What Happens When a Group of Strangers Spends a Day Debating Immigration?

About twenty-five minutes into a small group discussion of undocumented immigration, in Houston, something changed. The group was taking part in an experiment in “deliberative polling,” which measures the impact of discussion and information on public opinion. The conversation among the eighteen people in the room had hit all the standard talking points and deployed all the usual stereotypes of America’s current immigration debate: the law is the law; we can’t let everyone in; immigrants are a drain on resources; “they” have made “our” lives unreasonably difficult. Only three people in the group were white—all three happened to be men—and two of them made the first two statements in the discussion. The first question put to the group was whether “viable legal options should be provided for beneficiaries of the 2012 Deferred Action for Childhood Arrivals program”—the Dreamers. The first white man to speak said, “If you are here illegally, why should you benefit from any of the resources?” The second white man to speak agreed with him. Then a Latina woman suggested that punitive actions toward DACA recipients were expensive and unjustifiable. Then all hell broke loose, with people talking over one another.

The third white man, who said that in his retirement he was working as a substitute teacher, complained about immigrant children. “I can’t get them to assimilate,” he said. “They don’t want to learn English.” A middle-aged African-American woman, who said that she was a teacher, added, “We have to get E.S.L. certification, and I don’t think that’s fair!” A young African-American woman said, “They are breaking the law! There is something going on in their country. It’s been going on for a long time. That’s where you were born—you should be where you were born.”

This was the first session of the day-long exercise on undocumented immigration, and it made me dread the rest of the day. James Fishkin, the Stanford political scientist and communications scholar who had organized the event, held in June, on the campus of the University of Houston, had dreaded it, too. Fishkin has run a hundred and ten deliberative polls since he pioneered the format, in 1994. Each has involved gathering a representative sample of a population—of a city, state, or nation—for a day or a weekend, usually in person but on a few occasions online, to discuss an issue in an orchestrated sequence of small group and plenary sessions. Before the discussions start and again after they end, participants fill out detailed questionnaires about their views on the issue at hand.

Fishkin told me that a deliberative poll had made him this nervous only once before: in 2007, when he conducted an event on the rights and lives of Roma people in Bulgaria. The experiment ultimately appeared to show that more non-Roma Bulgarians were more accepting of Roma than expected, but the conversation had hinged on the participants’ realization that there were Roma among them. The Houston sample included registered voters only—there would be no undocumented immigrants among the participants. Fishkin couldn’t count on the Bulgarian effect.

The first group discussion I sat in on, with the three white men, seemed to bear out all of Fishkin’s fears, nearly going off the rails. But then it suddenly righted itself. An African-American man in his late thirties said, “Give me your tired, your poor, your huddled masses yearning to be free. That’s what it says on a sign in Ellis Island.” He was a little off on the location, but he got the attention of the room. “Our problem is we have made into a legal issue something that should be a humanitarian issue. Most people who are coming now, they are showing up at the border asking for asylum. They are following the law.” Almost immediately, there seemed to be a perceptible shift toward compassion, both toward undocumented immigrants and toward the other people in the room.

Before attending the event in Houston, I had heard Fishkin talk about his experiments, and I had heard other people, including the Harvard political scientist Jane Mansbridge, praise Fishkin’s approach to creating political conversation. In the format that Fishkin developed, each participant receives a folder with “briefing documents,”which contain concise summaries of arguments on different sides of every question, and basic fact sheets. Proponents of opposing arguments sign off on the documents: each side considers them accurate and balanced. There is no guarantee, of course, that participants will read the documents. Moderators are instructed to suggest looking up facts in the folders whenever a factual dispute or a factual inaccuracy arises, but moderators are not supposed to correct any inaccuracies themselves. They are, in fact, trained not to signal any authority: never to stand nor to speak louder than the participants. They are specifically cautioned not to take any straw polls or push for consensus on any position: Fishkin’s goal is to enable people’s views to evolve, avoiding anything that might have an “anchoring effect.”

Some deliberative polls last an entire weekend; the Houston event was just a day long, and packed. Participants filled out their initial surveys, then spent seventy-five minutes in a group discussion, then attended a seventy-minute plenary session, broke for lunch, and had another seventy-five-minute discussion and seventy-minute plenary session. Groups, accompanied by their moderators, stayed together throughout the day, and their joint project was to create one or two questions to ask experts during each plenary session. In Fishkin’s design, the experts do not lecture: they are there solely to answer questions posed by the ordinary people who are deliberating. The groups then met for another forty-five minutes of summarizing, venting, or tying up loose ends, and participants filled out the survey again. An optional dinner followed.

When Fishkin spoke about people working in small groups and “assemblies,” I envisioned grand amphitheatres in which people conducted measured, dignified discussions. What I found was a windowless hotel ballroom and a bunch of faceless conference rooms. Most of what I thought I saw was people talking past one another, and never going into any depth. There were sixteen groups altogether. During the moderators’ lunchtime debriefing session, the moderator of Group Thirteen, whose members included two men wearing red Make America Great Again hats, complained that she had been having issues with the “volume of the conversation, tone, and gestures.” I sat in on Group Thirteen after lunch. The questions under discussion concerned the economy and the workforce: Should the E-Verify system, which checks potential employees’ eligibility to work, be mandatory for all private businesses? Should various worker-visa programs be expanded? Of the fourteen people in the group, seven were African-American, six white, and one Asian. The white men dominated the conversation here, too. Positions, however, didn’t break down along the lines I expected. The younger of the red-hatted men played the pragmatist among conservatives.

“The idea of stealing jobs is minimal,” he said. He was first to speak on the topic.

“How many of you used to mow lawns when you were younger?” the older MAGA man objected. “So they are taking jobs.”

“So you are saying they are taking fourteen-year-olds’ jobs?” the younger man asked.

“But my brother cleans yards,” an older African-American woman said. “He’s retired . . . We keep reiterating that they [immigrants] take less money [for the same job]. That’s not true. My sister works at a company where she is the only black. Everyone else is not from here.”

“If we don’t let people work, we are creating crime,” a middle-aged African-American woman argued.

“I used to hire,”the older woman said. “I know for a fact that they were putting down ninety-nine dependents so they wouldn’t have to pay no taxes. They were saying, ‘I’m sending money back home.’ ”

“This is Mexico’s second biggest source of income,” a young white man said. “We are essentially propping up Mexico with our jobs.”

“Big corporations have to be held accountable for their exploitation of people,” the young Asian man said.

It seemed like something shifted again. Like the Emma Lazarus quotation in the morning group, the remark laying blame on corporations seemed to take the conversation to a different level, uniting the participants and creating potential for some solidarity with immigrants.

“What they are doing is perfectly legal,” the young MAGA man, now apparently referring to the employment of foreigners on work visas, at lower wages than a citizen might expect, said. “You can’t just end it.” He sounded not resentful but resigned.

“They don’t want to pay citizens more,” the older African-American woman said.

“It’s the corporate lobby on immigration,” the young Asian man summed up.

People nodded. At this point, they had a common enemy, and it wasn’t immigrants.

Plenary speakers came in pairs: for each area of expertise—legal options, higher education, crime and public safety, work and economy—organizers engaged one more-conservative and one more-progressive voice. With a couple of minutes per question, this too was far from the slow, in-depth discussion I had somehow imagined. But even at sound-bite length, the plenaries packed in a lot of information. The experts dispelled the myth that undocumented immigrants don’t pay taxes. In fact, many pay more than their share—because they don’t file tax returns and therefore don’t claim refunds and, more significantly, because they pay into the Social Security system, which most will never be able to use. The audience learned that the Trump Administration has suspended Temporary Protected Status programs and halted the program that allowed people to gain citizenship by enlisting in the military.

Toward the end of the first plenary session, one of the experts, Charles Foster, an immigration lawyer who served as an adviser to George W. Bush, during his 2000 campaign, and to Barack Obama, in 2008, did what he wasn’t supposed to do. “I’ll make a political statement,” he said. “The political rhetoric is so bad that there can be nothing passed today because they will be smeared with the A-word, and I don’t mean ‘adultery,’ I mean ‘amnesty.’ I talk to members of the Republican Party. They are good people, but they can’t do anything because they will be unseated by somebody who uses the word ‘amnesty.’ ”

During the second plenary, the Houston police chief, Art Acevedo, who came to the United States from Cuba as a child, served as one of the experts. He made no excuses about making a political statement. “Both parties are using this as a wedge issue, because if they solve it, they’ve got nothing to talk about to rile you up,” he said, to thunderous applause. “Let me tell you something else. We are all, here in this room, the hired help. And they are pitting us against one another. And God is watching.” People applauded, screamed, and jumped to their feet.

A few minutes later, participants were in small groups again, for their final discussion of the day. I sat in on one of them. The conversation kept coming back to resentment—and it was resentment of immigrants that seemed to unite people in the room across race, class, and education level. I left the session dejected. But when I looked at my notes a couple of weeks later I saw something else. A middle-aged white woman complained that her husband couldn’t find a job, and she blamed immigrants. Another middle-aged white woman said that she couldn’t get hired because she had a felony conviction, while immigrants “don’t have documentation” and therefore don’t have to disclose their criminal records. But interspersed among these expressions of resentment there were, it seemed, statements of solidarity and even hope.

“There are a lot of reasons someone can’t get a job that don’t have to do with immigration,” a young African-American man said.

“We don’t want to blame ourselves,” a young white man said. “We blame the out-group. And that prevents us from thinking about immigration the way we should, and that prevents us from fixing immigration.”

“I learned a lot today,” a middle-aged African-American woman said.

“We ought to all resolve to stop this immigration issue from dividing us,” an older white man said.

The surveys confirmed my impression. “It’s obvious that there were a lot of significant changes of opinion,” Fishkin said after reviewing the results, which he made available to me. “And they were all in the direction of the rights and interests of undocumented immigrants.” At the end of the event, more participants believed that undocumented immigrants should have legal options for staying and working in the country, that they should be able to pay in-state tuition rates, and that they should be provided with public counsel during immigration hearings. The most significant changes concerned police: 24.4 per cent more people came to oppose the practice of asking crime victims about their immigration status, and 17.6 per cent more now opposed asking crime witnesses whether they were in the country legally. Participants also gave significantly more correct answers to a set of seventeen “knowledge questions.”

The results were consistent with Fishkin’s experience: deliberative polling appears to shift people’s opinions. In several cases, Fishkin’s team has checked back with participants about a year after the experiment and found that, while some had reverted to their old positions, over-all the effect held. But in Houston, Fishkin got one surprising, anomalous result—a seven per cent decrease in the number of people who agreed with the following statement: “People with views very different from mine often have good reasons for their views, even when they are wrong.” On most topics, being exposed to people making arguments from the other side makes people more sympathetic, at least, to the possibility of a different position. “This stands out to me,” Fishkin said of the result. “This is really a hard issue.”

As Fishkin sees it, deliberative polling is a superior alternative to standard measures of opinion. “Normally, public opinion represents what people think when they don’t have a lot of time to think about an issue and they are talking to people like themselves,” he told me. Standard quantitative methods collect superficial responses—while standard qualitative methods place people in focus groups, in conversation with people who share their backgrounds and, often, views. Fishkin’s approach creates conversation across difference. In locations as diverse as Mongolia and the state of Texas, deliberative polls have also been used to make decisions—a thoughtful, interactive alternative to a referendum.

Fishkin likes to describe the results of deliberative polling as “what the people think under good conditions.” What struck me in Houston was that the conditions weren’t that great: the groups were large, the conversation often felt hurried and superficial, and the experts failed to maintain a posture of neutrality. Still, it turns out that spending a day in a generic hotel, engaged in often frustrating conversation with strangers, makes for better informed and more compassionate citizens than what passes for political conversation the rest of the time.