> The empirical evidence shows that, for example, NPR is more likely to interview a conservative economist about a story than a liberal.

To be meaningful, we'd need to compare the ratio of conservative economists on NPR to conservative economists in the population.