Use of ‘Sexist’ and ‘Racist’ in the New York Times Increased Over 400% Since 2012. Why?

In recent years, words and ideas used to describe discrimination against members of historically marginalized and disadvantaged groups have seemingly exploded into the lexicon: systemic inequality, privilege, white supremacy, the patriarchy, etc.

In some senses, the apparent novelty is deceiving: most of these concepts have been around for decades in academic and activist circles. None of these ideas are genuinely new.

Nonetheless, our new study published in Social Science Computer Review demonstrates that a substantial shift in discourse has occurred – at least in the US media. And it may have important social and political implications.

Analyzing 27m news articles published in 47 popular news media outlets between 1970 and 2019, we find that there was a rapid uptick in the use of words related to prejudice and discrimination beginning in the early 2010s. These shifts occurred in left- and right-leaning media alike.

Moreover, the changes are not limited to print media. Using data from the Stanford Cable TV News Analyzer, we found that the major shifts observed in print media seem to have been mirrored in television news coverage as well.

Just as striking as the magnitude of the shifts is when they occurred. The attitudinal and discursive changes don’t seem to be a response to anything in particular.

For instance, given that the shifts became increasingly pronounced in 2012, one might assume they are straightforwardly a response to the killing of Trayvon Martin. However, it is harder to explain why we see similar changes in discussion and public attitudes with respect to sexism, homophobia, transphobia, Islamophobia and antisemitism – all at roughly the same time as we observe the spikes related to race.

The shifts with respect to sexism and misogyny, for instance, were not a response to #MeToo. They preceded the emergence of #MeToo by a few years, and may help explain why the movement was able to achieve the impact it did – why it was able to find such a receptive audience when previous efforts to elevate the issues of sexual violence, harassment and discrimination were less able to gain traction.

Indeed, there may also have been some deeper antecedent shift that helps explain why the deaths of Trayvon Martin, Eric Garner, Michael Brown resonated so strongly when other comparable incidents of violence against African Americans (including other incidents caught on film) did not, and which may also explain why #MeToo and other social justice movements were able to gain such strong traction during this period as well.

One possible explanation: these movements occurred in the context of already-heightened concern and preoccupation with bias and discrimination among the literati.

Beginning in 2012 there was a rapid and dramatic shift in expressed beliefs regarding race, gender, and sexuality, which Vox’s Matt Yglesias popularly described as “The Great Awokening”. Critically, this transformation was most acute among the Americans who happen to be the primary producers and consumers of journalistic content.

Contemporary US media workers tend to hail from a very particular segment of the population. They tend to be urban dwelling and politically liberal. They tend to come from relatively wealthy backgrounds. They tend to be highly educated, and those who work at flagship outlets tend to hail from prestigious schools and programs. Reporters skew disproportionately white and male relative to the communities they are embedded in. Summarizing the primary producers and consumers of journalism in the US today, the media scholar Nikki Usher aptly described the current state of affairs as News for the Rich, White and Blue.

It may be that this rapid shift in attitudes among news producers and their primary audiences is what ended up driving the rapid changes in media discourse – setting the stage for social justice movements to “break through” into the mainstream in a way that they might not have been able to prior.

However, while the attitudinal shifts with respect to prejudice and discrimination may have begun among a specific subset of the population, they did not end up confined there. Broader shifts in perceptions of prejudice and discrimination seemed to follow shifts in media discourse.

It was not just the media discourse that has changed since the early 2010s – public opinion has shifted dramatically too. For most types of prejudice and discrimination (with the notable exception of homophobia), Americans grew far more likely to claim that identity-based bias and unfair treatment are widespread in the US as compared to polling from the turn of the century.

What’s more, our data suggest that it is changes in media coverage that seem to predict shifts in public perceptions around race and gender discrimination – not the other way around.

While we cannot definitively establish causation between media coverage patterns and changes in public attitudes, our findings are compatible with the possibility that the former affected the latter.

And insofar as this coverage influenced public opinion, at the margins it may have also altered the course of American politics.

The shifts in media discourse and public perceptions about identity-based prejudice and discrimination predate Trump’s 2016 presidential run. In other words, it was not a response to Trump. Instead, Trump seems to have emerged on the political scene in the context of already-heightened news media discussion about prejudice and discrimination, and public concern with the same. Both trends continued to accelerate over the course of Trump’s stay in the White House and, indeed, have continued unabated since his ouster.

With this in mind, many subsequent social and political developments become a bit more comprehensible:

Previous quantitative analyses have shown that there was an unprecedented focus on Trump in print media, television news and academia alike. This coverage was overwhelmingly negative, and focused heavily on his purported sexism, racism, Islamophobia and xenophobia – attributes assumed to be shared by his supporters, and that were purported to be key to his political success.

Yet under Trump’s tenure at the helm of the party, Republicans consistently lost ground with white people. The attrition was most pronounced among white people who were relatively well-off and highly educated; again – who are among the primary consumers of news media in the United States. What’s more, there is evidence that these disaffected white voters were alienated from the Republican party precisely as a result of Trump’s racialized rhetoric and policies. Overwhelming negative depictions of Trump in news media might have exacerbated this trend.

That is, prior to and over the course of Trump’s presidency, media discussion of prejudice and discrimination increased significantly. The public grew increasingly concerned with identity-based mistreatment as well – particularly key subsets of white voters. This increased awareness and sensitivity around bias and discrimination – including among key Republican blocs – may have led many white people who previously (if begrudgingly) supported Trump to turn against the former president and his party in 2018 and 2020.

Trump and his opponents seem to have been united in their belief that the former president’s racialized policies and rhetoric were key to his electoral success. Our data may provide important context for why Trump instead met his political demise in 2020. Despite consistent Republican gains among many ethnic and religious minority populations over the course of Trump’s candidacy and presidency, losses among key white constituencies proved fatal for his electoral prospects. Changes in media discourse around prejudice and discrimination seemed to both reflect and exacerbate the underlying attitudinal shifts among whites that may have led to this outcome.