Two Surveys, Two Divergent Findings
School Administrator, February 2022
In preparation for a school tax election seeking additional operating money to support classroom programs, the Anyville Public Schools contracted for a scientific, random-sample survey to test support for key components of its educational plan and the community’s tax tolerance.
Anyville (a pseudonym for a school district we’ve worked with), located 40 miles outside of a metropolitan area, serves multiple cities and townships. The district’s K-12 enrollment of just under 4,000 students is served in four elementary schools, one middle school and one high school.
The methodology in the district’s scientific survey included 350 random-sample interviews and a phone interview of about 30 probes, yielding a margin of error of plus or minus 5.2 percent at the 95th confidence.
The random interviews were controlled for demography, including age, gender, parent status, geography within the district and past voting frequency. The intent of these targets was to have the makeup of interviewees closely mirror the demographic profile of all registered voters in the district.
The survey used three benchmarks that initially tested support for the proposal before information, then measured support after information, and last evaluated reaction to potential tax increases. The initial reaction to the district’s proposal was tepid with 42 percent support and 58 percent closely split between opposition and undecided. After providing interviewees with nine examples of both positive and negative consequences related to the desired funding, support was tested a second time. Unfortunately, information only increased support 2 percent with 56 percent still either opposed or undecided.
The final test sought reaction to three potential tax increases of $25, $50, and $75 per year on a $100,000 of property value. Only the $25 impact measured above majority support at 51.8% (potentially below 47 percent given the margin of error). Complicating things further was the fact that the most active voters were 3 percent to 9 percent less supportive.
On a separate probe, more than half of respondents said they would not support a tax increase of any amount for any purpose.
Opt-in Survey
Anyville followed the scientific survey with an opt-in survey completed by nearly 1,100 residents. Demographic makeup of survey takers was unrepresentative with 22 percent more females than males, nearly 60 percent of respondents between ages 35 and 54, only 20 percent were ages 55 or older, and 67 percent of the respondents were parents of school-age children.
Not surprisingly, this unrepresentative sample — self-selected and based on self-interest — projected significantly higher support on all data points, measuring an average of more than 8 percent higher support for the three tax increases.
The Anyville case study reinforces the importance of school leaders matching survey methodology with key decision making. The opt-in survey estimated support at the $25 impact at a comfortable margin of 61 percent while the scientific survey projected between 47 percent and 57 percent. On Election Day, fewer than 48 percent of community residents voted to approve Anyville’s ballot question.
— DON LIFTO AND WILLIAM MORRIS