The data graphic below shows measure of volume and accuracy for newspapers based on data from 1980-2018, and television networks from 1990 to 2018. (See more detailed information below.)
Scatterplots show both volume (y-axis) and accuracy (x-axis). Volume is the average number of sentences about welfare spending appearing in a given outlet in a single fiscal year. In many policy domains, there are thousands of these sentences in a single fiscal year. And all of these sentences are coded based on whether they indicate increases, decreases or no change in spending. The average of these codings indicate the direction of change suggested in media coverage in a given fiscal year. Accuracy captures the strength of the relations between this annual ‘media signal’ and changes in actual spending. Higher values indicate greater media accuracy.
In the scatterplots, media outlets towards the upper right of the graphic produce a high volume of accurate content. Media outlets towards the bottom left produce lower volumes of inaccurate coverage. Note that no outlet produces perfectly accurate coverage by our measure, but there are certainly some that do better than others.
Bar graphs then show the same data separately, first for volume (on the left) and then for accuracy (on the right).