Jeremy Carl uses a National Review Online column to place Donald Trump’s presidential victory in a larger context for the Republican Party’s future.

Over Labor Day weekend in 2015, four months before the first primary votes were cast in Iowa, a somewhat obscure policy analyst and strategist wrote a confidential memo to a presidential campaign he was informally advising. He outlined what he believed was the path to victory for the GOP in the 2016 presidential election, based on election simulations he had run using a couple of publicly available models. He attacked the GOP’s official Election 2012 post-mortem as being politically motivated and divorced from actual voting data. A portion of the memo, edited lightly for style and length, is reproduced below.

The most important voters to win in the entire country are white voters in the Midwest and upper Midwest.

According to the modeling done here, if [Candidate X] could win white voters at Reagan 1984 percentages (66 percent) and at Bush 2004 turnout levels (67 percent) and we assume African-American turnout was to return to historical levels and percentages for the Democrats, we could win the presidency without winning a single Hispanic, Asian, Native American, or Arab vote. Think about that, because that is a staggering statement, and it’s a true one.

The converse is equally staggering: We could win 53 percent of the non-black minority vote (Hispanics, Asians, Native Americans, etc.), and if white and black turnout and voting percentages stayed the same as they were in 2012, the Democrats would win the presidency. These stats indicate the foolishness of the approach of pandering to chase after minority votes if that means ignoring the interests of the GOP’s white voter base. Of course the Democrats and the media are always concern trolls about this subject — they want us to lose. What is dismaying is how many in the party have absorbed the media/Dem concern-trolling. Of course, that doesn’t mean for a second we shouldn’t pursue minority voters or pay attention to their interests. We can do far better with minority voters by going into communities, solving problems, having a positive economic agenda, a focus on families, and a hopeful message. But the numbers don’t lie — if we engage in any amount of ethnic pandering, and in doing so alienate or discourage even a small number of our base white voters, we will lose the election, period.

The voters who stayed home and didn’t vote for Romney, particularly in the upper Midwest, are just the sort we need to reach. The Reagan/Bush scenario outlined above flips Ohio, Pennsylvania, Wisconsin, and Iowa, along with Virginia and New Hampshire, to give us a victory.

Nothing came of the memo, which was received politely but not acted on. That candidate did not become the GOP nominee, or even come close to doing so. But while Donald Trump may or may not have gotten a similar memo, his campaign certainly acted according to the principles expressed in this one. He took all the states listed above except New Hampshire, where he lost by a hairbreadth, and Virginia, and added Michigan, where he won by a similar margin, taking advantage of the same demographic forces. Had the author assigned even a trivially small (10 percent) Hispanic vote share for the GOP, which he refrained from doing only to support his broader point about the importance of white turnout, it would have forecast Trump’s Florida win as well.

Now for a confession: That somewhat obscure strategist was this author. And while what is written looks prescient, it really didn’t take a rocket scientist to figure it out — just someone who hadn’t already decided upon the answer (“Win Hispanics with amnesty,” as pushed by the GOP’s 2012 autopsy) before doing the analysis.