The answers to five questions

I've been at this a while, gang, this business of looking at the world around me and wondering what patterns hold it together. It's the seventh-grade-biology-class version of The Matrix; I see a thing and suddenly find myself trying to dissect the strings of numbers that make it what it is.

Hence my appreciation for data visualizations. With charts, graphs and maps, no dissection! The numbers are on the outside.

If, for example, you are perusing the 1910 census on a Friday evening and come across a map of the territorial expansion of the United States, you can spend 14 to 18 hours picking out patterns and datapoints, the way nature intended. (Two of those hours will be spent ruminating on the grim distinction between "ceded," "annexed" and "acquired.")

Over the course of this week, I realized that this extremely cool tendency to dig into data had provided me with the answers to several questions that I'd been mulling over for a while. So, instead of my normal approach to this newsletter (charts that I saw or made organized into "chapters" due to some inscrutable affectation), I thought I'd do something a little different (present those charts as the answers to the questions that prompted their creation).

Don't worry; I'm very confident that this abrupt departure from the norm will not prove to be too confusing.

So:

Where do losing presidential candidates' votes come from?

Last week, I wrote a column about the weirdness of the "will Texas turn blue" question, as though Democrats winning one Senate race would constitute an inversion in the state's deeply rooted Republican politics.

In writing that piece, I ran a calculation that I'd run a few years prior, determining if the losing candidate in the most recent presidential election (Kamala Harris in 2024 and Donald Trump in 2020) had won more votes in states they'd won or in states they'd lost. The answer, in both cases, was they won more votes in states they lost. In other words, most Harris voters lived in Trump-voting states and, four years prior, most Trump voters lived in Joe Biden-voting states.

Setting aside questions of the extent to which this fosters resentment — I am a dissector of numbers, not of psychologies — I was curious how common this divergence was. Pulling numbers from U.S. Election Atlas, I got my answer.

It's very common, indeed. Over the past century, losing presidential candidates have almost always won more votes in states they lost than in states they won.

How to read this chart: It took me a while to figure out how to effectively visualize this. In short, each election is colored according to the victor. The dots show the net percentage of votes cast in states won by the Democratic candidate and those in states won by the Republican. So dots to the right of the center line indicate that a candidate won more votes in Republican-voting states.

In almost every case, those dots line up with the actual election winner, meaning that both candidates won more votes in states that voted for the winner. Which makes sense; the winners of presidential elections generally win more states. (In 2020, both candidates won the same number of states, excluding D.C.; in 1960, Richard Nixon won more states.)

Note the exceptions, though, the circles with colored centers like delicious blueberry cookies. In 2000, 2004 and 2016, the losing candidate earned more votes in states they won. In 2000 and 2016, those losing candidates also won the popular vote. Making 2004 a rarity: a contest in which the candidate who got fewer votes overall earned more of them in states he won than in ones that he lost.

If that wins you bar trivia at some point, you owe me a beer.

Does higher primary voting lead to general election wins?

One of the reasons that Democrats are optimistic about "turning Texas blue" in November is that Democratic candidates won more votes in the recent primary than did Republican ones — 7 percent more, in fact. Doesn't that augur a victory in the general?

Happily, we have plenty of data about past Senate elections which can help answer that question. The Federal Election Commission, for example, has spreadsheets with the results of every recent federal election, including primaries.

Looking at that data, we see a clear pattern: the party that wins more votes in the primary usually wins the general.