A week and a half ago, Arizona Governor Brewer signed the nation's toughest--and most
controversial--immigration law. The law, recently revised, has generated daily news and analysis. But the public surveys emerging do not yet provide a complete picture of public opinion. (For a comprehensive, but clearly-worded summary of the bill, the National Conference of State Legislatures has one here.)
This Rasmussen survey drastically summarizes the Arizona law to only one of its provisions, and finds clear majority support (60%). Nate Silver critiqued it further here. Last week's Gallup survey doesn't even describe the bill at all. It shows voters who self-report reading or hearing something about the bill supported it (51% favor/39% oppose) . Without a bill description in the question, voters are responding to what they think they think the bill contains, and so likely have widely divergent perceptions of it. Chuck Todd and Media Matters criticized that survey over the weekend. Despite their flaws, both polls have informed subsequentmedia coverage.
This online Angus Reid poll does examine some of the individual components of the Arizona bill. Putting aside potential objections to online methodology, these questions at least describe the bill in some detail. However, it lacks a single question on the entire Arizona bill, and leaves out some key provisions, like making it easier to sue the state for insufficient enforcement. Despite these differences, the Angus Reid poll shows widespread support for tougher restrictions on immigration, much like the Rasmussen and Gallup surveys.
But to get a full picture of national attitudes, a survey should test supporting and opposing arguments to the bill, and see how, if at all, the arguments change voters' opinions. The goal should be to explore how a protracted national debate on immigration policy might affect voters' views, as well as measure the importance of questionnaire wording and policy details. For example, how do voters evaluate the costs of implementation? What about questions about racial profiling, or judgments based on clothing or shoes? How would support for Arizona's bill compare to support for moderate, yet comprehensive, federal legislation? As Tom Schaller wonders here, do poll respondents simply react to something sweeping being done?
These are all important research questions as we continue the national conversation
on immigration. As always, it's important for media outlets, bloggers and pundits to examine questionnaire language before taking a poll's results at face value. To describe national attitudes based only on the post-Arizona polling so far would be a mistake.
CORRECTION: The original version of the post incorrectly identified results from the Gallup poll collected among all adults as representing those who had not heard about the bill.