Review of MPR, Humphrey Institute poll recommends changes

Minnesota Public Radio and the University of Minnesota's Humphrey Institute of Public Affairs Friday released a review of the methodology used in this year's election polling, after the results of the poll in the Minnesota governor's race came under criticism.

The final MPR News-Humphrey Institute poll this election season came under scrutiny for showing Democratic gubernatorial candidate Mark Dayton ahead of Republican Tom Emmer by 12 points in late October. Dayton defeated Emmer by 8,770 votes, or just 0.42 points, less than two weeks later.

The review by University of Minnesota professors Larry Jacobs and Joanne Miller discussed several reasons why the poll may have skewed heavily in Dayton's favor.

Gallup editor in chief Frank Newport highlighted several areas in his independent audit of the review.

Both the poll's authors and Newport highlighted the poll's apparent failure to weight the survey sample by region. The review found higher participation rates from individuals in the 612 area code -- 81 percent compared to 67 percent statewide.

"Given that Democratic candidates traditionally fare well among voters in the 612 area code, the higher cooperation rate among likely voters in the 612 area code may explain why the estimate of Emmer's support by MPR/HHH was slightly lower than those by other polls conducted at around the same time," Jacobs and Miller wrote. In the future, they said, they would include geographical weighting.

Newport, of Gallup, wrote that weighting samples by region "is commonly done in state polls, particularly in states where voting outcomes can vary significantly by region, as apparently is the case in Minnesota."

Newport also questioned how the poll's authors determined "likely voters." He noted that the poll's authors didn't provide details about how they determined likely voters.

"This is an important area of focus," Newport wrote. "Screens for likely voters typically produce a more Republican estimate (than) if all registered voters' vote intentions are taken into account -- sometimes substantially so."

The review and Newport's audit also include recommendations to include cell phones in future polls, and to hold at least one final poll closer to Election Day.

The final poll this election season surveyed 751 likely Minnesota voters from Oct. 21 to Oct. 25. It had a conventional margin or sampling error of +/-3.6 percentage points, and a more conservative margin of 5.5 percentage points. Dayton's 12-point lead in the poll attracted considerable media attention.

Earlier polls indicated a tighter race. The MPR News-Humphrey Institute Poll conducted from Aug. 25 to Aug. 29 showed Dayton and Emmer tied with 34 percent of the vote. The second poll, conducted from Sept. 22 to Sept. 29, showed Dayton leading Emmer 38 percent to 27.

Both MPR News and professors Jacobs and Miller "jointly determined the content of the questions," the report said. The Humphrey Institute "oversaw the survey process and conducted the analysis of the polling data," the report said. MPR was responsible for deciding the timing of the surveys.

"This review was a necessary step, and we look forward to more conversation about the recommendations," said Chris Worthington, managing director of news at Minnesota Public Radio, in a statement released Friday.