Why Town Halls Fail And What Would Replace Them

The town hall format emerged in American political culture as an analog to the New England town meeting — a face-to-face forum where all citizens could participate in local governance. The New England town meeting, at its best, had some genuine democratic properties: it was small enough that most participants could speak, decisions were made at the meeting rather than by officials afterward, and participation was tied to actual decision-making power.

The modern town hall inherits the aesthetic of this tradition without most of its functional properties. The modern version is typically:

- Open to anyone (not a selected representative sample) - Not decision-making (advisory at best, performative at worst) - Microphone-at-the-aisle format (sequential speeches, not discussion) - Large enough that most participants will not speak - Often televised or streamed (changing participant behavior toward performance)

These structural features produce predictable dynamics:

Selection bias. Who attends a voluntary town hall? People with strong, usually negative, views about the subject at hand. People with the time, transportation, and social comfort to attend a public meeting. People with community organizations that mobilize attendees. The result is a room that wildly overrepresents motivated opponents and underrepresents the majority who hold moderate, ambivalent, or simply uninformed views. The sample is not remotely representative of the community.

Performance incentives. With an audience and a microphone, participants optimize for applause rather than inquiry. The goal is not to learn something or advance the conversation; it's to land the effective point that gets a reaction. This is a fundamentally different cognitive task than deliberation.

No listening mechanism. Deliberation requires that participants actually listen to and consider what others say. A standard town hall has no mechanism for this. When you're waiting for the microphone, you're rehearsing your statement, not processing others' arguments. When you're not scheduled to speak, there's no accountability for whether you're listening.

Adversarial frame. The physical setup — residents at the microphone, officials at the table — creates an inherently adversarial frame. Both sides perform for their respective audiences. Officials perform responsiveness and control; residents perform grievance and demand. Neither side is in the cognitive or social posture for genuine inquiry.

No iteration. A single town hall event is a snapshot. There's no opportunity to test proposals, collect evidence, reconsider in light of new information, or develop shared understanding over time. The format assumes that one evening of speeches is sufficient for complex community issues.

The research confirms these intuitions. Lee Drutman and colleagues at New America found that political events in adversarial format — including town halls — reliably increased participants' negative views of the opposing side. Christopher Karpowitz and Tali Mendelberg's research on deliberation found that standard group discussions, without structured deliberative conditions, systematically amplify the voices of those who already have social advantages and suppress women, racial minorities, and lower-status participants.

James Fishkin developed deliberative polling at Stanford's Center for Deliberative Democracy over several decades. The method has now been tested in more than 100 instances across six continents, including in national policy debates, local planning processes, and cross-national comparisons.

The core design:

1. Random sampling. A statistically representative random sample of the relevant public is recruited. This solves the selection bias problem: the deliberating group reflects the actual distribution of views, demographics, and information levels in the population.

2. Balanced briefing materials. Participants receive materials presenting multiple perspectives on the issue, prepared with input from a cross-partisan advisory group. No single perspective dominates the information provided.

3. Small-group deliberation. Participants deliberate in small groups (typically 8-12) with a trained moderator whose job is to ensure everyone speaks and the discussion is substantive, not to guide participants toward particular conclusions.

4. Expert and stakeholder questioning. Participants develop questions through their small-group discussions and pose them to panels of experts and stakeholders representing diverse perspectives. This gives ordinary citizens access to the kind of information and expertise that elites routinely access.

5. Measurement. Participants' views are measured before and after. The before-after comparison reveals whether and how deliberation changed views.

What the research shows: deliberative polling consistently produces significant opinion change — not in a single direction, but toward more nuanced, informed, and often more moderate positions. Participants consistently rate the experience as valuable. The views of deliberated samples are more stable over time than views formed through conventional media exposure.

Notable examples:

"What Do Americans Really Want?" A national deliberative poll on America's role in the world conducted in 1999 found that after deliberation, Americans significantly increased their support for multilateral engagement and were more willing to accept international agreements limiting U.S. sovereignty. The deliberated views were markedly different from polling of the general public, and markedly more aligned with expert assessment of U.S. long-term interests.

Mongolia's national deliberative poll (2017) brought together a representative sample of Mongolians to deliberate on constitutional reform. The exercise influenced actual constitutional amendments — an example of deliberative polling connecting to formal political process.

Europolis (2009) brought together a stratified sample of citizens from across EU member states to deliberate on European issues. The project demonstrated that cross-national deliberation was feasible and that it produced convergence on some issues along with persistent disagreement on others — a more honest picture of European public opinion than either elite consensus or unreflective polling.

Texas utility deliberative polls are among the most directly consequential. Fishkin partnered with Texas utilities to conduct deliberative polls on electricity generation options. The results — that deliberated Texas consumers supported renewable energy investment more strongly than unreflective surveys suggested — influenced actual utility planning decisions.

Citizens assemblies go beyond deliberative polling in their formal connection to political decision-making. A citizens assembly is convened by a government authority, given a specific question to address, and either tasked with producing formal recommendations or, in some cases, given direct decision-making authority.

The most studied examples:

Ireland's Citizens' Assembly (2016-2017) was convened by the Irish government and tasked with recommending whether Ireland should repeal the Eighth Amendment (constitutional ban on abortion). The assembly — 99 randomly selected citizens, meeting over multiple weekends — recommended repeal by 64 votes to 25, with specific recommendations on the framework for new legislation that were more permissive than most political observers expected citizens to support.

The Irish government put the question to a referendum in May 2018. The vote was 66.4% in favor of repeal. The citizens assembly had accurately predicted what a deliberated public would decide — and the deliberated public proved more willing to change than conventional political process had assumed.

Iceland's constitutional assembly (2010-2011) followed the financial crisis and used a similar model to produce a draft constitution via a 25-member elected assembly working in public view, with online input from citizens. The draft was approved in a non-binding referendum by 67% of voters. (Iceland's parliament ultimately did not adopt the constitution, for political reasons — an illustration of the limits of deliberative processes when political elites aren't committed to honoring outcomes.)

UK Climate Assembly (2020) brought together 108 randomly selected UK residents to deliberate on how the UK should achieve net-zero carbon emissions. The assembly's recommendations — published as "The Path to Net Zero" — provided detailed, citizen-developed policy recommendations that were more ambitious than government targets at the time and influenced subsequent policy development.

French Convention Citoyenne pour le Climat (2019-2020) convened 150 randomly selected citizens to produce proposals for reducing French greenhouse gas emissions by 40% by 2030. The convention produced 149 proposals. President Macron committed to putting some proposals to referendum and incorporating others into legislation — though the follow-through was incomplete, illustrating the accountability challenge.

The pattern across these examples: when random citizens are given time, information, expert access, and structured discussion, they produce recommendations that are more coherent, more long-term oriented, and often more ambitious than what conventional political process produces. The short-term electoral calculations that constrain politicians don't apply to citizens who aren't running for anything.

World Café was developed by Juanita Brown and David Isaacs in the mid-1990s as a format for large-group conversations. Participants rotate among small tables, discussing a shared question with different people at each table. Each table has a host who stays throughout, capturing key themes from each conversation. After several rounds, themes are harvested in a whole-group session.

The format excels at: generating collective intelligence in large groups, surfacing diverse perspectives, creating a sense of shared ownership of emerging insights. It's not a decision-making tool but a sense-making tool — most appropriate for early-stage inquiry rather than arriving at recommendations.

Open Space Technology (Harrison Owen) creates a self-organizing conference or meeting where participants create the agenda. The law of two feet — participants are expected to move to wherever they can contribute and learn most — prevents both captive audiences and disengagement. Works well for complex issues where the best questions aren't yet known.

National Issue Forums (Kettering Foundation) trains community organizations to host structured deliberative discussions on national policy issues. The materials are designed to present 3-4 genuine approaches to an issue (not a debate format with two sides) and participants work through tradeoffs. The goal is not consensus but a deeper understanding of the genuine tensions in the issue.

The Public Conversations Project — now Essential Partners — developed dialogue processes for deeply divided communities, initially working with abortion rights advocates and opponents in the wake of clinic violence. Their structured dialogue processes — which emphasize speaking to personal experience, asking genuine questions, and maintaining curiosity — have been applied to racial conflict, religious disagreement, and political polarization across dozens of communities.

Participatory budgeting is not purely a dialogue process but merits mention: it gives residents direct decision-making authority over a portion of a government budget, requiring them to deliberate about competing priorities and make real tradeoffs. Porto Alegre, Brazil pioneered it in 1989; it now operates in hundreds of cities globally, including New York City, Boston, and Chicago. Evaluation research shows participatory budgeting increases civic engagement, particularly among historically marginalized communities.

Research on deliberative mini-publics — the academic term for randomly selected citizen bodies — finds consistent results:

- Participants report high satisfaction and sense of value (typically 90%+ positive ratings) - Participants demonstrate measurable learning about the issue - Participants show reduced stereotyping of those with different views - Participants maintain engagement with civic issues after the process concludes - Recommendations are more nuanced than either elite positions or unreflective public opinion

Research on process quality and legitimacy (Fung, Curato, Dryzek) finds that when deliberative processes are well-designed and connected to genuine decision-making, they increase public trust in government — even when the outcomes aren't what individual participants wanted. The experience of being taken seriously, having a fair hearing, and seeing the reasoning behind decisions produces trust even in disagreement.

Cities and regions that have institutionalized deliberative processes — including participatory budgeting, citizens advisory panels with genuine authority, and regular deliberative polls — report higher civic engagement over time, not just in the processes themselves.

The strongest argument for deliberative democracy at the local level is not that it's morally superior to representative democracy. It's that it produces better decisions.

Representative democracy, at the local level, has significant failure modes: - Motivated minorities dominate local elections (turnout in local elections is often below 20%) - Well-organized interest groups systematically outperform diffuse public interest - Short electoral cycles create pressure toward short-term thinking - Local politicians lack political cover for difficult tradeoffs

Deliberative mini-publics address each of these. A randomly selected assembly isn't susceptible to the same motivated-minority capture. It doesn't have electoral incentives. It can take genuinely long-term perspectives. When its recommendations carry public legitimacy, politicians have cover for difficult decisions.

This is why forward-looking local governments — in places like Paris, Madrid, Edinburgh, Wellington — are institutionalizing deliberative processes alongside conventional representative structures. Not replacing representative democracy. Augmenting it with processes that produce more considered, more representative public judgment.

The town hall isn't just bad format design. It's an outdated institutional technology that was never actually good at deliberation, relying on a tradition (the New England town meeting) that had very different structural properties. The question is not how to fix the town hall. It's how to build the deliberative infrastructure that communities actually need.

Starting point for a community: a single pilot citizens panel or deliberative dialogue on one local issue — a park redesign, a budget priority, a school policy. Run it well. Document what it produces compared to the town hall. Build from there. The capacity for deliberation, like most capacities, develops through practice.