In these weeks following the 2020 election, the number one question I get asked about my work as a pollster is why the election results in many states seemed so different from the expectations from pre-election polling. When I answer, I always offer up four important facts to consider:
- First, remember that poll results reflect a snapshot of a moment in time (they don’t predict the future). Opinion changes, campaigns evolve, and the pool of likely voters is constantly changing.
- Once all the votes were counted, many polls ended up being more accurate than was expected on election night – and polls on ballot measures, as opposed to candidates, did not show as much variance from the final results.
- Polls are not all the same, and they’re not created equal. When evaluating polls, carefully review their methodology, timing and the questions they ask as you evaluate the results.
- There is a legitimate opportunity to learn from the missteps made in some of the polls, and as an industry we should be intentional about improving polling.
In this interview with Bazi Kanani, I address some of the concerns about the dependability of polling and offer thoughts on best practices that can obtain results that more accurately reflect the experiences and opinions of the communities surveyed. Many of these approaches are ones we take when conducting the annual Pulse: The Colorado Health Foundation Poll to hear the perspectives of Coloradans from across the state on issues affecting their health and well-being.
Bazi and I also discuss why polls like Pulse are so important. Even though public opinion research is rightly being evaluated for its effectiveness after this election, it’s critical that we remember its importance and usefulness as a tool to listen to people and reflect back their priorities. Without public opinion research, only those with power and access to decision makers can participate in shaping the policies that impact all our lives.