Given the magnitude of the Olympics issue in Boston we can expect continued post-mortems and “winners and losers” lists to emerge. One factor that seems beyond question is that polling played a large role – the USOC cited poor approval numbers as one reason it pulled the bid from Boston. With polling once again assuming a central position in our political discourse it is worthwhile to reflect upon the role it plays in our democracy.
One person who has thought about polling a great deal is MassInc Polling Group President Steve Koczela, who posted some of his insights in a recent The Topline, Special Edition: You Killed the Bid. Mr. Koczela is a topnotch professional. He’s visited my classes several times to discuss survey research with my students. Still, we have some differences.
One would be his argument that “Public polling, at its most basic, is a way for citizens to speak to their leaders and to one another.” Conversation and deliberation among citizens and with their leaders is essential to democracy, but if polling is a form of this at all it is a very weak and impoverished form.
Let’s suppose that television station WJM in Minneapolis wishes to broadcast poll results and hires Murky Research to conduct the poll. WJM and Murky want a sound and accurate poll, and WJM would also like something that will enhance the appeal of its broadcast to viewers and thus attract advertisers. They collaborate on what issues and personalities to poll about and Murky develops a script and goes about conducting a methodologically sound professional poll.
The script is sent to a call center not in Minneapolis but in California. In other words, this conversation will not be conducted between friends and neighbors with similar interests. For some odd reason WJM wants to broadcast something about Massachusetts politics so call center employees will be instructed on how to pronounce names like “Dukakis” and “Polito” (they still won’t get it right, however).
Some number of Minneapolis voters (400, 500, 1,000) will hear the exact same script with the same menu of answer options. Deviations from the script could bias the responses. The voters will not have the ability to change the topics, ask extensive questions, or seek additional information from their family members, social groups, unions, church communities, political parties, or community leaders.
When sufficient interviews have been completed the numbers will be crunched and answers aggregated and summarized, cross tabbed, and subjected to other statistical tests, and the poll released. Respondent 206 will never have discussed the poll topics with Respondent 347 nor with any other of the respondents, and none of them will engage in a conversation with other members of the community or their political leaders – at least not correspondence that is captured by Murky Research. The summary responses of Respondents 206 and 347 are somewhere in the topline results that will be broadcast by WJM but at this point they are merely data points, not citizens.
Democracy requires not only speech but listening, as the late political scientist Wilson Carey McWilliams argued. It is aided by the filtering and interpreting of social, work, faith, and community institutions. McWilliams had in mind messy and sometimes turbulent exchanges between and among citizens and their leaders, not summary statistics.
As to a second point raised by Mr. Koczela, “In a time when most parts of the political process are heavily weighted toward wealthy donors, Boston 2024 died at the doorstep of the everyday voter.” That seems pretty accurate. But, can it be replicated?
Boston2024, which had the money to use television advertising to improve the public’s opinions of the Olympics, did not do so. One of the endearing aspects of the story was the David and Goliath angle – NoBostonOlympics and NoBoston2024 operated on a financial shoestring; they couldn’t have matched any sizeable TV buy. To give one example of what can happen in such circumstances, in September 2012 the death with dignity referendum was leading in one poll by 64%-27%. After opponents broadcast a barrage of television spots the referendum’s proponents could not answer, the measure was narrowly defeated.
What decision-making by poll could point toward is a race to elevate poll numbers around particular issues or individuals. This is what we are seeing in the Republican presidential primary, as marginal candidates like Mike Huckabee make outrageous statements in hopes of elevating poll numbers enough to get chosen for the first tier debate. When such issues involve the moneyed versus the un-moneyed, how often will the wealthy decline to use their advantage?
Come Thursday there will be more to say about what numbers mean, knowledge and public opinion, polls and the republican form of government, and related topics.