Does Dispersion Help?
A few days ago, I was reading one of my listservs, and someone had shared a list of alleged coronavirus-related recommendations by a biologist.* One of the recommendations was "disperse population." Of course, much has been written about the alleged role of population density in spreading coronavirus: some say that New York's density is responsible for its high death and infection rates, and others point out that the most dense parts of New York are not in fact the most heavily infected, and that the very dense cities of Eastern Asia have very low coronavirus infection and death rates. (For a sampling of relevant literature read here.)
But the centralization of a nation's population in one or two urban areas is a very different issue. It could be argued that because epidemics spread from a city to its suburbs more easily than to different regions, a nation which (like the United States) is not dominated by one large metropolitan area is safer, whether that metro area is compact or sprawling. Certainly, evidence from North America does not support this view. The Greater Toronto Area has over 20 percent of Canada's population, while metro New York has about 7 percent of U.S. population. (All data on metro area sizes comes from here.) But Canada's fatality rate is only 20 per 1 million, less than 1/3 that of the United States. But because Canada and the Untied States are a sample of two, it may not make sense to generalize from these two cases.
So I looked at Western Europe: a group of affluent nations with no recent history of major epidemics (unlike East Asia). There are five Western European nations with over 40 million people: the United Kingdom (UK), France, Spain, Germany and Italy. The most centralized of these nations is the UK, where 22 percent of people live in metro London. The least centralized are Germany (where about 6 people of people live in metro Berlin) and Italy (where about 10 percent live in Milan). France and Spain are in between—16 people of French live in metro Paris, and 13 percent of Spanish in Madrid. If centralization was an extremely important variable, the UK by far would be the most dangerous place, and Germany and Italy the least.
In fact, there seems to be no correlation between coronavirus deaths and centralized population. The UK has the second lowest number of deaths among this group (167 per million) but less-centralized Germany has the lowest (36 per million). The two worst-off countries, Spain and Italy (both over 300 deaths per million), are in the middle.
Of course, five countries is a very limited sample size. What about smaller Western European countries? I found half a dozen nations with significant metro areas: Norway (where Oslo has about 35 percent of the nation's populace), Denmark (Copenhagen, 28 pct), Ireland (Dublin, 26 pct), Austria (Vienna, 25 pct), Portugal (Lisbon, 23 pct), Sweden (Stockholm, 21 pct). Although all of these countries are more centralized than the larger countries of Western Europe, all of them have relatively low death rates. The worst of them, Sweden, has 91 deaths per million—still far below all but one of the larger Western European nations mentioned above. Within this group of smaller affluent nations, the least centralized nation (Sweden) has the most deaths, perhaps because they have tried to avoid lockdowns. The most centralized nation, Norway, has only 25 deaths per million, fewer than any of the other nations I looked at. Ireland (74 deaths per million), Portugal (52 deaths per million), Denmark (49), and Austria (41) are all somewhere in between.
So it seems that there is no correlation between the dominance of a nation's largest city and coronavirus deaths. All of this is subject, of course, to many qualifications. First, this is still a pretty small sample of nations, and their public health policies differ in a variety of ways. Second, these figures may change as the epidemic matures; some nations may be more successful than others over the next few months.
*I say "alleged" because I have no idea of the alleged author of these ideas 1) really believes them or 2) is really an expert on any issue related to coronavirus.