Conclusion:
Our results suggest that newspapers do not treat all news the same. Given the same news, Republican presidents receive about 20 to 30 percent less positive coverage from all newspapers and 20 to 40 percent less positive coverage from the top 10 papers than was provided on average for Democrats and these results are quite statistically significant. This partisan bias is associated with about a seven to nine percent increase in respondents viewing the economy as getting better. Overall, the headlines affect people’s perception of the economy and, for at least some people, the impact does not appear to be offset by other news sources.
However, newspapers don’t treat all economic statistics treated the same way nor do all newspapers have the same partisan gap, but the vast majority of newspapers cover any given economic news more favorably under Democrats. Only one newspaper treated one Republican administration significantly more positively than the Clinton administration: the Los Angeles Times’ headlines were most favorable to the Reagan administration, but it still favored Clinton over either Bush administration. When combined with how the Houston Chronicle treated both the two Bush administrations, there is at least a possibility that Republican presidents receive at least some home state backing.
There is some modest support for the hypothesis that newspapers find it more difficult to negatively report on Republicans when Republicans control both the presidency and the congress, but even when Republicans control both branches the implied effect is that they still receive less positive coverage than Democrats. Despite the common perception, newspapers actually provide more coverage when the news is good and at least the top 10 papers are not monolithic in their biases. More durable goods, GDP growth, retail sales and jobs and less unemployment all produce more positive news coverage.
We have tried to account for various factors that could explain the partisan gap in
headlines, but variables measuring whether Republicans control the presidency or congress or even the different administrations could reflect other things that are not directly measured by our regressors, including “bias.” It is not obvious what these other factors are, but this is possible just as a lot of black-white, male-female differences can reflect factors other than race or gender.While we have provided some information on the incentives that the media has to cover political parties differently based upon how much of the government they control, the central question that remains: why does this partisan gap in news coverage persists over the fourteen years examined here? Others have started offering different theoretical explanations (e.g., Baron, 2004; Bovtiz, et. al., 2002), though there are still questions about why customers might prefer biased information.
Not all that “unexpected.”