Wanted to draw everyone’s attention to the Substack Midterms Hub which is going to have content from Slow Boring and many of this platform’s other fine outlets today, tomorrow, and beyond — a great place to go for updates, results, commentary, analysis and more.
One week before Election Day, Nate Silver’s model said that Democrats had a 76 percent chance of taking the Senate tomorrow, while Lakshya Jain’s model said they had a 65 percent chance.
I know the people who do this kind of work mostly focus on questions like “does this model give good gambling advice” and ignore metaphysical issues. But I’m always interested in what these claims really mean. Statistically speaking, to get enough power to reliably distinguish between these forecasts, you’d need to test them across 100 elections, which you obviously can’t do because this is a real-world event and not an experiment.
In other words, while we’ll soon know which party controls the Senate, we will never know which model was correct.
And the models themselves don’t clearly distinguish between claims about epistemic uncertainty (things we do not know about the underlying state of the electorate) and a more metaphysical uncertainty (the ability of events-to-come to shape the future). Almost nobody actually cares about this, but even as a devotee of quantitative models who consumes them heavily, I find something unsatisfactory about this.
So, with nothing more meaningful or satisfying to say about the events that will unfold this week, I thought it could be interesting to take a tour of some of the most significant midterm elections of American history.
Midterms lack the personalistic drama of American presidential elections, but outside of foreign policy, Congress really is the main driving force of the American constitutional system, and midterm votes often bring big changes. The phenomenon of “midterm backlash” is something we’re accustomed to, but while it now operates more or less as a law of the political system, it’s not an actual law of physics — whether any given backlash emerges is somewhat contingent and actually drives much of the drama of American politics.
So, with that in mind, here are my picks (in chronological order) for the seven most consequential midterm elections.