Turnout is persuasion
You have to PERSUADE people to TURN OUT
You may have seen lastweek the release of Deciding To Win, a new report highlighting why Democrats lost in 2024 and what they need to do to win more elections going forward.
There’s a lot of interesting stuff in there, and I encourage you to read it, but the graph that caught my attention the most was this one, showing the strong correlation between policies supported by non-voters and policies supported by swing voters:
There’s a really important takeaway here, regarding the intra-Democratic Party debates about how to win elections. It’s often assumed that there is a significant trade-off between strategies aimed at boosting turnout and those aimed at boosting persuasion. In this telling, you could run to the left and juice your base’s turnout at the cost of swing voters, or you could aggressively appeal to moderates but sacrifice high turnout from your own party. What this graph suggests, and what I’m going to look at more in this blog post, is that this is not true. Swing voters and non-voters are fundamentally quite similar and support similar things. There are not many policies that test well with one group and not the other. In general, there is not a difficult trade-off to be managed, but rather an optimal path forward, one that involves adopting popular positions on the issues, and highlighting issues where voters trust you more.
(1) What do I mean by turnout is persuasion
The phrase turnout is persuasion could be read many different ways, so let me be clear about what I am saying. I believe that, for the most part, differences in how a candidate positions themselves on the issues and how they conduct their campaign tend to have a positively correlated effect on both turnout and persuading swing voters. There are relatively few actions that they could take that would be a significant boost to turnout and a significant detriment to persuasion, or vice versa.
What I am not saying is that the party that does better in turnout in absolute terms will always do well in persuasion. For much of the last several decades, Democrats had a systematic disadvantage in turnout, due to their coalition being younger/poorer/less white. This did not necessarily mean they were bad at persuading voters. Today, Democrats have a systematic advantage in turnout due to their coalition being more educated and politically engaged, and this does not necessarily mean they are good at persuading voters.
However, if you control for geography and time period, I believe there is a strong correlation between success in turnout and success in persuasion. I think this claim most easily holds up in partisan general elections although it can often be applied to primaries (more on this later).
(2) Why would this be the case?
In some sense, turnout is persuasion is an unsurprising concept. After all, you have to persuade people to turn out for you. Both situations involve trying to convince someone that it is worthwhile for them to cast their ballot for you, the main difference is just in what their next-best alternative is. In another sense, it’s deeply counter-intuitive — aren’t Democratic base voters extremely different to swing voters, and shouldn’t they care about different things?
The important thing to understand is that turnout is mostly not about activating your hardcore base, nor is it generally about turning out someone who’s never voted before in their lives. By the same token, persuasion is not about convincing either a lifelong Democrat or a lifelong Republican to vote for you. In all four of these cases, why bother? In reality, turnout is about activating the marginal voter who is deciding between voting Democrat or staying home, and persuasion is about convincing the marginal swing voter. When viewed this way, it’s easier to see how these people are quite alike. They’re both just people who are potential Democratic voters but something is in their way.
My sense is that often when people talk about what is good for juicing Democratic turnout, they are talking about what would make hardcore Democrat voters more excited, but that’s totally the wrong framing. Those people are already guaranteed to vote for you, and they only get one vote!
(3) Evidence from general elections
If we look back at the recent history and general trends of US elections, persuasion and turnout being linked emerges as a clear theme.
For one thing, it’s generally taken as a given that the party who controls the White House will suffer a penalty in the midterms, this penalty functions through both turnout and persuasion. The 2018 blue wave is a classic example, where Democratic victories were powered both by flipping Trump voters and by turning out their own voters at a higher rate.
As for Presidential elections, Obama achieved record Black turnout in 2012, while also winning a crazy high share of the Black vote — in some Black precincts he got literally 100% of the vote. This is not a coincidence.
We’ve also got some more extreme case studies, such as the 2017 Alabama Senate special election, where Democrats flipped a Senate seat in one of the most conservative states in the country through an impressive combination of persuasion and strong turnout from their base. Again, the two go hand-in-hand.
Where we can find examples of turnout and persuasion going in opposite directions, it’s usually due to systemic/demographic factors that are driving turnout. For example, you’d always expect Democrats to have weak turnout in a special election in California’s Central Valley, just due to the demographics of the area. But when comparing apples-to-apples between the same kinds of elections, the association between turnout and persuasion is very real.
(4) Evidence from primary elections
Ideally, a good explanatory theory should also have good predictive power. In that vein, let me talk a bit about how turnout is persuasion helped me and my colleagues to predict Zohran Mamdani’s primary victory in New York earlier this year, several days before the election (and despite my earlier skepticism, as noted on this blog).
At the end of the early voting period for the NYC Mayoral primary, it became clear that Mamdani had fundamentally reshaped the electorate in a way that made it significantly more favorable to him than any previous primary electorate. What to do with this information? Well, taken at face value, one might say Mamdani is still an underdog. Maybe I previously thought Mamdani would lose by 10%, this electorate is on track to be 8% more favorable to him than I expected, so now I think he’ll lose by 2%. But again, because persuasion and turnout are so linked, the fact that Mamdani was such a beast at turning out left-leaning voters should also imply that he was really good at winning over persuadable voters. When you take into account both effects, he looked much better placed to win, and of course he did ultimately triumph by double digits.
Examining the data behind his win, it’s clear that Mamdani thrived or struggled in the same kinds of areas with both persuasion and turnout. For general elections, we can use party registration to quantify Democratic vs Republican turnout, here it’s a little harder but we know that the biggest determinant of Mamdani support was age, so we can use the share of the electorate aged under 45 as a proxy for Mamdani turning out his supporters.
There is a clear correlation between where Mamdani won and where the youth share of the electorate surged, compared to the Presidential election last November (registered Democrats only, to ensure a fair comparison). The main datapoints where this relationship comes apart are the ultra-Orthodox Jewish areas, where even the very young did not support Mamdani.
But hang on, you might say, this graph doesn’t necessarily support my point. All we’ve done is shown that youth turnout was high in the same places that Mamdani won. This seems like insufficient evidence for the overall claim. For this critique, I have a second graph. It’s the same, except it shows where Mamdani outperformed Maya Wiley, who was the left’s standard bearer in the same election four years prior.
This correlation is somewhat weaker, likely due to the idiosyncrasies of the two candidates and their elections, but nevertheless this shows that Mamdani outperformed his persuasion baseline in the same places he outperformed his turnout baseline.
As an example, let’s look at the 49th Assembly District, a heavily Asian part of South Brooklyn which used to be more reliably Democratic but voted in a landslide for Trump last year. Left-wing Democrats have struggled here — Maya Wiley got just 9% of the first round vote in 2021. Consequently, it was probably the most impressive part of Mamdani’s win when he crushed Cuomo here, getting 50% of the first round vote. This surge would’ve been almost entirely due to persuasion, as there’s only so much you can do with turnout. And yet it was augured heavily by turnout, for the youthquake was strong in this area. Early voting indicated that the electorate in this district would be significantly younger than even a Presidential election.
To be clear I am not relying on hindsight, this intriguing turnout bump was pointed out to me at the time by my colleague iabvek, who suggested that it had to be driven by Mamdani and therefore he might be on track to significantly outperform in this area.
Zohran Mamdani had two goals in this primary election. Get more young people to vote, and persuade enough old people to vote for him. His success in these two goals was extremely linked. In Black areas we saw the same weak youth turnout that we are used to seeing in local elections, and Mamdani did very poorly in persuasion. Meanwhile in Bushwick, Mamdani activated an eye-popping number of young voters, including many who’d never even voted before. Bushwick was Mamdani’s strongest area, and he probably won even old people there in a landslide. This is not a coincidence.
After using this theory successfully in NYC earlier this year, and considering all the evidence, I have become a strong turnout is persuasion believer. They are not literally the same thing, and this theory is not all-encompassing. But it is a useful heuristic. Doing well in persuasion looks a lot like doing well in turnout. Whatever trade-off exists between the two is usually minimal, especially in general elections. If you are worried that Democrats running to the center will depress turnout, fear not. Swing voters and non-voters both tend to live in the middle.
Thanks to iabvek and KJ for their thoughts on this topic.