first countries to have COVID-19 cases, due to direct travel from Wuhan. But it also has a tropical climate, with high humidity and temperatures in the 90s in February.

Here’s the complete list of states:

Region by Region

Let’s look more closely region by region. First, Europe:

Even allowing for the questionable reliability of, say, Russian reporting of deaths, the pattern in Europe is quite clear: The countries hit hardest have been the most connected and advanced economies, even though those are the countries with the most advanced health-care systems and the most sophisticated governments. A closer regional look at Italy offers a comparison point for New York:

Even in Italy’s brutalized north, only Lombardy had a higher death toll per capita than New York state, and not by much, despite the fact that a big chunk of New York is more spread-out upstate. The north–south pattern of the virus in Italy parallels that in the United States; the southern part of the country is hotter, wetter, poorer, less crowded, and less elderly.

Now, Central and South America, including the Caribbean:

Venezuela’s very low death toll is probably due to some of the same factors as in the poorest parts of Africa. Brazil, the region’s population colossus, has not suffered nearly as much as Ecuador, but for a country with so much tropical jungle, delta, and seacoast, it is doing quite poorly compared to the rest of the region.

The Middle East and Central Asia:

By now, we should not be surprised to see the wealthiest and most globally connected countries atop the list — even places such as Israel and Saudi Arabia that took dramatic early steps against the pandemic — while ravaged areas and police states such as Syria and Gaza are at the bottom.

East Asia and the Pacific:

This is where the virus started, yet every single country in the region — even the Philippines, hardly known for its efficient government, and Japan, with one of the world’s oldest populations — has suffered fewer than a fifth of the global average of deaths per capita. Even if we assume, reasonably, that Wuhan’s real death toll was understated by a fraction of ten, China would be well below the world average. (Then again, China has just reimposed a lockdown on over 100 million people in the northeast of the country).

South Asia:

The region has only a handful of countries — but almost 1.8 billion people — yet fewer than 5,000 reported COVID-19 deaths. India took early, draconian measures to seal its borders, but even Pakistan, which appeared badly unprepared for a public-health crisis, looks nothing like neighboring Iran.

Finally, Africa:

Sao Tome and Principe’s high death rate may be partly a statistical fluke due to its small size. Algeria’s presence near the top of the list is undoubtedly at least partly a factor of its continuing close ties with neighboring France.

The main takeaway from the massive disparities in death rates so far is their regional patterns. Those patterns may shift as the seasons turn over, but it seems likely from what we know so far that the pandemic will continue to have its greatest effect in the temperate and urbanized regions of the world.


The Latest

Is Trump buckable? &c.

Is Trump buckable? &c.

On internal GOP politics; a Chinese artist and the Olympics; Ukraine and its right to exist; Joan Didion and other Buckley-hired talents; and more.