| Popis: |
I construct a new measure of candidate quality using political endorsements made by local newspapers. Similar to expert opinions, newspaper editorial board endorsements are highly-informed judgements that reflect quality differences between candidates, once accounting for the partisan preferences of the newspapers. Using a new data set of over 22,000 local newspaper endorsements, I simultaneously estimate the quality differences between candidates in thousands of elections between 1950-2022 along with a dynamic measure of the partisan slant of hundreds of local newspapers across the United States. After validating the endorsement-based measures of quality and slant, I use the quality differential measure to assess the effect that candidate quality has on election results and governing performance. I conclude by discussing how the newspaper endorsement-based measures have a strong potential to help advance our understanding of the impact of candidate quality and media bias on political representation. |
