Scientists at UCLA and Washington State University are seeing America's polarization play out at the family dinner table, with Thanksgiving visits that were 30 to 50 minutes shorter after the presidential election of 2016.
Economists Keith Chen and Ryne Rohla also saw that visits were even shorter for travelers from media markets with intense political advertising. Their findings appear in the latest journal Science.
"Politics actually affects how long we spend at Thanksgiving dinner," said Rohla, a doctoral candidate at WSU. "Also, these partisan differences are inflamed by political advertising."
After the 2016 election, both Chen and Rohla had uncomfortable, politically divisive Thanksgiving dinners and read anecdotal reports of Americans cancelling or cutting short Thanksgiving visits "with their most politically problematic relatives."
They also noticed that most academic attention to the nation's political polarization was focused on public institutions: voting in Congress, fewer bipartisan bills being written, the gridlock of our political system.
"The fact that political polarization is increasingly degrading our close connections with friends and family didn't have much empirical study and attention from social scientists," said Chen, an associate professor of economics at the UCLA Anderson School of Management. "That's where we leapt off with this project."
Where other social scientists might have undertaken a poll, Chen and Rohla used anonymous cell phone data that included more than 10 million users' home locations in the weeks before Thanksgiving and their locations on Thanksgiving afternoon. They merged this with voting data and weighted users based on the proportion of Democratic and Republican votes in their home precincts.
When users visited precincts that tended to vote differently than theirs, they stayed 30 to 50 minutes less than the average Thanksgiving visit of 4.2 hours in 2015, a political off year. Travelers going from Republican precincts to Democratic ones tended to shorten their visits 50 to 70 minutes; those going from Democratic precincts to Republican ones tended to shorten their visits 20 to 40 minutes.
Visits were up to three times shorter when the travelers came from media markets with lots of political advertising.
"That really sealed it for us, that this was definitely a political effect," said Chen. "It's not just because families that have members that cut across political boundaries are different somehow. In 2016, those are the families that were hurt the most and they were hurt especially in places where there was a lot of political advertising. That same thing is not true in 2015, before these ads were run."
"Altogether," write Chen and Rohla, "an estimated 33.9 million person-hours of cross-partisan discourse were eliminated, perhaps creating a feedback mechanism by which partisan segregation reduces opportunities for close cross-party conversations."
The two posted an earlier version of their study last fall on the electronic preprint repository arXiv. They have since looked more closely at the data and found even greater effects.
Their insights are a testament to the power of large amounts of data, which in this case can offer a look into our domestic lives without ever entering our homes.
"This kind of data really opens up a whole bunch of doors that you never would have been able to open before," said Rohla. "We can ask people about Thanksgiving but generally speaking they're probably not going to keep track of the exact duration. People's memories are not perfect and people also have a tendency to lie in surveys if it's socially unacceptable to present a certain view. This way kind of cuts through all that." Still, they find their results disturbing. Both researchers see the dinner phenomenon as another iteration in the long decline of social capital, in which values like trust and cooperation are eroded as people participate less in civic activities.
"One of the things that's supposed to save you from this are deep civic institutions like close family ties and conversation over Thanksgiving dinner," said Chen.
"Some people don't see losing dinner with relatives as a particularly large cost," said Rohla. "I've talked to people who are, 'Well, so what?' Personally I think it is concerning. To me it's a symptom of a broader decline in the social fabric of the United States."