This post is motivated by the early study of measures under Eugene mayors Kitty Piercy and Jim Torrey. The question is: how correlated was approval of Kitty Piercy with support for the library measure.
The answer is: a lot!
This graphic shows the correspondence between support for the library measure and approval of Mayor Kitty Piercy in the first tracking poll – just before the ballots were mailed.
There was a very high correspondence. The distance between the “Don’t Knows” in part is due to their small number and the difference between uncertainty about the mayor and indecision about the measure.
This graphic provides a detailed picture of all question options. Those rating the mayor excellent are also those most in favor of the measure. Similarly, those rating her good or unsure are also yes voters.
Interestingly, those rating her fair are close to the four neighboring responses: somewhat yes, don’t know, lean no, and somewhat no. This implies that these four responses are all, to some extent, similar. Past analyses have indicated that “Lean Yes” responses on measures are those most likely to swing no over time.
Following up on these observations the next is natural: those rating the mayor poor are correlated with those who are strongly no.
QUESTIONS: 1. Eugene will have this measure on the November 3 ballot: Five-Year Library Local Option Levy. Shall Eugene levy $2, 700,000/year for five years, beginning 2016-2017, to increase library hours and expand other library services? This measure may cause property taxes to increase more than three percent? If the election were held today, would you vote yes or no? IF YES/NO: Is that strongly or somewhat? IF DON’T KNOW: Which way do you lean?
2. How would you rate the job Eugene Mayor Kitty Piercy is doing: excellent, good, fair, poor?
METHODOLOGY: 200 live telephone interviews of likely City of Eugene November 2015 Special Election voters were conducted October 14-15, 2015. The margin of error at the sample median for each survey was plus or minus 7%.
The technique of correspondence analysis was applied to the data. It is a form of visual cross tabulation (sometime called internals). The distance between points essentially measures their correlation.
“Approval” includes “Excellent” and “Good” responses. “Disapproval” includes “Fair” and “Poor” responses. Past analyses indicate this is a good measure of relative support.