Do people change their behaviour when face masks are mandatory?

Although face masks reduce the chance of transmission of coronavirus, some governments have been reluctant to make their use mandatory in public places. People’s behaviour is likely to change when wearing masks, but it is uncertain whether the risk of infections will rise or fall.

Face masks have increasingly been introduced as a tool to supplement other policy measures to control the Covid-19 pandemic, such as social distancing and formal lockdowns. In some countries, face masks are now mandatory in closed public spaces and on public transport, and they are generally encouraged. Elsewhere, notably in the United States, there has been strong reluctance among some to wearing face masks, prompting policy-makers to resist making their use mandatory.

At the same time, there is mounting evidence that countries where face masks are in widespread use have a better record of controlling the spread of the disease. This includes parts of the UK, which initially did not mandate their use.

The science on the effects of face masks on the spread of infection is still not clear cut. But several studies have now shown that face masks reduce the transmission of virus carrying droplets from infected individuals, and that most of the benefits from mask-wearing accrue to others, rather than to the person wearing the mask. In other words, people wearing masks are themselves still at risk of infection (for example, by exposure through their eyes), but reduce the risk that they infect others.

Yet even though masks reduce the chance of transmission, some public authorities have remained reluctant to make mask-wearing mandatory. This is because of the possibility that masks may cause behavioural changes, making their wearers become less cautious and reducing social distancing.

According to Henning Bundgaard, a professor at the University of Copenhagen, ‘If you feel safe, you might change your behaviour, maybe you get closer to people and maybe you don't wash your hands so often […] So the achievement might be none, or you might be worse off in fact.’

Why consider requiring face masks?

Short of pharmaceutical interventions such as vaccines and antiviral drugs, there are essentially two ways to try to reduce the transmission of the disease. The first is to reduce the extent of potentially transmissive contacts. This can happen in several ways, but typically relies on reducing the physical proximity of people and, by doing so, reducing the number of contacts between at-risk individuals and people who carry and spread the disease.

The second way to reduce transmission is to try to make each such potential contact less conducive to disease transmission. While not necessarily controlling the number of potentially transmissive contacts, this second approach makes it more difficult for the virus to spread when infected and susceptible people meet. The widespread use of face masks is intended to work in this latter way.

Covid-19 is a respiratory disease that spreads through droplets projected by sneezes and coughs by carriers of the virus. These droplets may be either breathed in by others in proximity or may be transferred through contact with surfaces. There is mounting evidence that face masks may catch droplets when carriers sneeze or cough and, in that way, make it less likely that people exposed to an infected person will themselves become infected.

Why might the use of face masks backfire?

There is concern that face masks could backfire and help to spread the disease. The reason is that mask-wearing could change the way people behave.

When people decide whether to expose themselves to the risk of infection, say, by going to a supermarket, by eating out or by going to work, they weigh their personal costs and benefits of doing so. For example, when deciding whether to attend a concert, the enjoyment of the event must be traded off against the elevated risks of contracting the disease. Thus the behaviour of people and how much they engage in self-protection depends on how they perceive the risks they face when exposing themselves to potential infection.

When face masks are in widespread use, an individual may perceive the risks involved in going out as lower than they otherwise would have been. This slightly alters the perceived trade-offs between the costs and benefits of self-protection and may thus lead to an overall increase in transmissive behaviour.

It may seem counter-intuitive that face masks both reduce person-to-person transmission and at the same time increase the overall number of new infections. The point is that it is not only the wearing of masks that changes: behaviour changes too.

In a world in which no-one wears masks, exposing oneself is riskier and so people are going to be more cautious. Once widespread mask-wearing becomes the norm, going out becomes ‘safer’, but only in the sense that the chance of transmission for a given number of contacts becomes smaller. But with more overall activity, the probability of getting infected may well go up once masks are introduced.

The effects described here are an instance of the ‘risk compensation effect’, formulated by Chicago economist Sam Peltzman in the context of safety equipment in cars (see Peltzman, 1975). The key difference between the classical ‘Peltzman effect’ and the one considered here is that in the case of self-protection against infectious diseases, there is a stronger ‘positive externality’ that individuals do not necessarily take into account when deciding how much effort to make to avoid infection. In other words, others may benefits from their self-protection.

In one study, I explicitly consider the interaction between behavioural responses and externalities in preventive behaviour (Toxvaerd, 2019). I find that the introduction of transmission-reducing measures such as face masks (or pre-exposure prophylaxis, in the case of sexually transmitted diseases such as HIV/AIDS) can indeed actually make everyone worse off. This possible effect is known as ‘immiserisation’, meaning that the adverse behavioural effects of a risk-reducing innovation outweigh the reduced risks and thus backfire (see Gersovitz, 2011).

So if mask-wearing is widespread, enough people may become less cautious in their behaviour that there is an increase in the overall number of new infections. But note that this argument cuts both ways. Infected people who refuse to wear face masks increase the chances that they infect others and therefore put others at risk. In doing so, being near them becomes riskier. But in keeping with the theory of disinhibition, this may prompt others to be more prudent in their behaviour, and this could thus lead to less overall infection.

It should be emphasised that the existence of a Peltzman effect is not in itself enough to conclude that there is immiserisation. It is all about trade-offs. A behavioural disinhibition response to mask-wearing is only a problem for society as a whole when the negative effects on others of the increase in some people’s risky behaviour outweigh the benefits of being able to reduce social distancing and the costs that go with that.

Should the wearing of face masks be encouraged?

Care is needed in drawing policy conclusions from the possibility of a behavioural response to mandating face masks. The risk compensation responses may make everyone worse off, but they need not do so. The judgement really depends on the properties of the disease and on the costs and benefits of self-protection, as perceived by people themselves. Although immiserisation is a theoretical possibility, it is not certain it will occur in practice.

On theoretical grounds, there are sound arguments on both sides of the debate. Only careful empirical research on the question of behavioural disinhibition can determine whether this should be a real concern for policy-makers when considering widespread wearing of masks.

Where can I find out more?

When safety measures lead to riskier behaviour by more people: Alex Horenstein and Konrad Grabiszewski writing at The Conversation.

Why making masks mandatory isn’t enough: Alex Horenstein and Konrad Grabiszewski argue that people who wear masks might take more risks without realising it and that public policy needs to take that into account.

Masks, seatbelts, and Peltzman effects: Vincent Geloso concludes that all sides of the debate on mandatory use of masks invoke their arguments with certainty and without the shadow of a doubt – some humility would be welcome.

Risk compensation: from face masks to credit, market and systemic risk: Overview from Open Risk, an independent provider of financial risk analysis tools and training.