A New York Times piece in the “Review” section this past Sunday really got my attention. Amy Sullivan, a Chicago-based journalist, wrote, “Now that Stormy Daniels has confirmed on national television that Donald Trump initiated sex with her just months after his third wife gave birth to their child, at least half the country is asking: Surely this is a porn star too far for white evangelical Christians, right?” Sullivan then answered her own question by saying, “Wrong.” She went on to note, “As we celebrate Easter Sunday, nearly 18 months after Mr. Trump won the presidency with about 80 percent of the white evangelical vote, surveys show him retaining nearly all of that support. In contrast, white evangelicals re-elected George W. Bush in 2004 with 78 percent of their votes, but by May 2006 their approval had slid to 55 percent.” Sullivan didn’t note, but I will, that President Bush was a born-again Christian. President Trump—well, not so much.

But the Sullivan line that grabbed me was her suggestion that 80 percent of evangelicals “would vote against Jesus Christ himself if he ran as a Democrat.” Sullivan’s 80 percent estimate was not too hyperbolic, given a March 7-14 Pew Research Center survey of 1,466 adults nationwide that showed 78 percent of white evangelical Protestants approved of the job Trump was doing, roughly double that of adults overall, and just 18 percent disapproved.

But this isn’t just about religion, and it’s not intended to be a criticism of white evangelicals. In meetings since the first of this year with farmers and others in agribusiness in California, central Illinois, Kansas, Minnesota, Nebraska, and North Dakota, I came out convinced that in small-town and rural America, despite their serious concerns about some of Trump’s specific policies and statements on immigration (availability of farmworkers) and trade (tariffs, trade wars, and crop prices), there seems to be little he can say or do that would alienate many of them.

Today’s political divisions are very different from what they were 30 or 40 years ago. Today race, religion (specifically church attendance), geography (rural and small-town versus suburban and urban), and attitudes on social and cultural issues (e.g. abortion, guns, the environment, and women’s issues) drive voters’ attitudes about party and Trump far more than class and economic self-interest. This dynamic—some would call it identity politics, for which conservatives often criticize Democrats—is alive and well on both sides of the partisan fence.

A similar irony exists in the manufacturing sector. Whether China and other countries are playing fair on trade or not, there are far more jobs in the sectors using aluminum and steel than in producing these metals, but don’t expect workers whose jobs could be adversely affected by tariffs to bail out on Trump and the GOP (even if Trump’s trade positions are not clearly part of traditional Republican doctrine). The same could be said for white, working-class people who may well end up paying higher prices because of the tit-for-tat tariff fights—don’t expect those folks to change sides on the president’s job approval or their preferences for Congress. Working-class voters in urban areas or farmers in rural areas, including those in Michigan, Pennsylvania, and Wisconsin who effectively determined the outcome of the last presidential election, may well not bail out on him now.

The pattern repeats itself: Upscale, college, and graduate-school-educated suburbanites and city dwellers whose pocketbooks might actually be helped by Trump and the Republican Congress are instead opposed to them—in part because these voters may support abortion rights, some gun controls, and strong environmental standards.

This is not at all to say that what Trump has said and done over his 14 months in office doesn’t matter, but rather that so much is already baked into the cake.

Sticking with the kitchen metaphor, according to the source of all knowledge (Wikipedia), “In cooking, reduction is the process of thickening and intensifying the flavor of a liquid mixture such as a soup, sauce, wine, or juice by simmering or boiling.” Trump’s behavior and actions have reduced his support to its essence. And the Wikipedia definition goes on to say, “While reduction does concentrate the flavors left in the pan, reducing too much will drive away all liquid in the sauce, leaving a ‘sticky, burnt coating’ on the pan.”

I have given up looking at the events of any given week and trying to anticipate how Trump’s approval rating might move up or down. If Trump were a stock, his trading range would be very narrow; it moved down a bit late last summer and through the fall, up some after passage of the tax bill, then started to settle a bit, and then more recently went back up a bit. It’s a good bet that over the next six months from now, his rating is not likely to be much lower or higher than it is today.

What’s more relevant than approval numbers is who shows up on Election Day. We know that Trump benefited from a greater intensity of support in the closing weeks of the 2016 campaign, particularly among white, working-class, small-town, and rural voters, most specifically across industrial, Rust Belt states, while Hillary Clinton’s side was handicapped by ambivalence. That intensity does appear to be greater on the other side of the political ledger today. It’s hard for a Republican not to feel a bit unsettled by the 45 to 48 percent who strongly disapprove of the job Trump is doing, the latest turn of events on the gun issue, the activism of younger people and a legion of moms, or the teacher walkouts or threats of job actions even in conservative states like Arizona, Kentucky, Oklahoma, and West Virginia.

There are large forces at work in American politics that channel voters in one direction or the other, forces very different than those that drove voters a generation or two ago. We just have to get used to it.

More from the Cook Political Report