July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling. Top Chart among all voters: Christopher Taylor (39%); Sabra Briere (19%); Stephen Kunselman (15%); Sally Petersen (13%); Undecided (15%). Bottom Chart (if the election were conducted among those who disapproved of current mayor John Hieftje’s performance): Christopher Taylor (19%); Sabra Briere (19%); Stephen Kunselman (32%); Sally Petersen (20%); Undecided (9%).

They show Ward 3 councilmember Christopher Taylor to be a clear favorite, with about a week to go before the Aug. 5, 2014 primary. Taylor polled at 39% compared to 19% for Ward 1 councilmember Sabra Briere.

Ward 3 councilmember Stephen Kunselman and Ward 2 councilmember Sally Petersen polled a few points behind Briere at 15% and 13% respectively.

The poll indicates that 15% of voters still haven’t made up their minds. Margin of error for the poll is plus or minus 4.7%.

The relatively large four-candidate field is attributable to the fact that no incumbent is in the race.

Kunselman was the first of the four candidates to declare his candidacy – before mayor John Hieftje announced last year he would not be seeking reelection to an eighth two-year term.

The PPP poll also asked respondents if they approved of the job that Hieftje was doing as mayor.

One of the patterns revealed in the analysis of the poll responses is that Kunselman would be a 12-point favorite if the election were held just among those voters who disapproved of Hieftje’s performance. But the poll indicated that only 27% of Ann Arbor voters disapproved of Hieftje’s performance.

A polling question that asked about favorable or unfavorable opinions of candidates – independently of an inclination to vote for them – showed Kunselman polling with the highest unfavorable opinion numbers, at 36%. But the “not sure” category for that question polled fairly high across all candidates, ranging from 29% to 43%.

The poll also included two questions about future growth – one about downtown development, and the other about the need for an improved train station. The poll indicated 46% support for the downtown projects that have been approved and built in recent years and 39% opposition. The need for a new train station polled at 52%, while the alternate view – that the current station is adequate – polled at 35%.

The content of the poll – which evinces some knowledge by its creator of the Ann Arbor political landscape – was not commissioned by The Chronicle or by any of the four mayoral campaigns. Tom Jensen grew up in Ann Arbor and is now director at Public Policy Polling, a firm based in Raleigh, North Carolina. But Jensen still follows Ann Arbor politics. It was Jensen who put the poll together – out of his own interest. And it was Jensen’s voice that was used in the interactive voice response (IVR) technology deployment of the Ann Arbor mayoral poll.

The poll drew as a sample all those who’d participated in any primary election (Democratic or Republican) since 2006. Poll respondents included 32% Republican or other non-Democratic affiliation.

In a telephone interview, Jensen stressed that any poll result should be viewed with a lot of caution, especially with local elections. “I would definitely, as a pollster, encourage people to take caution in over-interpreting one poll of a low-turnout race in the middle of the summer. You’re definitely prone to more error.”

But based on the results of this poll, he said he was 99% confident that Taylor was going to be the next mayor of Ann Arbor.

Additional charts and some additional background on the polling methodology are presented below.

Analysis: Hieftje as Frame

At the Ann Arbor Democratic Party primary forum for mayoral candidates held on June 14, 2014, Stephen Kunselman sought to distinguish himself from the other three candidates – by saying that he’d represent the working class, stressing that he’s the only candidate with policies and politics that differ from the current mayor, John Hieftje, and from Hieftje’s supporters. “I’m offering you a choice of someone that is not in that camp,” he said.

Hieftje appears to be a popular mayor – judged by outcomes of elections over more than a decade that have not seen Hieftje ever lose a precinct. And that popularity is supported by the PPP poll, which gives Hieftje 52% approval against 27% disapproval and 21% not sure:

Chart 1: July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling. 52% of Ann Arbor voters approve of mayor John Hieftje’s performance, 27% disapprove, and 21% are unsure about his performance.

The logical premises of campaigning against a sitting popular mayor – who is not running for reelection – hinge on the fact that this year’s primary is a four-way race: If Kunselman were to dominate among voters who are dissatisfied with Hieftje, and the other three candidates were to enjoy roughly uniform shares of support from voters who are satisfied with Hieftje, that combination could conceivably give Kunselman more votes than the other three candidates.

While Kunselman is the clear favorite among voters who are dissatisfied with Hieftje, Kunselman still polls at just 32% in that group. [See red bars in Chart 2 below.]

The second logical premise also does not appear to be supported by the polling data. The other three candidates do not roughly split the support of voters who approve of Hieftje’s performance. Instead, Taylor is clearly dominant in that category, achieving a majority of support at 56%, with only Briere doing better than single digits. [See green bars in Chart 2 below.]

Among those who aren’t sure whether they approve or disapprove of Hieftje’s performance, the split is more uniform – the kind of distribution that would be needed among those who approve of Hieftje’s performance, to give Kunselman a win. [See gray bars in Chart 2 below.]

Chart 2: July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

Who Are These People?

Kunselman’s campaign has also relied heavily on asking Ann Arborites to vote for him, the person. The PPP poll indicates that of the four candidates, Kunselman polls highest for those that have an unfavorable opinion of him – at 36% compared to 26% who have a favorable opinion. But even more voters (38%) aren’t sure of their opinion of Kunselman.

The same uncertainty applies to Briere, although more voters (40%) have a favorable opinion of her. Even more voters are unsure of their opinion of Petersen (43%) – who was first elected to city council two years ago. All the other candidates have at least six years of council service. Taylor’s favorable opinion numbers track closely with Hieftje’s approval numbers: 52% of voters have a favorable opinion of Taylor; and 52% of voters approve of the job Hieftje is doing as mayor.

Chart 3: July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

Other Demographics

When the data is cut in other ways, Taylor’s strength, according to the PPP poll, is still apparent, although some exceptions emerge.

In Ward 3 and neighboring Ward 4, Taylor polls stronger than he does citywide, doing better than 50% in those two wards. But in her home Ward 1, Briere hits 42% compared with 31% for Taylor. And in Petersen’s home Ward 2, the poll indicates 29% support for Petersen, compared with 24% for Taylor. The undecided vote in Ward 5 is high – at 21%. Polling data cut across wards is presented in Chart 4 below:

Chart 4: By Ward. July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

Among African-Americans, the poll shows Briere to be the strongest candidate, with 24% support compared to 21% for Taylor. Petersen’s 27% comes close to Taylor’s 32% in the “other” category for ethnicity. Polling data cut across ethnicity is presented in Chart 5 below:

Chart 5: By Ethnicity. July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

Across age, one result that emerges from the poll is that Briere appears to enjoy more support among the youngest voters than she does among the oldest voters. Taylor shows a similar pattern. Kunselman and Petersen are both stronger among the oldest voters than they are among the two categories of younger voters. Polling data cut across age is presented in Chart 6 below:

Chart 6: By Age. July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

With many Republicans expected to cross over and vote in the Democratic primary, the polling sample was selected from previous primary voters without regard to which primary they’d participated in – dating back to 2006. A large number of Republican and independent voters are still undecided – 27% and 26%, respectively. But Petersen is the strongest candidate among Republican voters, polling at 27%.

Chart 7: By Party Affiliation. July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

Candidates don’t show any striking differences across gender lines. Polling data cut across gender is presented in Chart 8 below:

Chart 8: By Gender. July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

Future of Ann Arbor

The poll indicates more support than opposition to recent downtown developments that have been approved and constructed – 46% to 39%. The poll also indicates that a slim majority of Ann Arbor voters (52%) think the city needs a new train station, while 35% think the current station is adequate.

Chart 9: July 28-29, 2014 survey of 435 likely Democratic primary voters by Public Policy Polling.

About the Poll

In the last few days, The Chronicle noticed reports on Twitter of telephone polling for the Ann Arbor Democratic primary mayoral race – citing Public Policy Polling as the polling organization. An inquiry with PPP, located in Raleigh, North Carolina, revealed that it had not been commissioned by any of the four campaigns. Responsible for the poll was PPP itself – in the form of director Tom Jensen.

Jenson spoke with The Chronicle by phone on July 29. Why did he design and implement this poll? Jenson explained: “I’m an Ann Arbor native. And even though I haven’t lived in Ann Arbor for 12 years now, I’m still really interested in Ann Arbor politics. Electing a new mayor for the first time in 14 years … I guess my curiosity killed me and I just decided to do a poll about it.”

Jensen said that he grew up in Burns Park, attended Burns Park Elementary, then Tappan Middle School, then Pioneer High School. He hasn’t lived in Ann Arbor since moving away to North Carolina to attend college.

“Since I was seven or eight years old following Ann Arbor politics, I can’t think of us ever having a truly contested mayoral primary, so there is no precedent for this.”

PPP uses an automated phone polling technique – with pre-recorded questions delivered to the respondents, who press numbers on their phone corresponding to choices. For this survey, Jensen recorded the questions himself. Readers who’d like to hear Jensen’s voice can listen to an extended interview about 2012 election results hosted on the Ann Arbor District Library website.

The survey started with the basic voting question, followed by the question about attitudes toward mayor John Hieftje’s job performance. At that point the demographic questions were asked. Respondents who answered at least the demographic questions had their responses counted for the mayoral voting poll. Additional questions came after the set of demographic items.

Asked if PPP tracks how many people hang up the phone without responding to the poll, Jensen explained that the positive response rate for polls in general these days is below 10%. The Ann Arbor mayoral primary poll had a roughly 5% response rate. Jensen attributed the general trend to flagging attention spans. A generation ago, he said, 40% of people who were called would answer the poll questions.

Jensen explained that the polling technique does account for the possibility of variable response rates across different groups of voters. So it’s always important to make sure that you have an appropriate gender, racial and age balance in the poll. Jensen said he also looked at what percentage of the vote came from each ward in the last three mayoral primaries – to make the poll response numbers were lining up in that. He concluded: “Even though the response rate was very low, I think it’s a pretty good representation of the Ann Arbor electorate.”

But he allowed that the more local a race gets, the less accurate the polling is. And that’s not something that is specific to PPP – it’s just something that is true across the board. He attributed that to the fact that it’s harder to model the electorate for smaller elections. It’s hard to say who is actually going to vote in a local election, he said. And because of that he’d cast a wide net for the polling sample. The poll had called anyone who’d voted in at least one primary since 2006.

But it’s possible that people might come out to vote who had never voted in a municipal primary before – and that can have an impact on the results, Jensen said. It’s also possible that the screening for the sample was too lenient – and that perhaps only people who had voted two or three times in a primary should have been called.

So he offered the results with a caveat: “I would definitely, as a pollster, encourage people to take caution in over-interpreting one poll of a low-turnout race in the middle of the summer. You’re definitely prone to more error.”

Still, Jensen ventured: “I am 99% confident that Taylor is going to be the next mayor of Ann Arbor – but I do think it’s good for people to always take poll results with a certain amount of caution, and not just 100% assume that every poll you see is right on the mark.”

The Chronicle could not survive without regular voluntary subscriptions to support our coverage of public bodies like the Ann Arbor city council. We sit on the hard bench so that you don’t have to. Click this link for details: Subscribe to The Chronicle. And if you’re already supporting us, please encourage your friends, neighbors and colleagues to help support The Chronicle, too!