Bee’s Budget Survey, Like Referendum, Lacked

Widespread Participation

By John Voket

While town officials were advised that there is no legal mechanism by which advisory questions could be placed on Newtown’s local budget ballot, The Newtown Bee attempted to provide an avenue for taxpayers to register not only how they would vote, but the reasoning behind their vote.

And it appears that The Bee’s unofficial budget advisory poll suffered from the same low participation rate as the actual budget.

According to local registrars, 3,747, or only about two of every ten eligible, voters bothered to cast a budget ballot this week. And the budget subsequently failed by a 51-vote margin.

As activity at the budget referendum wound down after 8 pm Tuesday, The Bee’s unofficial budget survey closed. At that time, of the 183 respondents who participated, 72 percent indicated that they were planning to vote against the budget proposal.

The largest number of respondents, by a ratio of 2 to 1, indicated they voted No so more reductions would be made to both the town and school proposals. While 13.7 percent said they voted No because they wanted to see more money restored to the school budget request, 12.6 percent said they voted No in hopes that funding to the school district would be reduced further.

Another 17.5 percent said they voted Yes because they thought a No vote would bring further cuts to the school’s budget request, and 13.1 percent said they thought the proposal that went to referendum Tuesday was “the right budget for Newtown this year.”

Newtown resident Paul Lundquist is a polling expert who volunteered his services in 2007 to produce a comprehensive poll covering myriad issues affecting the community. In regard to The Bee’s poll, he said first and foremost that he had “a great appreciation for the service The Bee has been trying to provide with community-based polling through its website utilizing an online questionnaire from SurveyMonkey.com.”

But Mr Lundquist said he responded after being contacted for comment, so he could “put the brakes on the interpretation/reinterpretation or quoting of these numbers by folks on either side of the issue.”

“I have some concerns about the resulting data that was collected, and I’d strongly suggest that it is not appropriate to use these results for advisory purposes or even as a source for directional insights into the possible reasons behind Newtown residents’ rejection of the budget this week,” Mr Lundquist said via an email. “There are two fundamental problems here: first, there were only 183 people who completed the online survey. Second, and more important, there is no way to determine whether this sample is even remotely reflective of Newtown’s voter population as a whole.”

In regard to the small sample of 183 respondents, Mr Lundquist said the biggest problem is that rules of statistics demand that there would need to be very large gaps in the proportion of people giving each answer in order to have reasonable confidence that the differences are real, rather than being due to random, natural “wobble” in the data.

“Statistical rules say, in essence, the larger the sample size, the smaller the difference needed for two numbers to be significantly and meaningfully different. And the opposite is also true — the smaller the sample size, the bigger the difference needed.

“Statistical rules would also say that the very same answers would have been considered significant if a larger number of people had provided those answers,” he added. “Mathematically, if there were 270 people included instead of only 183, the differences between these top reasons for voting No would in fact be statistically significant. But still not necessarily valid because one would still have to wonder… Who are these people? Do they accurately represent Newtown?”

The full results of The Newtown Bee’s budget poll and all commentary provided by participants can be reviewed at newtownbee.com.