Political scientists have long noted that politics is a competition between groups with diverse and competing interests. During campaigns, candidates actively attempt to sway certain groups and vilify others in order to garner support.

In this year’s election, scholars and commentators have argued that the success of Donald Trump’s campaign is a consequence of pitting racial groups against each other. Specifically, they argue that Trump is appealing to whites who feel they are losing their influence to other racial and ethnic groups. Scholars have also noted that the election of Barack Obama may have actually increased racialized thinking among whites.

There is considerable anxiety among whites about American national identity. The growth of the Hispanic population has created questions about what it means to be an American. Further, the increasing number of attacks on U.S. and European cities has heightened a sense of a domestic and international conflict with Muslims.

So, it is not surprising that since launching his campaign in June 2015, much of Republican presidential candidate Donald Trump’s campaign rhetoric has addressed these themes. His early rhetoric focused on immigration from Mexico, but subsequently broadened to include anti-Muslim rhetoric. He speculated Syrian refugees were terrorists and asserted that Muslims celebrated the collapse of the World Trade Center on September 11, 2001.

While racial and religious hostility has often been noted as a factor in Trump’s support, it has not necessarily been clear which of these is the most important: Is it race or religion?

My research looks at the role religious feelings are playing in this election. What I have found is that white Republican and white Democrat feelings toward Muslims is strongly associated with their candidate choice and their willingness to vote.

Here’s how I did the study

I used data from the American National Election Study (ANES) 2016 pilot to find out how group preferences were playing a role in support for Trump. The survey was conducted between January 22 and 28, 2016, just before the first primary.

To demonstrate the extent to which whites prefer their racial group over Muslims, blacks and Hispanics, I used a measure referred to as polar affect. To create the measure, I individually subtracted the Muslims’, blacks’ and Hispanics’ feeling thermometer scores from the whites’ feeling thermometer scores. Because my primary concern was to see how whites perceived other racial and religious groups and how that influenced their support for candidates, this analysis is limited only to white participants.

Negative scores indicated whites prefer the minority group over their racial group, zero indicated indifference, and positive scores indicated whites prefer their racial group over the minority group. For instance, if a respondent scores whites at 80 and Muslims at 50, she would be considered to prefer whites 30 points more than Muslims.

Here is what I found

My results show a preference among whites for their own racial group over these minority groups. What is noteworthy is that the preference was significantly higher when Muslims were the comparison group.

One reason for this could be attributed to the survey being administered six weeks after the San Bernardino terrorist attack in which 14 people were killed and 22 were seriously injured. However, that might not offer a full explanationnegative attitudes toward Muslims have been fairly consistent throughout this election.

These preferences were even more stark when I compared partisans: Republicans expressed the strongest preference for whites over Muslims, blacks and Hispanics, compared to independents and Democrats. These relationships held even when accounting for demographics, such as age, sex, education and income.

Preference for whites over minority groups. 2016 ANES Pilot

In a match-up between Hillary Clinton and Donald Trump, a strong preference for whites over Muslims was the most important relationship. This preference for whites over Muslims benefited Trump and hurt Clinton.

Clinton had a 47.7 percent chance of being chosen by whites indifferent to Muslims, Trump had only a 29.9 percent chance of being the choice. This reversed when it came to whites with a strong preference: Trump’s chances doubled, whereas Clinton’s chances dropped by half.

Democrats, with a strong preference for whites over Muslims, are less likely to choose Clinton as their presidential vote choice. These voters do not move to Donald Trump; rather, they chose a third-party candidate or abstention.

Republicans, with a strong preference for whites over Muslims, have an 84.4 percent chance of choosing Trump. Further, they have virtually no chance of abstaining.

These results reveal that a strong preference for whites over Muslims energized Republicans, but deflated Democrats.

  Presidential vote choice. 2016 ANES Pilot.

Is this about race or religion?

Given the association between attitudes toward Muslims and support for Donald Trump in this election, one could ask whether this is a story about religion or race.

What is important to note is while the importance of religion has decreased in other Western industrialized nations, the decline has been much slower in America. Religion remains a critical part of understanding American identity.

It is, in fact, possible that exposure to terrorist attacks and international conflicts might have increased importance of religion in America. In 1996, 50.9 percent of white Americans endorsed the idea that it was important for one to be a Christian in order to be an American. But, in 2004, 64.4 percent of Americans endorsed that criterion.

Religion is also tightly intertwined with race in America. Many of the shameful periods of American history, such as slavery and Jim Crow, were justified in religious terms.

Finally, data from the 2016 American National Election Study (ANES) pilot demonstrate that whites who believe they are being discriminated against are also likely to believe American Christians are being discriminated against. These results hold even when examining whites who do not identify as Christian.

All of these issues make it difficult to disentangle racial and religious attitudes.

A language of hostility?

Overt racial hostility is no longer acceptable; however, certain forms of religious hostility may be.

For example, critics of President Obama have shied away from using his race, but a significant portion of the electorate has used his religion.

The 2016 ANES Pilot found that 37.8 percent of whites believed President Obama to be a Muslim. Of those who held this belief, 71.8 percent chose Trump over Clinton.

Scholars have paid close attention to the role of racial hostility in elections. However, this election has demonstrated a need to pay closer attention to the role of religious hostility as well as how the experience influences the political and social engagement of religious minoritiesirrespective of the election results.

The Conversation

This article was originally published on The Conversation. Read the original article.