Republican presidential candidate Donald Trump speaks during a campaign rally at St. Norbert College in De Pere, Wis. The Wisconsin primary is Tuesday. (Scott Olson/Getty Images)

Whit Ayres is president of North Star Opinion Research, a GOP polling firm, and was the pollster for Marco Rubio’s presidential campaign. He is the author of “2016 and Beyond: How Republicans Can Elect a President in the New America.”

Some Republicans have been trying to talk themselves into believing that Donald Trump would not be so bad as the Republican nominee for president. Maybe he really would scramble the political deck sufficiently, they think, to allow the party to retake the White House.

Those Republicans are whistling past the graveyard. A Trump nomination has as much chance of success in the general election as Trump University, or Trump Mortgage, or Trump Shuttle, or Trump Vodka, or Trump Casinos. Trump is an electoral disaster waiting to happen.

Demographic trends make clear that a Republican nominee who hopes to win a majority of the popular vote in 2016 must gain either 30 percent of the nonwhite vote or 65 percent of the white vote, a level not seen since President Ronald Reagan’s 49-state landslide sweep in 1984. Trump doesn’t stand a chance of doing either one.

Gain 30 percent of the nonwhite vote? Trump’s favorable to unfavorable ratings in the latest Post-ABC News national poll among nonwhites are 16 percent to 81 percent (72 percent strongly unfavorable); among Hispanics alone — the largest and fastest-growing minority group — they are 14 percent to 85 percent (74 percent strongly unfavorable). A Trump nomination could lock Hispanics into the Democratic column for a generation or more, threatening Republican presidential victories as far as the eye can see.

Gain 65 percent of the white vote? Trump’s favorable to unfavorable ratings among white women are 29 percent to 68 percent. His strongly unfavorable rating of 55 percent among white women is more than four times his strongly favorable rating among that group of 13 percent. How in the world could Trump reach 65 percent among whites overall if more than two-thirds of white women give him an unfavorable rating today, and if most of that unfavorable rating is intensely negative?

And that’s not all. Millennials have now passed baby boomers to become the largest generation. Trump’s ratings among millennials are now 18 percent favorable to 80 percent unfavorable, with 70 percent strongly unfavorable. Partisan identifications that are formed when voters come of political age tend to stick with them the rest of their lives, and a Trump nomination could push an entire generation into the Democratic column.

Negative ratings at these levels are historic for a major party nominee, creating hurricane-force head winds. And just imagine what Trump’s unfavorable ratings would be once the Democrats’ ads finish running.

But what about “the missing white voters” whom Trump supposedly would energize and bring into the electorate? Weren’t there more than 4 million whites who voted in 2008 but not 2012? Yes, and Mitt Romney lost by 5 million votes. Had every one of the missing white voters turned out and voted for Romney, he still would have lost.

Those whites who did not vote were concentrated in the deepest red states — Arkansas, Oklahoma, West Virginia — where President Obama never had a chance and lack of competition drove down turnout. No evidence exists of a dramatic falloff among white voters in the swing states that decide the outcome of a presidential contest.

Conservatives are fond of calling for dynamic scoring of the financial impact of changes in tax policy, where individuals’ behavior is incorporated into the estimate. For example, doubling the tax on yachts does not generate twice the revenue, because people in turn do not buy as many yachts.

We should apply the same concept to voter turnout. Would a Trump nomination bring more white voters to the polls? Undoubtedly, based on the high viewership of Republican debates and the high turnout in Republican primaries so far. Higher white turnout could conceivably put some of the Democratic-leaning but overwhelmingly white states of the Great Lakes region into play.

But the other side gets to play this game, too. There may have been 4 million fewer whites who voted in 2012, but there were also 12 million eligible Hispanic voters who stayed home. We know it is easier to drive voters to the polls to vote against rather than for a candidate — witness the tea party’s success in 2010 and 2014 in trying to stop the Obama agenda.

How difficult would it be to increase Hispanic turnout, given Trump’s ratings and his threat to deport 11 million immigrants? The easiest job in U.S. politics in 2016 goes to the Democratic operative charged with doubling Hispanic turnout to stop Trump. And that could put some Republican-leaning states with higher nonwhite populations, such as Georgia, into play.

Trump has a serious Republican problem as well. Since 1984, no victorious Republican presidential candidate has received less than 91 percent support from Republicans. Trump’s favorable to unfavorable ratings among Republicans are 52 percent to 47 percent, with 34 percent strongly unfavorable. A candidate beginning a general election campaign with almost half of his party holding unfavorable views is a non-starter. Contrast that with Hillary Clinton’s favorable to unfavorable ratings among Democrats of 78 percent to 20 percent.

A Trump nomination would put a Democrat in the White House, seriously threaten Republican majorities in Congress and leave the Republican Party in shambles. Let’s hope Republicans wake up before it’s too late.