How do excess‑mortality estimates for the U.S. through 2025 compare with the confirmed COVID-19 death tally reported by CDC NVSS?
Executive summary
Excess‑mortality estimates for the United States through the pandemic and into 2023–2025 generally exceed the tally of confirmed COVID‑19 deaths that appear in CDC’s National Vital Statistics System (NVSS), reflecting deaths both directly attributable to SARS‑CoV‑2 and indirect losses driven by overwhelmed health systems, delayed care, and social disruption; however, the size of that gap varies by methodology and grows more uncertain as the pandemic period lengthens [1] [2]. NVSS provides a running, provisional count of deaths where COVID‑19 is listed on the death certificate, but those provisional counts are subject to reporting lags and incompleteness, especially for recent weeks and years, so direct comparisons must account for known data limitations [3] [4] [5].
1. How the two measures differ conceptually and in source data
NVSS’s confirmed COVID‑19 tally counts deaths coded to ICD‑10 U07.1 on death certificates and is produced from death records filed by states and local jurisdictions; these provisional counts are updated continually but can lag by weeks to months and are acknowledged by NCHS to be incomplete for recent periods [3] [4]. By contrast, “excess deaths” are estimated as the difference between observed all‑cause mortality and an expected baseline that models what deaths would have been absent the pandemic; excess estimates therefore capture direct COVID‑19 deaths that may not have been recognized as such, plus deaths from other causes influenced by the pandemic [1] [6].
2. What empirical analyses have found through 2023–2025
Multi‑method research and cross‑country comparisons find that U.S. excess deaths spiked during 2020–2021 and remained elevated afterward, producing an outsized mortality burden relative to peer high‑income countries; one large analysis reported excess US deaths peaking around 1.0 million in 2020 and 1.1 million in 2021 and still hundreds of thousands in 2022–2023, contributing to a long‑running US mortality disadvantage across decades [7] [8]. Academic excess‑mortality studies and WHO‑style global estimates consistently conclude that confirmed COVID‑19 death tallies undercount the pandemic’s full mortality impact to varying degrees, meaning excess totals often exceed NVSS COVID counts, particularly in early surge periods and in settings with limited testing or certification [2] [9].
3. Why the gap exists and why it changes over time
The gap between excess deaths and NVSS confirmed COVID‑19 deaths reflects several mechanisms: misclassification on death certificates, testing shortages early in the pandemic, and indirect mortality from strained health services and behavioral changes [9] [2]. Methodological choices — how expected baselines are constructed, whether pandemic weeks are excluded from baseline estimation, and how uncertainty bounds are handled — materially affect excess estimates; the CDC itself revised its excess‑death methodology in March 2023 to avoid inflating expected deaths after excluding many pandemic weeks, acknowledging increased uncertainty the longer the pandemic persists [1].
4. Data quality, debate, and competing interpretations
NVSS emphasizes that provisional counts are valuable but incomplete and that cause‑of‑death certifiers exercise judgement that can vary by jurisdiction, producing under‑ or over‑reporting relative to model‑based excess estimates [4] [5]. Some scholars and commentators caution that global or out‑of‑sample extrapolations can overestimate deaths in settings with limited data and argue for parsimonious modeling; rebuttals note systematic under‑reporting is also plausible and that multiple methods point to the same qualitative conclusion: a higher toll than confirmed tallies alone imply [10] [2] [11].
5. Bottom line for comparisons through 2025
Available peer‑reviewed and public‑health sources indicate that excess‑mortality estimates through the pandemic years commonly exceed NVSS‑reported confirmed COVID‑19 deaths, capturing both unrecognized COVID deaths and pandemic‑related indirect mortality, but the precise magnitude of the excess depends on methodological choices and is increasingly uncertain as analyses project further into 2024–2025 [1] [7] [2]. NVSS remains the authoritative source for certificate‑based COVID counts, while excess‑mortality analyses give a broader lens on total pandemic impact; responsible comparison requires acknowledging reporting lags, the CDC’s methodological updates, and ongoing scholarly debate about model assumptions [3] [1] [11].