SPRINGFIELD --  Less than a week before Springfield's big decision on a police and fire pension fund 3/4-percent sales tax, a new poll released exclusively to KY3 News shows a closely divided contest.  The Missouri State University poll shows a slight but barely measurable edge for those supportive of the sales tax. 

The public poll offer some hope for proponents and city leaders, who suffered a stinging defeat of a 1-percent sales tax for the same purpose last February.

Of the 275 registered voters polled, 45 percent signaled support and 44 percent said they're opposed.  Both are numbers that fall well within the 5.9-percent margin of error.  The other wild card is that 10 percent of registered voters remain undecided.

"It'll be interesting to see if those folks who are not sure end up siding significantly with one side or the other on election day," said Missouri State University polling director Brian Calfano.

Add in all 321 residents polled and support grows, 47 percent to 43 percent for the tax, with 11 percent unsure.

"I'm not that surprised," said Calfano of the results.  "I would've been more surprised if it looked like a blowout on either end," he said.

The data shows women are more likely to support the tax than men by a 12-point margin. Broken down by age, those 35 to 54 register the highest support for the tax at 56 percent, while 53 percent of younger people -- ages 18 to 34 -- oppose it.

MSU political science and graduate students spent two weeks, from Sept. 26 to Oct. 11 making calls and collecting the data.

"Everything we've done has been vetted by other survey researchers with great experience throughout the U.S," said Calfano.

He also notes the poll did not attempt to target what type of turnout that precincts will experience on Tuesday.

"The other caveat is that the poll is not a prediction of who is likely to turn out," Calfano said.

FULL POLL BREAKDOWN: RESPONDENT DEMOGRAPHICS

Poll was conducted from Sept. 26 to Oct. 11, 2009, with a scientifically selected random call sample using Random Digit Dialing (RDD) of 524 Greene County residents, 321 of whom were determined to be Springfield residents by matching computer generated respondent numbers with city addresses generated as part of the random call sample. 275 of the 321 Springfield residents were determined to be registered voters. The poll was conducted through the Center for Social Sciences and Public Policy Research on the campus of Missouri State University, under the supervision of poll director Brian Calfano.

It should, therefore, be viewed as a general public opinion survey of Springfield residents, including registered voters, about the issue(s) in question during the time the poll was in the field. It does not purport to identify likely voters, or isolate opinions among likely voters. Survey results were weighted according to respondent age, sex, and education level using an iterative weighting algorithm. Using weights helps to ensure that poll respondents are reflective of the Springfield population more generally (as determined by U.S. Census data). This is especially important concerning the age weight in relation to younger Springfield residents, who are more likely to have cell phone numbers only, which are not covered by standard RDD procedures. Unless otherwise noted, all results are reported with the applied weights for the 321 Springfield residents and the 275 registered voters. All surveys are subject to sampling error, which is the expected probable difference between interviewing everyone in a population (i.e., the City of Springfield), versus a scientific sampling drawn from that population. The sampling error for the Springfield resident sample of 321 adults is +/- 5.5 percent, at a 95 percent confidence interval. Therefore, if 50 percent of Springfield residents favored a particular position, one would be 95 percent sure that the true figure would be between 44.5 and 55.5 percent had all Springfield residents been interviewed. The sampling error for the Springfield registered voter sample of 275 adults is +/- 5.9 percent, at a 95 percent confidence interval. Therefore, if 50 percent of Springfield registered voters favored a particular position, one would be 95 percent sure that the true figure would be between 44.1 and 55.9 percent had all Springfield registered voters been interviewed. Sampling error does not take into account other sources of variation inherent in public opinion studies, such as non-response, question wording, or contextual effects.

Would you support a proposed sales tax increase in Springfield to address shortfalls in the police and fire pension fund?

1 Some aggregate cell counts may not equal 100% due to rounding.