My fundamental premise in thinking about democratic theory is that democracies need to make some meaningful connection between the will of the people and what actually gets done. And right now, the connection is severely compromised. Public opinion is increasingly shaped by propaganda and distortion, and the public sphere has decomposed into filter bubbles and ideological enclaves where errors never get corrected.

This fragmentation makes it very difficult to assess public opinion, the will of the people — or what they would think under better conditions. So, I propose some commonsense, good conditions for assessing what the people would really think: situations where people have a chance to reflect on important issues that affect their lives but which they often ignore due to social context or collective action problems.

In my view, deliberative democracy helps align the formation of public will — the will of the people — with what’s actually done. And in many cases around the world, we’ve seen it succeed. That’s the basic deliverable: it offers a better way to assess what people would think if given the opportunity to deliberate.

I am interested in people’s considered judgements — what they believe after they have weighed the pros and cons in a thoughtful way, in an evidence-based environment, with the best available information that can be practically provided to them. Most of the time most people don’t do that. In fact, political economist Anthony Downs years ago argued that voters were “rationally ignorant” because your one vote in millions is unlikely to matter, why should you invest time in understanding the complexities of policies? Your individual vote or opinion won’t have much effect.

That means a lot of the results you get from public opinion polls are just top-of-the-head impressions of sound bites and headlines. So, my central question becomes: What would people think if they really thought about it? When they do — when they have access to information, can ask questions, and engage in dialogue — we often see polarization decrease.

Well, we’re not trying to change their minds. We’re trying to empower a decision-making process through which people can make up their own minds. There are no predetermined outcomes in this process. We’re just trying to engage them under what we think are good conditions — where they can actually spend some time to think about an issue that affects their community or their state or their country and ultimately themselves.

Everything is designed to be balanced. We do our best to make sure that all the information is vetted by a balanced advisory committee that determines that the information is accurate. And if it’s not well established, then we note that it’s contested and present competing accounts. But the people are not just moved by exposure to information — they’re moved by discussion.

We’ve run controlled experiments and found that it’s the discussion process that gets people to revise their views and brings issues to life. Exposure to information alone has little effect. We’ve tested this with a control group that received only information, and opinions barely shifted.

The problem with public opinion as represented in ordinary polls is that it usually reflects off-the-cuff reactions. It’s shaped by a combination of rational ignorance and the extreme views of people in ideological enclaves who have very decided opinions. And it’s hard to break through that because our public sphere has decomposed into enclaves and people’s freedom of choice in terms of their media has expanded so much that it’s easy to use that freedom to engage with people and media sources whose viewpoints you already tend to agree with.

This is ironic for someone like me whose presumptions are basically those of John Stuart Mill. He believed that error would be self-correcting through exposure to opposing arguments. But that doesn’t happen if people never hear the other side. And if people are only getting one side of the argument because they’re choosing only media sources and interlocutors they find congenial, they may have no idea what people on the other side are thinking or why.

But in a properly structured dialogue, they find out. And it has very dramatic implications on how they end up thinking. We’ve seen this again and again — both in the United States and abroad.

In my book, I review the evidence from a series of national “America in One Room” projects we conducted with a great survey partner, NORC [National Opinion Research Center] at the University of Chicago, which included control groups. We found clear evidence of considered judgments, depolarization, and lasting effects. We published these findings in the American Political Science Review.

We also did a climate change project using an online deliberation platform developed here at Stanford. It works in any language, and it doesn’t require a human moderator — which opens the door to large-scale applications. We’ve already deployed it in about sixty countries, and we’re currently in serious discussions about expanding it further.

In the last chapter of my book, I speculate about how we could create a more deliberative society. These dialogues create lasting effects: they create greater mutual trust, dramatically reduce depolarization, and generate more informed judgments about what should be done. They also create the kind of thoughtful voter that many of my colleagues in political science dismiss as rare as a unicorn. And that’s exactly the kind of voters we need to solve our public problems.

In a healthy democracy, politics shouldn’t just be a contest of strength between parties — it should be a battle of ideas. People need to hear the best arguments on either side and decide for themselves what they really prefer, for the good of their communities.

So, how would we scale it? In the last part of the book, I outline a range of venues and opportunities — many of which we have already piloted. The book begins with the Athenians, which is a personal avocation of mine — to study both the Athenians and the American founding and the values at stake in both cases. In Athens, Aristotle talked about the rotation method among citizens — that each person, each citizen, would rule and be ruled in turn.

Now that’s been dismissed as an idea suitable only for a small-scale society. The Athenians used a machine for choosing citizen panels of five hundred or more who made important public decisions. In a small society, everybody had a chance to serve. But could something like that work in a large society?

Once I began cataloging all the possible venues for deliberation — many of which we’ve already tried, I think yes, it could be applied on a large scale. Most obviously it could be applied in schools and in universities as a form of civic education. That would reach very large numbers of people.

We’ve already been experimenting with that, as I document in my book and on our website, the Deliberative Democracy Lab at Stanford. Eventually, I’d like to build up to doing something like what Bruce Ackerman and I argued in a book a long time ago: Deliberation Day. The idea is that, before national or state elections, everybody has an opportunity to deliberate. And the platform provides a good opportunity for that because it lowers the cost.

When Ackerman and I first proposed the idea, it was treated as utopian because it was very expensive — but doing it online is not expensive. And we have now shown demonstrably that the results of our online platform are very similar to the results of face-to-face deliberation. We saw this in a project we ran with our colleagues in Finland. One-third of the participants deliberated face-to-face in Helsinki (I was there observing), one-third participated via Zoom — small groups with human moderators — and one-third used our online platform. The results were basically identical. We view that as proof of concept.

We’ve also identified all kinds of other contexts where deliberation could be embedded — like ballot proposition evaluation. In a large state like California, getting a proposition on the ballot requires millions of dollars, meaning the agenda can often be shaped by moneyed interests. That distorts the democratic intent behind these direct-democracy mechanisms. Deliberative democracy methods can be used both to help create ballot propositions and to evaluate them for the voters.

As I explain in the last part of the book, I’m very much an advocate of a complementary role for deliberation. And I outline a number of scenarios for how that can work. I have colleagues who argue that deliberative democracy should replace elections entirely. I think that’s a mistake.

Instead, I think we need to make our elections more thoughtful — to incentivize citizens to seriously engage with the issues. That’s entirely possible. But we don’t need to give up the value of mass political participation, which is most efficiently realized by voting. So, I’m emphatically against replacing electoral democracy and emphatically in favor of supplementing the process with deliberation.

In the book, I write about a remarkable case in Mongolia, where deliberative polling has been institutionalized by law. In order to change the constitution, they must first convene, according to the law, a national deliberative poll based on public suggestions. An independently elected advisory council supervises the process, with multiple safeguards to make sure it’s a rigorous process.

Recently, for the second time, they carried out this process: a national sample of more than seven hundred people, plus a control group who deliberated in the parliament building in Ulaanbaatar. And they came to certain recommendations about how to change the electoral system. The public wanted more than just the two parties that were butting heads against each other; they wanted more third-party representation in parliament. So, they supported adding proportional representation for a significant portion of seats. That proposal passed the deliberative poll and, as required by law, also passed with a two-thirds vote in parliament. It became a new constitutional amendment. And in the next election, sure enough, more third-party members were elected.

Let me begin by pointing to several cases in other countries where political leaders have actually embraced deliberative democracy. It’s interesting to puzzle out why and how.

For example, President Moon [Jae-in] in South Korea had a big problem: some nuclear reactors that were already half-built when he took office, but his party program was antinuclear. So, what was he going to do if he didn’t complete the reactors? They’d have to import fossil fuels — which he was also against. But if they did complete them, he’d be committing to nuclear power. So he convened a national deliberative poll.

We had already conducted a number of successful deliberative polls in South Korea, with local collaborators, so there was precedent. To my amazement, President Moon announced that he would follow whatever the public concluded through the deliberative process. The people recommended finishing construction — and the reactors were completed.

That was an example of a hard problem. And President Moon very wisely decided to share responsibility for solving the hard problem with the people. Japan did something similar after the Fukushima disaster when reevaluating their nuclear policy. A national deliberation was held to determine whether they were going to try to phase out nuclear power or maintain it. The government followed the public’s recommendations from that process as well.

In another instance, Bulgaria faced a hard choice about what to do with the Roma — sometimes referred to as gypsies — who are the most discriminated-against group in Eastern Europe. We helped organize a national deliberative project in Bulgaria on Roma-related issues, including education, housing, and the criminal justice system.

And one of the dramatic results was that the participants came to support desegregating the Roma-only schools and busing the Roma children to schools with everybody else. And that’s what the government ended up doing — based in part on the results of the deliberation. You can find out more about it on our website. The Roma schools have been desegregated.

That was a hard choice and nobody quite knew, given the antipathy toward the Roma, what the results would be. It was sort of a gamble. But once people actually discussed it, once they understood what it meant for Roma children to attend segregated schools where they didn’t even learn the dominant language in the country — and what that meant for their prospects — their views changed. That’s the power of deliberation.