Case Study 7.2: Writing the Immigration Questions — When Every Word Is a Political Decision

Case Study 7.2: Writing the Immigration Questions — When Every Word Is a Political Decision

The Challenge

Among the twelve policy questions in the Meridian/Garza-Whitfield survey, the immigration battery was by far the most difficult to write. Carlos Mendez had spent two hours on what he expected to be a thirty-minute task. He had four drafts open simultaneously, each with different question wording, and he couldn't decide between them.

The problem was not ignorance — it was the opposite. He knew too much about how every word would land differently with different respondents. He knew that "illegal immigrant" activated different associations than "undocumented immigrant." He knew that "border security" was a Republican frame and "comprehensive immigration reform" was a Democratic one. He knew that asking about enforcement first would prime enforcement-friendly responses to subsequent questions. Every choice felt like taking a side.

He brought all four drafts to the Tuesday meeting.

Draft 1: The Republican Frame

"How important is it to you that the federal government secure the border and enforce immigration laws? Very important, somewhat important, not very important, or not important at all?"

Carlos had found this wording in a Heritage Foundation survey from two cycles back. It produced very high "very important" numbers — consistently 65-70% in conservative-leaning samples.

Trish identified the problems immediately: "This question is so broad that it doesn't distinguish between someone who wants a physical wall and someone who wants better visa tracking. It's also entirely enforcement-framed — there's no acknowledgment that any other aspect of immigration policy exists."

More fundamentally, the wording "secure the border" is a phrase that appears in Republican campaign ads. Using it in a neutral poll borrows that frame.

Draft 2: The Democratic Frame

"Do you support or oppose creating a pathway to citizenship for immigrants who are already living and working in the United States?"

This question, Carlos noted, is structurally neutral — it's a simple support/oppose question without evaluative language. But it's entirely focused on the legalization element of immigration policy, with no acknowledgment of enforcement. It would predictably produce majority support — national polls on this framing consistently show 60-70% support — without capturing the significant public demand for enforcement-first policy.

"Using this question alone would make the state look much more pro-legalization than it actually is," Vivian said. "And that would be misleading to our client."

Draft 3: The Academic Frame

"Which of the following comes closest to your view on immigration policy? The government should: (a) increase the number of legal immigrants allowed into the country each year, (b) keep the current level of legal immigrants, (c) decrease the number of legal immigrants allowed into the country, (d) have no restrictions on immigration."

This question, drawn from the General Social Survey, asks about the level of legal immigration — an analytically important dimension, but one that is somewhat removed from the emotional and political center of current immigration debate, which focuses more on border crossings and undocumented status.

Vivian recognized it as a technically clean measure. "But it misses the unauthorized population question entirely. And 'no restrictions' is a position held by virtually no one — it inflates the apparent share of 'moderate' views because people who want some restrictions cluster in the middle three options."

Draft 4: The Forced-Choice Frame

"Thinking about immigration policy, which of the following comes closer to your own view? (A) The most important priority should be stronger border security and stricter enforcement of immigration laws. (B) The most important priority should be creating a path to legal status for immigrants who are already here. (C) Both are equally important."

This was the version the team had ultimately included in the questionnaire. Carlos thought it was the best of the four — but he wanted to understand why.

The Resolution: Why Draft 4

Vivian walked through the reasoning:

It presents both major frames. Unlike Drafts 1 and 2, which each represented only one side's frame, Draft 4 places enforcement and legalization in explicit tension. This reflects the actual political debate rather than privileging one dimension.

The "both equally important" option does real analytical work. When 48% of respondents say both are equally important, that's actionable information: it tells the Garza campaign that either/or framing is likely to lose more votes than it gains. Draft 1 and Draft 2 can't produce this insight.

It doesn't use inflammatory language. "Stronger border security" and "path to legal status" are less loaded than "secure the border" or "amnesty for illegal immigrants." They're not neutral — no immigration language is — but they're defensible as reasonable descriptions of the policy positions.

It's a priority question, not an ideological identity question. Asking about "priorities" rather than "support or oppose" acknowledges that many respondents agree with both goals in principle; the question is about which they believe government should focus on first.

Carlos pushed back on one element: "What about the word 'immigrants' rather than 'undocumented immigrants'? Are we losing precision?"

Vivian: "Yes. But specifying 'undocumented' changes the question. Many respondents who support a path to legal status for undocumented immigrants don't support it framed that way — because 'undocumented' activates law-breaking associations. We're measuring a politically real position: people who want legalization for people who are already here, whatever they call them."

The Broader Lesson

Carlos left the meeting understanding something that no methodology textbook had quite articulated. Every survey question is a political act, not because questionnaire designers are manipulating results, but because language in the political domain is inherently loaded. There is no neutral vocabulary for immigration — or for healthcare, race, taxation, or any other contested policy domain. Every word choice activates certain considerations and suppresses others.

The professional obligation is not to find nonexistent neutral language. It is to:

Understand the political connotations of the language you use
Make deliberate choices that serve your measurement purpose
Disclose those choices in your methodology statement
Consider whether split-sampling different wordings would yield more complete information
Never claim that your results are free of framing effects when they are not

Trish put it practically: "If someone asks me to defend any question in this questionnaire, I can tell them exactly why it's worded the way it is and what tradeoffs we made. That's the job. You don't have to be neutral. You have to be transparent."

The Results and Their Interpretation

When the immigration data came back, Draft 4's results were illuminating:

32% enforcement priority
20% legalization priority
48% both equally important
(Among Latino respondents): 18% enforcement, 35% legalization, 47% both equally

The Garza campaign's instinct had been to run hard on a pro-legalization message in the southern metro area. The data suggested this would activate the 35% of Latino respondents already there but risk alienating the 47% who wanted a balanced approach. Nadia Osei revised the messaging to emphasize a "secure and humane" frame — signaling both enforcement competence and a compassionate legalization pathway.

Would the Draft 1 or Draft 2 wording have led to the same strategic insight? No. Draft 1 would have made enforcement look dominant; Draft 2 would have made legalization look achievable without addressing the plurality preference for balance. Draft 4 told the Garza team something they needed to know.

That is what good question design does.

Discussion Questions

Carlos's four drafts reflect four different implicit theories of what immigration opinion is "really about." What theory does each draft embed, and how does that theory shape the results each would produce?
The team chose the forced-choice format (Draft 4) over the more conventional agree/disagree format (Drafts 1 and 2). What would a methodological purist say is the cost of this choice? What is the benefit?
"Both equally important" captured 48% of respondents — making it the plurality response. If a questionnaire designer wanted to avoid a dominant "both" response, how could they redesign the question? Is this a legitimate methodological goal, or a form of manipulation?
Vivian argues that "undocumented" would change the question in politically meaningful ways. Do you agree with her reasoning? When is terminological precision more important than minimizing loaded language?
If the Garza campaign had commissioned this survey, should that be disclosed? Does the identity of the survey's sponsor affect how we should interpret the question wording choices? Why or why not?