Category Archives: elections

reaching the opt-outs

In today’s New York Times, Rob Flaherty, who was Kamala Harris’ deputy campaign manager, argues that Democrats lose “opt-out voters,” people who distrust all politicians and all traditional media and who obtain their politically relevant information from other sources, such as online influencers or real-life contacts who follow the influencers. These “opt-outs” may start out looking for tips on health or nutrition or relationships or gaming (not politics), but they find their way to right-wing propaganda.

I can support some of these generalizations with data from the 2020 American National Election Study. (I don’t think 2024 data are available yet). For example:

61.5% of strong Republicans and 5.5% of strong Democrats expressed no trust in the media.
26% of strong Republcans and 8% of strong Democrats fully agreed that “Much of what people hear in schools and media are lies by those in power.”
6% of strong Republicans and 30% of strong Democrats trusted experts much more than ordinary people for public policy.
Of those who said they did not follow the 2020 campaign using any source listed on the survey, 57% said they intended to vote for Trump; 27% for Biden.

It would be possible to overstate this problem. If most Americans only got information from unreliable influencers, then Trump’s approval rating would not have declined across most of the population in 100 days, before his tariffs and cuts had directly affected many people. If influencers had persuaded everyone to hate civil servants and researchers, then DOGE’s personnel cuts would not be as unpopular as they are. Evidently, many non-Democrats are seeing hard news. Nevertheless, Flaherty’s diagnosis is important.

His recommendation is to build an alternative media environment that carries people from “culture” (their interests in regular things like health or relationships) to liberal political ideas.

I doubt this approach is realistic, and it creates more of conflict or even contradiction for the institutionalist center-left than it does for the MAGA right (or, indeed, for the radical left). Basically, it asks liberals who believe in institutions to use anti-institutionalist means, which looks hypocritical and may prove impossible.

Here is an alternative: People have reasons to trust big, impersonal systems only when the human representatives of those systems relate to them well. For example, I trust the mainstream scientific views of vaccines and climate change not because I understand all the science, but because human beings who represent science as an institution–my own k12 and college teachers, doctors and nurses, and now my academic colleagues–have generally earned my trust. They relate to me with respect, as a fellow citizen.

Actually, not even scientists understand the science, because the necessary knowledge exceeds any person’s capacity (and much of it is built into instruments and software and datasets that each user must simply trust). But some of us have confidence in the whole process because we have benefitted from most of the moments when it has touched us directly.

The sociologist Anthony Giddens calls this process “re-embedding”: contacts between abstract systems and ordinary people via professionals who represent the systems. To be honest, I have never read a significant amount of Giddens, but I take his vocabulary from a relevant article by Mills and St Clair (2025).

The employees who are points of contact between abstract systems and regular people include teachers and professors (and educational administrators), doctors and nurses, lawyers and police officers, local elected officials, and reporters.

Americans have widely differing experiences with these professionals and varying grounds for trust. If you are at risk of being stopped and harassed by the police on account of your race, you do not have a reason to trust the criminal justice system. If your doctor dismisses your concerns, or you can’t even afford to see one, then you have less reason to trust the health sciences. If you can’t get into college, can’t afford the tuition, or experience contempt for your home culture in a college classroom, then our trust in academia is bound to fall. If your kids’ k12 school is failing–or if it seems driven by standardized curricula and tests and there’s no way for parents to engage–then you have reasons to be skeptical of schools.

For center-left institutionalists, I don’t think there’s any shortcut. In an environment where it pays to attract outrage by attacking abstract systems, we must make these systems as accountable, caring, and interactive as possible so that people will have reasons to trust them more.

The goal is for people’s “influencers” to be their own kids’ teachers, their doctors, and the reporters for their local newspaper (among others). This requires not just encouraging them to trust people who often have more education, power, and income than they do, but also making these professionals more consistently trustworthy.

Source: Mills, M. Anthony, and Price St. Clair. “The Strange New Politics of Science.” Issues in Science and Technology 41, no. 3 (Spring 2025): 40–48. https://doi.org/10.58875/NDTQ1755. See also to restore trust in schools and media, engage people in civic life; my own trust in institutions; it’s no accident that people distrust institutions (2017); and many other posts.

affective partisanship and young people

Leave a reply

Affective partnership means disliking people who belong to a competing party. This attitude can be defensible in certain circumstances, but it certainly poses an obstacle to regular democratic processes.

To measure it, I use the American National Election Study’s “partisan thermometer” question, which asks people to rate members of the other party from 0-10.

The graph below shows the proportions of Democrats and Republicans who rate the other party at zero. These proportions have risen rapidly and basically symmetrically, although in 2020, Republicans were somewhat more hostile to Democrats than vice-versa. (Data from 2024 are not yet available.)

The bar graph with this post adds detail about young adults from the 2020 election. That year, young Democrats and Republicans were equally likely to rate each other at zero. Democrats’ levels of affective polarization did not change with age, but older Republicans were more polarized than younger ones. Indeed, a majority of Republicans over 60 rated Democrats at zero that year, and older Republicans accounted for the difference between the parties that is evident in the line graph.

In short, if affective polarization is a problem, it is not particularly a youth problem, which means that solutions cannot depend on civic education alone.

CIRCLE report: How Does Gen Z Really Feel about Democracy?

Leave a reply

My colleagues at CIRCLE have issued a report with Protect Democracy that is based on a nationally representative sample of citizens between 18 and 29 that they conducted soon after last fall’s election.* At the heart of the study is a categorization of young American citizens into three groups. In the words of the report, these groups are characterized by:

Passive Appreciation: The majority of youth (63%) value the basic values and practices of democracy, but they are relatively disengaged from civic action and may be passive in the face of current threats to democracy.

Dismissive Detachment: Nearly a third of young people express lower support for core democratic principles; they are “checked out” of a democracy that has not served them well or met their needs.

Hostile Dissatisfaction: A small (7%) but significant number of young people believe in the principles of our system of government, but are extremely dissatisfied with our democracy as it exists today, and they are willing to consider political violence in order to achieve change.

The first group is relatively likely to be conservative (or perhaps less likely to be on the left) and less likely to be queer. Members of this group not very involved with civic actions like volunteering and protesting.

Those who exhibit “dismissive detachment” tend to have less education and are more likely to be people of color. They “are less likely to value the basic principles of democracy such as free and fair elections, the protection of civil rights, and bipartisan cooperation.” They are more hostile to people who disagree with them and feel less efficacious in politics. They have the lowest media literacy and other civic skills.

People in the “hostile dissatisfaction” group are more likely to be on the left and more likely to be queer. “Youth with this profile do tend to value the principles and rights of our democracy, scoring slightly above average in those areas. But they have very low confidence in democracy as they are experiencing it today.” They express strongly negative views of people with the opposite political ideology, which could be labeled “affective polarization.” They are the most engaged in political action and also the most favorable to violence.

You can read the whole report here. It offers important insights for anyone who provides civic engagement and civic learning to young people in this moment.

*Apau, D., Suzuki, S., Medina, A., Booth, R.B. (March, 2025). How Does Gen Z Feel About
Democracy? Insights from Three profiles of Youth and Democracy. CIRCLE (Center for Information & Research on Civic Learning and Engagement) & Protect Democracy.

Where have lower-educated voters moved right? (a look at 102 countries over 35 years)

Leave a reply

I believe that left parties should draw their votes from lower on the socio-economic hierarchy, so that they can compete by offering more governmental support. Right parties should draw their votes from the upper end, so that they can compete by promising economic growth. This debate and competition is healthy.

In contrast, when left parties draw from the top of the social order, they tend to offer performative or symbolic policies, while the right promises low-SES voters some version of ethnonationalism. This debate is unhealthy because it blocks more effective and fair social policies, and it sets the right on a path whose terminus can be fascism.

Education is a marker of social class. We saw a social class inversion in the US 2024 election, with Harris getting 56% of college graduates and Trump getting 56% of non-college-educated adults.

Nowadays, we are used to assuming that Republicans have an advantage in the Electoral College because they are dominant in the states with the lowest percentages of college gradates, while Democrats win easily in the most educated states. But the opposite should be true.

I am fully aware that race is involved in the USA. Recently, less-educated white voters have formed the Republican base, whereas voters of color have preferred Democrats, regardless of their social class. However, in 2024, we saw a significant shift of low-education voters of color toward Trump.

Besides, race plays different roles in various countries, but many countries display a trend of lower-educated people preferring the right and moving in that direction .

The World Values Survey has periodically surveyed populations in 102 countries since 1989, for a total sample of almost half a million individuals in the dataset that I used for this post. The WVS asks most respondents to place themselves on a left-right spectrum, and the global mean is somewhat to the right of the middle. It also asks people their education level. For the entire sample, the correlation between these two variables is slightly negative and statistically significant (-.047**). In about two-thirds of sampled countries, the correlation is negative. This pattern is upside-down, suggesting the people with more education tilt mildly to the left around the world.

However, considering the heterogeneity of the countries and years in this sample (from Switzerland in 1989 to India in 2023), it is important to break things down.

The graph with this post shows the correlations for wealthy countries with democratic elections: the EU countries, the USA, Canada, Australia, New Zealand, South Korea, Japan, and Taiwan. A positive score indicates that people lower on the educational spectrum are more likely to vote left. The trend is slightly downward, meaning that the highest-educated have moved a bit left (and the lowest have moved right).

Among the countries that have recently demonstrated a class reversal (with the lower classes voting right) are Australia, Canada, Greece, Germany, Japan, the Netherlands, New Zealand, Northern Ireland, the UK and the USA. Czechia and Slovakia are the main exceptions. In Japan and South Korea, less educated people have consistently favored the right to a small degree.

By contrast, for a sample of Latin American and Caribbean countries, the trend has generally been toward what I consider the desirable pattern, with lower-educated people increasingly voting left. The mean for all voters in this region is distinctly left of center.

In a cluster of countries that were part of the Warsaw Pact and are not now members of the EU, the trend is flat. Interestingly, in these countries, the mean voter is on the right.

Finally, the WVS surveys in some countries from the Global South–from Bangladesh to Zimbabwe-but these countries do not seem representative of the whole hemisphere. For what it’s worth, the trend for this sample is just slightly upward, and the results vary a great deal among countries.

I am still in the “deliverism” camp, believing that left parties have not delivered sufficient tangible benefits to less advantaged voters since the 1990s. (One explanation could be their dependence on affluent voters, who do not really want them to do much.) Achieving more tangible change could turn things right-side-up again.

However, it should give us pause that the Biden Administration actually spent trillions of dollars in ways that will benefit working-class Americans, yet Trump won and drew an increasing proportion of lower-educated voters of color. The “deliverist” thesis now depends on the premise that Biden-Harris had too little time and suffered from post-COVID inflation.

Meanwhile, if your premise is that US working-class voters moved right due to (increasing?) racism and sexism, you need an explanation of similar trends in many countries, including some without substantial ethnic minorities.

complexity and nuance about public opinion

Leave a reply

Last Monday, I gave a talk at Colgate University. I claimed that if you read a lot of mainstream survey research, you’re likely to conclude that “people are stupid and they hate each other,” but this negative assessment reflects some bias. A student, Colgate senior Clementina Aboagye, told Maddie Koger of the Colgate Maroon-News:

“I think it was important that we had someone like Peter Levine who comes from an institution like Tufts University to present us [with the idea] that as much as we may disagree with each other, we still have complexities in how we think — that it’s important we search for gray areas because politics isn’t so black and white,” Aboagye said. “Those gray areas are important for us to not only converse about, but also to give each other space to speak — even when we don’t agree, because we can’t always agree — and we live in a world where people’s experiences and access to things determine what kind of ways in which they think — that deserves consideration.”

The very next day, a majority of American voters chose Donald Trump, concluding a campaign marked by polarized media, misinformation, hostility, and attacks (from one side) on basic liberal norms. Yet I still think there’s truth in the argument I offered at Colgate, which you could watch in full here.

The previous week, I had given American University’s annual Lincoln Scholars Lecture on “What Should We Do? A Theory of Civic Life.” Although my topic was quite different, this talk also offered a more positive view of civic life than one would glean by focusing only on an ugly and dangerous national election. According to Ridha Riyani’s summary in the AU Eagle newspaper, I said,

“We disagree because we care, and we need to do it better. …

[Levine] ended with a call to action, reminding attendees that civic life extends beyond national politics and requires thoughtful collaboration.

“In conversation, we can move towards greater wisdom,” Levine said. “We communities are capable of changing the rules.”

I would not argue now that all we need is to listen generously across differences and explore the complexity of other people’s views. We must also stand up against injustice. Confrontational nonviolent civil resistance was a major theme in my AU talk, and I have been preparing for a Trump victory for several years. Still, there remains a place for listening, bridge-building, and collaboration, and I strive to offer useful concepts and skills for those purposes.