An overview of research findings explaining mask compliance across countries
According to Francis Fukuyama, leading political thinker, institutional arrangements, including increasing judicialisation and influence of interest groups, constrict decision making in democracies. These along with legally mandated checks and balances can slow decision making in crises. Fukuyama was writing in the US context (The Origins of Political Order, 2011; Political Order and Political Decay, 2014). But, India too is no exception. This perhaps explains the need for seemingly autocratic decisions. Whatever be the method, public view and experience of governmental decisions, through law or regulation, are manifested in compliance.
The Covid-19 pandemic tested the ‘limits to medicine’ (a phrase popularised by Ivan Illich’s eponymous book) and the supporting infrastructure like none else did since the Spanish flu of 1918-20. The unavoidable delay in developing a vaccine, following due process of testing and validation, necessitated developing strategies to counter the spread of the virus. These are collectively known as social or community mitigation strategies. These strategies included personal protection measures (PPM). Apart from staying at home, the three most important measures were mask wearing, social distancing, and hand washing.
In this post, I consider mask wearing as a proxy for compliance with the entire gamut of personal protection measures. The title, Why People Wore Masks, is based on the seminal tract, ‘Why People Obey the Law,’ by Tom Tyler (Princeton University Press, 2006). Tyler is now Professor of Law and Professor of Psychology at the Yale Law School. I am sure he would have debated whether to title his book, alternatively, Why People Disobey the Law. It is a matter of perception and perhaps depends on whether obedience or disobedience is more dominant. Similarly, I could have titled my post, Why People Don’t Wear Masks? The past tense presumes that the worst of the pandemic is behind us.
If regulation or standards refer to mandated rules or suggested behaviour to realize certain desirable outcomes, compliance means the literal adherence to such rules and behaviour, the actual outcome, or the process of obtaining such outcome (this is an extempore definition, nevertheless rooted in the literature).
We can discuss compliance at four levels. International standards required compliance at the national level, as in nuclear safeguards or anti-money laundering standards. National level regulation could even require regional or state government level compliance, as in the three-language formula and other official language measures. At a disaggregated level, compliance can also be by organizations and individuals. Compliance at any of these levels is determined by a host of factors. There is a large body of literature on the subject cutting across several disciplines: law, political science, psychology, economics, sociology, organizational behaviour, organization theory, game theory, neo-institutionalism, criminology, and so on.
Pandemic and compliance
The pandemic offered a rare opportunity for studying compliance with similar regulation across the world all happening within the same timeframe. In a series of articles, the Regulatory Review of the University of Pennsylvania had tracked the legal responses to the pandemic. But, they don’t seem to have similarly tracked compliance. A search of the top 50 papers in Google Scholar showed that many have already attempted that. Some of these have compared the experience across select countries. But, barring a few, they don’t generally correlate with the existing body of literature on regulation and compliance. What are the factors that explained compliance with mask wearing requirements? Or noncompliance? My brief notes on some of the findings are in the following paragraphs. I have tried to categorise them according to some of the well-known constructs seen as explaining compliance.
Interestingly, mask wearing was much lower than expected in many jurisdictions, and they reported higher incidence of Covid, even among health care professionals. In Latin America, only less than half the population complied with all the community mitigation strategies (CMS). 82.9% in masks … it was higher for hand washing and lower for social distancing. Wearing a mask was important also as a social signal to others to maintain distance. As per some research it did not create a false sense of security regarding distance. Seres et al 2021
Personal protection measures mandated staying home, avoiding public places, limiting trips, and more conservative practices. As mask wearing require venturing out, the evidence is based on behaviour of those who defied or were compelled to defy these requirements.
As Koebele et al. (2021) put it, “Understanding the full range of determinants of mask-wearing is critical for promoting evidence-based public health solutions to slow the spread of COVID-19.”
Legitimacy and mask compliance
Legitimacy of regulation is an important determinant of compliance. Some extend this to mean legitimacy of the regulator/government. This is not to be confused with its literal meaning, it is determined by whether it is justified, is it the right way to address the problem, how well it is drafted, whether it considers market practices, etc. The research findings related to legitimacy are as below:
Individuals who believed wearing masks protected others were more likely to report voluntarily wearing them. The authors (Bir and Vidmar) attributed this to altruism. I would put it under legitimacy, which is also determined by the belief that the regulation will be effective to meet a valid objective.
Both organizations and individuals tend to reinterpret regulation in the context of their circumstances. Accordingly, some believed that face mask use is not based on hard evidence.
Perceived risk of contagion for the self and for others determined mask wearing compliance.
There was support for legitimacy as a determinant of mask wearing compliance. But, in another study, perceived legitimacy did not predict compliance with PPM. This could be attributed to lack of trust in health authorities in that country. Another possible explanation could be loss of faith in authorities to protect them. In such situations, instrumental factors such as enforcement, become more important as compared to normative factors where compliance is based on culture and attitude.
The other factors affecting perceptions as to legitimacy included the following:
- Lack of effectiveness of mask wearing
- Unnecessary or inappropriate for certain people
- Negative effects of mask wearing
Trust and mask compliance
A related construct emphasised by a host of authors, including Tyler and others, is trust in the issuing authority, whether the government or a regulator. Fukuyama devotes an entire book to Trust.
Aversion to any control was less common when trust in the government and the information it provides was greater. This dichotomy was nowhere more evident than in the two parts of Germany where those brought up under the coercive regime of erstwhile East Germany showed less trust. The higher the confidence in the government, the greater the compliance with PPM whether voluntary or enforced. Enforcement could be counterproductive where interpersonal trust and social capital are traditionally strong as in Denmark or Norway.
Those with higher trust in the government were more compliant with mask wearing requirements. In an unequal country such as Chile, where trust was low, compliance was low.
Trust building measures are important whether it is Kashmir or Koodankulam. It also points to the need for different strategies in enforcement and communication. This could vary according to whether a State is ruled by the same party as at the Centre, is ideologically linked, or is opposed, ideologically or otherwise.
The Role of Culture
That individual culture, and the norms and values that one is exposed to will influence compliance needs no elaboration. Assuming that the social benefits of mask wearing are made known widely and that individuals have internalised these messages, compliance with mask wearing should become universally prevalent.
Culture and mask compliance
But, there can be variations for the following reasons:
- personal comfort preceding societal interest
- perceived comfort costs
- anti-mask attitudes were also explained by physical discomfort
- safety culture or ‘predilection towards compliance with health-related directives.’ This probably explains the early favourable numbers coming out of Kerala which has had a culture of congenial health habits (Please see my article on the Kerala Model).
- personal hygiene factors
- handwashing and use of PPE predict face mask wearing compliance
- self-interest and health concerns
- pro-sociality or other-regarding concerns increased risk perception and, therefore, compliance. A viable compliance strategy would target this among people, like the young, who are otherwise not worried about the disease.
- personal costs of mask wearing, in particular comfort costs
- in one study, coefficient of concern for others contracting the disease was less than half of that for oneself
- personal and descriptive norms were the most important predictors of mask wearing.
- conditional cooperation (willing to incur prosocial costs if I perceive others also doing so). Experiments across cultures show that free riders disincentivise conditional cooperators and cooperation would be difficult to sustain.
Attitudes towards mask compliance
Indicative of the kind of attitudes people have towards mask compliance are the following questions used in a survey:
- Wearing a mask is a sign of weakness
- It is a sign of being a follower
- It is a sign of caring for others
- Science shows that wearing masks is effective to prevent disease transmission
- Government-mandated mask wearing is an attack on individual freedom
- Wearing a mask is uncomfortable
- My decision to wear a mask predominantly affects me
- My decision to wear a mask predominantly affects others
- Businesses have right to refuse entry and service to people NOT wearing
- I look down on people who wear masks in public
- I look down on people who do not wear masks in public
- Masks are for healthcare professionals, not me
- People look silly when wearing masks in public
- Scaled responses to how effective mask wearing is
One major difference between organizational compliance and individual compliance is that political inclinations are usually not a factor in the former while it is in the latter, even if it comes to something as essential and basic as mask compliance.
‘Anti-mask attitudes were stronger in conservatives, and were associated with increased concerns for in-group loyalty, national identity, and personal liberty.’ Kaplan et al 2021. Within this group, messages framed in terms of loyalty, appealing to protection of the community and America, reduced anti-mask beliefs, as compared to control messages based on pure scientific facts.
Framing messages according to the beliefs and values of the audience was quite effective in improving compliance. For example:
- Pure science-based message: Evidence is clear. Wear a mask.
- Liberty: Choose freedom. Choose to wear a mask.
- Loyalty: Help protect America. Wear a mask.
There could also be a combination of the above messages.
Being more liberal, risk-averse and with higher subjective numeracy went with higher mask compliance. The framing effect discussed elsewhere was also moderated by political ideology on mask-wearing, effect stronger in liberals than in conservatives. The study advocated depoliticising PPM for greater compliance.
Perceptions of inequities and civic values predicted mask-wearing.
Political conservatism was negatively related to mask-wearing. However, conservatives were more likely to wear masks if they reported greater perceptions of inequities.
Both liberals and conservatives report higher compliance if they had personal experience of the consequences of the virus.
A more liberal tendency was associated with greater mask compliance.
Various studies confirmed psychological differences as influencing compliance. Individual beliefs or attitudes could influence compliance. One study used checkboxes that included the following:
- masks are uncomfortable
- I don’t think it will prevent me from getting sick
- I don’t think it will prevent others from getting sick
- looks silly
- my choice not to wear one
- looks weak
- I don’t have one
- it fogs up my glasses
- my friends aren’t wearing them
- it is socially awkward to wear one
- my right not to wear one
- hard to exercise while wearing a mask
Of course, these overlap with various other categories.
Koebele et al. found that psychological factors, including threat- and efficacy-related perceptions, were significant determinants of mask compliance. According to them, “respondents’ perceptions of self-efficacy (e.g., ability to wear a mask) and response efficacy (e.g., effectiveness of mask-wearing in reducing COVID-19 transmission) better predict mask-wearing behaviour than a number of commonly cited sociodemographic factors.”
Self-efficacy and collective efficacy, psychology-based concepts popularised by the psychologist Albert Bandura, have been suggested in the context of compliance with PPM.
Some people found masks unnecessary or inappropriate. Greater perceived effectiveness indicated increased personal compliance.
Those in the loss-framed condition than in the gain-framed condition were more compliant.
Greater self-control, NFC (need for cognition, tendency to enjoy and engage in effortful, systematic thinking) and health/safety risk averse all predicted higher mask compliance. The relationship was stronger in liberal participants in the survey.
In the context of organizations, Amy Edmondson at Harvard underlined the importance of learning environment. I used the concept in my own research on organizational compliance. In the case of mask compliance, organizational and individual compliance overlap when it comes to health care professionals (HCP). Studies showed that active discussion, and increased communication from leadership was effective in increasing mask compliance among HCP.
The importance of communication and educating the public about the pandemic and the need for PPM were validated by many studies. Incomplete knowledge, and confusion about the role of masks led to spread of the virus. Exposure to instructions regarding face mask use was the strongest predictor of mask compliance.
When information about effective protection is inconsistent, compliance would vary. The level of inconsistency could also vary over the life of the pandemic with recommendations coming in from different sources starting with the World Health Organization (WHO) which itself changed its stance on the use of the mask.
The level of understanding of the pandemic, which changed with ongoing research and available information, impacted perceived effectiveness, and compliance.
Early implementation of stay-at-home orders and exposure to Covid-19- related news impacted compliance.
Both the time spent watching and searching for Covid-19-related news impacted outdoor mask use but not indoor mask use.
Less time on social media and internet browsing information on the pandemic meant low compliance.
Epidemiological data on test positivity rate and number of confirmed cases did not influence compliance.
Suggestions for learning
The suggestions for improving information availability to improve compliance included the following:
- Increase mass compliance education
- Centralise information to avoid conflicting messages
- Non-threatening communication
- Attitudes and norms as key tools in messaging
- Target males and those below 25 who were less compliant.
- Consider which factors influence which population
- The above implies different methods of communication with varying content
- Follow moral reframing, or align messages with the core values of the receiver
- Loss framing of messages will be more beneficial to increase compliance
- In spreading information, include cooperative learning and reflective learning, instead of traditional lectures
- Include social media in media outreach
Internal control and audit are factors not relevant for individual compliance. But organizational approaches influence individual compliance, as in the case of health care professionals (HCP). In such cases, audit and passive feedback were effective in improving compliance among HCP in the US.
The influence of others is very important in compliance, whether by organizations or by individuals. In a seminal paper, DiMaggio and Powell (Iron Cage Revisited) begin with a rhetorical question: Why do organizations look similar? Explaining this phenomenal which they called institutional isomorphism, they went on to list three factors: coercive, mimetic, and normative. The first is the influence of government and regulators through law and regulation. Second is the result of mimicking peers. The third is the influence of professionals such as auditors and consultants who through their work spread similar practices across organizations. Somebody else added a fourth reason, technological, being the influence of adopting similar technology, as in core banking solutions, across organizations. I had identified such isomorphic factors as significant factors determining compliance in my research. A similar approach would also apply for individual compliance.
The evidence from research supporting the influence of external factors could be summarised as below:
- Pro-sociality difficult to sustain without norm enforcement or other ways to nudge compliance
- People in counties where healthy behaviours were commonplace, reflecting collective values and norms around health, showed higher mask compliance.
- Prevalence of mask-wearing by others in same district positively influenced mask compliance.
- Relatives and friends comply too
- Social identity effects as influencing mask compliance. Thus, in an example of ‘in-group favouritism’, non-mask wearers cooperated significantly less with mask wearers as compared to mask wearers. They cooperated more with non-mask wearers as compared to mask wearers. (Powdthavee, 2021)
- High perceived social norms predicted higher compliance
- Social norms had stronger effects on compliance than instrumental concerns about being punished or getting infected.
- Promoting beliefs about compliance being the norm might have a stronger effect on mask compliance.
- In a paper titled ‘His lack of a mask ruined everything,’ Kostromitina, et al. (2021) showed how customer reviews on Yelp influences mask compliance.
- Greater norms related to indoor mask compliance.
- Normative concerns regarding duty to support the authorities dominated compliance decisions. Instrumental and normative factors predicted compliance with mandatory public health restrictions. (Murphy et al 2020)
These findings suggest adopting a range of strategies, such as normative nudges, and not just enforcement to increase compliance. Officials can alter governance structures and normative behaviours to improve healthy behaviours.
Mask compliance and Enforcement
Tyler, in his Why People Obey the Law, emphasised the role of instrumental factors and ‘procedural justice,’ in addition to normative factors. Some argued for no role for social shaming or enforcement in improving mask compliance. On the other hand, they suggested promoting altruism and improving governance structures. The evidence has been mixed.
Lessons from enforcement
Some of these are summarised below:
- Instrumental factors were relevant predictors of compliance.
- Significant effects of perceived legal enforcement on compliance
- As stated earlier, enforcement could be counterproductive where interpersonal trust and social capital are traditionally strong as in Denmark or Norway.
- While securing higher compliance, enforcement may crowd out voluntary motivation.
- People will support PPM more under voluntary rather than enforced implementation
- Control aversion, or opposition to enforcement, varied across policy measures: it was rare for masks but much higher for vaccination and a tracing app on the mobile. Considering the criticality of total or near-total compliance, some measures required enforcement while for others it was found to be unwise.
- Crowding out of voluntary support due to enforcement was more a concern in liberal democracies than in authoritarian regimes.
- Effective enforcement reduces concerns on crowding out is not valid. The challenge is whether the policies can be sustained over a long period. If enforcement has provoked an adverse response, the challenge would be greater.
- Enforcement without people’s consent provokes discontent and aggressive behaviour, “which may be channeled in destructive acts such as domestic violence … or the riot nights in Paris, Stuttgart, Frankfurt, or Bogotá.” (Schmelz 2021)
- “…worry about freedom loss, opposition to surveillance tactics, police heavy-handedness and perceptions of procedural injustice from police during the pandemic all drive bounded-authority concerns… bounded-authority concerns are associated with reduced duty to obey and mediate the relationship between procedural justice and the duty to obey authorities’ enforcement of COVID-19 restrictions.” (Williamson et al., 2022)
Governments are more effective when they combine enforcement with voluntary compliance. Schmelz (2021) echoes Fukuyama when he writes that the enforcement power in a liberal democracy may be limited especially with regard to vaccination and enforcing contact tracing apps on mobile phones. Full enforcement may infringe upon privacy and personal autonomy especially in countries like Sweden. This explain why democratic government delayed the more restrictive measures.
Availability of masks and other such personal protection equipment was a constraint at least as per one study (Looi, 2022).
Mask compliance across countries/cities
PPM, including mask compliance and hand washing, was high even in cities with no such mask wearing culture.
Mask wearing was 20% higher in Hong Kong than in Vermont, USA. Compliance was higher in Wisconsin than in Vermont probably reflecting their differences in income, education, and political affiliation. (Beckage and Buckley, 2021).
Counties where health-related physical environment was poor showed higher mask compliance even after allowing for demographics and political leanings.
There were significant differences in compliance even in areas just five kms apart such as Honolulu and Waikiki. These could be because downtown locations allowed exemptions. Other reasons could be differences in the number of residents, their age, employment backgrounds, and planned activities.( Tamamoto et al, 2020)
Frey et al. (2020) from the Oxford University observed that democratic countries were not as successful in restraining mobility as autocratic countries. In India, this was famously visible in large scale labour migration on foot starting a few days after the lockdown was imposed.
It is interesting to see whether locations had anything to do with mask compliance. In one study, lowest compliance was found at golf courses followed by gas stations. Maybe the social distancing that naturally goes with a golf course legitimised a no-mask approach. The mandatory measures were more effective in hospitals, retirement homes, and elderly communities. In Argentina, mask compliance was highest among people walking followed by people cycling, and the least compliance by runners. Is it that the physically active assumed that they would be the least vulnerable?
Stage of the pandemic
As expected, mask compliance rates varied as the pandemic progressed with the end nowhere in sight. A study in Indonesia showed that compliance decreased in November and December 2020, before the peak of the second wave. It increased between January and May 2021, during and after the second wave.
In Italy, when PPM requirements were extended for longer periods than expected, people were less likely to comply reflecting a pandemic fatigue or compliance exhaustion.
In Turkey, images of pedestrians crossing at signals showed high mask compliance at the beginning of the pandemic. People wore masks less correctly after 40 days of the pandemic, despite increasing Covid cases. This probably reflected a decrease in disease risk perception.
In a study across five cities in Australia, US, and the UK, pandemic fatigue was more common among the young. This reflected changing epidemiology, risk perception and pandemic fatigue, more common among males and the young.
In later stages, better judgment and decision making were better determinants of mask compliance than epidemiological information.
The effect of demographic factors on mask compliance has been shown by many authors. These could be summarised as below:
- Higher compliance by females and those above 60 whether US or Saudi Arabia. Looi et al rationalised that “men might see face mask as infringing upon their independence, whereas women might only perceive face mask as being uncomfortable and be more willing to wear due to their self-protective instinct…In addition, women have a stronger tendency to feel shame from deviating from the norm and are more influenced by moral limitations…”
- Contrary to experience elsewhere, one study showed less mask compliance among older adults in Australia, probably reflecting cultural differences.
- Another study in South Africa also found significantly lower compliance among the elderly despite the higher mortality risk.
- Ethnic variations
- Significant differences according to affluence/income
- Clinical care, and social and economic factors did not influence mask compliance
- Older people, people with higher educational levels, and those who do not work are more likely to comply with personal protective measures.
- Conflicting evidence in relation to gender and health risks.
- Females showed greater compliance.
- White individuals were less likely to engage in outdoor mask use compared to non-White individuals. As the racial minority had higher morbidity rates due to systemic inequality (e.g., access to medical care), they are more likely to take steps to minimise infection. (Gette et al.)
- Compliance varied by political beliefs, gender, living situation (alone, with parents, with family, partner, etc.), and race.
- More education and higher belief/ understanding in science meant higher compliance.
- hypertension, obesity, or being overweight did not imply mask compliance.
- Stress, anxiety, and negative affect indicated greater compliance.
- Those with underlying medical conditions were more compliant
This overview of the literature on mask compliance broadly parallels my experience researching organizational compliance. Legitimacy and trust are there. So are culture, learning environment, and external influence. Enforcement is also low in priority with the emphasis being on nudges and persuasion. So, what is new? Governance, compliance structure, and internal control and audit are out, for obvious reasons. Political views and demographics are in. Do these therefore call for a unified approach to compliance? The regulatory pyramid, describing stages in enforcement or compliance, after all has at its core the centuries-old stages in Kautilya’s sama, dana, bheda, and danda.