Think critically, think depth when reading about polls

If it’s election season — and it is, with the March 18 primary barely more than 60 days away — it must be time for a public-opinion poll, and with it, this warning: reader beware. Much is written and said about the reliability of public polling and the effects of reporting on public polls. Whether the results involve a candidate or an issue, they can suggest a body of support or opposition that sways many people to a certain way of thinking or diminishes their willingness to advance their own point of view or sometimes even to vote.

For this reason, the Daily Herald takes great care in deciding whether to report stories about polls, especially during election campaigns, and particularly works to provide context when we do report on them. But poll aversion shouldn’t apply just to political candidates. It also extends to broader issues on which you may be forming an opinion. And here, the issue isn’t so much that you should be skeptical as that you really must read beyond the headline.

The Illinois Policy Institute, a conservative foundation monitoring political trends in the state, demonstrated this maxim vividly last week with a description of one of its own polls. In a statewide survey on attitudes toward the state income tax, the group emphasized the vast difference that can occur in results depending on how a question is worded. When, for example, people were asked whether the state should move to a “graduated income tax that would lower tax rates for 90 percent of Illinois taxpayers while increasing rates on higher earners,” 57 percent said yes. But when the same people were asked simply whether they support changing to taxation according to a “graduated rate that requires a higher percentage paid as your income increases,” support fell to 33 percent.

So, do you believe the latter, IPI-preferred language or the former, “push left” language, as IPI describes it? That may depend on your own political leanings, but the really important conclusion is that you should know the question when you’re evaluating the merits of a public-opinion poll. On this topic, you might well see a headline or statement claiming, “Poll finds 57 percent back graduated income tax” or stating “Poll finds 33 percent back graduated income tax.” Partisans may use one approach or the other in their debates, essays and letters to the editor. And both statements would be accurate as far as they go. But you must go further in order to really understand the story.

A similar discrepancy is emerging in polls about the minimum wage. Depending on the wording some polls have found both strong support and solid opposition for raising the ceiling on basic pay. If you want to understand the truth, you need to look deeper.

And that includes considering not just how questions are worded, but who worded them and who answered them. A poll being floated this week found that 95 percent of respondents are more likely to vote for candidates who have “a strong record on improving opportunities for people with disabilities.” Respondents to the poll, conducted by an organization for people with disabilities, were themselves people with disabilities. So, on that particular question, you’re not likely to be too surprised by the result. But you may find other questions and responses more valuable precisely because of the population polled.

It’s all a matter of detail and context. Whether you’re seeing a poll about who’s ahead or behind in an election campaign or one about an issue elected officials may vote on, don’t necessarily ignore the results. Just be sure to dig into them.

Jim Slusher, jslusher@dailyherald.com, is an assistant managing editor at the Daily Herald. Follow him on Twitter and Facebook.