Fake News in addition to Bots May Be Worrisome, although Their Political Power is actually Overblown

How easy is actually the item to change people’s votes in an election?

The answer, a growing number of studies conclude, is actually that will most forms of political persuasion seem to have little effect at all.

This specific conclusion may sound jarring at a time when people are concerned about the effects of the false news articles that will flooded Facebook in addition to different online outlets during the 2016 election. Observers speculated that will these so-called fake news articles swung the election to Donald J. Trump. Similar suggestions of large persuasion effects, supposedly pushing Mr. Trump to victory, have been made about online advertising via the firm Cambridge Analytica in addition to content promoted by Russian bots.

Much more remains to be learned about the effects of these types of online activities, although people should not assume they had huge effects. Previous studies have found, for instance, that will the effects of even television advertising (arguably a higher-impact medium) are very modest. According to one credible estimate, the net effect of exposure to yet another ad shifts the partisan vote of approximately two people out of 10,000.

In fact, a recent meta-analysis of numerous different forms of campaign persuasion, including in-person canvassing in addition to mail, finds that will their average effect in general elections is actually zero.

Field experiments testing the effects of online ads on political candidates in addition to issues have also found null effects. We shouldn’t be surprised — the item’s hard to change people’s minds! Their votes are shaped by fundamental factors like which party they typically support in addition to how they view the state of the economy. “Fake news” in addition to bots are likely to have vastly smaller effects, especially given how polarized our politics have become.

Here’s what you should look for in evaluating claims about vast persuasion effects via dubious online content:

How many people actually saw the questionable material. Many alarming statistics have been produced since the election about how many times “fake news” was shared on Facebook or how many times Russian bots retweeted content on Twitter. These statistics obscure the fact that will the content being shared may not reach many Americans (most people are not on Twitter in addition to consume relatively little political news) or even many humans (many bot followers may themselves be bots).

Whether the people being exposed are persuadable. Dubious political content online is actually disproportionately likely to reach heavy news consumers who already have strong opinions. For instance, a study I conducted with Andrew Guess of Princeton in addition to Jason Reifler of the University of Exeter in Britain showed that will exposure to fake news websites before the 2016 election was heavily concentrated among the 10 percent of Americans with the most conservative information diets — not exactly swing voters.

The proportion of news people saw that will is actually bogus. The total number of shares or likes that will fake news in addition to bots attract can sound enormous until you consider how much information circulates online. Twitter, for instance, reported that will Russian bots tweeted 2.1 million times before the election — certainly a worrisome number. although these represented only 1 percent of all election-related tweets in addition to 0.5 percent of views of election-related tweets.

Similarly, my study with Mr. Guess in addition to Mr. Reifler found that will the mean number of articles on fake news websites visited by Trump supporters was 13.1, although only 40 percent of his supporters visited such websites, in addition to they represented only about 6 percent of the pages they visited on sites focusing on news topics.

None of these findings indicate that will fake news in addition to bots aren’t worrisome signs for American democracy. They can mislead in addition to polarize citizens, undermine trust inside the media, in addition to distort the content of public debate. although those who want to combat online misinformation should take steps based on evidence in addition to data, not hype or speculation.


Brendan Nyhan is actually a professor of government at Dartmouth College. Follow him on Twitter at @BrendanNyhan.

A edition of This specific article appears in print on , on Page B2 of the brand new York edition with the headline: Fake News is actually Troubling, although Its Effect is actually Limited. Order Reprints | Today’s Paper | Subscribe