The United States experienced a staggering surge in excess deaths during the COVID-19 pandemic, peaking at over one million above what would be expected in both 2020 and 2021. But the country’s excess-death problem neither started with the pandemic nor ended with it. A 2025 study comparing U.S. mortality to that of peer nations found that excess deaths attributable to the broader American mortality disadvantage had been rising for decades and, even in 2023, accounted for roughly 23 percent of all deaths in the country. The pandemic amplified a crisis that was already building, and the forces behind it run far deeper than any single virus.
How Excess Deaths Are Counted
Excess deaths represent the gap between the number of people who actually died in a given period and the number who would have been expected to die based on recent trends. Researchers build a baseline using years of historical mortality data, adjusting for population growth, aging, seasonal flu patterns, and the long-running decline in death rates from improved medical care. When observed deaths exceed that baseline, the difference is the “excess.” The concept is simple, but the modeling choices matter enormously. One widely used approach fits historical death counts to a statistical model that accounts for slow-moving secular trends, seasonal cycles, and even day-of-week effects, then projects that model forward to produce the expected count for any new time period.
Different baseline methods can yield meaningfully different estimates. Approaches range from straightforward rate-difference calculations to more complex regression and time-series models, and each handles things like flu-season variability and longer-term mortality trends differently. A study examining four separate methods for estimating influenza-related deaths found that the choice of model influenced estimates substantially.
This matters because excess-death tallies are only as reliable as the baseline they are measured against. If the baseline does not properly account for, say, the aging of the baby-boom generation or the pre-existing rise in drug-overdose mortality, it may over- or understate the pandemic’s true toll. Researchers analyzing U.S. deaths from 2009 to 2022 have used national- and state-level data to carefully project baselines and then measure the gap during the first two years of the pandemic.
The Pandemic’s Direct Toll
From March through July 2020, the U.S. recorded roughly 225,500 excess deaths. About two-thirds of those, around 150,500, were attributed directly to COVID-19 on death certificates. The remaining third were coded to other causes, a pattern that would persist and deepen throughout the pandemic.
That one-third gap raises an immediate question: if the pandemic killed far more people than appeared on COVID-19 death certificates, what else was going on? The answer involves a tangle of delayed medical care, surging substance use, worsening chronic disease, and the ripple effects of an overwhelmed health system. Each of these contributed to a death toll that was broader than a single respiratory virus.
Healthcare Disruption and Its Consequences
When hospitals filled with COVID-19 patients, routine care for everything else suffered. Elective surgeries were postponed, cancer screenings dropped, and people with chronic conditions like diabetes and heart disease missed appointments or avoided emergency rooms out of fear of infection. A study of populations in the southeastern U.S. with high rates of chronic disease and socioeconomic disadvantage found that ongoing disruptions in routine care were likely contributing to a disproportionate share of non-COVID excess deaths. Researchers warned that delayed diagnoses could mean more advanced disease at the time of eventual treatment, increasing both illness and death for years afterward.
This was not only an American problem. An analysis of England’s National Health Service found robust evidence of excess mortality among non-COVID-19 hospital patients during the pandemic, estimating at least one additional excess death among these patients for every 42 COVID-19-related deaths in the general population. The mechanism was straightforward: when hospitals are under extraordinary pressure, the quality of care for everyone else degrades. Staffing thins, wait times grow, and patients who would normally survive their conditions do not.
The Drug Overdose Crisis
The pandemic collided with an already devastating overdose epidemic and made it dramatically worse. From March 2020 through May 2023, drug-overdose deaths in the U.S. exceeded baseline levels by roughly 67,600, a jump of about 24 percent. The sustained elevation was tied to several factors, including unemployment, higher COVID-19 case rates, and the proliferation of illicitly manufactured fentanyl.
Synthetic opioids, primarily fentanyl, had been climbing as a cause of death for years before COVID-19 arrived. Age-adjusted overdose death rates from illicit fentanyl and related synthetics rose from about 1 per 100,000 in 2012 to 7.5 per 100,000 by 2016, overtaking every other drug category. By 2022 that figure had reached 27.4 per 100,000. Fentanyl also increasingly contaminated other drug supplies: its involvement in heroin, stimulant, and prescription-pill deaths grew substantially, and western states that had previously been somewhat insulated saw fentanyl-related deaths climb sharply even before the pandemic hit.
The pandemic accelerated these trends. Social isolation disrupted treatment access and harm-reduction services, while economic stress pushed more people toward substance use. Methadone distribution, which was inversely associated with overdose levels, became harder to maintain under lockdown conditions. The result was a spike in overdose deaths that, while partially receding after 2023, has not returned to pre-pandemic levels.
Alcohol-Related Mortality
Alcohol-related deaths followed a similar pandemic-era trajectory. An estimated 8,700 excess alcohol-induced deaths occurred in 2020, rising to nearly 13,800 in 2021, reflecting rates of about 2.8 and 4.4 per 100,000, respectively. Alcohol-use-disorder-related mortality rates jumped roughly 25 percent in 2020 and 22 percent in 2021 compared with projected trends. The surge hit different demographic groups in distinct ways. Among women, the spike was driven largely by alcoholic liver disease. Among men aged 35 to 44, it split between liver disease and mental and behavioral disorders linked to alcohol.
These deaths do not make headlines the way overdose deaths do, but they added meaningfully to the excess-death count and reflected a broader pattern of behavioral health deterioration during the pandemic years. The age profile of alcohol-related excess deaths skewed older than overdose deaths, with most falling in the 35-to-74 range.
The Reversal in Cardiovascular Progress
For decades, heart-disease death rates in the U.S. had been falling. From 2010 to 2019, the age-adjusted cardiovascular disease mortality rate declined by about 9 percent, from roughly 457 to 416 per 100,000 people. Then the pandemic reversed that progress almost overnight. The rate climbed to about 435 per 100,000 in 2020, then to 452 in 2021, and reached roughly 455 in 2022, a level the country had not seen since around 2010. That three-year period represented an average annual increase of about 3 percent, erasing more than a decade of steady improvement.
Some of this reflects the cardiovascular damage caused by COVID-19 infection itself. But much of it ties back to the healthcare disruption discussed earlier: missed medications, delayed diagnoses, skipped follow-ups for hypertension and high cholesterol, and a general pullback from preventive care. Diabetes, kidney disease, and related metabolic conditions also became a more prominent cause of excess U.S. deaths, rising from the fifth to the third leading contributor when the country’s death rates are compared with those of peer nations. Excess deaths from these conditions did not increase significantly from 1999 to 2010 but then accelerated sharply from 2010 to 2019, adding roughly 3,500 excess deaths per year.
Who Died, and Who Was Spared
The pandemic’s toll was starkly unequal across racial and ethnic lines. During the early pandemic period, adjusted excess all-cause mortality was about 6.8 per 10,000 for Black Americans, 4.3 for Hispanic Americans, 2.7 for Asian Americans, and 1.5 for White Americans. Put another way, age-standardized excess death rates for Black, American Indian/Alaska Native, and Latino men and women were more than double those for White and Asian men and women. The disparity extended beyond COVID-19 itself: non-COVID excess deaths from conditions like diabetes, heart disease, and stroke were two to four times higher among Black, American Indian/Alaska Native, and Latino populations than among White populations.
For Black Americans, the pandemic abruptly reversed years of progress. Among Black men, excess mortality relative to White men had declined from 404 per 100,000 in 1999 to 211 in 2011 before plateauing through 2019. In 2020 it shot back up to 395, a level not seen since 2000. Among Black women, the rate had dropped from 224 in 1999 to 87 in 2015 before stagnating; in 2020 it jumped to 192, wiping out fifteen years of gains in a single year. These disparities reflected differences in occupational exposure, housing density, access to healthcare, and the prevalence of underlying chronic conditions, all of which were themselves shaped by longstanding structural inequities.
The Rural-Urban Divide
Geographic disparities in excess deaths predate the pandemic. An analysis of the five leading causes of death from 2010 to 2017 found a persistent gap between the most rural and most urban counties. Noncore (highly rural) counties had a higher share of potentially excess deaths every single year of the study period, while large metropolitan counties consistently had the lowest share. The gap reflected rural America’s older population, higher rates of chronic disease, fewer hospitals and physicians per capita, and greater distance from specialty care.
COVID-19 did not erase this rural disadvantage. After an initial phase in which the virus hit dense urban areas hardest, later waves swept through rural communities that had fewer ICU beds, less public-health infrastructure, and lower vaccination rates. The rural-urban mortality gap widened rather than narrowed.
How Politics Shaped the Death Toll
One of the more striking findings from pandemic-era research is the relationship between political affiliation and excess mortality. A study of voter-registration records in Florida and Ohio found that after vaccines became widely available to all adults in mid-2021, the excess death rate gap between Republican and Democratic registered voters widened dramatically, reaching 7.7 percentage points in adjusted analyses. The excess death rate among Republican voters was 43 percent higher than among Democratic voters during this period. The gap was largest in counties with lower vaccination rates.
At the state level, excess death rates were higher in states with Republican governors and greater Republican representation in state legislatures between mid-2020 and early 2022. The mechanisms behind these findings include differences in vaccination uptake, mask-wearing behavior, and the stringency of public-health measures, all of which tracked closely with partisan identity during the pandemic. This is a rare case in which political orientation became a measurable predictor of survival.
A Problem That Predates COVID-19
The pandemic spike obscures a longer and arguably more troubling trend: the United States has been losing ground on mortality for years. Even before 2020, researchers Anne Case and Angus Deaton documented rising midlife mortality among White non-Hispanic Americans without college degrees, driven by suicide, drug overdoses, and alcohol-related liver disease. They called these “deaths of despair” and linked them to worsening labor-market opportunities, family instability, and cumulative social disadvantage across successive birth cohorts. All-cause mortality for this group was climbing even as it continued to fall for Black and Hispanic Americans, narrowing what had been a long-standing racial gap in death rates.
The international picture reinforces the scope of the problem. In 2018, before anyone had heard of SARS-CoV-2, American life expectancy lagged the leading countries by more than five years for both men and women, and trailed the average of comparable high-income nations by roughly three and a half years. Applying European age-specific death rates to the U.S. population in 2017 would have prevented an estimated 400,700 deaths that year, accounting for about 13 million years of life lost. The U.S. was already an outlier among wealthy nations, and the pandemic only widened the gap: American men lost 2.2 years of life expectancy in 2020 alone, more than males in any of the 29 countries in one international comparison.
Death Certificates and Data Quality
Excess-death estimates partly exist because death certificates are imperfect. The cause-of-death information on a certificate depends on clinical judgment, available test results, and local reporting practices, all of which vary. A 2021 analysis noted that cause-of-death discrepancies are common in respiratory-illness-related mortality and that a significant share of excess deaths during the pandemic could not be clearly attributed to any specific reported cause, raising questions about both COVID-19 undercounting and miscategorization of other conditions.
A more recent study examining U.S. death certificates from 2003 to 2023 found that systematic differences between reported and reclassified causes of death can profoundly alter disease-burden estimates, with major implications for how resources are allocated and which health problems receive attention. The excess-death metric sidesteps this issue by looking at total deaths regardless of listed cause, which is one reason epidemiologists rely on it so heavily. But it also means that when someone asks “what are all these people dying of,” the answer is murkier than it should be.
Where Things Stand After the Pandemic
COVID-19 death rates in the U.S. fell by about 93 percent from 2021 to 2024. That is the good news. The less reassuring finding is that excess non-COVID death rates declined only about 48 percent over the same period, with sustained excess mortality likely for many natural causes going forward. The pandemic appears to have left a lasting mark on the country’s health trajectory.
A comprehensive 2025 analysis quantified the broader American mortality disadvantage by comparing U.S. deaths with what would be expected if the country matched peer nations. By this measure, excess deaths peaked at just over one million in 2020 and roughly 1.1 million in 2021, then fell to about 820,000 in 2022 and 705,000 in 2023. But that 2023 figure was still higher than the 631,000 excess deaths estimated for 2019, and the overall trend was consistent with the upward slope established from 2014 to 2019. In 2023, these excess deaths accounted for about 23 percent of all American deaths and 46 percent of deaths among people younger than 65.
That last number deserves emphasis. Nearly half of all deaths among working-age Americans in 2023 were excess deaths, deaths that would not have occurred if the country’s mortality profile matched its economic peers. The pandemic did not create this gap. It supercharged a trend that was already heading in the wrong direction, and the country has not yet returned to its prior trajectory, let alone closed the underlying gap.
Violence, Accidents, and Other External Causes
During the first pandemic summer, from March through August 2020, the U.S. recorded roughly 2,000 excess homicide deaths, accounting for about 2.4 percent of non-COVID excess mortality. Motor vehicle crash deaths showed a more complicated pattern: they initially dropped below expected levels in April 2020 when roads emptied during lockdowns, but then rose above expected levels from June through August, a likely result of faster driving on less congested roads. Together, these external causes contributed a smaller share of excess deaths than chronic disease or substance use, but they reinforced the picture of a society experiencing disruption across every dimension of health and safety.
Maternal Mortality and Reproductive Health
A separate line of research has begun examining whether state-level abortion restrictions have affected maternal and pregnancy-associated mortality. An analysis of 14 states with abortion bans found some evidence of a potential 9.2 percent increase in the number of pregnancy-associated deaths above what would be expected, equivalent to roughly 68 excess deaths, though the estimate had wide uncertainty and did not reach conventional statistical significance. The study found no detectable increase in maternal mortality as traditionally defined, but the authors noted substantial measurement challenges in this area. This remains an active and politically charged area of investigation, with data limitations making firm conclusions difficult.
Nursing Homes and Congregate Settings
Nursing homes were among the deadliest settings during the pandemic, and quality ratings did not protect residents the way families might have hoped. A study examining the relationship between nursing home quality and COVID-19 deaths found that, as of late April 2021, five-star-rated facilities had experienced about 8.4 percent more total deaths than one-star facilities. The finding seems counterintuitive, but higher-rated homes tended to be larger, have more residents, and serve populations with more complex medical needs, all of which increased exposure risk during a respiratory pandemic. The result complicates any simple narrative about quality of care and pandemic outcomes in long-term-care settings.
Years of Life Lost
Raw death counts do not capture the full human cost because they weigh a death at age 90 the same as one at age 35. Researchers have tried to address this by calculating years of life lost, which accounts for how much remaining life expectancy each death erased. An early estimate from the first months of the pandemic put COVID-19-related years of life lost in the U.S. at roughly 1.2 million, with the toll concentrated in a handful of hard-hit jurisdictions. On a per-capita basis, the years-of-life-lost burden for women was generally higher than for men across the country, reflecting longer female life expectancy and thus a larger gap between actual and expected age at death.
When the lens is widened beyond the pandemic to the entire American mortality disadvantage, the numbers grow staggering. The 400,700 excess deaths estimated for 2017 alone, using European death rates as the benchmark, translated to about 13 million years of life lost in a single pre-pandemic year. These are not abstract statistics. They represent decades of life that Americans lose, year after year, to conditions and risks that peer countries manage more successfully.