We find no evidence that partisan newspapers affect party vote shares, with confidence intervals that rule out even moderate-sized effects.Yes, newspapers increased voting turnout. But they did not influence the direction of the votes.
This finding arises from the voting behaviour in presidential elections across US counties between 1868 and 1928 - before radio and TV, when newspapers were the dominant media.
Even in the era when Charles Foster Kane was allegedly deciding the results of presidential elections and the Daily Beast was causing revolutions, newspapers seem to have had much less effect than everyone thought. And if this was the case then, how much more likely is it to be true today, when newspaper influence is offset by the influence of other media?
It could be that “media influence” is a post hoc ergo propter hoc fallacy, an egomaniacal fantasy of journalists and proprietors which has sucked in politicians.
For lefties, this is a mixed blessing. On the one hand, it means we shouldn’t care so much about the bile spewed out by the Mail, Sun or Express. But on the other hand, it raises a worrying question: could it be that the public are stupid, bigoted and illiberal for reasons other than the influence of newspapers?