Competing polls: Does handpicked data deepen political divides?

There are many polls – but no consensus – on how Americans feel about President Trump’s immigration order.

How do Americans feel about President Trump’s controversial immigration order?

Well, it’s not so easy to say, really.

While thousands of protesters have crowded airports, city and town halls, and streets to loudly and visibly decry Mr. Trump’s executive immigration order over the past two weeks, polls have revealed varying levels of support or opposition to the orders overall. Discrepancies in polling data is nothing new, and results are generally contextualized with a degree of uncertainty. But using the data to support a policy, viewpoint, or agenda can be a polarizing, and sometimes misleading, tactic.

On Wednesday, Trump tweeted a graphic with numbers from a Politico/Morning Consult poll, showing that 55 percent of respondents approved of the order, while 38 percent disapproved. He added the quote, “Immigration Ban Is One Of Trump's Most Popular Orders So Far.”

The measure, which has raised national security and legal questions, temporarily bars green card and visa holders from seven predominantly Muslim nations including Syria, Iraq, Iran, Libya, Somalia, Sudan, and Yemen as well as other refugees from entering the United States. A federal Seattle judge has temporarily halted the orders, and an appeals court is reviewing that decision.

Other recent polls seemingly contradict the findings tweeted by Trump, creating dissonance in gauging public opinion. Yet, just as the president has chosen to cry “fake news” when confronted with reports he views as opposition to his agenda and allowed those in the administration to provide “alternative facts” when discussing his inauguration crowd size, Trump’s tweeted poll reflects a tendency to promote information that confirms his success at the exclusion of evidence of any unpopularity.

Public opinion polling is far from a perfect science, as a majority of pollsters proved during the 2016 election after data wrongly predicted that Hillary Clinton would win the race. Misleading wording or even constructions that lack proper nuance for complex issues and scenarios can create discrepancies in responses to questions that are similar at their core. Timing, question order, and response choices can also lead to very different results. Conducting polls regarding policy proves even trickier than trying to gauge whether a candidate will win an election, adding another layer of uncertainty to results surrounding the immigration orders.

Still, polls can show how Americans feel about policies, which in turn can inform leaders and allow them to make decisions that more closely reflect the will of the people. But if politicians handpick poll results that affirm their own agendas to the exclusion of contradicting data, the findings can further polarize the already divisive political climate, says Kirby Goidel, a communications and public policy professor at Texas A&M University.

“The polls themselves are not the danger,” Dr. Goidel tells The Christian Science Monitor in a phone interview. “The danger is their misrepresentation by politically interested parties who are trying to use them in ways that they shouldn’t be used, and attempting to provide a level of certainty around an issue where there is probably not a level of certainty.”

Recent polls from Gallup, CBS News, Quinnipiac University, and CNN/ORC have all showed that between 51 and 55 percent of respondents are opposed to the immigration orders, contrasting the Politico/Morning Consult poll. All except for the Politico and Quinnipiac polls were conducted prior to the temporary restraining order instituted last Friday.

The Quinnipiac poll found that 50 percent of voters oppose "suspending immigration from 'terror prone' regions, even if it means turning away refugees,” while 44 percent support the measure. That revelation contrasts the findings of an early January Quinnipiac poll, taking before Trump’s inauguration. Then, voters supported the measure by 48 percent, with 42 percent opposing.

These results come even after 70 percent of American voters said it is "very likely" or "somewhat likely" that there will be a terrorist attack in the United States resulting a large loss of life, according to the poll. A majority agreed, however, that homegrown terrorists were more likely to perpetrate mass, organized attacks than immigrants or refugees, as a recent study examining post-9/11 attacks shows.  

"Message to President Donald Trump: 'Give me your tired, your poor, your huddled masses yearning to breathe free' still has profound resonance with Americans," Tim Malloy, assistant director of the Quinnipiac University Poll, said in a statement. "Significant pushback on immigration tells the president that many voters are not on board with a ban on refugees and that voters are strongly opposed to holding back those most threatened, Syrian refugees."

And while Democrats in opposition to the orders as well as protesters who have gathered nearly daily in cities and towns across the nation to decry them as un-American would likely take that view, it’s just as easy for the Trump administration and its supporters to use the Politico poll to argue for their agenda.

Another poll from public polling agency Rasmussen Reports also found a majority supported Trump’s order after posing the question: “Do you favor or oppose a temporary block on visas prohibiting residents of Syria, Iraq, Iran, Libya, Somalia, Sudan and Yemen from entering the United States until the federal government improves its ability to screen out potential terrorists from coming here?”

By placing an emphasis the possible presence of terrorists, or as Quinnipiac did, on Syrian refugees who have been caught in a widely reported crisis, pollsters can see variation in results.  

“Even really minor question wording changes can lead to very different responses by citizens,” Rosalee Clawson, head of the political science Department at Purdue University, tells the Monitor. “That’s not liberal bias or conservative bias, that’s just a fact of polling.”

Still, those answers matter, even when taken with an eye to uncertainty.

“We do get a sense of whether policy is following along with public wishes, even with the squishiness of public opinion,” Goidel says. “[We see] where people are in a democracy and if their wishes are being followed.”  

For citizens who want the full picture, it’s important to pay attention to a question’s context and when the poll was taken, experts say. Many respondents are swayed by the last information they heard on an issue and could poll differently depending on the day and current news.

But boiling a complicated issue down to just one of several polls and portraying its vast findings in a simple graphic can create more division, Goidel and Dr. Clawsons say. Contextualizing public opinion is crucial to understanding how polls derived their data, and what portions of a complex order are resonating with the public.

Trump's tweet, on the other hand, lacks that nuanced approach, Clawson says.

“The kind of way in which [Trump] has used technology to go around the media, that’s the new phenomena. By using Twitter, this is a technological way to communicate directly with citizens where you don’t have journalists acting as a filter,” she says. “With Twitter, he does have a platform there were he can pick one poll, and put it out. Citizens, if they want to be informed, need to go beyond that.”