Bad immigration polls, like the one published by Public Religion Research Institute (PRRI) and the Brookings Institution that I analyzed the other day, fail because they ask narrow, "either/or", and loaded questions for a purpose: to get the result they support.
However, that is not the only tool in their arsenal. Consider another more recent (July 2014) Public Research Religious Institute survey on the border surge of illegal migrants that took place this past summer.
The second question in that survey asks respondents what to do about illegal migrants already living in the United States. It is essentially the same question they had asked before, but then had only provided two options, one of which made no sense.
In this survey, however, they provided three options: (1) "Allow them a way to become citizens provided they meet certain requirements"; (2) "Allow them to become permanent legal residents, but not citizens"; and (3) "Identify and deport them." The figures for support of these three options were 58 percent, 17 percent, and 22 percent, respectively. They are worth reporting because, unlike their earlier surveys, this one provides three, not two, mutually exclusive choices, one of which is unlikely in the extreme.
In the earlier PRRI survey, 68 percent chose the only plausible option available and that one included a "pathway to citizenship". In the July 2014 survey, when respondents were offered three choices, support for the citizenship option dropped 10 percentage points.
Note, however, the vague wording of the "pathway to citizenship" option. The wording asks respondents if they support a "pathway to citizenship" "providing they [illegal migrants] meet certain requirements."
What are those "certain requirements"? The survey never says, choosing not to list any of them. Rather, respondents are invited to believe that the requirements they might wish to see enforced for any legalization offer to be balanced and fair are the requirements the survey question is alluding to.
Moreover, the phase that introduces the option —"allow them a way to become citizens" — is also vague. Which way would that be? The question never says, and the options are numerous.
So, in effect, we have a vague introductory statement linked to a vague conditional conclusion whose elements are unspecified, producing a double dose of ambiguity.
More specific requirements would be more informative. They might also be useful in beginning to really assess where any common ground lies for breaking the current immigration impasse.
There are many ways in which such a question could be phrased. Here's one:
Allow the legalization of only those who have been in the country a decade or more, have not committed serious criminal offenses (included DWIs), are on record as having paid their taxes, and have not applied for benefits to which they are not entitled.
While the survey questions are ambiguous for the option that got the most support, there is little ambiguity in the third option: "identify and deport them." What we might term the draconian option suffers from the same all-or-nothing phrasing built on a foundation of improbable assumptions. How will the government "identify" all 11-12 million illegal migrants? Will ICE agents go door to door to find them? Will the government mail inquiries to every address in the United States? The absurdity of these suggestions only underscores the phantasmagoric nature of the options.
It's a throw-away option designed to be so extreme and improbable that that it would be hard to reasonably select, except as an affirmation of despair.