Co-published by the Economic Hardship Reporting Project and the Quantitative Methods in the Social Sciences program at Columbia University, with support from research assistants Laura Uguccioni, Guo Jing, and Burui Chen

Introduction

The presidential election in 2016 brought intense discussions of the “working class” as class was perceived to be important to its outcome. This marked a turn in  political discourse, which had long eschewed class-centered identity politics . By the mid-1990s, the political Left had embraced the idea that new times required new strategies and moved away from class as an economic category. Scholar Christopher R. Martin documented this “abandonment of the working class” in his 2019 book “No Longer Newsworthy”.  Indeed, as newspapers dismantled labor beats in the naughts, coverage of unions and workplace issues plummeted. 

Yet, while class-based political narratives were abandoned, it seems that at least some of the print media did not abandon the term “working class”. Robyn Muncy, a historian at the University of Maryland, examined the frequency of New York Times articles in which the term appeared and showed a steady increase, decade over decade, since the 1950s. In her interpretation, the growing inequality of the post-Reagan years came with an increasing recognition of the existence of class, in a sharp reversal from the post-WWII years when the consensus was that the working class was disappearing into the general middle class. 

Number of articles and editorials in the New York Times using the term “working class” by decade | Source: Robyn Muncy; The Strange Career of “the Working Class” in US Political Culture Since the 1950s. Labor 1 December 2018; 15 (4)

How can we reconcile these two findings? How is it that “working class” as a category is gaining momentum in the mass media’s coverage while, at the same time, it is eclipsed in reporting on the ground?  How do American newspapers describe the working class: how do they talk about it, in what contexts, and in what terms? If we look at existing research on the subject, we can see three possibilities. The cliche white male factory workers of yore is still one version.  Another is those with less than a college degree. A final description is the working class as a cultural category denoting tastes, values, and beliefs, rather than a structural economic reality. Is there evidence for any of these three descriptions, or is there a new one emerging?

The growing development of text mining techniques and the digitized historical records of newspaper articles, allow us to answer these questions for a period of forty five years, or since the beginning of what has become known as the neoliberal era inaugurated by the Reagan presidency.  We collected all articles for the period 1980-2024 in which the expression “working class” is mentioned at least once in The New York Times, Pittsburgh Post-Gazette, Philadelphia Inquirer, Dayton Daily News, Minnesota Star Tribune, and Chicago Daily Herald. The five regional papers were selected on the basis of locations in traditional industrial areas in the Midwest and Northeast. We compare the coverage of the New York Times against these five papers in order to check whether, given their locations, the regional papers might be addressing more working class issues and topics than the NYT. We excluded obituaries, articles from the foreign desks or international sections, and we de-duplicated articles (due to multiple formats). The result is a dataset with a total of 35,005 articles across all publications and years. Note that the dataset of digital articles for the New York Times starts in 1980, but for all other papers it starts in 1990.

Recognition Gaps and Representation of Working Class in American Culture

In her 2017 address to the American Sociological Association, as well as in her book “Seeing Others”, Michele Lamont drew attention to growing “recognition gaps”, or disparities of worth and cultural membership between groups in society.  In the neoliberal economic order of the last four decades, growing economic inequalities have also produced growing claims over recognition of moral worth. Recognition gaps arise when groups are ignored, stigmatized, or denied moral standing in society.  The rise of identity-based movements like MeToo, Black Lives Matter, the Dreamers, LGBTQ+, but also the Tea Party on the right, can be understood as making such claims to recognition and against neglect and stigma. The growing use of prejudice-denoting words such as antisemitic, racist, supremacy, gender, homophobic, bigotry, etc. in major US newspapers in the last twenty years (Rozado et al, 2021) is another sign of increasing demands for recognition of cultural and moral worth.

In the moral scripts of  the “neoliberal self,” people and groups are judged according to their individual achievements, self-reliance, and material success. By definition, the working class, facing the loss of well-paying jobs, deskilling, and de-unionization does not meet the neoliberal idea of worth as economic success. As the sociologist Nancy Fraser put it two decades ago, the politics of redistribution has shifted towards politics of recognition: the invisibility of the working class in the media that Martin describes in his book is an act of social and cultural exclusion.

Growing inequality and stagnating class mobility has not gone unreported in the media, for sure. The typical framing has been around access to opportunities and specifically, education. Reports have been focused on the education gap, or the so-called “diploma divide” (Grossman and Hopkins, 2025), a proxy for class divide, and the resulting achievement disparities. News consumption has changed dramatically with the rise of social media and the divide between the “information rich” and the “information poor” has grown (Kalogeropoulos, Fletcher and Nielsen, 2026). This is mostly driven by online news consumption, while offline, college degree is not correlated with the number of news sources consulted. Offline, as Muncy observed, there is a growing awareness and referral to a category called “working class”. Does that mean the print media has seen an opportunity to address the working class recognition gap? Unfortunately, we do not find that to be the case. 

Increased Recognition: Yes but…

First, let us look at our own data and check the claim for increasing recognition of class politics in the media. In our own dataset, we do see the increase in use of the term in the New York Times that Muncy documented but we do not observe a similar increase in the five regional papers we included. Only during times of economic and political upheavals, such as around the 2008 Great Recession and the election of Barack Obama and the subsequent presidential elections in 2012 and 2016, do we observe an uptick in articles which mention “working class”. The biggest divergence in coverage happens around 2015 with the start of the Trump campaign and its emphasis on factory workers, when the sheer volume of NYT articles mentioning the working class doubles and remains high, while for regional papers the 2016 focus on the working class remains a temporary political event.  

This finding might be surprising, given that these newspapers are based on old industrial areas. For people living in and around Pittsburgh, Chicago, Philadelphia, Minneapolis, and Dayton, there is no increase in visibility and attention to the working class as a salient identity in the last thirty years, except as a political actor around elections. It is the national New York Times that shows a sustained recognition beyond election peaks. 

Where is the attention to class showing up? To dig deeper, we looked into the sections of the papers where the term appears over time and again compared the national NYT coverage to the regional papers (It is important to note that for regional papers, the number of articles mentioning “working class” is below 200 for the first three years in the 1990s). We standardized the various sections across the different papers and years into eleven categories and tracked them over time.  

What emerges is a sharp contrast between the regional papers and the NYT, as well as changes over time. For the NYT, there is a shift from a predominance of articles in the Metro, Local, d and Arts and Culture sections in the 1990-2010 to a rise in articles in the Opinion and U.S. or national sections. A similar rise in Opinion sections is observed for regional papers, but a proportional decline in the US/ National section. The very high proportion of lifestyle-oriented articles in the early 1990s is just an artifact of the small overall number of articles mentioning “working class” in that period. However, it is notable that after 2015, the two most common sections in regional papers after Opinion are Lifestyle and Metro/Local.

For a different look into topics, we ran the transformer algorithm on all articles for the two groups of papers which classified them according to the main topic of each article. While there is an overlap between section and topic, the latter allows for greater precision: rather than opinion, or metro vs national sections we get to see specific topics such as US politics and elections and various topics that are salient for specific periods of time such as public health  and COVID 19.

What emerges is even more pronounced than what we see in the analysis of sections: for regional papers and The New York Times alike, the proportion of articles on US politics and elections heavily  dominate after 2015. Between 70% to 80% of all articles in regional papers and  about 60% of all NYT articles are classified in this topic. Both the NYT and other publishers show a long-run shift toward political framing, but the timing and intensity differ.

In the NYT, working-class coverage in the 1990s and early 2000s remains topically diverse: arts and culture, urban life, housing, and lifestyle topics each account for roughly 20 – 35% of coverage in some years. Political topics grow gradually and do not dominate until the mid-2010. Perhaps surprisingly, regional papers which might have been more attuned to local issues especially around outsourcing, deindustrialization, and housing, show an earlier and more pronounced consolidation around political topics. 

One interpretation of this finding could be that the New York Times spends a lot more coverage on political campaigns, election results, and election prediction and the appearance of the term “working class” is driven largely by the language used by conservative politicians and personalities. While this is partly the case, it is interesting to note that the rise of the mentions of the working class predates the 2015 campaign all the way back to 1980s and even before if we take Muncy’s findings. More importantly, it does not account for all of the mentions of “working class”. As we can see, the second most prominent category in arts and entertainment. This topic covers reviews, interviews and essays about movies, books, TV shows, and the theater. Typically, “working class” would appear in the description of the author’s or the characters’ backgrounds. It is notable that for the NYT, this was the most frequent topic (among a much more diverse set of topics) up to the 2008 housing crisis. Politics and elections articles were about 20-25% of all articles, comparable to those in the Housing and NYC Life section. 

Overall, we show that over the last 45 years, the attention to the American working class has shifted from a fragmented, regional topic into a centralized, highly political and cultural identity. Where it is not specifically mentioned in political frames, it is an identity marker for characters and neighborhoods in works of art, rather than in stories about housing, education, work, family, or religious life. 

While these trends certainly speak to a growing recognition of an entity called “working class” that is worth mentioning and discussing. For instance, how do the papers talk about it: what descriptions and what language is being used, in what settings and what topics are being discussed? 

Before we turn to the empirical findings, it is worth revising the concept of representation that the British media scholar Stuart Hall introduced in studying exclusion in the media. If Lamont asks,“Who is seen as worthy and legitimate?” Hall’s question might be paraphrased as “How is worth made meaningful and by whom?”. Recognition can run from neglect, stigmatization, and even praise. We might even think of recognition in terms of a scale of inclusion. Representation is not about the scale of inclusion/exclusion but about the terms and the quality of inclusion. To represent means to describe and portray a social reality from a particular perspective. More than just mirroring reality, re-presentation is about who gets to define the group definition and reconstruct a lived experience through a particular lens. Who speaks to what issues and what kind of language is used is part of this process of re-presenting identities and social realities. 

As Reece Peck argued in his 2019 book, Fox Populism, the first media entity to start framing political issues in terms of class was cable news. Specifically, Fox News adopted a working class language and style. Peck argues against Thomas Frank’s thesis that class conflict was reinvented by the Right to foment a culture war and deny the economic basis as well as class inequality. Instead, Peck describes the shift not as a false consciousness but as a transmutation of class consciousness into a cultural-educational formation where the “elite” are educated professionals rather than the wealthy. This shift was the result of the rising educational levels throughout the 20th century, the shift from industrial economy to one based on cultural – informational production with an educated class best positioned to benefit from it. That was not enough on its own, however, as it took a lot of work performed by the conservative media to construct a category of working class, purely along the lines of cultural tastes and anti-elitism. Peck demonstrates that Fox News performs cultural norms and practices that appeal to those without college degrees and that beyond the content of their programming, it is the style and symbols they use to align themselves with their audience. From that point of view, they are the primary definers of what “working class” means even if they are not doing the work of representation. What the Fox News hosts did was perform “working class” as a cultural practice while at the same time eschewing topics about economic inequality, affordability, health care, education, and other economic concerns.

Is the print media any better at doing representation work to go along with recognition of a new group category? In analyzing newspaper articles that mention the term “working class” we find that while there may be a growing recognition of the working class as an important identity in America today, there is hardly any representation. 

Working Class in Abstract

The most significant finding here is that the vast majority of our articles are not about the working class but only mention it to signify a general category of identity in a story about something else: most often elections and the arts. Nevertheless, we wanted to see whether there is anything we could say about the places where these stories were taking place, or what gender and racial characteristics were most likely associated with the term. To accomplish this, we used the transformer models embedded in large language models, specifically in ChatGPT.  The prompts we used instructed the algorithm to look for specific regions, after giving it definitions of the commonly accepted regions in the United States. We validated a random sample of articles from each region to verify that the algorithm has classified it appropriately. 

We show that early NYT coverage emphasizes the Northeast and Midwest, reflecting its historical institutional geography. Over time, however, NYT coverage shifts toward region-unspecified framing, exceeding 30% by the 2010s. In other words, in its reporting and coverage most of the articles mentioning the working class are no longer rooted in a specific location. They are likely to be articles about general economic conditions or election analysis that does not go into specific geographies.

Other publishers, by contrast, retain stronger regional signatures well into the 2000s, and then see a slight decline in unspecified regions and a slight increase in Midwest and Northeast locations in the 2020s. They become even more regionally-oriented than in the previous two decades, reflecting the geographic locations of these newspapers. Coupled with increased political and electoral coverage, the trend for regional newspapers seems to be to mention “working class” in the context of local elections and campaigns.

Gender representation is strikingly similar across outlets. In both the NYT and other publications, 50% – 65% of articles leave gender unspecified. Men appear more frequently than women when gender is specified, but women rarely exceed 5% – 7% of coverage in either outlet.

Similarly unspecified are other demographic markers such as race or immigrant status, with the vast majority of articles talking about working class in the aggregate.

To further probe the idea that class is used in the abstract, we performed concreteness analysis to capture the average level of specificity in words used in any given article (Brysbaert and Kuperman, 2014). The concreteness scale runs from 1 to 5, where 1 is very abstract and 5 is very concrete. For example, the words “apple” and “tree” have a concreteness score of 5, while the word “justice” has a score of 1.45,  “idea” has a score of 1.61 and “democracy” has a score of 1.78. 

Both the NYT and other publishers show rising concreteness through the early 2000s, peaking around 3.30 on the concreteness scale. After 2008—and especially after 2015—both decline. On average, the decline in specificity from 3.31 at its highest point to 3.16 at its lowest for the NYT is 4.5%. 

Recognition without Representation

Class identity and particularly working class appears to be something to be analyzed, appealed to, or explained rather than described in situ. It is mainly a political actor and perhaps also a marker of authenticity and character when it comes to the arts, i.e. something that exists statistically and rhetorically but not in any concrete social reality that these actors get to define themselves. 

In other words, the working class is talked about more, but described less. It is not centering voices and experiences from communities that can sociologically be described as working class, whether one relies on economic or cultural delineations. The discussion remains at a distance, the identity ill-defined and thin. It will not be right to call this “othering” of the working class, but what we get from the coverage is a hum, a low-level rumbling rather than a full voice.

REFERENCES

Brysbaert, M., Warriner, A.B. & Kuperman, V. Concreteness ratings for 40 thousand generally known English word lemmas. Behav Res 46, 904–911 (2014). https://doi-org.ezproxy.cul.columbia.edu/10.3758/s13428-013-0403-5 

Grossman, M., & Hopkins, D. A. (2025). Polarized by degrees: How the diploma divide and the culture war transformed American politics. Cambridge University Press. 

Hall, S. (1997). The work of representation. In S. Hall (Ed.), Representation: Cultural representations and signifying practices (pp. 13–58). London: Sage. 

Kalogeropoulos, A., Fletcher, R., & Nielsen, R. K. (2026). Inequalities in offline and online news media environments across six countries: The role of social class and interest in news. Digital Journalism, 28(1), 1–20. 

Lamont, M. (2023). Seeing Others: How Recognition Works and How It Can Heal a Divided World. New York: Atria/One Signal Publishers.

Martin, C. R. (2019). No longer newsworthy: How the mainstream media abandoned the working class. Ithaca, IL: ILR Press, an imprint of Cornell University Press. 

Peck, R. (2019). Fox populism: Branding conservatism as working class. Cambridge University Press. 

Robyn Muncy; The Strange Career of “the Working Class” in US Political Culture Since the 1950s. Labor 1 December 2018; 15 (4): 37–58 

Rozado, D., Al-Gharbi, M., & Halberstadt, J. (2023). Prevalence of prejudice‑denoting words in news media discourse: A chronological analysis. Social Science Computer Review, 41(1), 99–122.