On Friday, I’ll be speaking on a panel about “Educating for Democratic Resilience” at a conference on Democratic Resilience at Boston College’s Clough Center. Here are some notes:
All my work is based on the theory that democracy is more resilient when many people belong to self-governing, autonomous associations. That was Alexis de Tocqueville’s insight, subsequently developed in different ways by John Dewey, Hannah Arendt, and others, and tested in recent decades by illustrious social scientists like Elinor Ostrom (who won the Nobel Prize for her work on this topic) and Robert Putnam.
I believe in it–not as an article of faith, but as a useful model. It does not always turn out to be right empirically. Democracies depend on many things, not simply on associations; and not all associations support democracy. But the model often explains phenomena that we observe in the world. More importantly, it often generates practical insights that we can use to act. Basically, by strengthening associations, we can improve the condition of our democracy. This is one of the main levers that we can pull.
Here is just one example of an empirical finding. The most recent available American National Election Study (2020) asked several items about civic participation, including this one: “During the past 12 months, have you worked with other people to deal with some issue facing your community?” That is a measure of Tocquevillian civic participation.
The ANES also asked several items relevant to the resilience of democracy. For instance, it asked whether “Much of what people hear in schools and the media are lies designed to keep people from learning the real truth about those in power.”
When controlling for education, age, gender, race, and ideology, respondents who participated in groups were much less likely to hold a hostile view of media and schools. Conservatives were more likely to be hostile, but when I included ideology in the model along with civic participation, ideology was no longer significant. In other words, irrespective of ideology, people who work with others to address local issues are more likely to trust schools and media.
From my perspective (which is contestable, obviously) approving of Donald Trump is a problematic sign. A person may vote for him for various reasons, but appreciating him as a leader suggests a lack of support for democracy. In the 2020 ANES, people who worked with others on community issues strongly disliked Trump. As expected, conservatives were more likely than liberals to approve of Trump. However, once I controlled for participation in local groups, conservatives felt no differently from liberals about Trump.
These are selected statistics, which should never be persuasive on their own. However, they illustrate the common patterns that are central to the work of Ostrom, Putnam, and others.
What does this model mean for education?
To form and sustain groups requires practical know-how. Traditionally, the most common way to obtain such knowledge was by growing up around successful organizations, but such experience has become rare as civil society has shrunk. Although US schools still teach American history and civics (with a focus on government), they do not regularly teach how to manage effective groups. Meanwhile, changes in the economy and media have created new challenges for voluntary associations, so that traditional know-how may no longer suffice.
The weakness of associational life has been evident recently. For example, when Donald Trump was elected president in 2016, informal groups popped up almost everywhere. They often attracted people who had never been involved in politics before. Stereotyped in the media as suburban white women, these citizens were informally named “The Resistance.” About a half-million of them attended the Women’s March in Washington on Jan. 21, 2017, with another 5 million marching in their home communities.
But the Resistance proved evanescent because the nascent groups mainly encouraged their members to support famous, large national organizations. About 350,000 people donated to the ACLU in just one weekend during Trump’s first month as president. People also shared and encouraged each other to follow news from national outlets, and digital subscriptions for The New York Times and The Washington Post tripled under Trump. A bit later on, many people gave money and time to Democratic Party candidates.
But this generosity and energy did not build local associations. Most–although not all–of the nascent groups faded away. As Erica Chenoweth and Zoe Marks observed, many committed and skilled activists emerged, yet we have “no established, organizational infrastructure that can facilitate sustained collective action across a multiracial, multiclass constituency.”
I believe one reason is that too few Americans understand the nuts-and-bolts of associations. An initial meeting often draws many concerned people, who may use the time to express strong feelings and to share ideas, some of which are excellent. But nothing concrete is decided. The second meeting draws some of the same people along with many newbies and recapitulates the first discussion. By the third meeting, most people are too frustrated to continue. The group needed an agenda, a committee of accountable, volunteer leaders, a decision-making process, and a budget.
The resistance in Minneapolis this winter may suggest a more optimistic story. In January, The Atlantic’s Robert F. Worth reported that local groups there had trained 65,000 residents in nonviolent civil resistance since the previous month. That is evidence of impressive organizational muscle.
Thomas Friedman recently celebrated the Minneapolis movement as “neighboring.” He quoted a local business executive who described its decentralized and participatory structure. “There were hundreds of leaders of this movement,” he said, “and I don’t know a single one of their names.” Geneva Cole argues that Minnesota groups that launched or expanded in the wake of George Floyd’s murder in 2020 have become strong enough–both as organizations and as a network– that they were able to accomplish disciplined mass action this winter.
That is a promising sign, but in general, associations remain too weak, and too few Americans have the necessary practical skills. Schools and colleges should devote more attention to the nuts-and-bolts of effective groups.
I would also love to teach older adults how to do those things, but it’s unrealistic to offer civic education to millions of older people. An alternative is to make processes easier so that groups can focus on the substance of their work.
After all, associations depend on documents: recruitment emails, agendas, minutes, job descriptions, and budgets, among many others. Say what you will about Large Language Models (LLMs), but they can quickly draft documents. I’ve launched an experimental helpdesk that uses an LLM that is trained on specific documents about voluntary organizations and instructed to respond to queries in specific ways.
The target user of this helpdesk is a newcomer to civic life who is highly concerned about a current problem: I call these people “Alarmed Complete Newbies” or ACNs. The helpdesk encourages them to find and join groups that already exist. If users needs to launch a new group or help to strengthen an existing one, the AI nudges them to request useful documents, which it then drafts for them.
The goals of this project are, first, to enable people to make progress together even if they never learned how to manage voluntary groups; and, second, to learn principles and skills from this experience.
The helpdesk is just an experiment, not a panacea. The underlying idea is that democracy and local communities will be stronger once more groups of concerned people gell into effective organizations that collect time and money from their own members, use their own resources to build their own capacity, make collective decisions, and act effectively. People must learn how to do those things. That is what I mean by “education for democratic resilience.”
