Is American Democracy Broken?
What Research Shows about Polarization, Representation, and Elections
Read or watch the news about the 2026 US midterm elections and it can seem as though the country is coming apart at the seams. Republicans and Democrats see each other as enemies rather than opponents. Partisanship has become a core part of Americans' identity, determining where we live and whom we marry, as well as who we vote for. At the same time, Americans’ trust in the federal government to address both domestic and international problems is at an all-time low. And even as politicians cast doubt on the integrity of our electoral system, they engage in unprecedented partisan gerrymandering to claim or remain in power.
The Harvard political scientist Steve Ansolabehere, PhD '89, says his research tells a more hopeful, if complicated story. Below the surface of partisan division, Americans share many of the same concerns, struggles, and hopes for the future. Our political system faces challenges, but most often still represents the will of the people. And, despite attempts to tip the scales, our decentralized elections are still overwhelmingly free and fair—at least for now. He spoke with Harvard Griffin GSAS's Office of Communications about his recent work on these topics, and why the US may actually be in a “golden age” of elections.
You and your colleagues have surveyed more than 700,000 respondents over two decades through the Cooperative Election Study. But in recent years, you started to ask a new question. What was it?
We started asking people what important things had happened in their lives in the past year. We wanted to understand what it was about people’s lives that could help us understand the context within which their political ideas and attitudes evolve. One of the things that struck us right off the top was how similar people’s lives are across the political divide. It doesn’t matter if you’re a Democrat or a Republican. People’s top concerns are the same: their health, their families, the cost of living and making ends meet, and their jobs. If I gave you the text of somebody’s statement, you could not tell whether the person was a Democrat or a Republican.
So, what did you take from the responses to the new question, in the context of all the other data from the study?
It was an important reminder that our society really isn’t divided in some of the most fundamental sociological ways—by how people are living their lives. There’s a lot written in books like The Big Sort about how party polarization is infecting everything: where we live, how we eat, how we live, and so forth. We just don’t see that. That’s not really what emerges as where or how we’re divided. That’s not to say that there aren’t differences in political opinions and values between ardent Democrats and Republicans, but we’re not living in a world where everything is labeled D or R when you go to the grocery store, how you live your life, or what your family is. That was the first solid reminder.
The other fact that popped out of the data is that we’ve asked hundreds and hundreds of questions about hundreds and hundreds of issues over the last 20 years. Across all those questions—everything from taxes to war to abortion to immigration—about 40 percent of the time, a majority of Democrats and a majority of Republicans are on the same side. If you take the other 60 percent, the divisions on most issues are pretty modest, around 60–40. We’re not in an 80–20 world on most issues. Maybe a quarter of issues are 75–25 or more between Democrats and Republicans. There’s a big mass of people in the middle who don’t consistently line up as Democrats or Republicans, and they tend to be much more moderate in their issue preferences.
[Survey data shows that] people’s top concerns are the same: their health, their families, the cost of living and making ends meet, and their jobs. If I gave you the text of somebody’s statement, you could not tell whether the person was a Democrat or a Republican.
If there’s so much common ground, why do our elections seem so polarized? Think about the unprecedented wave of partisan gerrymandering in the last year or so, in both red states and blue. Is that just skin-deep?
It might be worth separating questions about issues from questions about voting. Questions about issues are really questions about what’s on the agenda, how those issues are framed, and how they might divide us in the way they’re framed. That is very much an elite-driven dialogue and rhetoric, and it’s definitely a contributor to the way in which we’re divided.
The academic consensus on this question is that the parties in Congress put forward issues that purposefully alienate the other side and bring their voters and independents over to them. They’re purposefully trying to put the other side in a politically less tenable position or force them to cast embarrassing votes. That creates polarization, or at least this sense of polarization.
Interestingly enough, when we’ve looked across all the issues, we find that about 70 percent of the very big issues on Congress’s agenda are supported by a majority of people. They’re very much majoritarian, so polarization is a secondary part of the picture compared with majoritarianism. When we look at what bills actually got passed, about 80 percent are supported by a majority of people. It’s a highly majoritarian system through all this period of political polarization.
When you talk about 70 percent of the issues on Congress’s agenda enjoying majority support, can you give an example? For instance, you mentioned that people are broadly concerned about their health. Do you see widespread agreement on health care reform?
Health care is probably the defining debate of the first quarter of the 21st century. There’s been more legislation, and more important legislation, approached and tackled on health care than just about anything else. President George W. Bush spearheaded the Prescription Drug Act. That was the biggest reform in Medicare since the Medicare Act. That was superseded by Obamacare and so forth.
Starting with the prescription drug bill, which was a very popular bill nationwide, that’s a good example of Congress taking up and passing something that enjoyed majority support. It also wasn’t a highly polarizing bill. Democrats and Republicans were in the majority for it on both sides of the issue.
Obamacare was highly divisive. Republicans were very much against it, Democrats were very much for it, and independents were sort of torn. It evolved over time as it was debated. By the end, it was somewhat more popular than unpopular, but it wasn’t a wildly popular bill. It was probably a great example of a very divisive, very polarizing piece of legislation.
When the Republicans tried to repeal it, they often found themselves on the wrong side of the majority of people. The Democrats opposed repeal, and over time the independents came to accept Obamacare, so they also opposed repeal. In the end, repeal did not succeed. It’s a really interesting case in which the evolution of public opinion around Obamacare made it very hard to repeal, even though it was also a fairly highly polarized bill. The Republicans would have liked to repeal it.
They eventually repealed unpopular parts of Obamacare, such as the individual mandate. That was an interesting case in which the individual mandate was unpopular even with Democrats. We have both parties playing both sides of the issue over time, taking all possible positions on it. In the end, the decisions the government made were really in line with the majority of people.
We find that about 70 percent of the very big issues on Congress’s agenda are supported by a majority of people. . . . When we look at what bills actually got passed, about 80 percent are supported by a majority of people.
You mentioned that 80 percent of the legislation Congress passes has majority support. There’s a broad-based feeling on both the right and the left that our democracy is dysfunctional. But you published a study earlier this year with Yale Professor Shiro Kuriwaki, PhD '21, that looked at public opinion on 103 important issues on the congressional agenda from 2006 to 2022. What did you find, and what did it tell you about how well or poorly our government enacts the will of the governed?
There are many layers to that project. The first layer was whether the average person could articulate a preference—whether people knew enough to say that they supported or opposed something. One of the striking things that comes out of most survey research is the degree to which Americans have pretty clear, consistent, and stable views on public policies. They’re not pushed around willy-nilly depending on who’s saying what. They know what they want from government, and they have a pretty good sense of what different proposals will do for them or to them. Members of Congress then vote on those bills. One concern political scientists have long had is that members of Congress may not be highly responsive to their constituents. In fact, they’re quite responsive to their constituents—not perfectly so, but quite responsive. Where representational success and failure seem to be operating is in the US House and Senate.
The US House is highly responsive to public opinion. If a bill is very popular, it’s very likely to pass. If a bill is very unpopular, it’s very unlikely to pass. The House is doing what you’d expect. It doesn’t matter if it’s a Democratic House or a Republican House, and it doesn’t matter who’s president at the time. It’s just very responsive. The degree of polarization on bills is unrelated to whether the House passes a bill. Polarization on an issue has nothing to do with what the House is doing. It’s really a majority institution.
The Senate looks very different. It’s only mildly responsive to public opinion. The Senate is very responsive to polarization, but in a negative way: the more polarized a bill is, the less likely the Senate is to pass it. That makes sense when you think about the institutions of the Senate, which are more removed from public opinion because of the six-year term. The Senate also has institutional checks, such as cloture, that require supermajorities. Those supermajority rules, procedures, and norms mean that in order to get anything done in the Senate, you have to reach across the aisle.
A highly polarizing bill is never going to get a Republican vote in the Senate if the Democrats are in charge, or a Democratic senator’s vote if the Republicans are in charge. They have to have things that are more moderate or less polarizing. The Senate actually acts as an interesting check against polarization affecting our laws.
If that’s true, then why doesn’t legislation like universal background checks for gun purchases make it through Congress? It has overwhelming support among voters in both parties.
Gun-control legislation is the one area where you would point, among all the different policy domains, to a case in which majority rule and the functioning of government don’t seem to be working. I suspect that interest-group politics, party-based politics, and so forth are controlling the outcomes.
The infrastructure bills passed under Biden are good examples of one part of the dynamic. Majorities of Democrats and Republicans in the electorate supported those bills. The Republicans in Congress, especially in the House, held the line against them. They didn’t want them to prevail. The Democrats managed to steer them through the Senate and get enough support along the way to pass them. It was a case in which party elites, pursuing a strategy of trying to polarize things and oppose the president, were making it almost impossible to pass legislation despite their own voters wanting it. In the end, the bills succeeded in both the House and the Senate. That includes the CHIPS Act and the Build Back Better bills. Those are good examples of how a lack of polarization helped push bills through Congress.
The repeal of Obamacare is a good example of a bill that was highly polarizing and unpopular, so it was not going to pass the House or the Senate. If it did pass the House, which it did a couple of times when the Republicans controlled it, the Senate effectively killed it. The House passed it for partisan reasons, understanding that the Senate was going to do its job and kill the bill. The Senate killed it because it was polarizing and unpopular; procedurally, it never made it to the floor for a vote.
If the Senate is less responsive to popular support for legislation, is that the good news or the bad news? Think of the potentially species-level threats of climate change and AI, or the broken healthcare system, or the upcoming shortfall in the Social Security trust fund. These issues are literally life and death, and the federal government seems unable to address them.
Our yardstick throughout is a democratic yardstick, with a small “d”: Is Congress doing what a majority of people want it to do? That was a question political scientists had sitting in front of them for fifty or sixty years and hadn’t been able to answer until this project. We reflect on another set of questions about whether representation is good, whether democracy is good, and whether our system is good in other senses. Does it produce good laws according to some other metric? That isn’t the metric we take up. We take up majority rule in principle.
One might argue that there are other things we’d like, such as farsightedness and equity, and that awaits further analysis. I’m not sure we could lay the lack of response to some of these issues at the doorstep of Congress. One could fault every institution in our society for not responding to them, including corporations and churches. One might ask where the religious leadership is on some of these issues. Congress is doing its job as it is laid out. It is responding to us. The question is why we are not bringing enough of these issues to the fore.
More people—a higher percentage of the population—are voting than ever before. A higher percentage of votes are being counted than ever before. Elections are more secure, and fewer are being overturned by recounts. . . . [Elections are] actually in a pretty good place right now.
Finally, many on both the left and right are concerned about the integrity of the upcoming election, particularly in light of some of the administration’s rhetoric and executive actions over the last year. How concerned are you that voting will actually be free and fair?
Recently, the Department of Homeland Security went to the state of Nevada and said, “Here are around 16,000 people that we’ve identified on the voter files as noncitizen voters.” Nevada then checked those names person by person and asked for documentation. Just 185 were positive matches. On further investigation, all these turned out to be citizens as well.
My colleague, Charles Stewart, who runs the MIT Election Data and Science Lab, was blogging with glee about it. Something like 0.000003 percent of the electorate was even questionable—not definitively noncitizen, just something we had a question about. The lesson was that if you let the experiment run all the way to the end, you learn what the result is.
To draw out a bigger lesson, my colleagues Justin Grimmer, PhD ’10, and Eitan Hersh, PhD ’11, are in the middle of writing a book that begins with a very important insight: We are in the golden age of elections.
More people—a higher percentage of the population—are voting than ever before. A higher percentage of votes are being counted than ever before. Elections are more secure, and fewer are being overturned by recounts. They go through piece of evidence by piece of evidence. We’re actually in a pretty good place right now.
The thing I worry most about is that the rhetoric around this has been sufficiently unhealthy, and I don’t think social media has helped. A lot of election administrators are getting tired of the job. These officials are often in low-level county positions, and they do the work because they love it and care about it.
It would behoove everybody who’s listening—and everybody in the United States—to take a more realistic view of this. Problems do emerge, but these officials work incredibly hard to make elections work. They’ve been under a huge amount of pressure since the 2000 election, and they’ve responded really well. I’m heartened by how well our county and town election officials have managed things. The last eight years have put enormous pressure on them, and they’ve stuck by their guns admirably.
Banner image: Shutterstock
