Trump’s Tricky Math

By Henry Olsen

Monday, October 21, 2024

 

The Trump campaign released a polling memo on October 13 contending that he was ahead by two points in the seven swing states. A close read of what the memo said — and didn’t say — nonetheless shows that the race still hangs by a thread.

 

The memo was authored by campaign chiefs Susie Wiles and Chris LaCivita and polling maestro Tony Fabrizio. It argues that their polling shows that Harris’s favorability is down slightly since Labor Day in the battleground states and that Trump now leads 49–47 in those. So far, so good for Trump.

 

It then compares Trump’s internal numbers among blacks and Hispanics in those states with the numbers in those states from the 2020 exit polls. It states that Trump’s margin of defeat among blacks has shrunk by 20 points (81 to 61). It also says he leads among Latinos by seven points whereas he had lost them by 25 points against Joe Biden.

 

Both pieces of data, if correct, are fantastic news for Trump. But given the large numbers of both groups in those states, such massive gains should have him more than two points ahead.

 

Assume that the campaign’s definition of battleground states is the same as the conventional, seven-state grouping (Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania, and Wisconsin). Trump actually carried the four states with the largest shares of minority voters (all but the three Midwestern states). If Trump is doing 20 points better with blacks and 32 points better with Latinos, he should be romping to victory in those states.

 

Georgia’s electorate was 29 percent black and 7 percent Hispanic, according to the exit poll. Improving by 20 points on the margin with blacks would shift the statewide margin 5.8 points in his favor. Doing 32 points better with Latinos would shift it another 2.2 points.

 

Those changes would therefore turn a painful 0.2-percentage-point loss into a comfortable 7.8-point lead.

 

The same math applies in Arizona, whose electorate was 19 percent Hispanic in 2020. A 32-point gain on the margin would shift Trump’s statewide margin 6.1 points in his favor. That 0.3-point loss turns into a 5.8-point win.

 

Nevada and North Carolina show similar swings to Trump if the memo’s numbers are right. Applying the memo’s figures to the shares of each demographic in that state’s electorate yields a pro-Trump statewide shift of 6.9 points in Nevada and 6.4 points in North Carolina.

 

These shifts would also help Trump in the Midwestern swing states, although to a smaller degree. Applying the memo’s swings to the shares of blacks and Latinos in those places shifts the margin 3.9 points in his favor in Pennsylvania, 3.1 points in Michigan, and 2.4 points in Wisconsin.

 

If true, all of that is great news for Trump fans. But here’s the rub: This analysis assumes that Trump’s margin with whites in the swing states is unchanged.

 

Trump lost the seven swing states by a mere 238,278 votes out of 31.1 million cast. If we apply the average swing in the four non-Midwestern states and assume identical turnout, he gains 1.065 million votes on the margin. That alone would give him an 800,000-vote lead in the battleground states, or a nearly three-point lead.

 

Applying the average swing in the Midwestern states expands his lead further. The average swing was 3.1 points, which would increase his margin by another 489,000 votes, assuming 2020 turnout. He would therefore lead all seven states by nearly 1.3 million votes out of 31.1 million cast.

 

That’s a lead of over four percentage points. Yet the memo says he only leads by two. What gives?

 

There’s only one explanation: Trump’s own polling must show that he’s losing ground among white voters.

 

That’s what the public polling averages show. The Cook Political Report keeps a running average of the polls by key demographic groups, and it shows Harris with a 17-point lead among whites with at least a four-year college degree.

 

That’s nearly double the nine-point advantage the data on the Cook side say Biden had in 2020.

 

College-educated whites were between 23 (Nevada) and 36 (Pennsylvania) percent of 2020 voters in the battleground states. Lose eight points on the margin with them and you lose somewhere between two and 2.5 points overall.

 

In other words, factor this unstated movement into Trump’s memo, and the numbers all make sense.

 

The problem is that college-educated whites are most important in the Midwest, where Trump’s minority-driven gains are weakest. Factor these losses into the equation, and while Trump still leads or is tied in each of those states, he leads by frighteningly close margins.

 

And the memo’s figures show much larger gains among non-whites than most public polling. If Trump is really only 15 points better on the margin with blacks and 20 among Hispanics, for example, those narrow leads turn into narrow losses — again.

 

The Trump memo is an important window in the campaign’s intelligence and strategy. It shows he’s got a real shot, but he has to do a lot right over the next two weeks to fulfill that promise.