Americans support a border wall more than the media wants to believe

As Americans debate the necessity of a border wall — or lack thereof — it’s not surprising if they’re confused. President Trump hasn’t always been exacting and consistent in expressing his vision. Yet a lot of confusion is being caused by media reports.

Take the polls.

Reporting on polls typically implies America is soundly against a border wall. “Sorry, Donald: Pew Poll Finds Large Majority Oppose Border Wall,” Mediaite wrote in April 2016, reporting on the results of a Pew Research poll.

{mosads}Last February, Pew reported similar findings: 62 percent of Americans oppose building a wall. Only 35 percent support it.

But are we telling the whole story?

First, it’s worth looking at what Pew asked: “All in all, would you favor or oppose building a wall along the entire border with Mexico?” To me, it’s a confusing question. After all, there already is a wall or fencing along approximately 700 miles of the southern border. It might make more sense to ask, “Would you favor or oppose building a wall along the remaining, unwalled portion of the border with Mexico?”

Second, why ask about something that’s not under consideration: a wall along the “entire” border? If you think President Trump favors such a thing, that, too, might be blamed on confused reporting.

In July, President Trump told reporters, “It’s a 2,000-mile border, but you don’t need 2,000 miles of wall because you have a lot of natural barriers.” News outlets reported that as if they were hearing it for the first time.

“President Donald Trump doesn’t see the need for a proposed border wall to stretch the length of the roughly 2,000-mile frontier with Mexico,” the Associated Press reported. 

“Trump says wall may not need to cover entire U.S. Mexico border,” echoed Reuters.

“No need for a full border wall, Trump says,” wrote the Los Angeles Times.

The New York Daily News called it a “radical departure” on Trump’s part.

BuzzFeed said Trump had suddenly “scaled back” his wall.

They didn’t do their homework.

Trump has actually been making similar statements for a long time. In February 2016, he told MSNBC, “Of the 2,000, we don’t need 2,000, we need 1,000 because we have natural barriers etc. etc.” He repeated himself in March, and in November, he told Sixty Minutes that part of the border barrier could be fencing rather than a wall.

It turns out Trump even said it early in his candidacy. “You also have natural terrain which is automatically a barrier, which is a good thing,” the Republican presidential candidate told Fox Business Network’s Maria Bartiromo on August 20, 2015. “So you’re talking about a thousand… a little more than a thousand miles.” 

It was Pete Saenz, the Mayor of Laredo, Texas, who recently alerted me to the fact that Trump has long talked of building only a partial wall. (I was interviewing the mayor for this Sunday’s edition of Full Measure.) Saenz told me that Trump formulated some of his ideas during a July 2015 visit to Laredo.

“[Trump] asked about how I felt about the wall. I told him with all due respect it was offensive to Mexico,” Saenz told me. “He listened…he did give an interview at the end basically saying that maybe areas with natural barriers like the river possibly may not take a wall.”

Polls have been consistently unrepresentative of the electorate since Trump entered the political stage

While we’re in the weeds, assuming there’s value to asking a poll question about something that nobody is proposing, there’s additional nuance to consider. Pew ended up with a Democrat-heavy sample: 38 percent Republican/Republican leaning and 52 percent Democrat/Democrat leaning. The 14 percentage point difference means Pew interviewed 38 percent more Democrat thinkers than Republican thinkers. I can’t find any estimate that says the actual U.S. population is politically lopsided along those lines.

Other polls, such as ones exploring the president’s popularity, have similar political breakdowns. In fact, of the polls I looked into during campaign 2016, they always interviewed more Democrats than Republicans; never the other way around. This is partly what led me to predict Trump would win the race for president despite nearly every official and unofficial prediction in the media.

“Of course, there’s no way Trump will become president,” an opinionist named Sally Kohn declared on CNN in June 2015. “Polling suggests Trump is, in fact, the most disliked political candidate since 1980.”

Skewed polls and irrelevant questions are a weak way to gauge public opinion

The pollsters who took my calls during the campaign were forthright and transparent in answering my questions. They have a tough job. They rely on samples that are scientifically sound but, to an outsider like myself, seem shockingly small: often 1,000 people or less. They apply a lot of magic weighting and adjustments that sometimes takes pages of disclosures to explain, in order to try to make sure those 1,000 people somehow reflect the beliefs of 323 million Americans. It’s the pollster’s clients—often news organizations—that may be responsible for one-sided or incomplete reporting of results.

There are two things we could do to provide more meaningful reporting. First, when addressing polls on political topics, we should disclose the breakdown of Democrats and Republicans upfront. To state the obvious: findings from a sample that’s made up of 98 percent Republicans will be entirely different than findings from a sample of 98 percent Democrats. How can meaning be put behind results on any political topic without the partisan makeup of the sample being considered?

Second, our reporting could include opposing findings and trends, if they exist. For example, in the most recent Pew poll, “three-quarters (74 percent) of Republicans and Republican-leaners supported a border wall” and that support had grown substantially in recent months. Conservative Republican support for a wall was up nine points since Trump was elected President (from 71 percent to 80 percent). 

Support also grew among moderate and liberal Republicans (from 51 percent to 60 percent). An accurate headline could just as well have been: “Poll shows growing Republican support for a wall under a Trump presidency.”

All things considered, I came up with my own headline that’s more transparent than many of the ones I saw: “In polls with Democrat-heavy sampling, there’s overwhelming opposition to building a wall along the ‘entire’ border; a concept that nobody is, in fact, proposing.”

Sharyl Attkisson (@SharylAttkisson) is an Emmy-award winning investigative journalist, author of the New York Times bestsellers “The Smear” and “Stonewalled,” and host of Sinclair’s Sunday TV program “Full Measure.”