The US news media is an important disseminator of political information. In doing so, it shapes how issues, candidates, and other “objects” of interest are defined and perceived. This is particularly important for emerging language and buzzwords, which appear with increased frequency in today’s fast-paced, polarized environment. This article discusses the trajectory of the phrase “identity politics” and how it relates to the political science discipline. Then, using second-level agenda setting as a theoretical framework, we examined how newspapers use and shape the phrase “identity politics” for the public. Overall, we argue that vague phrases like this are ideal candidates for the media to inject meaning into and that political communication theories should be used to study similar terms going forward.
TRAJECTORY OF “IDENTITY POLITICS” AND WHY IT MATTERS TO POLITICAL SCIENCE
The phrase “identity politics” was first coined around 1977 by Barbara Smith and other Black feminist and LGBT activists in the 1983 Combahee River Collective Statement. The document proposed the political framework that oppression—which is structural and interlocking—affects separate identity groups in different ways and can be overcome only after this is recognized. The original document states, “This focusing on our own oppression is embodied in the concept of identity politics. We believe that the most profound and potentially most radical politics come directly out of our own identity, as opposed to working to end somebody else’s oppression. In the case of Black women, this is a particularly repugnant, dangerous, threatening, and therefore revolutionary concept because it is obvious from looking at all the political movements that have preceded us that anyone is more worthy of liberation than ourselves” (Combahee River Collective Reference Collective1983, 4). Thus, the origins of the phrase are unambiguously rooted in the struggles of marginalized communities, particularly Black women.
Since it was coined, academics across disciplines have used the phrase loosely. Bernstein (Reference Bernstein2005, 47) noted in her review of the sociology literature that it has been used “…to describe phenomena as diverse as multiculturalism, the women’s movement, civil rights, lesbian and gay movements, separatist movements in Canada and Spain, and violent ethnic and nationalist conflict in postcolonial Africa and Asia as well as in former communist countries of Eastern Europe.” Sociologists and others in the humanities also note that the phrase has assumed a negative connotation in some scholarship to describe Leftist politics (Fraser Reference Fraser1997), political silos, and divisiveness (Bow et al. Reference Bow, Brah, Goeman, Harriford, Keating, Lin and Pérez2017). This negative connotation perhaps culminated in Lilla’s (Reference Lilla2017) book, The Once and Future Liberal: After Identity Politics, in which he argued that the American Left will flounder as it splits into increasingly narrow and isolated identity-based subgroups that obsess over movement politics rather than winning raw power. The conception of this phrase that Lilla draws on—that the Left excessively caters to racial and gender minority groups—was likely its most common use when he wrote his book. This conceptualization even appeared as a wedge issue between the progressive, activist wing of the Democratic Party and the establishment wing, having played an explicit role in the party’s post–2016 Unity Reform Commission (Masket Reference Masket2020).
Discussion of the phrase “identity politics” also has permeated political science discourse since the 2016 election. Although the study of identities and their place in power structures is not new, discussion of the phrase has become more commonplace. As Masket (Reference Masket2020) explained in his book, Learning from Loss, some political scientists have taken recently to stating that “all politics is identity politics” as a way to push back on the pejorative and narrow use of the term described previously. These academics suggest that “identity politics” includes everything from explicitly courting racial demographics to local governments fixing potholes. This line of reasoning is unsurprising; our scholarship suggests that identity can be found everywhere: in mobilizing Latinos (Barreto Reference Barreto2007), Black political behavior (Austin, Middleton, and Yon Reference Austin, Middleton and Yon2012; Dawson Reference Dawson1994), political campaigns by women (Plutzer and Zipp Reference Plutzer and Zipp1996), and more. Crucially, however, identity is not simply a feature of politics for minority groups; rather, it similarly affects majority “default” groups. These groups include Christians (Albertson Reference Albertson2011, Reference Albertson2015), whites (Jardina Reference Jardina2019; Knuckey and Kim Reference Knuckey and Kim2020), and men (Carian and Sobotka Reference Carian and Sobotka2018). However, identity extends beyond the demographic boundaries of race, gender, and religion. Partisanship is understood to be a social identity and part of one’s self-concept (Greene Reference Greene1999, Reference Greene2004), as is gun ownership (Lacombe Reference Lacombe2019). Moreover, Republican and Democratic legislative offices are biased by racial identity in how they respond to constituent requests (Butler and Broockman Reference Butler and Broockman2011). Identity is omnipresent at the mass and elite levels on both sides of the political aisle. In summary, academics and progressive activists seem to be speaking past one another; there is a definitional disconnect in elite circles. This led us to ask: How is this confusing phrase portrayed to the public on a mass level? One way to answer this question is through the lens of second-level agenda setting.
SECOND-LEVEL AGENDA SETTING
Agenda-setting theory is commonly used in political communication studies. It suggests that the public will deem certain topics as “important” if the media covers them (McCombs and Shaw Reference McCombs and Shaw1972). Thus, the original theory is about issue salience in the news. However, second-level agenda setting—sometimes called attribute agenda setting (McCombs et al. Reference McCombs, Llamas, Lopez-Escobar and Rey1997)—is less about issue salience and more about the evaluation and interpretation of issues or “objects” by the media. This type of agenda setting has two components: (1) a cognitive attribute, which connects substantive components to the object; and (2) an affective attribute, which connects positive, neutral, and negative tones to the object (McCombs et al. Reference McCombs, Llamas, Lopez-Escobar and Rey1997). This theory typically is applied to specific issues, such as the development of a local park (Kim, Scheufele, and Shanahan Reference Kim, Scheufele and Shanahan2002) or to candidates (Kiousis, Bantimaroudis, and Ban Reference Kiousis, Bantimaroudis and Ban1999; McCombs et al. Reference McCombs, Llamas, Lopez-Escobar and Rey1997). However, emerging terms and phrases that lack an obvious definition—at least for the general public—can operate as relatively empty vessels, ready to be imbued with information and feeling. They also are “objects.”Footnote 1
…emerging terms and phrases that lack an obvious definition—at least to the general public—can operate as relatively empty vessels, ready to be imbued with information and feeling.
Tracking the media’s use of terms and phrases is not new. The phrase “climate change” is a prime example; a substantial number of academic papers discuss the frequency and use of this phrase in the media (e.g., Badullovich, Grant, and Colvin Reference Badullovich, Grant and Colvin2020). It is unsurprising that scientific terms, which can be confusing for laypersons, would be monitored by academics. In recent years, however, new and vague political terms have emerged. Granath and Ullén (Reference Granath and Ullén2019) studied the proliferation of the phrase “political correctness” in Time Magazine, also documenting the tone and to which issues and political groups it was connected. Lewis and Reese (Reference Lewis and Reese2009) examined how the media proliferated and framed the phrase “war on terror.” Levendusky and Malhotra (Reference Levendusky and Malhotra2016) tracked the uptick in the term “polarization” and how it was used in US newspapers from 2000 to 2012. Cunha et al. (Reference Cunha, Magno, Caetano, Teixeira, Almeida, Staab, Koltsova and Ignatov2018) examined how media in 20 countries shaped the phrase “fake news.” Although these studies were excellent at tracking emerging political language, few connected the phenomenon to political communication theories (for an exception, see Lewis and Reese Reference Lewis and Reese2009).
METHODS
To examine the frequency and portrayal of “identity politics” in the media, a quantitative content analysis was conducted. Any article that used the exact phrase “identity politics” between 1977 (when the phrase was coined) and 2020 was pulled from newspapers that were classified as “major publications” in the LexisNexis database. (The online appendix lists all publications from the article population and how the phrase breaks down by publication region.)
Figure 1 shows that these publications did not use the phrase until the 1990s. Even then, the number of mentions did not surpass 50 during that decade. We observed a noticeable spike around 2007/2008, when Hillary Clinton and Barack Obama sought the Democratic nomination. Afterwards, the number decreased a little and then skyrocketed around 2016, when Clinton became the first female nominee for a major party, pitting her against Donald Trump. The delta formations in the number of mentions indicate a sign of first-level agenda setting (i.e., increased salience).
Figure 1 Raw Mentions of “Identity Politics” in Major American Newspapers
Note: There are no mentions until 1990. When coding began in July of 2020, there were only 112 mentions of the phrase that year. The number incresed to 185 by the end of the year, which is included in this graphic for completeness.
To determine how the news media was injecting attributes into this phrase, we focused on articles between January 1, 2016, and December 31, 2019,Footnote 2 which produced a population of 1,256 articles. The analysis was nuanced and therefore had to be done by hand. To make it manageable, 50 articles were randomly selected from 2016, 2017, 2018, and 2019 for a total sample of 200 articles. In July 2020, a research assistant was trained to code the articles for the following features:
-
1. Source: From which publication the article came.
-
2. Group: Whether “identity politics” was tied explicitly to the American “Right” (i.e., Republicans, Republican Party or candidates, and Conservatives); the American “Left” (i.e., Democrats, Democratic Party or candidates, and Liberals); both groups about equally; or neither. “Neither” includes a general use of the phrase, an application to another country, or to people not explicitly connected to the political Left or Right.
-
3. Valence: Whether the article had an overtly positive tone, neutral/balanced tone, or overtly negative tone.
-
4. Use: Whether the author was clearly editorializing when mentioning the phrase, reporting on it without editorialization (including using another person’s quote), or both.
The second and third features map onto the cognitive and affective components of second-level agenda setting. They also capture the debates and discussions among political scientists and political consumers. Is this phrase mainly used in connection with the American Left, despite many forms of “identity politics” on the Right? Is it usually leveraged in a pejorative way? The “Use” feature indicates whether the affective (i.e., valence) component is a result of authors taking a stance on identity politics or trying to report more straightforwardly on it. Because “identity politics” is a new phrase in mainstream political discourse and has not been studied, there is no expectation for how the results should emerge; therefore, this analysis is exploratory. When the coding was completed, a second coder—who was blind to the coding and identity of the first coder—was recruited in January 2021 to code a random sample of 20 articles to determine inter-rater reliability. The codes had 77% agreement and a kappa statistic of 0.62, which is considered “substantial agreement” (Fleiss and Koch Reference Fleiss and Koch1977).Footnote 3
RESULTS
Some basic aggregate trends emerged from the “identity politics” spike from 2016 through 2019, as shown in figure 1. First, 68.5% of articles were from the New York Times and 11.5% from The Washington Post, which means that 80% of the sample was concentrated in two widely circulated national publications. About 20% of articles came from other newspapers including the Los Angeles Times, USA Today, and regional papers such as the Pittsburgh Post-Gazette. It is clear that this phrase was used mostly by influential newspapers known for their highly visible analyses and editorials. This suggests a simple difference in the jobs of national-versus-regional journalists. If “identity politics” is so pervasive that it indicates an underlying shift in how we engage consciously with politics, then discussion by national journalists whose job it is to assess macro political trends is warranted. Conversely, regional journalists remain focused on smaller-scale issues that affect their local landscape. This finding also may reveal that “identity politics”—the divisive issue within elite and activist party circles mentioned previously (Masket Reference Masket2020)—is being pushed into the consciousness of the moderately engaged simply by being covered so often by major publications, a form of first-level agenda setting (McCombs and Shaw Reference McCombs and Shaw1972). It may not trickle down to those who are less engaged with national politics and who read local sources.
Second, the phrase most often is tied (32%) to people and groups associated with the political Left.Footnote 4 For example, Burris (Reference Burris2019) stated that it is “just plain nuts” for Democrats to discuss changing the name of their Jefferson–Jackson Day dinner. He suggested that “The heritage of Jefferson and Jackson is what pulls the Dems out of interest-group and identity politics and gives them a link to something great and lasting.” However, the American Right is tied to “identity politics” about 21.5% of the time, as shown in Caldwell’s (Reference Caldwell2016) article, “What the Alt-Right Really Means.” He stated that “Donald Trump is the first step toward identity politics for European Americans in the United States.” A subsequent quote stated that Alt-Right movement leader Richard Spencer replied, “Yuh. You’re right,” when asked if his group is practicing identity politics (Caldwell Reference Caldwell2016). In the article, “Nationalism and Heckling Take Spotlight at Conservative Conference,” Republican consultant Mike Madrid was quoted as saying, “It’s horrifying but not surprising…the Republican Party is devolving into the home of white identity politics” (Weigel Reference Weigel2018). “Identity politics” is tied to both groups about 15% of the time and to neither side about 31% of the time. These findings demonstrate that the media affixes cognitive associations to an object, which is one component of second-level agenda setting. In this case, they are mostly associating identity politics with the political Left.
Caldwell (Reference Caldwell2016) stated that “Donald Trump is the first step toward identity politics for European Americans in the United States.”
During our selected timeframe, about 60% of articles used the phrase “identity politics” in a way that was unambiguously negative (see the previous Madrid quote for an example). About 37% were neutral and only 3% were unambiguously positive. This is perhaps the most telling trend to emerge from the analysis: “identity politics” was conveyed to the public in an unflattering light (or at best not positively). Finally, when discussing “identity politics,” journalists were making their own opinionated statements about it 24% of the time. A strong example of this is Brooks’ (Reference Brooks2016) column titled, “Identity Politics Run Amok,” in which he stated, “Identity politics distorts politics in two ways. First, it is Manichaean. It cleanly divides the world into opposing forces of light and darkness. Second and most important, identity politics is inherently the politics of division….It corrodes the sense of solidarity. It breeds suspicion, cynicism, and distrust.” Conversely, authors were strictly reporting on the phrase without inserting their own opinion 60% of the time and doing both 16% of the time.
This is perhaps the most telling trend to emerge from the analysis: “identity politics” was conveyed to the public in an unflattering light (or at best not positively).
ADDITIONAL ANALYSES
We might wonder if the Trump years and simultaneous rise of white nationalist and supremacist movements forced the media to tie this phrase more to the Right, despite identity long being central to both sides of the political spectrum. In addition to analyzing 2016–2019, the original coder also coded 50 randomly selected articles from 2007, when we begin to observe a smaller spike long before Trump’s candidacy. As shown in table 1, the percentages indicate that “identity politics” was tied to the political Right about three times as often beginning in 2016 as it was in 2007. However, it dips back down (and simultaneously increases on the Left) when the presidential primaries began in 2019. Although 2007 is only one year of data, it provides some indication that Trump’s presidency could have affected the cognitive component of second-level agenda setting such that journalists tied “identity politics” more to the Right during this period.
Table 1 Which Political Groups Are Associated with “Identity Politics”
DISCUSSION
This analysis is limited in scope: only about 250 articles from a population of 1,320 in the five years under study were coded. Hand-coding is labor intensive, making large sample sizes difficult to achieve. However, it can be necessary when the analysis requires nuanced reading and fine distinctions that computers cannot identify (Conway Reference Conway2006; Simon Reference Simon, Hart and Shaw2001). The analysis also was limited to major newspapers because these publications—although likely biased to some degree—are not overtly political outlets that cater to readers in search of partisan or ideological viewpoints (e.g., National Review, The Daily Wire, and Jacobin). They better represent how the phrase is portrayed to a more general—albeit at least moderately engaged—audience. This study also did not test whether the media’s interpretation of “identity politics” affects the public—it simply documents that the media has imbued cognitive and affective meaning onto the phrase. Future studies should test the effect of this phenomenon, especially because there is evidence that these patterns matter for public attitudes (Levendusky and Malhotra Reference Levendusky and Malhotra2016).
As terms and phrases that tap into our political climate (e.g., “neoliberal,” “Alt-Right,” “cancel culture,” and “woke”) continue to emerge, political scientists should consider analyzing them under the second-level agenda-setting umbrella. This framework can provide important theoretical context and expand the literature such that it keeps up with modern political discourse. Future studies also can illuminate whether certain terms (e.g., “woke” and “cancel culture”) have been appropriated (Adams Reference Adams2020; Clark Reference Clark2020) and redefined by elite actors.Footnote 5
Supplementary Materials
To view supplementary material for this article, please visit http://doi.org/10.1017/S1049096522000452.
