The <10% of Americans who regularly consume news vastly overestimate the amount of news consumed by everyone else. This may explain why the information elite[1] are so \baffled by public opinion. We assume opinions are informed by headlines or lived experience – not American Idol. In her book The American Mirage, Eunji Kim explains why that’s wrong.

Upper Middle spoke to Kim, now a professor at Columbia, about how reality TV fuels belief in economic and class mobility among those least likely to experience either. The conversation has been edited for length and clarity.
Your book attacks the most persistent bias of editors, authors, and academics: that they’re part of a politically meaningful discourse. Did you get pushback? Did your peers accept that American Idol might be a worthy topic?
I started grad school as a Korean student claiming puzzles in American politics could be solved by looking at non-political media. So… people thought I was crazy. But we know many Americans – the ones who don’t live on campuses – watch five hours of TV a day. And we have analytics that allow us to push back on the assumption – built on self-reporting in survey data – that people watch news. That real data says people rarely watch or read news. At the time, people thought studying reality TV was frivolous. That has changed.
Upper Middle is a member-supported publication. Want access to the data and analysis that has helped 5,300+ anxious professionals re-negotiate their relationship with status, taste, and money?
Already a member? Sign In.

➺ Weekly research-focused "Class Notes" essays looking at learned behaviors, cultural/social biases, and internalized expectations.
➺ Biweekly survey data-driven "Status Report" teardowns of why people like us act like this. (Future state: Access to raw data via AI portals.)
➺ Invitations to "Zero-Martini Lunch" live Q+As with scholars, authors, experts and minor celebrities with insight into the Upper Middle lifestyle.
➺ Other stuff. Access to deals. Introductions…. We're always trying shit.



