Skip to content Skip to navigation

Don’t blame the web for the growing political divide

March 23, 2017

All the blame cast upon the internet and social media for intensifying America’s political strife does not dovetail with real data, according to a new study by researchers at the Stanford Institute for Economic Policy Research.

Political polarization has been growing the most among older Americans, and that’s precisely the demographic group that is the least likely to go online, the study found. 

“You certainly see polarization on the internet and social media, but that doesn't mean the internet is causing the polarization you see,” said Levi Boxell, a 23-year-old predoctoral researcher at SIEPR. Boxell co-authored the paper with SIEPR Senior Fellow and economics Professor Matthew Gentzkow and Jesse Shapiro of Brown University.

The researchers examined data from the American National Election Study and found that the biggest increases in political polarization from 1996 to 2012 occurred among Americans aged 65 and older. In fact, their polarization growth was six times higher than that of Americans aged 18 to 39 during that period. 

Polarization levels — indicators of partisan behavior between Democrats and Republicans — for middle-aged Americans also measured consistently higher than for the younger demographic.

As the nation grappled with deepening divisiveness, the researchers tested a popular theory of how the rise of the internet and its digital echo chambers — where like-minded people can buttonhole themselves to selected sources of online information — could be perpetrating political acrimony.

“None of this is to say that what’s happening in social media and digital media is not important and couldn’t change things down the line. I just think it’s not the whole story,” Gentzkow said.

The empirical evidence, which used nine different measures of polarization, suggests that social media and the internet are not the main culprits for the widening political chasm. In assessing polarization, researchers used measures ranging from voting along party lines to a thermometer-like rating of how “warm” or “cold” a respondent feels toward others on the opposing side of their political party or ideology. 

“Our evidence,” the study stated, “rules out what seem like the most straightforward accounts linking the growth in polarization to the internet.” 

Earlier research by Gentzkow has shown how political polarization existed long before the days of the internet, and the new study here affirms how political polarization has increased since at least 1972. 

Indeed, millennials are more plugged-in than their elders: 60 percent of 18– to 39-year-olds reported consuming political information online during the 2012 presidential campaign, compared to only 26 percent of those aged 75 or more, the study found.

But even if tweets and Facebook posts are aiding and abetting the political rancor among Generation Y, then what’s behind the surge in polarization for those who are registering limited internet use and almost negligible use of social media?

The study cites a few possible contributing factors to explore: It could be that young social media users are directly affecting the views of older adults, or they could be indirectly asserting their influence to older adults by their roles in the selection of political candidates and the creation of content in traditional media.

“I think it’s much more likely there are some deeper structural factors that have nothing to do with media — like increasing income inequality — that could be related to this,” Gentzkow said. “And other media trends — like the rise in partisan cable (television content) — might play a significant role.” 

The paper by Gentzkow, Boxell and Shapiro was published March 20 by the National Bureau of Economic Research.