The British physicist Lord Kelvin liked numerical measures: “… when you cannot express it in numbers, your knowledge is poor and unsatisfactory …” To which one modern statistician replied: “and even when you express it in numbers, your knowledge is often poor and unsatisfactory “.
This is the case for many indicators of COVID, including “excess mortality,” a demographic that some hope will eventually resolve many of the current controversies about the virus’s real health implications. This is a very useful tool, but you need to understand its limitations.
Excess mortality is “more deaths from all causes during a crisis than we would expect to see under ‘normal’ conditions.” It is a measure, albeit always ex post facto, of the total deaths, for example, as a result of war, epidemic, famine or other large-scale disaster. It is a measure by which some of the current controversy over COVID may eventually become clear – for example, whether the same criteria are used across states or countries to report infections and deaths from the virus.
Understand, however, very simply: inferences about excessive mortality are examples of the “posterior error” that philosophy and economics professors warn students about. The fallacy is to assume that because one happened earlier than the other, the former caused the latter: Lotterman went to work at the Federal Reserve Bank of Minneapolis, and our country had eight years of prosperity and high employment. He moved on, and in less than two years we were in a recession. Cause and investigation? No. Funny? Yes.
Thus, during COVID, the number of deaths all over the world has skyrocketed. Many of them are directly related to the infection of people with this virus. However, there are other factors, possibly related, or not. The number of cases of domestic violence and murder has increased. The number of drug overdose deaths has increased. And some people have died of cancer or heart problems – perhaps because crowded COVID offices, real or perceived, have kept people from seeking or getting help for chest pain or possibly cancerous tumors; or maybe these people would have died anyway. However, over the same period, the number of deaths associated with travel to work and vacation decreased.
Trained researchers have ways to play it all off, but the layman should be wary of comparing global deaths in 2019 to 2020 and conclude that the whole difference was due to COVID.
More generally, examining the social and economic consequences of catastrophic events, such as pandemics, wars, severe floods and famines, often means disaggregating deaths by sex and age.
The Population Pyramid is a useful graphical tool that separately shows the number of males and females in age cohorts, usually 0-5, 5-10, and so on up to 95 and over. A pyramid consists of two histograms at the end with touching bases and bars going in opposite directions for the two genders.
The graph is a snapshot in time that usually looks like an evergreen tree – male and female populations on both sides. A wealthy county with a low birth rate and low death rate would have cool sides. A poor country with high fertility and still high mortality would be like a broad pine, very wide at the bottom but tapering rapidly. The First World War caused a large split in the male section of the adolescent and mid-twenties cohorts, with little or no effect on women. As the time snapshot progresses through the years, the proportion of male deaths increases with the age of this cohort.
The “natural flaw” of the United States during the Great Depression at first looked as if the tree had been placed on a narrow stand. As these individuals age, subsequent graphs show that this fossa on both the male and female sides moves upward. The 18-year baby boom, which began in 1946, manifested itself as the opposite, significantly expanding the base, which increases in later photographs.
The once-broad baby boom base is now close to the top, but as the population grew over the next five decades, the base remained fairly broad, even as the number of children born declined. Now in terms of “natural growth” we are approaching the level of reproduction. Except for immigration and higher fertility of immigrant women and their daughters, the US population will decline.
Epidemiologists and demographers are now preparing to study excess mortality from COVID, just as they looked back at the Spanish flu epidemic of 1918-20. There will be more clarity on many COVID issues, including whether state or country governments have taken steps to decrease or increase the number of officially tabulated infections and deaths, possibly for political reasons.
This works in countries like ours, with reliable “vital statistics” – numbers of births, deaths and diseases. But it also helps if there are no simultaneous events that muddy the waters. Consider the world a century ago. North America and Western Europe kept good records. But further east, political and social chaos reigned, as border shifting, ethnic cleansing, the dismemberment of the Austro-Hungarian Empire, the Russian revolution and civil war raged. Tracking who died, when, and where was a low priority. Hundreds of thousands of targeted killings have occurred, and millions have died from malnutrition or other diseases such as typhus among refugees. All of this happened at the same time as a deadly new flu.
It is difficult, at best, for many countries to precisely divide the huge excess mortality of those years into its component parts. Moreover, on the way to Central Asia from Turkey’s eastern border to China, vital statistics often did not exist at all. Pandemic historians write things like “… and 10 to 25 million deaths in Central Asia.” Hopefully we will do better with COVID, but there will be problems.
There are many more examples of how historical events such as pandemics or wars have dramatically affected economic and social structures over many decades. They stay the next day.
St. Paul economist and writer Edward Lotterman can be reached at [email protected]
