Understanding the "Sometimes Masker": Political Orientation and Trust in the Media During the COVID-19 Pandemic.

OBJECTIVES: The COVID-19 pandemic in the United States has brought to light the problematic way partisan politics interferes with public health prevention and control measures. This study aims to investigate how Americans responded to the novel coronavirus with respect to their sociopolitical identity and masking habits. STUDY
DESIGN: This mixed-methods study incorporated three ethnographic projects and surveys together, from two rural areas (in Iowa and California) and one suburban community in California.
METHODS: We interviewed 156 Americans about how masking habits related to six themes: participants' perceived risk level, concern for themselves and others, support for President Trump, trust in scientific organizations, and confidence in major news outlets. We conducted content analysis of qualitative interviews and evaluated survey questions to understand how and why people masked or engaged in public health prevention practices.
RESULTS: Greater perceived risk, concern for others, and trust in health and media institutions was correlated with increased masking, while support for Trump was predictive of anti-masking sentiments. Participants who diverged from these trends, specifically those who sometimes wore masks, but not always were called "sometimes maskers". These sometimes maskers often identified as politically moderate and were more likely to mask due to concern for a vulnerable person or group in their lives.
CONCLUSIONS: Since one in three Americans are political moderates, understanding what promotes their adherence to public health guidelines is essential for policy makers interested in pandemic containment. Relatedly, the conservative tendency to distrust mainstream media is what separated those who reported sometimes masking from those who reported always masking.

Introduction

“Politics do play a big role in it. I think Trump and his tweets and his arguing with the professionals and doctors and stuff, I think that hurts their case because unfortunately people hear him and believe him.” -Small business owner, late thirties, in Okoboji, Iowa

The American response to COVID-19 has been overtly political. [1] How and why people followed public health measures, or did not, was linked to how people vote. [2] This was particularly concerning during the first year of the American COVID-19 pandemic when President Donald Trump was reluctant to admit that the disease posed a significant risk, downplaying its severity and emphasizing its detrimental effects on the economy. [3] Yet, how these messages affected people’s perception and experience of viral risk has received limited attention in the social science literature. [4] This study investigates how masking behaviors were associated with perceived risk, personal politics, trust in health institutions, and faith in the media in three American communities, including two rural areas and one suburb.

While many argue that Trump’s dismissal of COVID-19 influenced his political support and ultimate election loss in 2020, [5] his leadership style shaped how some Americans perceived and subsequently followed public health measures in the midst of the first respiratory pandemic in over a century. [6] For instance, white evangelical Christians were most likely to get their information on and develop strong views about COVID-19 risk from President Trump, imbuing meaning in both politics and religion and often perceiving Trump as a Messianic Figure [7]. This overlap of politics and personal values had a demonstrable impact on the health behaviors of the white, conservative, Christian public. [8].

Masking became a contentious topic, and in some ways a weapon and symbol in defining partisan politics; masking became a way for people to visibly demonstrate their political views. [4] In a study of mask-acceptance and the experience of COVID-19, Cherry et al. (2021) found that knowing someone who tested positive for COVID-19 was associated with support for masking and mask mandates, even if the person themselves had never tested positive. They found that there was a significant, but very small, association between exposure to CDC guidelines and wearing a mask, even among some Trump supporters. However, they also found that people who maintained moderate politics and tested positive for COVID-19 were less likely to mask. They suggest that mass testing, and finding cases among those with minor symptoms may have had the unintended consequence of suggesting that COVID-19 posed a more muted risk. [9] We found this in one of the study sites (Okoboji) investigated for this study, where people coined the term “Corona Light” due to the low mortality rates and mild symptoms experienced by many people in the community. [4].

While there seems to be a clear relationship between political affiliation and masking, actions of political moderates are less predictable. One explanation for this is that moderates tend to frame COVID-19 as a constraint to personal or professional growth while also expressing concern for themselves or others. [4] In general, these moderates lean farther left with regards to certain social issues and right for others, making them a particularly interesting population to study when it comes to pandemic prevention. Political moderates comprise a sizable portion (34%) of the United States, and as such, understanding what promotes their adherence to public health guidelines is essential for policy makers interested in pandemic containment. In what follows, we investigate how social and political factors impact masking habits, particularly among political moderates, and in doing so consider what hinders ‘sometimes maskers’ from always masking around others.

Methods

This mixed-methods study incorporated three ethnographic projects and surveys together, from two rural areas (in Iowa and California) and one suburban community in California during the summer of 2020 during the COVID-19 pandemic. The first site was Okoboji, Iowa, a small tourist district of 17,000 permanent residents. [4] The second site was in Redding, California, a large rural tourist town of 89,000. The third site was in Morgan Hill, California, a suburb of San Jose with a population of around 45,000. While the first two sites are majority white, conservative, and Christian, Morgan Hill is home to a large Latinx population. We primarily interviewed year-round residents at each site to ensure that we garnered an authentic understanding of local perceptions and experiences. The study was led at each site by co-authors who had been full-time residents to optimize familiarity with the communities.

The two rural areas, Okoboji and Redding, demonstrated elements of “rugged individualism” seen in many former frontier regions; [10] this individualism manifested in community members pushing back against government proclamations and ensuring that few public health measures were left in place. However, state level mandates differed in Iowa and California. In Iowa, the conservative government intervened minimally and relied on local authorities. In contrast, the Democratic governor of California issued state-wide stay-at-home orders and mask mandates. Similarly, in Morgan Hill, which leans Democrat, there was less push-back against the governor’s public health initiatives in part because of political alignment.

Study Samples and Methods

We employed snowball sampling to recruit study participants for each site. [11] We began by reaching out to people from each community through Facebook, email, text message, and phone calls; we also met some participants informally and recruited them for the study. We also sought business owners’ views amidst periodic economic shutdowns to understand how and why businesses were following or disregarding public health measures. Those who declined to interview - be it due to availability conflicts or a lack of interest - often put us in touch with others who would potentially be willing to speak with us.

We met most participants virtually over Zoom, though we conducted some interviews outdoors and socially distanced. We asked for consent from participants prior to each interview (via email or Facebook Messenger) and acquired verbal consent before beginning each interview. The study was reviewed and approved by the Institutional Review Board at Georgetown University.

We interviewed 156 Americans about how masking habits related to six themes: participants’ perceived risk level, concern for themselves and others, support for President Trump, trust in scientific organizations, and confidence in major news outlets. Our interview guide was similar across study sites, although each interviewer adapted the interview guide to adjust to local businesses, environments, and symbols. For instance, when fires raged in Redding, California, we started asking about fires and masking, which introduced new findings and meanings around masking. We inquired about the timing of people’s experiences with COVID-19 (When did you hear about, quarantine for, respond to, and adapt for coronavirus?), the risk that the virus poses to them (Who are you most concerned for in your family? Yourself? A family member? What are you doing to mitigate risk?), and their sense of personal responsibility, which focused explicitly on what COVID-19 prevention behaviors they were engaging in and for what reasons [4, 18, 19].

Data Analysis

We conducted content analysis of qualitative interviews and evaluated survey questions to understand how and why people masked or engaged in public health prevention practices. We focused on six core themes that emerged from 156 study participants. First, we coded each interview to evaluate if, and how often, people reported masking daily. We coded frequency of responses relating to masking habits, ranging from ‘never (0)’ to ‘sometimes (1)’ and ‘always (2)’. Second, we surveyed, “Do you perceive yourself to be ‘high (2)’, ‘medium (1)’, or ‘low (0)’ risk for coronavirus?” We surveyed “Are you concerned about yourself or others getting sick?”, with options: no, concern for self, and concern for self and/or others. Fourth, we coded how people talked about their support for President Trump and categorized their responses as none or supportive. Fifth, we coded for trusting or untrusting of medicine and health institutions, with focus on discussions about National Institutes of Health (NIH) and Centers for Disease Control and Prevention (CDC). Finally, we coded for trust or lack of trust in the media. Exemplar quotes for each theme are in Table 1.

Table 1

Representative Quotes of Each Theme

	Exemplar Quote
Perceived Risk (low)	“The “majority of [young] people are just like ‘Whatever, who cares. If I get sick, I get sick.’ Everyone is just thinking of it as the common cold. They will be tired for a few days, and then they will be fine.”
Perceived Risk (medium)	“I am 21…I actually consider myself like a medium risk. I was nervous about it because I vaped for three years and before that, when I was a kid, I had asthma.”
Perceived Risk (high)	“Especially with the younger group from ages 18 or younger to 30… They seem to want to live life the way they enjoy it. Ok, partying and socializing. And that’s spreading the virus. And who has to pay the price? It’s the older people that are vulnerable.”
Masking (Never)	“I have been verbally attacked with the most heinous comments wishing that I die because I’m not wearing a mask. And everywhere you go, you know, the masks are a symbol of fear… I’ll go anywhere and everywhere without restriction. I have no fear, I fear people. I don’t fear this virus.”
Masking (Sometimes)	“I go out and I go sometimes with a mask and sometimes without a mask. I’m not a big advocate for masks. I think that they have some value, but I also think that … a virus [is] so small that it’ll go through the cloth.”
Masking (Always)	“I don’t want to give it to my daughter. I mean as you know I’ve lost a child I’m not going to lose another one. So, that is why I keep my mask on me pretty much wherever we go.”
Concern for Self/Others (none)	“At one point I really gave into it; I gave into the fear. And then, you know, the fakeness that was happening. And it kind of took me going into that stream of fear to really realize that I feel like I’m being controlled right now.”
Concern for Self/Others (self)	“I think if I ever got it I would be absolutely terrified because I think they all just make you think you’re gonna die.”
Concern for Self/Others (others)	“I almost never get sick. So, if I did contract it, I would be able to beat it I feel. But I don’t want to give it to my daughter. I’ve lost one child; I’m not going to lose another one. So, that is why I keep my mask on me pretty much wherever I go.”
Trump Support (yes)	“I do believe that president Trump is doing the best that he can. I’m thankful he is in office because of his strength. We elected him to do a job and I really believe he is not doing perfectly but I believe he is doing the best that he can.”
Trump Support (no)	“And then the president, who has all of his subjects just blindly listening to everything he says, says ‘Oh yeah disregard that I know that like every scientist that works for me says to wear a mask, but we live in a free country and you should be able to decide for yourselves.”
Trust in Medicine (yes)	“That is what scientists do…they look at something that they think they know, and they learn more and they change their thinking with evidence. That is how science works.”
Trust in Medicine (no)	“I think any lung disease or sickness at this point, they’re just calling it covid. I honestly believe that some of the numbers could be rigged. I honestly don’t know what to believe.” “[Dr. Fauci] originally [said] “No, there’s no need for a mask. We don’t need masks.” That was basically the gist of it: that masks don’t do anything. Well, come to find out he’s telling a lie.”
Trust in Major News (yes)	I: Where do you get your news? P: “The New York Times, Rachel Maddow, MSNBC, CNN, Foreign Affairs…just everywhere…The New York Times for sure. I feel like that’s my main news source.”
Trust in Major News (no)	“I feel like it’s mainstream news just cherry-picking stories, you know, to go with what their agenda is. And, you know, I feel like I’m not given enough information.”

Table 2 shows demographics for the Okoboji, Redding, and Morgan Hill samples. Table 3 presents how each theme associated with risk, politics, and trust were distributed among people who never, sometimes, or always wore masks. Table 4 shows significant correlations between masking habits and each of the six themes.

Table 2

Demographics

	Location
	Morgan Hill, CA (n = 29)		Okoboji, IA (n = 97)		Redding, CA (n = 30)		Total (n = 156)
	N	%	N	%	N	%	N	%
Age Range
0–30	6.00	21%	8	8%	10	33%	24	15%
30–60	9.00	31%	75	77%	11	37%	95	61%
60+	14.00	48%	14	14%	9	30%	37	24%
Gender
Man	8	28%	38	39%	17	57%	63	40%
Woman	21	72%	59	61%	13	43%	93	60%
Other	0	0%	0	0%	0	0%	0	0%
Income
Low/Low Middle	1	3%	12	12%	6	20%	19	12%
Middle	15	52%	47	48%	12	40%	74	47%
Upper Middle	12	41%	14	14%	8	27%	34	22%
High	0	0%	24	25%	4	13%	28	18%
Politics
Republican	4	14%	40	41%	7	23%	51	33%
Democrat	17	59%	41	42%	12	40%	70	45%
Other	8	28%	16	16%	11	37%	35	22%
Religious
Yes	4	14%	66	68%	5	17%	75	48%
No	25	86%	31	32%	25	83%	81	52%
Race/Ethnicity
White	15	52%	94	97%	25	83%	134	86%
Other	14	48%	3	3%	5	17%	22	14%

Table 3

Masking and Perceptions of Risk, Politics, and Trust

	Masking
	Never (23, 15%)		Sometimes (31, 20%)		Always (102, 65%)		Total (n = 156)
	N	%	N	%	N	%	N	%
Perceived Risk
High	1	4%	1	4%	24	92%	26	17%
Medium	1	3%	10	32%	20	65%	31	20%
Low	21	21%	20	20%	58	59%	99	63%
Concern
None	13	62%	6	29%	2	10%	21	13%
Self	1	3%	4	11%	32	86%	37	24%
Others	9	9%	21	21%	68	69%	98	63%
Trump
Support	21	39%	10	19%	23	43%	54	35%
No Support	2	2%	21	21%	79	77%	102	65%
Trust Med
No	15	56%	9	33%	3	11%	27	17%
Yes	8	6%	22	17%	99	77%	129	83%
Trust Media
No	16	26%	19	31%	26	43%	61	39%
Yes	7	7%	12	13%	76	80%	95	61%

Table 4

Masking Correlations with Risk, Politics, and Trust

	Masking Habits
	Never Mask		Sometimes Mask		Always Mask
Perceived Risk	-0.219**		-0.095		0.242**
Concern	-0.507***		-0.122		0.480***
Trump Support	0.406***		0.256**		-0.517***
Trust in Medicine	-0.527***		-0.154		0.522***
Trust in Major News	-0.260**		-0.226**		0.383***
p < 0.05*		p < 0.01**		p < 0.001***

Results

In Table 2 we show that among 29 participants from Morgan Hill, 97 from Okoboji, and 30 from Redding, most identified as women (60%), were white (86%), and from 30 to 60 years old. The sample was roughly split between religious and non-religious individuals and almost half identified as middle class. 12% identified as low/lower middle class and 40% identified as upper middle/upper class. Most reported Democratic political beliefs (45%), one third were Republicans (33%), and others identified as Libertarian, Moderate, Independent, or unaffiliated (22%).

Tables 3 and 4 display how various identities correlated with masking habits. Table 3 shows that most study participants (102, 65%) reported always masking when in public, regardless of whether it was requested or mandated. In contrast, 23 people reported never masking (15%) and 31 people reported sometimes masking (20%). Most participants (63%) reported that they perceived themselves to be at low-risk for COVID-19, compared to those who reported high (17%) and medium (20%) risks. Table 4 illustrates a significant positive relationship between higher perceived risk and always masking (r = 0.242, p = 0.01), and a significant negative relationship between perceived risk and never masking (r=-0.219, p = 0.01).

Table 3 shows that most respondents were more concerned about the risk that COVID-19 posed for their family and friends than they were about how COVID-19 could impact their own health. Most were concerned about others getting sick (63%) compared to those who were only worried about themselves (4%) or themselves in addition to their family and friends (20%). Table 4 shows that as participants’ concern for others increased, they were significantly more likely to mask (r = 0.480, p = 0.001). Those who expressed no concern for others masked significantly less often (r=-0.507, p = 0.001).

Table 3 shows that participants who supported President Trump were significantly less likely to mask (r = 0.406, p = 0.001). Among those who never masked, 91% expressed support for President Trump. In contrast, as demonstrated by Table 4, less support for President Trump correlated with increased masking (r=-0.517, p = 0.001). Of the ‘sometimes maskers’, 32% supported Trump, while only 23% of those who always masked did. The majority of those who always masked (78%) did not approve of Trump.

Table 3 illustrates that most people trusted science (129, 83%) compared to those who expressed mistrust in science and health institutions (27, 17%). Table 4 shows how greater trust in medicine correlated significantly with increased masking (r = 0.522, p = 0.001), while diminished confidence in science aligned with less masking (r=- 0.527, p = 0.001). Most (97%) of those who always masked expressed trust in medical authorities, including the CDC, NIH, and the World Health Organization. Contrastingly, the majority of those who did not mask expressed distrust (65%) of these institutions. Of those who ‘sometimes’ masked, 70% expressed trust in science, while 30% voiced skepticism.

Lastly, Table 3 indicates that the majority (80) of participants who expressed trust in the media always wore masks, while only 43% of respondents who distrusted the media always wore masks. Table 4 shows that those who do not mask were significantly less likely to trust major news sources (r=-0.260, p = 0.01), while always masking correlated significantly with faith in mainstream media (r = 0.383, p = 0.001).

Discussion

This study shows the perceived risk, politics, and trust people residing in rural and suburban areas in the United States grappled with in relation to their COVID-19 beliefs. We found expected and significant relationships between masking behaviors and how people perceived risk COVID-19 posed to themselves and their loved ones, their support for President Trump, and their trust in science and in major news outlets. [4] There were particularly interesting findings among the 20% of people who sometimes masked; these individuals only wore masks when mandated by the government or by private businesses, or when masking was explicitly requested by others. Some believed that masks were not entirely effective, and many masked only to avoid shaming and conflict with others.

These ‘sometimes maskers’, therefore, had less polarized beliefs about politics and public health compared to those who always masked or to those who never masked. It is this small number of non-partisan respondents that we believe presents the most opportunity for targeted public health interventions to shift the population towards more careful attentiveness to public health. Many of these individuals align with what we have elsewhere called the “concern frame,” where people mask to protect those who are vulnerable or because it is required, even when they are not necessarily worried about their own risk. [4] Political moderates comprise a sizable portion (34%) of the US population, and as such, communicating public health messages for this target audience is crucial to elevate public health understanding across the country.

Yet, there may be something more revealing about those sometimes maskers, which underscores the importance of strong institutional leadership in the midst of a pandemic. Table 4 shows that the tendency to “sometimes mask” was not significantly related to perceived risk or concern for other people. While sometimes masking was significantly correlated with support for President Trump, this correlation was weaker than the opinions of Trump held by the ‘never maskers’ and ‘always maskers,’ indicating relative political moderacy. Sometimes masking was not significantly correlated with trust in medicine, but had a significantly negative relationship with trust in media. This trend contrasts those seen in participants with more extreme views, such as those who subscribe to conspiracy or crisis frames; [4] those who reported ‘always masking’ or ‘never masking’ tended to either trust both science as well as the media (for always masking) or trust neither institution (for never masking). It is only the moderate 20% that had unremarkable views on science but significant distrust in the media.

This deviation seems to relate to how political ideology relates to trust. The 20% of participants who sometimes masked were significantly more likely to have a non-partisan political affiliation, such as Libertarian, Moderate, or Independent. Unlike self-identifying Democrats and Republicans, the behaviors of political moderates are less predictable, as they encapsulate both liberal and conservative values. Our participants who were ‘sometimes maskers’ neither trusted nor distrusted scientific institutions; there was no significant relationship between sometimes masking and trust in health institutions. However, participants who sometimes masked leaned farther right with regards to the media, sharing a distrust in news outlets with many Republican respondents.

Therefore, the gap between always masking and sometimes masking appears to relate to the conservative ideologies surrounding trust in the media, rather than political beliefs. Republican distrust of the media is not a new phenomenon; a study conducted in 2004 found that only 60% of Republicans and right-leaning moderates had faith in major news outlets. [12] Trump’s use of Twitter to convey presidential correspondence to the public rather than utilizing traditional news outlets served to further undermine the legitimacy of mainstream media in the eyes of his supporters and spur distrust. [13] By 2021, the percent of conservatives that retained faith in the media had dropped to 35%, which starkly contrasts the 78% of Democrats that expressed trust in the media. [14] We found many study participants reported sentiments aligning with this distrust; one individual from Okoboji stated, “I think Fox and CNN are some of the biggest pushers on fake news, so I try not to watch them.” Even those that held more moderate views than self-identifying Trump supporters had similar opinions: “I hate to say it, but Trump is absolutely right. He coined fake news…it’s like…maybe this idiot is actually on to something.” Decreased confidence in the media amongst Trump supporters impacted how they received news reports on COVID-19; specifically, Republicans are less likely to believe COVID-19 coverage than Democrats. [15].

Due to this mistrust of major news outlets, many participants reported supplementing their news intake with Facebook and other forms of social media. A similar approach was taken by politically moderate people, who felt that mainstream media presents only partisan viewpoints and neglects objectivity in reporting; therefore, these individuals turned to social media and other websites as well in order to obtain minority viewpoints. [12] However, alternative news sources present issues of their own, including misinformation and conspiracy theories. This, in conjunction with studies that show conservatives perform worse at discerning real news from fake news than liberals, [16] makes those with conservative beliefs more susceptible exposure to and belief in the misinformation disseminated online.

The misinformation shared by alternative news sources throughout the pandemic often “demonized” the medical community, [17] calling COVID-19 a hoax and suggesting government leadership was lying regarding its severity. In this way, distrust in the media may have lessened political moderates’ faith in public health (which was also observed in the lack of significant association between sometimes masking and trust in medicine). Many study participants held similar beliefs that COVID-19 case numbers were being “rigged” by the media, and dire cases were being “cherry-picked” to exaggerate the severity of the disease. Yet, few thought that the pandemic was a government-sponsored mechanism to “decrease the population worldwide”.

We found Americans with diminished confidence in mainstream media are more likely to believe that COVID-19 is not severe and are subsequently less likely to mask. It is this conservative tendency to distrust major news sources that we believe may explain the discrepancy between the ‘sometimes maskers’ and those who always mask. This belief that COVID-19 is not severe may decrease people’s perceived risk and reduce concern for others among moderates as well [18].

There were meaningful limitations to this study. For instance, we cannot do a comprehensive analysis comparing the three sites, in part because this was a qualitative study where we used purposive sampling as opposed to a large representative survey where we might have used random sampling. Instead, we look at how individual self-reported politics and behaviors collide. Moreover, during the period between June and October of 2020 when interviews were being conducted, several economic, political, and viral challenges emerged, including a summer surge of COVID-19 cases, struggling tourist economies, and public exhaustion due to repeated quarantines. Nevertheless, this ethnographic snapshot of a period in American history provides important understanding and potentially can inform future public health approaches.

Conclusions

This study illustrates that when and why people mask is not consistent, particularly across spectrums of political beliefs. Our results suggest that the spectrum of masking habits - ranging from ‘never’ to ‘always’ - parallels the political spectrum, with political moderates falling in between. We attribute this to the tendency of moderates to echo the conservative skepticism of mainstream media. To ensure greater compliance with masking in future outbreaks, we must restore faith in media and government, and public health experts and policymakers must continue to work on combating the spread of misinformation online.