Life Expectancy is down. So what? – Twin Cities

U.S. life expectancy fell in 2021 according to a report from the National Center for Health Statistics. And death rates rose. Neither of these indicators are good news, but some droop was to be expected given that some 460,000 deaths were associated with COVID-19 in 2021.

Yet what exactly does the data mean? What do these adverse changes say about health in our country going forward? For example, does it mean that babies born in 2021 will, on average, live fewer years than ones born in 2020 or 2019?

The answer to that question is no. People born in those three years probably will average about the same ages at death. And those born in all three are likely to live longer than the “life expectancy” just announced. But if that is all true, why is the data release bad news?

These good news/bad news anomalies reflect common misunderstandings about what “life expectancy” is and is not. Start with the full title of the report: “life expectancy at birth.” That “at birth” is included means the same data can show one’s life expectancy from any other age.

The full definition as used by the World Health Organization is: “Life expectancy at birth indicates the number of years a newborn infant would live if prevailing patterns of mortality at the time of its birth were to stay the same throughout its life.”

What exactly does that mean?

It is helpful to start by noting why several of the common assumptions about what it means are not correct:

• It is not the average age of all the people who died in that year.

• It is not a prediction of how long someone born in 2021 will live.

• This average does not mean that most people do die or will die at that age or close to it.

Understand then that this indicator primarily tells us something about the state of health of a population for the year it’s tabulated. It looks at the present, not the future. But this important snapshot of the present does involve calculations from a hypothetical projection over time.

Take the death rates for each year of life from the first after birth to 110 years out or more, for a specific geographic area, the U.S., a specific year, 2021, and specific genders. Assume that these would remain frozen in time, even though we know they won’t. On that assumption, start with a large number of new births, say 100,000, and calculate how many would die in the first year at 2021 death rates for that age. List that number, subtract it from the total and see how many of these survivors would die in the second year of life. List this number, subtract and see what happens to the second-year survivors in their third year. Keep doing this for another 110 years or so, until everyone is dead. Then take a weighted average of the age at death over the whole 100,000 newborns you started with.

With the death rates for each year, you can start at any age and calculate the expectancy at that age. (Use the same numbers of dead each year, but when computing the weighted average, use the number alive at the beginning of the year instead of 100,000. Or start the process all over again from 100,000. You will get the same answer.)

The expectancy for 2021 was down because age-adjusted death rates were up, from 835.4 deaths per 100,000 population to 879.7. This was due to COVID and to other causes best discussed another time. Understand that this average is across all genders, ages, races and locations. But specific groups vary widely.

Not all groups had increases. Hispanic and Black male death rates actually dropped slightly. But death counts for American Indian and Alaskan Native males and females both rose sharply. The increase was greater for ages 25 through 54 than for those 5-14 or over 85. (There reportedly are differences by state, but final data on that is not yet available.) And there was virtually no difference by gender.

Death rates due to COVID, the third-largest cause of death, rose, but the only comparison was to nine months in 2020. Deaths to heart disease and cancer and to unintentional injuries and stroke, which bracketed COVID, also rose. But deaths due to Alzheimer’s and chronic lower respiratory diseases fell.

For specific numbers behind this news, search for NCHS No. 496 or “Mortality in the United States, 2021.”

For a better understanding of the impact of life expectancy data generally, consider a common political misconception about Social Security. One sometimes hears, “Social Security was a cynical scam. In 1935, when Congress passed the Act setting retirement at 65, life expectancy was under 60, so almost no one collected benefits.”

That is flat wrong because the data are based on expectancy at birth. This was low back then because high infant and young-child mortality skewed the average. But the “life expectancy at reaching adulthood” was much higher. In 1930, there were 6.7 million people aged 65 and over in a population of 123 million. By 1940, when the first checks were mailed, it was 9 million out of 132 million. And in that year, if you had gotten to age 65, men could expect to live another 12.7 years and women 14.7. So yes, there were beneficiaries. Moreover, with higher death rates during working years, survivors’ benefits were extremely important to millions of households, including mine after my father died at age 51 in 1950.

Again, life expectancy is a snapshot of the health of the population at one point in time. It is not a forward-looking prediction for an individual or age group.

When I was born in 1950, my life expectancy was 65.6 years, or to 2016. But by 2004, improvements in nutrition and health meant I could expect to make it to 2030. By 2016, it was 2034. And in 2019, the last year for which I could get a full “period life table,” I gained another eight months. But understand that these expectations are averages for a large number of men and not a prediction for one individual who is overweight, hypertensive and was Agent Orange saturated 50 years ago. So should I replace my rusting 2003 Ford with a new pickup, or one with 100,000 miles?

(To make your own calculations, search the Social Security Administration website. Look especially for “Life Expectancy for Social Security” and “Actuarial Life Tables.”