Presidency Scholar, specializing in elections, White House organization, decisionmaking, American politics.
Joined January 2012
- Tweets16.6K
- Following1.5K
- Followers2.2K
- Likes1.9K
UNH poll director Andrew Smith defends his UNH Vermont polls, noting (I think) that the likely voter weights they used were not the same as the age distribution of all Vermonters - which is the point I made as well in my comments on these polls. wcax.com/news/2026/11/06/que…
The problem is that, as far as I can see, they didn't publish the weighted age estimate of likely voters, so it is hard to understand why they were so off in the governor's race. Smith argues (I think) that it was something specific to @PhilScott4VT's coalition. #vtpoli
Per my earlier discussion of the dueling VT polls, the State Navigate likely voter model estimated 41% turnout in 65+ category. The UNH published sample had it 27%. Actual turnout in this age group, per NYT, was 39%. #Vermont
The difference mattered because this age group voted substantially for @PhilScott4VT. As I noted earlier, if you reconfigure the age groupings for the State Navigate sample to match UNH's age distribution, State Navigate would also have Janoo winning 53% of the vote. #Vermont
To be clear, I did not mean that @PhilScott4VT attacked Janoo personally in his closing statement. He attacked the lack of specifics in how she would pay for her policies. In my view, it was a very effective closing statement, but it did not veer into the personal. #vtpoli vtdigger.org/2026/11/04/in-a…
Matthew Dickinson retweeted
VT polls disagreed on the Scott/Janoo race: UNH (Janoo +8) vs. Vermont Public largely came down to projections of how many young people would turn out. The exit poll that the NYT is reporting for Balint race suggests young voters simply did not show up, and were swamped by 65+
With 88% of the vote in, worth noting that Scott is above the 50% vote threshold necessary to win outright. #Vermont
1
118