If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to: Ying Liu, PhD, Center for Economic and Social Research, University of Southern California, 635 Downey Way, Room 501K, Los Angeles CA 90089.
Center for Economic and Social Research, University of Southern California, Los Angeles, CaliforniaDepartment of Sociology and Spatial Sciences, University of Southern California, Los Angeles, California
This study examines COVID-19–associated discrimination regardless of infection status. It evaluates the contribution of various risk factors (e.g., race/ethnicity and wearing a face mask) and the relationship with mental distress among U.S. adults in March and April 2020, when the pandemic escalated across the country.
Methods
Participants consisted of a probability-based, nationally representative sample of U.S. residents aged ≥18 years who completed COVID-19–related surveys online in March and April (n=3,665). Multivariable logistic regression was used to predict the probability of a person perceiving COVID-19–associated discrimination. Linear regression was used to analyze the association between discrimination and mental distress. Analyses were conducted in May 2020.
Results
Perception of COVID-19–associated discrimination increased from March (4%) to April (10%). Non-Hispanic Black (absolute risk from 0.09 to 0.15 across months) and Asians (absolute risk from 0.11 to 0.17) were more likely to perceive discrimination than other racial/ethnic groups (absolute risk from 0.03 to 0.11). Individuals who wore face masks (absolute risk from 0.11 to 0.14) also perceived more discrimination than those who did not (absolute risk from 0.04 to 0.11). Perceiving discrimination was subsequently associated with increased mental distress (from 0.77 to 1.01 points on the 4-item Patient Health Questionnaire score).
Conclusions
Perception of COVID-19–associated discrimination was relatively low but increased with time. Perceived discrimination was associated with race/ethnicity and wearing face masks and may contribute to greater mental distress during early stages of the pandemic. The long-term implications of this novel form of discrimination should be monitored.
INTRODUCTION
The coronavirus disease 2019 (COVID-19) pandemic threatens global health and national economies. Both the UN and the Centers for Disease Control and Prevention have called for increased attention to prevent stigma associated with COVID-19,
The Health Stigma and Discrimination Framework: a global, crosscutting framework to inform research, intervention development, and policy on health-related stigmas.
Discrimination toward people who share social or behavioral characteristics with COVID-19 patients but may not carry the novel virus, termed COVID-19–associated discrimination (CAD), was first seen in heightened anti-Chinese rhetoric online
Soon after, CAD expanded to a broader subset of Americans regardless of race/ethnicity. For instance, people wearing face masks were perceived as more likely to carry the virus,
For black men, fear that masks will invite racial profiling.
The New York Times.April 14, 2020; (https://www.nytimes.com/2020/04/14/us/coronavirus-masks-racism-african-americans.html. Updated May 26, 2020. Accessed June 22, 2020)
This study uses national survey data to examine CAD among U.S. residents when the pandemic escalated across the country. It explores the contribution of risk factors including (1) social and behavioral identifiers that might tie individuals to COVID-19 (e.g., race/ethnicity and mask wearing) and (2) in-person and social media encounters during which discriminatory acts might occur. Additionally, the study examines the association between perceived CAD and mental distress.
METHODS
Study Sample
The survey cohort came from the Understanding America Study (UAS), a nationally representative panel maintained by the Center for Economic and Social Research at the University of Southern California. The panel consists of a random sample of residents aged ≥18 years across the U.S., recruited using address-based sampling since 2014. Respondents were invited by mail and provided a tablet and broadband Internet if they did not have Internet access. They then answered survey questions on a computer, tablet, or smartphone. The design of the UAS is documented extensively elsewhere,
Angrisani M, Finley B, Kapteyn A. Can Internet match high-quality traditional surveys? Comparing the Health and Retirement Study and its online version. In: The Econometrics of Complex Survey Data (Advances in Econometrics, Vol. 39). United Kingdom: Emerald Publishing Limited; 2019:3–33.https://doi.org/10.1108/S0731-905320190000039001.
The Understanding America Study: coronavirus in America (COVID) survey. University of Southern California. https://uasdata.usc.edu/page/Covid-19+Home. Updated 2020. Accessed May 1, 2020.
with a first survey fielded between March 10 and March 31, 2020 (March survey) and a second from April 1 to April 28 (April survey). Among the 8,502 invited to the March survey, 6,884 (81%) completed it. The April survey was part of a tracking series
requiring a separate consent. Although all panelists were invited on March 29, only 5,891 consented by April 1, with 5,450 (93%) completing the April survey. Individuals who completed both surveys were included for analysis.
Because prior discrimination experience could be a potential confounder in this study, participants without such information were excluded from the analysis. Roughly 1 year before COVID-19 surveys, respondents were asked about discrimination experience in day-to-day life. Panelists at the time (6,708) were invited to this survey, with 5,569 (83%) completing it. The analytic sample was then restricted to the 3,665 respondents who completed all 3 surveys. The parent study protocol was approved by the IRB of University of Southern California.
Measures
This study assessed CAD using a 4-item scale adapted from the Everyday Discrimination Scale Short Version,
Validation of the Detroit area study discrimination scale in a community sample of older African American adults: the Pittsburgh Healthy Heart Project.
Respondents answered yes, no, or unsure to the question of whether they had the following experiences due to people thinking they might have the coronavirus: treated with less courtesy and respect than others; receiving poorer service at restaurants or stores; people act as if they were afraid of them; threatened or harassed. Perceived CAD was identified if one reported yes to any of the experiences. Sensitivity analysis showed similar results when unsure was coded as no or missing.
Mental distress was assessed using the Patient Health Questionnaire (PHQ-4) validated in the general population.
Respondents answered how often in the past 14 days they felt bothered by the following: feeling anxious; not being able to stop or control worrying; feeling depressed; having little interest in doing things. Response options included: not at all (0), several days (1), more than half the days (2), and nearly every day (3). Although the first 2 items assessed anxiety and the latter 2 assessed depression, the sum score measured overall mental distress.
The study also included social and behavioral risk factors of CAD. Both COVID-19 surveys asked respondents whether they had worn a face mask or covering in the past 7 days. Social media exposure was only assessed in March, as number of minutes a respondent spent on social media in a day on average, which was then categorized as: none (0), 1–30 minutes (1), 31–60 minutes (2), 61–120 minutes (3), and more than 120 minutes (4). Assuming that respondents’ social media usage was similar across months, responses were carried over to analyze the April data.
Working status was assessed in April as number of days in the past week a respondent worked at a job and days worked from home. The ratio between the 2 was then derived as: not working (0), working outside (1), working partially from home (2), and working completely from home (3). The April survey also asked about in-person interaction in the past 7 days (e.g., going to the grocery store or pharmacy or visiting others).
Discrimination experience before the pandemic was assessed between December 2018 and February 2019 by asking respondents if they had been treated unfairly or harassed because of their parents’ education or possessions. Prior mental distress was evaluated using the 8-item Center for Epidemiologic Studies–Depression Scale
on a rolling basis—for this sample, 53% in 2018 and 39% in 2019. As prior data are often strong predictors of later observation on a similar construct, controlling for these prior measures was useful to parse out the impact of the pandemic.
Additional covariates were used in this study. The COVID-19 surveys asked about respondents’ symptoms experienced in the past 7 days (e.g., cough and fever). Although COVID-19 cannot be diagnosed based on symptoms alone, a combination of self-reported shortness of breath and fever, the 2 cardinal symptoms of the disease,
was considered indicative of a respondent potentially having COVID-19. Sociodemographic data were collected periodically in the UAS—for this sample, between December 2019 and March 2020.
Two contextual measures were constructed from publicly available data, quantifying the progression of and policy response to the pandemic in the state in which the respondent resided on the date they responded. Disease prevalence was constructed by dividing the total number of confirmed cases documented by the New York Times
Analyses were performed using Stata, version 16 in May 2020 and restricted to respondents who completed both COVID-19 surveys and a prior survey on perceived discrimination (n=3,665). Differences in sample characteristics by whether one perceived CAD were tested using Pearson chi-square test for categorical variables and independent sample t-test for continuous variables.
Two distinct modeling strategies were used. The cross-sectional approach analyzed each month's data separately because people's exposure to CAD could take different forms before (March) and during (April) the implementation of SIP orders requiring most people to stay home. The longitudinal approach then compared across months by predicting the April dependent variable while controlling for its March counterpart.
Multivariable logistic regression was used to predict the probability of a person perceiving CAD as a function of risk factors (e.g., race/ethnicity and wearing face masks) and prior discrimination experience, with AOR quantifying association and predicted probability assessing absolute risk. Multivariable linear regression was then used to model mental distress (PHQ-4 score), using CAD as a predictor rather than an outcome as in the logistic regression, while controlling for prior mental distress. All models adjusted for COVID-19 prevalence in each state, number of days under the SIP order, self-reported symptoms, and demographic characteristics. SEs were clustered at the state level using the Eicker–Huber–White estimator,
Abadie A, Athey S, Imbens GW, Wooldridge J. When should you adjust standard errors for clustering? Cambridge, MA: National Bureau of Economic Research. https://doi.org/10.3386/w24003. Published November 2017.
accounting for nonindependence of respondents who resided in the same state. Less than 5% of data were missing for each variable and list-wise deletion was used in analyses.
RESULTS
The study sample consisted of 3,665 U.S. adults aged ≥18 years. Appendix Table 1, available online, shows that three quarters (75%) were non-Hispanic White; more than half (55%) were female; three quarters (73%) were aged 18–64 years; and less than half (41%) reported having a bachelor's degree or higher. More than half (59%) had a household income <$75,000.
Compared with the excluded sample who completed COVID-19 surveys but not the prior discrimination survey, the analytic sample was more likely to be non-Hispanic White; male; older; and with higher income. Because the UAS oversampled Los Angeles County residents in the past year,
46% of the excluded sample came from Los Angeles, driving the differences. Appendix Table 2, available online, shows a similar composition of included and excluded non–Los Angeles samples.
The overall percentage perceiving CAD doubled, from 4% in March to 10% in April. Table 1 contrasts data by perceived CAD. In both months, relative to respondents who did not experience CAD, those who did were more likely to be racial/ethnic minorities, immigrants, and younger, and to have disease-related symptoms, face mask usage, and prior discrimination experience. Additionally, those who perceived CAD in April tended to have a lower household income and reside in states with higher COVID-19 prevalence. Mental distress, during and before the pandemic, was higher for those perceiving CAD than for those who did not.
Table 1Comparison of Respondents Perceiving and Not Perceiving COVID-19–Associated Discrimination (n=3,665)
March survey
April survey
Characteristics
No perceived discrimination (n=3,460)
Perceived discrimination (n=157)
No perceived discrimination (n=3,285)
Perceived discrimination (n=371)
Race, n (%)
White
2,627 (76)
86 (55)
***
2,499 (76)
234 (63)
***
Black
229 (7)
23 (15)
226 (7)
36 (10)
Hispanic
317 (9)
16 (10)
289 (9)
54 (15)
Asian
96 (3)
21 (13)
92 (3)
26 (7)
Other race/ethnicity
185 (5)
11 (7)
176 (5)
21 (6)
Age, years, n (%)
18‒34
411 (12)
33 (21)
***
382 (12)
63 (17)
***
35‒44
1,253 (36)
74 (47)
1,177 (36)
166 (45)
55‒64
825 (24)
32 (20)
782 (24)
85 (23)
≥65
971 (28)
18 (11)
944 (29)
57 (15)
Education, n (%)
High school or less
739 (21)
36 (23)
703 (21)
93 (25)
Some college
1,301 (38)
58 (37)
1,236 (38)
135 (36)
Bachelor or higher
1,420 (41)
63 (40)
1,346 (41)
143 (39)
Household income, n (%)
<$25,000
607 (18)
33 (21)
571 (17)
92 (25)
**
$25,000–$49,999
722 (21)
36 (23)
692 (21)
76 (21)
$50,000–$74,999
694 (20)
32 (20)
643 (20)
72 (19)
>$75,000
1,430 (41)
56 (36)
1,374 (42)
130 (35)
Sex, n (%)
Female
1,914 (55)
89 (57)
1,832 (56)
194 (52)
Fever/chills and shortness of breath, n (%)
Yes
36 (1)
<10 (<6)
**
18 (1)
<10 (<3)
*
Prior perceived discrimination, n (%)
Yes
954 (28)
66 (42)
***
854 (26)
169 (46)
***
Immigrant status, n (%)
Nonimmigrant
2,053 (61)
85 (54)
***
1,976 (62)
190 (52)
***
First-generation immigrant
252 (7)
25 (16)
249 (8)
31 (8)
Second-generation immigrant
356 (11)
27 (17)
319 (10)
66 (18)
Third-generation immigrant
714 (21)
20 (13)
661 (21)
78 (21)
Wear a face mask, n (%)
Yes
229 (7)
41 (26)
***
1,584 (48)
215 (58)
***
Time spent on social media in a day on average, n (%)
None
727 (21)
21 (14)
701 (22)
60 (16)
1‒30 minutes
1,192 (35)
52 (34)
1,126 (35)
130 (35)
31‒60 minutes
722 (21)
41 (27)
688 (21)
83 (22)
61‒120 minutes
535 (16)
23 (15)
494 (15)
63 (17)
>120 minutes
259 (8)
17 (11)
242 (7)
34 (9)
Mental distress (PHQ-4), mean (IQR)
1.70 (0‒2)
3.05 (0‒5)
***
2.33 (0‒4)
3.60 (1‒5)
***
Prior mental distress (CES-D 8), mean (IQR)
1.50 (0‒2)
2.39 (0‒4)
***
1.48 (0‒2)
2.23 (0‒4)
***
Number of COVID-19 cases per 100,000 population in the state, mean (IQR)
8.21 (0‒4)
12.16 (1‒6)
921.76 (286‒787)
875.38 (298‒872)
Number of days under state SIP order, mean (IQR)
0.16 (0‒0)
0.27 (0‒0)
13.13 (7‒19)
14.33 (9‒20)
**
Work from home, n (%)
Do not work
1,705 (53)
158 (45)
**
Work but no day WFH
655 (20)
97 (28)
Some days WFH
156 (5)
20 (6)
All days WFH
681 (21)
75 (21)
Social activities, n (%)
Go out to a bar, club, or other places where people gather
30 (1)
<10 (<3)
Go to grocery store or pharmacy
2,535 (77)
302 (81)
Go to a friend, neighbor, or relatives’ residence (that's not your own)
691 (21)
84 (23)
Have visitor such as friends, neighbors, or relatives at your residence
694 (21)
96 (26)
*
Attend a gathering with more than 10 people such as a reunion, wedding, or funeral
49 (1)
10 (3)
Go outside to walk, hike, or exercise
2,455 (75)
275 (74)
Note: Boldface indicates statistical significance (*p<0.05; **p<0.01; ***p<0.001) using chi-square test for categorical variables and independent sample t-test for continuous variables.
CES-D 8, 8-item Center for Epidemiologic Studies–Depression Scale; PHQ-4, 4-item Patient Health Questionnaire; SIP, Shelter-in-Place; WFH, working from home.
Cross-sectional analyses (Table 2) show the AOR of perceiving CAD in March. In addition to race/ethnicity, risk factors, including prior discrimination experience, immigrant status, mask wearing, and social media usage, entered the model one at a time to illustrate their incremental contribution and potential mediation of the racial/ethnic effects. Compared with non-Hispanic Whites, more non-Hispanic Blacks perceived CAD (AOR=2.69, p<0.001 in Model 4). Asians also had a higher likelihood, but the magnitude reduced when immigration status and face mask usage entered the model (AOR=5.79, p<0.001 in Model 1; AOR=3.71, p<0.001 in Model 2; AOR=3.04, p<0.001 in Model 3), suggesting their potential to mediate the association between being Asian and perceiving CAD. Those who experienced discrimination before also experienced more CAD than those who did not (AOR=1.74, p<0.01), as did respondents who wore face masks in March relative to those who did not (AOR=3.51, p<0.001).
Table 2AORs of Perceiving COVID-19–Associated Discrimination in March (n=3,665)
Spent no time on social media in a day on average as reference category.
1‒30 minutes
—
—
—
—
1.28 (0.79, 2.07)
31‒60 minutes
—
—
—
—
1.67 (0.98, 2.85)
61‒120 minutes
—
—
—
—
1.09 (0.70, 1.69)
>120 minutes
—
—
—
—
1.75 (0.91, 3.34)
Note: Boldface indicates statistical significance (*p<0.05; **p<0.01; ***p<0.001) using multivariable logistic regression. Values in parentheses are 95% CI. All models adjusted for age, education, sex, household income, self-reported symptom, COVID-19 prevalence in the state a respondent resided, and the number of days under the state's Shelter-in-Place order on the day of the response, with Eicker–Huber–White robust SEs clustered at the state level.
a Non-Hispanic White as reference category.
b Nonimmigrant as reference category.
c Spent no time on social media in a day on average as reference category.
Table 3 presents the AOR of perceiving CAD in April including risk factors presented in the March analysis (Model 4), whereas Appendix Table 3, available online, includes the summary for Models 0–3. Because the April survey collected data on respondents’ working status and social activities, these measures were added to Models 5 and 6. Compared with non-Hispanic Whites, non-Hispanic Blacks (AOR=1.80, p<0.001 in Model 6) and Asians (AOR=2.02, p<0.05) were more likely to experience CAD, but the racial/ethnic gaps were smaller than those observed in March. Prior discrimination experience remained a strong predictor of CAD (AOR=2.18, p<0.001).
Table 3AORs of Perceiving COVID-19–Associated Discrimination in April (n=3,665)
Did not work as reference category. WFH, working from home.
Work but no day WFH and no mask
—
1.29 (0.76, 2.20) 0.35
1.26 (0.75, 2.11)
1.32 (0.74, 2.33)
Some days WFH and no mask
—
1.61 (0.84, 3.09) 0.15
1.55 (0.79, 3.00)
1.51 (0.74, 3.07)
All days WFH and no mask
—
1.12 (0.85, 1.48) 0.42
1.11 (0.83, 1.48)
1.16 (0.83, 1.61)
Work but no day WFH and wear mask
—
1.52 (1.13, 2.05) 0.01
1.58 (1.17, 2.14)**
1.73 (1.25, 2.41)**
Some days WFH and wear mask
—
1.10 (0.58, 2.12) 0.76
1.12 (0.59, 2.11)
1.08 (0.57, 2.07)
All days WFH and wear mask
—
1.20 (0.88, 1.63) 0.26
1.23 (0.89, 1.69)
1.25 (0.87, 1.78)
Social activities
Go out to a bar, club, or other places where people gather
—
—
1.17 (0.47, 2.89)
0.88 (0.25, 3.05)
Go to grocery store or pharmacy
—
—
1.10 (0.82, 1.47)
1.10 (0.80, 1.50)
Go to a friend, neighbor, or relatives’ residence (that's not your own)
—
—
0.94 (0.73, 1.22)
0.94 (0.73, 1.23)
Have visitor such as friends, neighbors, or relatives at your residence
—
—
1.32 (1.08, 1.60)**
1.29 (1.04, 1.61)*
Attend a gathering with more than 10 people such as a reunion, wedding, or funeral
—
—
1.79 (0.89, 3.60)
2.09 (1.06, 4.11)*
Go outside to walk, hike, or exercise
—
—
1.18 (0.89, 1.57)
1.22 (0.86, 1.73)
Note: Boldface indicates statistical significance (*p<0.05; **p<0.01; ***p<0.001) using multivariable logistic regression. Values in parentheses are 95% CI. All models adjusted for age, education, sex, household income, self-reported symptom, COVID-19 prevalence in the state a respondent resided, and the number of days under the state's Shelter-in-Place order on the day of the response, with Eicker–Huber–White robust SEs clustered at the state level.
a Also controlled for March perception in COVID-19 associated discrimination in the longitudinal analysis.
b Non-Hispanic White as reference category.
c Nonimmigrant as reference category.
d Spent no time on social media in a day on average as reference category.
e Did not work as reference category.WFH, working from home.
In April, respondents who were second- (AOR=1.86, p<0.01) and third- (AOR=1.65, p<0.001) generation immigrants experienced more CAD relative to nonimmigrants. Frequent social media users, who spent >2 hours in a day on average, were also at a higher risk than those who did not use social media (AOR=1.58, p<0.05). To understand the association with use of face masks, reports of mask wearing were interacted with respondents’ working status, given that most states implemented SIP orders in April requiring everyone except essential workers to stay home. Compared with people who did not work, only those who worked outside and wore face masks perceived more CAD (AOR=1.58, p<0.01). No difference was found for those who worked outside but did not wear face masks (AOR=1.26, p>0.05) nor those who worked partially or fully from home regardless of face mask usage. Engagement in social activities in which in-person discrimination might occur (e.g., going to the grocery store or pharmacy) was not associated with more discrimination except for those with visitors at their own residence (AOR=1.32, p<0.01).
Figure 1 illustrates the predicted absolute risk by race/ethnicity, adjusted for risk factors and covariates in the cross-sectional analyses. In both months, non-Hispanic Blacks (absolute risk [AR]=0.09, 95% CI=0.059, 0.122 for March, Model 4; AR=0.151, 95% CI=0.115, 0.187 for April, Model 6) and Asians (AR=0.105, 95% CI=0.070, 0.140 for March, Model 4; AR=0.165, 95% CI=0.086, 0.244 for April, Model 6) had higher risk than other racial/ethnic groups (AR ranged from 0.027 to 0.054 in March and 0.090 to 0.108 in April). From March to April, the risk increased for all racial/ethnic groups including non-Hispanic Whites, although it was less evident for Asians. Mask wearing also increased the likelihood of perceiving CAD, persistent in both March (AR=0.112, 95% CI=0.082, 0.142 for wearing; AR=0.037, 95% CI=0.031, 0.042 for not wearing) and April (AR=0.144, 95% CI=0.110, 0.178 for wearing and working completely outside home; otherwise, ARs ranged from 0.079 to 0.111).
Figure 1Predicted probability for racial/ethnic groups experiencing COVID-19–associated discrimination (n=3,665). All models adjusted for age, education, sex, household income, self-reported symptom, COVID-19 prevalence in the state a respondent resided, and the number of days under the state's Shelter-in-Place order on the day of the response, with Eicker–Huber–White robust SEs clustered at the state level. NH, non-Hispanic.
Longitudinal analysis (Table 3, last column) associated the increase in perceiving CAD over time with several aforementioned risk factors, including being non-Hispanic Black; a second- or third-generation immigrant; and a frequent social media user (i.e., spending >2 hours in a day on average), as well as having prior discrimination experiences and a combination of working outside and using a face mask. Different from the cross-sectional analysis (Table 3, Model 6), attending a gathering with >10 people in April significantly predicted the increased CAD perception over time. Relative to non-Hispanic Whites, Asians did not experience more CAD in April after accounting for their experience in March (AOR=1.35, p>0.05).
In addition to assessing factors associated with perceived CAD, subsequent analyses examined the relationship between CAD and mental health factors to determine potential impact (Appendix Table 4, available online). Cross-sectional analysis revealed that perceived CAD predicted a higher PHQ-4 score in March (b=0.77, p<0.001) and April (b=1.01, p<0.001) after adjusting for the prior score on 8-item Center for Epidemiologic Studies–Depression Scale and person- and state-level covariates (e.g., demographics and COVID-19 prevalence). The longitudinal approach associated perceived CAD with the PHQ-4 score in April after controlling for PHQ-4 score in March (b=0.50, p<0.001) and the same set of covariates.
DISCUSSION
Anecdotal discriminatory acts amid the COVID-19 pandemic have been widely documented in media reports. To the authors’ knowledge, this study provides the first systematic assessment on how perceived CAD is associated with potential risk factors (e.g., race/ethnicity and social media usage) and mental distress. Given the representativeness of this sample, the authors believe that these results are generalizable to U.S. residents.
The findings suggest that perceptions of CAD are increasing although the overall prevalence remains relatively low. Although all racial/ethnic groups have experienced CAD, disparities are observed. Elevated risk among Asians in March confirms the media reports on hate crimes,
and it was in part mediated by their immigration status and use of face masks. Although their risk also appeared higher than non-Hispanic Whites in April, it was mostly carried over from March. Non-Hispanic Blacks are another group who persistently experience more CAD, and their risk increased from March to April. This could reflect the longstanding stereotypes associating non-Hispanic Blacks with the spread of infectious diseases,
Smaller racial/ethnic gaps in April than March could be due to the increase in CAD among non-Hispanic Whites.
Besides race/ethnicity, wearing face masks is another risk factor consistently associated with CAD. This confirms media reports on the stigmatization of mask wearing during the pandemic,
For black men, fear that masks will invite racial profiling.
The New York Times.April 14, 2020; (https://www.nytimes.com/2020/04/14/us/coronavirus-masks-racism-african-americans.html. Updated May 26, 2020. Accessed June 22, 2020)
SARS related preventive and risk behaviours practised by Hong Kong-mainland China cross border travellers during the outbreak of the SARS epidemic in Hong Kong.
perceived CAD because of mask wearing could be attributable to the spotlight effect, in which individuals overestimate others’ attention directed toward them when their appearances are different from people around.
As many states transit to lifting SIP orders and reopen businesses and public activities, CAD associated with mask wearing, a highly recommended disease control metric,
The results also point to CAD's potential association with other risk factors such as heavy social media usage (i.e., spending >2 hours on a day on average), being an immigrant, and engagement in some social activities. However, such relationships seem less robust and need further confirmation.
As for potential health outcomes of CAD, this study showcases substantial evidence linking CAD to increased mental distress, consistent with literature associating general discrimination with poorer mental health, especially among racial/ethnic minorities.
Moderating effects of three coping strategies and self-esteem on perceived discrimination and depressive symptoms: a minority stress model for Asian international students.
attributable to concerns of disease spread, austere social restrictions, and financial stress.
Limitations
Findings from this study should be interpreted in light of several limitations. First, the analysis was based on observational data. The reported correlational relationships need further confirmation in stronger designs such as a field experiment. Additionally, CAD was operationalized as people's self-reported perception instead of direct observation. Other data types detailing occasions or types of discrimination might provide an in-depth view. Moreover, risk factors examined here are limited. For example, coughing in public was not covered in the data although anecdotal evidence associates it with disease avoidance. Finally, mental health was assessed by perceived mental distress instead of mental illness as in other studies.
The impact of the COVID-19 pandemic on public health, including mental health, will continue to reverberate for years to come. This study's initial findings of perceived CAD and its association with poorer mental health underscore the existing racial/ethnic inequalities experienced by many non-Hispanic Blacks and Asians. The findings also suggested discrimination toward those who wear face masks, signaling the need to raise public awareness, especially as economies reopen and wearing a face mask is highly recommended to prevent transmission.
ACKNOWLEDGMENTS
The project described in this paper relies on data from surveys administered by the Understanding America Study (UAS), which is maintained by the Center for Economic and Social Research (CESR) at the University of Southern California (USC). The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of USC or UAS. For questions or additional information about the UAS, contact Tania Gutsche, Project and Panel Manager, CESR, USC, at [email protected] UAS is funded from several sources, including the Social Security Administration and the National Institute on Aging under grant 5U01AG054580. The UAS coronavirus surveys referred to in this paper were initiated and financed by USC and funded in part by the Bill & Melinda Gates Foundation. The parent study protocol was approved by the USC IRB, and data access was granted by UAS’ data user agreement.
Ms. Brenneke's work was supported by the National Institute on Drug Abuse (T32DA007292) and Dr. Le's work was supported by a training grant from the National Institute of Mental Health (T32MH103210). The research presented in this paper is that of the authors and does not reflect the official policy of NIH. No financial disclosures were reported by authors of this paper.
YL and BKF conceived the study and analyzed and interpreted the data; all authors contributed to drafting and revising the article. The contents of the article have not been presented or published elsewhere.
The Health Stigma and Discrimination Framework: a global, crosscutting framework to inform research, intervention development, and policy on health-related stigmas.
For black men, fear that masks will invite racial profiling.
The New York Times.April 14, 2020; (https://www.nytimes.com/2020/04/14/us/coronavirus-masks-racism-african-americans.html. Updated May 26, 2020. Accessed June 22, 2020)
Angrisani M, Finley B, Kapteyn A. Can Internet match high-quality traditional surveys? Comparing the Health and Retirement Study and its online version. In: The Econometrics of Complex Survey Data (Advances in Econometrics, Vol. 39). United Kingdom: Emerald Publishing Limited; 2019:3–33.https://doi.org/10.1108/S0731-905320190000039001.
The Understanding America Study: coronavirus in America (COVID) survey. University of Southern California. https://uasdata.usc.edu/page/Covid-19+Home. Updated 2020. Accessed May 1, 2020.
Validation of the Detroit area study discrimination scale in a community sample of older African American adults: the Pittsburgh Healthy Heart Project.
Abadie A, Athey S, Imbens GW, Wooldridge J. When should you adjust standard errors for clustering? Cambridge, MA: National Bureau of Economic Research. https://doi.org/10.3386/w24003. Published November 2017.
SARS related preventive and risk behaviours practised by Hong Kong-mainland China cross border travellers during the outbreak of the SARS epidemic in Hong Kong.
Moderating effects of three coping strategies and self-esteem on perceived discrimination and depressive symptoms: a minority stress model for Asian international students.
30273-7&id=gr1.jpg){kind=link}