Political variations in pandemic lifestyles and COVID-19 vaccination by age cohort in the United States

Abstract

While recent scholarship suggests that political affiliation is a robust predictor of pandemic behaviors and COVID-19 vaccination status, research has yet to examine whether the impact of political affiliation on these outcomes vary by age. Drawing on health lifestyles theory, we contribute to the social epidemiology of infectious disease behaviors by testing whether the impact of political affiliation on risky pandemic health lifestyles and COVID vaccination varies by age cohort. We employ data collected from the 2021 Crime, Health, and Politics Survey (CHAPS), a national study of adults from the United States, to formally assess this understudied association. In all models, Democrats reported less risky pandemic lifestyles compared to their Republican counterparts. Moreover, Democrats displayed greater odds of being vaccinated than Republicans or Independents. Further, the impact of political affiliation on vaccination status varied by age cohort, such that the impact of political affiliation was stronger among the oldest adults in our sample. Our analyses contribute to the growing study of politics and health lifestyles by challenging theoretical perspectives and cultural narratives that claim that older adults are less swayed by political influence when it comes to healthcare decisions. Our results help better our understanding of the ways in which political discourse shapes adopting public health recommendations.

1. Introduction

Since July 15, 2022, there have been 1.02 million deaths and over 4 million hospitalizations due to COVID-19 in the United States (US) (CDC, 2022a; CDC, 2022b), mostly concentrated among older adults. Individuals 65 or older account for nearly 75% of COVID deaths (CDC, 2022c; Bosman et al., 2021). In 2021, COVID-19 became the leading cause of death for those aged 45–54 – (Shiels et al., 2022). The Biden administration continues to strongly urge older Americans to vaccinate and take precautions against COVID to prevent new cases, hospitalizations, and deaths (CDC, 2022d; Ibssa, 2022). Similarly, doctors continue to encourage masks, hand-washing, and social distancing in addition to vaccination to reduce the overall rates of serious morbidity and mortality (Bradfield, 2022; Chiu et al., 2020; Doung-ngern et al., 2020).

It is important to identify the subpopulations of adults who engage in the riskiest pandemic lifestyles. For example, studies indicate that individuals who identify as Trump voters, Republicans, or political conservatives exhibited poorer pandemic lifestyles, including lower rates of sheltering-in-place, social distancing, mask usage, and COVID-19 vaccination (Allcott et al., 2020; Cai et al., 2021; Chan, 2021; Corcoran et al., 2021; de Bruin et al., 2020; Fazio et al., 2021; Gollwitzer et al., 2020; Gonzalez et al., 2021; Grossman et al., 2020; Hamilton and Safford, 2021; Hill et al., 2021; Kaushal et al., 2022; Kerr et al., 2021; Perry et al., 2020; Shepherd et al., 2020; Travis et al., 2021). Even though older Americans have been consistently shown to be vaccinated at much higher rates than younger age groups (Bosman et al., 2021), it remains to be tested whether older conservatives engage in these riskier pandemic behaviors more than their left-leaning peers.

1.1. COVID-19 & age

While some studies have shown that political conservatism is a robust predictor of pandemic lifestyles (Hill et al., 2022), subgroup variations by age cohort remain understudied. Some evidence suggests that older adults may be more likely to adopt protective pandemic behaviors due to their increased mortality risk. Research has shown that older adults are more willing to trust medical professionals on topics related to prescription drugs (Donahue et al., 2009) and vaccinations (Zimmerman et al., 2004), as well as trust their doctors over information they find online (Czaja et al., 2009). The fundamental question is do threats to political ideology outweigh threats to personal health among older populations. Drawing on health lifestyles theory, we contribute to the social epidemiology of infectious disease behaviors by testing whether the impact of political affiliation on pandemic health lifestyles varies by age.

Some research suggests that the impact of political ideology may be attenuated among older adults, as seen in pre-pandemic polls finding political ideology was unrelated to flu vaccination (Enton, 2021). Instead, the likelihood of adopting a specific intervention centered more on the patient's risk profile and whether a physician had encouraged the intervention (Butterworth and Campbell, 2014; Nowalk et al., 2004; Zimmerman et al., 2004). Given the politicized environment of COVID-19, it is unclear whether these patterns are generalizable to current vaccination practices. Since older adults perceived COVID to be a special threat to their health, older adults, regardless of political ideology, may adopt healthier pandemic lifestyles than younger adults (Barber and Kim, 2021). This is seen in research that indicates that a majority of older adults took greater precautions during the pandemic (Barber and Kim, 2021).

1.2. The politicization of COVID-19

Many conservative politicians downplayed the risks of COVID-19, i.e., Trump minimizing COVID-19 by claiming it was no worse than the flu (Beer, 2021). While Trump consistently supported the vaccine, he suggested that COVID would disappear with warmer weather and promoted unproven treatments like hydroxychloroquine (Neuman, 2020). Other politicians downplayed the necessity of being vaccinated. While discussing the reopening of schools, Rand Paul argued that individuals who have had a previous infection did not need to be vaccinated because “naturally-acquired immunity is as good as the vaccine” (Paul, 2021). However, the immune response varies by individual; moreover, vaccine-induced immunity can be more reliable than natural immunity (Tu et al., 2023). For example, one study indicated that roughly 1/3 of individuals who were recovering from SARS-CoV-2 infection did not develop antibodies (Liu et al., 2021). Additionally, previously infected individuals who did not get vaccinated had twice the risk of being reinfected than those who vaccinated after infection (Cavanaugh et al., 2021). Still, others called the adoption of healthier pandemic lifestyles political “theater” (DeSantis, 2022). And some suggested that older Americans valued the nation's economy more than their own lives (Rodriguez, 2020).

Some conservatives were quick to weaponize public health failures related to changing guidelines for masks, lockdowns, and immunity, accusing Dr. Anthony Fauci of “flip-flopping” (Faulders and Santucci, 2020). Some of these leaks presented partial statements made by Fauci leading to distrust in the reliability of public health information (Faulders and Santucci, 2020). Moreover, “Fire Fauci” was a popular political rhetoric aimed at Trump supporters in the 2022 midterms (Stolberg, 2022) and research suggests that some conservative Americans were more willing to trust Trump on vaccines than the World Health Organization (Kreps et al., 2020; Peng, 2022). For all of these reasons, the choice to follow healthier pandemic lifestyles may be heavily influenced by the rhetoric of political elites.

1.3. Pandemic lifestyle behaviors

In 1997, William Cockerham and colleagues defined health lifestyles as the “collective patterns of health-related behavior based on choices from options available to people according to their life chances.” Health lifestyle theory emphasizes the role of both agency and structure regarding the choices that we make about our health behaviors (Cockerham et al., 1997; Cockerham, 2005). Traditionally, health lifestyles have included behaviors like tobacco usage, alcohol consumption, and exercise that are generally apolitical. New infectious disease lifestyles include pandemic-specific behaviors like social distancing, wearing masks in public, hand washing, and vaccination (Hill et al., 2022). For clarity, we separate mitigation techniques, such as masking or social distancing, from vaccine uptake.

In this paper, we use national survey data collected from US adults to formally assess whether the associations of political ideology with healthy pandemic lifestyles and COVID-19 vaccination vary across age cohorts. Based on the previous research detailed above, our first hypothesis is that affiliating as a Republican, rather than as a Democrat, will be associated with riskier pandemic lifestyles and lower rates of COVID-19 vaccination. Because previous research shows that older adults may be less impacted by political ideology (Barber and Kim, 2021; Zimmerman et al., 2004), our second hypothesis is that political affiliation differences in pandemic lifestyles and vaccinations will be attenuated for older adults as compared to younger adults.

2. Data

In this study, we use data from the 2021 Crime, Health, and Politics Survey (CHAPS). CHAPS is based on a national probability sample of 1771 community-dwelling adults aged 18 and over living in the United States. Respondents were sampled from the National Opinion Research Center's (NORC) AmeriSpeak© panel, which is representative of households from all 50 states and the District of Columbia (https://amerispeak.norc.org/Documents/Research/

AmeriSpeak%20Technical%20Overview%202,019%2002%2018.pdf). Sampled respondents were invited to complete the online survey in English between May 10, 2021, and June 1, 2021. The study met the data curators' guidelines for Human Subjects and is on file with the institution. The data collection process yielded a survey completion rate of 30.7% and a weighted cumulative response rate of 4.4%. Our cumulative response rate is within the typical range (4–5%) of high-quality general population surveys (see: https://www.pewresearch.org/politics/2021/05/17/scope-of-government-methodology/).

3. Measures

3.1. Pandemic lifestyles

Our first variable of interest is whether individuals were less likely to adopt behaviors that could reduce the spread of COVID-19. This measure reflects an individual's willingness to follow the Centers for Disease Control & Prevention guidance on mitigation practices. It is measured with the mean response to three items: (a) during the pandemic, how often did you wear a mask in public? (1 = always to 5 = never); (b) during the pandemic, how often did you use hand sanitizer in public? (1 = always to 5 = never); during the pandemic, how often did you attend large gatherings? (1 = never to 5 = always).

3.2. Vaccination status

Our second variable of interest is vaccination status. It is measured with a single item: have you been vaccinated against COVID-19? (1 = yes, 2 = no, but I plan to be, 3 = no, and I do not plan to be, 4 = no, and I am undecided whether I will get vaccinated). This measure was recoded into a binary outcome (1 = currently vaccinated, 0 = not currently vaccinated for any reason).

3.3. Political affiliation

In order to test our hypothesis of whether political affiliation is associated with negative pandemic lifestyle behaviors, we use a measure that asks respondents how they identify politically (1 = Republican, 1 = Independent, 0 = Democrat).

3.4. Conservative news consumption

Because media rhetoric may impact an individual's decision to engage in healthier pandemic lifestyles, we also include a measure of conservative news consumption. This is measured with a single item: Which of the following news networks do you watch most often to learn about current events? (1 = ABC, 2 = CBS, 3 = CNN, 4 = Fox News, 5 = MSNBC, 6 = NBC, 7 = Newsmax, 8 = One America News Network, 9 = other [specify], 10 = None of these). All media sources, including those written into the other category, were compared to Media Bias / Fact Check and coded as conservative if the source at least leaned toward the right. We recoded this as a binary measure (1 = primarily use conservative media, 0 = primarily use some other media).

3.5. Other controls

Multivariate analyses also control for background variables that are known correlates of political behaviors, including: gender (1 = women, 0 = men); race/ethnicity (1 = non-Hispanic Black, 1 = Hispanic, 1 = other racial minority, 0 = non-Hispanic white); age (in years); level of education (1 = less than high school, 1 = some college, but no degree, 1 = college graduate, 1 = postgraduate education, 0 = high school degree or equivalent); religious affiliation (1 = mainline Protestant, 1 = Catholic, 1 = other Christian, 1 = other religions, 1 = no religious affiliation, 0 = conservative Protestant); church attendance (1 = never to 5 = several times a week); importance of religious beliefs (1 = not important to 5 = very important); employment status (1 = employed, 0 = unemployed); household income (in dollars, categorical ranging from 1 = $5 K or less to 18 = $200 K or more); marital status (1 = married, 0 = all others); owning a home (1 = yes, 0 = no); type of community of residence (1 = urban area, 0 = small town/rural area); region of residence (1 = South, 1 = Midwest, 1 = West, 0 = Northeast).

3.6. Statistical procedures

Our analyses begin with descriptive statistics for all variables, including ranges, means or percentages, and standard deviations (Table 1 ). In Table 2 , we use ordinary least squares regression to model our risky pandemic lifestyle outcome. Table 2 assesses the association between risky pandemic lifestyle and political affiliation for all respondents (Model 1), respondents under 50 (Model 2), respondents 50–64 (Model 3), and respondents 65 and up (Model 4).

Table 1.

Descriptive Statistics (n = 1771).

	Range	Mean / Percent	SD
Dependentvariables
Adverse pandemic lifestyle behaviors	1–5	1.91	0.76
Vaccination status	0–1	0.66
Politicalvariables
Republican	0–1	0.39
Independent	0–1	0.14
Democrat	0–1	0.47
Conservative media consumption	0–1	0.20
Sociodemographiccontrols
Religious involvement	1–5	2.25	1.35
Evangelical Protestant	0–1	0.23
Mainline Protestant	0–1	0.14
Catholic	0–1	0.19
Other Christian	0–1	0.15
Other religion	0–1	0.05
No religion	0–1	0.25
Religious importance	1–5	3.15	1.58
Male	0–1	0.48
Female	0–1	0.52
Non-Hisp. White	0–1	0.67
Non-Hisp. Black	0–1	0.11
Hispanic	0–1	0.16
Other race	0–1	0.06
Age	18–94	49.67	17.00
Less than HS	0–1	0.03
HS graduate	0–1	0.17
Some college	0–1	0.44
College graduate	0–1	0.21
Postgraduate	0–1	0.14
Employed	0–1	0.62
Income	1–18	10.05	4.14
Married	0–1	0.51
Home owner	0–1	0.62
Urban	0–1	0.83
South	0–1	0.33
Midwest	0–1	0.28
West	0–1	0.25
Northeast	0–1	0.14

Source: Crime, Health, and Politics Survey (2021).

Table 2.

Ordinary Least Squares Regression Predicting Riskier Pandemic Lifestyle Behaviors by Political Affiliation of US Adults.

	All respondents	Respondents under 50	Respondents 50–64	Respondents 65 & up
	Model 1	Model 2	Model 3	Model 4
Political variables
Republican	0.51***	0.58***	0.50***	0.32***
Independent	0.20***	0.24**	0.04	0.21*
Sociodemographic controls
Religious attendance	0.05**	0.07**	0.01	0.04
Mainline Protestant	−0.14*	−0.22*	−0.05	−0.07
Catholic	−0.13*	−0.11	−0.22*	−0.04
Other Christian	−0.12*	−0.14	−0.01	−0.22+
Other religion	−0.29**	−0.38**	−0.13	−0.22
No religion	−0.16*	−0.28**	−0.01	−0.02
Religious importance	−0.04*	−0.08**	0.01	−0.01
Conservative media consumption	0.13**	0.12	0.92	0.32***
Female	−0.15***	−0.05	−0.33***	−0.18**
Non-Hisp. Black	−0.09	−0.01	−0.18	−0.12
Hispanic	−0.04	−0.02	−0.00	−0.17
Other race	−0.09	−0.06	−0.20	−0.00
Age	−0.01***	−0.01**	−0.01+	−0.00
Less than HS	0.07	0.04	0.19	−0.13
Some college	0.01	−0.03	−0.06	−0.01
College graduate	−0.05	−0.01	−0.34**	0.08
Postgraduate	−0.06	−0.11	−0.07	−0.02
Employed	0.05	−0.01	0.20**	0.16
Income	−0.00	−0.00	0.00	0.01
Married	−0.06+	−0.04	−0.07	−0.15*
Home owner	0.03	0.03	0.07	−0.03
Urban	−0.08+	−0.12	−0.04	−0.05
South	0.10*	0.15+	−0.03	0.12
Midwest	0.15**	0.19*	−0.02	0.23*
West	0.09	0.12	−0.07	0.15
N	1749	883	461	425
R2	0.22	0.20	0.26	0.27

+p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001. Reference categories include democrat, non-Hispanic white, men, high school, conservative religious identity, northeastern residence, and rural residence.

Source: Crime, Health, and Politics Survey (2021).

In Table 3 , we use binary logistic regression to model our COVID vaccination uptake outcome. In Table 3, we assess the association between vaccination uptake and political affiliation l for all respondents (Model 1), respondents under 50 (Model 2), respondents 50–64 (Model 3), and respondents 65 and up (Model 4). In Table 2, Table 3, we present our results as odds ratios. Odds ratios are interpreted as the estimated difference in the odds of practicing poorer “pandemic lifestyle behaviors” and vaccination status for those who are Republican or Independent compared to being Democrat, adjusting for other predictors in the model. Table 4, Table 5 formally test the interaction between age cohort and political affiliation. These models allow us to examine whether the impact of political ideology on adverse pandemic lifestyle behaviors and vaccination status varies across age cohorts.

Table 3.

Binary Logistic Regression Predicting Vaccination Status by Political Affiliation of US Adults Presented in Odds Ratios.

	All respondents	Respondents under 50	Respondents 50–64	Respondents 65 & up
	Model 1	Model 2	Model 3	Model 4
Pandemic lifestyle behaviors	0.43***	0.42***	0.58**	0.26***
Political variables
Republican	0.28***	0.30***	0.23***	0.08***
Independent	0.39**	0.47**	0.29**	0.12**
Sociodemographic controls
Religious attendance	1.22**	1.27**	1.21	0.94
Mainline Protestant	1.40	1.49	0.96	0.91
Catholic	1.46+	1.36	2.55*	0.97
Other Christian	1.34	1.48	1.09	1.25
Other religion	1.74	2.82*	1.78	0.10**
No religion	1.44	1.45	1.16	1.03
Religious importance	0.91	0.91	0.80+	0.93
Conservative media consumption	0.78	0.66	1.12	1.21
Female	0.79+	0.58**	1.41	2.06+
Non-Hisp. Black	0.35***	0.21***	1.08	0.98
Hispanic	0.86	0.84	0.64	0.84
Other race	0.53*	0.37**	1.19	1.34
Age	1.04***	1.01	1.12***	1.14**
Less than HS	1.41	1.23	1.35	1.49
Some college	1.44*	1.24	2.25*	1.65
College graduate	2.18***	2.11**	6.62***	1.79
Postgraduate	2.91***	4.14***	3.34*	1.34
Employed	0.93	1.31	0.99	0.67
Income	1.06**	1.06*	1.04	1.16
Married	0.82	0.77	1.33	1.03
Home owner	1.32+	1.25	1.65	0.86
Urban	1.63**	1.83*	2.03*	0.86
South	0.78	0.87	0.58	0.55
Midwest	1.10	1.04	1.34	0.68
West	0.94	1.07	0.63	0.69
N	1745	883	459	423

+p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001. Reference categories include democrat, non-Hispanic white, men, high school, conservative religious identity, northeastern residence, and rural residence.

Source: Crime, Health, and Politics Survey (2021).

Table 4.

Ordinary Least Squares Regression Predicting Adverse Pandemic Lifestyle Behaviors by Political Affiliation of US Adults with Interaction Between Age & Political Ideology.

	All respondents
	Model 1
Political variables
Republican	0.43***
Independent	0.22+
Sociodemographic controls
Religious attendance	0.05**
Mainline Protestant	−0.14*
Catholic	−0.12*
Other Christian	−0.11+
Other religion	−0.30**
No religion	−0.15*
Religious importance	−0.04*
Conservative media consumption	0.13**
Female	−0.14***
Non-Hisp. Black	−0.09
Hispanic	−0.03
Other race	−0.08
Resp under 50	0.50**
Resp 50–64	0.05
Less than HS	0.08
Some college	0.00
College graduate	−0.06
Postgraduate	−0.08
Employed	0.08*
Income	−0.00
Married	0.81*
Home owner	0.03
Urban	−0.08+
South	0.15*
Midwest	0.15**
West	0.08
Interaction of age & Politics
Resp. under 50 with republican ideology	0.13
Resp. under 50 with independent ideology	0.04
Resp. 50–64 with republican ideology	0.05
Resp. 50–64 with independent ideology	−0.12
N	1749

+p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001. Reference categories include democrat, respondents aged 65 and over, non-Hispanic white, men, high school, conservative religious identity, northeastern residence, and rural residence.

Source: Crime, Health, and Politics Survey (2021).

Table 5.

Binary Logistic Regression Predicting Vaccination Status of US Adults with Interaction Between Age & Political Ideology.

	All respondents
	Model 1
Adverse pandemic lifestyle behaviors	0.41***
Political variables
Republican	0.12***
Independent	0.14**
Sociodemographic controls
Religious attendance	1.20**
Mainline Protestant	1.36
Catholic	1.50*
Other Christian	1.37
Other religion	1.66
No religion	1.39
Religious importance	0.92
Conservative media consumption	0.82
Female	0.80+
Non-Hisp. Black	0.37***
Hispanic	0.82
Other race	0.55*
Resp under 50	0.09***
Resp 50–64	0.20**
Less than HS	1.30
Some college	1.46*
College graduate	2.17***
Postgraduate	3.04***
Employed	1.03
Income	1.07***
Married	0.89
Home owner	1.33+
Urban	1.65**
South	0.73
Midwest	1.07
West	0.92
Interaction of age & Politics
Resp. under 50 with republican ideology	2.69*
Resp. under 50 with independent ideology	3.69*
Resp. 50–64 with republican ideology	2.39+
Resp. 50–64 with independent ideology	1.98
N	1745

+p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001. Reference categories include democrat, respondents aged 65 and over, non-Hispanic white, men, high school, conservative religious identity, northeastern residence, and rural residence.

Source: Crime, Health, and Politics Survey (2021).

4. Results

4.1. Descriptive statistics

According to Table 1, the average respondent reported low levels of adverse pandemic lifestyle behaviors (1.91 on a 5-point scale). Most respondents (66%) reported being vaccinated.1 The political composition of the sample included Republicans (39%), Democrats (47%), and Independents (14%). Approximately 20% of the sample got their news from conservative media sources. The average age of the sample was approximately 49 years, and more than half identified as female (52%). The race and ethnic composition of the sample included non-Hispanic whites (67%), non-Hispanic blacks (11%), Latinos (16%), and respondents of other races and ethnicities (6%). Many respondents reported living in an urban area (83%) and it was more common for respondents to live in the South (33%). Over one-third of respondents reported having a four-year college degree or higher (35%), and over half of the sample reported being employed full- or part-time (62%). The average respondent reported an annual household income between $50,000 and $74,999.

In Table 2, we assess the relationship between political ideology and pandemic lifestyles. For all respondents (Model 1), being a Republican is positively associated with engaging in riskier pandemic lifestyle behaviors (β = 0.51, p < 0.001), while being an Independent is positively associated with engaging in poorer pandemic lifestyle behaviors (β = 0.20, p < 0.001) compared to those who identify as a Democrat.

For respondents under 50 years of age (Model 2) being a Republican is positively associated with engaging in poorer pandemic lifestyle behaviors (β = 0.58, p < 0.001). Identifying as politically independent was positively associated with engaging in poorer pandemic lifestyle behaviors (β = 0.24, p < 0.01) compared to those who identify as a Democrat. For respondents 50–64 (Model 3), being a Republican is positively associated with engaging in poorer pandemic lifestyle behaviors (β = 0.50, p < 0.001) compared to those who identify as a Democrat. There was no statistical difference between Democrats and Independents for this age group.

For respondents 65 and older (Model 4), being a Republican is positively associated with engaging in poorer pandemic lifestyle behaviors (β = 0.32, p < 0.001) compared to those who identify as a Democrat. There was a statistical difference between Democrats and Independents among the oldest respondents in our survey, with Independents being positively associated with engaging in riskier pandemic lifestyle behaviors (β = 0.21, p < 0.05) compared to those who identify as a Democrat.2

In Table 3, we assess the relationship between political affiliation and the odds of being vaccinated, accounting for negative pandemic lifestyle behaviors. For all respondents when compared to Democrats, being a Republican was associated with 72% lower odds of being vaccinated (OR = 0.28, p < 0.001), while being an Independent was associated with 61% lower odds of being vaccinated (OR = 0.39, p < 0.01). For respondents under 50 years of age (Model 2) being a Republican is associated with 70% lower odds of being vaccinated (OR = 0.30, p < 0.001), while being an Independent was associated with 53% lower odds of being vaccinated (OR = 0.47, p < 0.01), as compared to being a Democrat. For those 50–64 when compared to Democrats (Model 4), being a Republican was associated with 77% lower odds of being vaccinated (OR = 0.23, p < 0.001), while being an Independent was associated with 71% lower odds of being vaccinated (OR = 0.29, p < 0.01). For those aged 65 and older when compared to Democrats (Model 4), being a Republican was associated with 92% lower odds of being vaccinated (OR = 0.08, p < 0.001), while being an Independent was associated with 88% lower odds of being vaccinated (OR = 0.12, p < 0.01).

While the coefficients in Table 2, Table 3 suggest that the effect of political affiliation on negative pandemic lifestyles and vaccination status may vary across age cohorts, Table 4, Table 5 formally test these interactions. Table 4 examines the interaction between political affiliation and age cohorts predicting negative pandemic lifestyle behaviors. Results from this table suggest that the effects of political affiliation on pandemic lifestyle behaviors do not vary across age cohorts, as none of our interaction effects were statistically significant. Conversely, Table 5 suggests that the impact of political affiliation on vaccination status does vary across age cohorts, although not in the direction that we expected. The effects of political affiliation appear to be most pronounced among the oldest respondents in our survey as compared to the youngest respondents. The effects of both identifying as a Republican and an Independent, as compared to identifying as a Democrat, on COVID-19 vaccination are more pronounced for those 65 and older as compared to those under 50.

5. Discussion

Drawing on a national probability sample of 1771 community-dwelling adults, we formally assessed whether political affiliation influenced the decision to engage in risky pandemic lifestyle behaviors and COVID-19 vaccination and whether this association varied by age cohort. Consistent with our first hypothesis, our results show that political affiliation is associated with both negative pandemic lifestyles and vaccination status. However, our second hypothesis was not supported. On the contrary, the influence of political affiliation on vaccination status was more pronounced for older Americans.

The association between political affiliation and both negative pandemic lifestyles and vaccination status persists while controlling for several important control variables, including conservative media consumption; moreover, the impact of political affiliation on vaccination status endures when accounting for negative pandemic lifestyles. During the COVID-19 pandemic, Trump and other political figures promoted unscientific treatments and contradicted advice offered by the CDC (Kreps et al., 2020). This may explain why the politically conservative Americans in our study were less likely to adopt health behaviors aimed at reducing the spread of COVID-19. These findings were consistent across age cohorts. Future research should examine mediators of this relationship, including sources individuals trusted for advice about COVID-19 vaccination and other COVID-19-related health behaviors.

Unexpectedly, we found that the influence of political affiliation on vaccination status was more pronounced for older Americans. This finding may suggest that older adults are more vulnerable to misinformation campaigns than their younger counterparts. Research has indicated that older adults are especially susceptible to fake news on social media, which may make them less able to discern credible content (Brashier and Schacter, 2021). During the 2016 election, adults over 65 were exposed to and shared more “fake news” than any other age group (Brashier and Schacter, 2021). If we consider that older adults may have more trust in sources of misinformation, including friends, clergy, or favored politicians, they may be resistant to public health information. While other research has indicated that older adults are less likely influenced by political messages that affect their health (Barber and Kim, 2021; Zimmerman et al., 2004), this study shows that political messages can have great influence over older adult's healthcare choices. This suggests that political beliefs can have potentially life-threatening consequences for older Americans, especially considering that the vast majority of COVID-19 deaths are concentrated among individuals over 65 (Nania, 2022). Future research, perhaps employing mixed-methods and qualitative approaches, should examine why political affiliation is a particularly powerful influence on the vaccination behaviors of older adults.

This study has several limitations. First, the data does not allow us to determine if an individual is partially vaccinated or has a medical condition that prevents vaccination. Future research should focus on those who fully intend to not be vaccinated. There may be fewer differences by political affiliation once we are able to parse out partially vaccinated individuals from never vaccinated individuals. Finally, our data did not include respondents' specific sources of information about the COVID-19 vaccine and other health behaviors meant to reduce the spread of infection. Without these measures, we can only speculate reasons for age variations in the influence of political affiliation on vaccination status.

6. Conclusion

Despite these limitations, this study has broken new ground. To our knowledge, this is the first systematic analysis of the influence of politics on older adults' decision to practice COVID-19 mitigation strategies in the US. Viewed broadly, this study adds to the growing literature on politics and the COVID-19 pandemic. Our results help better our understanding of the ways in which political discourse shapes the adoption of public health interventions. Specifically, these findings can help inform public health campaigns to address misinformation and vaccine hesitancy through targeted interventions and community-based strategies that account for age and political beliefs. Moving forward, we should continue to examine the impacts that politicization has on the healthcare choices of older Americans.

Funding

Funding for this research was provided through a grant made available by the Claude Pepper Center at Florida State University.

CRediT authorship contribution statement

Benjamin Dowd-Arrow: Conceptualization, Writing – original draft, Writing – review & editing, Methodology, Formal analysis. Amy M. Burdette: Conceptualization, Writing – original draft, Writing – review & editing, Methodology. Alyssa Smith: Writing – original draft, Writing – review & editing.

Declaration of Competing Interest

All authors declare that they have no conflicts of interest.

Appendix A. Supplementary data

Distribution of Adverse Pandemic Lifestyles and Vaccination Status by Age Cohort

Data availability

Data will be made available on request.