Minnesota’s Border Battles: COVID-19 edition

Last year, we released a report titled ‘Minnesota’s Border Battles‘ in which we compared the economic outcomes in Minnesota counties bordering other states with the outcomes in the border counties in the neighboring states. The idea was that this would allow me to analyze areas that have more in common so that any observed differences could be attributed with more certainty to which state they were in.

Would the same approach be useful for looking at COVID-19 outcomes? Minnesota and its neighbors have followed very different paths during the pandemic. According to Wallethub‘s stringency index, these range from a current ranking of 40 for Minnesota and 1 for Iowa, meaning that our neighbor to the south is the least restricted state in the union.

We could look at the outcomes for Minnesota and its neighbors in terms of cases, hospitalizations, or deaths and judge these policies accordingly, but there could be other factors which drive some of variation in the observed outcomes. Can we compare COVID-19 outcomes in Minnesota and North Dakota without taking into account the fact that Minnesota has a population density of 68.1 residents per square mile while the figure for North Dakota is 10.5? By contrast, the population density of Pembina County in North Dakota was 6.6 residents per square mile in 2010 and 4.1 in its neighbor in Minnesota, Kittson County. The essential point is that neighboring counties are more similar than neighboring states.

The headline numbers are summarized in Table 1. So, for example, it shows that, in the Minnesota counties bordering Wisconsin, there were 94.9 cases per 1,000 people and 104.2 cases per 1,000 in the Wisconsin counties bordering Minnesota. Looking below that, to deaths, we see that the Minnesota counties have suffered 1.1 deaths per 1,000 and the Wisconsin counties have suffered 0.9.

Table 1: Border county COVID-19 outcomes per 1,000 (up to April 20)

Source: The New York Times and Center of the American Experiment.

How should we interpret these numbers?

First, it is worth noting that, while Minnesota has recorded a lower number of cases per 1,000 in each instance, it has only recorded fewer deaths than Iowa. In terms of mortality, the border counties of Wisconsin, South Dakota, and North Dakota have all seen better results then their neighboring counties in our state. And that is in spite of a greater number of cases.

Second, the differences between neighbors is much greater when we look at cases than when we look at deaths. This could be interpreted as success for the state’s stringent measures. But if these measures have slowed the spread of cases, this does not seem to have had any positive impact on mortality.

John Phelan is an economist at the Center of the American Experiment.