Why We Can’t Trust Polls On Immigration, In One Horribly Xenophobic Chart

While polls consistently show that Americans overwhelmingly favor the tech industry’s #1 legislative goal, comprehensive immigration reform, I wondered if I could also get them to agree to a horribly xenophobic plan that was the opposite of the proposed bill: deporting all 11 million undocumented immigrants. We conducted a CrunchGov poll with Google Surveys to find out and the results, you’ll find, are surprising.

First, remember that Americans like to agree to things: they say yes to bigger cars, more debt, and, of course, more croutons on their all-you-can-eat Olive Garden salad. So, when pollsters ask Americans if they “support” an idea, the answer is, on average, yes.

In our poll, a majority of all Americans (53%) and a whopping 74% of Republicans want to kick out every undocumented immigrant (details here*). Bear in mind, no one, not even the most anti-immigrant members of Congress, are proposing any remotely close to this idea. At most, Congressmen disagree over whether current undocumented immigrants should be permitted to become citizens, not whether they can stay in the country.


More importantly, the results from our on-going CrunchGov Poll find that roughly the same percentage (64%) want to grant a pathway to citizenship as kick them all out. Confused? You should be, because both of those views are diametrically opposed.


Survey methodologists lovingly refer to this as the “forbid/allow asymmetry“, where respondents are more likely to support ideas that “allow” action, rather than forbid it, even if the outcome is the exact same.

As I’ve written before, “In 1989, a poll found that far more respondents seemed to support interracial marriage when they were asked whether the government should “allow” the marriage (32%) vs. “forbid” such marriage (19%), even though, legally speaking, it’s the exact same. The polling irregularity, known as the “forbid/allow asymmetry,” showed how seemingly innocuous changes could cause a big difference in the outcome.”

Surveys are a tricky beast, and seemingly innocuous difference can result in massively different outcomes. Congressional leaders know this, which is why playing games with surveys is often ineffectual in convincing them to support an idea. Instead, let’s stick to arguing the merits of a plan; it’s more honest–and definitely more reliable.

*Methodology notes: poll was conducted using Google Surveys of 1,000 respondents. Results are currently finishing up, but numbers holding steady and are statistically significant. For more details on why we use Google Surveys, read here.