I start this column by admitting that I just don’t know what the right policy recommendation is for state leaders concerning this pandemic. That is a change from the early days of January through April, when we knew much less about the disease. Uncertainty is an input to decisions, and so many months ago, vigorous efforts to contain the disease’s spread were clearly warranted. Every serious benefit-cost analysis came to this conclusion. 

While many epidemiologists still believe we can control the disease, I am less sanguine. This is not because I know more about the disease than they. I do not. Rather, it is because I think the politicization of basic public health measures leaves too many Americans scoffing at masks, social distancing and other steps to contain this global pandemic. Quite simply, the amoral buffoonery that animated the anti-mask crowd makes effective policies untenable. 

This is reminiscent of the early days of World War II. It took more than six months after Pearl Harbor to convince all East Coast mayors to enforce blackouts. The last holdouts came around only after the flotsam of U-boat attacks cluttered their ports. We Americans are stubborn people, for both good and ill. 

Even with vaccine availability, much more vigilance, sacrifice and heartbreak remain before us. Of course, whatever actions our government takes balance the risks of the disease with the cost of government restrictions on businesses. Today, many Americans claim that government ‘shutdowns’ are the cause of our deep economic downturn. Others believe it is the disease itself, not government action that has suppressed our economy to Great Depression levels. A bit of rigorous thinking is in order. 

A number of groups maintain records of government actions during COVID, and at least two have produced measures of restrictions across all 50 states and the District of Columbia. This lends itself well to a statistical test on these data. Here’s what that reveals. 

It is true that more restrictive government shutdowns are correlated with an increase in unemployment rates since the start of the pandemic. One might be tempted to conclude that shutdowns did the economic damage, but that’s not quite sufficient to assert causation. It turns out that the share of a state’s economy in at-risk sectors like tourism are also correlated with increases in the unemployment rate. For example, a place like South Dakota, which has largely ignored COVID, has much lower unemployment than Hawaii, which has taken the most aggressive actions to prevent the spread of the disease. 

Fortunately, it is simple to tease out which of these factors matter most. One need only include both government restrictions and the share of the tourism economy in the same statistical test. It is even more useful to add other factors that might influence the disease and economy. So, including the share of urban population and average annual temperature capture some elements of both. 

This was informative, and it turned out that the statistical test clearly removed the correlation between unemployment rate changes and government actions. What remained was simply how urban the state was, and how much of their economy was in tourism-related sectors.

The conclusion from this analysis is clear. Government restrictions appeared to have no statistically meaningful effect on unemployment rate changes since late last year. Instead, it was the disease effects on restaurants, bars, hotels and similar businesses. But, you don’t have to believe my work.

Harvard’s Opportunity Insight project has a superb website that tracks daily expenditure data by state across the economic sectors most likely impacted by COVID. There are three obvious lessons to be learned from these data. The first is that every state experienced very large shifts in consumer spending prior to any state actions. So, here in Indiana, grocery stores saw a 60 percent spike in sales. Restaurants, accommodations and retail stores all began a rapid and deep descent in sales before schools shut down or Governor Holcomb took any executive actions. 

The second lesson is that this pattern played out everywhere, at the same time, in every state. So, places with heavy restrictions saw the same pattern as those with no real restrictions. Not surprisingly, no matter where they live, many folks prefer to avoid contracting a potentially fatal disease. The third lesson is that spending in all these sectors started to recover long before any government eased their restrictions. In other words, businesses figured out how to safely provide goods or services, so consumers went back to buying. 

There are other studies of the early government actions. The best paper to look at this issue estimated restaurant, bar and non-essential business closures last March explained about 12.5 percent of all unemployment claims. So, even in the very early days, close to 90 percent of unemployment was caused by the disease effects on household spending, not government action. 

In the end, the belief that government restrictions are responsible for the current economic crisis are easily debunked, and many economists have done so. But, we live in a time when many folks proudly avoid wearing masks and believe Mr. Trump lost the election due to fraud. We should not expect the tools of mid-20th-century economic analysis to make much of an impact on rational thinking.