Thinking Like A Strategist: Tipping Points
There’s too goddamn many seats in too many chambers in this country. And while I’m probably as politically informed as anyone, I can’t keep track (except in a spreadsheet, which I have, obviously). Any given race has some complicated internal dynamics happening, its own special brand of weird, and there are thousands of these. So which ones do you have to care about? How do you prioritize? If you’re from a particular brand of national-focused political strategy, as I am, you do this by thinking about tipping points.
(I feel like I need a disclaimer that I love and cherish all our extremely miscellaneous state and local elections, and there being “too many” is a joke)
Individual Races
Every election in the country has some odds of going for one candidate or the other. In most races, this means that either a Democrat or Republican will win, or sometimes a candidate who will caucus with either the Democrats or Republicans. While the specifics of the winner may matter a lot in races that aren’t in some legislative chamber (city councils in highly Democratic areas, for example, have a ton of Dem on Dem fights with policy stakes), I’m primarily concerned with elections that feed into some legislative body where partisan control matters very much. This can include state legislatures, the House, Senate, and so on.
For those races that feed into a partisan-controlled legislative body, it’s all well and good to worry about the specific candidate in the partisan primary, but for the purposes of getting anything done, you mostly care about which party controls the chamber. This means you mostly care about the party ID of the candidate who wins a given district, or at least who they’ll caucus with and vote with. I find this kind of calming, as I laid out in my Heterodoxy Towards Moderation piece, because I really do not have to care about the weirdo views of the back bench Democrat from wherever, as long as they vote with us.
I imagine each race as a set of scales with D on one side and R on the other, with many outside actors trying to sprinkle sand on each side and tilt the scales. If you look at a whole chamber, you’ve got a ton of these scales tilted in different directions.
Chamber Level
You’ll have to excuse my graphic design here.
On a macro scale, all the component elections are inputs to a much larger set of scales that determine who controls the chamber.
Using the Senate as an example, if you control 51 seats, you have a majority. That 51st seat makes the difference between no control of the chamber, and control of the chamber. If you for example got 57 seats, that’s great, but the difference between 56 and 57 seats is very small compared to the difference between 50 and 51, or 49 and 50. (There’s some added complexity about supermajorities here, mostly relevant in state legislatures, but this logic all applies if you just shift the target from 50+1 to 60+1 or whatever)
The 57th seat would be like adding another block to a scale that looks like this- it’s nice to have, but it doesn’t substantially change the overall result.
The 50th and 51st seat are like adding a block to a scale that looks like this, which is a huge deal.
That 50+1 seat is the “tipping point”, the seat that puts you over the line into majority control. That seat is the MOST important seat in the entire chamber. Everything before it is necessary to get to 50, everything after it is gravy.
Estimates and Seat Ordering
So if we care the most about the tipping point seat.....which one is it?
We can broadly sort elections from very Democratic to very Republican using polling and past election results. Our “safe seats” for either party go at the ends of the distribution- an MA senator, or a WY senator. These could be designated as seat 1 and seat 100, respectively. In the middle, we have the entire spectrum of other senate seats, going from “pretty safe Democratic” to “ehhhhhh” to “pretty safely Republican” [these are scientific terms, of course].
Once we’ve set up that distribution, our 51st seat is quite literally whatever seat falls 51st in our ordering.
HOWEVER, this does not mean that that election itself must be close, or 50/50. The balance of the chamber can be tilted by even a very safe seat, if one party is dominating the election. In an election season where you end up with 70 Republicans in the senate, the tipping point 51st seat could well be 80% Republican. This is what you might call “a bad time”.
In advance of an election, we don’t know exactly which seat will end up being the tipping point. It’s easier in the Senate, where we have a very small number of races, but the House has much more seats that could potentially be our tipping point. Luckily, we can make smart guesses about which set of seats are very likely to be the tipping point, and what seat has the best odds of it.
Worked Example
For the purposes of argument, let’s take a very small hypothetical chamber, with 8 seats.
If we’re pretty sure that seats A, B, and C are likely to elect Democrats, seats D and E are a tossup, and seats F, G, and H are likely Republican, we can guess that one of D or E are going to be the tipping point.
Similarly if we think seat A is safe democratic, B and C are tossups, D and E are likely Republican but not 100% and F, G, H are safe republican, one of D or E is *still the tipping point seat*, we’re just very likely to lose.
What Does This Give You?
Thinking about tipping points can give you a meaningfully different prioritization of races than just looking at the races without this logic. It might make intuitive sense to spend money and attention in races that are close to 50/50. And in a vacuum, that’s right! However, when we talk about something like the senate, we’re trying to gain control of the chamber, not maximize the number of individual races we win.
We know that elections can swing wildly- stuff happens. The economy tanks or recovers, wars start or end, etc. Sometimes you get a -6 point national swing, oof ow, and have a rough election. When those big events happen, they tend to make the partisan leanings of the nation swing one way or the other in a way that’s pretty similar across states or races. We call this “uniform swing” (which is the basic way I explained the 2024 election results).
The idea of uniform swing is after a large event, our set of seats is very likely to shift like this:
And *not* likely to do this:
Focusing on the tipping point seat is really helpful in cases where you get a large unexpected uniform swing, especially if that swing is in your direction. If, for example, you thought you were completely hosed in our 8 seat election and barely going to scrape out 2 seats, the tipping point logic would still have you focusing on seats D and E despite how grim it looked. If you suddenly got a uniform swing that put chamber control up for grabs, you’d feel good about having put maximum effort (and money) into the seats that put you over the top, rather than conceding to protecting your safer seats. Essentially, you always want to act like control of the chamber is up for grabs. If it extremely wasn’t, then nothing you did much mattered no matter how you allocated your focus. Grim, but, give minimal campaign effects, this is how politics works.
Another “fun” effect of focusing on the tipping point seat is that if you unexpectedly have negative uniform swing, you can look like a real asshole who spent all their money in long shots and didn’t protect your at-risk seats. I explicitly care more about chamber control than about any specific seat, so I’m not much affected by this argument. There’s decent logic around making sure all your candidates run a bare minimum campaign, which I’m fine with given the amount of money sloshing around an average election, but you do eventually have to make prioritization decisions. I’m okay with looking like an asshole occasionally if it means we have the highest possible odds of controlling the chamber in question.
Senate 2026
Using this general logic, let’s take a look at the 2026 Senate elections. It’s not great! There’s 33 seats up, 13 currently in Democratic hands and 20 in Republican hands. That would generally mean it was a slightly harder map for Republicans (since they have more seats to defend). Unfortunately, Republicans have an edge on total seats held (given the whole Senate majority thing). This means Democrats need to defend all 13 seats and pick up 3 to have a chance at retaking the majority. On top of that, take a look at some of those Dem-held seats, they’re far from certain.
The probable tipping point seat here, under the laziest possible analysis of ordering everything by 2020 Senate margin, is in Maine. You can make an argument that Nebraska might perform differently, since Pete Ricketts is a recent appointee and Osborne did relatively well there. I look forward to having any polling or better sense of the race, which will tell us more about if that or Maine is going to be our main focus.
I would put our odds of taking Maine and thus the chamber at somewhere between “bad” and “not good at all”. Perhaps the most important conclusion of thinking through the tipping point seat is that once you’ve identified it, you have to meaningfully contest it. Even if that feels like a long shot, and especially if that means running a sort of campaign you’re not excited about, perhaps one that focuses more on moderation and state-specific appeals. I look forward to learning way more about Maine than I ever cared to know and experiencing Susan Collins’ personal oddities. I suppose this is slightly better than our tipping point being in Texas, the place where money goes to get lit on fire, so I’ll take that (right up until someone drops polling predicting that Texas is closer than Maine).
Happy holidays, and I hope you get a chance to log off for a bit.
