Deep Dive on Voter Communication with Todd Rogers

by Elizabeth Rowe (She/Her)

Voter Communication - Todd Rogers

7 Questions with Todd Rogers about Voter Communication

Todd Rogers is a professor of public policy at the Harvard Kennedy School of Government and is a behavioral scientist who has done research on increasing communication skills between families, educators, and students, employees, consumers, and voters. He is the faculty director of the Harvard Behavioral Insight Group, which focuses on using behavioral science for public good. Todd co-founded the Analyst Institute, which focuses on measuring and improving the impact of progressive campaigns on voter communication and everyday labs, which works with school districts to reduce absenteeism and increased learning by better communication with families. He is also the author of the new Writing for Busy Readers, which came out in September of this year, and is a practical guide for writing more effectively and practically to increase communication skills. This week we asked him 7 questions about his work and career. 

You can check out the Podcast interview here

How has our political climate affected voters' attention span and their willingness to listen?

I am going to take that question in a much more micro way than you might generally ask people and they answer. So, people's attention spans in general I think are shorter than they used to be, and I think there are a few reasons for that. The first is it's easier to produce things that get in front of people and demand their attention than it ever has been, and to deliver them to people through written text, TikTok, whatever. And then on the other hand, the switching costs are lower than they've ever been. So, we're getting an inflow of stuff demanding our attention. And the moment we have just the slightest interest in anything else, we can switch at the click of a button or less. Those two combine to make it so that when we present people with anything, any kind of communication, and in the case that I'm focused on, any kind of written communication, they are skimming and very quickly triggered to move on.

Has the introduction of short-form content had any impact on engagement?

I think it all contributes to it. We have text messages, slack, email, printed mail, TikTok, social media, Twitter, and then if we go to the news it's just an infinite supply of things demanding our attention right now with exclamation marks next to it and them.

You had been in the politics and the education space, and now you've written this book which is Writing For Busy Readers. Talk about the journey that led you to write this book.

I spent 10 years working on voter communication. Co-Founded the Analyst Institute and ran hundreds of randomized experiments looking at different forms of voter communication, and then returned to academia and shifted entirely to how to communicate with families working with school districts, and ran dozens of these large field experiments with school districts. And in the process discovered that it shouldn't have required a discovery at all. Everybody's busy and no one is reading what we're sending them and everyone is skimming. So, then during the pandemic started advising a bunch of state and federal leaders, could be mayors, state bureaucrats, federal bureaucrats, and administrators on how to communicate this really important information to constituents and busy stakeholders.

Across the three categories, it turns out there are a handful of common principles. Jessica and I started doing experiments on it and we're like, wow, there aren't that many and we've identified six principles for how you write. So busy people read and respond to the TLDR version, because that's the world we're in. If we make it easy for readers, they're more likely to read and respond. We should have a round of edits for everything we do where we ask ourselves at the end, how do I make it easier for a skimmer? And if we do, it's more effective, it's kinder to the reader and it's accessible to more people.

We are taught how to write well, and those continuously flowing paragraphs read beautifully if someone devotes themselves to reading it. But, the reality is that everyone is skimming and that doesn't mean they're skimming everything. The headings give them guideposts where if, oh, I wanna learn more about that, dive in and then pop out again and start jumping around. And we wanna make it easy for skimmers. It doesn't just have to be cutting content, it's also about how we make it easy for skimmers to navigate so they can figure out what, what they're interested in and when to move on.

Is that what led you to motivate people to take action based on what they are reading?

Yes, it has gone from increasing people's likelihood of voting, to donating, to volunteering, to supporting a candidate. So let's think about vaccines, sign up for vaccines or testing for Covid, of course, there is a deep political resistance to these things. There is also just a basic busyness that makes it so that anything that's difficult we're just less likely to do. It's just like voting. We found that when you back, I mean now it's prevalent, but in the early days of this kind of research on get out the vote, prompting people to make a voting plan made the ‘get out the vote’ script more effective, what time will you vote? How you get there, where will you be coming from? And by doing that, we made it more likely people are gonna follow through. We didn't persuade anyone to vote. These were people who already wanted to vote. Their days are just busy and things get in the way. That's not the whole population, but for a big chunk of them, it's about how we make it as easy as possible for them, 'cause then they're more likely to follow through on it.

Talk about some more of these ways to make it easier for voters. Were there tests you did or things that are featured in the book, especially ones that we might not think are intuitive?

Okay, there's two that I particularly like, and what's intuitive and what's not? Everything Is obvious after the fact. So Sure, yes. We've now done lots and lots of these randomized experiments where we send messages with the writer's final version, or we cut it in half in some way. Either I arbitrarily delete paragraphs or I ask the writer to cut it in half. Consistently people, the reader, like when we hire readers to read both think that the longer ones are gonna be more effective. And we have never found that to be the case from three sentence text messages, turning them into two sentences, to writing, to elected school board members asking to fill out a survey, to my favorite, which will be relevant to this audience, working with one of the Democratic national committees.

We're working with them on improving a fundraising email that they had shared with us. I was like, well, I don't wanna edit it because they gave us a five minute turnaround, unsurprisingly, right? So, what we proposed was, why don't you arbitrarily delete the second, fourth and sixth paragraph? just delete every other paragraph so it doesn't make sense. Then we, again, hire people to read both, and they think that the full message is more coherent than the long message. We made the short message incoherent, and then we ran an AB test and it increased donations by 16% just by cutting words. I think the underlying mechanism is one, people are more likely to quit when there's a lot of words and they're in the middle of it, but before that, they're more likely to just be deterred from engaging at all when they see it's a wall of words like, I'll deal with this later, or, it's just too much.

Now, when you're doing things though for search, digitally is funny. Multiple studies show that on the internet some longer posts always get more engagement than shorter posts. So some of this is not because of the reader, it is actually because of bots, right? And also because of the way that a search engine works versus how a human works?

Yeah. I don't think that the Google bot gets depleted and decides to move on because it's tired, right? No. And so I think it's a really good point that I have not thought enough about, which is that when we're writing for humans, it's different than when we're writing to optimize SEO, right?

It is so interesting because when we started the campaign workshop blog, it was our view that we could just do this thing in 300 words, not a problem. Then to be honest, it just wasn't showing up in searches until we started writing at 800, 1000, 1500 words because it was more engaging for the search.

I'm gonna make a seamless transition to the second principle that I wanted to share which is, okay, so you've decided you're gonna write your thousand words, right? That is honestly mean to the reader when it could have been 200 words, right? Which is like the thing that the second order effect of what we're talking about is our goal is to help writers achieve their goals. That said, in the process, it's also just more respectful of our readers and to write in ways that make it easy for them. It saves them time. It's just kinder, that said, so you're gonna do your a thousand words to all the people who click on the link that they only see because it's a thousand words, but you can make it navigable, right? You wanna make it navigable and so you can still have your thousand words. We ran an experiment where there was an eight paragraph message and we had headlines every two paragraphs or, you know, subject headings every two paragraphs or not. We more than doubled the likelihood that anyone clicked on anything after the second paragraph.

So, talk a little bit about your principles. One of the ones that really jumped out to us was the fifth principle among the six central principles, which is telling readers why they should care and why that is so important, right?

Sure. The best example of this that we use in the book and that we've ever come across is a Rock the Vote experiment. Rock the Vote is interested in first getting volunteers to then get young people to register to vote, and they often do them in cool places like concert venues and festivals and things like that. So they ran an experiment with Lauren Keen, who is now one of the leaders at the Analyst Institute. She ran this experiment with them where the subject line that they had written was volunteer to register voters, and they're recruiting, in this case, volunteers to then show up at these registration events. And there was another condition that Lauren suggested, which is their interest.

The writer's goal is we want people to volunteer to register voters, but why would they volunteer and care about this? Those of us who care a lot about youth civic engagement, it's because they really wanna save the republic and increase youth voter turnout. Others just wanna go to a Beyonce concert. So, the next subject line was ‘attend free concerts registering voters’, and in that condition, she gets like four to eight times as many volunteers unsurprisingly, right? Because instead they write the message, and I honestly think that's the first draft. And then you go through it and you're like, okay, well now I don't wanna change my content, my goals, which of these things can I raise up for the reader that they might care about? I mean, there's other steps I would go through too, but in this case it's, well, let's just raise up the part that the reader might care about and we're not perverting or undermining our own goals. We're serving them, but we're just trying to take the perspective as best we can of the reader on the other side.

There are people of different backgrounds who are all reading the same thing. What are ways that we can make these things more readable and engaging for everyone if it's not just a very specific audience?

If you have multiple audiences and you have to communicate through some fixed text, then what you wanna do is make sure that it is navigable for the different readers. A great example for us, I'm in a university and our deans will write these 18 paragraph long responses to every world event. And I am not joking, I do not know anyone who has ever read the entirety of any one of them. And I've worked with the dean's office, I've worked with the president's office on this, and they say, well, we have so many different constituencies and each paragraph is pointing to someone else and we can't really cut it. We could have a different discussion about that, but if that's the fact, what they can do is they can add structure below. 

They have their headings exactly as you've been talking about. And if you're interested in these different things, you can jump around. I think that's the best you can do if you can't cut it down, is make it easy to navigate. But I wanna be really careful that our orientation in this is inspired by user-centered design. 

If a person interacts with an object you have created and gives up before understanding how it works, it's always the designer's fault. It's never the user's fault. I think the same is true of writing where if a person looks at what you have written gives up within one second and has no idea what it's about and you move on, often we're like, well, it was in there, it's on you. I actually think that we should shift it. Here's the fact: everyone is busy and everyone is skimming. So the burden is on us to make sure it is super clear from the very get go what this is about. If they move on, they should have at least pulled out the key information. And if they didn't, it's on us, not on them.

Often people think they know their audience, so they won't try different things to that audience. I think that idea of testing and looking to say, how can we make it better? How can we increase engagement over time? It usually is to really increase that engagement over time.

People have expectations about what you write. They come to have expectations about what other people write. Other people converge on one style. And so if you wanna stand out, it has to look different. There isn't an equilibrium that's stable except here's the fact they're busy and their goal is to move on as quickly as possible sometimes without understanding what you're saying.

Is there someone that you feel particularly does this really well, who is consistent at doing this in communicating in this style in a way that we can learn from?

I don't know that many of you would have encountered this person, but there is an alum of the school that I teach at the Harvard Kennedy School government who is joint at the Harvard Medical School. He's an emergency room physician named Alistair Martin. He was a White House fellow a couple of years ago, started an organization called Vote ER that registers voters in ERs and has a bunch of different arms now. It's called a Healthier Democracy. And I don't know how he got there, but every communication I have ever seen from him I have forwarded on to my co-author Jessica Lasky Fink and been like, wow. His name is Dr. Alistair Martin. He's the head of a Healthier Democracy. It's like you stumble upon people in your life that are actually very good at it and are very good at engaging in ways that are public facing or odd. It's an incredible skill that very few people have.

Honestly, a lot of it comes from being very good at perspective taking. It's really about how good you are at imagining what the reader's experience is. I think a good heuristic is assuming they have no interest in what you're saying, or at least that they care a lot less than you do about what you're writing. I have to use this checklist that we made for how to write so busy people read because I would've assumed that what I'm writing are really important words. People would love to read them. I don't think that's the case now.

Do you have any tips for our listeners around program evaluation or testing that you still use to this day? things that you think are like good tips for folks? 

We should approach assessing whether something works with humility, even when we think we have very good results. I know that this is not what we tell our funders, but it's really hard to know what's gonna work and what's not. The tip is, I try to find outcome measures that are already comprehensively and costlessly collected. So in education I worked on student absenteeism because it is administratively collected, it's super easy to get that data and then all I have to do is randomize my treatment and the outcome measure is costlessly and comprehensively collected. If it were attitudes which many of your listeners and our friends care about, it turns out that's really expensive and hard to collect.

Whereas, if it was attitudes towards school or something like that in schools or in politics, where turnout is super duper easy to get the outcome measure on. Not timely though. That's one of the challenges with get out the vote research was always that you, you find out after the fact. And so you have to have a multi-cycle horizon. And even then effect sizes may vary depending on what's going on in the election. But fundraising is great because you can, you get this instant feedback, the dependent variable. We are very invested in collecting that outcome measure, which is why someone gave money or not.

Do you have a favorite GOTV tactic that we should be thinking about for 2024?

There was a study in Virginia during the 2005 gubernatorial race, and I think I saw it replicated somewhere recently, but I remember they randomly assigned people to six weeks of getting the Washington Post in print or six weeks of the Washington Times in print or nothing. And then after the election they did surveys and they looked at turnout records. They found that both increased turnout for the Washington Post, which is a more liberal newspaper, increased support for the Democratic candidate relative to the Republican. And they were more informed with the Washington Times, which is a more conservative paper, increased support for the Democratic candidate relative to the Democratic candidate. And they were more informed. It turns out that being more informed in that situation made them more progressive. And of course that was a particular moment in time during the Iraq war. I love the idea of investing in this intermediate term time horizon of informing and not just thinking of persuasion, but it's thinking of informing. It turns out that facts are often on our side.

Do you have a favorite movie, podcast, TV show or book that our listeners should be reading or listening to?

You mean other than Writing For Busy Readers by Todd Rogers and Jessica Lasky Fink? I just finished Recode America by Jennifer Polka and I loved it in so many ways. She wrote Recode America and tells the story of the centrality of technology and basically the last mile in good policy, which is that we can design great policy and that implementation matters a ton. She tells amazing examples of heroic people in administrative and bureaucratic roles and how we can do better in the implementation in a pretty nonpartisan way. Like to the extent that we want the government to do stuff, which I don't think is a partisan position or it shouldn't be, then an incredible book about it. There's a part of it that really resonated with, she says, complexity favors the sophisticated or complexity favors the privileged.

And it's exactly one of the principles we talk about. If it's important for us, we wanna make it easy for the readers to do, we wanna reduce the number of steps, reduce the friction. And one of the things she talks about is just the more complicated we make things, the more we advantage the people who have the resources to navigate and handle these things. And so if the federal government is allocating money for cities and it's a super complicated process, it's hard for Akron to apply and it's easy for Chicago.

Todd, how can folks get ahold of you if they want to?
You go to our website,, where we actually have a tuned large language model that is trained on these principles and a bunch of examples of pre and post where we'll edit your emails and suggest how you could have rewritten an email. So it's easy for skimmers. It's fun. I am a big fan of the pod, and thanks for having me.

Let me end by reiterating what I think the big theme is when we write anything, whether it's a text message about Thanksgiving dinner, a fundraising letter or a webpage, we should go through one round of edits where we ask ourselves, how do I make it easier for the reader? Because when we make it easier for the reader, it's more effective, it's kinder and it's accessible to more people.

Thank you Todd! 

Have a question on voter communication? Drop us a note!