America’s widening political divide stands out above other nations, according to a new cross-country polarization study by Stanford University economists.

Over the past four decades, the chilly chasm of negative sentiment between Democrats and Republicans in the U.S. has nearly doubled, and it has grown faster and larger compared to the partisan climates of eight other established democracies analyzed in the study.

In three other nations — Canada, New Zealand and Switzerland — political polarization rose as well but to a lesser extent. During the same period, polarization decreased in the five other countries — Australia, Britain, Norway, Sweden and Germany.

The study — released as a working paper Jan. 20 by the National Bureau of Economic Research — is the first to provide cross-country evidence on long-term trends in inter-party sentiment. The research was conducted by Stanford economist Matthew Gentzkow, Jesse Shapiro of Brown University, and Levi Boxell, a Stanford PhD student. The researchers also teamed up for a 2017 study challenging the idea that the internet played an outsized role in causing polarization.

Today’s divisive discourse in and outside of government halls is a sign that partisanship has grown in the U.S., but the researchers wanted to evaluate whether or not this was a distinctly American phenomenon, and to shed light on possible factors.

“Understanding this trend is really important. There’s a lot of discussion on the rise of polarization and its implications,” said Gentzkow, a senior fellow at the Stanford Institute for Economic Policy Research (SIEPR). “But the reasons why polarization is increasing are unclear, and knowing whether a reason is specific to the U.S. or if it’s more universal is helpful in assessing those explanations.”

The trio of economists set out to understand trends and potential explanations behind the feelings of dislike or animosity between citizens aligned with opposing parties — an aspect of partisanship that notable political scholars, including Shanto Iyengar, Morris Fiorina and Neil Malhotra, all of Stanford, refer to as “affective polarization.”

Gentzkow and his colleagues compared nine member nations of the Organization for Economic Cooperation and Development that had longstanding national public opinion surveys on politically related questions.

Using survey data from 1975 through 2017, they analyzed the thermometer-like ratings of how “warm” or “cold” a respondent feels toward others in the opposing political party, or responses indicating the extent one sympathized with a given party.

“The average American used to feel positive about their own party and kind of neutral about the opposite party,” Gentzkow said. “But now they are positive about their party and quite negative about the opposing party.”

According to the study, in 1978, the average gap of negative sentiment between Democrats and Republicans was 27 points. By 2016, that difference grew to 46.

To evaluate possible explanations for the growing gap, the researchers examined some concurrent societal trends. If, for instance, the emergence of the internet was a driver of polarization, then the universal rise in internet usage and broadband penetration should have resulted in a greater degree of partisanship across the countries, the researchers say.

But contrary to what some may think, those particular factors did not appear to play an important role, as more than half of the countries in the analysis instead showed a shrinking political divide.

Similarly, the researchers found that partisanship trends were not apparently swayed by the relatively universal increases in global trade, immigration, or economic inequality.

Growing racial divisions, however, could have played a role, the study suggests. Countries with rising polarization had also experienced double the increases in the share of their non-white population compared to countries where polarization decreased.

Another potential factor could lie with how the U.S. has had a more distinctive shift in party coalitions over time. Lines along ideological issues deepened, and the makeup of political parties became more aligned along racial, geographical and economic attributes, Gentzkow said.

The researchers also suggested that the rise of 24-hour partisan cable TV networks may have played a larger role in the U.S. than elsewhere, citing a 2017 study by Gregory Martin and Ali Yurukoglu of Stanford. In addition, more public funds were spent per capita on public broadcast media in the five countries where polarization fell.

“We don’t have any way in this study to isolate the causes of polarization,” Gentzkow said. “We’re just looking for facts that might help set the stage for that future research more clearly.”

This article was originally published by Stanford Institute for Economic Policy Research (SIEPR) on its website.