5 Keys to Campaign Strategy #2: Elasticity
Welcome back to my series on the five essentials of strategy for political campaigns and the math, stats, and economics concepts behind them. My first piece was on categorizing voters and applying different vote-getting strategies to each using the persuasion-GOTV matrix.
This time, we’re going to zoom out to answer a more fundamental question: how hard will it be for my favored candidate to win? We’ll introduce a new concept, elasticity, to explain why looking at raw vote totals isn’t enough, and why so many candidates easily reach the mid-40s but can never make it to 51%. We’ll also learn why there are two kinds of swing states, each demanding a very different strategic approach.
There’s more to election results than meets the eye
Suppose you’re a Democrat running for Congress and found that, in the last cycle, the Democrat lost 55% to 45%. (That’s not too far off from the Florida campaign I’m working on.) We can illustrate this by assuming there were just 100 voters in your district and assigning 45 to the Democrat and 55 to the Republican:
To win this time around, all you have to do is keep your 45 and grab 6 away from your opponent. Seems straightforward enough, right?
Wrong. Just looking at the 55–45 margin obscures a very important detail: how many of those voters are actually swing voters?
Perhaps your district actually has an equal number of dedicated Democrats and Republicans and a lot of swing voters, and the Democrat just so happened to lose them last time:
Winning this rematch won’t be easy, but it’s certainly possible, since swing voters are (by definition) willing to switch sides. You just need to get 6 of the 10 swing voters to change their mind, and you’ll get to 51.
In this situation, your district is elastic: due to the large number of swing voters, the outcome can swing wildly between elections based on the relative strengths of the candidates or shifts in the national mood. (Elasticity is a concept from economics that explains how dramatically the output changes when the input changes slightly. If you’ve ever used a toaster that does nothing when you set it to level 3 but turns your bread to a smoldering pile of charcoal when you set it to level 4, you’ve experienced an unfortunately elastic toaster.)
But your 55–45 disadvantage may not be that easy to overcome. Suppose, instead, that your district actually had a lot of hardcore Republicans and only a few swing voters:
This yielded the same 55–45 outcome as before, but this time it’s a lot harder for you to win. There are only a handful of easily-persuadable voters, and even if you got them all, you’d still fall short:
In order to win, you have to start digging into your opponent’s base and convincing some diehard Republicans to switch sides. It’s not impossible, but it’s a lot, lot harder than getting swing voters to tilt toward you.
If this describes your district, you’re in an inelastic district. Because there are almost no swing voters, the outcome won’t actually change all that much between elections. Overcoming a 55–45 disadvantage is far harder than it would be in a highly elastic district.
In short, it’s not enough to look at the last election’s results. Depending on the elasticity of the district — the number of swing voters and the sizes of each party’s base — the difficulty can vary wildly. You could have won by 5 last time and still be in a safe seat; you could have lost by 15 last time and still have a shot.
The two flavors of swing state
We hear a lot of talk about how important swing states are. But just saying “swing states” omits a very important factor: not all swing states are created equal. Elasticity theory shows us that there are actually two different flavors of swing states; each one behaves very differently and demands a very different tactical approach.
Some swing states are highly-elastic states that swing freely between the parties due to their high numbers of swing voters. States with independent streaks, like quirky Vermont and “Live Free or Die” New Hampshire, tend to be elastic, as do states with a lot of white suburban voters (the main swing group in America). As a rule of thumb, the New England states and other mostly-white states tend to be the most elastic.
Not all of these elastic states are swing states. Rhode Island is the second-most-elastic state in the Union, by FiveThirtyEight’s math, but the Democratic base is so much bigger than the Republican base that it’s never competitive. Meanwhile, North Dakota is the eight-most-elastic state but still ruby-red thanks to its gigantic Republican base.
Other elastic states have more balanced populations and are thus perennial swing states: consider New Hampshire (the most elastic state), Maine (#4), Iowa (#7), and Colorado (#12).
Other elastic states fall somewhere in the middle: they’re not usually competitive but can become so in “wave” elections. Alaska and Montana, for instance, feature surprisingly competitive Senate and Presidential races this year thanks to the Democrats’ 10-point national edge.
The other type of swing state is the inelastic state that just happens to be evenly divided between the two parties. There aren’t many swing voters, but the Democratic and Republican base are roughly the same size, thus making for consistently tight contests.
Deep Southern states are the classic example of inelastic states: they’re sharply divided between rural white voters (a reliably Republican block) and Black voters (staunch Democrats), with few people in between. Mississippi is the most inelastic state in the Union; other inelastic states include Alabama (#2), Georgia (#3), Maryland (#4), South Carolina (#5), and North Carolina (#10).
As you’ll see from that list, most of these inelastic states lean heavily toward one side or the other. North Carolina and Georgia are evenly balanced, though, resulting in elections that always sit on a knife’s edge.
One interesting corollary from this analysis is that Democratic candidates can get tantalizingly close in some of these inelastic states but have a really hard time getting to 51%. (Think back to the 51–45–4 example.)
For instance, Democrats regularly pick up 40% of the vote in deep-red places like Mississippi and Louisiana — that’s their base. The rest of the way to 51% is an uphill climb: even with a huge push, Democratic candidate Mike Espy only won 46% of the vote in the 2018 Mississippi Senate election. That, in fact, was Democrats’ high-water mark in Mississippi: the party hasn’t gotten more than 46% in a Senate election since 1982.
Don’t be fooled by candidates polling in the mid-to-low 40s in inelastic states. Getting to 51% will be very, very hard for them. Their best hope is to pull off something like a 47–46 win.
Thus, we have two types of swing states:
- The elastic ones, which we’ll call “New Hampshire-class”
- The inelastic but evenly-balanced ones, which we’ll call “North Carolina-class”
Each type demands a very different campaigning strategy. NH-class states are won and lost by persuading independent voters, which is why television ads are so common in places like Iowa and Colorado.
Meanwhile, persuasion won’t take you very far in NC-class states; the election becomes a turnout battle. Victory depends on getting out the vote (GOTV’ing) among your base before your opponent can turn out their base — which is why helping your supporters get to the polls and fight through 11-hour lines is so essential in close Southern states.
This might remind you of the persuasion-GOTV matrix we discussed in the first piece in this series. Indeed, these concepts are related: NH-class swing states have more persuadable voters in the middle of the matrix, while NC-class swing states have more voters on the edges and thus more need for GOTV.
With this piece, we’ve seen why the label “swing state” is incomplete. You need to figure out whether you’re in an elastic state like New Hampshire or an inelastic state like North Carolina and apply different tools in your toolbox: persuading swing voters or GOTV’ing your base.
In part three, we’ll look at another takeaway from this analysis: why the way we look at polls is wrong, and why we overlook one of the most important numbers in electoral analysis.