Then-candidate Donald Trump greets supporters after finishing speaking during a campaign event at the Jacksonville Equestrian Center in Jacksonville, Fla., in 2016. (Jabin Botsford/The Washington Post)

The Deseret News and the Center for the Study of Elections and Democracy at Brigham Young University are out with results of a survey on the American family. With 3,000 respondents, the survey provides a wealth of information on everything from cellphone use to drug addiction to political behavior. We’ll look at a few of the findings that have particular insight into hot-button public policy issues and political voting patterns.

Views on health care confirm that Republicans have been out of tune with ordinary Americans:

Consistent with other surveys, many elements of the Republican plan proved highly unpopular, regardless of income or family medical challenges  . . Few respondents to our survey wanted to give states control over the choice to cover pre-existing conditions, and even fewer wanted to cut federal funding for Medicaid. A little less than half of respondents wanted to eliminate the insurance mandate. The only exception to this unpopular slate of reforms is allowing for a greater variety of plans, though again, we did not offer respondents any detail about what that diversity of insurance offerings might look like. Though the differences tend to be modest, lower income respondents tend to be less supportive of the Republican plan than middle- or upper-income Americans. …

Though very few [Hillary] Clinton voters or nonvoters wanted state control over insurance rules, nearly 45 percent of [Donald] Trump voters wanted to empower the states – still not a majority, but much greater support than we find among other respondents. Similarly, about one-third of Trump voters wanted to cut federal funding for Medicaid. This, too, is far short of a majority, but far more than the 4 percent of Clinton voters and 18 percent of nonvoters. Still, because these core elements of the plan did not garner majority support even among Trump supporters, these results may provide some insight as to why the House bill did not ultimately become law. Other elements of the plan were, however, highly popular among Trump voters, including ending the insurance mandate and providing a larger variety of plans. The idea of eliminating the mandate was extremely unpopular among Clinton voters, and nonvoters were evenly split. Finally, clear majorities of all different political perspectives supported the idea of greater diversity in insurance plan offerings.

Republicans might want to consider, then, that paying for an already unpopular tax bill with Medicaid cuts would be a political loser.

Political watchers have looked at economic conditions and racial views to explain voting patterns in 2016. But there is another critical element that gets less mention:

Trump voters are simply less socially connected than are Clinton voters. On the survey we asked a series of questions about whether or not people turned to others for help with things like childcare, child rearing advice, relationship advice, help with finances, taking care of property or transportation. Though this data can be cut in several different ways, for those who reported only one such connection the probability of voting for Trump was higher than for those who reported three or more such connections. …

Trump voters appear to be more or less like Clinton Voters in terms of spending Thanksgiving with their immediate families (the same was true of other options like friends or being alone), but they look much more like nonvoters when it comes to spending time with extended family. Neither group is as likely to get together with extended family as are the Democratic voters. The inescapable conclusion is that support for Trump is related to social connectedness. His voters have relatively fewer connections outside of their immediate families, fewer connections to the welfare state and even tend to look more like nonvoters with respect to Thanksgiving plans and extended families.

This may explain why Trump voters cared less about policy results than the emotional feeling that Trump was going to understand and take care of the “forgotten man and woman.” When he makes constant references to cultural markers (e.g. Merry Christmas, loathing of liberal media), he is in effect reminding these otherwise disconnected voters that he’s “one of them.” Put differently, the extreme tribalism that characterizes Trump’s politics can be seen as a critical organizing factor for people looking for a sense of belonging. It’s beyond logic and reason, which explains why his core supporters (roughly a third of voters) are unmoved by political outcomes.

Finally, it turns out that attitudes on immigration change considerably when the family element gets considered, but not for a whole lot of Trump voters. The pollsters asked “whether or not respondents favor or oppose ‘deporting unauthorized immigrants [even when it separates family members].’ However, we varied whether or not the respondent was shown the text in brackets. Essentially, this question tests how much people favor deportation of immigrants when they are reminded that it may break up families.” The results were mixed:

The family reminder drops the percentage who favor the policy [of deportation] by about 18 percentage points and raises opposition by the same amount. Respondents clearly change their opinion when faced with the reminder of how immigration impacts families (at least potentially). …[However] the effects of the experiment are overwhelmed by the electoral choices of the respondents. Some Trump voters are moved by the plight of immigrant families, but by more than three out of four of them still favored deportation when shown the text about separating families. In contrast the number of Clinton voters who favor deportation drops from about three in ten to about one in ten when they see the text. Nonvoters are the most affected by the reminder. About half of them favor the policy when it is shown without context. Adding the text cuts that proportion in half.

In short, “family values” to a large degree has faded for Trump voters. Trump’s raging nativism seems a greater motivator for these voters. Maybe Democrats should start embracing — and redefining — the notion of “family values.”

There is plenty more for data junkies to pore over. What stands out here is that Trump’s emotional, nativist appeal has hit the sweet spot for his base, although that doesn’t guarantee support for specific policies (e.g. Trumpcare). For all the talk of “values voters” on the right, the main concerns in 2017 are not cultural. The study finds that “more people believe that the most serious problems facing marriages and families are economic, and fewer believe that the challenges are primarily cultural.” The survey reports: “Since 2015, there has been an 11 percentage point increase in the people who say the top problems facing families are economic and a 17-point decrease in the percentage of respondents choosing cultural issues. In particular, the costs of raising a family and high work demands on parents seem to be a greater consideration for more survey respondents in 2017 than they were in 2015.” That certainly seems to be an opening for Democrats to claim the mantle of the party of economic security, the party that reduces stress and strain on American families.