Skip to Main Content

As the Covid-19 pandemic took hold, life expectancy in the United States dropped one full year during the first half of 2020, according to a new Centers of Disease Control and Prevention report, with even greater declines seen among Black and Hispanic people.

Preliminary data from the agency’s National Center for Health Statistics released Thursday mark a reversal of incremental gains over the past few years. Life expectancy at birth for the total U.S. population declined from 78.8 years in 2019 to 77.8 years for January through June 2020. During that same time period, life expectancy for non-Hispanic Black people decreased by 2.7 years (74.7 to 72); for Hispanic individuals, 1.9 years (81.8 to 79.9); and for non-Hispanic white people, 0.8 years (78.8 to 78).


The Covid-19 pandemic is likely responsible for “the majority of the decline,” said Elizabeth Arias, a health scientist at the Mortality Statistics Branch of the NCHS and the lead author of the report. 

Arias said it’s also possible that some of the decline can be indirectly attributed to the pandemic, such as the rise in drug overdose deaths. José Manuel Aburto, a demographer at the Leverhulme Centre for Demographic Science at the University of Oxford who was not involved in the analysis, said that many other factors related to the pandemic could influence deaths, such as general delays in receiving medical care and effects on other hospital patients, such cardiac care patients, at facilities overwhelmed by Covid-19.



The new life expectancy projection is the lowest it’s been since 2006 — and it had been on the rise in both 2018 and 2019

“Life expectancy has mostly been increasing from year to year,” said Arias. Even before 2018, Arias said life expectancy estimates tended to be almost the same or decrease very slightly, for instance by a few tenths of a year.

At the same time, Aburto said, life expectancy in the U.S. has not significantly increased in the last decade, which is already a deviation from much of the world and particularly unusual for a high-income country like the U.S. A decline of a full year, Aburto said, is even more unheard of, which he said points to the “unprecedented burden on life expectancy” the pandemic has taken.

Among males, life expectancy at birth fell 1.2 years, from 76.3 years in 2019 to 75.1 during the first half of 2020. For females, it dropped 0.9 years, from 81.4 years to 80.5. The gap in life expectancy between sexes increased from 5.1 years in 2019 to 5.4 in 2020.

Life expectancy -- Figure_1.001
National Center for Health Statistics

Looking further at racial disparities in the data, the gap in life expectancy between non-Hispanic white and Black people widened from 4.1 years in 2019 to six years in the first half of 2020 — the largest gap since 1998.

“The disparities between those populations have been declining consistently, since we began estimating life expectancy, which goes back to 1900,” she said. 

The gap between Hispanic and white non-Hispanic individuals narrowed, however, from three years in 2019 to 1.9 in 2020.

Life expectancy -- Figure_3.001
National Center for Health Statistics

Non-Hispanic Black males experienced the highest decline in life expectancy — three years — of any race and sex subgroup in the new analysis, followed by Hispanic males (2.4 years), non-Hispanic Black females (2.3 years), and Hispanic females (1.1 years).

While Aburto said he was not surprised by the racial disparities, he did not expect them to be so stark. “What is baffling to me is the magnitude of these decreases,” he said.

Aburto emphasized that these disparities have been driven by systemic inequities in access to health care and a higher likelihood of being exposed to environmental risks. These effects were only exacerbated by the pandemic, as people of color are more likely to have to work in jobs that expose them to sick people and less likely to be able to work from home.

The data come with several caveats: The provisional count means that the report doesn’t capture the full impact of the Covid-19 pandemic in 2020. Arias said data from the entire year will be released in May or June, and finalized versions of 2020 data on mortality and life expectancy, which the NCHS publishes annually, will be released by the end of this year or early next year.

And because the early part of the Covid-19 pandemic largely affected urban areas, where people of color are more likely to live, overall life expectancy could be an underestimate. Arias said data from the full year will likely better reflect the impact of the pandemic on life expectancy, since the pandemic eventually spread throughout both urban and rural areas. 

Aburto said he thinks it’s likely life expectancy will decrease for the rest of 2020. “My intuition, as a researcher usually analyzing life expectancy trends, is that life expectancy will drop even more,” he said. He pointed to the country-wide increase in Covid-19 deaths in the second half of the year, and also numbers he’s seen from the Human Mortality Database, which has more recent data available.

Like Aburto, Arias said the differences in life expectancy among racial groups were particularly striking, as the Covid-19 pandemic took a disproportionate toll on communities of color.

“There’s a big difference there,” Arias said. “That was one of things that stuck out the most.” 

This story has been updated with additional comment.

  • Looking up the Wikipedia definition of “Life Expectancy” there is a retrospective use of the term – using records, we can calculate the life expectancy of a baby born in 1900 – and a prospective, or predictive use of the term – based on current mortality rates, we can guess how long a newborn is likely to live, a 20 year old, 30 year old, 40 year old – etc.
    Using this predictive term, it is still impossible the Wuhan CoV could lower life expectancy by as much as a year, without assuming future death rates, from CoV related causes, which are now known to be wrong.
    According to my county stats, for example, the death rate, even among people over 80, has been about 1 % – the case fatality rate is 35% – but people who are over 80 go to a lot of trouble to avoid catching it apparently -so, even in that age group, the virus did not reduce life expectancy by a year starting in 2020 = that is, if you could see the future of the over 80 group, and see a 1% mortality rate, ie, accurately predict the future – it would not be as bad as they thought in early 2020.
    Their modeling was erroneous – no shame in that, back at the beginning of 2020, but publicizing it now that they should know better is wrong.
    So far, the average reduction in lifespan in the US is about a week.

  • As to some reader’s comments below : the number / percentage of Covid deaths is highest in the higher age groups. So as this pandemic takes out a sizeable chunk of US elder (who therefore get vaccination priority), it logically brings down the life expectancy. If Covid deaths were equally spread among all age groups, there would be no change in life expectancy.

    • I do not understand your Comment, the last part ” If Covid deaths were equally spread among all age groups, there would be no change in life expectancy”.

      I may be a off base a bit, but the average reduction in life, so far, is about a week – but, can you steer me to a definition of “Life Expectancy” ?

  • Rebecca, what effect does it have on the life expectancy result, that the CDC is using abridged life tables that cut off at age 85? Doesn’t lopping off the upper end of the age categories automatically exert a downward pressure on the weighted mean? Thanks, Tracie

  • I find this data questionable at best, and sensational at worst. How the covid pandemic be responsible for a decline in life expectancy when it has a 0.5-1% mortality rate? And in the US, somewhere between 42-45% of deaths were in nursing homes–meaning death among people who were old and sick to begin with–and about 95% of ICU admissions were those with comorbidities which probably would have killed them sooner if not later. And did anyone not bother to actually look at the cause of deaths in 2020? How many people died of cancer, heart disease, complications of diabetes, etc–that number dwarfs those who died of covid.

    Even more reprehensible is that STAT has not interviewed anyone who has counteracted these bizarre claims, or disputed this hypothesis or these numbers–I’m not a statistician but I can do so quite easily. Doesn’t make any sense. Is the number of opioid related deaths higher than it was last year or the year before? What about delays in getting medical care–has that really impacted the mortality rate. Some studies have said no, that delays in care did not cause most people any harm.

    Overall, this is junk science, and embarrassing that STAT would publish this without more input from other experts who can point to the giant holes in this theory. STAT has been publishing more of these pseudo reports lately, I’ve noticed, and it may be time to cancel my subscription. If this trend is going to continue.

    • Roxanne – it seems to me you are right about all you have said. But I want to say, I think Stat should not be blamed for not looking for people to contradict a CDC study. Or, I think Stat should publish results without trying to doublecheck them – that is not exactly Stat’s job when this is a CDC people, if it was some bogus think tank then I think Stat would be remiss to not factcheck.

      I have suspicions this is the CDC trying to scare people, because CDC itself is scared by the new variants. Now they tell us, “wear two mask, that is safer” – why now? Because now, the UK Kent Variant is set for a big outbreak here, and it is about 3 times worse than the previous Wuhan type.

      So, my guess is, this figure on life expectancy is meant primarily to scare people before the variants hit. My belief is, the variants will hit very hard. CDC does not want that – they also do not want to be blamed for the variants, which may eventually happen, in my view they did not do nearly enough to keep them out.

    • You may have been referring to me -and I admit I did not even try to read it – but, looking at my hypotheticals- how can 1 in 500 people dying, about 85% of them over 70, about 40% over 80, possibly create a 1.3% reduction in lifespan?
      I do understand there are a lot of drug overdoses and those are much younger people, but they more or less only counterbalance the old folks- meaning, to get to over 1% reduction in average lifespan, you still need about 1% of people dying, and we have only about 1/5th that number. Or add in the drug overdoses and get up it to 3/10ths- if that even makes sense, I mean, IF someone was using opiates before the pandemic, even if young i doubt their life expectancy is anywhere near average.
      Or, if you do not want to educate me on why I am wrong, can you at least tell me the most tenuous assumptions they made? It is just way to far off to be right.

  • I do not pretend to understand this claim of a lost year of life expectancy, so I can not refute it, but I have to say it is very highly counter-intuitive.
    Right now, let us say for the sake of argument, we assume some of the estimates of further death tolls are more or less correct, and we get to 660,000 dead by the time vaccination is nearly complete -= and we stay on top of the virus and it’s mutations to keep further annual deaths very low.
    So, 660,000 dead is 1 in 500 people in the US – most of them are over 70 – but forget that – let’s say all 500,000 were newborns – and they ALL died on one day – to maximize the lost years of life, which I think is the same as life expectancy – as much as possible.
    so, instead of living an average of roughly 80 years, all the babies born that day have their average life reduced by 1/500th – 80 years is roughly 1,000 months – so, 1/500th of 1,000 = 2 months, not a year.
    In reality of course, the people who have died are more like, on average, age 65 or so. so the epidemic has caused nowhere near a 2 month loss in life expectancy, it’s more like a week I think.

    Or, again, maybe I do not understand this at all, but if I am wrong, can someone explain why?

    • Reading the article more carefully, I see it does not explain the bases for changing the estimate of life expectancy an entire year, or for attributing such a big change to an epidemic which was killing far less than 1% of the population, when life expectancy is less than 100 years. Going back to my original example, if you end 1 in 100 lives at birth, it appears you reduce life expectancy by 1% – which is not even a year, so you do not get a full year loss in life expectancy, and of course, it did not kill 1%, more like .2%.

      I have to think this study is complete bunk, absolute nonsense, at this point. Or, I should say, attributing such a big change to the epidemic is bunk. The only way the epidemic could have done that is if a very large number of people did not get life saving medical treatment – which no doubt did happen to tens of thousands, but not enough to lose a whole year.
      Since one of study author spoke of drug overdose as a further possible loss that was not in their calculations – and in the beginning of 2020, the riots and big jump in crime – including murders – had not happened -so, any estimate on that would have to be a guess – – and again, they attribute the big change to Covid only – I am not buying this claim at all. We did not have good US based data on death rates by age in early 2020, and we were told there was up to a 5% death rate in China, AND we were told it would be several years to a vaccine- so, I am just speculating – maybe they put assumptions into their models, best guesses at the time – which were 100 times too high? Not that hard to do – assume 5% of patients die when it is more like .2% – assume 5 years to vaccine, rather than 15 months and now you are off by a factor of 100.

  • In a few years time, once this covid19 undergoes further mutations and recombination to create more deadly and contagious variants, life expectancy will further fall.

    The world government should do what can do to vaccinate as many people as possible to prevent dangerous mutations and recombinations.

Comments are closed.