Trendspotting: In First Half of 2020, US Life Expectancy Fell by One Year

NEWSWIRE: 2/22/21 

• Life expectancy in the United States fell by a full year in the first half of 2020, to 77.8 years. This was the largest drop since World War II and highlights stark racial and ethnic disparities: Among blacks, life expectancy fell by 2.7 years. (The New York Times)
• NH: Prompted by the magnitude of the pandemic's death toll, the CDC issued an early and unusual "half year" mortality report for the first six months of 2020. Finding: Life expectancy at birth declined by a full year, from 78.8 in 2019 to 77.8 years--the shortest lifespan since 2006.
• The NYT quotes many "shocked" researchers. IMO, the results should not have surprised anyone. After all, the CDC keeps a running total of "excess deaths" in the United States and the total for CY 2020 is coming in at just over half a million (512K). That is 25% larger than total U.S. military deaths in World War II. Who would not expect such a toll to have a major impact on national mortality?

• Broken down by race and ethnicity, the report notes that while life expectancy for whites fell by 0.8 years, it fell by 1.9 years for Hispanics and by 2.7 years for African-Americans.

• As I have noted in previous NewsWires and Covid Calls, there are many compelling reasons for this disparity. Nonwhite Americans are more likely to live in urban areas (with Covid-19, density kills) and are less likely to live in households or work in jobs where social distancing is possible. They have less access to healthcare. They may be physiologically more at risk due to higher rates of comorbidities like diabetes and hypertension and lower levels of serum Vitamin D (25(OH)D). And their death rates are disproportionately higher at younger ages, which has a bigger impact on life expectancy.
• As a result, the Hispanic advantage over whites in life expectancy has narrowed. And the white advantage over blacks has widened. This shift comes after a couple of decades in which nonwhite lifespans have been improving steadily relative to those of whites. (See "Life's Ups and Downs.") Back in the late 1990s, whites had a seven year advantage over blacks in life expectancy at birth--versus over four in 2019. Some of this gain has been attributed to lower death rates among blacks (e.g., from youth violence and hypertension) and some to higher death rates among whites (e.g., opioids and, more broadly, midlife "deaths of despair.")
• Once the Covid-19 death rate has fallen for good, we should expect a rebound in life expectancies--both overall and by race and ethnicity--back to something close to where we were before. Hopefully, that happens by CY2022.
• Before that happens, though, the numbers will get worse before they get better. That's because the rate of Covid-related deaths during the 1H 2020 (21.4K per month) was only about half of what it was in 2H (41.4K per month). So the fall in life expectancy over the entire year is likely to be about 50% higher. Thus, the final drop in life expectancy for 2020 YoY is likely to be around 1.5 years. The drop for blacks will also be larger, though not proportionally larger since whites and Hispanics were relatively harder hit in 2H than in 1H.
• Now let's take a longer perspective on U.S. life expectancy.

• Assuming a final decline of 1.5 years in 2020, we're looking at a retrogression back to where we were in 2002, 18 years ago. That bad news doesn't just reflect the exceptional virulence of Covid-19. It also reflects the stalling of life expectancy gains over the last decade. Even before the pandemic, in fact, U.S. life expectancy had shown no improvement since 2014--for reasons I have often discussed. (See "Death Becomes Us... Mortality Increases.")
• But one thing surely has changed. Before 2020, the biggest drag on life expectancy gains was rising mortality among nonelderly Americans while the elderly continued to live longer. (See "Adults Under Age 65 Driving Decline in U.S. Life Expectancy.") In 2020, that pattern has radically reversed: It's seniors who are taking the hit. To date, per the CDC, 81% of all U.S. Covid-related deaths have been at or over age 65.
• Have we seen a decline like this before? Certainly not since World War II. Before then, it's hard to say since NCHS data retrieval and calculation methods seem to have generated a much higher YoY sigma in the annual number.
• How does it compare with the Spanish Influenza? In 1918, according to the NCHS, life expectancy at birth fell by a massive 11.9 years--from 50.1 to 39.1. And judging by the trend before World War I and after 1920, that may be an underestimate. Absent the war and the pandemic, life expectancy should have been in the mid-50s. So we may be looking at a drop of around 15 years.
• Now let's plug in some numbers. It's usually estimated that the Spanish Flu killed roughly 650K Americans--and let's say two-thirds (or 500K) of them died in 1918 rather than in 1919. That's about the same number as those killed by Covid-19 in 2020. The big difference is that the U.S. population is 3.2X larger today (331M rather than 103M). So, other things being equal, a 1.5 year decline in 2020 should have translated into a 4.8 year decline in 1918.
• So far so good. But of course other things aren't equal. And the biggest of these other things is the age of people dying. Take a look.

• In 1918, the modal age of death was just under age 30 and mortality rates for young children were actually higher than for the elderly. In 2020, by contrast, half of all deaths occurred at or above age 77 or 78. (Why were 30-year-olds so hard hit in 1918? Last May, I offered one theory based on "antigenic imprinting" by cohort: See "Covid-19: May 14 Update.")
• Any precise calculation of the life expectancy shift would require knowing the exact weighting of deaths in every age bracket. As an approximation, I would estimate that the average remaining life expectancy for a pandemic victim in 2020 was about 12 or 13. (Life expectancy in 2020 at age 78 is just over 11 years.) And the equivalent life expectancy for a pandemic victim in 1918 was about 35 to 40 years. (Life expectancy in 1918 at age 30 was 37 years.) That's about 3X more years per victim in 1918. Which would push the 4.8 year decline to a 15 year decline. 
• The number of dead was just one of the horrors of the Spanish Flu. Another was the youth of those who died. H.L Mencken, age 38, was living in Baltimore at the time. Decades later he tried to explain why so few historians wrote about it: They preferred to skip that horrible winter and move directly from the Armistice in November to Woodrow Wilson's signing of the Versailles Treaty in Paris the following June.
• Mencken wrote: "The influenza epidemic..., though it had an enormous mortality in the United States and was, in fact, the worst epidemic since the Middle Ages, is seldom mentioned, and most Americans have apparently forgotten it. This is not surprising. The human mind always tries to expunge the intolerable from memory, just as it tries to conceal it while current."