Four More Years! or So What?

Abstract

This study investigated the mental health significance of Barack Obama’s 2012 presidential re-election among Blacks. Upon his re-election, we hypothesized Blacks would either feel symbolic empowerment or relative deprivation. They would feel symbolic empowerment because a man who identifies as Black won re-election to the nation’s highest office. His second victory should generate optimism, given his status as a historic first. Alternatively, they would feel relative deprivation because The Great Recession from 2007 to 2009 curtailed what Obama could achieve. More important, he withered when afforded opportunities to challenge White supremacy and championed individual responsibility. Using a quasi-experimental design with nationally representative survey data from the 2012 Behavioral Risk Factor Surveillance System (BRFSS), we predicted Blacks’ preelection and postelection poor mental health days. We found no time period main effects. However, Black men with less than a college degree experienced 1.11 more poor mental health days postelection whereas Black men with a college degree or more experienced 2.93 fewer poor mental health days postelection. These findings support relative deprivation theory.

Introduction

Barack Obama’s 2008 presidential election was a milestone moment in U.S. history, particularly for Blacks (Bobo and Dawson, 2009; Brooks 2012; Coates 2017; Collins 2012; Dyson 2016; Esposito and Finley, 2009; Hunt and Wilson, 2009; Logan 2014; Moore and Bell, 2010; Parker 2016; Wingfield and Feagin, 2013). Enslaved Africans and their descendants spent 246 years (1619–1865) in chattel slavery—dehumanized, raped, abused, branded, battered, and discarded. They spent 100 additional years (1865–1965) as victims of unmatched institutional discrimination and racial terrorism, barred from equal protection under the law, including voting. Then forty-three years later (1965–2008), a man who identifies as Black became president of the United States. Upon Obama’s presidential election on November 4, 2008, White supremacy seemed vulnerable (Brown, et al., 2021; Bonilla-Silva and Ray, 2009; Gorman 2023; Hunt and Wilson, 2009; Ikard and Teasley, 2012; Teasley and Ikard, 2010; Parker 2016). Right then, an overdue promissory note compensating Blacks for racial injustice seemed redeemable (Brooks 2012; Coates 2017; Dyson 2016; Moore and Bell, 2010; Teasley and Ikard, 2010). However, expectations for Obama outsized what any elected official could accomplish.

If Obama’s 2008 presidential election was the wedding, then his 2012 presidential re-election was the wedding vow renewal. By then, the honeymoon was long over. Evidence existed for Blacks to judge Obama’s contributions beyond descriptive representation (Fletcher 2012; Gleason and Stout, 2014; Hunt and Wilson, 2009; Moore and Bell, 2010; Simien 2015; Stout and Tate, 2013; Teasley and Ikard, 2010). Evidence existed regarding whether he would challenge White supremacy (see Dyson 2016; Lowenthal 2010; Wallace 2012). Evidence existed regarding the harsh reality of governing during and after The Great Recession alongside White politicians who seemed opposed personally to a Black president (Kessler 2012; Parker 2016; Wingfield and Feagin, 2013). In fact, it appeared Obama underestimated White supremacy. For example, in July 2009, a White police officer arrested Henry Louis Gates, Jr., a black Harvard University professor, whilst he attempted to open a jammed door at his home. Someone reported a burglary. Obama suggested race might have played a role in the arrest stating, “…there’s a long history in this country of African-Americans and Latinos being stopped by law enforcement disproportionately.” That statement caused White backlash. Consequently, he backtracked quickly and invited Gates and the arresting officer to the White House for what pundits called a “Beer Summit,” a group-hug-response to racial profiling. In September 2009, during a televised joint session of Congress on health care reform, Rep. Joe Wilson, a White South Carolina Republican, shouted “You lie!” at Obama. It was a breathtaking display of disrespect and White privilege. In January 2010, on his first MLK, Jr. holiday as president, Obama spoke at Vermont Avenue Baptist Church in Washington, DC—a place MLK, Jr. had spoken. He reflected on difficulties collaborating with Congress and distractions stemming from remarks about his race. Referencing the post-racial shift many predicted his presidency would cause, Obama said, “That didn’t work out so well.” In April 2011, the birtherism movement forced the White House to release copies of Obama’s long-form birth certificate to prove he was born in Hawaii. Finally, in February 2012, George Zimmerman murdered Trayvon Martin, a seventeen-year old child whose only crime was being Black in a White space. In response, Obama said, “When Trayvon Martin was first shot, I said that this could have been my son.” Those words enraged many Whites and did nothing to relieve the Black community’s anguish (Coates 2017; Dyson 2016).

Juxtaposing representation (i.e., there is a Black president) against reparations (i.e., there is a Black president challenging White supremacy), we investigate the mental health significance of Obama’s 2012 presidential re-election among Blacks. We invoke symbolic empowerment theory and relative deprivation theory to frame our hypotheses. Using a quasi-experimental design with nationally representative survey data from the 2012 Behavioral Risk Factor Surveillance System (BRFSS), we predict poor mental health days Blacks experienced preelection versus postelection. Findings support relative deprivation theory and imply the thrill (and mental health benefit for Blacks) of a Black president faded by fall 2012.

Symbolic Empowerment Theory

Symbolic empowerment captures a sense of inclusion prompted by an underrepresented individual occupying an authoritative position (Brown et al., 2021; Gorman et al., 2023; Jackson et al., 1996; LaVeist 1992). Their presence confers power vicariously to aggrieved constituents excluded typically from authoritative positions (Goldman and Mutz, 2014; LaVeist 1992; Simien 2015). It taps descriptive representation (Gleason and Stout, 2014; Philpot and Walton, 2007; Stout and Tate, 2013) where an elected official shares physical, social, or experiential characteristics of aggrieved constituents. Further, it signals to aggrieved constituents what might be possible despite strictures of a racialized social system (Bonilla-Silva 1997; West 2017). Descriptive representation manifests best when examining historic firsts in elections (see Simien 2015).

Symbolic empowerment explains changes in political attitudes and behaviors (Gleason and Stout, 2014; Parker 2016; Simien 2015; Stout and Tate, 2013). For example, Shane A. Gleason and Christopher T. Stout (2014) found Blacks’ external efficacy—belief the government is receptive to constituents’ demands—was higher in districts with a Black congressional representative, compared to districts without a Black congressional representative. They concluded descriptive representation facilitated Blacks’ trust in institutions and elected officials. High voter turnout likewise reflects symbolic empowerment. For instance, voter turnout among Blacks, Hispanics, and Asians increased at the 2008 presidential election, compared to 2004 (Lopez and Taylor, 2009; Roberts 2009). In fact, the gap between Black and White voter participation rates shrank in 2008. Black voter turnout reached a record high (66%) during the 2012 presidential election surpassing White voter turnout for the first time (Krogstad and Lopez, 2017). Finally, Tasha S. Philpot and Hanes Walton, Jr. (2007) found Black women most supported Black women candidates, again suggesting representation among political candidates encourages political participation.

Regarding symbolic empowerment and historic firsts, we must discuss Barack Obama. He won election in fall 2008 and re-election in fall 2012 to the nation’s highest office. There were prior historic firsts in Congress (e.g., Hiram Revels), and among presidential candidates (e.g., Shirley Chisolm and Jesse Jackson, Sr.) and vice presidential candidates (e.g., Charlotta Bass). But Obama broke through twice. His biography and broad appeal (i.e., adoration by young White voters) led many to believe the United States would become post-racial (Bobo and Dawson, 2009; Brown et al., 2021; Dyson 2016; Esposito and Finley, 2009; Goldman and Mutz, 2014; Hunt and Wilson, 2009; Ikard and Teasley, 2012; Moore and Bell, 2010; Logan 2014; Parker 2016; Teasley and Ikard, 2010; Valentino and Brader, 2011; Wingfield and Feagin, 2013).

Three studies exemplify the health significance of symbolic empowerment. First, James S. Jackson and colleagues (1996) analyzed nationally representative panel survey data spanning thirteen years and overlapping with Jesse Jackson, Sr.’s 1988 presidential campaign. They found Blacks’ views of racial progress improved and Blacks reported reduced levels of psychological distress and physical health disability around 1988. The authors attributed this salubrious pattern to Jesse Jackson, Sr.’s visibility and campaign message promoting social and racial justice. Second, Jennifer Malat and colleagues (2011) analyzed data from the Ohio Family Health Survey, whose field period lasted from August 6, 2008 until January 24, 2009. They found improvements in self-rated health for Blacks and Hispanics immediately following Obama’s nomination by the Democratic Party, but not after his presidential election or inauguration.

Third, Tony N. Brown and colleagues (2021) developed symbolic empowerment theory to explain why the 2008 presidential election of Barack Obama would demonstrate mental health significance for Blacks. Specifically, they proposed symbolic empowerment increases optimism because “members of an aggrieved group experience a sociopolitical context where a progressive redistribution of power seems possible and possibly ineluctable” (2021, p. 102). They argued Obama’s 2008 presidential election ignited a “collective racial effervescence” benefitting Blacks’ mental health. Using the same outcome and quasi-experimental design as here, they found Black men experienced 1.01 fewer poor mental health days postelection, compared to preelection. Brown and colleagues (2021, p. 110) speculated Black women did not experience improved mental health postelection in fall 2008 because of: (1) internal conflict related to supporting Hillary Clinton as a woman, (2) identification with and fear for Michelle, Sasha, and Malia Obama, (3) realistic conflict with Black men over gendered racism, and (4) awareness the racial status quo would soon return. We suspect Black women in fall 2012 still doubted whether racial progress would happen, hence, we stratify our analyses by gender. Invoking symbolic empowerment theory, we hypothesize (H1) Blacks would report improved mental health after as compared to before Barack Obama’s 2012 re-election.

Relative Deprivation Theory

Samuel A. Stouffer and colleagues (1949) proposed and others honed relative deprivation theory (see Runciman 1966; Smith et al., 2012; Vanneman and Pettigrew, 1972). It describes a situation in which individuals sense, despite being entitled to scarce resources, they are deprived of them relative to others and/or themselves at some previous time. Walter G. Runciman (1966) theorized an individual “…is relatively deprived of X when (i) he does not have X, (ii) he sees some other person or persons, which may include himself at some previous or expected time, as having X (whether or not this is or will be in fact the case), (iii) he wants X, and (iv) he sees it as feasible that he should have X” (p. 10). Importantly, relative deprivation can occur at the social category level (i.e., fraternal relative deprivation, see Smith et al., 2012 for further elaboration). Presently, we investigate relative deprivation experienced by low versus high socioeconomic status Blacks.

We invoke relative deprivation theory for two reasons. First, the U.S. housing market collapse and subsequent global financial crisis sparked by mortgage-backed securities initiated The Great Recession—the most severe recession since The Great Depression. It happened near the start of Barack Obama’s first presidential term, but its impact lasted through his second presidential term. During The Great Recession, the U.S. unemployment rate doubled. The GDP dropped 4.3 percent. Foreclosures were commonplace. Unsurprisingly, the economic downturn harmed Blacks disproportionately. For example, home equity decreased nine percent for Whites but twelve percent for Blacks. Household wealth dropped furthest for Black families (Hall et al., 2015; Thomas et al., 2018). From about 2007 to 2011, Black median income declined 15.8 percent. In contrast, Hispanic median income declined 11.8 percent, Asian median income declined 7.7 percent, and White median income declined just 6.3 percent (DeNavas-Walt et al., 2012). Almost twenty-four percent of Blacks were under-employed and sixteen percent were unemployed during The Great Recession (Kochhar and Cilluffo, 2017; Thomas et al., 2018). During the “recovery period” from 2009 to 2012, Black households still continued to lose wealth (Hall et al., 2015; Thomas et al., 2018). We assert The Great Recession’s racial disproportionality and the fact high socioeconomic status Blacks were more protected generated a sense of relative deprivation among low socioeconomic status Blacks.

Second and more important, we argue many Blacks (but especially low socioeconomic status Blacks) expected Barack Obama to pursue a vigorous civil rights agenda (Brooks 2012; Coates 2017; Dyson 2016; Price 2016). More pointedly, they expected Obama to challenge White supremacy (Bonilla-Silva and Ray, 2009; Coates 2017; Dyson 2016; TMZ 2023). In contrast, Obama’s presidency ushered in a new racial politics (Esposito and Finley, 2009; Fletcher 2012; Gorman 2023; Ikard and Teasley, 2012; Logan 2014; Price 2016; Teasley and Ikard, 2010). He downplayed how structural inequality produced nihilism in the Black community (Dade 2012; Fletcher 2012; Saul 2008; West 2017). In fact, some scholars assert an implicit racial pact enabled his political ascendancy (Bonilla-Silva and Ray, 2009; Coates 2017; DiAngelo 2011; Dyson 2016; Esposito and Finley, 2009; Ikard and Teasley, 2012; Logan 2014; Price 2016; Moore and Bell, 2010; Teasley and Ikard, 2010; Wingfield and Feagin, 2013). This racial pact prescribed how Obama should behave in exchange for access to the White House, a historically White institution: He must affirm his election meant racism receded, race-targeted policies are unfair, racial discrimination is rare, the political system works, and candid conversations about White supremacy are divisive. Indeed, Daniel Q. Gillion (2016) found Obama spoke less about race during his first two years as president than any other Democratic president since the early 1960s. Moreover, colorblindness and abstract liberalism characterized his remaining six years in office (Coates 2017; Dyson 2016; Esposito and Finley, 2009; Moore and Bell, 2010; Price 2016; Wingfield and Feagin, 2013). Applying relative deprivation theory, we argue low socioeconomic status Blacks (i) lacked access to power because of White supremacy, (ii) experienced disparities in health, wealth, education, and incarceration relative to high socioeconomic status Blacks because of power disparities, (iii) wanted Obama to be an anti-racist hero and empower them, and (iv) saw it as feasible that a man who identifies as Black elected to the nation’s highest office would challenge White supremacy.

Evidence of the disappointment low socioeconomic status Blacks felt is scarce because the Black community avoided airing dirty laundry (Brooks 2012; Dade 2012; Fletcher 2012). Given the frequency of Whites’ attacks on Obama (and his family), many Blacks refused to criticize him publicly but rather embraced what he symbolized (e.g., the value of voting and political participation). Still, we did find anecdotal support confirming Blacks’ disappointment.

First, in a politically bizarre media mashup (TMZ 2023), Tucker Carlson interviewed Ice Cube and the topic of Barack Obama’s presidency came up. Ice Cube stated he was proud America had elected a Black man, and thought maybe “this guy is gonna be the guy.” But ultimately, Ice Cube felt Obama, akin to his White predecessors in the White House, did nothing to improve Black people’s standing. Ice Cube stated not much has changed “for the people he knows or cares about.” He concluded Obama’s presidency was simply a “symbolic victory.” Along those lines, at a January 2012 Black poverty forum organized by Tavis Smiley and Cornel West, Roger A. Clay, president of the Oakland, CA-based Insight Center for Community Economic Development said:

I’m extremely disappointed [in Obama], more so than I ever thought I could be. I think part of the reason I’m disappointed is because I had hoped for a lot….Some of my hope was probably based on unrealistic expectations. But because he’s black, I still have very high expectations….I don’t think you go around talking about race, but I do think you have to go around talking about issues that affect black people.

Second, Rakim Brooks (2012) argued linked fate prevents Blacks from criticizing Obama, writing that Black politicians and activists “…should avoid open criticisms of Obama, which only leads to their political marginalization. The administration may have proven a disappointment to most black voters, but they know that a second term will be a good deal better than what they can expect from Mitt Romney” (p. 45). Finally, Bill Fletcher (2012) wrote, “…the content of Obama’s politics, while reform oriented, did not represent anything approaching the consistently progressive. For Black America this presented a complicated picture. While Obama did not run away from being an African American, neither did he integrate race and racial justice into his program” (pp. 5, 8). More bluntly, Fletcher (2012) opined,

…Obama has done little to deal with the structural problems that face Black America….While it is absolutely correct that various reforms initiated by the administration (e.g., healthcare) are of benefit to African Americans, and while it is certainly true that the initial stimulus helped many African Americans, what was missing was attention to the structurally dispossessed in addition to those affected by the Great Recession (p. 6).

Theoretically, the correlation between relative deprivation (measured using socioeconomic status) and poor health is a function of social comparisons, social engagement, and autonomy (Marmot 2005; Wilkinson 1997, 2005). Michael Marmot (2005) argued we find socioeconomic status gradients in morbidity, mortality, and stress responses because hierarchies distribute autonomy differentially and disturb social engagement. Basically, some of us make demands of the world and the world responds. Others of us make demands that go ignored and we feel disappointed. Some of us are rarely deprived. We enjoy full participation in society, in part because others, who compare themselves to their more fortunate counterparts, do not (Wilkinson 1997; 2005). For instance, Bruce P. Kennedy and colleagues (1996) found age-adjusted all-cause mortality, infant mortality, coronary heart disease mortality, malignant neoplasms mortality, and homicide were lowest in U.S. states with less pronounced income disparities between the poor and wealthy, as measured by the Robin Hood Index. Other studies replicated this finding at the country level (Marmot 2005; Kennedy et al., 1996; Wilkinson 2005). Richard G. Wilkinson (1997) wrote, “The most plausible explanation is that mortality is lower in more egalitarian societies because the burden of relative deprivation is reduced” (p. 593).

When Barack Obama won re-election in fall 2012, we suspect low socioeconomic status Blacks felt relative deprivation because they expected their lives would be better after a man who identifies as Black had served as president for four years. But their lives were not, and they felt disappointment. Racially disproportionate devastation caused by The Great Recession explains a small part of it. Obama’s inclination to be “the president of all America,” not Blacks’ anti-racist hero, explains the remainder. Many Blacks, but especially those with low socioeconomic status, expected Obama to challenge White supremacy and discredit individual responsibility as the cure for structural inequality (Bonilla-Silva and Ray, 2009; Coates 2017; Dyson 2016). However, his re-election would bring more of the same: support for White fragility (DiAngelo 2011), colorblind rhetoric (Dyson 2016; Esposito and Finley, 2009; Moore and Bell, 2010; Price 2016; Wingfield and Feagin, 2013), and respectability politics (Coates 2017; Collins 2012; Crenshaw 2014). Invoking relative deprivation theory, we hypothesize (H2), relative to Blacks with a college degree or more, Blacks with less than a college degree would report worse mental health postelection, compared to preelection.

Study Contribution

Our study makes at least three novel contributions. First, mental health research relies overwhelmingly on the stress process model (Pearlin et al., 1981; Pearlin 1989; Turner 2013), a micro-level framework theorizing why certain negative experiences arising from routine circumstances of life overwhelm an individual’s capacity to respond. In contrast, we focus on Barack Obama’s fall 2012 re-election, a macro-level event and whether it influences the mental health of Blacks. Second, our design could be replicated to examine other macro-level events because the BRFSS collects data year-round and has since 1984. Third, we contrast symbolic empowerment theory and relative deprivation theory. The former focuses usually on political attitudes and behaviors, whereas the latter focuses usually on physical health outcomes. Hence, our analysis of mental health is unique.

Methods

Data

We analyzed nationally representative survey data from the 2012 Behavioral Risk Factor Surveillance System (BRFSS). The BRFSS is this nation’s premier system of health-related telephone surveys and collects data year-round from adults (eighteen years of age and older) regarding their risk behaviors, chronic health conditions, and use of preventative services. Established in 1984 with fifteen states, the BRFSS presently collects more than 400,000 survey interviews each year across all fifty states as well as the District of Columbia and three U.S. territories, making it the largest continuously conducted health surveillance system in the world.¹

The available sample included self-identified Blacks (i.e., “Black only, Non-Hispanic” on the calculated variable _IMPRACE [i.e., imputed race] in the 2012 BRFSS [n = 40506]). We aimed to isolate the contemporaneous effect of Barack Obama’s presidential re-election on November 6, 2012. We therefore restricted the available sample to Blacks (n = 4589) surveyed during these time periods: 2012-10-07 to 2012-11-05 (i.e., a thirty-day period before the re-election) or 2012-12-06 to 2013-01-04 (i.e., a thirty-day period starting thirty days after the re-election). The second time period started thirty days after the re-election to accommodate an incubation interval, which permits significance of the re-election to concretize (see Brown et al., 2021, pp. 104-105). Additionally, the second time period started thirty days after the re-election to account for the dependent variable’s “during the past thirty days” framing, thus bolstering potential causal claims. Finally, we kept Black respondents with complete information on study variables, resulting in an estimation sample of 3807.

Variables

Our outcome was the number of poor mental health days reported. The question read: “Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?” It is the only mental health indicator in the BRFSS and has been widely used to examine trends in population-level mental health (see Slabaugh et al., 2016). There were two predictors: (1) time period (0=preelection vs. 1=postelection) and (2) college degree or more (0=no; 1=yes). Time period captured whether the survey interview occurred from 2012-10-07 to 2012-11-05 (i.e., a thirty-day period before the re-election) or 2012-12-06 to 2013-01-04 (i.e., a thirty-day period starting thirty days after the re-election). Replicating the quasi-experimental design used in similar studies (see Brown et al., 2021; Gorman et al., 2023, Malat et al., 2011), we treat the time periods as a control condition and an experimental condition, respectively, with the treatment being Barack Obama’s re-election on November 6, 2012. College degree or more captured whether the respondent attained education equivalent to a college degree or more (see Table 1).

Table 1. Descriptive Statistics Stratified by Time Period and Gender for Poor Mental Health Days, College Degree or More, and Control Variables among Black Adults in the 2012 Behavioral Risk Factor Surveillance System (BRFSS)

Note: Standard deviations shown next to means. Bold indicates marginally and statistically significant differences between men’s or women’s preelection and postelection estimates, respectively. Estimation sample size equals 3,807. Available sample size equals 4,589.

^a Represents excluded groups in the negative binomial regression models in Tables 2 and Table 3.

+ p <. 10 * p <. 05 ** p <. 01 *** p <. 001 (two-tailed tests)

Control variables included established social determinants of health and sociopolitical context variables. Sociodemographic controls included: gender (i.e., men, women), age (range eighteen to ninety-nine years of age), number of children in the household (range=0 to 5 or more), and marital status (i.e., married, not married). Socioeconomic controls included: household income (i.e., less than $25,000; $25,000 to $49,999; $50,000 to $74,999; $75,000 or more) and employment status (i.e., paid employment, retired, unable to work, other). We also controlled for urbanicity (i.e., not in metro area, inside the center city, outside the center city) and whether respondents lived in a state where Obama won the popular vote (i.e., no, yes). We retrieved information regarding which 2012 presidential candidate received the most votes by state from the Massachusetts Institute of Technology Election Data and Science Lab (https://electionlab.mit.edu/data). The BRFSS does not collect data on partisanship (i.e., party identification or political orientation) or voting behavior, or attitudes that may serve as proxies for them (e.g., views on birth control, abortion, affirmative action, etc.). These omissions are intentional because non-partisan federal agencies sponsor the BRFSS. Consequently, to capture information about the sociopolitical context during the 2012 presidential election, we constructed a variable representing the difference between Barack Obama’s state vote winning percentage and Mitt Romney’s state vote winning percentage (i.e., Obama-Romney percent differential). The Obama-Romney percent differential ranged from negative forty-eight percent (i.e., Romney defeated Obama by forty-eight percent) to eighty-four percent (i.e., Obama defeated Romney by eighty-four percent) and we used it as a control variable.

Analytic Strategies

To meet the BRFSS standard for sample designs, participating states must confirm sample records represent a probability sample of all households with telephones in their respective state. Consequently, a sample record is one telephone number in the list of all telephone numbers randomly selected via Random Digit Dialing (RDD) techniques.² Further, BRFSS respondents are surveyed randomly and continuously across an entire calendar year. That means respondents have the same probability of being contacted by telephone any day of the calendar year, barring major holidays. We treat this feature of the BRFSS design as a form of randomization and describe our methodological approach as a quasi-experimental design. Presumably, any unmeasured characteristics (e.g., job loss, poor mental health, voting behavior, political orientation, etc.) are distributed randomly across the time periods. Data in the estimation sample come from all fifty states and the District of Columbia. We omitted respondents from Puerto Rico, Guam, and the U.S. Virgin Islands. We also omitted respondents completing interviews by cell phone because the BRFSS does not capture urbanicity for them.

We present unweighted analyses consistent with a quasi-experimental design (see Brown et al., 2021; Gorman et al., 2023). Patterns in Table 1, where study variables tend not to differ significantly by time period and gender, support treatment of the time periods as control and experimental conditions, respectively. Item non-response was low (1.78 percent per variable, on average) but after listwise deletion the estimation sample was 3807. Household income was the culprit with 674 missing cases. However, missing cases should not affect the results because of randomization across the time periods. Recognizing the dependent variable originates from a count probability distribution with overdispersion (i.e., alpha significantly greater than zero), we ran negative binomial regression models. Each model in Tables 2 and 3 includes a lnalpha estimate and constant. The lnalpha estimate is the log of the dispersion parameter (i.e., alpha), which can be recovered by exponentiating lnalpha. The constant represents the expected log count when all variables in the model are evaluated at zero, if zero represents a plausible value for every variable. To address state-level clustering and inclusion of the Obama-Romney percent differential variable, we estimated robust standard errors. We conducted the analyses in Stata 17.

In Table 1, we generated descriptive statistics. To determine whether study variables’ distributions differed significantly by time period and gender, we conducted two-tailed two-sample t-tests for the continuous variables and Pearson’s chi-squared tests for the categorical variables—see columns 3 through 10. In Table 2, we presented estimates from gender-stratified negative binomial regressions of poor mental health days on time period and college degree or more, net of control variables. To address H1, we examined time period main effects—see row 1, columns 1 and 2 of Table 2. A statistically significant time period coefficient where poor mental health days decrease postelection would support H1. To address H2, row 3, columns 3 and 4 of Table 2 include a statistical interaction between time period and college degree or more. A statistically significant interaction where respondents with less than a college degree report more poor mental health days postelection and those with a college degree or more report fewer poor mental health days postelection would support H2.

Table 2. Estimates from Negative Binomial Regression Models Stratified by Gender Predicting Poor Mental Health Days Using Time Period, College Degree or More, and Control Variables among Black Adults in the 2012 Behavioral Risk Factor Surveillance System (BRFSS)

Note: Standard errors shown in parentheses underneath coefficients. Standard errors are corrected for within-state clustering. Estimation sample size equals 3,807. Available sample size for these analyses equals 4,589.

^a Represents excluded groups.

+ p <. 10 * p <. 05 ** p <. 01 *** p <. 001 (two-tailed tests)

Table 3. Sensitivity Analyses: Estimates from Negative Binomial Regression Models Stratified by Gender Predicting Poor Mental Health Days Using Time Period, College Degree or More, and Control Variables in the 2011 or 2012 Behavioral Risk Factor Surveillance System (BRFSS)

Note: Standard errors shown in parentheses underneath coefficients. Standard errors are corrected for within-state clustering.

^a Represents excluded groups.

^b All models include control variables from Table 2.

+ p <. 10 * p <. 05 ** p <. 01 *** p <. 001 (two-tailed tests)

Results

Table 1 displays descriptive statistics for Blacks in the full sample, and stratified by time period and gender. Blacks reported about four poor mental health days during the past thirty days. Over a quarter of respondents held a college degree or more. Over two-thirds (69%) were Black women. Approximately thirty-four percent were married. Respondents had roughly one child, on average. Over two-thirds held paid employment (42%) or were retired (28%), and almost half earned less than $25,000 per year. They were about fifty-six years old. More than seventy-five percent lived inside the center city or outside the center city. Half lived in a state Obama won. The average Obama-Romney percent differential equaled two percent. More respondents were interviewed from 2012-10-07 to 2012-11-05 (i.e., a thirty-day period before the re-election) than from 2012-12-06 to 2013-01-04 (i.e., a thirty-day period starting thirty days after the re-election) because fewer respondents were interviewed around the Christmas, Kwanzaa, and New Year holidays.

Table 1 shows a small number of study variables’ distributions differed significantly by time period. For example, men reported higher numbers of poor mental health days postelection (3.95), compared to preelection (3.02). The proportion of men who reported other employment was higher postelection (20%), compared to preelection (15%). The proportion of men who reported household income of $75,000 or more was marginally lower postelection (16%), compared to preelection (20%). The proportion of women not living in metro areas was marginally lower postelection (21%), compared to preelection (24%). Finally, more men and women lived in a state Obama won postelection (57% and 53%, respectively), compared to preelection (50% and 47%, respectively). Overall, respondents’ characteristics as indicated by the control variables did not shift across the time periods in a way that would cause changes in mental health.

Table 2 presents estimates from gender-stratified negative binomial regression models. Row 1, columns 1 and 2 addressed H1. Net of control variables, Black men’s expected log count of poor mental health days was. 01 higher postelection, compared to preelection. Black women’s expected log count of poor mental health days was. 03 lower postelection, compared to preelection. However, these coefficients were not statistically significant (95% CI [-.36,. 39] and [-.20,. 11] respectively). Thus, contrary to H1, we found no evidence supporting symbolic empowerment theory.

In terms of control variables, Black men and women unable to work reported more poor mental health days than those with paid employment. Black men in the other employment category reported more poor mental health days relative to those with paid employment. Black men with household income between $50,000 and $74,999 reported fewer poor mental health days relative to those with household income less than $25,000. Black women with higher household income levels reported fewer poor mental health days relative to their counterparts with household income less than $25,000. Each year increase in age associated with fewer poor mental health days for Black women. Black men who lived inside the center city reported more poor mental health days, relative to those not in a metro area. Black men living in a state Obama won reported more poor mental health days. Finally, each unit increase in the Obama-Romney percent differential associated with fewer poor mental health days for Black men.

Row 3, columns 3 and 4 addressed H2. Black men with a college degree or more reported significantly fewer poor mental health days (-1.37, p <. 001) than those with less than a college degree, postelection compared to preelection. Equivalently, Black men with a college degree or more reported seventy-five percent fewer poor mental health days (IRR =. 25, p <. 001), relative to those with less than a college degree. Black women with a college degree or more reported no significant mental health change postelection. Thus, we found evidence supporting relative deprivation theory, but only among Black men. Figure 1 plots the statistical interaction. The left panel displays the relationship between time period and poor mental health days for Black men with less than a college degree. The right panel displays the relationship between time period and poor mental health days for Black men with a college degree or more. Among respondents with less than a college degree, there was a non-statistically significant increase of 1.11 poor mental health days, from 3.19 to 4.30 days (Wald test p =. 18). However, respondents with a college degree or more experienced a statistically significant decrease of 2.93 poor mental health days, from 4.45 to 1.52 days (Wald test p =. 02). Thus, the postelection disparity between those with less than a college degree and those with a college degree or more was 2.78 (4.30-1.52) poor mental health days.

Fig. 1. Predicted poor mental health days among black men by college degree before and after Barack Obama’s 2012 presidential re-election, Behavioral Risk Factor Surveillance System (BRFSS)

In terms of control variables, Black men and women unable to work reported more poor mental health days than those with paid employment. Black men in the other employment category reported more poor mental health days relative to those with paid employment. Black women with higher household income levels reported significantly fewer poor mental health days relative to their counterparts earning less than $25,000. Similarly, Black men with household income between $50,000 and $74,999 reported fewer poor mental health days than their counterparts earning less than $25,000. Each year increase in age associated with fewer poor mental health days for Black women. Men who lived inside the center city reported more poor mental health days relative to those not in metro areas. Black men living in a state Obama won reported more poor mental health days. Finally, each unit increase in the Obama-Romney percent differential associated with fewer poor mental health days among Black men.

Sensitivity Analyses

Table 3 presents sensitivity analyses. It shows estimates from six, separate negative binomial regression models predicting poor mental health days, net of control variables. In the “Black Respondents in 2011” panel, we addressed whether Black men or women typically experience fewer or more poor mental health days around the winter holidays (i.e., a seasonal effect). Specifically, we replicated analyses in Table 2 in fall 2011 and found no statistically significant evidence supporting symbolic empowerment or relative deprivation. In the “White Respondents in 2012” panel, we replicated the analyses in Table 2 with Whites in fall 2012 and found no statistically significant results. Null findings for Whites suggest the mental health significance of Barack Obama’s re-election is specific to college-educated Black men. In the “Black Respondents in 2012, States Obama Lost” panel, we addressed whether hypothesized associations would be stronger in a state Obama lost. We speculated symbolic empowerment or relative deprivation may be more intense in sociopolitical contexts where Blacks are surrounded by voters who did not support Obama. We did not find any time period main effects. However, we again found a significant interaction between time period and college degree or more among men. College-educated Black men in a state Obama lost reported fewer poor mental health days relative to those with less than a college degree, postelection compared to preelection. Although larger in magnitude, the statistical interaction coefficient (-1.68, p <. 05) did not differ significantly (Wald test p =. 61) from the interaction coefficient in Table 2 (-1.37, p <. 001). In the “Black Respondents in 2012, Time Period x Obama-Romney Percent Differential” panel, we addressed whether the association between time period and poor mental health days depended on the Obama-Romney percent differential, a measure of sociopolitical context. In a recent study, Quintin Gorman, Jr. and colleagues (2023) found an amplifying effect for symbolic empowerment among Black veterans living near others who voted for Obama in the 2008 presidential election. Thus, we modeled an interaction between time period and the Obama-Romney percent differential. However, it was not significant. Next, we wondered if the relative deprivation finding was specific to education. We therefore modeled statistical interactions between time period and household income categories in the “Black Respondents in 2012, Time period x Household Income” panel. They were not statistically significant. Further, in analyses not shown, we compared the lowest category of household income against all other categories (i.e., less than $25,000 vs. else). Time period still did not interact with household income to predict poor mental health days. Thus, the moderating effect was specific to education, which did not fluctuate as income did during The Great Recession. Further, low education may associate more strongly than low income with the expectation Obama should challenge White supremacy. Finally, we addressed whether symbolic empowerment or relative deprivation emerged in the thirty-day period immediately following the 2012 presidential election. We realize the thirty-day period immediately following the election means respondents are technically evaluating their poor mental health before the election, in part. Still, we would expect some association to emerge during this incubation interval (see Brown et al., 2021, pp. 104-105). Along those lines, we found Black men with a college degree or more reported a marginally significant decrease in poor mental health days in the thirty-day period immediately following the 2012 presidential election (.62, p = .07; 95% CI [-1.29, .05]).

Discussion

This study examined the mental health significance of Barack Obama’s 2012 presidential re-election among Blacks. Unlike most sociological work examining mental health, we do not invoke the stress process model (Pearlin et al., 1981; Pearlin 1989; Turner 2013), which is a micro-level framework theorizing why certain negative experiences arising from routine circumstances of life overwhelm an individual’s capacity to respond. We instead address how a macro-level sociopolitical shift predicts the mental health of Blacks. Hence, our independent variable represents a social determinant of health far above the micro-level (see Brown et al., 2021; Morey et al., 2021). Two theories framed our hypotheses: (1) symbolic empowerment (Brown et al., 2021; Gorman et al., 2023; Jackson et al., 1996; LaVeist 1992; Stout and Tate, 2013) and (2) relative deprivation (Runciman 1966; Stouffer et al., 1949; Wilkinson 1997, 2005).

On the one hand, symbolic empowerment theory led us to hypothesize Blacks would report improved mental health after Obama’s re-election. Though some luster wore off Obama by fall 2012 (e.g., he received fewer Black votes in 2012 than 2008; Taylor 2012), his re-election represented continuation of a supposed progressive shift. Blacks should feel optimism when a man who identifies as Black earns re-election to the nation’s highest office. On the other hand, relative deprivation theory led us to hypothesize low socioeconomic status Blacks, relative to high socioeconomic status Blacks, would experience worsened mental health postelection, compared to preelection. Our reasoning revolved around constraints on what Obama could accomplish because of The Great Recession and its racial disproportionality, but more so on his reluctance to challenge White supremacy. Many Blacks, but especially those with low socioeconomic status, expected Obama to act as an anti-racist hero (Bobo and Dawson, 2009; Bonilla-Silva and Ray, 2009; Brooks 2012; Coates 2017; Dyson 2016; Gorman 2023; Moore and Bell, 2010; Teasley and Ikard, 2010). Related, relative deprivation predicts negative views of Black elected officials. For example, Brown and colleagues (2018) found Blacks who felt their economic situation had not improved tended to believe Black elected officials hurt or did not help the Black community’s standing. To sum, low socioeconomic status Blacks should feel relatively deprived by Obama’s re-election.

We found no evidence supporting symbolic empowerment theory—Black men and women did not report significantly improved mental health postelection, compared to preelection (see Table 2, row 1, columns 1 and 2). This null finding contradicts what Tony N. Brown and colleagues (2021) found regarding the mental health benefit of Obama’s 2008 presidential election. However, we found evidence supporting relative deprivation theory. Black men with less than a college degree experienced 1.11 more poor mental health days postelection, whereas Black men with a college degree or more experienced 2.93 fewer poor mental health days postelection. The postelection disparity between those with less than a college degree versus a college degree or more was 2.78 poor mental health days. Black women’s poor mental health days did not differ by time period and education (see Table 2, row 3, column 4).

To situate the present effect, consider a study analyzing data overlapping Obama’s second presidential term. With nationally representative survey data (n = 103,710) from the 2013-2015 Behavioral Risk Factor Surveillance System (BRFSS), Jacob Bor and colleagues (2018) examined exposure to police shootings of unarmed Blacks occurring three months prior to respondents’ survey interviews and poor mental health days (our outcome). Half their respondents were exposed to police shootings of unarmed Blacks and exposure predicted .14 additional poor mental health days. We found an association of substantively greater magnitude.

Regarding predictions derived from symbolic empowerment theory and relative deprivation theory, findings supported the latter. Alternatively, one could argue Obama’s re-election symbolically empowered college-educated Black men who weathered The Great Recession, and were less concerned with structural inequality but more concerned with assimilation into White society. Again, Obama personified a type of model minority (Bobo and Dawson, 2009; Brooks 2012; Dyson 2016; Esposito and Finley, 2009; Fletcher 2012; Logan 2014; Moore and Bell, 2010; Price 2016; Teasley and Ikard, 2010; Wingfield and Feagin, 2013). The type was sanitized, safe, and would endorse individual responsibility and downplay White supremacy.

Black women’s null findings underscore that on rare occasions when Obama acknowledged White supremacy, his approach was gendered (Crenshaw 2014). For example, in remarks after exoneration of Trayvon Martin’s murderer, he referenced challenges Black men and boys face in the United States. Nowhere in his remarks did Obama acknowledge Black women and girls also face violence. Further, Obama inserted himself into the incident involving Henry Louis Gates, Jr. He also mentioned regularly his respect for Eric Holder (the Attorney General of the United States from 2009 to 2015). Both were high socioeconomic status Black men. In addition, when his policies were not colorblind, he favored Black men. Consider, for example, the “My Brother’s Keeper” initiative. Obama crafted this 200 million dollar initiative to improve the lives of young men of color after Trayvon Martin’s murder. His decision to exclude Black girls could have signaled his belief Black men and boys are “exceptionally endangered” by White supremacy (Crenshaw 2014). Finally, when addressing the Black community in public and homogeneous racial settings, Obama frequently discussed responsible fatherhood. For example, in June 2008, he spoke on Father’s Day at Apostolic Church of God in Chicago, IL. He lambasted (presumably low socioeconomic status) Black men for forsaking parental responsibilities and the subsequent suffering it brought upon Black communities.

Obama’s responsible fatherhood rhetoric resonated with those Blacks seeking to redefine Black masculinity in line with respectability politics (Coates 2017; Collins 2012; Crenshaw 2014). To sharpen the point, his efforts to fix Black men appeased those who love and hate them (Crenshaw 2014). For instance, Barack Obama gave the 2013 commencement speech at Morehouse College (a historically Black institution serving only men and overwhelmingly Black men). We suspect the speech’s content probably made the brand-new college graduates feel superior to Black men with less than a college degree. Here are a few excerpts (Time 2013): “Sometimes I wrote off my own failings as just another example of the world trying to keep a black man down. I had a tendency sometimes to make excuses for me not doing the right thing. But one of the things that all of you have learned over the last four years is there’s no longer any room for excuses.” “…we have individual responsibilities. There are some things, as black men, we can only do for ourselves.” “Nobody cares how tough your upbringing was. Nobody cares if you suffered some discrimination. And moreover, you have to remember that whatever you’ve gone through, it pales in comparison to the hardships previous generations endured—and they overcame them. And if they overcame them, you can overcome them, too.” “…if you stay hungry, if you keep hustling, if you keep on your grind and get other folks to do the same—nobody can stop you.” “At the turn of the last century, W. E. B. DuBois spoke about the ‘talented tenth’—a class of highly educated, socially conscious leaders in the black community. But it’s not just the African American community that needs you. The country needs you. The world needs you.” “So it’s up to you to widen your circle of concern—to care about justice for everybody, white, black and brown. Everybody. Not just in your own community, but also across this country and around the world.” Unsurprisingly, scholars (Dyson 2016) reported Obama avoided allusions to respectability politics in commencement addresses at other places such as the Naval Academy, University of Notre Dame, and Barnard College.

Obama’s Morehouse College commencement speech occurred after his 2012 presidential re-election, but exposed his beliefs about how Blacks should overcome structural inequality (Bobo and Dawson, 2009; Bonilla-Silva and Ray, 2009; Coates 2017; Collins 2012; Fletcher 2012; Ikard and Teasley, 2012; Logan 2014; Moore and Bell, 2010; Price 2016). Themes in the speech suggested middle-class values produce upward mobility. However, those themes are problematic because structural inequality routinely trumps hard work and meritocracy (Bonilla-Silva 1997; Dade 2012; Esposito and Finley, 2009; Gorman 2023; Saul 2008; Wingfield and Feagin, 2013). Obama implied low socioeconomic status Blacks must learn to help themselves (Logan 2014). This view appealed to high socioeconomic status Blacks while disappointing low socioeconomic status Blacks—the group who bore the brunt of The Great Recession. Related, we recall a hot mic moment involving Jesse Jackson, Sr. in July 2008. He whispered to UnitedHealth Group executive Dr. Reed V. Tuckson (Reuters Staff 2008), “See, Barack’s been talking down to black people…I want to cut his nuts off.” Jackson apologized quickly. Recounting his remarks, he admitted, “I said it can come off as speaking down to black people. The moral message must be a much broader message. What we need really is racial justice and urban policy and jobs and health care.” Interestingly, Jackson had previously critiqued Obama for endorsing colorblindness and failing to challenge White supremacy (Saul 2008).

Limitations and Future Directions

This study is the first to examine the mental health significance of Barack Obama’s 2012 presidential re-election among Blacks. Moreover, this study’s limitations provide a roadmap for future research. First, using BRFSS data, scholars could investigate whether Donald Trump’s 2016 presidential election confers mental health benefit, via symbolic empowerment, for low socioeconomic status or politically conservative White men (see Morey et al., 2021). Second, given what transpired since Obama left office in 2016, Blacks’ current views of him may be more positive than they were in 2012. Thus, symbolic empowerment may be operating to protect Blacks’ mental health today, even though Obama is no longer in the White House. Third, there are other noteworthy dates from Obama’s second presidential term (e.g., his second inauguration occurred on January 21, 2013). Future studies of symbolic empowerment could examine those dates. Fourth, self-report measures are subject to present state bias. That means respondents feeling awful might over-report poor mental health days, whereas those feeling wonderful might under-report the same. But this bias should be distributed randomly across the time periods (see Analytic Strategies). Fifth, universal exposure to Obama’s 2012 presidential election disqualifies statistical techniques that might isolate a causal effect (e.g., propensity score methods or difference-in-differences modeling). Related, it would be ideal to investigate symbolic empowerment, relative deprivation, and mental health using panel data collected from the same individuals preelection and postelection. To our knowledge, no such data exist and therefore BRFSS data are the best available option. Moreover, the sensitivity analyses (see Table 3) should increase confidence in our conclusions. Sixth, BRFSS data suffer from coverage bias, just like all community-based social surveys. For example, the most relatively deprived Blacks (i.e., the incarcerated, homeless, impoverished, etc.) are unlikely to participate in social surveys, thus our estimates are probably conservative. Finally, although ninety-three percent of Blacks voted for Obama in fall 2012 (Roper Center for Public Opinion Research n.d.), BRFSS data do not include variables such as voting behavior and political orientation. Again, we argue those and other omitted variables’ effects should be distributed randomly across the time periods (see Analytic Strategies).

Conclusion

We investigated whether Blacks experienced fewer or more poor mental health days after Barack Obama’s 2012 presidential re-election. We found Black men with less than a college degree experienced 1.11 more poor mental health days postelection, compared to preelection. Whereas Black men with a college degree or more experienced 2.93 fewer poor mental health days postelection, compared to preelection. We suspect Obama alienated some low socioeconomic Black men (and possibly many Black women) by endorsing respectability politics, the power of interracial political coalitions, and racial progress narratives. We conclude Obama’s colorblind policies, underestimation of White supremacy, affinity for individual responsibility rhetoric, and disregard for structural inequality disempowered and deprived select groups in the Black community.