The enforcement of statewide mask wearing mandates to prevent COVID-19 in the US: an overview

Face masks have become the bulwark of COVID-19 prevention in the US. Between 10 April and 1 August, 2020, 33 state governors issued orders requiring businesses to require their customers and employees to wear face masks, and persons outdoors who could not social distance to do the same. We documented the policies and enforcement actions for these policies in each of the states. We used governors’ orders and journalists’ news reports as our sources. Our results show that the states used a variety of state and local (county and municipality) agencies to enforce business prevention behaviors and primarily local law enforcement agencies to enforce outside mask-wearing behaviours. In particular, law enforcement officers demonstrated a strong preference for educating non-mask wearers, and indicated a reluctance to resort to civil penalties that were enacted in the state orders. Businesses expressed a preference to have government agencies enforce non-mask wearing behaviours. But there was also a widespread reluctance on the part of local law enforcement to resort to legal remedies.


Introduction
In the United States, the Centers for Disease Control first recommended the wearing of face masks on April 3, 2020. On April 8, The Governor of New Jersey was the first to issue a general, statewide mandate, which required "workers and customers to wear cloth face coverings while on the premises". By August 1, 33 state governors had issued state-wide mask wearing orders. The resulting pattern of regulations regarding mask wearing across the states is a welter. There is wide variation in the design of the ordinances, including whose behavior is being targeted, what is expected of them, when they must observe the ordinances, where their behavior is to be regulated, why they must submit to these orders, and how violations will be enforced. In this paper we summarize the introduction of state policies in those states which issued statewide mandates, and how these mandates are being enforced.

Methods
Our subjects were the 33 states which had mask mandates in effect by August 1.
We used newspaper and broadcasting articles as our primary sources. We identified these articles using Google searches ending August 4 with the following keywords: "COVID-19", "State" (e.g., "Pennsylvania") AND "mask order" or "mask mandate" AND "enforcement" OR "education". We used newspaper or broadcast articles as the primary source because they contained essential information about the mandates, as well as additional information about enforcement. We identify primary sources for each state in the Underlying data, Appendix 1 1 . Most news sources came from local broadcast stations, local and national news services, networks including the Public Broadcasting System, CBS, and ABC. We used the actual ordinances as secondary sources, but these often did not contain information about enforcement.
We abstracted the following items for each state: Are businesses given first-line responsibility for their customers and employees? By reading news articles, we determined if businesses were directly responsible for mask wearing behaviors of customers and employees on their premises. Answers were "Yes," "No," or in a few cases, "Unsure." What authorities enforce whether businesses apply mask wearing orders on their premises? We determined from the news articles which government agencies were responsible for ensuring that businesses were applying the governors' orders. This information might also be obtained from the governors' orders. For each state, we listed these authorities: "LLE" indicates "Local Law enforcement," "LPH" signifies "Local Public Health," and "DOH" signifies "Department of Health." What authorities enforce mask wearing behaviors outside of business premises? We identified the enforcement authorities from the news articles. Abbreviations are the same as the previous question.

Is education viewed as a primary tool to encourage mask wearing behaviors?
We obtained this information from news articles. Answers were "Yes" or "No".
Do local governments (counties and municipalities) pass their own orders? We obtained this information by searching news articles which identified additional county or municipal mask orders within individual states. Answers were "Yes" or "No".
Has local law enforcement shown any resistance to enforcing statewide orders? We obtained this information by searching for news articles that identified local law enforcers' comments on "enforcement." Answers were "Yes" or "No".

Results
The publications on which the analysis was based are shown in the Underlying data, Appendix 1 1 . These data underlying the analysis are shown in the Data Availability section.

Which states issued statewide mandates?
We identify the 33 states with statewide mandates (green) in Figure 1. We did not cover states with county or municipal orders (yellow), states with only municipal orders (brown) or states with no orders (red).
When did the mandate become effective? Mandates became effective in the months of April (8 states), May (7 states), June (4 states), July (12 states) and August (2 states). See Figure 1.
What was the scope of the mandate (indoor, indoor and outdoor, or "in public")?
The mandates covered indoor only (8 states: Colorado, Hawaii, Indiana, Maine, Montana, Minnesota, Montana, Oregon, and West Virginia), outdoor and indoor or "public" places usually where recommended spacing was not available or there were large crowds (25 states). Some states excluded certain industries (e.g., gyms) if they did not meet public health requirements. See Table 2 for individual state information.
Who was the primary target of the violation: businesses or individual non-mask wearing violators? In 26 states, businesses and establishments formed the primary targeted group. Three states (Arkansas, Indiana, New Jersey) targeted only individual mask non-wearers. In two states (Pennsylvania, Wisconsin) we were unsure of the targeted group. See Table 2. What is the maximum penalty of a violation? Ten states had provisions for fines only. The usual amount was between $100 and $200, but NY had a maximum of $1,000. Nine states had provisions for fines and/or jail sentences. The maximum fine in this group was $5,000 (Hawaii, Maryland) and the maximum jail sentence was one year (Maryland), but these amounts were outliers. More typical values were $500 fines and/or 30 days to six months in jail. In five states fines varied by county. The other states did not specify penalties, and some of these could have no penalty: in a number of states sheriffs expressed a strong preference for education over criminal or even civil proceedings. See Table 2 for individual state information.
Who is the enforcement agency? In most of the states (27 states) the government relied on private businesses to enforce mask wearing behaviors. Three states (Arkansas, Indiana, New Jersey) focused directly on non-mask-wearers. We were unsure of the governors' focuses in three other states (Massachusetts, Pennsylvania, Wisconsin). In addition, because many states extended the mandate to outdoor non-business settings, there was enforcement in these settings as well. The pattern of enforcement in the two settings was very different.
Agencies which enforced businesses include state occupation safety and health agencies (Nevada, Ohio, Oregon, Washington, Kentucky); state or local public health agencies (Conneticut, Delaware, Indiana, Kentucky, Montana, New Mexico, New York, Ohio, Virginia, Wisconsin), city or county law enforcement (Connecticut, Hawaii, Kansas, North Carolina), alcohol and beverage control (California, Colorado), business regulators and licensers (Maine, Mississippi, Rhode Island). In some cases regulators were not specified, or the area was not enforced by  the government. The enforcement of non-mask wearing behavior outside of businesses was usually assigned to local law enforcement (19 states) or public health (3 states). In other cases, there was limited to no enforcement outside the business setting. Many local sheriffs or local police chiefs stated their objections to enforcement on grounds that there were limited resources; a few also mentioned constitutional grounds.

Discussion/conclusions
Based on public and journalist reports, we analyzed the introduction and enforcement of statewide mask wearing mandates in the 33 US states that imposed such orders between April 10 and August 1, 2020. Most states relied on businesses to enforce customer and employee mask wearing behavior. However, both the contents of the governors' orders and the type and degree of enforcement varied widely between the states. Enforcement responsibility for personal (outdoor) mask wearing behavior was handed over to local law enforcement.
The business sector of the economy has been very active in COVID-19 prevention during this time period by introducing mask wearing regulations for its customers and employees. A survey of large US retail chains showed that 16 chains had introduced cross-country mask wearing policies in May, 2 in June, and 34 in July. Despite the adoption of storewide prevention policies, businesses, their trade associations, and employee trade unions have expressed concern at taking on primary enforcement roles. Private companies still relied on state laws to provide them with a rationale for requiring customers to wear masks. These businesses are subject to degrees of enforcement that have been inconsistent across states and that have often been lax. Many governors' orders also covered outdoor areas, and most of these were nominally enforced by local law enforcement agencies. However, senior law enforcement officers in all states issued statements that they would not enforce mask wearing orders. For example, 38 sheriffs in Montana issued an op-ed which stated that a mask wearing directive "is not a mandate for law enforcement to issue citations and arrest violators." Although such statements were not universal, examples can be found in every state.
We used both governors' orders and news reports for our data sources. Policies like engaging in public education are subjectively described and data for them are not collected; instead we used sheriffs' interviews with journalists to document the policies. Also, we could not obtain data on citations written by local law enforcement. Nevertheless, information on education and enforcement was widely reported in the press across the nation, and strongly suggests a national trend. It also indicates that mask wearing behavior, now considered a bulwark against the spread of COVID-19, is not being strongly enforced, especially in the outdoor sectors.

Open Peer Review
The study methodology involved conducting online searches using Google for newspaper and broadcast articles as the primary data source. The actual text of the orders (described as "ordinances") was identified as a secondary data source. The documents were reviewed, and specific characteristics of interest were abstracted from the text.
The study found that policies varied substantially among the 33 states that had issued mask mandates, particularly with respect to enforcement and penalty provisions. State policies varied also by authority of local governments to issue mask mandates and exceptions (e.g., age, medical contraindications). While the degree to which law enforcement officials had demonstrated resistance to enforcement was not characterized in the study, the authors noted that examples of resistance could be found in every state.
The article clearly describes the methodology used to identify which states had mask mandates, including the specific online search terms and other parameters. While general descriptions of the abstract categories were provided (e.g., enforcement requirements for businesses, which authorities enforce business requirements, whether education is viewed as a primary tool, etc.), the article did not include detailed definitions for each category. The article also did not describe how the actual text of the orders was utilized in the abstraction process. These methodology details might provide additional clarity with respect to how the states were categorized. For example, Kansas is categorized as having a statewide mask law. While that is correct, the law also allows local boards of county commissioners to adopt orders that are less stringent than the state order, thereby giving counties the option to opt out, which many have done 1 , 2 . The article provides a link to the source data in the form of a Microsoft Word document with URLs for each article or order by state.
Reproducibility of the study is uncertain given the qualitative nature of the methods, the lack of coding method details and reliance on news articles as the primary data source. In addition, while URLs are provided for each source document, full citation details are not provided. As a result, links might be obsolete, or the content might have changed. In our own government mitigation policy research, we have found it challenging to interpret and consistently categorize aspects of public health orders and other official governmental mitigation policy actions. In addition, we have often found discrepancies between published news reports and the actual text of the orders. Nonetheless, the article is clearly written and well-organized, the conclusions appear sound, and the study makes an important contribution to understanding the challenges of implementing and enforcing population-level policies to control the spread of COVID-19.