Introduction
This research note aims to revisit the subjective social position framework introduced by Martinez-Ebers, Branton, and Calfano (Reference Martinez-Ebers, Branton and Calfano2021) and originally used to predict differences in evaluations of policing. Specifically, we assess the framework’s applicability in another debated policy matter by examining Asian American and Latino attitudesFootnote 1 toward the government’s role in addressing income inequality. Extant research commonly notes the correlation between attitudes regarding policing and income inequality (e.g. Algan et al. Reference Algan, Guriev, Papaioannou and Passari2017; Foster and Frieden Reference Foster and Frieden2017). If the framework functions as found in Martinez-Ebers et al. (Reference Martinez-Ebers, Branton and Calfano2021), it seems reasonable to expect it should explain attitudes regarding redistributive policy. In essence, this is a test of the construct validity of the framework. The findings herein indicate those in lower subjective social positions are more likely to express support for governmental policy addressing income inequality than those in higher subjective social positions, which lends further support for the construct validity of the subjective social position framework.
A Subjective Social Position Framework
In its original formulation, the subjective social position framework synthesized three competing explanations for varying evaluations of police performance: ascribed characteristics, psychological mindset, and personal experiences. It successfully predicted more negative evaluations of police performance from persons with lower subjective social positioning. Subjective social position refers to the individual’s self-identified location in the social hierarchy and this plays a key role in how the individual perceives the world (Adler et al. Reference Adler2000; Brown-Iannuzzi et al. Reference Brown-Iannuzzi2015; Duman Reference Duman2020; Lindemann Reference Lindemann2007). An individual’s subjective social position not only depends on objective characteristics but also on how she experiences society and perceives her situation in comparison to others. Herein, using the framework, we expect persons in lower subjective social positioning will be more supportive of government intervention to assist those in poverty than those who see themselves in higher subjective positions.
Ascribed Characteristics
According to Omi and Winant (Reference Omi and Winant2015), minority classification is the primary factor animating social perception and relations. Though interracial marriage is gradually altering the racial hierarchy, Anglos remain at the top of the hierarchy, while African Americans remain at the bottom (Gans Reference Gans2012). Asians and Latinos are located somewhere between Anglos and African Americans, with Asians usually positioned closer to Anglos and Latinos positioned closer to African Americans (Bonilla-Silva Reference Bonilla-Silva2002).Footnote 2 However, Latinos are viewed by Whites as “more American” than Asians when immigrant populations increase (Fouka and Tabellini Reference Fouka and Tabellini2022). Class and gender distinctions are also important in social position construction, but in this application are secondary, particularly given Hero’s (Reference Hero1992) two-tiered pluralism argument about the US socio-political structure (with its lack of substantive equality for racial/ethnic minorities). We hypothesize people of color, especially Latinos who are closest to African Americans in the racial hierarchy, will be more supportive of redistributive policy than Asians but the difference should be small at best given their middling position in the racial hierarchy.
H 1 : Latinos will be more likely to agree it is the government’s responsibility for addressing income inequality than compared to Asian Americans.
Regarding the expectations for age, younger people are commonly seen as being more progressive or liberal while seniors are more conservative regarding the role of government. Indeed, studies show as individuals age they become significantly less supportive of redistributive policies (Lin, Kamo, and Slack Reference Lin, Kamo and Slack2018; Pittau, Farcomeni, and Zelli Reference Pittau, Farcomeni and Zelli2016). Theoretically, we expect age, to matter in assessments of the role of the government in addressing income inequality, even when other individual attributes are controlled for. Specifically, younger individuals who are at the lower end of the social order will be more supportive of redistributive policy than older people.
H 2 : Younger respondents will be more likely to agree it is the government’s responsibility for addressing income inequality than compared to older respondents.
Psychological Mindset
One’s psychological mindset includes feelings of social marginalization, the process of pushing a particular group or multiple groups to the edge of society by not allowing them an active voice, identity, or place in it. Through both direct and indirect processes, individuals are relegated to a secondary position and made to feel as if they are less important than those who hold more power or privilege in society (Kagan, Tindall, and Robinson Reference Kagan, Tindall and Robinson2010; Silva et al. Reference Silva, Esparza, Martinez-Ebers and Branton2022). This suggests individuals who feel marginalized—of not being accepted or respected by society—will be more supportive of policy helping a marginalized group, i.e. poor people.
H 3 : Socially marginalized individuals will be more likely to agree it is the government’s responsibility for addressing income inequality.
Personal Experiences
A key experiential factor associated with support for government action to address income inequality is personal experiences of income insecurity, such as a reduction or loss in wages or increased living expenses without an increase in salary (ILO 2016). Income insecurity poses challenges for daily life, such as the inability to pay for rent/mortgage, adequate food, utilities, healthcare, childcare, etc. Those who have experienced income insecurity are expected to be more supportive of the government addressing economic inequality than individuals who have not experienced income insecurity (Melcher Reference Melcher2023).
H 4 : Those who are income insecure will be more likely to agree it is the government’s responsibility for addressing income inequality than those who are not income insecure.
Data and Methods
To evaluate Asian American and Latino attitudesFootnote 3 regarding one’s perception of the government’s responsibility for addressing income inequality, we use data from the 2020 Collaborative Multiracial Post-Election Survey (CMPS). The CMPS includes interviews collected online in a respondent self-administered format between December 3, 2020 and February 15, 2020. The sample used in this study includes 3,003 Latinos and 3,006 Asians (Barreto et al. Reference Barreto2017). The survey was available in multiple languages including English, Chinese (simplified and traditional), Korean, Spanish, and Vietnamese. The CMPS features survey weights for the U.S. adult population based on age, ancestry, education, gender, nativity, and voter registration status.
The dependent variable measures respondent attitudes regarding the perceived responsibility of the government in addressing income inequality.Footnote 4 The respondents were asked: “It is the responsibility of the government to reduce the differences in income between people with high incomes and those with low incomes?” (Responses: 1-“Strongly Disagree,” 2-“Somewhat Disagree,” 3-“Neither Agree or Disagree,” and 4-“Somewhat Agree,” and 5-“Strongly Agree”). Given the dependent variable is based on a 5-point Likert scale measure, we utilize ordered logistic regression to estimate the models. Further, we estimate two sets of models to evaluate both intergroup and intragroup differences in public opinion regarding the government’s responsibility for addressing income inequality. First, the pooled model considers the differences in evaluations of the government’s responsibility for addressing income inequality across Asian Americans and Latinos. Second, the subgroup models examine the factors that predict evaluations of attitudes about the government’s responsibility for addressing income inequality among Asian Americans and Latinos.
In terms of the subjective social position framework, the key objective (i.e., ascribed) characteristics included in the pooled model are a respondent’s race/ethnicity and age. To account for racial/ethnic differences in evaluations of the government’s role in addressing inequality, we include a variable reflecting the respondents’ race/ethnicity category. The model includes a dichotomous variable denoting if a respondent is Latino with Asian serving as the baseline (or omitted) category. To account for the influence of age on attitudes about the government’s responsibility for addressing income inequality, both the pooled and subgroup models include a continuous measure of respondent age ranging from 17 to 108.
The key psychological factor is marginalization, which is measured using the following four items: “How much do you believe that you belong in the United States?” “How much do you believe you are an insider/outsider in the United States?”Footnote 5 “How much you believe that others in U.S. society value and respect you?” and “How much you believe that you are accepted and included by others in U.S. society who are not like you?” The response options for each item range from 1 (“A lot”) to 3 (“Not at all”), where higher values reflect greater marginalization. We factored and analyzed the four items and then created the index using the factor scores.Footnote 6 The factor score-based marginalization index ranges from −1.20 to 2.48, with higher values reflecting greater levels of marginalization from society. The Cronbach alpha score containing all four measures of marginalization is .72.
The key experiential measure is a respondent’s income insecurity. Income insecurity is measured using responses to the following item: “Please indicate whether any of the circumstances apply to you or someone in your household”: (1) lost job, (2) Had work hours cut, or pay cut, but kept job, (3) Had to temporarily or permanently close my business or my self-employment, (4) Currently unemployed and looking for work (5) Lost access to healthcare, and (6) Was not able to pay monthly rent or mortgage. The response options for each item are (1) “Yes, me,” (2) “Yes, someone else in my household,” and (3) “No, nobody in my household.” We factored and analyzed the six items and then created the index using the factor scores.Footnote 7 The factor score-based income insecurity measure ranges from −.63 to 5.91, with higher values reflecting greater levels of income insecurity. The Cronbach alpha for the income insecurity measure is .73.
The models also include five individual-level control variables: female, education, income, partisanship, foreign-born, and political interest. Female is coded 1 if the respondent is female. Education is measured using a categorical variable ranging from 1 (eighth grade or less) to 4 (post-graduate education). Income is measured using a categorical variable ranging from 1 (less than $20,000) to 12 ($200,000 or more). The model also includes No Income, which is coded “1” if respondents did not report their income. Partisan affiliation is measured by two binary variables: Democrat and Independent (coded 1 in each case). Republican identification serves as the baseline (omitted) category. Foreign-born is coded 1. Political interest is measured by: “Some people are very interested in politics while other people can’t stand politics, how about you? Are you: Very interested in politics, Somewhat interested, Not that interested in politics, Not at all interested in politics.” Higher values reflect heightened interest in politics.
Results
Given the ordered categorical nature of the dependent variable, we use ordered logit to estimate the income inequality models.Footnote 8 The results for the pooled model are presented in Table 1 and the results for the subgroup models are presented in Table 2.
Table 1. Government role in addressing income inequality: Pooled model
*** p < 0.001;
** p < 0.01;
* p < 0.05. Coefficients are ordered logistic regression coefficients with standard errors clustered on the zip code.
Table 2. Government role in addressing income inequality: Subgroup models
*** p < 0.001;
** p < 0.01;
* p < 0.05. Coefficients are ordered logistic regression coefficients with standard errors clustered on the zip code.
Pooled Model
The first column in Table 1 presents the coefficients; the second column shows the standard errors. We begin by discussing the impact of the first of the objective (i.e., ascribed) characteristics—respondent race/ethnicity—on attitudes about the government’s responsibility for addressing income inequality. The findings indicate there are slight, yet significant differences across racial/ethnic groups. Asian Americans are significantly less likely to disagree and significantly more likely to agree it is the government’s responsibility to address income inequality than Latino respondents, although the differences are marginal. The probability of an Asian American strongly disagreeing it’s the government’s responsibility to address income inequality is .08, while the probability for a Latino is .09 (∆ = −.01). The probability of an Asian American strongly agreeing it’s the government’s responsibility to address income inequality is .19, while the probability for a Latino is .17 (∆ = .02). Substantively, this finding, while muted, is not consistent with H1.
Next, we turn to the other variables that are key components of the subjective social framework. To assist with the interpretation, we present the change in predicted probabilities for each of the remaining variables that comprise the subjective social framework (holding all other variables at their means) as they increase from their minimum to maximum values in Figure 1. The top left graph presents the probabilities for the second ascribed factor: age. The top right graph presents the probabilities for the first psychological factor: marginalization. The bottom left graph presents the probabilities for the experiential factor: income insecurity. The lines with the circle represent disagree/strongly disagree and the lines with the square symbol represent agree/strongly agree the government is responsible for addressing income inequality.Footnote 9
Figure 1. Subjective social framework: Pooled model.
First, we consider the impact of the second objective (ascribed) characteristic—respondent age—on attitudes about the government’s responsibility for addressing income inequality. Older respondents are significantly more likely to disagree and significantly less likely to agree it’s the government’s responsibility to address income inequality, lending evidence for H2. As age increases from minimum to maximum value, the probability of agreeing it’s the government’s responsibility to address income inequality decreases from .56 to .26 (∆ = −.30). Further, as age increases from minimum to maximum value, the probability of disagreeing it’s the government’s responsibility to address income inequality increases from .16 to .41 (∆ = .25).
Next, the findings indicate the psychological factor—marginalization—is associated with attitudes about the government’s responsibility for addressing income inequality. Individuals who feel the most socially marginalized are significantly less likely to disagree and significantly more likely to agree the government is responsible for addressing income inequality, lending evidence to support H3. For example, as marginalization increases from minimum to maximum value, the probability of agreeing it’s the government’s responsibility to address income inequality increases from .44 to .54 (∆ = .10). Further, as marginalization increases from minimum to maximum value, the probability of disagreeing it’s the government’s responsibility to address income inequality decreases from .24 to .17 (∆ = −.07).
Finally, the results indicate the experiential factor—income insecurity—influences attitudes about the government’s responsibility for addressing inequality. Respondents with greater income insecurity were significantly more likely to agree and significantly less likely to disagree the government is responsible for addressing income inequality than compared to respondents with less income insecurity, lending evidence to support H4. Specifically, as income insecurity increases from minimum to maximum value the probability of agreeing it’s the government’s responsibility to address income inequality increases from .46 to .55 (∆ = .09), while the probability of disagreeing it’s the government’s responsibility to address income inequality decreases from .22 to .15 (∆ = −.07).
Subgroup Models
We now turn to the subgroup models in Table 2 to determine if the subjective social position influences attitudes about the role of the government in addressing income inequality within racial/ethnic groups. The first two columns reflect the results for the Asian American respondents and the last two columns reflect the results for Latino respondents.
Asian American Respondents
Figure 2 presents the change in probabilities across the income inequality measure for each of the remaining variables that comprise the subjective social framework (holding all other variables at their means) as they increase from their minimum to maximum values for Asian Americans. The layout of the graphs and the symbols are consistent with the layout of Figure 1.
Figure 2. Subjective social framework: Asian American model.
First, we consider the impact of the (ascribed) characteristic—respondent age—on attitudes about the role of the government in addressing income inequality. Older Asian respondents are significantly more likely to disagree and significantly less likely to agree it’s the government’s responsibility to address income inequality, lending evidence to support H2. For example, as age increases from minimum to maximum value, the probability of agreeing it’s the government’s responsibility to address income inequality decreases from .57 to .25 (∆ = −.32). Further, as age increases from minimum to maximum value, the probability of disagreeing it’s the government’s responsibility to address income inequality increases from .16 to .43 (∆ = .27).
Next, the findings indicate the psychological factor—marginalization—is associated with Asian Americans’ perceived role of government in addressing income inequality. Asians who feel the most socially marginalized are significantly less likely to disagree and significantly more likely to agree it’s the government’s responsibility to address income inequality, lending evidence to support H3. For example, as marginalization increases from minimum to maximum value, the probability of an Asian respondent agreeing it’s the government’s responsibility to address income inequality increases from .43 to .55 (∆ = .12). Further, as Asian American marginalization increases from minimum to maximum value, the probability of disagreeing it’s the government’s responsibility to address income inequality decreases from .24 to .18 (∆ = −.06).
Finally, the results indicate the experiential factor—income insecurity—is not associated with Asian American attitudes about the government’s responsibility for addressing income inequality. This indicates that Asian American perception of the attitudes about the government’s responsibility for addressing income inequality is not associated with the experiential component of the subjective social framework, lending no evidence to support H4.
Latino Respondents
Next, we turn to the results for Latino respondents, which are presented in Figure 3.
Figure 3. Subjective social framework: Latino model.
First, we consider the impact of the (ascribed) characteristic—respondent age—on attitudes about the role of the government in addressing income inequality. Older Latino respondents are significantly more likely to disagree and significantly less likely to agree it’s the government’s responsibility to address income inequality, lending evidence to support H2. For example, as age increases from minimum to maximum value, the probability of a Latino agreeing it’s the government’s responsibility to address income inequality decreases from .55 to .31 (∆ = −.24). Further, as age increases from minimum to maximum value, the probability of a Latino disagreeing it’s the government’s responsibility to address income inequality increases from .15 to .34 (∆ = .18).
Next, the findings indicate the psychological factor—marginalization—is associated with Latinos’ attitudes about the role of the government in addressing income inequality. Latinos who feel the most socially marginalized are significantly less likely to disagree and significantly more likely to agree it’s the government’s responsibility to address income inequality, lending evidence to support H3. As marginalization increases from minimum to maximum value, the probability of a Latino respondent agreeing it’s the government’s responsibility to address income inequality increases from .43 to .55 (∆ = .12). Further, as Latino marginalization increases from minimum to maximum, the probability of disagreeing it’s the government’s responsibility to address income inequality decreases from .23 to .16 (∆ = −.07).
Finally, the results indicate the experiential factor—income insecurity—influences Latino attitudes about the government’s responsibility for addressing income inequality. Latino respondents with greater income insecurity are significantly more likely to agree and significantly less likely to disagree the government is responsible for addressing income inequality than compared to those with less income insecurity, lending evidence to support H4. Specifically, as income insecurity increases from minimum to maximum value the probability of a Latino respondent agreeing it’s the government’s responsibility to address income inequality increases from .45 to .63 (∆ = .18), while the probability of disagreeing it’s the government’s responsibility to address income inequality decreases from .22 to .12 (∆ = −.10).
Conclusions
Our study advances the literature in three ways. First, we show that the subjective social framework effectively explains individual attitudes beyond evaluations of local police. This framework highlights how ascribed characteristics, psychological mindset, and personal experiences shape one’s views on the government’s responsibility for addressing income inequality. By focusing on personal perspectives rather than just solely on objective conditions, this approach provides a deeper understanding of how attitudes are formed.
Second, we observe differences in the applicability of the subjective social framework in predicting Latino and Asian American perceptions of the federal government’s role in addressing income inequality. All three components—ascribe, psychological, and experiential—work in the expected manner for Latinos, while only the ascribe and psychological components function as expected for Asian Americans. This difference prompts questions about the underlying reasons for these variations, particularly whether there are within-group differences among Asian American respondents that contribute to the findings.
Third, this study contributes to the research on public support for redistributive policies. Despite decades of rising income inequality, the public remains largely opposed to greater redistribution, likely due to a lack of knowledge or indifference (Condon and Wichowsky Reference Condon and Wichowsky2019; Shaw and Gaffey Reference Shaw and Gaffey2012). Research finds concern about inequality has a trivial impact on beliefs in economic meritocracy (Hayes Reference Hayes2014; Trump Reference Trump2018; Trump and White Reference Trump and White2018). However, Condon and Wichowsky (Reference Condon and Wichowsky2019) suggest support for redistribution is contingent on one’s social standing relative to others, indicating factors associated with attitudes regarding income inequality may extend beyond objective factors. We propose applying a subjective social framework, which includes additional elements, that could provide a more comprehensive explanation of attitudes regarding the government’s role in addressing inequality.
In future research, we hope to include all major ethnic/racial groups in our analysis to further test the applicability of subjective social framework. Future research should also further consider the intragroup applicability of the framework regarding attitudes toward the government’s role in addressing income inequality. Extant research highlights differences within the Asian American population with regard to public opinion, which may emerge in the applicability of the subjective social framework (Wong et al. Reference Wong, Ramakrishnan, Lee, Junn and Wong2011). Finally, future search should also seek to test our framework on more policy topics and dealing with different levels and functions of government.
Funding statement
This research did not receive any specific financial support.
Competing interests
None.
Appendix A: Government Role in Addressing Income Inequality: Including White and Black Respondents
The model presented in Appendix A presents the results of the model predicting attitudes about the role of the government in addressing income inequality including African American and White respondents. Note, the model does not include the marginalization measure as the 2020 CMPS did not ask African American and White respondents the related questions.
The results indicate there are significant differences in attitudes regarding the role of the government in addressing income inequality as a function of the ascribed factors of the subjective social framework. For instance, uniformly minority respondents are significantly more likely to agree and significantly less likely to disagree the government is responsible for addressing income inequality than compared to White respondents. Further, African American and Asian American respondents are significantly more likely to agree and less likely to disagree the government is responsible for addressing income inequality than compared to Latino respondents. Finally, there is no difference in attitudes about the role of the government in addressing income inequality when comparing African American and Asian American respondents. Additionally, the findings indicate older respondents are significantly less likely to agree and significantly more likely to disagree it is the government’s responsibility to address income inequality than compared to younger respondents.
Finally, the findings indicate there are significant differences in attitudes regarding the role of the government in addressing income inequality as a function of the experiential factor of the subjective social framework—income insecurity. Substantively, the findings indicate those who experience income insecurity are significantly more likely to agree and significantly less likely to disagree the government is responsible for addressing income inequality than compared to those did not experience income inequality.
Taken together, the findings in the model including African American and White respondents are largely consistent with the results presented in Table 1 which includes only Asian American and Latino respondents.
Appendix B: Descriptive Statistics
Appendix C: Government Role in Addressing Income Inequality: OLS Model
The models presented in Appendix C present the OLS pooled and subgroup models predicting attitudes about the role of the government in addressing income inequality. Substantively and statistically, the results are consistent with the ordered logit results presented in Tables 1 and 2.
***p < 0.001; **p < 0.01; *p < 0.05. Coefficients are OLS regression coefficient with standard errors clustered on the zip code.
Open access
