Why is Trump campaigning in blue states?

Maybe he’s onto something.

Madison Square Garden, in an earlier iteration, when it frequently paid host to presidential candidates. (Photo by Frank Cooper)

It was once a tradition, in an earlier era of American politics, for presidential candidates to close out their campaigns with a rally at Madison Square Garden.

Between 1912 and 1968, Theodore Roosevelt, Woodrow Wilson, James Cox, John Davis, Herbert Hoover (twice), Al Smith, Franklin Roosevelt (three times), Alf Landon, Wendell Willkie, Harry Truman, Thomas Dewey, Dwight Eisenhower (twice), Adlai Stevenson (twice), John F. Kennedy, Lyndon B. Johnson, Barry Goldwater, and Richard Nixon all spoke at the legendary New York City arena in the final days of their bids for the White House.

There was a certain logic to this. Not only was MSG a legendary venue — “the world’s most famous arena,” as it is known — but New York was also, at the time, the country’s largest swing state. In the 15 consecutive presidential races represented above, Democrats won New York eight times; Republicans won the state seven times.¹ The margins were often very thin.

Donald Trump’s campaign announced on Monday that he will hold a rally at Madison Square Garden on October 27, following in the grand tradition of MSG closing speeches. This time, though, the logic makes less sense. New York is no longer a swing state. It has voted for the Democrat in each of the last nine presidential races; Trump himself has never received more than 38% of the vote there.

So why is Trump spending precious time at the end of a close election campaigning in a deep-blue state?

Actually, make that several deep-blue states. In the last week, while Kamala Harris has raced around the seven battleground states, Trump has also held rallies in California and Colorado,² both states where he is destined to lose. Plus, his late-October rally at the Garden will be his third New York event of the cycle, following previous events there in May and September.

Several theories have been floated to explain Trump’s unexpected detours. In Colorado, perhaps he was trying to emphasize migrant crime. In New York and California, perhaps he was trying to lift Republicans in contested House races. Maybe, one Republican strategist floated, he is trying to ensure a victory in the popular vote. In the case of the MSG rally, maybe the native New Yorker just wants to play the Garden.

Here’s another theory for you: Maybe Trump is reading up on his political science.

---

Christopher Devine, a political scientist at the University of Dayton, has literally written the book on presidential campaign travel: his tome, “I’m Here to Ask for Your Vote: How Presidential Campaign Visits Influence Voters,” was published last year by Columbia University Press.

In researching the book, Devine assembled a comprehensive database of presidential campaign stops in recent elections, and analyzed how the candidates performed in the places they visited. His finding? “The basic assumption is that if you make a campaign visit, you’re going to pick up votes,” he told me in an interview. “And generally what my research shows is that that does not happen typically.”

Using regression analysis, Devine found that — between 2008 and 2020 — only one presidential or vice presidential candidate’s campaign visits had a statistically significant effect on how they performed in the counties they visited: Barack Obama in 2008.

“On average,” according to Devine, “each additional Obama visit in 2008 increased his share of the presidential vote in a given county by 0.77 percentage points.” No other candidate in the last four cycles, he added, can be proven to have “won votes via campaign visits.”

Graphic from “I’m Here to Ask for Your Vote: How Presidential Campaign Visits Influence Voters” by Christopher Devine

Of course, this doesn’t mean candidates should just stay off the trail and quit campaigning altogether.

Interestingly, Devine found that — contrary to conventional wisdom — campaign visits are slightly more effective at persuasion (convincing previously opposed/undecided voters to support you) than mobilization (convincing supportive voters to actually cast ballots), although they aren’t particularly strong on either metric. Campaign visits also provide a boost in positive local media coverage, which is helpful considering people trust local news a lot more than national media.

Unilateral campaigning disarmament would also be unwise. “While you’re unlikely to gain much, if anything, from an individual visit,” Devine told me, “if you just give up on it and let the other campaign flood the information space in that area, whether that’s through ads, visits and other things, you actually might see more of an effect.” (In other words: Hillary Clinton not visiting Wisconsin made no provable impact on her Wisconsin vote share — Devine tested it — but it is still risky for candidates to hand a narrative to their opponent that they aren’t robustly campaigning.)

But, in general, maybe Trump is onto something. As long as he’s still holding events — so Kamala Harris can’t claim that he’s, I don’t know, hiding in his basement — why not just campaign across the country, if there isn’t much evidence that an umpteenth battleground state rally will make much of a difference in the race?

Certainly, that’s what the political science literature would say. “I’d like to think that they’re reading our stuff,” Devine said with a laugh, referring to himself and other academics who have found the county-level effects of presidential campaign visits to be quite small.

Still, Devine said, Trump’s strategy comes with little precedent. He could name only two other candidates who made a similar string of stops in states they were destined to lose: John McCain in 2008, who held a series of events in states he was stationed in his military career, to highlight his biography, and Bill Clinton in 1996, who campaigned in red states in order to boost Democratic House candidates.³

But McCain’s cross-country tour came towards the beginning of the campaign, not at the end, like Trump’s is. And Clinton turned his focus to down-ballot races only because he was confident that he would win the White House. (Indeed, he won in a 379-electoral-vote landslide.⁴) It’s possible that Trump is feeling just as cocky, but the incredibly tight public polling would suggest he has little reason to.

---

In addition to everything else, Trump’s three campaigns for the White House have often functioned as real-life political science experiments, offering tests on a grand scale for vexing questions like whether a candidate needs to raise the most money to win (no, the experts — and Trump’s 2016 victory — say) and whether voters adhere to fixed ideologies or merely follow their party leader (the latter, it would seem).

Now, he’s offering a real-time test of the research by Devine and others that suggests that where the candidates campaign matters less than we might think. More than that, he’s betting on a fully nationalized theory of elections. While the Harris campaign is going hyper-local — targeting specific counties in the battleground states where she hopes to “lose by less” — Trump’s on-the-ground operation has been shaky and his rallies in California and New York seem more aimed at receiving national than local media attention.

The theory could end up being borne out. It’s striking, in Devine’s research, that the only candidate who boosted their vote share with campaign visits was Obama, in an election that took place 16 years ago. You might think that’s a credit to his unique candidacy — but the only other contender whose visits came closest to having a statistically significant impact was McCain, Obama’s 2008 rival.

Perhaps changing minds via campaign visits was simply more feasible in a less polarized, less politically nationalized era. According to Devine’s calculations, presidential candidates since 2008 have generally made fewer campaign stops — potentially suggesting that campaigns no longer feel the need to barnstorm with their former ferocity.

As Trump shows, in the modern era, a rally in California can be seen by viewers across the country on national TV — and go viral on social media as easily as a rally in Pennsylvania. If the Pennsylvania rally can’t be proven to move many votes, Trump is gambling that his best play is to maximize the amount of time he spends doing interviews and other flashy events that will guarantee national coverage.

If the gamble works, it will make for a fascinating case study — and potentially influence how future candidates campaign. But, at this point, it’s hard to imagine many other candidates taking such a plunge, especially this late in a razor-thin election.

“Having a really spectacular rally at Madison Square Garden, to the extent that impresses people, it’s probably going to impress them just as much in Oklahoma as in Wisconsin or Alabama as much as Arizona,” Devine said. “So it’s just not well targeted… If you’re going to spend your time trying to win votes, why not go for maximum impact in a swing state rather than spreading your impact out over the entire country, including some states that, frankly, don’t mean much in the Electoral College?”

It’s true that the return on one more battleground state visit is, quantifiably, quite marginal. But that’s still a risky finding for Trump to hang his hat on, considering — his apparent confidence aside — that this is a race that will almost certainly be decided on the margins.

1. New York is also the state that has most frequently been the “tipping point state” in presidential elections: eight times, it has been the state that swung the Electoral College to the winning candidate. (Wikipedia defines “tipping point state” thusly: “In a list of states ordered by decreasing margin of victory for the winning candidate, the tipping point state is the first state where the combined electoral votes of all states up to that point in the list give the winning candidate a majority in the Electoral College.”) New York’s status as the most frequent tipping point state connotes its historic status as a swing state — although it faded as a swing state in the 1980s, and has not been the tipping point state since 1944.

2. OK, Colorado isn’t really a “deep” blue state — but it’s certainly no longer competitive at the presidential level. No Republican has won the state since 2004; Biden won it by 13 points in 2020.

3. Another canonical example of a candidate traveling to unexpected places is Richard Nixon, whose 1960 pledge to visit all 50 states meant that — in the closing days of a close campaign — he wasted time visiting far-off Alaska. Nixon’s travel is sometimes blamed for his defeat; 1960 was outside the scope of Devine’s research, but his findings would suggest (at least in a modern context) that a strategy like Nixon’s wouldn’t be expected to have an enormous impact on the race.

4. At least, that’s certainly a landslide by modern standards.