Introduction
In the 2020 presidential election, voter turnout increased to unprecedented levels despite potential threats to the voting booth for many Americans (Morris and Grange Reference Morris and Grange2021). This effect could largely be due to the increase in mail-in ballot options that were offered to registered voters in select states that saw an increase in turnout for all voters by 4 percent on average (McGhee, Paluch, and Romero Reference McGhee, Paluch and Romero2021). Even more notably, there was an increase in voter turnout for Black as well as non-college-educated Whites (Frey Reference Frey2021; Morris and Grange Reference Morris and Grange2021). However, historical gaps between Black and White voter turnout continue to persist at around a 7-point difference since the 2012 election. Given that there are a multitude of factors that could be contributing to this persistent turnout gap between Black and White Americans, it raises the question of how models within political science measure the factors that influence and affect Black voter turnout in today’s political world. These factors include structural factors such as restrictive voting laws and access to socioeconomic resources, but often do not include psychological factors such as resilience or apathy.
The study of Black political behavior has a long history of exploring the disparate nature of Black political participation and how it is distinct from that of White populations. Black political behavior is unique because of its strong belief in group consciousness and linked fate regardless of socioeconomic status (SES) (Dawson Reference Dawson1994). It is also true when it comes to conceptions of political ideology where scholars find that Black voters’ preference for the Democratic party cannot be explained by SES or ideology measured on a liberal-conservative scale because racial group identity supersedes ideological sorting (Abramowitz Reference Abramowitz2010; Hajnal, Lajevardi, and Nielson Reference Hajnal, Lajevardi and Nielson2017; Jefferson Reference Jefferson2020; Philpot Reference Philpot2017). In addition, it is found to be true in the conception of group efficacy, where scholars find racial group efficacy to be a stronger predictor of participation for Black voters over any other racial group (D. Phoenix and Chan Reference Phoenix and Chan2019; D. L. Phoenix and Chan Reference Phoenix and Chan2022). Considering how these factors that are said to influence voter turnout in the extant literature differ for the Black community, this work considers if traditional models hold in recent elections since 2008.
Considering this, I ask: In what ways do prominent models of political behavior explain the racial voting gap between Black and White voters? In addition, how do these turnout theories hold in a post-Obama political context? This research note argues that differences in Black turnout in a post-Obama context are due to a lack of significant political change for Black Americans after the Obama presidency that reduced the impact of Black politics (Gilliam and Kaufmann Reference Gilliam and Kaufmann1998; F. Harris Reference Harris2012; Spence and McClerking Reference Spence and McClerking2010). Specifically, it improves upon and updates Leighley and Vedlitz’s (Reference Leighley and Vedlitz1999) work, which challenged the notion that preexisting models of political behavior are equally effective regardless of race. This note argues for greater theorizing around Black voter turnout in the modern political era to better understand the increasing Black-White turnout gap as not just based on structural factors, but also attitudinal shifts about politics. I argue that there have been major socializing events that particularly impact Black individuals’ political psychology as well as turnout in the post-Obama political context. These events include the Shelby County v. Holder decision and resulting voter suppression, the Protest for Black Lives in response to the death of George Floyd, and the disproportionate impact of a global pandemic in 2020 that have not been sufficiently acknowledged in the extant literature about Black political participation (Hajnal, Lajevardi, and Nielson Reference Hajnal, Lajevardi and Nielson2017; Kuk, Hajnal, and Lajevardi Reference Kuk, Hajnal and Lajevardi2022; Reyes Reference Reyes2020).
In 1999, Leighley and Vedlitz posed a similar inquiry as they suggested that various models present a distinction between White and non-White voter turnout have not been extensively empirically tested (Leighley and Vedlitz Reference Leighley and Vedlitz1999). To begin to empirically test these models, the authors evaluate five models of political participation using a Texas-based survey and find that the models work differently among each of the racial groups and aid in the explanation of differences in turnout for the different racial groups. My study seeks to employ their methods to assert findings about Black and White turnout nationally in today’s political world.
Using multivariate models of voter turnout, this research note seeks to update Leighley and Vedlitz’s (Reference Leighley and Vedlitz1999) findings using a mini meta-analysis of regression coefficients from national data collections that used self-reported voter turnout data. Mini meta-analysis summarizes effect sizes of a small number of studies to glean more succinct and persuasive trends, which I argue is useful to analyze Black and White turnout from the American National Election Study (ANES) and Collaborative Multiracial Post-election Survey (CMPS) in 2016 and 2020. These new findings will contest and contribute to Leighley and Vedlitz as well as provide support for the argument that there needs to be alternative explanations for turnout for the Black electorate.
Established Predictors of Black-White Turnout
Socioeconomic Status (SES)
Socioeconomic status (SES) theory suggests that those with higher levels of educational attainment and income are more likely to participate (Verba et al. Reference Verba, Schlozman, Brady and Nie1993). For decades, much of the existing literature on Black political participation finds that accounting for socioeconomic factors such as income and education, Black voters participate in some political activities as much or at even higher rates than non-Hispanic Whites (Krogstad and Lopez Reference Krogstad and Lopez2017). In scholarship specific to Black voters, for example, Dawson found that regardless of SES, members of the Black electorate still vote similarly (Dawson Reference Dawson1994). This example raises the question of whether SES increases turnout among both White and Black voters in a similar way. These racial differences have often been attributed to theories of racial group identity, like Dawson’s theory of linked fate, which notes that racial solidarity is as impactful for African American turnout as socioeconomic resources.
Psychological Resources
Leighley and Vedlitz evaluated the theory of psychological resources using measures of political efficacy and political interest and found that increased levels of psychological resources resulted in greater turnout amongst all racial and ethnic groups within their sample. Voters with lower psychological resources have been found to view elected officials as effective or active in making the policy decisions that are important to them or their community. However, these voters also believed that politicians just do what they need to do to get elected and dislike how candidate-centered elections have become rather than issue-oriented (Dalton Reference Dalton2015; Hill Reference Hill2006).
Efficacy has been discussed as the belief that an individual’s actions can bring about the desired outcome (Almond and Verba Reference Almond and Verba1963; Campbell et al. Reference Campbell, Converse, Miller and Stokes1960; Campbell, Gurin, and Miller Reference Campbell, Gurin and Miller1954). It has been conceptualized as either internal or external, where internal efficacy speaks to an individual’s ability to approach the complex nature of politics to engage effectively, while external efficacy looks at the government being responsive to an individual’s political demands (Campbell and Converse Reference Campbell and Converse1972; Verba and Nie Reference Verba and Nie1987). Early studies discuss efficacy as static and related to individual factors such as SES, race, and gender, while others point to it being environment-dependent (Abramson Reference Abramson1972; Campbell et al. Reference Campbell, Converse, Miller and Stokes1960).
Regarding political interest, early studies found that increased ease of political information increases political interest, but only among a small group of people use this information to constantly update their policy positions (Campbell et al. Reference Campbell, Converse, Miller and Stokes1960; Downs Reference Downs1957). In addition, these early scholars utilized cross-sectional data to find that interest in politics increases as young adults grow older and become middle-aged (Glenn and Grimes Reference Glenn and Grimes1968). Average political interest is said to increase as people continue to age and is correlated with political participation (Glenn and Grimes Reference Glenn and Grimes1968). Moreover, within the United States, aggregate political interest was found to be more stable, especially within the 1980s and 1990s (Prior Reference Prior2010). Within the age of the Internet, young and old people alike report higher levels of political interest after viewing websites that they deem as effective in sharing political information (Lupia and Philpot Reference Lupia and Philpot2005). However, young and old respondents differ in which website they evaluate as more effective. Higher levels of political interest are said to influence knowledge about politics, make a person more likely to vote, and more likely to be politically active outside of voting (Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995). Moreover, those who are more politically interested tend to respond more to mobilization efforts and attempts to increase political participation (Brady, Schlozman, and Verba Reference Brady, Schlozman and Verba1999).
Specifically, looking at people of color, scholars find that Black respondents have lower rates of political efficacy in comparison to White respondents (Cohen Reference Cohen2012). In addition, Dawson presents the Black utility heuristic, which discusses how Black Americans use racial group identity as a proxy for political attitudes and preferences (Dawson Reference Dawson1994). Similarly, Phoenix and Chan report that when looking at measures of racial efficacy and political participation that non-White respondents who have increasing levels of racial efficacy engage in more traditional political action as well as system-challenging action (D. L. Phoenix and Chan Reference Phoenix and Chan2022).
Social Connectedness
A wide range of scholars from interdisciplinary fields discuss the importance of social interaction and connection to politics, including alienation, trust, prejudice, and apathy (Allport Reference Allport1954; Dean Reference Dean1960; Feldman Reference Feldman1986; Lane Reference Lane1959). Teixeira identifies that, how voter apathy in the past twenty years is correlated with a lack of connection between individuals and social and political organizations (Teixeira Reference Teixeira2011). He specifically turns to the decline of the political parties’ influence. When this theory of social connectedness is observed with race in mind, there are a few studies that focus on social connectedness specifically in Black communities.
When Black communities are looked into, scholars find that participation in a Black church or a Black political organization leads to increased political participation (F. C. Harris Reference Harris1994; Liu, Austin, and Orey Reference Liu, Wright Austin and D’Andrá Orey2009; Tate Reference Tate1993). Cohen (Reference Cohen2012) finds that despite participation, Black young people feel strong feelings of political alienation. Political alienation is conceptualized with three dimensions: government orientation, political community, and equal opportunity. The early socialization of being distrustful of the government as well as the pathologizing of Black youth leads to greater feelings of these dimensions (Cohen Reference Cohen2012).
Group Consciousness
Group consciousness is defined as the awareness among a group of their shared status as an unjustly deprived and oppressed group (Shingles Reference Shingles1981). It is directly linked to the concept of linked fate, which suggests that Black individuals formulate policy and voting decisions based on how it would impact the improvement of the Black community as a whole (Dawson Reference Dawson1994; Sanchez and Vargas Reference Sanchez and Vargas2016). It also leads to increased political participation of Black Americans (Dawson Reference Dawson1994; Sanchez and Vargas Reference Sanchez and Vargas2016). This perspective has proven true because the shared racial history of Black Americans exists across economic groups in ways not seen in other racial groups (Dawson Reference Dawson1994). Recent scholarship, however, has questioned the measurement of linked fate and group consciousness and shows generational fluctuations in the levels of linked fate over time (Gay, Hochschild, and White Reference Gay, Hochschild and White2016; C. W. Smith et al. Reference Smith, Bunyasi and Smith2019; J. C. Smith et al. Reference Smith, Clemons, Krishnamurthy, Martinez, Mclaren and White2023).
Scholars like Ashley Jardina also argue that White Americans have a racial group-based identity that is not associated with racism (Jardina Reference Jardina2019). Instead, it is associated with maintaining status among other racial groups. These models remain dominant within political science literature and should be evaluated and updated to better understand the current political sphere, more specifically, to improve the understanding of the Black political sphere.
Hypotheses
The previous paper hypothesized that the aforementioned five predictors of voter turnout are expected to result in increased turnout in all the racial groups assessed. Recent research has continued to discuss predictors of minority political behavior and has similar findings (Fraga Reference Fraga2018). However, Fraga (Reference Fraga2018) argues that socioeconomic factors are posttreatment characteristics that develop as a result of race. Similar to Fraga, I hypothesize that the aforementioned predictors are not effective for Black individuals because they are largely influenced by race. This research note suggests that race has only become more salient in recent elections, therefore resulting in less of an effect of traditional predictors of turnout on voting. In addition, it argues that more research needs to be done on the predictors of Black voting that is specific to Black individuals because of their place in the American racial hierarchy (Kim Reference Kim1999; Omi and Winant Reference Omi and Winant2014; Zou and Cheryan Reference Zou and Cheryan2017) in relation to White individuals and the differential effects of these characteristics on voting and mobilization specific to Black individuals (Davis Reference Davis2023; Fraga Reference Fraga2018; Scott, Michelson, and DeMora Reference Scott, Michelson and DeMora2021; Southwell and Pirch Reference Southwell and Pirch2003).
Moreover, I hypothesize that traditional predictors of voter turnout are no longer effective, specifically in a post-Obama context because of recent racialized experiences in the United States like the Black Lives Matter movement and voter suppression. Previous research has found that race has differing impacts on electoral politics depending on geography and demographics (Carmines, Huckfeldt, and McCurley Reference Carmines, Huckfeldt and McCurley1999; Fraga Reference Fraga2018). However, I argue that recent political events in Black politics have altered this in recent elections, with race becoming more salient regardless of geography and demographics. I argue that the experience of protesting against police brutality and recognizing the disparate impact that voter suppression has on Black Americans socialized Black individuals to understand that their political experience remains distinct from that of Whites and has effectively demobilized some Black Americans to participate in voting. Although voter suppression and police brutality have been discussed as issues for the Black community previously, I argue that the impact of these events are different as they received nationwide and even global attention and because they occurred after the presidency of Obama, which was often described as the beginning of a post-racial society (ACLU 2021; Diallo and Shattuck Reference Diallo and Shattuck2020; Kennedy Reference Kennedy2016; Roth and McCracken Jarrar Reference Roth and McCracken Jarrar2021).
Instead of the post-racial society that was suggested after the Obama presidency, race has become even more salient in American politics and has made it more evident that the Black political calculus is distinct from that of White individuals. Related to Phoenix (Reference Phoenix2019), who finds that political anger produces different psychological effects in Black and White individuals, I argue that Black individuals are identifying with this distinct political calculus in experiencing voter suppression, mass protest, a global pandemic, and a lack of political change in the years since the Obama presidency. I argue that this distinct experience will result in differential effects in traditional turnout models and create a need for new or updated conceptualizations of Black political behavior in today’s political world.
The prominent predictors of voter turnout are income, education, political interest, political efficacy, length of residence, marital status, group threat, group closeness, and nativity. Considering that White individuals’ political calculus is distinct from Black individuals, I hypothesize that for White individuals, seven or the eight predictors—income, education, political interest, political efficacy, length of residence, marital status, and group threat—will have a more significant and positive relationship on turnout in comparison to Black individuals (H1). These factors have been traditionally established as pertinent to political behavior, especially for those who have had their political grievances acknowledged throughout American history. On the other hand, given their distinct psychological calculus due to anti-Black racism, I hypothesize that these seven prominent predictors will have less of an effect on Black turnout. Instead, for Black individuals, I hypothesize that one key predictor, group consciousness will have a more positive and significant effect on Black turnout in comparison to White individuals (H2). I expect this to be true given preexisting research on racially linked fate (Dawson Reference Dawson1994).
These hypotheses were preregistered via as predicted (#104290).
Methods
In Leighley and Vedlitz’s paper, the data used to evaluate the five aforementioned models of voter turnout—SES, psychological resources, social connectedness, and group consciousness—were taken from a Texas-based survey that included an oversample of African Americans, Mexican Americans, and Asian Americans. Although other explorations of Black-White turnout have been conducted, I employ Leighley and Vedlitz’s methods because it allows me to extensively compare several prominent predictors of voter turnout rather than just limiting it to specific predictors (Fraga Reference Fraga2018; Philpot, Shaw, and McGowen Reference Philpot, Shaw and McGowen2009; Schneider and Robnett Reference Schneider and Robnett2015). In addition, the combination of their methods and the usage of mini meta-analysis strengthens the findings and understandings of Black eligible voters.
Their dependent variables were a measure of self-reported voter turnout. To conduct an analysis similar to this paper and determine the state of prominent political behavior models, I used a sample from the American National Election Study from 2016 and 2020 and the Collaborative Multiracial Post-election Survey (CMPS) 2016 and 2020 of only Black and White respondents. The sample sizes included: the ANES 2016 White sample (n = 2,429), CMPS 2016 White sample (n = 810), ANES 2020 White sample (n = 5,442), CMPS 2020 White sample (n = 3,150), ANES 2016 Black sample (n = 322), CMPS 2016 Black sample (n = 2,033), ANES 2020 Black sample (n = 656), CMPS 2020 Black sample (n = 2,607) (American National Election Studies 2017; American National Election Study 2021; Frasure et al. Reference Frasure, Wong, Vargas and Bareto2021, Reference Frasure, Wong, Vargas and Barreto2024).
Bivariate models of voter turnout were conducted that replicate those as closely as possible to the ones reported in Leighley and Vedlitz (Reference Leighley and Vedlitz1999). After running those models, a mini meta-analysis of these coefficients was conducted to detect any summary trends across datasets (Goh, Hall, and Rosenthal Reference Goh, Hall and Rosenthal2016). The dependent variable in this study is self-reported turnout or validated turnout to replicate Leighley and Vedlitz (Reference Leighley and Vedlitz1999), who used voter turnout in the 1992 Presidential election as well as an index of political activities within the last 12 months. Self-reported turnout involved variables that asked respondents to answer if they voted in either the 2016 or 2020 Presidential elections.
The mini meta-analysis in this paper employs Goh, Hall, and Rosenthal’s (Reference Goh, Hall and Rosenthal2016) approach, which is typically used when there are only a few surveys available. To conduct this mini meta-analysis, the effect size and the sample size for each model were collected to determine correlations. After correlations were determined, they were converted into Cohen’s D values to display the magnitude of effect, where values of .20, .50, and .80 are perceived as small, medium, and large effects, respectively. D values are standardized mean differences that are measured in standard deviation units. For example, a D value of .5 means that the variable is associated with 1/2 a standard deviation shift in the dependent variable. A mini meta-analysis was useful for this research to discover trends across different election years and racial groups to compare to preexisting research on voter turnout. It also allows comparison between different data collections that feature the same variables of interest to begin to make larger theoretical claims.
As far as variable operationalization, I replicated the project as much as possible using the ANES and CMPS datasets. Within the SES model, educational attainment and household income were utilized in both Leighley and Vedlitz (Reference Leighley and Vedlitz1999) and this paper. For the psychological resources models, political interest and political efficacy were measured with a four-point self-report and a five-point self-report, respectively. Social connectedness was measured with a dichotomous variable for marital status, years living in the community, measuring the length of residence, and a dichotomous variable for home ownership. Group consciousness was observed through group closeness and intergroup threat. In Leighley and Vedlitz (Reference Leighley and Vedlitz1999), group closeness was measured with a 5-point scale of closeness to an ethnic group, and intergroup distance was asked similarly just for other ethnic groups. For this paper, group closeness was measured by how much the respondent’s life was affected by what happened to their racial group. For White respondents, the feeling thermometer about White respondents people was measured, and the same was done for Black respondents. To prevent issues that arise with the usage of feeling thermometers, like a lack of similar reference points, the intergroup distance measure included how one group rated the other group minus the rating of their own group. The group conflict model was perhaps the most difficult to replicate, with the original models including group threat, being born in the United States, and speaking English at home. The replication includes the group threat measure in a similar way to Leighley and Vedlitz (Reference Leighley and Vedlitz1999) by observing the percentage of ethnic groups other than the respondents in the zip code. Specifically, from the CMPS 2016 and 2020, there was a measure that identified the estimated percentage of Black and White respondents in the respondent’s zip code. The data from those born in the United States measures came from the question asking where the respondents were born.
Results
Replication Models
To best evaluate the models, full models of voter turnout were conducted within each dataset within this analysis. From the 2016 CMPS and 2016 ANES model (Appendix A), homeownership is statistically significant despite only being positive and statistically significant among Asian Americans in the original paper. These findings support the more recent investigations about homeownership and political participation, which also find that homeownership increases political participation (Fischel Reference Fischel2001; Hall and Yoder Reference Hall and Yoder2022; Leighley and Nagler Reference Leighley and Nagler2013). However, when comparing the 2016 ANES and 2016 CMPS, homeownership has a statistically significant and positive effect on White and Black turnout the White group threat measure, but statistically significant negative effects for both groups in the 2016 CMPS. These disparate findings for the same election year support a meta-analysis because it can ensure precision in the understanding of the effect of homeownership on participation among Black and White respondents in 2016.
Among the White samples of the 2016 ANES and 2016 CMPS, the White group threat is positive and statistically significant in the 2016 ANES. Scholars have recently explored the impact of White racial identity and it having a mobilizing effect among White Americans, especially in the 2016 election (Jardina Reference Jardina2019; Sides, Tesler, and Vavreck Reference Sides, Tesler and Vavreck2019). Political efficacy is significant in the Black sample for the CMPS 2016, while it was not in Leighley and Vedlitz (Reference Leighley and Vedlitz1999). Recent research has found that expressions of external efficacy have not been responsive to changing political contexts or evaluations of government performance (Chamberlain Reference Chamberlain2012; Reference Chamberlain2013). Scholars have since attempted to build on measures of group efficacy (D. L. Phoenix and Chan Reference Phoenix and Chan2022).
In the 2020 ANES model (Appendix B), all the models of social connectedness were significant in 2020 for the White model, while only the length of residence was significant in Leighley and Vedlitz (Reference Leighley and Vedlitz1999). Unlike the original paper, homeownership was significant and positive in all models in 2020. Given the variation we see in the 2016 and 2020 years and how they compare to Leighley and Vedlitz models, the analysis suggests a mini meta-analysis of the three data sets would be useful to best evaluate the models of turnout. Tables 1 and 2 are summary tables that depict the replication of findings.
Table 1. Summary table of Replicated Findings in 2016
Table 2. Summary table of Replicated Findings in 2020
Mini Meta-Analysis Results
To conduct the mini meta-analysis, the correlation of predictors was meta-analyzed by racial group and Presidential election year to ensure transparency. The results can be found in Table 3 and are presented in D-values. Notably, education produced small effects in 2016 and 2020 for White individuals while it produced medium to large effects for Black individuals, especially in 2016. Political interest had large effect sizes across all groups and election years with Black individuals in 2020 reporting the lowest magnitude at .89. Marriage followed a similar pattern to political interest across all years and groups. Interestingly, political efficacy had a negative effect among Black individuals in 2020. Group closeness also only had a positive magnitude in the Black 2020 group, which had implications for group closeness for Black people during that period. Lastly, the length of residence measure seems to have no effect among models, while being born in the United States had a large magnitude in 2016 for both groups but small effects in 2020, which is supported by scholarship on immigrant political participation (Greer Reference Greer2013; Jones-Correa Reference Jones-Correa2018).
Table 3. D-values for meta-analysis of Black and White turnout in 2016 and 20201
1Homeownership was not included in this section because there was not enough variation on the independent variable of homeownership for Black respondents.
My second hypothesis stated that Black respondents would have higher levels of turnout in comparison to White respondents in the group closeness model only. Table 4 presents a meta-analysis of racial groups in both the 2016 and 2020 elections. It provides varied support for this hypothesis and reports that education and being born in the United States are also significant predictors for Black voter turnout. Specifically for group consciousness, one unit increase in group closeness reports a very small shift in turnout. However, the opposite is true for group closeness for White respondents, as it has a small, negative relationship with turnout. Regarding the relationship between education and turnout for Black respondents, one unit increase in educational attainment results in almost one standard deviation shift in turnout. For White respondents, education results in more than ¼ of a standard deviation in White voter turnout. Being born in the United States for Black respondents reports greater than one standard deviation in voter turnout, but has the opposite effect for White respondents, with greater than ⅖ of a standard deviation in nonvoting.
Table 4. D-values for meta-analysis of Black and White turnout
The findings of this study featured mixed support for H1, which stated that income, education, political interest, political efficacy, length of residence, marital status, group threat, and nativity will have a greater influence on White voter turnout. Political interest, political efficacy, and marriage proved to be significant factors for White voter turnout, especially in comparison to Black turnout. Specifically, for political interest, a one-unit increase in political interest results in two-standard-deviation shifts in White voter turnout. In comparison, for Black voter turnout, a unit increase in political interest results in one standard deviation in Black voter turnout, while sizable is not as large as the finding for White turnout. Political efficacy produced a small unit increase (.11) in voter turnout, while the effect on Black voter turnout produced a small, negative value, showing a different relationship between Black and White turnout. Marriage also produced a significantly large increase in White voter turnout, where the difference between being single and married resulted in greater than one standard deviation in voter turnout. For Black turnout, marriage resulted in a moderate unit increase (.66) where the difference between being single and married resulted in greater than ½ a standard deviation in voter turnout.
Interestingly, income, length of residence, and group threat had little to no effect on either Black or White turnout. All coefficients were small and under .10. For group threat, however, there was a negative sign, which shows a small, but negative correlation with voter turnout for both racial groups. These findings suggest that these models of voter turnout have a decreased effect on recent turnout in Presidential elections and provide an update to Leighley and Vedlitz’s findings.
Conclusions
To conclude, previous theories of voter turnout do not explain voter turnout in the same way as previously discussed in Leighley and Vedlitz (Reference Leighley and Vedlitz1999), which was identified using mini meta-analysis. The mini meta-analysis in this paper allowed greater precision in the analysis of the effect of prominent models of voter turnout during the unprecedented Presidential elections in 2016 and 2020. In addition, it suggests that the methods employed by the Leighley and Vedlitz (Reference Leighley and Vedlitz1999) paper can be beneficial and improved upon to investigate novel phenomena in political behavior during consequential election years. In this analysis, the findings suggest that traditional models of political behavior have differed in their effect since the Leighley and Vedlitz (Reference Leighley and Vedlitz1999) paper and function differently for Black individuals and White individuals.
For example, income and length of residence seem to have little to no effect on turnout for Black and White respondents against Hypothesis 1. This effect could specifically be a result of identity politics becoming more prevalent in electoral contests after 2012. In 2016, Donald Trump made race, immigration status, and other identities very salient in his campaign strategy. In 2020, protests for racial justice in the wake of the murder of George Floyd and a global pandemic also resulted in increased racial rhetoric that could render more sociodemographic factors less important to voter turnout. In addition, in 2020, a Black woman, Kamala Harris, was on the Democratic ticket, making her the first Black woman to secure that office and changing the context of the 2020 election (Smith et al. Reference Smith, Clemons, Krishnamurthy, Martinez, Mclaren and White2023).
Education was also a significant predictor for Black voter turnout only, which goes against preexisting scholarship that suggested group consciousness or other factors were more important to Black individuals (Coveyou and Pfeiffer Reference Coveyou and Pfeiffer1973; Murray and Vedlitz Reference Murray and Vedlitz1977). As more Black individuals have been able to pursue higher education, this effect has also resulted in increased political resources that motivate political participation, which could influence the relationship between education and political participation to be similar to that of Whites. However, this could also produce a chasm between Black individuals with higher educational attainment and Black individuals with lower educational attainment that is distinct from White individuals. This chasm could largely be influenced by an era of deracialized campaigns by descriptive representatives, like President Obama, who are only represented by highly educated political elites. Instead of feeling represented by these figures, Black individuals with lower educational attainment may feel ostracized and ignored, and that could result in decreased voter turnout.
Against prior expectation, neither group closeness nor group threat significantly predicted turnout for Black or White individuals, supporting preexisting work by Gay et al. (Reference Gay, Hochschild and White2016) and Watts-Smith et al. (Reference Smith, Bunyasi and Smith2019) and going against Hypothesis 2. This finding suggests that Black and White individuals in 2016 and 2020 are expressing less connection to their racial group identity, as previous theories of political behavior have considered. There may be generational differences in respondents’ understanding of race and their political opportunities that could call for an updated understanding of individuals’ relationship to their racial group that was produced by the unprecedented context of the Trump presidency, a global pandemic, mass protests, and other significant factors.
In addition, political interest and political efficacy proved to be significant factors for White voters, but not for Black voters when meta-analyzed. This observation supports the argument that disparate socializing political events like voter suppression or the global pandemic could be decreasing Black individuals’ interest and belief in the American political system. Rather than believing the United States is a post-racial society that was implied after the Obama presidency, Black Americans are more aware of the inequities that exist on a day-to-day basis in education, healthcare, wealth, and other factors. Instead of the racial resilience that some scholars theorize, this effect could cause Black individuals to not express interest in politics or believe that politics can benefit them (Slaughter Reference Slaughter2021).
The findings within this paper do suggest that more exploration needs to be done for alternative or updated models of turnout that are specifically fit for Black voter turnout as political interest and political efficacy differ in their effect on Black and White turnout. These updated models include racial group efficacy and emotions like anger and political apathy (Phoenix Reference Phoenix2019; D. L. Phoenix and Chan Reference Phoenix and Chan2022). Racial group efficacy would be a good addition due to Phoenix and Chan’s finding that conventional measures are not as sufficient predictors of political behavior in African Americans as racial group efficacy (Phoenix and Chan Reference Phoenix and Chan2022).
In addition, this paper suggests that an interdisciplinary approach is necessary to best evaluate Black voter turnout and recent elections in American politics. Without the precise analysis that was provided by mini meta-analysis, the political science discipline would not observe how traditional models of political behavior are functioning differently for different racial groups or not even functioning in the way they were originally intended. Moreover, these findings encourage novel interdisciplinary approaches to understanding political behavior from fields such as political psychology and social psychology to improve or update preexisting scholarship. Future research could test these new models and improve the comprehension of Black political behavior to potentially diminish the consistent Black-White turnout gap exhibited in recent Presidential elections.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/rep.2025.10027
Funding statement
The work was generously supported by the National Science Foundation Graduate Research Fellowship awarded to Alexandria Davis (NSF ID: #000883302).
Competing interests
The author declares none.
Open access
