Want your vote to really count? See how AI meets democracy in a Princeton lab.
The intersection of artificial intelligence, neuroscience, and political science has produced a revolutionary tool that could transform how you participate in democracy.
In this week’s WhoWhatWhy podcast, Princeton neuroscientist Sam Wang explains how Vote Maximizer, created by his team at the Electoral Innovation Lab, applies sophisticated AI to solve a fundamental challenge:
Where can your voice, vote, time, and dollars make the most difference?
In an era of billion-dollar campaigns and overwhelming political noise, Wang claims his Vote Maximizer can empower individual voters with unprecedented strategic leverage.
Once you have inputted your location and interests, this platform, using AI, analyzes everything from local school board races to state ballot initiatives, showing exactly where your political engagement can have maximum effect.
Wang, whose journey from mapping neural pathways to charting democratic engagement spans two decades at Princeton, argues he’s created more than just another political tool. Vote Maximizer’s goal, he explains, is to democratize the kind of strategic analysis typically reserved for campaign war rooms.
The platform can identify high-leverage races where a few votes could tip the balance, and even quantify the impact of proposed democratic reforms.
Wang, who has done groundbreaking work leading both the Princeton Election Consortium and the Princeton Gerrymandering Project, shows how this fusion of cutting-edge technology and civic engagement offers a powerful new path to real political influence.
Apple Podcasts
Google Podcasts
RSS
Full Text Transcript:
(As a service to our readers, we provide transcripts with our podcasts. We try to ensure that these transcripts do not include errors. However, due to a constraint of resources, we are not always able to proofread them as closely as we would like and hope that you will excuse any errors that slipped through.)
Jeff Schechtman: Welcome to the WhoWhatWhy podcast. I’m your host, Jeff Schechtman. With just 10 days until election day, we find ourselves drowning in a sea of political ads, urgent appeals, and dire predictions. The presidential race alone has consumed over a billion dollars and countless hours of our attention, but in this frenzy, it’s easy to forget the myriad other races and initiatives that profoundly impact our daily lives, from Congressional seats to school boards, from ballot measures to state legislatures.
As we try to care about democracy to engage meaningfully, we’re left wondering how can we possibly know where to focus our attention, our efforts, our contributions? Which of these races truly matter to us, to our communities, and to the issues that we care about? In this whirlwind of political noise, our guest today offers a beacon of clarity.
Professor Sam Wang, a neuroscientist with a passion for democracy, has developed a tool that cuts through the confusion and can empower voters like never before. It’s called Vote Maximizer, and it has the potential to revolutionize how we engage with our local political landscape. Here’s the core of it: You input where you live, where you vote, what issues and/or candidates you care about, and Vote Maximizer does the rest.
Using AI, it analyzes your local races and ballot initiatives using data from polls and past elections. Then it shows you precisely how your vote, your volunteer time, your donations can have the most significant impact in your community and beyond. This groundbreaking tool is the brainchild of the Election Innovation Lab at Princeton University.
Under Sam Wang’s leadership, this nonpartisan think-tank is not just studying democracy: It’s actively working to enhance it, providing individual voters like you with the same level of strategic insight usually reserved for political campaigns. From mapping neural pathways to charting the course of civic engagement, Sam Wang’s journey is as fascinating as it is crucial for our democracy.
With degrees in physics from Caltech and a PhD in neuroscience from Stanford, he’s been a Princeton faculty member since 2000. His work has spanned from the Princeton Election Consortium to the Princeton Gerrymandering Project, consistently at the forefront of applying data science and cutting-edge technology to strengthen the democratic process. It is my pleasure to welcome Professor Sam Wang here to the program. Sam, thanks for joining us here on the WhoWhatWhy podcast.
Sam Wang: Wow. With an introduction like that, what can live up to it? Thank you so much for having me on.
Jeff: [Chuckles] Well, thank you so much for doing this today; appreciate it. First of all, talk a little bit about the Electoral Innovation Lab at Princeton and how that evolved and what its mission is.
Sam: Sure. The Electoral Innovation Lab began here within the walls of the university. It’s a project to help build a science of democracy reform to help find ways in which democracy has gone off-track in terms of its rules and its mechanisms and then to maybe find science-based ways to make it work better. Since that time, the Electoral Innovation Lab has spun off from the university. It’s now a separate nonprofit.
It’s still here in town. If you know Princeton at all, it’s not a big town, and it’s about half a mile off-campus — maybe a few blocks off-campus. And we have a team — of scientists, political scientists, biological researchers, cognitive science — working together with students to try to find ways to help our democracy be a little bit less rickety — to find ways to make it more representative.
Jeff: And in doing that, there is this focus on technology and the use of technology to accomplish that. Talk a little bit about how that evolved, what those conversations were like early on in terms of ways to bring technology and science and data science, really, to enhance democracy.
Sam: Despite the proliferation of data being available to people in political life, and despite democracy being so well studied, it’s remarkable how little of the research and data on that has really percolated through to the general public. I mean, of course, we’re all very interested in politics, and certainly, listeners to your program are, and there’s a lot that we know about democracy being a bit rickety, whether it be gerrymandering or voting, where you only get to make one choice, and you don’t get to make a second choice.
What we thought we would do is just do things from the simple to the complicated. The simple is just taking polling data and past voting history to understand how close the race might be to show people where they might have more power. And then the complicated is long-term, building past just the next election but helping understand what rules might stand a little improvement.
Is it better to have partisan primary, top-two primary, top four primary. and to think of ways in which that can be made to work a little bit better? And that second thing requires computer simulation. It requires cognitive science. It requires getting a deep knowledge from political science — a lot of factors working together. And so it’s everything from just, as I said, simple calculations for this year to long-term calculations to help plan the future.
Jeff: And how did you get into this work from the background that you have?
Sam: Well, I have a pretty bad sense of self-preservation.
Jeff: [Laughs]
Sam: And so, when I first arrived at Princeton in 2000, I was up for tenure. And before tenure, I got really interested in the 2004 presidential election. And I started thinking of ways to apply simple statistical analysis to figure out where voters could have the most impact. And I remember in 2000 realizing that Florida was going to be important — that certainly was true.
In 2004, again, flagged a few key states, and that held up. And then, in 2012, realizing that gerrymandering was becoming a bigger and bigger distortion on representation of the level of Congress and in state houses. And in each case, it started as, I would say, curiosity and doing a little bit of math programming on my desktop computer and growing into something that turned into advocacy and shining a light on offenses as they arose and shining light, hopefully, to show people where they have the most power.
Jeff: And talk about the gerrymandering and the work that you’ve done and that the lab has done in that area.
Sam: Of course. In 2013, it became clear in the wake of the 2012 election that it was possible for one side, the Democrats, to get more votes for Congress and for the other side to… the Republicans to end up with a 33-seat advantage. And I, frankly, writing about it at the Princeton Election Consortium, didn’t believe that that was true.
But I had a lot of excellent readers who were quantitatively-inclined — financial traders, social scientists, mathematicians — and they said, “Oh, no, no, it’s bad this year.” And they were right. And I thought, “Wow, that’s amazing.” And so I did a little bit of computer simulation, wrote about it in The New York Times, and called attention to what I called at the time “the great gerrymander of 2012.”
And it became clear that single-party control of legislatures leads to an overwhelming temptation to self-deal. And when one party gets in power, it builds itself very favorable maps. And at that point, I realized that it was time to do something about it, at least in my own little mathy way. And so I formed a small team here at Princeton.
We started doing things like writing analyses, law articles, court briefs, looking for a way to move the needle a little bit. We failed at the Supreme Court, but then we started working at the state level. And so now, I think there was a wave of people who are working on gerrymandering, and there’s been progress in the last 10 years among people who finally realize what a problem it is. And we’ve been part of that wave of effort.
Jeff: And talk about the ways in which you’ve been part of it and how technology has played a significant role?
Sam: So one way is a number of people around the nation, experts, have gotten interested in how you can mathematically identify a gerrymandering. And so, for example, a simple way might be to… So one side gets more votes, and the other side gets more seats. But that can happen by chance. And so one can then do a statistical test to see whether it occurred by chance.
It’s possible to run computer simulations of maps to draw hundreds of thousands or even millions of maps to see whether a particular map is an outlier. And if it’s a severe outlier, it probably didn’t get there by chance. My team at Princeton has worked with people around the country. We’ve worked in partnership with reformers in Michigan, in Georgia, in Virginia, in Ohio bringing what we can to the table.
I was an expert in a court case in North Carolina that at least temporarily got better Congressional and legislative maps to North Carolina. This year on the ballot, there’s an initiative in Ohio that, if passed, will take the power of redistricting away from politicians and put it to the hands of citizens. And so whether it’s by court action in North Carolina or ballot initiative and the power of the vote in Ohio, it’s been a little bit of progress over the last five years.
Jeff: And all of that brings us to Vote Maximizer and what that is. I described it a little bit in the introduction. Talk a little bit about how it evolved and what the intent and mission was.
Sam: Vote Maximizer: If you go to votemaximizer.org, you can see what it is. It’s basically using the power of math and also some pretty deep dives into state politics to understand where voters have the most power. Here in New Jersey, I do not have a ton of power to influence the presidential race because it’s a safe blue state. You’re in California, and there’s no doubt about how California will go.
So power is distributed unevenly, whether it’s because of the electoral college or just because some races are closer than others. And so what we did at Vote Maximizer was we gathered up all that data. We asked the general question: How do individual voters have power? In Nevada, they have power over the presidency, they have power over democracy because they’ve… are going to be voting in ranked choice voting in a few weeks.
In Ohio, as I said, they can vote on redistricting reform. In Florida, voters can vote on a ballot initiative to guarantee reproductive rights. And so, everywhere you go in the country, you are not very far from a place where voters have leverage. And so what we did at votemaximizer.org is we cataloged all these things — presidential power, senate power, voter power — to influence the House of Representatives, voter power to influence state legislatures, and ballot measures, and democracy. It’s all in the ballot, and in our calculation in at least 40 states in the District of Columbia, 80% of people in the US have a direct role in shaping the future of the country.
Jeff: And in each of those places, in each of those communities, it’s really fundamentally different because people don’t have a choice where they vote. It really is what’s available to them in the place where they live.
Sam: That’s right. So you and I live in a particular zip code. We could type it in, and we can see how much or how little power we have. I can do either mine or yours. What would you like?
Jeff: Mine is 94558.
Sam: 94558. And we go there. And it is not fantastic for the presidency because we know how California’s going to go. But if we go to votemaximizer.org, I can see that you are on the northern coast of California. There’s Congressional districts in the Central Valley of California that are swing districts. There’s a number of swing districts in Nevada. Right there in California, you’ve got ballot initiatives to increase in the minimum wage — wow, look at that — to $18 an hour, and then have annual cost of living increases after that. Oh, that’s big.
And then, let’s see, what else do we have? Local bond measures: The threshold for passage is going to go from 60 — from a two-thirds requirement — to 55%. And finally, there’s establishing a right to marry in the state of California. So, gosh, I was wrong. I thought that California voters… My own database tells me that you guys have actually a fairly large number of reasons to vote in California, and then it also tells you where nearby, once you have voted, you can go get out the vote elsewhere.
Jeff: Right. Talk about that because there’s two parts of this equation. One is what you can individually do based on where you live, whether it’s measures that are on a state ballot, candidates that are on a state ballot, or whether it boils down to local communities, legislative districts, or school boards or whatever. But there is also the ability that people have, particularly in this day and age, to be able to work on campaigns in other places, in other states, in other communities. And this has a value in that regard as well.
Sam: That’s right. So, people say, “Oh, I vote.” But voting is, in some ways, the absolute minimum. It’s like brushing and flossing: We should do these things, but I’m not sure we should go around fishing for compliments for it. And so I think it’s good to vote, but voting is the starting point. Voting is an act of faith that your vote makes a difference.
But if you really want to live your faith, or if you live your beliefs, then you take it a step further. You say, “Okay, I’m going to phone bank. I’m going to send postcards, I’m going to go someplace.” I have one friend who relocated from the East Coast to Nevada for the month of October because he saw in the data that Nevada was a really great place to get out to vote because of the Senate race there because of three swing Congressional districts there because of ranked choice voting on the ballot and because of the legislature being possibly in the balance.
And so he went there; I was very impressed with that. And likewise, if you have money to spare, then look, money plays a pretty big role in politics, and you can project voice by giving money to a swing Senate race in Nebraska, of all places. So, Vote Maximizer shows you all those things. Some of them you read about in the headlines, if you read political news, and some of them you might not know about. So, we tried to put it all in a place that’s as convenient as possible to make it easy to find.
Jeff: Does it prevent or get in the way of potential upsets? If the money and the effort and all of the accoutrements of a campaign are directed at those campaigns that have a chance of winning that people see, there’s a real chance if they do something, it can win. Does it make other races that might have had a different outcome, if people had put undue attention to them? Does it make that more impossible?
Sam: Yes, that’s a good question. Whenever green-eyeshade types like us get into calculating voter power, the question does come up: Are you discouraging people from voting in a place like, oh, I don’t know, let’s say in Congressional races in Oklahoma, where in fact the races are fairly likely to turn out one way or the other. The thing about that is that voting is this activity where you generally fill out a whole ballot.
And so one way of looking at this is that if you are motivated to vote for any reason whatsoever, there you are with your ballot, and you’ve got the entire ballot in front of you, and now you’ve got other things to vote for. Like, look, here in New Jersey, in my Congressional district, there’s not a whole lot of suspense in how it’s going to turn out. But you know what? When we filled out the ballot, my wife and I looked at it and said, “You know, the school board election is here, and we have a lot of say in that,” and we made sure to vote on that. We informed ourselves.
And so, generally speaking, as I said a few minutes ago, by our calculations, at least 80% of people in the US have some reason that we can identify where their vote is likely to be pivotal or fairly influential in the outcome. And so 80% is a lot more than what a lot of people say, which is, “Oh, my vote doesn’t count.” And I would say that what we’re trying to do is give, as I said, voting is an act of faith. We are trying to provide a burning bush to say, “Actually, you should vote.”
Jeff: How does polling enter this equation? Because that’s part of the raw data that gets put into this.
Sam: Polling is a significant part of what we do at Vote Maximizer. We use publicly available polling data. Usually, polls get used to drive news stories. So journalists use polls. Obviously, campaign operatives use polls to figure out where to put resources. An unfortunate consequence is that polls usually get used for horse race, and people often become passive.
Our philosophy here is that polls should be used to drive active citizenship. And so, if a poll shows that a ballot question is, say, at a 20-point margin, that’s good information to know, as you plan your day, as you plan your voting strategy. But if something is within 10 points, that’s getting closer. And so what we try to do at Vote Maximizer is take voting history, take polling data, and look for races that are as close as possible.
If a race is close to tied, then that is the moment when your door-knocking makes the most difference. A friend of mine here was expressing concern about the presidential race in Pennsylvania, which is just over the Delaware River from here in Princeton. And he was on the fence about whether to go door-knocking. And I said, “Look, the polls are showing a near tie. That is the moment when your door-knocking makes the most difference. So why don’t you get out there and go do something?” And he said, “Well, that’s persuasive.” So I felt like that was a small win.
Jeff: In small races, talk about the way in which this amount of information, this kind of data while it’s being made available to individuals to motivate them in the ways that we’re talking about, it also can provide a huge resource to individual candidates, particularly in local races that don’t have the resources to do this kind of data collection.
Sam: Voting nerds know where the data can be found. And so one of my favorite resources is a thing called Dave’s Redistricting App. And if you go to davesredistricting.org, you can find district-by-district information about Congressional districts, state house, state senate districts, and it’s possible to look at that data and to see which districts are swing districts.
And so what we’ve done at Vote Maximizer is taken data of all types, Dave’s Redistricting is one source, voting history and voting records are another source. Public polls are another source and our own internal analytics, which began originally at the Princeton Gerrymandering Project. We have all this information available to us, and it’s possible to understand where people might make a difference.
I’m right now sitting here, looking at a map of Arizona, and I see that, in fact, there’s a couple of state house districts between Tucson and Phoenix where there’s some districts that may play a critical role in who controls the Arizona Legislature in 2025. And I’m seeing here that Pima County, which surrounds Tucson, voters are pretty powerful there. That might be a good place to go door-knocking, if you happen to be in Tucson.
Jeff: Have you any data at this point on the impact that this has, that the impact of people having and being empowered by this information? To what extent is it determining outcomes?
Sam: That is a great question. I think that campaigns do this all the time. Whenever “get out to vote” efforts are ramping up, any campaign organization worth its salt is going to be using this kind of information to figure out where to go door-knocking. Now, those people have pretty granular information, they have voter files; they know, honestly, disturbingly large amounts of information about many of us.
This is less granular than that, but it shows where to make a difference. So, for example, if I’m, again, thinking back about my area here near the Pennsylvania border, I can go here, and I can say, “If I go here from Princeton, what’s close to Princeton,” and then I see some legislative districts and Congressional districts, and I discover… I mean, not to be too granular about it, but I discover that all along the Delaware River, almost every Congressional district and almost every legislative district is fairly pivotal in Pennsylvania politics, and also in national politics.
And so I think, “Well, the leaves are turning, it’s nice out. Maybe it’s a good time to go for some door-knocking along the Delaware River.”
Jeff: To what extent is this information being used, do you think, by the media, and what impact does that have? It’s almost a kind of multiplier effect.
Sam: The media is not a monolithic entity. And so there are horse race people who just want to know who’s up and who’s down, and maybe they’ll just look at some polling site or odds-making site to look at what’s going on. I would say that one use case of the Vote Maximizer is to help local journalists, who are obviously strapped for money and time, figure out what might be near them; that’s really exciting to talk about.
So, for instance, if you’re a reporter in, oh, I don’t know, in say, Wisconsin, you might want to know which districts are close. You might want to go see for yourself who’s door-knocking; you might want to see what’s on the ballot. And so, what we’ve tried to do here is create a resource that’s not just good for citizens but also good for anyone who wants to tell stories anywhere in the US.
And so, if you go to votemaximizer.org, you can click on, I don’t know, Colorado, and find out in Colorado that there are ballot initiatives about abortion, same sex marriage. There’s a couple of swing Congressional districts, there’s ranked choice voting. There’s all kinds of stuff. And so, in Colorado, there’s a lot going on there.
Jeff: Do things like ranked choice voting and primaries and other situations impact how the information is dealt with on Vote Maximizer?
Sam: Well, let’s see. So what we’re really focused on at Vote Maximizer is what’s on the ballot. So I will say that in 10 states around the country, there are… I guess we count DC as a state; there are places where ranked choice voting is on the ballot. There are places where the rules of voting will be changed so that instead of having to be a Democrat or a Republican or other party member to vote in a primary, everybody can vote in one great big primary.
And what that should do is it changes the incentive structure. It can change the incentive structure to make it attractive to try to appeal for votes in a positive way to the supporters of other candidates. And so what we’ve cataloged here if you go to Vote Maximizer and click on “Democracy,” you can see all the places where democracy, in one way or another, is on the ballot.
It’s not just ranked choice voting, it’s not just redistricting reform. It can also be judicial elections. There’s all kinds of ways in which the future of democracy is on the ballot this November.
Jeff: Is there a danger that it leads to passivity in any way in some communities, where you might look at it and say, “Well, my vote’s really not going to make a difference. It’s very blue, or it’s very red, and there’s just no impact.”?
Sam: We already have that problem. Imagine… There’s research that’s shown that presidential swing states have higher turnout than nonswing states. And so if you vote in Pennsylvania, turnout’s higher than in, say, West Virginia, so we already have that problem. And what Vote Maximizer does is it addresses that problem by finding more states where you have a say.
So I’ve given you examples where people have a say in California, where they might otherwise think they don’t — in, say, Utah or Idaho, which I didn’t get into. Again, ballot initiatives and ballot questions where people can make an impact. So I would say it’s the opposite, I think that knowing more about the ballot makes you think about your ballot, makes you think how you might make a difference and hopefully, gets you to go mail in your ballot or go vote because as you’re recording this, voting doesn’t… It’s not that election day is on November 5th; voting ends on November 5th, and so the time to vote is now.
Jeff: What, if any, pushback have you gotten to this from anywhere?
Sam: I think surprisingly little pushback. I wrote about this in The Washington Post a few weeks ago. That drove a lot of interest; we’ve been getting thousands and thousands of people visiting Vote Maximizer. It’s a small fish compared with a lot of stuff that’s going on in this billion-dollar campaign. I would say that dollar-for-dollar, Vote Maximizer is this real bargain.
Maybe a little bit of pushback in the sense that voting is ideally this romanticized thing. And so what we’re trying to do is we’re trying to bring a little bit of hard analysis to understanding how people have power, and that’s a little nuanced. And so, what we’re trying to do is help people understand that the math can help them be active and not passive.
Jeff: And finally, where do you want to take this? Do you want to make it bigger? Does it get to be stronger with more compute? What’s the future of it?
Sam: Well, let’s see. So in the short term, at the Electoral Innovation Lab, we are very interested in understanding the impacts of those reforms that may be enacted, whether in Arizona, in Nevada, Idaho, Ohio, or other states. and so in the short term, we’re doing computer simulations and using cognitive science to try to understand the impacts of democracy reforms.
In the longer term, I’m hoping that we’ll have this integrated program, where as we build the science of democracy reform, we’ll understand the math of both making people more powerful in the short term and also building a better democracy in the long term. And that turns us back into the Electoral Innovation Lab in postelection mode, where we’ve been very focused on the campaign this year, but next year what we really would like is to understand how to draw better maps in Ohio or how to run better primaries in Arizona, or Colorado, or really anywhere.
Jeff: Professor Samuel Wang, votemaximizer.org. Sam, I thank you so much for spending time with us today here on the WhoWhatWhy podcast.
Sam: Thanks for having me on today.
Jeff: Thank you. And thank you for listening and joining us here on the WhoWhatWhy podcast. I hope you join us next week for another WhoWhatWhy podcast. I’m Jeff Schechtman. If you like this podcast, please feel free to share and help others find it by rating and reviewing it on iTunes. You can also support this podcast, and all the work we do by going to whowhatwhy.org/donate.
