Two public polls of the Vermont electorate have been released in the past two weeks. One, conducted by the Castleton Polling Institute for Vermont Public Radio, showed the governor’s race essentially tied, with Republican Phil Scott at 39 percent and Democrat Sue Minter at 38 percent. The second, conducted by Braun Research of Princeton, N.J., for WCAX-TV, showed Scott leading by 47 percent to 40 percent, just outside the margin of error.
Both polls showed that Hillary Clinton and Sen. Patrick Leahy have very large leads among Vermont voters, and that Democratic lieutenant governor candidate David Zuckerman is ahead of Republican Randy Brock by a margin in the low- to mid-teens.
There are several explanations for the different results of the two polls for governor. First, the interviews for the Vermont Public Radio poll were conducted in early October, while the WCAX poll was in the field in mid-October. Campaigns do matter, and Scott might very well have gained a few percentage points between the time the two polls were conducted.
The second difference is methodological, and may explain why Castleton Polling is given a B+ grade in 538.com’s ranking of polling firms, while Braun Research is given a B- grade. Castleton used random digit dialing, to both landline and cellphone numbers, to draw its sample. Interviewers then asked screening questions to determine which respondents were registered voters and were likely to vote in November.
Braun Research used a voter registration list to draw its sample. There are two difficulties with that approach. The first is that, according to the Secretary of State’s office, there have been thousands of newly registered voters this fall. The list that Braun used to draw the sample for the WCAX poll was likely out-of-date, and did not include new registrants. If new registrants are disproportionately supporters of one candidate or political party, that could skew the poll results.
The second difficulty is that accurate telephone numbers are generally not part of publicly available voter registration databases. Even privately produced voter registration lists often omit current phone numbers. This poses particular problems in reaching voters who have recently changed phone numbers, have unlisted numbers, or rely primarily on a cellphone with a non-public number.
My sense is that both Phil Scott and Sue Minter have a plausible path to victory in the gubernatorial race, with Minter perhaps having somewhat greater momentum in the final weeks, as she did in the last weeks of the Democratic primary campaign.
Scott needs to convince about a quarter of those who will vote for Hillary Clinton for president to split their ticket and vote for him as governor. These voters believe that it is time for a change in Montpelier after six years of a Democratic governor, and that having a Republican governor would be a useful balance to a legislature with large Democratic majorities.
Minter can rely on the strong performance of Democratic candidates above her on the ballot — Clinton, Leahy and Peter Welch — as well as her endorsement by, and active support from, Bernie Sanders, currently by far the most popular political figure in Vermont. Also, President Obama, highly regarded in Vermont, has recorded a radio commercial endorsing Minter.
Minter will benefit from the better organizational and financial resources of the Vermont Democratic Party compared to the Vermont Republican Party. These organizational and financial advantages make a big difference when it comes to identifying, mobilizing, and turning out supportive voters. Turnout in this year’s election in Vermont is likely to be high, and high turnout in presidential years generally favors Democratic candidates.
Whether Scott’s campaign can overcome Minter’s organizational and endorsement advantages in the final week will likely determine who will be the next governor of Vermont.
Eric L. Davis is professor emeritus of political science at Middlebury College.