Lay-theories are everyday reasonings people construct to help them make sense of their social worlds (Levy et al., 2006). These theories are influential in shaping individuals’ attitudes and behaviors (Molden & Dweck, 2006). Prior research has taken both descriptive and predictive approaches in understanding lay-theories about a variety of psychological attributes, groups, and social issues (Zedelius et al., 2017).

In the health domain, lay-beliefs can influence perceptions of health and health-related outcomes. For example, lay-theories about obesity as due to diet (versus exercise) are related to lower food consumption (McFerran & Mukhopadhyay, 2013). People who endorse more incremental lay-theories (i.e., one is largely in control of one's own health) report greater COVID-19 protective behaviors like mask-wearing (Zhang & Kou, 2022). Importantly, lay-theories are not limited to views towards one’s own health and behaviors. The lay-theories individuals hold about the causes, cures, and contributors to physical and mental health phenomena can guide group perceptions and attitudes. For example, lay-theories of self-control as fixed predicts negative perceptions of smokers and obese individuals (Freeman et al., 2013). Clinicians who endorse lay-theories of mental illness as genetically-based see their clients as less blameworthy for their condition, but also express less empathy (Lebowitz & Ahn, 2014).

Individuals also hold lay-theories about race and racism (e.g., Haslam et al., 2006; Rosner & Hong, 2010). For example, White and Black Americans may hold different lay-theories on what constitutes racism in the first place – White people are less likely to see subtle behaviors that reflect discomfort/unfamiliarity with Black people, or denials of current racism, as constituting racism (Sommers & Norton, 2006). Lay-theories about race can carry influential personal and social consequences. For example, biological and essentialist lay-beliefs about race predict more stereotypical attitudes towards racial minorities (Haslam et al., 2006) and greater acceptance of current racial disparities (Williams & Eberhardt, 2008). Given that lay-theories dynamically shape and reflect our understanding of health and racial perceptions, this integration seems valuable for understanding views on racial health disparities within COVID-19.

Through the lens of a Critical Race Psychology (CRP) framework, we use lay-theories to examine how Americans reflect, reproduce, deny, or acknowledge racism in their beliefs about the causes of health disparities between marginalized groups and White people. CRP engages Critical Race Theory (CRT; Crenshaw, 1995) as an analytic framework to be applied in psychological science (e.g., Jones, 2024; Perez & Salter, 2020; Salter & Adams, 2013; Volpe et al., 2019). Articulations of CRP have highlighted six core tenets for understanding how race and racism continue to manifest in society including: racism as a systemic force, how neoliberal individualist discourse masks racism’s role in society, societal change and stagnation as a function of interest convergence, possessive investment in Whiteness, intersectionality and matrices of oppression, and counter-storytelling as a tool for combating racism (Salter et al., 2024). All six tenets are relevant to this work, but two key intersecting ideas were especially helpful for our analysis: (1) racism is systemic, deeply embedded in the fabric of US society and (2) culturally-dominant narratives in the US—e.g., neoliberal discourses about individual choice, color-blindness, personal responsibility, and meritocracy— reproduce racism by obscuring racism’s central role in structuring society. These dominant narratives appear race-neutral but functionally obscure the ongoing significance of racism and need for reparative action during crisis (e.g., Bonilla-Silva, 2022). CRP explicitly challenges these ostensibly identity-neutral narratives that minimize the impact of race and racism on our perceptions and behaviors. Moreover, a CRP analysis prompts consideration of how perceptions of Black and Latinx communities are racially and culturally complex, shaped by both overlapping and divergent histories, experiences, and representations in the US.

Culturally-dominant narratives about race and racism can shape individual-level views of the world, which are then reproduced in individuals' preferences, attributions, and behaviors (see Salter et al., 2018; Skinner-Dorkenoo et al., 2023). While dominant narratives and lay-theories are distinct–the former being largely expressed in cultural products, institutions, and social representations, the latter being expressed by individual actors–they are mutually constituted (e.g., Markus & Kitayama, 2010). For example, existing dominant narratives about personal responsibility in the US government, media, and popular health campaigns promote views of poor health outcomes as the result of individual choices that are independent of structural forces (Hook & Markus, 2020). In turn, people offer advice, support policies, and troubleshoot problems in line with those views. People’s beliefs and choices develop within the cultural and social context and the context reflects those views. In the US, colorblindness can manifest in four key ways: using abstract language to discuss race or racism, naturalizing racialized outcomes, attributing racial differences to cultural practices, and minimizing racism (Bonilla-Silva, 2022). Thus, people may generate individual level lay-theories that integrate existing narratives about race to fit the present context of COVID-19.

Given the long history of discriminatory practices, abusive experimentation, implicit bias, and institutionalized policies that harmed Americans of color (Feagin & Bennefield, 2014), a CRP framework may be particularly useful for examining racial health disparities. Recent research suggests greater perceptions of health disparities (i.e., perceiving COVID-19 as impacting People of Color worse) among White Americans predicts less COVID-19 fear, which in turn predicts reduced support for COVID-19 safety precautions (Skinner-Dorkenoo et al., 2022). A critical implication of this research is that the consequences of conversations around COVID-19 are not race neutral; the racialized context in which they are discussed have social, psychological, and policy implications. Volpe and colleagues (2019) suggested that psychology could address racial health disparities by using CRT to conceptualize health disparities within historical, social, and contextual frames that acknowledge racism. While Volpe and colleagues focus on psychology as a field, it is also important for psychologists to consider how racial health disparities are discussed and deployed broadly.

We conducted a qualitative study where we asked people to share possible reasons for racial health disparities during COVID-19²2

This qualitative study relies on data that were originally collected for a larger research project. The project included quantitative data that were not analyzed in this study.

A sample of 298 U.S. Americans were recruited for a larger research project. We included 295 participants (M_Age = 38.1 years, SD = 14.4; range 18-77 years) who answered the key open-ended question. Psychology undergraduates (n = 42) received participation credit and Amazon Mechanical Turk workers (n = 253) received $1.25³3

We retained responses from these two sources because we wanted to consider a range of possible responses from US adults. The residential college sample provided access to a sample experiencing the COVID-19 pandemic often away from home, while the crowdsourced sample allowed for a wider age range. The college student (M = 3.81, SD = 1.74) and crowdsourced samples (M = 3.47, SD = 1.82) were similarly politically-oriented (near midpoint, but slightly liberal on 1-7 scale; t(290) = 1.12, p = .26).

After providing consent online, participants either responded to a prompt about disparities between Black and White Americans (n = 150) or disparities between Latinx and White Americans (n = 145). This open-ended prompt read:

Then, participants answered questions for the quantitative survey. Data were collected June through July 2020 (prior to vaccine development; coinciding with nationwide Black Lives Matter protests). Texas A&M University’s Institutional Review Board approved the study. For full text responses see Rivera et al., 2025S. Responses ranged from one word (e.g., “Racism”) to 181 words (e.g., a paragraph detailing five distinct reasons; M = 35.54 words, SD = 26.36 words). Participants often mentioned at least three reasons; even among the 29 responses that were 10 words or less, participants identified an average of two reasons, with 11 participants identifying three reasons.

We used a reflexive thematic analysis (Braun & Clarke, 2022) and a collaborative coding approach (Cornish et al., 2013; Saldaña, 2013). Philosophically, our analytic approach is aligned with constructivist perspectives in that we understand our participants, their responses, and ourselves as actively co-constructing meaning in relation to existing systems of power (e.g., Ponterotto, 2005). Although our data was collected within a positivist research paradigm (see General Discussion), we approached our analysis with relativist assumptions that there would be multiple meanings embedded in the lay-theories. Our goal was to extract meaning from the different lay-beliefs expressed with respect to our research questions. Using the systematic procedures of thematic analysis, we analyzed the data in six phases (Braun & Clarke, 2006, 2013).

Phase 1 included familiarization with the data by team members who made notes and observations about the data. After Phase 1, we decided the approach to coding would be primarily inductive and data-driven, and that we would consider each listed reason on its own, rather than focusing on each participant (i.e., one participant could generate multiple, seemingly incompatible lay-theories). We also decided to use collaborative coding where two non-independent coders worked together to sort and code each dataset. Coding teams were made up of insider-outside pairs (Cornish et al., 2013) with one member identifying with the group experiencing the disparity; Rivera and Noor coded participant responses to the Latinx-White disparities prompt, and Salter and Schlegel coded participant responses to the Black-White disparities prompt. Pairs were able to discuss responses from different vantage points as a part of the analytic process and whenever they came up, dual meanings or divergent interpretations were part of our analyses (e.g., “They stay together” could mean a living situation or they are always around each other). When notable divergent interpretations were generated within a coding pair, they were brought to the research team. When multiple interpretations were deemed meaningful, we included them in our analysis. In other instances, a shared interpretation emerged after discussion.

Phase 2 included three steps: (1) coding teams generated initial codes from the first third of respondents with QDA Miner-Lite software; (2) the research team met to discuss initial codes across the two coding teams; (3) the remaining two-thirds of responses were coded incorporating feedback and insights from the research team. In Phase 3, the research team met to sort the codes into potential themes and track overarching patterns. To audit our assumptions, values, and positionalities, we wrote and discussed what surprised us, intrigued us, and disturbed us about the codes as a part of our reflexive, analytic process. After the initial sorting was completed, individual team members added notes and brief descriptions of the potential themes. In Phase 4, coding teams checked whether the proposed thematic categories worked in relation to the data, and vice versa. Coding teams then drafted descriptions of the thematic categories and proposed names, drawing upon excerpts of data. Descriptions were discussed in team meetings and individuals provided feedback.

In Phase 5, we further developed the potential themes. First, team members considered RQ2 for each potential theme individually (i.e., writing their own notes), we then discussed and subsequently included these insights into our description of themes. Second, to codify defining features of each theme, we again checked tagged data against the overarching thematic categories. During our final write-up (Phase 6), our analysis remained iterative. We examined the Black/White and Latinx/White responses in conversation with one another; refined or pruned themes as needed; and where relevant, broadened our analysis from a descriptive to an interpretative level in relation to prior findings and public discourse.

Across both response sets, participants referenced a lack of behavioral compliance to CDC guidelines to explain the disparities (e.g., not social distancing, wearing masks, washing hands, nor seeking healthcare). This first theme was perhaps unsurprising given our collective understanding of COVID-19 as a socially communicable disease and Western tendencies to draw on internal attributions for others’ behaviors (e.g., Morris & Peng, 1994). If there are differential health outcomes between two groups, zeroing in on behaviors that have been publicized as increasing negative outcomes seems like a fairly straightforward application of attributing problems to something about the group rather than considering the context. Without critical interrogation, this fits with an individualistic understanding of health disparities seen in previous literature (e.g., McFerran & Mukhopadhyay, 2013).

However, through a critical analytical lens, these perceptions of differential compliance appear to run counter to empirical evidence. According to a nationally representative sample of the US, Black, Latinx, and Asian respondents were more likely to report mask-wearing in response to COVID-19 compared with White respondents (Hearne & Niño, 2022). Moreover, their results show White men were least likely to wear a mask from late April 2020 to early June 2020. This potential disconnect between perceptions and empirical evidence—and those highlighted in later themes—is one way these lay-theories can reflect and reproduce racism.

The following sub-themes offer insight into perceptions of where this perceived non-compliance stems from.

Theme two suggested beliefs that Black and Latinx communities were differentially impacted because biological vulnerabilities make them more susceptible to contracting the virus and experiencing adverse impact. The sub-themes offer insight into where the perceived biological vulnerabilities come from.

Theme three captures the social and cultural contexts in which Black and Latinx people were assumed to have increased risk. The sub-themes describe perceived cultural differences and included general references to “lifestyle choices” [White man, 51, Black response set].

Another theme referenced precarious working conditions in terms of increased exposure risk and longer-term risks associated with low wages and lack of benefits. Responses in this theme highlighted the entanglement of healthcare and employment in the U.S.

Regarding immediate risk, participants indicated that members of both communities were more likely to be essential, on-site workers (“many black people have blue collar jobs and were unable to work from home, therefore putting them on the "front line" and exposing them to greatest risk” [White woman, 45]). Other participants mentioned the role of poor working conditions in many service industry jobs (e.g., crowded, unsafe, lacking PPE; “Some Latins work in jobs that pack folks together like sardines” [White man, 58]). These responses are in line with research showing workers of color were overrepresented in risky “essential jobs” during COVID-19 (Rho et al., 2020).

Regarding longer term risks, participants tied working conditions to a lack of benefits, inadequate health insurance, and the struggle to afford insurance/medical treatment. One participant explained, “Lack of good jobs that provide paid sick days. Lack of health care, because many low-wage jobs don't provide it” [White woman, 60, Black response set].

Notably, some made explicit connections between working conditions and systemic racism,

The above quote acknowledged racism’s role in structuring working conditions, and therefore, health. However, many simply referred to working conditions as a base rate issue without such explicit connections, leaving acknowledgements of systemic racism implied or unsaid. We also noted within the Latinx response set, explicit references to particular occupations reproduced stereotypes via alignment with stereotypical media portrayals such as manual labor and maids (e.g., Reny & Manzano, 2016; “Many of the jobs they must do are in working with others (maids, food service, picking crops)” [White man, 56]) versus other high exposure occupations (e.g., nurses, doctors).

In both response sets, we saw explanations highlighting differences in access to resources, healthcare, information, and insurance benefits. One participant summed things up by saying, “Black people have less access to healthcare. They also have higher unemployment rates which denies them access to adequate insurance.” [White man, 77]. Participants indicated that even when members of Black and Latinx communities are able to access healthcare, there are differences in healthcare quality. Some responses linked financial resources to the quality of care received (“Most [Black Americans] cannot afford quality health care due their low income status” [Black man, 37]) and others linked these differences in healthcare quality to racism and discrimination (see Theme #6). Reduced access to COVID-19 supplies (e.g., PPE, testing, and treatment) also mattered. For example, “Latino communities are worst impacted by this virus because they don't have easy access to supplies so they have to constantly go out to get stuff.” [Latino, 19].

Responses also highlighted inadequate information resources in Latinx communities (“Lack of reliable communication about the virus and prevention” [White/Latino, 44]). Discussions included society’s failure to provide information in Spanish (“lack of materials in Spanish/Portuguese” [White woman, 51]). However, some responses discussed language barriers in a way that seemed focused on capabilities (“Latinos can't understand some things if they only speak spanish” [Asian man, 20]). Responses considering our societal responsibility to address language barriers seemed distinct from remarks about lacking education, understanding, and intelligence (see Theme #1).

Responses acknowledging social conditions, access, and structural barriers to good health align with calls to dismantle and address racism in the healthcare system. Overall, we were surprised by the centrality of structural mentions. We frequently returned to this theme because when responses mentioned “access” or “SES”, it was not always clear whether they were referring to structural racism, specifically. Acknowledging economic disparities can be anti-racist, but this acknowledgement can also reproduce racism if used as a counterpoint to anti-racist arguments (e.g., it is not race, but class driving health disparities: “There is some noise bubbling up about racism, but I am not too sure about that. But being poor can also mean poor insurance and therefore poor medical care” [White Man, 69, Black Response set]). An exclusively economic focus can reflect a stereotyped, essentialized view of communities of color (“they’re poor”) or ultimately fit within meritocratic, individualistic frameworks that continue to blame anyone who fails to overcome poverty as lacking the will (e.g., “poor decisions”). This aligns with prior work suggesting that efforts to address structural racism may be met with resistance by healthcare providers who highlight other social and behavioral determinants (e.g., access to healthcare, insurance, health literacy) as more valid explanations than racism and discrimination (Gollust et al., 2018).

Both response sets named racism in different forms (i.e., systemic racism, prejudice, racial bias, unconscious racism, discrimination). Some responses used individual-level frameworks, referring to racial inequalities in terms of who is prioritized for treatment or subjected to racial biases (“Latino people sometimes cannot get health care as easily as White people due to things like income or location or prejudice of medical staff” [White woman, 43]; “Also [Black Americans] are definitely discriminated against by healthcare workers, at least the racist ones,” [White man, 27]). Others explicitly linked disparities to historical forces or systemic racism (e.g., “The reasons have less to do with [Latinx Americans’] health and instead with the trickle down of systemic racism.” [Black woman, 20]. Salient responses directly named racism in a way that highlighted its impact on the health and well-being of Black and Latinx communities.

Sometimes responses alluded to contextual forces beyond the individual without naming racism explicitly (“This illness impacts Black communities more because historically Black communities are less wealthy than White communities” [Black-White Biracial woman, 48]). This response points out economic disparities between racial groups; but, naming racial differences in outcomes (as a state of affairs) and naming ‘racism’ (as an account of why those differences exist) are not exactly the same thing. Responses such as this acknowledge race as a central issue, but only imply that racism may be the problem. Notably, some participants acknowledged, but expressed skepticism, about racism attributions themselves (“Correlation vs. Causation. We can't be sure. Cultural? Less access to healthcare? Less personal responsibility? Genetic? Lower socio-economic conditions? Racism by doctors and nurses?" [White man, 57, Black response set]).

One important consideration is that our qualitative data come from a larger quantitative research project. The positivist methods used (e.g., large sample sizes, random assignment) do not necessarily align with constructivist approaches. However, our research question and philosophical assumptions were aligned with perspectives arguing for surfacing hidden meanings through interactive dialogue (Ponterotto, 2005). There are limitations to our approach (e.g., we could not probe participant responses; the format likely prompted short responses), but the heterogeneity of responses, iterative process of reflexive thematic analysis, and our collaborative coding approach facilitated a different sort of interactive dialogue within our coding pairs, during team analysis meetings, and within participant responses that considered one's own and others’ beliefs.

Because of the prompt, participants varied on whether they claimed beliefs as their own, positioned them as others’ beliefs, or did not say. While “others might say” was an initial subcode, it was not extracted as a salient pattern in the data by our team; but it is an interesting dimension to consider in future research. For our purposes, this prompt allowed participants to share possibly unfavorable beliefs and/or attributions they may have heard but disagreed with, providing a wider picture of the explanations for health disparities that exist in the US cultural milieu. Our goal was not to determine the most popular lay-theories, who might be most likely to produce different lay-theories, nor whether individuals are racist or not in their lay-theories. Our goal was to describe different ways people explain COVID disparities and examine what kinds of racialized discourse shows up in those explanations. While lay-theories may traditionally be thought of as individual-level mental processes and are typically studied in ways that prioritize methodological individualism (Adams & Stocks, 2008), our approach positioned lay theories as socially and culturally co-created. Although responses were brief, they corroborate, in participants' own words, different ways people make sense of health disparities.

Another consideration is that the prompt presented Black and Latinx communities separately when more than 1.3 million people in the U.S. identify as both. Future research could examine related lay-theories in Afro-Latinx communities or other minoritized communities. COVID-19’s negative impact on Asian American communities health-wise —in addition to the racism— was largely ignored (Yee, 2021) and Native Americans also experienced high incidence rates (Hatcher et al., 2020). When, how, and whose lay-theories are represented in psychological research is an important conversation in epistemic exclusion (Fryberg & Eason, 2017; Settles et al., 2020). Approximately 73% of our sample identified as White, and all of our participants were U.S. citizens with English as a first language. Identifying counter-narratives would benefit from analyzing lay-theories among different groups, and across multiple waves of the pandemic.

Finally, in the early waves of the pandemic, U.S. news outlets were generally focused on why People of Color were hit harder by the pandemic than White communities. The larger project associated with this study was designed to have a similar framing as these impactful news stories. Alternative approaches could have focused on explanations of why White Americans were faring better, likely producing differently racialized explanations (Powell et al., 2005).

Lay-theories are helpful tools for examining “mind-in-context” because they are both shaped by culture and have the power to shape culture (Salter et al., 2018). Studying the lay-theories individuals generate helps us understand conceptions of health disparities on a cultural level. Rather than treat the reasons generated by participants as a sample of individual beliefs, we consider these reasons as a sample of available narratives about racial health disparities in COVID-19 present within the US.

Lay-theories can feed into institutionalized public narratives and can sustain support for policies that either maintain or remedy racial injustice. They are entangled with the institutional practices and cultural narratives that are contributing to, if not causing, health inequalities themselves (Hook & Markus, 2020). While popular accounts described the ongoing global pandemic as “unprecedented,” the factors that contributed to the occurrence of COVID-19 disparities, and the racism-relevant attributions that people shared, were not “unprecedented” at all. There is a long history of systemic racism in healthcare and our analysis simply underscores the intersections of race, racism, and health during the pandemic.

A critical direction for researchers and practitioners who want to promote equity in healthcare is to address the questions of how individuals, groups, and societies explain the causes of health disparities. Understanding perceived causes provides much needed insight on the cultural ideas about Black and Latinx communities that may continue to function as barriers to equitable treatment. For example, if people believe COVID impacts Black and Brown “bodies differently than other races”, then it likely reduces the perceived need for systemic intervention or structural solutions. Some responses conveyed understandable frustration with the entanglement of employment, healthcare, and racial disparities. However, when does such frustration facilitate endorsing policy reform (e.g., universalizing healthcare) versus resigned acceptance in a culture of personal responsibility (i.e., work hard to get good jobs and health insurance)? As demonstrated by Senator Cassidy in the opening section of this paper, “experts” and policy makers are not immune to holding such theories (Leeman et al., 2011). It is crucial for scholars and practitioners to understand the lay-theories that they may need to challenge to mobilize resources for addressing racism as a barrier to equitable health outcomes.

Lay-explanations represent and reproduce narratives and explanations of COVID-19 disparities that are–at the very least–culturally and cognitively available among participants. Our qualitative approach highlights the utility of using critical, race-conscious approaches in psychological research. Without it, one could miss the ways in which lay-theories about COVID-19, and health disparities more broadly, resonate with existing racial frames and rhetoric. When viewed through a race-conscious lens, lay-theories become an important source of discussion for how racism is reflected and reproduced in everyday beliefs about health.