The rapid spread of COVID-19 in 2020 prompted a series of unprecedented policy measures. One such measure was mandating the use of face coverings or masks. Many countries and sub-national authorities resorted to the policy. Widespread mask wearing – unlike strict lockdown measures – allows for continued human interaction and economic activity. It is thus of both human and macroeconomic relevance.
With the rollout of vaccines in primarily advanced economies, mask mandates were phased out in many places. However, as the COVID-19 delta variant became prevalent, health authorities are learning that breakthrough infections in vaccinated people are more common than anticipated.1 This coupled with vaccine hesitancy in some parts of the world has put mask mandates again at the centre of the debate.
So do mask mandates save lives?
The re-emergence of mask mandates in the policy debate has underscored the need for evidence on their effectiveness. Mask mandates may reduce the transmission of COVID-19 directly by inducing greater mask usage, which lowers the transmission of airborne diseases (Greenhalgh et al, 2020 provides a survey of the evidence). Mask mandates can potentially also work indirectly, by inducing behavioural changes that reduce infection risks (such as physical distancing). However, for the policymaker, the important question is whether mandating mask wearing is effective in curbing the pandemic.
The experience in 2020 provides precious data for identifying the effect of mask mandates. However, using these data to get credible estimates is not trivial. The greatest challenge for identification is that the decision to impose a mask mandate is likely not exogenous and depends on the severity of the pandemic at the local level. Several existing studies have tackled the question using a variety of methods (Chernozhukov et al. 2021, Leech et al. 2021, Lyu and Wehby 2020, Mitze et al. 2020, Renne et al. 2020).
In a recent working paper (Hansen and Mano 2021), we contribute to the debate by assessing the effects of US state-level mask mandates using a quasi-experimental setup that allows for a causal interpretation – a regression discontinuity design. Conceptually, we also add to existing studies by finding that mask mandates have markedly different effects depending crucially on the attitudes towards mask wearing. These conditional effects are critical to obtaining credible effects.
Figure 1 illustrates our approach. The map shows counties in and around New Mexico. The colour scheme relates to mask mandates in place on 1 June 2020, as an illustration. Counties with a mask mandate are shown in green; at this time, only New Mexico had a mandate in this region, so all of the ‘greens’ are in New Mexico. Neighbouring counties that do not have a mask mandate and are shown in red. You can think of the border between greens and reds as a ‘mask border’. So, what is the identification idea? We zoom into these ‘mask borders’ and look at COVID-19 outcomes across reds and greens, and we ignore counties in white that are far from the ‘mask border’. Counties around these mask borders are likely to be in the same local stage of the pandemic. The border is random from their perspective, and so is the mask mandate policy.
Figure 1 Counties subject to state-wide mask mandates in and around New Mexico on 1 June 2020
Notes: The map shows counties in New Mexico and the neighbouring states of Arizona, Utah, Colorado, Kansas, Oklahoma and Texas. Counties are coloured (i) white if they are beyond 150 miles from the “mask border”; (ii) red if they are within 150 miles from the ‘mask border’ and do not have a mask mandate; and (iii) green if they are within 150 miles from the ‘mask border’ and have a mask mandate.
Source: Hansen and Mano (2021).
Thinking beyond that particular mask border, Figure 2 shows raw data for COVID-19 outcomes for all weeks in 2020 up to December 19, across all mask borders. The horizontal axis shows a county’s distance to a county without a mask mandate. Imagine driving through the US and starting at the far left on the horizontal axis at -150 miles – the reader would be in a county coloured red in Figure 1. As one drives towards a state that has a mask mandate, the distance to the mask border declines until it reaches zero, meaning a mask border has just been crossed. Now the reader is in a county coloured green in Figure 1. Continuing to drive, one leaves the mask border behind and gets deeper into ‘green’ territory. What happened when the reader crossed the mask border? The vertical axis shows the number of new weekly covid-19 cases per 100,000 inhabitants. There seems to be a discontinuity at the border – the number of new cases per 100,000 inhabitants drops from around 160 to around 110. Similarly, the number of new deaths per 100,000 inhabitants also drops markedly. These are just raw data, without any controls.2
Figure 2 New weekly COVID-19 cases (left) and deaths (right) per 100,000 inhabitants
Note: These charts show raw data of new weekly COVID-19 cases and deaths and thus do not account for county or time fixed effects as done in our econometric analysis. Data are binned in intervals of three miles.
Source: Hansen and Mano (2021).
Going from raw data to an econometric model, we find that mask mandates on average reduced new weekly COVID-19 cases, hospital admissions, and deaths by 55, 11, and 0.7 per 100,000 inhabitants, respectively (Figure 3). To put these numbers in perspective, the average new cases, hospital admissions, and new deaths in our sample are 166, 24, and 2.6. This implies that mask mandates reduced the average weekly COVID-19 cases, hospital admissions, and deaths by roughly a third, a half, and a quarter, respectively. Our results are robust to several robustness checks. Importantly, our results are not driven by other contemporaneous state-level health and containment policies or the existence of county-level mandates in states without state-wide mandates.
Figure 3 Effects of masks mandates on new weekly COVID-19 cases, hospital admissions, and deaths per 100,000 inhabitants
Note: The figure shows the estimated effect (and 90% confidence interval) of state-wide mandates on new weekly COVID-19 cases, hospital admissions, and deaths.
Source: Hansen and Mano (2021).
We also find that estimated effects vary strongly with attitudes towards mask wearing, which themselves vary significantly across the population (Milosh et al., 2020). Such conditional effects are important enough that they must be taken into account. Indeed, the estimated effect of mask mandates on weekly deaths per 100,000 varies from around -2.5 to 0 across our sample of counties, compared to the average effect of -0.7 mentioned above.
Our results imply that mask mandates saved 87,000 lives through 19 December 2020, while an additional 58,000 lives could have been saved in the same period if all states had put in place a mandate starting in April 2020. The magnitude of these effects is large – COVID-19 deaths in the same period amounted to around 309,000 in the US. And yet, these are likely to be lower bound estimates. Indeed, states imposing mask mandates are often reacting to relatively worse outbursts of the pandemic. This would lead us to find smaller differences between counties with and without mandates.
Since 2020, the deployment of vaccines against COVID-19 is a major advance in the battle against the pandemic. Vaccines are highly effective at reducing serious complications from a COVID-19 infection, and can even reduce transmission. Still, mask mandates remain a highly relevant tool in the years to come. First, not all countries have the same access to vaccines and, even in countries with ample supply, the demand for vaccines is hampered by hesitancy. Second, there is already some evidence that newer variants of the virus can generate breakthrough infections even in vaccinated individuals, meaning vaccine efficacy may decline over time.
Authors’ note: The views expressed in this column are those of the authors and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.
Chernozhukov, V, H Kasahara, and P Schrimpf (2021), “Causal impact of masks, policies, behavior on early Covid-19 pandemic in the US”, Journal of Econometrics 220(1): 23–62.
Hansen, N-J H and R C Mano (2021), “Mask Mandates Save Lives”, IMF Working Paper No. 21/205.
Lyu, W and G L Wehby (2020), “Community Use Of Face Masks And COVID-19: Evidence From A Natural Experiment Of State Mandates In The US”, Health Affairs 39(8): 1419–1425.
Milosh, M, M Painter, D Van Dijcke, and A L Wright (2020), “Unmasking Partisanship: How Polarization Influences Public Responses to Collective Risk”, Becker Friedman Institute for Economics Working Paper 2020-102.
Mitze, T, R Kosfeld, J Rode, and K Wälde (2020), “Face masks considerably reduce COVID-19 cases in Germany”, Proceedings of the National Academy of Sciences 117(51): 32293–32301.
Renne, J-P, G Roussellet, and G Schwenkler (2020), “Preventing COVID-19 Fatalities: State versus Federal Policies”, arXiv preprint 2010.15263.
2 The reader may wonder why cases appear to be increasing in distance on the right side of the cut-off. Notice however that the figure shows raw data without controls such as county and time effects. Once these are added, only the jump at the border remains robustly significant.