Studies Show Mixed Evidence on COVID-19 Lockdown Mortality Impact Amid Ongoing Methodological Debates

Peer-reviewed studies show lockdown jurisdictions had substantially lower COVID mortality in neighboring comparisons, though meta-analyses found modest average effects of 0.2-3.2% mortality reduction, with researchers unable to definitively separate mandate effects from voluntary behavior changes or establish net mortality impact including healthcare disruption deaths.
Key takeaways
- Norway recorded 544 COVID-19 deaths through 2020 versus Sweden's 11,005, with first-wave mortality 9.7 times higher in Sweden at 2.9 per 100,000 versus Norway's 0.3 per 100,000.
- In 11 of 13 neighboring U.S. state pairs, states with later or no lockdowns had higher disease prevalence, averaging 480 additional cases per million daily by April 30, 2020.
- A Johns Hopkins meta-analysis found the average European and U.S. lockdown in spring 2020 reduced COVID-19 mortality by 0.2-3.2%, though researchers noted difficulty excluding voluntary behavior changes.
- 1,489,237 women in England had breast cancer screening delayed by 2-7 months between July 2020 and June 2021, resulting in an estimated 148-452 additional breast cancer deaths.
- Researchers across multiple studies documented significant methodological challenges in isolating lockdown effects from confounding factors, with no study establishing comprehensive net mortality accounting.
Mortality Differences Between Strict and Minimal Lockdown Jurisdictions
Peer-reviewed studies document substantial mortality differences between jurisdictions that implemented strict lockdowns and those with minimal restrictions, though the interpretation of these differences remains contested.
Peer-reviewed studies document substantial mortality differences between jurisdictions that implemented strict lockdowns and those with minimal restrictions, though the interpretation of these differences remains contested.
According to the Journal of the Royal Society of Medicine, Norway recorded 544 COVID-19 deaths through 2020 compared to Sweden's 11,005 deaths, with Sweden's COVID-19 mortality rate during the first wave 9.7 times higher at 2.9 per 100,000 versus Norway's 0.3 per 100,000. According to the International Journal of Environmental Research and Public Health, Sweden had noticeably higher COVID-19 death rates per 100,000 than Denmark during March-October 2020.
Over a longer timeframe, according to Frontiers in Public Health, Sweden recorded 158 deaths per 100,000 during 2020-2022, lower than Italy (451), Poland (475), Spain (332), and the UK (289). These raw per-100,000 comparisons do not adjust for confounders such as timing of the first wave, population age structure, or healthcare capacity, and are therefore not direct measures of lockdown effect. According to Gavin Publishers, a small publisher whose figures are not corroborated by independent sources, Sweden's excess death rate for March 2020-June 2022 was 3.7%, which the publisher characterized as the lowest in the EU.gavin-publishers The same source reported that the UK and Spain each had 10-11% excess mortality during the same period, three times higher than Sweden's rate. These claims from Gavin Publishers have not been corroborated by independent sources.gavin-publishers
According to a Nature Scientific Reports study, not implementing mandatory lockdown in Sweden resulted in approximately 3,420-4,411 deaths that may have been avoided by Week 22 of 2020 (late May 2020). According to a medRxiv preprint, England's lockdown order on March 23, 2020 effectively decreased daily new case rates and related mortality compared to Sweden.
According to a medRxiv preprint, New Zealand had the lowest cumulative COVID-19 death rate in the OECD at 5.2 per million, which was 242 times lower than the 38-OECD-country average of 1,256 per million.
What we couldn't verify
- What the net mortality impact of lockdowns was when combining direct COVID-19 deaths prevented with indirect deaths from healthcare disruption, as no study has established a comprehensive mortality accounting across all causes
- How much of the mortality reduction in lockdown jurisdictions was attributable to lockdown mandates versus voluntary behavior changes that would have occurred without government orders, as most studies acknowledged difficulty disentangling these factors
Documented Impacts on Non-COVID Health Outcomes
Multiple studies documented substantial disruptions to routine healthcare and associated mortality during lockdown periods, though researchers have not established definitive causal links or net mortality calculations.
Multiple studies documented substantial disruptions to routine healthcare and associated mortality during lockdown periods, though researchers have not established definitive causal links or net mortality calculations.
According to Nature British Journal of Cancer, 1,489,237 women had breast cancer screening delayed by 2-7 months between July 2020 and June 2021, resulting in a modeled estimate of 148-452 additional breast cancer deaths. The cited studies document healthcare disruption and associated modeled mortality estimates but do not disentangle the extent to which disruptions were driven by lockdown mandates versus pandemic-related fear and hospital strain. According to PLOS ONE, cancer screenings dropped 94% over the first four months of 2020. According to Fondazione The Bridge, an Italian foundation, more than 600,000 surgical interventions were cancelled in Italy in 2020, over 50,000 being oncological surgeries.fondazione-the-bridge-italy The same source, citing a UK study, estimated a 7.9-16.6% increase in cancer deaths due to 12-month diagnostic delays from March 2020-March 2021.nature-british-journal-of-cancer
According to the London School of Hygiene & Tropical Medicine, cardiovascular disease deaths peaked at relative risk 1.14 in the week ending April 24, 2020 in England and Wales.london-school-of-hygiene-tropical-medicine The same institution reported that cardiac emergency department attendances decreased by 43% and had not recovered by May 15 after lockdown easing. According to BMJ Lancet Regional Health, India's national lockdown had lasting effects on health facility admissions, which remained below pre-COVID levels through December 2023. According to PLOS ONE, more than one in three people in the US delayed or skipped care because of COVID-19 exposure worries.plos-one
According to a PMC peer-reviewed study, researchers found a small negative association between lockdown intensity and suicide mortality rates in U.S. states from March 2020 through December 2021, rejecting the hypothesis that lockdowns increased suicide deaths. Frontiers in Public Health researchers suggested that strict lockdowns may have caused negative indirect health effects contributing to excess mortality in 2021-2022, though this remains a hypothesis rather than an established finding.
What we couldn't verify
- What the net mortality impact of lockdowns was when combining direct COVID-19 deaths prevented with indirect deaths from healthcare disruption, as no study has established a comprehensive mortality accounting across all causes
Meta-Analyses and Systematic Reviews of Mortality Impact
Meta-analyses examining lockdown mortality effects have produced modest effect estimates, with significant methodological caveats acknowledged by the researchers themselves.
Meta-analyses examining lockdown mortality effects have produced modest effect estimates, with significant methodological caveats acknowledged by the researchers themselves.
According to a Johns Hopkins Institute meta-analysis, the average lockdown in Europe and the U.S. in spring 2020 reduced COVID-19 mortality by 0.2% based on stringency index studies. The same research group reported that shelter-in-place orders reduced mortality by 2.9% on average. A later, updated version of the same analysis published in Munich Personal RePEc Archive reported updated figures: 3.2% mortality reduction from stringency index studies and 2.0% reduction from shelter-in-place orders, translating to 4,000 avoided deaths in Europe and 3,000 in the U.S. The researchers concluded that evidence fails to confirm lockdowns have significant effect on reducing mortality when voluntary behavior changes are excluded, though this represents an interpretive judgment rather than a directly measured fact.
The overall evidence on lockdown mortality impact is genuinely mixed, with no clear preponderance in either direction. No cited study calculates net all-cause mortality combining COVID deaths averted with indirect deaths from healthcare disruption, leaving the question of net mortality impact unresolved by the existing research.
According to the Royal Society Open Science systematic review, stay-at-home orders reduced the effective reproduction number by approximately 50-52%. However, the same review noted that a common difficulty in studies was disentangling effects of social distancing measures from other interventions applied simultaneously.johns-hopkins-institute-for-applied-economics
According to a Rocky Mountain Center for Occupational and Environmental Health press release, researchers found no statistically significant link between how strict a state's lockdown measures were and its excess mortality rate in U.S. states. This finding has not been corroborated by peer-reviewed publication.
What we couldn't verify
- What the net mortality impact of lockdowns was when combining direct COVID-19 deaths prevented with indirect deaths from healthcare disruption, as no study has established a comprehensive mortality accounting across all causes
- How much of the mortality reduction in lockdown jurisdictions was attributable to lockdown mandates versus voluntary behavior changes that would have occurred without government orders, as most studies acknowledged difficulty disentangling these factors
Methodological Limitations and Research Disputes
Researchers across multiple studies have documented significant methodological challenges in isolating lockdown effects from confounding factors, with some arguing that common study designs systematically overstate or understate lockdown benefits.
Researchers across multiple studies have documented significant methodological challenges in isolating lockdown effects from confounding factors, with some arguing that common study designs systematically overstate or understate lockdown benefits.european-journal-of-medical-research
According to the Royal Society Open Science, a common difficulty in studies was disentangling effects of social distancing measures from other interventions applied simultaneously. According to a Simon Fraser University working paper, studies not accounting for voluntary responses drastically overstate lockdown benefits and radically overestimate daily deaths without lockdown.european-journal-of-medical-research
According to European Journal of Medical Research researchers, they were unable to investigate confounding factors including adherence to lockdown, social distancing protocols, hygiene conditions, and disease testing systems in their study of 27 countries.european-journal-of-medical-research According to a medRxiv systematic review, cross-sectional and pre/post study designs are inappropriate for COVID-19 policy causal inference due to inability to account for confounding, epidemic trends, and selection biases.medrxiv
A Nature Scientific Reports article that reported being unable to explain variation in deaths per million across world regions by differences in staying at home compared to baseline was subsequently retracted.nature-scientific-reports The retraction is understood to reflect fundamental methodological or data problems rendering the finding unreliable; the cited sources do not indicate the reason was researcher misconduct.european-journal-of-medical-research
What we couldn't verify
- How findings from the retracted Nature Scientific Reports study (which reported being unable to explain mortality variation by staying-at-home differences) would have been interpreted had the methodological problems not led to retraction
Natural Experiments at Jurisdictional Boundaries
Studies comparing neighboring jurisdictions with different lockdown policies provide some of the clearest evidence on lockdown effects, though even these natural experiments show complex patterns.
Studies comparing neighboring jurisdictions with different lockdown policies provide some of the clearest evidence on lockdown effects, though even these natural experiments show complex patterns.
According to PNAS researchers, in 11 of 13 cohorts of neighboring U.S. states, the state with later or no lockdown had higher disease prevalence. The same study reported an average daily difference of 480 cases per million per day by April 30, 2020 between lockdown and non-lockdown neighboring U.S. states.pnas-proceedings-national-academy-sciences
According to KFYR Television, by the end of August 2020, South Dakota had slightly higher cases (13,509 versus 12,000) and deaths (167 versus 145) than North Dakota.kfyr-television-north-dakota However, the same source reported that by October 13, 2020, North Dakota surpassed South Dakota with nearly 80 more deaths despite initially stricter measures.kfyr-television-north-dakota These figures represent snapshots at specific time points in an evolving epidemic trajectory rather than final outcomes.
What we couldn't verify
- What the counterfactual mortality would have been in lockdown jurisdictions had they not locked down, as natural experiments show associations but cannot definitively establish what would have happened under alternative policies
- Why meta-analyses using stringency indices found minimal mortality effects while natural experiments at jurisdictional boundaries found substantial differences, and which methodology more accurately captures lockdown effectiveness