« a new progressive era? | Main | me, briefly, on CNN »

October 30, 2008

media bias and election outcomes

The conservative world is abuzz with the idea that liberal news media are either hurting McCain or making his polling results look worse than his real support in the public. I know plenty of liberals who believe that the rise of Rush Limbaugh and Fox News accounts for conservative electoral victories after 1992. These two claims don't cancel out; one or both could be true. But a full statistical model of election outcomes would have to factor in at least:

a. The fit between candidates' positions and public opinion
b. The candidates as communicators/symbols
c. Strategic and tactical political decisions by campaigns
d. Grassroots activity by citizens
e. Campaign finance
f. Changes in the real economic and social circumstances of voters before the election
g. The real performance of the incumbent administration
h. Media bias

I can imagine that (h) would account for some of the variance in election results. But I don't think it can explain too much, because there is a lot of evidence about the importance of (f). Specifically, changes in inflation-adjusted, disposable household income before an election remarkably predict whether the "in" or "out" party wins. And people know their own income; they don't need the media to tell them.

We should wish that (a), (d), and (g) would explain most of the variance in election results. Those are the democratic inputs. Political reforms should maximize the importance of these factors. (H) is unfortunate, but I doubt it's very important once the other factors are considered.

October 30, 2008 10:56 AM | category: press criticism | Comments


Site Meter