The real reason why Democrats are so unpopular

The Democratic Party has spent most of the past decade deciding to lose.

Or so argues a new report from Welcome PAC, an organization that backs center-left candidates, so as to build “a big-tent Democratic Party.”

It is no secret that the Democrats are in a sorry state. They’ve lost to an exceptionally unpopular Republican presidential nominee twice in the last nine years. They have long odds of regaining control of the Senate next year and aren’t even certain to retake the House of Representatives. They’re historically unpopular, having lost ground with much of the party’s traditional base — including voters who are working class, people of color, young, or all of the above.

To determine how this happened, Welcome’s Simon Bazelon conducted six months of polling with nearly half a million voters, examined the results of hundreds of recent elections, analyzed shifts in the Democratic Party’s legislative priorities, and crunched various other data points. In so doing, Bazelon produced a rigorous and thorough accounting of what the centrist organization already knew: The Democratic Party has veered too far left, effectively choosing to prioritize progressive orthodoxy over electoral success.

Welcome’s report has already resonated with many of the Democratic Party’s leadership. And if Bazelon’s analysis proves persuasive to Democratic insiders, it could shape the trajectory of the party’s 2028 presidential primary — and thus, the future of American democracy.

To see how well that analysis stands up to scrutiny, I spoke with Bazelon about various progressive objections to his arguments. Our conversation has been edited for length and clarity.

Give me the short version of your story: How did Democrats end up in their present state?

Since Barack Obama won reelection in 2012, the Democratic Party has undergone two really major shifts. First, we have shifted our priorities, focusing less on kitchen table economic issues and more on issues that are less concrete and more abstract to voters: climate change, democracy, abortion, and other identity and cultural issues.

Meanwhile, the Democratic Party became a lot more left wing than it used to be across the board. In 2013, 24 percent of Democrats in Congress co-sponsored Medicare for All. In 2023, that was 47 percent. In 2013, 41 percent of Democrats in Congress co-sponsored an assault weapons ban. Now, it’s 88 percent. Only 1 percent co-sponsored a reparations study bill in 2013. Now, that’s a majority. And I think these two shifts are primarily responsible for the situation that Democrats are in today.

At the same time, polls also show that voters believe Democrats have been focusing too much on social issues and not enough on concrete economic issues.

Elections are complicated and multi-causal. In 2024, there was obviously a big anti-incumbent surge around the world. Inflation played a big role, but immigration also became incredibly salient to voters. They really disapproved of the way that the Biden administration handled immigration. And I think you can draw a direct line from the party’s leftward shift to the Biden administration’s handling of immigration to voter disapproval of Biden to Trump gaining votes.

And there’s another indication that the ideological shift mattered: Democratic candidates who have resisted these shifts, who are more in line with where the party was 10 or 15 years ago, do better electorally.

Some political scientists have disputed that and suggested the benefits of moderation have all but disappeared in recent elections. In their view, when we look at results from 2024 or 2022, we simply don’t see moderate candidates significantly outperforming extreme ones. What do you think they get wrong?

It’s going to be hard to answer this question without really getting into technical specifics. But I’ll say that Grumbach and Bonica’s papers are one data point. They have a view. I think looking at the full sweep of the literature over the last decade, the primary thrust is that there still is a penalty for more ideologically extreme candidates and a benefit for more ideologically moderate candidates. I think it is true that that benefit is smaller than it was 20 years ago. But also, elections are a lot closer than they used to be.

Adam Bonica’s own paper from last spring found that running a more moderate candidate would be worth a 1 percentage point increase in Democrats’ share of the vote in a presidential election.

Well, if Democrats did one point better in the 2016 and 2024 presidential elections, they would have won both. If they did 1 percentage point worse in the last three presidential elections, Republicans would’ve won all three.

What’s your theory of why Democrats shifted their focus and ideology since 2012, if they were not responding to what the electorate actually wanted?

I think it’s primarily a story about Democratic elites. It’s about donors, both large and small, about the staffers who run Democratic campaigns. It’s a story about the pundits who are influential on Twitter. It’s a story about the advocacy groups that have a lot of policy sway in Democratic spaces. All these groups have themselves moved left over time.

Meanwhile, as the party became increasingly a high-socioeconomic-status party, those groups got increasingly influential. And I think there’s a ton of polling and academic evidence showing that Democratic elites have systematically different priorities and preferences than the general electorate.

So, compared to swing voters, college-educated Harris supporters were 20 percentage points more likely to prioritize climate change, 16 points more likely to prioritize voting rights, 17 points more likely to prioritize gun issues, and 15 points more likely to prioritize the environment.

And then, on the other side, they just care a lot less about border security, immigration, crime, gas prices, and budget deficits. So, the increasing influence of these increasingly out-of-touch Democratic elites has pulled the party away from the views and priorities of regular voters.

I think many people are skeptical that policy positions shape the parties’ popularity in a straightforward way. After all, voters often don’t know a candidate’s precise policy positions. And their impressions of where the parties stand are informed by broad media narratives, which aren’t necessarily grounded in reality. Thus, if conservatives dominate the podcast and social media scenes, then they can promote the idea that Democrats are radically left wing, no matter what positions the party actually takes. For example, one poll suggested that about half of swing voters falsely believed that Kamala Harris supported defunding the police.

So, to some, this suggests that Democrats should worry less about moderating their positioning than increasing their media influence.

I don’t think that there’s anything wrong with trying to get more attention on social media. But it’s really important to be getting attention for the right reasons.

In our data, how well a Democratic candidate did in 2024 was completely uncorrelated with the size of their social media following. To the extent there was any relationship between those two variables, it was negative. Democrats with the most followers on social media did slightly worse than Democrats with fewer followers. Jared Golden [a Democratic Congressmember from Maine who has repeatedly won in a pro-Trump district] isn’t some TikTok star. Marie Glusenkamp-Perez [a Democratic Congressmember from Washington who has repeatedly won in a pro-Trump district] isn’t a TikTok star. Instead, these are candidates who are to the right of the median Democrat on issues like immigration and who laser-focus on economic issues.

More broadly, the idea that voters don’t notice substantive shifts is belied by what I said up top: The Democratic Party has moved left since 2012 and — at the same time — the share of voters who think the Democratic Party is too liberal has increased a lot.

Many liberals believe that there is a trade-off between making the party more palatable to swing voters and maximizing turnout among the Democratic base. And they contend that in modern elections, mobilization is what really decides races, since there are so few nonpartisan voters left in America’s polarized climate. Why do you think that’s wrong?

So, first, the lion’s share of the academic literature and data on this topic shows that more progressive Democrats tend to underperform electorally, which I think directly contradicts the theory of mobilization. But more importantly: The data is really clear that there’s no trade-off between appealing to non-voters and appealing to swing voters. Those two groups have very similar priorities and policy opinions.

There’s this myth in Democratic politics of who the nonvoter is, which is that the nonvoter is a latent socialist who is disaffected from the Democratic Party because they see the Democratic Party as sellouts of true progressivism.

That describes a lot of people who work in Democratic politics very well, since they are more left wing than the party as a whole and are frustrated when the party tries to meet voters where they are. But the fact that people who work in Democratic politics see that kind of archetype around them a lot does not mean that it is, in fact, a super common outlook among the general electorate.

I think many progressives would say that your theory of the case was already tried and found wanting. Kamala Harris ran to the center in 2024, advertising herself as a border hawk and tough prosecutor. She campaigned with Liz Cheney and eschewed bold progressive promises. And she lost. Meanwhile, in 2020, Joe Biden actually ran on the most progressive platform in the party’s history and won.

I think both Biden in 2020 and Kamala Harris in 2024 show us that voters are evaluating candidates in large part on their record over a period of several years, not just in the three months before the election.

Kamala Harris was a senator between 2016 and 2020. I believe she was the second-most left-wing senator in the country, based on her voting record. When she ran for president, she endorsed a whole slew of left wing policies, from banning fracking to banning plastic straws. Then, she became vice president for a president that the overwhelming majority of Americans thought was too liberal. And then, she refused to distance herself from any of that president’s actual policy positions. And when she moved away from some of her own older stances, she did it quietly and never articulated an explanation for why she thought those old ideas were bad.

I don’t think that Kamala Harris’ efforts to moderate hurt her electorally. We saw, during her short-lived campaign, that her approval rating went up a lot from the beginning to the end. So, I think it’s sort of hard to argue that her efforts to moderate were the reason she lost. But, at the same time, I think it really just shows that you can’t expect to talk a big game on border policy in the three months before the election if your actual border policy when you’re in office allows millions of unauthorized immigrants to come into the country.

And then, with Biden in 2020, he was coasting off of his longstanding reputation of being a moderate Democrat and being to the right of other Democrats. When he actually took office, he governed as progressive; progressives acknowledged this at the time. And then, during his administration, the data shows that the share of voters who thought that Joe Biden was too liberal increased dramatically. And so, I think it’s just a good example of, again, your substantive positions really matter for how voters perceive you.

Your report suggests that Democrats need to moderate their positions on immigration, public safety, energy, and some identity and cultural issues. Could you get specific? On immigration, which of the party’s current policies are too left wing?

On immigration, the party is a lot closer to public opinion than it was years ago, when you had almost every Democratic presidential candidate raising their hand to say, “We want to decriminalize border crossings.” So, the party has learned a bit of its lesson. I think the problem is that Democrats are facing an absolute crisis of credibility and trust. Even as Trump is unpopular in polls, the Republican Party continues to have an enormous advantage on the question of which party you trust more to handle immigration, to handle border security, to handle crime. And so, I think a big step for Democrats would be to acknowledge what happened and give Trump credit for securing the border.

Some Democrats have been doing that. Bernie Sanders had a really good line about this the other day.

I think what makes some proposals for moderation difficult, from a progressive standpoint, is that they entail accepting certain substantive costs for the sake of hypothetical political benefits.

If Democrats abandon their commitment to increasing refugee admissions, then that directly constrains the potential number of stateless people who could one day find safety and prosperity in the United States. If the party foreswears a carbon tax, then that limits its capacity to reduce climate change.

On the flip side, there’s no guarantee that moderating on any particular issue will deliver a significant electoral benefit. And really, the upside of shifting any one issue position is likely to be tiny if not negligible.

Yeah, I think that really gets to the core of what has driven a lot of the Democratic Party’s problems in the last couple of years. During the Biden administration, we had government by a “coalition of the Left,” where every individual faction had veto authority over any policy position taken by the administration, at least during the first years. And I think what you’re saying is exactly right: Moderating on any individual policy doesn’t necessarily swing the election. The problem is that if you apply that logic to every single scenario, you end up with a party that is dramatically more left wing than it used to be and is going to have a much harder time winning elections. And then, the consequences of losing those elections are absolutely enormous.

So, while progressives may feel like there are some substantive costs to moderating on all these issues, when the alternative is governance by a faction of the country that has a disregard for both the economic well-being of lower-income Americans and the rule of law, I think the trade-offs just become quite clear.

Do you think that the Democratic Party should retain any policy positions that are unpopular with the electorate, merely because they are right on the merits?

I think foreign aid. Public opinion is against it, but the humanitarian benefits are so large that the electoral costs of supporting it — which I don’t think are huge — seem clearly worth it.

Isn’t there a tension between delivering the electorate’s preferred outcomes and appealing to its policy instincts? For example, tariffs often polled well before Trump actually implemented them. Yet, when his tariffs raised prices, voters turned against them. So, what should Democrats do when the imperatives of good governance and electoral expedience come into conflict?

It’s much harder to figure out how to govern a country well than it is to figure out how to win an election.

The state of the economy is the single most important factor in determining election outcomes. And so, it’s really, really important for Democrats to get macroeconomic management correct. I think the problem is that macroeconomic management is really hard. The Biden administration messed up on inflation, but the Federal Reserve didn’t predict that inflation, markets didn’t predict that inflation. And under the Obama administration, Democrats undershot on stimulus. So, it wasn’t irrational for the party to do a lot of stimulus under Biden. It just didn’t work out.

But, we should be focused on delivering stable prices, low unemployment, and growth. And that requires not just always playing to public opinion per se.

Your report argues that moderation doesn’t mean being “feckless and weak.” Rather, you say the party should “stand firm against Trump and the Republican Party” but be “disciplined and strategic in which fights we pick, and how we pick them, by focusing our opposition on issues where public support is most on our side (like protecting Social Security, Medicare, and Medicaid, opposing tax cuts for the wealthy, and opposing Trump’s tariffs).”

Reading that, some liberals may think: Okay, but Democrats can’t always pick the fights they want to have. When the president sends an immigrant to a foreign prison without due process, or revokes legal status from immigrants based on their activism, or throws trans people out of the military, Democrats can either stand firm against Trump — on those pressing matters — or keep their opposition focused on bread-and-butter issues. But they can’t do both. So, in practice, they’d argue, the discipline you propose entails fecklessness on vital moral questions.

I’m not saying that the Democratic Party needs to start endorsing the extreme policies of the Trump administration.

But there are absolutely trade-offs in politics, in your prioritization. Every tweet you send, every speech you give, every ad you run about one topic is a forfeited opportunity to speak about another topic. Every time the Democratic Party engages in political discourse, we are influencing voters’ perceptions of what we stand for and what we care about. And the reality is, yes, there are trade-offs, and if we want to win elections, then we need to show voters that we feel that issues like health care and the cost of living are the most important issues.

Some would agree with you that Democrats should center their messaging on economic issues but argue that they should do this by moving left — embracing a more robustly populist economic agenda. After all, there is a lot of anti-establishment sentiment in the country, a lot of disaffection with the performance of our institutions and elites. By failing to provide a sufficiently distinct economic vision, this argument goes, Democrats haven’t given working-class voters with culturally right-leaning views a reason not to prioritize their social intuitions.

But if the party adopted more egalitarian policies — such as a large wealth tax on billionaires, the breakup of large tech companies, big minimum wage hikes, and Medicare For All — it could regain credibility on economics and win back some of those voters without making big cultural concessions.

I’m not opposed to economic populism. I have a lot of sympathy for economic populists in all factions of the party, whether it’s folks like Jared Golden or Bernie Sanders.

But I think that there are a couple things that the populist argument gets wrong. The main thing is just the idea that a more aggressive approach on economics can distract voters from other issues, whether it’s immigration, or crime, or energy, or cultural issues. And I think Bernie Sanders’s own career shows that’s not true. In 2006, and in 2012, Bernie Sanders was to the right of the median Democrat on issues like immigration and guns. He was endorsed by the police union earlier in his career.

But then, he moved. He endorsed decriminalizing border crossings. He endorsed a very expensive climate policy that included big restrictions on energy production. As that happened in 2024, we ended up with a situation where Bernie Sanders actually did worse in his home state — despite being a multi-term incumbent in a small state — than Kamala Harris. Meanwhile, Elizabeth Warren ran well behind Kamala Harris in her home state. I think that shows that combining economic populism with social views that are really out of touch does not work.
Voters aren’t going to just stop thinking about those issues because we talk a lot about oligarchs and billionaires. They’re still going to care about immigration.

And then, there’s a big difference between focusing on economic issues and moving left on economic issues. In our polling, many left-wing economic policies that Bernie Sanders supports — significantly expanding prescription drug price controls and expanding Medicare to cover dental, vision, and hearing — are popular. But policies like student loan forgiveness, big new universal social programs like “free college” are not popular, at least when you tell voters how much they cost.

On the whole, means-tested programs are more popular than universal ones. Not in every case; universal free school lunch is more popular than means-tested school lunches. But increasing Social Security disability insurance, expanding Social Security benefits for low-income elderly Americans — these things are very popular. In general, voters prefer incremental safety-net expansion.

And this is borne by which candidates do better in elections. If being really left wing on economic policies were popular, we’d see someone like Kara Eastman — who supported Medicare For All — beating Don Bacon in Nebraska’s Second District. But instead, what we see is that the Democrats who win swing districts are absolutely populist on economics but not left wing in the way that Elizabeth Warren is.

So, who should Democrats nominate in 2028?

I don’t want to speak for the voters.

But when looking at candidates, it’s really important to evaluate how they did in the past. And so, we did an analysis of how every Democratic candidate did in their most recent election, relative to expectations. And what we found is that Democrats have a wide number of strong options, whether it’s Josh Shapiro, Andy Beshear, Ruben Gallego, or Amy Klobuchar.

But, at the moment, the candidates leading in Democratic primary polls are AOC, Gavin Newsom, and Kamala Harris, who were all under-performers in their elections. So, I would strongly urge every Democrat who wants to win in 2028 to look really carefully at the policy positions of the Democratic candidates, to look at their track record and their ability to win over the kinds of swing voters who backed Donald Trump in 2024.