RALEIGH – Sometimes the conventional wisdom isn’t wise.

That’s not to say that one should never pay attention to the prevailing wisdom, or to public opinion as measured by credible polling organizations. But whether you are a politician or an advocate for a particular point of view, polling should never be considered the last word on anything. If the public agrees with you, that doesn’t necessarily mean you’re right. If they disagree, that doesn’t mean you’re wrong. The reason it’s important to know what voters think about issues and why is because such information can help you frame persuasive arguments, set priorities, and make choices about investments of time and money.

Consider the case of education policy. There are few issues voters say they care more about, particularly at the state and local level. Unfortunately, many voters know surprisingly little about the details of proposed policies. They have tendencies and preferences, but the more specific the question, the more likely it is that many voters truly have no strong preferences, having never before thought about the matter in question. You may get an answer, but it doesn’t represent a deeply held belief. Ask the question with different words and emphasis, and the answer may be different, too.

That’s not an argument against polling the public on education. It is an argument for interpreting the results carefully.

I was struck, for example, by several of the findings of a poll commissioned by the Harvard Program on Education Policy and Governance and the journal EducationNext. For example, there is fairly widespread agreement among education-policy researchers of all ideological stripes that teacher quality is probably the most important variable affecting school effectiveness. While average class sizes have been shown to have a modest effect on performance in some well-designed studies, teacher quality has a much-greater effect.

Based on the evidence, it would be reasonable to conclude that scarce resources should be devoted first to recruiting and retaining better teachers, and secondarily to reducing the number of students exposed to the average teacher. The public doesn’t agree, however. Asked whether it is a better use of educational dollars to improve teacher compensation or decrease class size, poll respondents voted overwhelmingly (77 percent to 23 percent) for class-size reduction.

To choose another example where there is a fairly broad consensus among education policy analysts: while most of the experts believe that school districts ought deploy more of their resources towards differentiated pay to attract teachers to fields where there are severe shortages, such as math and science, only 33 percent of respondents agreed. Two-thirds said they would rather use the same money to give smaller salary increases to all teachers equally.

It’s less surprising, but still significant, to see that on contentious issues such as parental choice in education, a large number of voters not only lack familiarity with the relevant research findings but also the basic information they need to offer an informed opinion. For example, while a plurality of respondents in the Harvard/EducationNext poll supported charter schools (44 percent), almost as large a share had no strong opinion on them (42 percent). Perhaps that is because only 13 percent of respondents knew that charter schools cannot be religious and only 24 percent knew that charters don’t charge tuition.

All sides of the education-policy divide ought to tread cautiously here. Daily Journal readers know that I strongly support parental choice measures such as tuition tax credits and tax-funded scholarships for low-income students to use at any school, public or private. While a majority of respondents endorsed tax credits (53 percent for, 25 percent opposed) and a plurality endorsed private-school scholarships (45 percent for, 34 percent opposed), more than a fifth of respondents didn’t offer an opinion one way or the other. That’s a good chunk of folks who, on any given day, may vote for or against a candidate or ballot proposition depending on how the issue is worded and explained to them.

To have a productive public debate on education, the public needs more education so they can make informed judgments.

Hood is president of the John Locke Foundation.