Abstract
Objectives. To determine the prevalence of COVID-19–related discrimination among major US racial/ethnic groups and estimate associations between discrimination, race/ethnicity, and other sociodemographic characteristics.
Methods. We conducted a nationally representative online survey of 5500 American Indian/Alaska Native, Asian, Black/African American, Hawaiian/Pacific Islander, Latino (English and Spanish speaking), White, and multiracial adults from December 2020 to February 2021. Associations between sociodemographic characteristics and COVID-19–related discrimination were estimated via multinomial logistic regression.
Results. A total of 22.1% of the participants reported experiencing discriminatory behaviors, and 42.7% reported that people acted afraid of them. All racial/ethnic minorities were more likely than White adults to experience COVID-19–related discrimination, with Asian and American Indian/Alaska Native adults being most likely to experience such discrimination (discriminatory behaviors: adjusted odd ratio [AOR] = 2.59; 95% confidence interval [CI] = 1.73, 3.89; and AOR = 2.67; 95% CI = 1.76, 4.04; people acting afraid: AOR = 1.54; 95% CI = 1.15, 2.07; and AOR = 1.84; 95% CI = 1.34, 2.51). Limited English proficiency, lower education, lower income, and residing in a big city or the East South Central census division also increased the prevalence of discrimination.
Conclusions. COVID-19–related discrimination is common, and it appears that the pandemic has exacerbated preexisting resentment against racial/ethnic minorities and marginalized communities. Efforts are needed to minimize and discredit racially driven language and discrimination around COVID-19 and future epidemics. (Am J Public Health. 2022;112(3):453–466. https://doi.org/10.2105/AJPH.2021.306594)
Historically, infectious disease outbreaks have often been accompanied by discrimination, stigma, and xenophobia.1,2 How these diseases are named and discussed can have a major impact on subsequent discrimination. Because of this, both the World Health Organization and the Centers for Disease Control and Prevention have guidelines that recommend against attaching locations or ethnicity to a disease to minimize backlash against members (and perceived members) of the identified community.3,4 Despite these recommendations, some public officials in the United States repeatedly referred to COVID-19 as the “Chinese virus” or “Wuhan virus” instead of COVID-19,3,5 and reports of racist and xenophobic incidents directed toward those perceived to be Chinese or of Asian descent have increased.6–9 Because of the broad scope of systemic racism in the United States, we hypothesized that attributing blame for the pandemic could also extend to other minority and marginalized communities.
To date, 4 studies to our knowledge have attempted to measure the prevalence of COVID-19–related discrimination in the United States. However, 2 focused on Asians only10,11; 1 was restricted to Asian, Black, Latino, and White individuals12; and 1 combined several racial/ethnic minority groups into a single category (“other race”).13 Thus, discrimination among other racial/ethnic minority groups (e.g., American Indian/Alaska Native) has yet to be assessed, and a comparison of all groups in one study is needed. Also, although other sociodemographic characteristics, such as age, household income, and immigration status, have been linked to a higher prevalence of discrimination,13 additional research is needed.
Thus, the goals of this study were to (1) estimate the prevalence of COVID-19– related discrimination among all major US racial/ethnic groups (as defined by the US Bureau of the Census), (2) estimate the association between COVID-19– related discrimination and race/ethnicity after adjusting for sociodemographic characteristics, and (3) identify other sociodemographic characteristics associated with COVID-19–related discrimination among a nationally representative and diverse sample of US adults.
METHODS
The COVID-19’s Unequal Racial Burden (CURB) survey was administered by YouGov, a consumer research firm based in Palo Alto, California, that uses a proprietary, opt-in survey panel composed of more than 1.8 million US residents to conduct nationally representative online surveys. Panel members are recruited through a variety of methods to ensure diversity, including Web advertising, permission-based e-mail campaigns, partner-sponsored solicitations, telephone-to-Web recruitment, and mail-to-Web recruitment. Participants receive incentives through a loyalty program to complete individual surveys.
To obtain nationally representative estimates, YouGov randomly matches eligible panel members with matching demographic characteristics (matched sample) to a theoretical cohort (target sample) identified by sampling nationally representative data. The target sample for the CURB study was drawn from the 2018 American Community Survey 1-year sample and included 1000 Asian, 1000 Black/African American, 1000 Latino (including 500 Spanish-speaking), 1000 White, 500 American Indian/Alaska Native, 500 Hawaiian/Pacific Islander, and 500 multiracial adults 18 years or older (overall n = 5500). A proximity matching method was then used to match YouGov panel members (matched sample) to the target sample according to race/ethnicity, gender, age, education, and language preference (Latino sample only). YouGov invited matched panel members to participate via e-mail until sample quotas were met for each racial/ethnic group. Online surveys were completed between December 8, 2020, and February 17, 2021.
After survey completion, survey weights were calculated. Briefly, within each racial/ethnic group, the matched sample and American Community Survey 1-year data were combined and multivariable logistic regression adjusting for age, gender, education, and region was used to estimate probability for inclusion in the study. Probabilities were then grouped into deciles and poststratified on gender, age, education, and region to produce a final weight for each participant. Ultimately, this combination of matching and weighting allowed for the generation of national estimates.14,15 Weights generating nationally representative populations within each racial/ethnic group were used in this analysis (e.g., Asian participants represented all Asian adults in the United States). YouGov has been used previously to conduct nationally representative survey-based research.16–18
The CURB survey was designed to assess the social, behavioral, and economic effects of the COVID-19 pandemic among diverse populations, including experiences of discrimination. The survey was created in English, translated into Spanish by an American Translators Association certified translator, and finalized by 4 bilingual/bicultural researchers via team reconciliation19 and decentering methods.20
Dependent Variable
Four items assessed experiences of COVID-19–related discrimination. Three were adapted from the Everyday Discrimination Scale: (1) people acting afraid of you, (2) being called names or insulted, and (3) being threatened or harassed.21 On the basis of news reports that people of Chinese descent were hearing racist comments from people thinking they were the cause of COVID-19, we created a new item that asked participants how often they heard racist comments because people thought they belonged to a group that contracts COVID-19 more often. For all 4 items, we asked how often participants had experienced the specific type of discrimination (e.g., people acting afraid of you) “because they think you might have COVID-19” using a 4-level response scale (1 = never, 2 = rarely, 3 = sometimes, 4 = always). Complete data for all 4 items were available for 5494 participants (more than 99%).
According to a multitrait scaling analysis,22 the people acting afraid of you item was not highly correlated with the other 3 items (r = 0.49). Thus, we developed 2 measures of COVID-19–related discrimination: a single-item measure (people acted afraid of you) and a 3-item scale (discriminatory behaviors). The discriminatory behaviors scale was scored as the mean of nonmissing values; the internal-consistency reliability for the total sample was 0.88, with similar results in each racial/ethnic group. The continuous scale was then categorized according to the original response scale: never (score of 1), rarely (scores from above 1 to 2), sometimes (scores from above 2 to 3), and always (scores above 3). The people acted afraid of you measure ranged from 1 to 4 (original response scale).
The 2 measures were then categorized into never, rarely, and sometimes/always; sometimes and always were combined into a single category owing to the small percentage of participants reporting “always” experiencing discrimination (4.2% and 1.7%, respectively). In sensitivity analyses, a composite “any” discrimination (sometimes/always or rarely) was also assessed. (For a full description of the survey questions and analysis metrics, see Table A, available as a supplement to the online version of this article at http://www.ajph.org.)
Independent Variables
All eligible panel members were asked “Which one of the following would you say best represents your race/ethnicity?” Response options were Latino/a/x or Hispanic, American Indian or Alaska Native, Asian, Black or African American, Pacific Islander, White, and multiracial. Among Asian participants, we included a question on national origin (Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, or other Asian).
Self-reported sociodemographic characteristics included age (categorized as 18–34, 35–49, 50–64, 65 years or older), gender (male, female, transgender or nonbinary), immigration status (US-born citizen, foreign-born citizen or legal resident, undocumented), English speaking proficiency (limited vs not limited), highest education level (less than high school, high school, more than high school), employment status (employed vs not employed), family annual income (< $20 000, $20 000–$59 999, $60 000–$99 999, ≥ $100 000), census division, and urbanicity (big city, smaller city, suburban, small town, rural). Limited English proficiency was defined as being able to speak English “not at all,” “poorly,” or “fairly well.” Amounts of missing data were minimal for all variables other than family annual income (659 [unweighted] participants selected “prefer not to say”).
Statistical Analyses
Descriptive statistics were used to estimate the prevalence of COVID-19– related discrimination across racial/ethnic groups. Multinomial logistic regression was used to estimate the independent association between race/ethnicity, sociodemographic characteristics, and the prevalence (sometimes/always or rarely vs never) of discriminatory behaviors and people acting afraid of the participant. Models included race/ethnicity, age, gender, immigration status, limited English proficiency, educational attainment, employment status, family annual income, census division, and urbanicity.
We conducted a secondary analysis restricted to Asian respondents to assess whether demographic characteristics associated with COVID-19–related discrimination differed within Asian subpopulations. Multinomial logistic regression models included the same variables listed earlier, with national origin included instead of race/ethnicity. As a result of the large proportion of Asian participants with college degrees, education was recategorized as high school or less, some college/vocational degree, bachelor’s degree, and postgraduate degree in this analysis. Census region was used instead of division to assess geographic differences.
As a sensitivity analysis, we used multivariable logistic regression to estimate the association between race/ethnicity and other social determinants and the odds of experiencing any discriminatory behaviors or people acting afraid (rarely/sometimes/always vs never).
We used SAS version 9.4 (SAS Inc, Cary, NC) for all of the analyses. All analyses were weighted to produce nationally representative estimates within each racial/ethnic group, and counts were rounded for interpretation.
RESULTS
There were 5804 online survey respondents (response rate: 20.0%) who were matched down to a sample of 5500 to produce the final weighted data set. Demographic characteristics, stratified by race/ethnicity, are reported in Table B (available as a supplement to the online version of this article at http://www.ajph.org).
Prevalence and Frequency of Discrimination
Overall, 22.1% of participants reported experiencing discriminatory behaviors (sometimes/always: 12.4%; rarely: 9.7%), and 42.7% reported experiences of people acting afraid of them (sometimes/always: 22.6%; rarely: 20.1%). A full breakdown is included in Table C (available as a supplement to the online version of this article at http://www.ajph.org).
The prevalence of discriminatory behaviors was highest among Asian participants (sometimes/always: 12.6%; rarely: 17.4%; Figure 1 and Table D, available as a supplement to the online version of this article at http://www.ajph.org). More than one quarter of Latino (sometimes/always: 10.6%; rarely: 16.3%) and American Indian/Alaska Native (sometimes/always: 16.8%; rarely: 9.4%) participants reported discriminatory behaviors, followed by Hawaiian/Pacific Islander (sometimes/always: 10.8%; rarely: 12.0%), Black/African American (sometimes/always: 9.1%; rarely: 11.4%), and multiracial (sometimes/always: 3.8%; rarely: 14.6%) adults; only 10% of White participants reported experiencing discriminatory behaviors (sometimes/always: 5.4%; rarely: 5.0%).
FIGURE 1—
Prevalence of Self-Reported Experiences of COVID-19–Related Discrimination Experiences Including (a) Discriminatory Behaviors and (b) People Acted Afraid of You Thinking You May Have COVID-19 or Belong to a Racial/Ethnic Group That Gets COVID-19 More Often: United States, December 2020–February 2021
Notes. Discriminatory behaviors were defined as being called names or insulted, being threatened or harassed, and racist comments. Percentages are weighted to be nationally representative within each racial/ethnic group.
Similar trends were seen within the individual items in the discriminatory behaviors scale (Figure A, available as a supplement to the online version of this article at http://www.ajph.org). Reports of people acting afraid were common, with half of participants reporting that such discrimination occurred sometimes/always (Figure 1b). The prevalence of people acting afraid was highest among Hawaiian/Pacific Islander (sometimes/always: 27.7%; rarely: 21.2%), Latino (sometimes/always: 29.5%; rarely: 18.4%), and American Indian/Alaska Native (sometimes/always: 25.5%; rarely: 21.7%) adults, although the prevalence was similarly high among other racial/ethnic minority groups (Asian: sometimes/always: 22.5%; rarely: 21.5%; Black/African American: sometimes/always: 21.8%; rarely: 17.3%; multiracial: sometimes/always: 18.8%; rarely: 25.4%).
Race/Ethnicity and Discrimination
After adjustment, all racial/ethnic minority groups were substantially more likely to experience discriminatory behaviors (rarely vs none; adjusted odds ratios [AORs] = 1.86–3.61), but only American Indian/Alaska Native and Asian participants were significantly more likely than White adults to report sometimes/always experiencing discriminatory behaviors (AOR = 2.67; 95% confidence interval [CI] = 1.76, 4.04; and AOR = 2.59; 95% CI = 1.73, 3.89; Table 1).
TABLE 1—
Prevalence of Self-Reported Experiences of COVID-19–Related Discriminatory Behaviors, Stratified by Sociodemographic Characteristics, and Adjusted, Independent Associations With COVID-19–Related Discrimination: United States, December 2020–February 2021
| Characteristic | Rarely | Sometimes/Always | ||
| No. (%) | OR (95% CI)a | No. (%) | OR (95% CI)a | |
| Race/ethnicity | ||||
| American Indian/Alaska Native | 47 (9.4) | 1.86 (1.19, 2.91) | 84 (16.8) | 2.67 (1.76, 4.04) |
| Asian | 174 (17.4) | 3.61 (2.45, 5.31) | 126 (12.6) | 2.59 (1.73, 3.89) |
| Black/African American | 114 (11.4) | 1.97 (1.35, 2.88) | 91 (9.1) | 1.24 (0.84, 1.85) |
| Latino | 163 (16.3) | 2.20 (1.48, 3.29) | 106 (10.6) | 1.13 (0.74, 1.74) |
| English speaking | 58 (11.7) | . . . | 29 (5.8) | . . . |
| Spanish speaking | 104 (20.7) | . . . | 78 (15.5) | . . . |
| Hawaiian/Pacific Islander | 60 (12.0) | 1.99 (1.27, 3.12) | 54 (10.8) | 1.39 (0.87, 2.24) |
| White | 50 (5.0) | 1 (Ref) | 54 (5.4) | 1 (Ref) |
| Multiracial | 73 (14.6) | 2.13 (1.40, 3.23) | 19 (3.8) | 0.44 (0.24, 0.82) |
| Age group, y | ||||
| 18–34 | 341 (17.4) | 1.54 (1.25, 1.91) | 264 (13.5) | 1.50 (1.18, 1.90) |
| 35–49 | 189 (13.0) | 1 (Ref) | 163 (11.3) | 1 (Ref) |
| 50–64 | 127 (9.6) | 0.70 (0.54, 0.91) | 88 (6.7) | 0.62 (0.46, 0.84) |
| ≥ 65 | 24 (3.1) | 0.27 (0.17, 0.43) | 18 (2.4) | 0.14 (0.07, 0.27) |
| Gender | ||||
| Male | 336 (13.0) | 1 (Ref) | 302 (11.7) | 1 (Ref) |
| Female | 316 (11.4) | 0.82 (0.68, 0.98) | 198 (7.1) | 0.53 (0.43, 0.66) |
| Transgender or nonbinaryb | 27 (20.5) | 1.80 (1.09, 2.98) | 34 (25.6) | 1.99 (1.16, 3.40) |
| Immigration status | ||||
| US-born citizen | 476 (11.1) | 1 (Ref) | 377 (8.8) | 1 (Ref) |
| Foreign-born citizen/legal resident | 149 (15.8) | 1.03 (0.80, 1.33) | 123 (12.9) | 0.96 (0.72, 1.29) |
| Undocumented | 54 (19.8) | 0.86 (0.57, 1.30) | 31 (11.5) | 0.47 (0.28, 0.77) |
| English proficiencyc | ||||
| Limited | 127 (20.5) | 2.14 (1.59, 2.88) | 141 (22.8) | 4.06 (3.02, 5.47) |
| Not limited | 554 (11.3) | 1 (Ref) | 393 (8.1) | 1 (Ref) |
| Highest educational level | ||||
| < high school | 89 (17.8) | 1.38 (1.01, 1.89) | 80 (16.1) | 1.77 (1.26, 2.49) |
| High school or equivalent | 190 (10.6) | 0.87 (0.70, 1.08) | 197 (11.0) | 1.09 (0.85, 1.38) |
| > high schoold | 402 (12.5) | 1 (Ref) | 257 (8.0) | 1 (Ref) |
| Employment status | ||||
| Employed | 327 (13.4) | 0.97 (0.80, 1.18) | 271 (11.1) | 1.26 (1.01, 1.58) |
| Not employede | 353 (11.5) | 1 (Ref) | 263 (8.6) | 1 (Ref) |
| Family annual income, $f | ||||
| < 20 000 | 163 (14.9) | 1.36 (0.98, 1.89) | 138 (12.6) | 2.02 (1.35, 3.04) |
| 20 000–59 999 | 242 (12.6) | 1.25 (0.94, 1.67) | 205 (10.6) | 1.83 (1.27, 2.63) |
| 60 000–99 999 | 110 (11.3) | 1.07 (0.78, 1.46) | 91 (9.4) | 1.66 (1.12, 2.45) |
| ≥ 100 000 | 88 (10.8) | 1 (Ref) | 46 (5.6) | 1 (Ref) |
| Prefer not to sayg | 77 (11.1) | . . . | 55 (7.9) | . . . |
| Census division | ||||
| New England | 17 (10.4) | 0.81 (0.44, 1.48) | 25 (15.7) | 1.33 (0.72, 2.45) |
| Middle Atlantic | 81 (13.0) | 1.12 (0.79, 1.58) | 61 (9.8) | 1.11 (0.75, 1.66) |
| East North Central | 59 (10.7) | 1.02 (0.70, 1.47) | 43 (7.8) | 0.97 (0.63, 1.51) |
| West North Central | 26 (11.4) | 1.15 (0.69, 1.92) | 22 (9.7) | 1.24 (0.71, 2.16) |
| South Atlantic | 121 (11.3) | 1 (Ref) | 79 (7.4) | 1 (Ref) |
| East South Central | 39 (15.6) | 1.99 (1.28, 3.07) | 41 (16.3) | 2.43 (1.50, 3.92) |
| West South Central | 92 (13.7) | 1.12 (0.81, 1.55) | 63 (9.5) | 1.01 (0.69, 1.47) |
| Mountain | 69 (12.5) | 1.03 (0.71, 1.48) | 64 (11.6) | 1.13 (0.75, 1.71) |
| Pacific | 177 (12.7) | 0.92 (0.69, 1.23) | 135 (9.7) | 0.82 (0.58, 1.15) |
| Urbanicityh | ||||
| Big city | 193 (13.4) | 1.01 (0.80, 1.29) | 174 (12.1) | 1.47 (1.11, 1.94) |
| Smaller city | 148 (14.5) | 1.05 (0.81, 1.36) | 105 (10.3) | 1.05 (0.77, 1.44) |
| Suburban area | 196 (11.8) | 1 (Ref) | 132 (7.9) | 1 (Ref) |
| Small town | 73 (11.1) | 0.91 (0.65, 1.26) | 63 (9.6) | 0.98 (0.67, 1.43) |
| Rural area | 67 (11.8) | 1.15 (0.82, 1.62) | 56 (10.0) | 1.44 (0.98, 2.10) |
Note. CI = confidence interval; OR = odds ratio. Discriminatory behavior includes being called names, being threatened/harassed, and hearing racist comments because people think you might have COVID-19. Data are weighted to be nationally representative within each racial/ethnic group. The study sample size was 5500.
aModeled with multinomial logistic regression (sometimes/always, rarely, and never [reference]); all ORs adjusted for all other variables in the table.
bNonbinary includes individuals who reported being nonbinary, gender fluid, gender queer, “other,” and no gender.
cLimited English proficiency was defined as speaking English “not at all,” “poorly,” or “fairly well.”
dIncludes some college/vocational school, bachelor’s degree, master’s degree, and doctoral or postgraduate education.
eNot employed includes temporarily laid off, unemployed, retired, permanently disabled, taking care of home or family, student, and other.
fCollected at enrollment into panel and updated every 6 months.
gA total of 659 (unweighted) participants selected “prefer not to say” and were dropped from the model; when household income was not included in the analysis, similar effect estimates for the other covariates were seen (data not shown).
hA total of 125 participants (unweighted) did not provide information on residential urbanicity and were not included in the analysis.
Fewer racial/ethnic differences were seen across the people acted afraid of you item. Relative to White adults, only American Indian/Alaska Native (rarely: AOR = 1.40; 95% CI = 1.03, 1.91; sometimes/always: AOR = 1.84; 95% CI = 1.34, 2.51) and Hawaiian/Pacific Islander (rarely: AOR = 1.42; 95% CI = 1.02, 1.97; sometimes/always: AOR = 1.90; 95% CI = 1.37, 2.64) adults were significantly more likely to report incidents in which people acted afraid of them across both frequency levels (Table 2). Asian adults appeared to also be more likely to report incidents of people acting afraid of them at both frequencies, but confidence intervals were wide (rarely: AOR = 1.22; 95% CI = 0.92, 1.62; sometimes/always: OR = 1.54; 95% CI = 1.15, 2.07). Latino participants were more likely to report frequent (sometimes/always) incidents of people acting afraid of them (OR = 1.45; 95% CI = 1.08, 1.96), and multiracial participants were more likely to report rare incidents of people acting afraid of them (AOR = 1.44; 95% CI = 1.07, 1.95). No differences were seen between Black/African American and White adults.
TABLE 2—
Prevalence of Participants’ Self-Reported Experiences of People Acting Afraid of Them Because of Suspected COVID-19 Infection, Stratified by Sociodemographic Characteristics, and Adjusted, Independent Associations With COVID-19–Related Discrimination: United States, December 2020–February 2021
| Characteristic | Rarely | Sometimes/Always | ||
| No. (%) | OR (95% CI)a | No. (%) | OR (95% CI)a | |
| Race/ethnicity | ||||
| American Indian/Alaska Native | 108 (21.7) | 1.40 (1.03, 1.91) | 128 (25.5) | 1.84 (1.34, 2.51) |
| Asian | 215 (21.5) | 1.22 (0.92, 1.62) | 225 (22.5) | 1.54 (1.15, 2.07) |
| Black/African American | 173 (17.3) | 0.93 (0.70, 1.22) | 218 (21.8) | 1.18 (0.90, 1.56) |
| Latino | 184 (18.4) | 0.95 (0.71, 1.28) | 295 (29.5) | 1.45 (1.08, 1.96) |
| English speaking | 94 (19.0) | . . . | 98 (19.8) | . . . |
| Spanish speaking | 90 (17.9) | . . . | 197 (39.2) | . . . |
| Hawaiian/Pacific Islander | 106 (21.2) | 1.42 (1.02, 1.97) | 138 (27.7) | 1.90 (1.37, 2.64) |
| White | 190 (19.0) | 1 (Ref) | 144 (14.4) | 1 (Ref) |
| Multiracial | 127 (25.4) | 1.44 (1.07, 1.95) | 94 (18.8) | 1.17 (0.84, 1.64) |
| Age group, y | ||||
| 18–34 | 454 (23.2) | 1.27 (1.05, 1.53) | 514 (26.2) | 1.26 (1.05, 1.51) |
| 35–49 | 298 (20.6) | 1 (Ref) | 366 (25.3) | 1 (Ref) |
| 50–64 | 260 (19.6) | 0.95 (0.77, 1.18) | 277 (20.9) | 0.91 (0.74, 1.12) |
| ≥ 65 | 92 (12.0) | 0.44 (0.32, 0.59) | 85 (11.0) | 0.39 (0.29, 0.52) |
| Gender | ||||
| Male | 575 (22.2) | 1 (Ref) | 608 (23.5) | 1 (Ref) |
| Female | 500 (18.0) | 0.76 (0.65, 0.89) | 596 (21.5) | 0.84 (0.72, 0.97) |
| Transgender or nonbinaryb | 29 (21.5) | 0.81 (0.49, 1.35) | 38 (28.9) | 0.89 (0.55, 1.44) |
| Immigration status | ||||
| US-born citizen | 860 (20.1) | 1 (Ref) | 882 (20.6) | 1 (Ref) |
| Foreign-born citizen/legal resident | 190 (20.1) | 0.97 (0.77, 1.22) | 252 (26.6) | 1.21 (0.98, 1.50) |
| Undocumented | 53 (19.3) | 1.15 (0.76, 1.73) | 107 (39.2) | 1.54 (1.08, 2.21) |
| English proficiencyc | ||||
| Limited | 132 (21.4) | 1.51 (1.14, 1.99) | 229 (37.1) | 1.68 (1.30, 2.15) |
| Not limited | 971 (19.9) | 1 (Ref) | 1013 (20.7) | 1 (Ref) |
| Highest educational level | ||||
| < high school | 110 (22.1) | 1.37 (1.03, 1.83) | 159 (31.9) | 1.53 (1.18, 2.00) |
| High school or equivalent | 307 (17.2) | 0.93 (0.77, 1.11) | 452 (25.2) | 1.11 (0.93, 1.31) |
| > high schoold | 686 (21.4) | 1 (Ref) | 632 (19.7) | 1 (Ref) |
| Employment status | ||||
| Employed | 552 (22.6) | 1.12 (0.95, 1.32) | 562 (23.0) | 1.07 (0.91, 1.26) |
| Not employede | 551 (18.0) | 1 (Ref) | 680 (22.2) | 1 (Ref) |
| Family annual income, $f | ||||
| < 20 000 | 201 (18.4) | 0.90 (0.69, 1.19) | 332 (30.3) | 1.98 (1.50, 2.62) |
| 20 000–59 999 | 396 (20.6) | 0.98 (0.78, 1.23) | 449 (23.4) | 1.58 (1.24, 2.03) |
| 60 000–99 999 | 196 (20.1) | 0.90 (0.71, 1.15) | 198 (20.3) | 1.35 (1.04, 1.77) |
| ≥ 100 000 | 191 (23.4) | 1 (Ref) | 117 (14.3) | 1 (Ref) |
| Prefer not to sayg | 119 (17.3) | . . . | 146 (21.1) | . . . |
| Census division | ||||
| New England | 32 (20.0) | 1.17 (0.74, 1.86) | 35 (22.1) | 0.93 (0.58, 1.48) |
| Middle Atlantic | 146 (23.5) | 1.33 (0.99, 1.77) | 125 (20.2) | 0.99 (0.74, 1.33) |
| East North Central | 98 (17.8) | 1.02 (0.75, 1.38) | 111 (20.0) | 0.90 (0.67, 1.21) |
| West North Central | 52 (23.4) | 1.22 (0.81, 1.84) | 48 (21.3) | 1.11 (0.74, 1.66) |
| South Atlantic | 189 (17.7) | 1 (Ref) | 224 (20.9) | 1 (Ref) |
| East South Central | 43 (17.0) | 1.19 (0.79, 1.80) | 73 (28.9) | 1.58 (1.10, 2.26) |
| West South Central | 133 (19.9) | 1.09 (0.82, 1.45) | 153 (22.9) | 0.91 (0.69, 1.19) |
| Mountain | 144 (25.9) | 1.42 (1.06, 1.91) | 133 (24.0) | 1.04 (0.77, 1.39) |
| Pacific | 266 (19.1) | 0.93 (0.73, 1.20) | 340 (24.4) | 0.91 (0.72, 1.16) |
| Urbanicityh | ||||
| Big city | 308 (21.5) | 1.24 (1.01, 1.53) | 358 (25.0) | 1.12 (0.92, 1.37) |
| Smaller city | 237 (23.2) | 1.31 (1.05, 1.64) | 240 (23.5) | 0.94 (0.75, 1.17) |
| Suburban area | 316 (19.0) | 1 (Ref) | 345 (20.8) | 1 (Ref) |
| Small town | 130 (19.8) | 1.19 (0.92, 1.55) | 154 (23.3) | 1.09 (0.84, 1.40) |
| Rural area | 89 (15.7) | 0.92 (0.68, 1.23) | 129 (22.8) | 1.02 (0.78, 1.34) |
Note. CI = confidence interval; OR = odds ratio. Participants were asked: How often have you experienced the following since the start of the pandemic: people acted as if they were afraid of you because they think you might have COVID-19. Data are weighted to be nationally representative within each racial/ethnic group. The study sample size was 5500.
aModeled with multinomial logistic regression (sometimes/always, rarely, and never [reference]); all ORs adjusted for all other variables in the table.
bNonbinary includes individuals who reported being nonbinary, gender fluid, gender queer, “other,” and no gender.
cLimited English proficiency was defined as speaking English “not at all,” “poorly,” or “fairly well.”
dIncludes some college/vocational school, bachelor’s degree, master’s degree, and doctoral or postgraduate education.
eNot employed includes temporarily laid off, unemployed, retired, permanently disabled, taking care of home or family, student, and other.
fCollected at enrollment into panel and updated every 6 months.
gA total of 659 (unweighted) participants selected “prefer not to say” and were dropped from the model; when household income was not included in the analysis, similar effect estimates for the other covariates were seen (data not shown).
hA total of 125 participants (unweighted) did not provide information on residential urbanicity and were not included in the analysis.
Other Sociodemographic Characteristics
Among the sociodemographic variables, having limited English proficiency was most strongly associated with experiencing both discriminatory behaviors (rarely: AOR = 2.14; 95% CI = 1.59, 2.88; sometimes/always: AOR = 4.06; 95% CI = 3.02, 5.47) and people acting afraid (rarely: AOR = 1.51; 95% CI = 1.14, 1.99; sometimes/always: AOR = 1.68; 95% CI = 1.30, 2.15; Tables 1 and 2). Being less than a high school graduate (relative to having more than a high school education) was also consistently associated with higher odds of experiencing discriminatory behaviors (rarely: AOR = 1.38; 95% CI = 1.01, 1.89; sometimes/always: AOR = 1.77; 95% CI = 1.26, 2.49) and people acting afraid (rarely: AOR = 1.37; 95% CI = 1.03, 1.83; sometimes/always: AOR = 1.53; 95% CI = 1.18, 2.00).
Lower annual income was associated with sometimes/always experiencing discriminatory behaviors (e.g., for < $20 000 vs ≥ $100 000, discriminatory behaviors: AOR = 2.02; 95% CI = 1.35, 3.04; people acting afraid: AOR = 1.98; 95% CI = 1.50, 2.62). Adults living in the East South Central division (Alabama, Kentucky, Mississippi, and Tennessee) were most likely to experience discriminatory behaviors (rarely: AOR = 1.99; 95% CI = 1.28, 3.07; sometimes/always: AOR = 2.43; 95% CI = 1.50, 3.92) and sometimes/always experience people acting afraid of them (AOR = 1.58; 95% CI = 1.10, 2.26); minimal differences were seen across other census divisions.
Younger age (18–34 years), being male or transgender/nonbinary, and living in a city or rural area also appeared to be associated with higher odds of experiencing COVID-19–related discrimination (discriminatory behaviors or people acting afraid). Similar trends were seen when discrimination was modeled as any versus never (Table E, available as a supplement to the online version of this article at http://www.ajph.org).
Discrimination Against Asian Adults
When we restricted our analysis to Asian participants, Vietnamese adults reported higher levels of COVID-19–related discrimination and Japanese adults reported the lowest levels (Figures B and C, available as supplements to the online version of this article at http://www.ajph.org). However, no substantial differences in COVID-19–related discrimination by national origin were seen either before or after adjustment, but confidence intervals were wide (Table 3).
TABLE 3—
Prevalence of Self-Reported Experiences of COVID-19–Related Discrimination, Stratified by Sociodemographic Characteristics, and Adjusted, Independent Associations With COVID-19–Related Discrimination Among Asian Participants: United States, December 2020–February 2021
| Characteristic | Rarely | Sometimes/Always | ||
| No. (%) | OR (95% CI)a | No. (%) | OR (95% CI)a | |
| Discriminatory behaviors | ||||
| National origin | ||||
| Asian Indian | 18 (10.3) | 0.41 (0.20, 0.80) | 25 (14.4) | 1.48 (0.76, 2.90) |
| Chinese | 58 (21.1) | 1 (Ref) | 35 (12.7) | 1 (Ref) |
| Filipino | 31 (19.3) | 1.00 (0.55, 1.81) | 21 (13.2) | 1.02 (0.46, 2.26) |
| Japanese | 11 (8.1) | 0.42 (0.19, 0.93) | 10 (7.2) | 0.83 (0.34, 2.03) |
| Korean | 17 (22.2) | 1.03 (0.50, 2.12) | 8 (10.9) | 0.88 (0.35, 2.21) |
| Vietnamese | 16 (27.2) | 1.26 (0.58, 2.73) | 10 (17.8) | 1.59 (0.62, 4.06) |
| Other Asian | 23 (18.7) | 0.87 (0.46, 1.63) | 17 (13.7) | 0.99 (0.46, 2.14) |
| Age group, y | ||||
| 18–34 | 84 (24.0) | 1.07 (0.68, 1.70) | 53 (15.2) | 0.93 (0.55, 1.57) |
| 35–49 | 55 (19.1) | 1 (Ref) | 43 (15.0) | 1 (Ref) |
| ≥ 50 | 34 (9.5) | 0.40 (0.23, 0.69) | 30 (8.3) | 0.40 (0.21, 0.75) |
| Genderb | ||||
| Male | 83 (18.3) | 1 (Ref) | 58 (12.8) | 1 (Ref) |
| Female | 87 (16.6) | 0.95 (0.64, 1.40) | 62 (11.8) | 0.82 (0.51, 1.29) |
| Immigration status | ||||
| US-born citizen | 100 (19.4) | 1 (Ref) | 67 (13.0) | 1 (Ref) |
| Foreign born | 74 (15.2) | 0.65 (0.43, 0.98) | 60 (12.3) | 0.58 (0.36, 0.94) |
| English proficiencyc | ||||
| Limited | 25 (20.2) | 2.04 (1.09, 3.83) | 30 (24.2) | 2.60 (1.40, 4.83) |
| Not limited | 149 (17.0) | 1 (Ref) | 97 (11.0) | 1 (Ref) |
| Highest educational leveld | ||||
| High school/equivalent or less | 36 (13.4) | 1.16 (0.57, 2.33) | 44 (16.5) | 1.36 (0.63, 2.92) |
| Some college/vocational school | 48 (22.2) | 2.02 (1.07, 3.83) | 27 (12.6) | 1.47 (0.69, 3.16) |
| Bachelor’s degree | 61 (20.7) | 1.68 (0.95, 2.94) | 35 (11.7) | 1.37 (0.70, 2.67) |
| Postgraduate degree | 29 (12.9) | 1 (Ref) | 21 (9.3) | 1 (Ref) |
| Employment status | ||||
| Employed | 86 (17.4) | 0.82 (0.54, 1.26) | 64 (12.9) | 1.38 (0.84, 2.29) |
| Not employede | 88 (17.3) | 1 (Ref) | 63 (12.4) | 1 (Ref) |
| Family annual income, $f | 0.00 | |||
| < 20 000 | 20 (19.2) | 1.17 (0.56, 2.42) | 19 (17.9) | 4.11 (1.74, 9.74) |
| 20 000–59 999 | 48 (17.2) | 1.04 (0.61, 1.78) | 46 (16.5) | 2.68 (1.32, 5.42) |
| 60 000–99 999 | 35 (15.2) | 0.95 (0.56, 1.60) | 27 (11.8) | 1.95 (0.97, 3.91) |
| ≥ 100 000 | 43 (17.8) | 1 (Ref) | 16 (6.7) | 1 (Ref) |
| Prefer not to sayg | 27 (19.0) | . . . | 18 (12.8) | . . . |
| Census region | ||||
| Northeast | 34 (16.0) | 0.81 (0.46, 1.41) | 38 (17.7) | 1.37 (0.75, 2.51) |
| Midwest | 18 (17.5) | 1.31 (0.69, 2.49) | 15 (14.4) | 1.48 (0.69, 3.18) |
| South | 39 (17.1) | 1.32 (0.80, 2.17) | 32 (13.9) | 1.57 (0.87, 2.84) |
| West | 81 (18.1) | 1 (Ref) | 41 (9.2) | 1 (Ref) |
| Urbanicityh | ||||
| Big city | 47 (17.8) | 0.98 (0.61, 1.57) | 32 (12.0) | 1.44 (0.81, 2.54) |
| Smaller city | 31 (19.4) | 0.99 (0.56, 1.74) | 24 (15.0) | 1.62 (0.86, 3.05) |
| Suburban area | 73 (17.5) | 1 (Ref) | 44 (10.6) | 1 (Ref) |
| Small town/rural area | 22 (14.2) | 0.93 (0.48, 1.79) | 27 (16.7) | 1.18 (0.57, 2.43) |
| People acted afraid of you | ||||
| National origin | 40 (23.1) | 0.92 (0.53, 1.60) | 42 (23.8) | 1.28 (0.73, 2.23) |
| Asian Indian | 63 (23.0) | 1 (Ref) | 50 (18.1) | 1 (Ref) |
| Chinese | 40 (24.6) | 1.76 (0.99, 3.12) | 42 (26.1) | 1.60 (0.88, 2.91) |
| Filipino | 21 (16.0) | 0.80 (0.42, 1.54) | 17 (12.6) | 0.60 (0.29, 1.22) |
| Japanese | 16 (20.5) | 0.92 (0.44, 1.91) | 20 (26.0) | 1.20 (0.59, 2.45) |
| Korean | 13 (21.7) | 1.10 (0.48, 2.50) | 20 (34.2) | 1.89 (0.89, 4.01) |
| Vietnamese | 23 (18.5) | 0.75 (0.39, 1.41) | 36 (28.8) | 1.17 (0.63, 2.15) |
| Other Asian | ||||
| Age group, y | 84 (23.9) | 0.86 (0.55, 1.35) | 92 (26.2) | 1.14 (0.73, 1.80) |
| 18–34 | 74 (25.7) | 1 (Ref) | 66 (23.1) | 1 (Ref) |
| 35–49 | 58 (16.0) | 0.64 (0.40, 1.03) | 67 (18.5) | 0.93 (0.57, 1.50) |
| ≥ 50 | ||||
| Genderb | 103 (22.7) | 1 (Ref) | 93 (20.5) | 1 (Ref) |
| Male | 110 (20.9) | 1.12 (0.78, 1.62) | 128 (24.5) | 1.40 (0.97, 2.02) |
| Female | ||||
| Immigration status | 117 (22.8) | 1 (Ref) | 105 (20.3) | 1 (Ref) |
| US-born citizen | 98 (20.3) | 0.71 (0.48, 1.04) | 120 (24.8) | 0.96 (0.65, 1.41) |
| Foreign born | ||||
| English proficiencyc | 33 (26.9) | 2.46 (1.38, 4.39) | 36 (29.2) | 1.25 (0.70, 2.24) |
| Limited | 182 (20.8) | 1 (Ref) | 189 (21.6) | 1 (Ref) |
| Not limited | ||||
| Highest educational leveld | 44 (16.4) | 0.75 (0.40, 1.40) | 74 (27.5) | 1.35 (0.74, 2.47) |
| High school/equivalent or less | 47 (21.7) | 0.97 (0.54, 1.73) | 48 (22.5) | 1.15 (0.63, 2.12) |
| Some college/vocational school | 72 (24.5) | 1.14 (0.70, 1.84) | 62 (21.0) | 1.11 (0.65, 1.89) |
| Bachelor’s degree | 52 (23.7) | 1 (Ref) | 41 (18.5) | 1 (Ref) |
| Postgraduate degree | ||||
| Employment status | 112 (22.8) | 1.10 (0.73, 1.64) | 118 (24.0) | 1.54 (1.03, 2.32) |
| Employed | 103 (20.3) | 1 (Ref) | 107 (21.1) | 1 (Ref) |
| Not employede | ||||
| Family annual income, $f | 17 (16.5) | 0.82 (0.39, 1.71) | 39 (36.5) | 3.29 (1.68, 6.44) |
| < 20 000 | 52 (18.6) | 0.83 (0.50, 1.39) | 80 (28.4) | 2.10 (1.24, 3.56) |
| 20 000–59 999 | 48 (20.9) | 0.73 (0.46, 1.17) | 38 (16.7) | 0.94 (0.55, 1.61) |
| 60 000–99 999 | 65 (27.1) | 1 (Ref) | 39 (16.1) | 1 (Ref) |
| ≥ 100 000 | 40 (23.1) | 0.92 (0.53, 1.60) | 42 (23.8) | 1.28 (0.73, 2.23) |
| Prefer not to sayg | 33 (22.7) | . . . | 30 (20.9) | . . . |
| Census region | ||||
| Northeast | 55 (25.5) | 1.49 (0.92, 2.41) | 41 (19.0) | 0.94 (0.56, 1.59) |
| Midwest | 21 (20.3) | 1.07 (0.57, 2.02) | 29 (27.5) | 1.44 (0.80, 2.59) |
| South | 47 (20.5) | 1.61 (0.99, 2.63) | 65 (28.2) | 2.05 (1.30, 3.24) |
| West | 92 (20.5) | 1 (Ref) | 90 (20.1) | 1 (Ref) |
| Urbanicityh | ||||
| Big city | 70 (26.4) | 1.25 (0.81, 1.94) | 56 (21.0) | 1.10 (0.71, 1.73) |
| Smaller city | 30 (19.0) | 0.83 (0.48, 1.44) | 36 (22.7) | 0.76 (0.44, 1.30) |
| Suburban area | 90 (21.7) | 1 (Ref) | 95 (22.8) | 1 (Ref) |
| Small town/rural | 25 (15.6) | 0.92 (0.50, 1.69) | 38 (24.1) | 1.02 (0.58, 1.79) |
Note. CI = confidence interval; OR = odds ratio. Discriminatory behavior includes being called names, being threatened/harassed, and hearing racist comments because people think you might have COVID-19. Also, participants were asked: How often have you experienced the following since the start of the pandemic—people acted as if they were afraid of you because they think you might have COVID-19.
aModeled with multinomial logistic regression (sometimes/always, rarely, and never [reference]); all ORs adjusted for all other variables in the table.
bNonbinary and transgender participants were excluded from all analyses.
cLimited English proficiency was defined as speaking English “not at all,” “poorly,” or “fairly well.”
dBecause of the large proportion of Asian participants with college degrees, education was categorized as high school graduate or less (< high school n = 40), some college/vocational degree, bachelor’s degree, and postgraduate degree.
eNot employed included temporarily laid off, unemployed, retired, permanently disabled, taking care of home or family, student, and other.
fCollected at enrollment into the panel and updated every 6 months.
gA total of 141 (unweighted) participants selected “prefer not to say” and were dropped from the model; when household income was not included in the analysis, similar effect estimates for the other covariates were seen (data not shown).
hOne participant (unweighted) did not provide information on residential urbanicity and was not included in the analysis.
Among Asians, limited English proficiency was strongly and consistently associated with experiencing discriminatory behaviors (rarely: AOR = 2.04; 95% CI = 1.09, 3.83; sometimes/always: AOR = 2.60; 95% CI = 1.40, 4.83) and people acting afraid (rarely: AOR = 2.46; 95% CI = 1.38, 4.39; sometimes/always: AOR = 1.25; 95% CI = 0.70, 2.24). Lower household income was also associated with experiencing discrimination sometimes/always (Table 3). Being employed and living in the South were associated with sometimes/always experiencing people acting afraid (AOR = 1.54; 95% CI = 1.03, 2.32; and AOR = 2.05; 95% CI = 1.30, 3.24, respectively). No differences in discrimination were seen across gender, educational level, or urbanicity.
DISCUSSION
In a nationally representative online survey conducted from December 2020 to February 2021 that included adults from the 6 major US racial/ethnic minority groups (as defined by the US Census Bureau) and White adults, we found that all racial/ethnic minorities experienced higher levels of COVID-19– related discrimination than White adults, with American Indian/Alaska Native, Asian, Hawaiian/Pacific Islander, and Latino adults having the highest prevalence. Despite news and social media reports on targeting of Chinese individuals, similar COVID-19–related discrimination trends were seen across all Asian adults regardless of national origin.
In comparison with White adults, all racial/ethnic minorities were more likely to report that people acted afraid of them because of suspected COVID-19 infection. Having limited English proficiency, less than a high school education, an annual income below $60 000, and living in a big city, rural community, or Alabama, Kentucky, Mississippi, or Tennessee were also associated with experiencing increased discrimination. To the best of our knowledge, this is the largest, most racially diverse, and most recent assessment of COVID-19–related discrimination in the United States.
We found that experiencing COVID-19– related discrimination was common among Asian adults (discriminatory behaviors: 30%; people acting afraid: 44%) and that half of those experiencing discrimination reported that the discrimination occurred sometimes or always. These rates are substantially higher than estimates obtained earlier in the pandemic, suggesting that COVID-19–related discrimination has not improved over time. In surveys conducted in March and April 2020, the prevalence of COVID-19–related discrimination was 18% and 22%, respectively, among Asian adults.13 Also, Liu et al. found that those who reported COVID-19–related discrimination in March were more likely to report such discrimination in April, supporting our finding that discrimination occurs repeatedly.13
In May and June 2020, 13% of Bhutanese and Burmese refugees reported being threatened or harassed, and 28% reported feeling that others were afraid of them owing to COVID-19.11 In a survey conducted from March to May 2020, Cheah et al. found that 50% of adults reported at least one incident of in-person COVID-19–related discrimination, but they adapted the Racial and Ethnic Microaggressions Scale, which includes less severe forms of discrimination (e.g., “people were unfriendly or unwelcoming”).10 It has also been estimated that 42% of adults living in the United States are extremely likely to engage in anti-Asian behaviors during the pandemic,23 and more than 2800 incidences of anti-Asian hate were reported in 2020 alone.7 These estimates of the prevalence and frequency of COVID-19–related discrimination targeted at Asian individuals, including our own, represent a call for action.
The prevalence of discriminatory behaviors was higher among all racial/ethnic minorities than among White adults, and most (American Indian/Alaska Native, Asian, Hawaiian/Pacific Islander, multiracial) were also more likely to report that people acted afraid of them. To the best of our knowledge, we are the first to report this substantial level of COVID-19–related discrimination toward American Indian/Alaska Native adults. In fact, American Indian/Alaska Native adults were just as likely to face frequent (sometimes/always) discriminatory behaviors as Asian adults (OR = 2.67 vs OR = 2.59) and were potentially more likely to report that people acted afraid of them (OR = 1.84 vs OR = 1.54). Hawaiian/Pacific Islanders were also at higher risk of reporting frequent incidents of people acting afraid of them (OR = 1.90).
Given these findings, it appears that the COVID-19 pandemic has exacerbated preexisting resentment against racial/ethnic minority groups in the United States. Future studies and public health efforts focused on COVID-19–related discrimination should explicitly include all major racial/ethnic groups, as most appear to be at equally high risk as Asian adults but have thus far been largely ignored in antidiscrimination efforts.
Both overall and among Asian participants, we found that limited English proficiency, lower household income, and lower education were the strongest predictors of reporting sometimes/always experiencing discriminatory behaviors and people acting afraid, even after adjustment for race/ethnicity. Liu et al. also assessed other predictors of discrimination during COVID-19.13 Although they did not include English proficiency as a predictor, they did find that immigrants (first or second generation vs nonimmigrant) were more likely to experience discrimination, as were those with lower household incomes. Interestingly, both our study and the Liu et al.13 investigation showed that older adults were less likely to experience COVID-19–related discrimination. Given the recent reporting of violence targeting older Asian adults,7 additional research is needed to assess whether older adults are truly less likely to experience COVID-19–related discrimination or they are less likely to report it.
Public health and media messaging must aim to reduce racism and xenophobia during COVID-19 and future pandemics. One recent study showed that COVID-19 messaging that focused on China or Chinese cultural practices as the origin of the pandemic (i.e., food markets) led to high levels of xenophobia and anti-Chinese sentiments, whereas information that did not mention China did not increase these negative beliefs.24 In a recent analysis of Twitter data, half of the tweets that referred to COVID-19 as the “Chinese virus” had anti-Asian sentiments; moreover, anti-Asian sentiments associated with COVID-19 on Twitter increased by more than 700% after the first “Chinese virus” reference by former president Donald J. Trump in March 2020.3 A similar increase in implicit bias toward Asian Americans was seen in the United States after the first “Chinese virus” reference by the former president.25 These findings provide further evidence that the language used by individuals in positions of influence (online and offline) can have a substantial impact on racism and xenophobia during public health emergencies.26
Limitations
This study involved some limitations. First, the survey was administered online, and individuals with limited Internet access or familiarity with technology were less likely to be included. Although we did match and weight participants to obtain a nationally representative sample, it is possible that some selection bias existed. In our analysis, we found that both lower income levels and lower educational levels were associated with higher rates of discrimination, suggesting that we may be underestimating the burden of COVID-19–related discrimination in the United States. Second, the survey was administered only in English and Spanish (Latino participants only), and thus non-Latino individuals with limited English reading proficiency were more likely to be excluded. Limited English proficiency was the strongest predictor in our analysis, again suggesting that we may be underestimating the burden of COVID-19–related discrimination.
Third, although our survey was designed to be representative of the major US racial/ethnic groups, stratified results for Asians by national origin may not be representative, and sample sizes were small in some groups. Finally, discrimination measures were based on individuals’ perceptions of the motivation behind others’ behaviors, and we did not ask about the perpetrators of discriminatory behaviors. Ethnographic approaches would enable a more nuanced understanding of these encounters. The extent to which our findings reflect actual discriminatory acts based on systemic racism, awareness or misperceptions of higher COVID-19 infection risks among certain racial/ethnic groups, a desire to protect oneself, and other factors needs to be investigated.
Public Health Implications
To our knowledge, this is the largest and most diverse survey on COVID-19– related discrimination in the United States to date, and it provides a critical update. Overall, in this nationally representative survey of US adults, we found that COVID-19–related discrimination was more prevalent than indicated in prior estimates and that all racial/ethnic minorities are at risk, with American Indian/Alaska Native, Asian, and Hawaiian/Pacific Islander adults experiencing the most discrimination. Limited English proficiency, lower education, and lower income were also significant predictors of discrimination. It appears that the COVID-19 pandemic has exacerbated preexisting resentment against racial/ethnic minorities, immigrants, and other marginalized communities. Moving forward, better efforts will be needed, especially from public officials, to minimize racially driven language around COVID-19 and future pandemics to stop targeted discrimination and xenophobia.
ACKNOWLEDGMENTS
This research was supported by the Division of Intramural Research, National Institute on Minority Health and Health Disparities, National Institutes of Health. A. L. Stewart was supported by the National Institute on Aging, National Institutes of Health (grant 2 P30 AG015272).
Note. The opinions expressed in this article are those of the authors and do not reflect the views of the National Institutes of Health, the Department of Health and Human Services, or the United States government.
CONFLICTS OF INTEREST
The authors declare no conflicts of interest.
HUMAN PARTICIPANT PROTECTION
The institutional review board of the National Institutes of Health determined that this study did not qualify as human subjects research because data were deidentified.
Footnotes
See also Crawford and Lewis, p. 354.
REFERENCES
- 1.Saeed F, Mihan R, Mousavi SZ, et al. A narrative review of stigma related to infectious disease outbreaks: what can be learned in the face of the COVID-19 pandemic? Front Psychiatry. 2020;11:565919. doi: 10.3389/fpsyt.2020.565919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Clissold E, Nylander D, Watson C, Ventriglio A. Pandemics and prejudice. Int J Soc Psychiatry. 2020;66(5):421–423. doi: 10.1177/0020764020937873. [DOI] [PubMed] [Google Scholar]
- 3.Hswen Y, Xu X, Hing A, Hawkins JB, Brownstein JS, Gee GC. Association of “#covid19” versus “#chinesevirus” with Anti-Asian sentiments on Twitter: March 9–23, 2020. Am J Public Health. 2021;111(5):956–964. doi: 10.2105/AJPH.2021.306154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.World Health Organization https://www.who.int/docs/default-source/coronaviruse/covid19-stigma-guide.pdf2021.
- 5.Villa S, Jaramillo E, Mangioni D, Bandera A, Gori A, Raviglione MC. Stigma at the time of the COVID-19 pandemic. Clin Microbiol Infect. 2020;26(11):1450–1452. doi: 10.1016/j.cmi.2020.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tessler H, Choi M, Kao G. The anxiety of being Asian American: hate crimes and negative biases during the COVID-19 pandemic. Am J Crim Justice. 2020;45(4):1–11. doi: 10.1007/s12103-020-09541-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Stop AAPI Hate https://secureservercdn.net/104.238.69.231/a1w.90d.myftpupload.com/wp-content/uploads/2021/02/Press-Statement-re_-Bay-Area-Elderly-Incidents-2.9.2021-1.pdf2021.
- 8.Chen JA, Zhang E, Liu CH. Potential impact of COVID-19-related racial discrimination on the health of Asian Americans. Am J Public Health. 2020;110(11):1624–1627. doi: 10.2105/AJPH.2020.305858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wang D, Gee GC, Bahiru E, Yang EH, Hsu JJ. Asian-Americans and Pacific Islanders in COVID-19: emerging disparities amid discrimination. J Gen Intern Med. 2020;35(12):3685–3688. doi: 10.1007/s11606-020-06264-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cheah CSL, Wang C, Ren H, Zong X, Cho HS, Xue X. COVID-19 racism and mental health in Chinese American families. Pediatrics. 2020;146(5):e2020021816. doi: 10.1542/peds.2020-021816. [DOI] [PubMed] [Google Scholar]
- 11.Zhang M, Gurung A, Anglewicz P, Baniya K, Yun K. Discrimination and stress among Asian refugee populations during the COVID-19 pandemic: evidence from Bhutanese and Burmese refugees in the USA. J Racial Ethn Health Disparities. 2021 doi: 10.1007/s40615-021-00992-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pew Research Center https://www.pewresearch.org/social-trends/2020/07/01/many-black-and-asian-americans-say-they-have-experienced-discrimination-amid-the-covid-19-outbreak2021.
- 13.Liu Y, Finch BK, Brenneke SG, Thomas K, Le PD. Perceived discrimination and mental distress amid the COVID-19 pandemic: evidence from the Understanding America Study. Am J Prev Med. 2020;59(4):481–492. doi: 10.1016/j.amepre.2020.06.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.YouGov https://smpa.gwu.edu/sites/g/files/zaxdzs2046/f/downloads/YG_Matching_and_weighting_basic_description.pdf2021.
- 15.Kennedy C, Mercer A, Keeter S, Hatley N, McGeeney K, Gimenez A.https://www.pewresearch.org/methods/2016/05/02/evaluating-online-nonprobability-surveys2021.
- 16.Dixon G, Garrett K, Susmann M, Bushman BJ. Public opinion perceptions, private support, and public actions of US adults regarding gun safety policy. JAMA Netw Open. 2020;3(12):e2029571. doi: 10.1001/jamanetworkopen.2020.29571. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kohl PA, Brossard D, Scheufele DA, Xenos MA. Public views about editing genes in wildlife for conservation. Conserv Biol. 2019;33(6):1286–1295. doi: 10.1111/cobi.13310. [DOI] [PubMed] [Google Scholar]
- 18.Twyman J. Getting it right: YouGov and online survey research in Britain. J Elections Public Opin Parties. 2008;18(4):343–354. doi: 10.1080/17457280802305169. [DOI] [Google Scholar]
- 19.Pan Y, Del la Puente M.https://www.census.gov/srd/papers/pdf/rsm2005-06.pdf2021.
- 20.Brislin RW, Lonner WJ, Thorndike RM.Cross-cultural Research Methods. New York, NY: John Wiley & Sons Inc; 1973. [Google Scholar]
- 21.Williams DR, Yan Y, Jackson JS, Anderson NB. Racial differences in physical and mental health: socio-economic status, stress and discrimination. J Health Psychol. 1997;2(3):335–351. doi: 10.1177/135910539700200305. [DOI] [PubMed] [Google Scholar]
- 22.Weech-Maldonado R, Carle A, Weidmer B, Hurtado M, Ngo-Metzger Q, Hays RD. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) cultural competence (CC) item set. Med Care. 2012;50(9, suppl 2):S22–S31. doi: 10.1097/MLR.0b013e318263134b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Dhanani LY, Franz B. Unexpected public health consequences of the COVID-19 pandemic: a national survey examining anti-Asian attitudes in the USA. Int J Public Health. 2020;65(6):747–754. doi: 10.1007/s00038-020-01440-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Dhanani LY, Franz B. Why public health framing matters: an experimental study of the effects of COVID-19 framing on prejudice and xenophobia in the United States. Soc Sci Med. 2021;269:113572. doi: 10.1016/j.socscimed.2020.113572. [DOI] [PubMed] [Google Scholar]
- 25.Darling-Hammond S, Michaels EK, Allen AM, et al. After “the China virus” went viral: racially charged coronavirus coverage and trends in bias against Asian Americans. Health Educ Behav. 2020;47(6):870–879. doi: 10.1177/1090198120957949. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Chou W-YS, Gaysynsky A. Racism and xenophobia in a pandemic: interactions of online and offline worlds. Am J Public Health. 2021;111(5):773–775. doi: 10.2105/AJPH.2021.306230. [DOI] [PMC free article] [PubMed] [Google Scholar]