Ever since the Senator Bill Doyle Town Meeting Survey made its annual appearance, and attracted the usual amount of media attention, I’ve been doing some thinking about this uniquely Vermont phenomenon. And I’ve come to some conclusions I’d like to run by the rest of you.
The Doyle survey can be filled out by anyone anywhere, and yet it’s always taken as some kind of bellwether of public opinion. Bill Doyle does his annual turn as the Sage of Vermont, and the political class chews over the meaning of the results with all the fervor of a pack of haruspices examining freshly-liberated goat entrails.
This is the same political class that has a rather poor record of predicting actual outcomes. And there’s good reason for that: the politicos and State House media live in a little golden bubble, they spend a lot of time talking to each other, and conventional wisdom is reinforced.
These are the people who lend credence to the Doyle Survey. The Freeploid’s Terri Hallenbeck writes that “often the results smell right,” and the Mitchell Family’s Man in the Bubble Peter Hirschfeld quotes a not-at-all self-serving Bill Doyle contending that “I think it does give us a sense of how people are thinking.”
Really? Well, now that we actually have a professionally-conducted polling operation in Vermont, the Castleton Polling Institute, we have a useful point of comparison.
The 2012 Doyle survey reported 48% approval for Governor Shumlin. This year it’s 42%. This has been interpreted as a rise in anti-Shumlin feelings, even though (a) the Doyle survey is completely unprofessional, and (b) even if it were, the six-point drop is barely outside the statistical margin of error for a professional survey. The results are precisely meaningless.
Castleton, meanwhile, showed the Governor sailing comfortably along in the 60% range. He won the election with 58% of the vote. Castleton was right, Doyle was wrong.
So why does the chattering class go on and on about the Doyle Survey? Well, we Vermonters love our traditions, the little things that make Vermont different and supposedly better. And everybody in the news business likes Bill Doyle.
All right, if the Doyle Survey is not to be trusted, is it simply because of the poll’s random nature? Or does it misrepresent reality in identifiable ways?
I do not believe that anyone is stuffing the ballot boxes. There’s nothing to gain from doing so, and it would take a lot of effort. No, I think the innate, ungoverned selection process produces a skewed sample. Here’s how.
Who fills out the vast majority of Doyle Surveys? People who attend Town Meeting.
Who attends Town Meeting? Well, to start with, not the residents of our largest communities. They may show up to vote in local elections, but they are much less likely to stop and fill out a Doyle than people who are sitting down for a meeting. I think I’m safe in saying that our smaller communities are over-represented in the Doyle Survey.
Okay, what kinds of people are more likely to attend a Town Meeting? Those with an appreciation for Vermont’s political traditions, and those who want to keep a close eye on their local government. Many of these folks are conservative. Some are liberal, progressive or even radical, but they share that belief in the things that make Vermont unique. I suspect that the Town Meeting crowd is substantially skewed toward lifetime, or longtime, Vermonters.
(Confession: I’ve lived in Vermont for seven years, in a community with a real Town Meeting. I’ve never attended one. And, Frank Bryan hagiography notwithstanding, I don’t feel like I’ve really missed anything, to be honest.)
Take all those inferences together, and what do you have? A survey that overrepresents those who are somewhat more conservative, and very much more invested in Vermont tradition, than the general electorate. (That’d be a very good dictionary definition of “Bill Doyle,” as it happens.) The survey results would tend to overamplify the uniqueness and quirkiness that Vermont is widely believed to possess.
Especially among the political media. I’ve found that many members of the media, even if they tend to be well-educated and personally outsiderish, have a deeply-held fondness for stereotypical Vermont.
Looking at the whole picture, Doyle Survey participants tend to be more Vermonty than the real-life Vermont. The results reflect this. The predispositions of the political class and media tend to accept, or even welcome, the survey results as more “proof” that Vermont is a special and different place.
It’s the Bill Doyle Feedback Loop.
And I believe it has a negative effect on our political discourse. Those who live in the hothouse of state politics think that Vermont is more conservative and eccentric than it really is, and that makes them more cautious than they should be. They tend to over-validate conventional wisdom on issues like taxes and gun control, as well as the lesser controversies that seem to tie politicians in knots, such as vaccination, smart meters, and wind energy.
I mean, as an essentially meaningless bit of springtime fun, the Doyle is an innocuous thing. But to the extent it’s taken seriously as a measure of public opinion, we’d be better off without it.
Don’t have the time to research, but I’m curious what Doyle showed about civil unions and marriage equality and how that compared to election results.
be retired people and people who have the flexibility to take time off from work (such as farmers and the self-employed). In my town, that’s heavily weighted toward retirees. I can say that I’ve never seen a 18-20 year-old at town meeting in the nine years I’ve been attending.
I’d also like to add that even in small rural towns of around one to two thousand people, town and school board meetings only draw about 100 people. So in my town of 1,200, that represents only 10 percent of the population. And if the number of people who fill out the survey break toward town meeting attendees rather than voters only, well, that suggests that the survey is taking the pulse of a Very Small Sample of Vermonters. So talk about missing the big picture.
JV. If honest intelligent Vermonters knew the details of the poll, such as polling stations are not the only outlet, those who like facts & accurate info would recognize it’s inherent flaws. But, since VT is also a religion Mr Doyle does a wonderful job of maintaining the status quo & mythology with all of its friendly home-spun errors.
As with the SurveyMonkey-gate propagated by legislator & Dr. Till when he chose a completely unscientific poll to peddle his wares namely that most Vt doctors were against Shummycare. This may be true however the poll could never prove this since anyone could take it, over & over again.
As it gathered credentials, such as beng associated with UVM, all VT mainstream monkey-see, monkey-do media in their typical lemming-like herd mentality began to report & quote the results of it as a scientific poll. It’s A-Ok, because hiding in the safety of “we” and safety of “numbers” Vermonters once again were not served well.
Only “one” lonely msm source failed to do so-AP said they did not report the results & so-clled “findings” because…they were not scientific. The only msm class-act in VT.
GMD also didn’t buy the spoiled goods, odum sez:
When the good doctor was caught with pants around ankles, claimed survey had been ‘hacked’. Heh.
http://www.greenmountaindaily….
doesn’t strike me as something to boast about, just the opposite. you have a chance to participate in absolutely direct democracy, a system that’s worked well since before the state was a state. i hope you’ll go some year and see.