South Bend, Ind., Mayor Pete Buttigieg stopped by a public radio station while in California on Thursday as part of his outreach for a possible run at the 2020 Democratic presidential nomination.

He was asked by hosts from San Francisco’s KQED what he had noticed while in California. He mentioned some tech-related things at the outset, then moved on to discussing supporters of President Trump.

“The other thing I’ve noticed, though, is there are some folks I encounter here who seem to have trouble believing that things like Trump voters actually exist,” Buttigieg said. “And so I feel sometimes like I’m an emissary from the middle of the country just pointing out that things look a little bit different in rural communities, industrial communities like mine and that we really need to find ways to knit this picture back together into one America.”

Asked if people were receptive to the message, he said they were.

“I think people are just puzzled by why people where I’m from make the political choices sometimes that they do,” he said.

It’s obvious what Buttigieg is saying here: There are obvious cultural differences between rural Indiana and San Francisco. But it’s still worth pointing out that the idea that people in San Francisco might not believe in the existence of Trump voters is undercut by the fact that there were tens of thousands of Trump voters in San Francisco.

Granted, Hillary Clinton won the city easily. But Buttigieg’s comment raises an interesting point. While the general conversation about Trump’s base of support focuses on the upper Midwest and, in particular, voters who flipped from voting Democratic to voting for Trump, that narrative — and those maps of heavy red counties in the middle of the country — miss the fact that Trump also earned an enormous amount of support on the coasts.

While Clinton won 40 percent of her total votes in the states on the West Coast, and from Virginia north on the East Coast, Trump won 26 percent of his votes there. If we visualize not how the candidates did everywhere but, instead, highlight the vote totals each candidate received in counties that the other candidate won, the number of hidden Trump votes, if you will, becomes apparent.

Or, we can look at it another way, isolating Trump’s support in Clinton counties from his support overall. Notice that a lot of the circles indicating his support in places Clinton won are much larger than the circles in places that he won — indicating that he got a lot of votes in those places where Clinton nonetheless carried the day.

Creating the same map for Clinton shows a bigger gap between her best counties (like Los Angeles) and how she fared in places that she lost.

In more than a third of states that Trump won, he received fewer votes than he received in at least one county where Clinton won. Usually that was Los Angeles County, where he got more votes than he did in Alaska, Arkansas, Idaho, Kansas, Mississippi, Montana, Nebraska, North Dakota, South Dakota, Utah, West Virginia and Wyoming.

In fact, there were 34 counties where Trump lost to Clinton but in which he earned more votes that he did in Alaska. That includes six counties in California.

Buttigieg would have needed to travel from San Francisco on I-80 for only an hour or so to get to Sacramento County, where Trump got about 190,000 votes. That’s fewer than Clinton but more than the 163,000 Trump got in Alaska.

Again, Buttigieg’s point was that the general stereotype of Trump voters — rural, blue-collar, white — might have seemed foreign to San Franciscans. But the issue is that the stereotype is often wrong. There are hundreds of thousands of people on the coasts who may have voted for Trump for very different reasons than people in Buttigieg’s home state of Indiana.

If Buttigieg really wants to understand Trump voters, incidentally, he was in the right place. For every Trump voter in Indiana, there were three in California.