Albert Cheng, Michael B. Henderson, Paul E. Peterson and Martin R. West:

Education’s political landscape has shifted dramatically over the past year. To the consternation of most school-district officials, Secretary of Education Betsy DeVos used the bully pulpit to promote charter schools, vouchers, and tax credits for private-school scholarships. To the distress of teachers unions, the U.S. Supreme Court struck down an Illinois law requiring government workers who elect not to become union members to pay representation fees. To the chagrin of civil-rights groups, the U.S. Department of Education said that it was reviewing a letter sent to school districts by the Obama administration informing them that they were at risk of incurring a civil-rights violation if students of color were suspended or expelled more often than their peers. To the relief of Common Core enthusiasts, the politically charged debate over the standards moved to the back burner. And to the dismay of parents, teachers, and policymakers across the political spectrum, students demonstrated almost no gains in reading and math on the 2017 National Assessment of Educational Progress (NAEP) over the 2015 test.

All these events were consequential, but none penetrated into the thinking of the American public as sharply as did teacher strikes in six southern and western states. Those walkouts seem to have lent new urgency to teacher demands for salary raises and increased financial support for schools.

The status of public opinion on these and other topics comprises the 12th annual Education Next (EdNext) survey of public opinion, administered in May 2018. The poll’s nationally representative sample of 4,601 adults includes an oversampling of parents, teachers, African Americans, and those who identify themselves as Hispanic. (All estimates of results are adjusted for non-response and oversampling of specific populations. See methods sidebar for further details.)

On several issues, our analysis teases out nuances in public opinion by asking variations of questions to randomly selected segments of survey participants. Respondents were divided at random into two or more segments, with each group asked a different version of the same general question. For example, we told half of the respondents—but not the other half—how much the average teacher in their state was paid before asking them whether they thought salaries should be increased, be decreased, or remain about the same. By comparing the differences in the opinions of the two groups, we are able to estimate the extent to which relevant information influences public thinking as to the desirability of a pay increase.