Home On the Agenda Is All This Polling Even Good for Democracy?

Is All This Polling Even Good for Democracy?

January 15, 2016

Public opinion research has never been more central to our political process. Polls provide the lifeblood of reporting during our seemingly endless presidential election. Candidates use polls to test messages, positions and attack ads. More polls tell them if they’re succeeding. Polls have even become a deciding factor of who, among the bevy of Republican candidates, gets to debate in prime time.

Interest groups not only use polls to frame high-impact messages, they also often use results (valid or not) as their message as they claim to serve the will of the American people.

But is all this polling good for democracy? Public opinion research can be both a boon for democracy and a bane. It all depends on whether it is done well and how it is used.

Public opinion research can be both a boon for democracy and a bane. It all depends on whether it is done well and how it is used.

On the quality of polling, the industry currently faces difficult methodological challenges. Too many people refuse to participate, have no landline, or both, making poll results less precise and reliable. This response rate problem is currently plaguing the profession.

While these methodological challenges are serious, they are technical and will likely be overcome. But several more fundamental problems will also need attention if polling is to serve democracy well.

First, pollsters often fail to distinguish between superficial responses versus those that are more stable. Second, communication about and reporting on results tends to sensationalize or oversell them. Finally, there is the question of whether public opinion can and ought to stand in for the public’s voice in our democracy.

Researchers often do a poor job of distinguishing between superficial, unstable, top-of-the-head responses and deeply felt, thought-through, stable ones. This distinction makes a huge difference for how meaningful and reliable polling results are, as well as their usefulness for the democratic process.

Social scientist and public opinion research pioneer Dan Yankelovich best conceptualized this problem in his classic Coming to Public Judgment.

Yankelovich, who is also a co-founder of my organization Public Agenda, observed that the public progresses through stages as it comes to terms with a public issue over time, moving from raw opinion to stable “public judgment.” Public judgment comes about when people have wrestled with an issue, worked through their conflicts about it, and are willing to accept the trade-offs involved in pursuing a solution.

Polls on a topic that people haven’t spent much time thinking about often capture superficial results that can easily change. These results can therefore be misleading. Consider, for example the difference in polling results related to public views on the Trans-Pacific Partnership versus those related to gay marriage. The former is a highly technical, fairly recent policy that few voters have had reason or time to consider or explore. Few could name the trade-offs involved in the trade agreement, never mind articulate their views on them.

Gay marriage is a social issue that most of us have had reason to consider and wrestle with in our daily lives, over the course of many years. If a public official wanted to look to polling for guidance on policy, a poll on gay marriage would yield far more reliable results that provide an accurate reflection on where the public stands than would a poll on the trade agreement.

When exploring the value of polling in democracy, we must also look at the vital question of how information about polling responses is used and communicated. It is highly questionable, for instance, that horse-race reporting early in the primary season, when name-recognition is the main factor fueling poll results, is a democratically meaningful and useful practice. It conflates the fact that people have heard of someone with the notion that they think they’re the best candidates, which is frankly ridiculous. On the other hand, if we again consider gay marriage, polling in recent years has provided a reliable lens on a changing social norm.

Finally, our poll-saturated culture unfortunately tends to position public opinion research results as the voice of the people. They are not. Polls provide a sampling of people’s opinions at a given point in time, useful for gaining a degree of insight into people’s current views and values. They become democratically meaningful only when leaders use results to launch better public conversations and create better-informed public policy rather than savvier communications meant to keep the public at bay.

Polls ought to be used to spur active civic engagement and better working relationships between citizens and candidates or government, rather than be treated as a proxy for those democratic essentials and an excuse to avoid them.