Poll Wars and Horse Races: Why We Need Less Polls and More Policy

(This is my second blog today so please look at this and the other one).
So what’s in a poll? Over the last two weeks four polls on the Minnesota governor’s race have been released, offering different interpretations of who is ahead or behind. The main news story has thus become the horse race, and the results of these polls may impact the fortunes of the different candidates.

As far as I am concerned the polls have become a distraction and we would be better off if we spent less time on them and more time on the substantive public policy proposals the different candidates have broached.

Ok, I will confess to have gotten sucked into the polling frenzy. Several weeks ago I criticized one MPR-Humphrey Poll and in the last week I have also commented on the Rasmussen Fox 9, Star Tribune, and now the most recent MPR poll. Some of my blogs address polling methodology issues and last Tuesday I discussed polling issues on Fox 9 at 6PM.

All of this polling is problematic on several scores. First from a geek point of view (of which I may be one), they may be giving inaccurate pictures of the race because of their basic methodological flaws, differences, or assumptions upon which they are based. My core criticism of the most recent MPR poll–and Steve Schier has said the same thing in a recent e-mail–is that it makes assumptions about what the electorate looks like that are wrong.

I agree that there is no way that the percentage of people who identify themselves as DFLers has grown in the last few weeks, and also would argue that there is no way that the percentage of individuals who identify themselves as DFL is equal to or greater than it was four years ago. Their numbers have shrunk, as have those who consider themselves Republicans.

To state again what I have been arguing for several months, here is my estimate of the partisan alignment in MN.
DFL 35%
GOP 30%
Independence Party 10%
Unaffiliated 25%

These numbers reflect a disenchantment and disaffiliation with the two major parties and an increase in those who are unaligned. Some unaligned are fringe right and left but most are centrists or moderates.

These percentages are my best guess based on previous elections and assumptions about trends this year. They should be treated as approximates. If the polls do not accurately capture these percentages the polls will over, under, or misrepresent candidate strength and support.

Moreover, to be accurate, polls also need to weigh other issues such as clarify who is likely to vote, balance among critical subpopulations such as Iron Rangers, suburbanites, and those in Minneapolis and St. Paul, and then give a better read on who the swings are. Minnpost pieces have done a good job recently discussing some of these issues.

However, beyond these geek issues, there are two other problems with the polls. First, they have come to replace substantive coverage of the three candidates. We no long hear of where candidates stand on jobs, education, or the deficit. This displacement of substantive policy with political strategy has a place, but it should be limited. Voters are not well served with this focus. Second, if voters are simply making their decisions based on polls this is not good. Should Tom Horner be judged as a contender based simply on how he is doing in the polls or should he be judged by the merits of his proposals? The same should be asked of Mark Dayton and Tom Emmer. The question that should be asked is not “Can you win based on the polls?” but “What do you plan to do if elected governor?”