**1. Introduction**

The COVID-19 pandemic has caused widespread suffering and death around the world. Hundreds of millions of people have been infected by the virus, and at least 4.9 million people have died from COVID-19. However, the effects of the virus have not been equally distributed, with countries that explicitly prioritized economic incentives above public health having greater numbers of cases and fatalities [1]. Furthermore, there are widespread inequities in countries such as the United States, with the COVID-19 pandemic having disproportionate effects on communities of color, further exacerbating existing healthcare disparities [2–4]. Yet, while increasing evidence points to the greater threat posed by COVID-19 to Latinx communities, less is known about how identification as Hispanic/Latinx and migration status influence individuals' perception of risk and harm.

In this article, we use cross-sectional data from a large national probability sample to demonstrate a large positive association between ethnic identity and migration status and perceptions of harm from COVID-19 in the US. Through an OLS regression of over 150,000 survey responses, we find that individuals identifying as Hispanic/Latinx and firstgeneration immigrants perceive themselves to be at significantly greater risk of becoming infected with the virus, and dying if they contract the virus. Collectively, our results provide evidence about how the pandemic places increased stress on people from Latinx and immigrant communities relative to non-Hispanic White individuals in the US. As a result, not only do Hispanic/Latinx individuals and immigrants face greater health consequences of COVID-19, but they also face a disproportionate mental burden.

**Citation:** Jamieson, T.; Caldwell, D.; Gomez-Aguinaga, B.; Doña-Reveco, C. Race, Ethnicity, Nativity and Perceptions of Health Risk during the COVID-19 Pandemic in the US. *Int. J. Environ. Res. Public Health* **2021**, *18*, 11113. https:// doi.org/10.3390/ijerph182111113

Academic Editors: Anthony R. Mawson and Paul B. Tchounwou

Received: 15 July 2021 Accepted: 18 October 2021 Published: 22 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

This article proceeds in four further sections. First, we provide a potted review of the scholarly literature on public health among Latinxs and communities of color, especially during the COVID-19 pandemic. We conclude this section by introducing our hypotheses based on this literature. Second, we outline the data and methods employed to run our tests, including describing the data and variables from the Understanding America Study's Understanding Coronavirus in America surveys run from the Center for Economic and Social Research at the University of Southern California. Next, we describe our results and their significance for the understanding of health disparities in pandemics. Finally, we summarize our results in the context of the broader literature, describe some limitations of the manuscript, and provide recommendations for future research to build on the findings of this article.

#### **2. Determinants of Health Risk Perception**

Public perception of health risks is influenced by multiple factors including the possible feelings of dread, comprehension of the complexity of the situation, uncertainty about its effects, familiarity with the risk, the possibility of solving the situation by oneself, perceived incentives for accuracy, and news coverage of public health threats [5–7]. However, as these risks are often domain-specific, people can differentiate between particular threats and outcomes in the process of assessing the nature of the risk [8]. Context also plays an important role in risk perception [8]. This is particularly important for minority populations in the US. Following Ferrer and Klein, we can argue that racists and xenophobic attacks on African Americans and US Latinxs can influence risk perceptions, increasing pessimistic feelings toward the possible effects of threats.

Pessimistic feelings towards possible effects of threats are furthered evidenced by Martinez Tyson, Arriola, and Corvin's [9] research which finds that Latinx individuals across certain subgroups have comparable perceptions of risks posed particularly by mental health symptoms and diseases. Martinez Tyson and colleagues [9] also mention that economic and social discrimination could be responsible for a lack of healthcare visits by Latinx community members even though they accurately perceive symptoms and risks of diseases. Meanwhile, Bucay-Harari et al. [10] suggest that there may be a correlation between Latinx individuals and more severe mental health symptoms. Following this line of thought, we could argue that Latinx individuals are more likely to have severe mental health issues that often are exacerbated by the economic and social discrimination they experience which causes them to perceive a higher risk to their own health.

First, it is important to discuss how discrimination itself has been linked to adverse mental health outcomes in order to show its effects on the Latinx community and minority communities broadly, and to paint a linkage between discrimination and perceived health risk during COVID-19. Discrimination has been shown to be a determinant of health risk perceptions among minority groups. A number of studies have linked experienced and perceived discrimination to adverse mental health outcomes which are associated with higher perceptions of health risk [11]. Thompson [12] found that experienced discrimination is related to intrusion and avoidance symptoms regarding Black Americans, an issue that could lead towards social isolation. In particular, Thompson [12] found that appraisal of the stressfulness of the discriminatory experience was associated directly with the experiences of intrusion or avoidance symptoms. Similarly, Salgado de Snyder [13] found that experiencing discrimination for being Mexican among Mexican women was correlated with higher scores in depression on the CES-D depression scale. Williams et al. [14] showed that experiencing discrimination is linked to lower levels of subjective well-being and high distress, particularly among Black Americans.

Some research has found specific variables that are linked with psychological distress in Latinx individuals and other minorities. For instance, Brown et al. [15] found that, among Black Americans in particular, financial security is correlated with lower levels of distress, age is negatively correlated with distress (younger individuals and women generally had higher levels of distress), and higher levels of formal education are negatively correlated

with distress. Brown and colleagues [15] also found that higher self-reporting of discrimination was not associated with prior mental health issues and, instead, self-reported experiences of discrimination were the factor indicating adverse mental health and distress. Bucay-Harari et al. [10] have indicated that being underinsured/uninsured is an important factor in distress, anxiety, and greater severity of mental health problems, particularly in the Latinx community. Bucay-Harari and colleagues [10] also indicate that migrants and Latinx individuals are more likely to be uninsured because of increasing barriers to these individuals in accessing private or public health insurance due to immigration status, socioeconomic factors, and a lack of political representation. At the same time as Latinx individuals are more likely to have adverse health outcomes, distrust in public institutions broadly makes reaching out to people in distress due to disasters difficult [16].

These studies help to demonstrate that there is a certain stress or distress associated with discrimination and how it can exacerbate increasing sociopolitical barriers to healthcare that could lead to communities more likely to experience or perceive discrimination having higher perceptions of health risk during COVID-19. One way that discrimination can correlate with higher perceptions of health risk is through the concept of stigma. This concept describes the labeling of others with attributes that are devalued or discredited by those in a position of power, stereotyping such negative attributes, and using them to separate these others from the dominant group. As a result of this separation, the stigmatized group suffers from loss of status and discrimination. Direct consequences of this stigmatization are mental and physical illness at the individual level, and unequal access to healthcare and socio-economic inequality at the macro level [17,18].

Oaten, Stevenson, and Case [19] found that while stigma surrounding fear of disease outbreak can cause heightened hygiene and disease avoidance at first, generally stigmatization can become a barrier to health care access. Earnshaw and Chaudoir [20] find that stigmatized groups tend to have internalized senses of inferiority compared to unstigmatized groups. Fischer et al. [21] also found that stigma can cause a significant reduction in public health measure compliance or generally impede outbreak controls.

Stigmatization also tends to affect minority groups above others. Health-related stigmas are generally found to have adverse effects on Latinx or African American adults. Darrow, Montanea, and Gladwin [22] found that perception of HIV-infection among Latinx or African American adults is correlated with having never received an HIV test. Rueda et al. [23] demonstrated that health-related stigmas also are correlated with higher levels of anxiety, stress, and avoidance strategies. Nadeem and her colleagues [24] show that this is particularly relevant for immigrant women who are more likely than nonimmigrant women to report stigma-related concerns over care and particularly over mental health care. Perreira and Pedroza [25] argue that anti-immigrant sentiments can produce higher mortality, poorer self-reported overall health and mental health specially among Latinx children and adults living in mixed-status families. Finally, Faccini et al. [26] found that, generally, stigma hinders contract tracing efforts which can then exacerbate risk in stigmatized communities. With mental health distress and a lack of healthcare access prominent with both discrimination and stigma, we can begin to discuss more direct vulnerabilities that the Latinx community has in relation to COVID-19 specifically. Due to persistent systemic inequalities in the United States, minority populations tend to have disproportionately high hospitalization rates associated with COVID-19 while also being more likely to abide by regulations or change behavior to stop the spread of disease [27,28]. Olivo et al. [29] posit reasons for these communities' high rates of contraction of infectious diseases as being economically related to the struggle to get personal protective equipment (PPE), a dominating presence in service industry jobs that are unlikely to shut down, and generally from having to take on riskier jobs in exchange for money to leverage vulnerable economic situations, and culturally related to a higher likelihood to live in multi-generational homes. Political leaders and the public often focus on cultural factors to blame vulnerable communities for disease [30–32].

The view of cultural inadequacy rather than the general inequities driving risk factors may cause majority populations to stigmatize minority populations [30]. The reality of Latinxs' and immigrants' specific vulnerability to COVID-19 is compounded by their higher likelihood to avoid contact with educational or health care services due to increasing number of raids, federal immigration enforcement, and immigration surveillance at all levels of government disrupting their daily lives [33–35].

Immigrants in general are also more susceptible to infectious disease compared to native populations [36,37]. Limina et al. [36] found that the reasons for this increased likelihood of contracting infectious disease are the socioeconomic situations in the country in which they are living. Some of these situations are exacerbated by immigration status in the host country, social exclusion, discrimination, language difficulties, gender, and access to medical services, among other things [2,36]. Distress might be higher on undocumented immigrants having to choose on a day-to-day basis between employment status, financial security, and their health and well-being [37] as legal residency status permeates immigrants' position in a stratified system [25,34].

From the perspective of the state, increased discrimination toward minority groups regarding their access to health benefits also makes them more susceptible to increased negative effects in health crises. Perreira and Pedroza [25] argue that a decline in public assistance coverage increased poverty and food insecurity among immigrant households and mixed-status families, simultaneously decreasing health utilization among immigrant women and their children.

Far less is known about the extent to which the public perceives health risks associated with public health emergencies and infectious disease outbreaks such as COVID-19 [38]. In this area, studies have primarily come from multiple global pandemics such as the SARS and avian influenza epidemics [39], the H1N1 swine flu [40–43], and the Ebola outbreak [44,45]. Most of this literature, however, relies on nationally representative and cross-sectional data, which provides challenges in analyzing minority populations such as foreign-born groups, and racial and ethnic minorities.

In this article, we build on prior research to develop a thorough understanding of the mental health toll of the COVID-19 pandemic on Latinx and migrant individuals. Given prior work that demonstrates the disproportionate stress of public health threats for the Hispanic/Latinx community and for immigrants in the United States, we expect that perceptions of risk and harm will be greater among individuals from Latinx and first-generation immigrants than non-Hispanic/Latinx and non-immigrant individuals. Building on previous work that identifies greater levels of exposure to COVID-19 risk and harm among Latinx communities than other communities, we expect this to also be reflected in greater degrees of worry about becoming infected or dying from the virus. As a result, our hypotheses are:

**Hypothesis 1 (H1).** *Latinx individuals report a higher chance of becoming infected and dying from COVID-19 compared to non-Latinx individuals.*

**Hypothesis 2 (H2).** *First-generation immigrants report a higher chance of becoming infected and dying from COVID-19 compared to non-immigrants.*

**Hypothesis 3 (H3).** *Latinx individuals who are also first-generation immigrants report a higher chance of becoming infected and dying from COVID-19 compared to other individuals.*

#### **3. Materials and Methods**

All data for this study came from the Understanding America Study's (UAS) Coronavirus in America Survey conducted by the Center for Economic and Social Research (CESR) at the University of Southern California. We used Waves 1–25 of the Understanding COVID-19 national studies conducted between 10 March 2020 and 31 March 2021. In total, most variables in our analysis have over 140,000 observations over this entire period, although, given not all questions were asked in each survey wave and nonresponses, we end with varying numbers of observations in each model.

#### *3.1. Dependent Variables*

Our two dependent variables of interest reflect participants' perception of health risks from COVID-19. Participants were asked to indicate their perceived chance of contracting COVID-19 in the next three months and their chance of dying from COVID-19 if they do get infected by the virus. For both questions, they were asked to place their risk on a scale from 0 percent to 100 percent. To address the positive-skewed distribution of both variables and make the results more substantively meaningful, we divided these variables into quintiles.

#### *3.2. Independent Variables*

We have two independent variables of interest. First, we are interested in how identifying as Hispanic/Latinx affects perception of health risks from COVID-19. This was measured through a dichotomous variable where participants were coded as 1 if they identified as Hispanic/Latinx, and 0 if they did not.

Second, we expected that an individual's immigration status would influence their perception of health risks from COVID-19. This was measured through participants' identification as a non-immigrant (0); first-generation immigrant (immigrant who migrated to the US) (1); second-generation immigrant (US-born children of at least one foreignborn parent) (2); third-generation immigrant (US-born children of at least one US-born parent, where at least one grandparent is foreign-born) (3); or unknown immigrant status (4). Given the absence of information about how to interpret the unknown immigrant status response, we recoded this response to indicate that these are missing values, and they were excluded from our analysis. As a result, this variable ranges from 0 to 3 with each number corresponding to the generation.
