If you understand polls, some November likelihoods are clear. And it looks like there may be more change in Olympia than in D.C.
The political pre-season is almost over. The Republicans have their nominee for president. The major candidates for federal and statewide office have launched their campaigns. By May 18, candidates will have officially filed for office.
Soon the money will start to fly and we will be treated to a new poll every few days. Now, in the calm before the storm, let’s take a look at how polling really works, and where the major races stand in advance of the real battle soon to come. (Below, we will also survey the big races for Washington state.)
Polls taken by reputable pollsters are generally accurate. Too many people, however, don’t understand how to read and understand the results of a poll. And, increasingly, arguments arise due to the differing techniques pollsters use to “weight” their likely voters survey in an attempt to predict the makeup of the actual electorate.
When reading a poll, two things must be understood. First, polls do not predict the future, they simply show a snapshot of where things stand at the time the poll was taken. Second, although a single number is reported, what a pollster really identifies is a range within a margin of error.
For example, on Feb. 16 of this year, Survey USA released a poll of the governor’s race and the headline was, “McKenna leads Inslee by 10%.” The poll was done Feb. 13 to 15 and had a reported margin of error of 4.2%. This poll, therefore, does not mean Rob McKenna is going to win the governor’s race by double digits. What it means is that it is highly likely that, when the poll was taken, McKenna had a lead of somewhere between 6% and 14%. (That's accounting for a 4% margin of error above or below the reported 10% lead.)
A 6% McKenna lead would be consistent with other polls in this race taken at that time, and, therefore, not a surprising result if you know how to read polling.
There is another problem in the polling world, however. Pollsters must make educated guesses about the makeup of the population that will actually vote. Often they adjust their raw data based on these assumptions. I wrote about this weighting process when the polls were so inconsistent during the 2010 U.S. Senate race ("Why are the polls so wildly different in the Rossi-Murray race?," Crosscut).
To be accurate, a poll must be done on a statistically valid sample of the population. Your sample must have the right number of men and women, young and old, and the right geographic mix. As a poll is being conducted, pollsters will work with their callers to get this right mix, or they will weight the results after the poll is done. If you didn’t get enough responses from men, for instance, you can look at the responses from the men sampled, extrapolate, and adjust (or weight) the results accordingly.
This weighting is easy to do when you have known demographics, such as age. But what about trying to weight a sample to reflect party identification? Party identification shifts over time, and no one really knows how many Republicans and Democrats are going to vote in each election (the so-called enthusiasm factor). Quite often assumptions have to be made, and these assumptions can dramatically affect results.
In 2010, the factor pollsters were trying to weight was the partisan makeup of the off-year electorate. This year it is race.
Earlier this month, Gallup released a tracking poll that showed Mitt Romney leading President Obama 47% to 45%. Almost immediately the president’s strategist, David Axelrod, attacked the validity of Gallup’s sample, saying it underestimated the percentage of non-white voters who will participate in November. Sean Trende at RealClearPolitics.com responded with a vigorous defense of Gallup’s numbers.
The argument comes down to this: Will the percentage of the white vote be in the high 70s or the low 70s? Pollsters have to make an educated guess, and the differing answers yield slightly different polling results. The actual final answer will likely determine whether this will be a close presidential election or a comfortable Obama victory.
Moving to the big picture so far, the good news for Republicans is they will nominate a candidate for president who has a credible chance at victory, and who should not cripple the rest of the ticket. And President Obama's approval rating remains under 50%. The bad news for the GOP is all the data indicates that if the election were held today President Obama would defeat Mitt Romney. Obama is ahead by roughly 4%.
More importantly, Obama leads in all the battleground states a Republican must win in order to have any chance at getting to 270 electoral votes: Ohio, Florida, Virginia, and North Carolina.
Republicans, however, have a slight lead in the generic ballot polling. As we turn the calendar to May, it appears we are headed to a status quo election. Democrats hold the White House, Republicans control the House, and the Senate remains gridlocked. Will events change the trajectory of the election? Can Mitt Romney reintroduce himself to swing voters?
A continuation of the status quo will certainly not be the result of our elections here at home. By January of next year, we will have a new governor, three new members of Congress, at least two new statewide officials, and dozens of new legislators.
When the filing period closes in two-and-a-half weeks we will be able to do a thorough analysis of where things stand. Until then, here are some snapshots from the trail:
- We have no new polling in the governor’s race so there is no way to know if Jay Inslee was damaged by all the media attention paid to his resignation from Congress and the dispute over the costs of a special election to fill his seat. Both campaigns continue to snipe at each other with press releases and web videos while we await new polls and the beginning of the ad wars.
- In the highly competitive 1st Congressional district, the Democratic establishment appears to be lining up behind Suzan DelBene, but the other Democratic candidates, most notably state Sen. Steve Hobbs, Laura Ruderman, and Darcy Burner, all appear to have the resources and desire to stay in the race. Republican John Koster awaits the survivor.
- Sen. Maria Cantwell, and incumbent House members Rick Larsen (D), Jaimie Herrera Beutler (R), Doc Hastings (R), Cathy McMorris Rodgers (R), Jim McDermott (D), Dave Reichert (R), and Adam Smith (D) are all cruising to re-election.
- Democrats Derek Kilmer and Denny Heck, running for the open 6th and 10th congressional districts, are still prohibitive favorites to win those seats.
- In the race for state Attorney General, Republican Reagan Dunn and Democrat Bob Ferguson remain virtually tied in terms of fundraising. Dunn has raised $684,000, Ferguson, $697,000. There is no new polling in this race either.
- Democratic State Rep. Roger Goodman has dropped out of the 1st district race for Congress and will run for re-election to his state House seat, taking away an open seat target for the GOP in the 45th legislative district (Redmond and Woodinville).
- In the 28th district (Lakewood), Democratic Rep. Troy Kelley will not seek re-election (he's running for state auditor). This creates an open seat opportunity for the Rs in a swing district, and takes pressure off Republican Sen. Mike Carrell, as his previously announced Democratic opponent has decided to run for the open House seat instead.
- Democratic State Rep. Tim Probst announced he will run against GOP Senator Don Benton in the 17th district near Vancouver. Good news for the House Rs; they will likely pick up the open House seat. Bad news for the Senate Rs. This is a Republican district, but the GOP wasn’t planning on having to spend resources to defend the Senate seat.
- Finally, Olympia insiders have anticipated for some time that the Senate Republicans had a chance of gaining three seats and winning the majority. Now it appears that the House Rs, have done well in recruiting and might put enough seats in play to give them a realistic chance at winning the majority.
