[H]ave the predictions of widespread media echo chambers really come true? It’s hard to tell using questions like Pew’s, which ask people to self-report where they get their news. People can be biased in what outlets they choose to name or forgetful of the media they did consume in different settings and contexts. In particular, liberals or conservatives may be prone to exaggerating their exposure to ideologically consistent news outlets. Naming Fox or MSNBC in response to a question like the one Pew used may thus be more of a marker of tribal affiliation than a direct measure of news consumption.
The picture looks a lot brighter when social scientists have analyzed measures of people’s news consumption in the real world. It turns out that the media people are actually exposed to both online and offline is much more diverse and heterogeneous than people’s self-reports suggest.