The LGBT community should respect and use polling and focus groups to understand how we are doing with voters – but not to the exclusion of doing other valuable research, including learning from campaign history and from empirical experimentation.
Intellectually honest research on LGBT issues is a challenge—not only because the work itself requires thinking and stamina, but also because it demands consistent curiosity. Sometimes, it is emotionally taxing to sustain that high level of curiosity because part of the reward is the discovery that anti-gay prejudice is harder to change than we might ever have initially expected.
Yet there is no substitute for curiosity in these ballot measure fights. When we have an unquenchable desire to understand voters—how they think, feel, and respond—we are much less likely to assume we already know who is persuadable and what will persuade them to stand with us.
Polling alone—even polling combined with focus groups—is simply not adequate to satisfy that curiosity. This is not to denigrate polling or focus groups. They are useful tools. Future campaigns should use them. But they are not a panacea.
If you, the reader, want to know more about research experiments that can supplement polling and compensate for its blind spots, see Recommendation 3 for some examples.
If you, the reader, want to know how to extract truth from a poll, optimal performance from a pollster, and know whether and when to commission a poll, go to Appendix J, which explores these topics in more detail.
The most important points about polling that every LGBT person and ally needs to know are these:
· Polling is both invaluable and inadequate, both insightful and misleading. If you want to understand whether the campaign you care about is making smart decisions, take the time to learn how to read a poll, how to spot its potential weaknesses, and how to ask questions that will uncover assumptions that can limit any poll’s value.
· Don’t poll when it won’t help. Read Appendix J before you commission a poll.
· Polling is not a substitute for doing work that is harder.
· Polling is not a substitute for knowing our history on these ballot measures and our experience fighting anti-gay prejudice. Respect the value of both polling and history and the next campaign will not have to repeat past mistakes.
Polling performed well has the potential to provide insights that could otherwise easily elude us. The Lake tracking polling is just one example; it was able to measure which voters moved and when with significant precision.