(Randall Hill/Reuters)

A number of pundits, political scientists, and politicians have attributed Donald Trump’s lead in the polls to the media’s incessant coverage of his candidacy. This claim undoubtedly has some truth. Had Trump’s candidacy been covered in only a perfunctory fashion, it seems unlikely that he would have soared so quickly to a lead in national polls and many state polls.

But our new research suggests that we should not be too quick in attributing Trump’s success to his ability to dominate the news cycle. Moreover, the media may be more important to several other candidates than to Trump. Here’s what we’ve found.

We started by collecting daily data on the Democratic and Republican presidential candidates’ support in polls, percentage of media attention (relative to other candidates in their party), and percentage of Google searches. Of course, the primary is not over yet, but we have examined the period between June 1 and Dec. 30, 2015, during which there were many important trends and moments in the campaign thus far.

The graph below shows the trends in media attention based on data from the Internet Archive’s Closed Caption Database. We counted the number of mentions in closed captioning on all three major cable news networks (CNN, Fox News, MSNBC) along with a sample of local TV News programs across the United States. We include here only the percentage of mentions dedicated to each GOP candidate.

We find, as others have, that Trump consistently dominates media coverage after his announcement in June. In fact, from July to mid-October, Trump was the most talked about candidate, taking up between 40 percent and 50 percent of the coverage of all Republican candidates.

The graph below plots Trump’s share of media coverage alongside his national polling numbers and his share of Google searches.

There is some evidence that Trump’s initial surge in both polls and Google searches came after his surge in media coverage. However, throughout the rest of the summer and fall, there is no easily visible relationship. In fact, near the end of our analysis, support for Trump appears to increase in spite of media coverage declining.

Our working paper analyzes the relationship among Trump’s polls, Google traffic, and media coverage in more depth. What we find is striking: increases in Trump’s media coverage do drive Google searches, but not his poll numbers. And neither do his poll numbers appear to drive media coverage, as we might think it should if the media focuses on traditional horserace metrics.

For other candidates, the story is much closer to our traditional expectations for candidate media coverage. This is particularly true for candidates who were not as well-known as Trump going into the race.

It is instructive to compare Trump to Ben Carson, both of whom saw clear surges of support and changes in media coverage but started the campaign with different levels of notoriety. Below are Carson’s poll numbers and share of media coverage and Google searches.

Support and coverage for Carson gradually rise in late summer to early fall. Then, in early November, the Carson campaign was plagued with scandals regarding his background. During this time, we observe an increase in both media coverage and Google searches for Carson, followed by a corresponding decline in support.

Perhaps the starkest difference between the two is search interest. Trump receives 60 percent to 70 percent of the searches for Republican candidates during most of the summer. Carson, like the other Republican candidates, is searched at a much lower rate.

Statistical models suggest that media coverage works differently for Carson than for Trump. Bumps in coverage of Carson do drive bumps in his poll numbers, whereas media coverage plays a smaller role in driving Carson’s search volume than it does for Trump.

Our findings differ to some extent with previous findings. Although we wouldn’t suggest that media coverage had nothing to do with Trump’s rise, over the entire period since he entered the race we don’t find much relationship between media coverage of Trump and his polling numbers. Why?

We suspect that media coverage of Trump has outpaced his levels of support. The media continues covering Trump at very high rates, but there are no more sympathetic voters for him to reach. A bump in media coverage drives more interest for him in general, but does not seem to convert supporters. Contrast this with candidates like Ben Carson, who does gain significantly from bumps in media coverage because he has not yet reached the same level of media saturation.

This is not to say that Trump’s media saturation is not having an effect. It works to suppress support for other candidates by preventing them from gaining media attention — which, according to our analysis, would likely increase their support.

Kevin Reuning and Nick Dietrich are PhD students in the Department of Political Science at Penn State University.