In the slew of critcism of PPP this week, the most insidious suggestion has been that we 'copy' our results off of other pollsters. The other criticisms are basically differences of opinion about our methodology, but this one goes to a whole different level because it implies seriously unethical behavior on our part.
That attack neglects the fact that we have polled more races that no one else wanted to poll than anyone else in the country over the last 5 years, and generally gotten it right. Here are 15 examples of where we were either the only pollster to look at a race, or the first to pick up on a surprising shift in a contest:
2) When we went into the field for our final Minnesota poll last year, the most recent publicly released poll on the state's voter ID amendment had found it leading 53/41. That made it shocking when we put out our last poll before the election and found it actually failing by a 51/46 margin. But we were right- in the end it failed 54/46.
3) When there was a special election in California's 36th District in 2011 no one was really sure how it would turn out because only 3 points had separated Democrat Janice Hahn and Republican Craig Huey in the primary, and we weren't that far removed from the 2010 Republican wave election. We polled the race for Daily Kos and found Hahn up 52/44...she won 55/45.
4) David Dewhurst beat Ted Cruz 45/34 in the Texas Republican Senate primary last year so it was generally assumed that Dewhurst was the favorite in the runoff too, and with a week to go before the election his internal polling was showing him with a 5 point lead. So it certainly shook up that race when we came out with a poll 2 nights before the election not only showing Cruz ahead, but showing him ahead by double digits at 52/42. He ended up taking the runoff 57/43.
5) 2 polls came out of the 2010 Massachusetts Senate race two weekends before the election. One, by the Boston Globe, showed Martha Coakley leading by 15 points. The other, by us, showed Scott Brown winning by one. When we released our poll the closest one on the race to that date had still shown Coakley ahead by 9 points. But soon thereafter pretty much all the polls started finding what had been shocking when we first put it out, and our final poll finding Brown up 51/46 matched the 52/47 final outcome almost perfectly.
6) While most of the attention on the 2011 Wisconsin State Senate recalls was focused on attempts to recall Republican Senators, there were also 2 attempts the next week to recall Democratic State Senators. The Washington Post reported that 'anyone in the know' in the 12th District would tell you it was a toss up. But we found that Democrat Jim Holperin held a 55/41 lead over his Republican opponent, and he won 55/45. In the 22nd District we found Democrat Robert Wirch winning 55/42 and he won 57/43.
7) In its preview of last year's special election to replace Gabby Giffords in Arizona's 8th Congressional District, CNN reported that all the political rating services saw it as a toss up. So when we went in the field and found Democrat Ron Barber leading easily- by a 53/41 margin- it raised a lot of eyebrows. But he went on to win the next day 52/45.
8) In 2009 Maine had a referendum on whether to keep gay marriage legal in the state. The general consensus from live interview polling was that it would remain legal- one found the effort to overturn gay marriage failing 41/50 and another did so by a 42/53 spread. So when we came out the weekend before the election and found that voters were planning to invalidate gay marriage by a 51/47 margin, there was a lot of skepticism. But the final result was 53/47 to overturn the law, and it was another 3 years before gay marriage finally became legal in Maine.
9) Earlier this year there was a special election to fill an at large seat on the Washington DC City Council. We found Anita Bonds leading Patrick Mara and Elissa Silverman by 6 points each, which set off lots of discussion about how impossible it was to poll a special election in Washington DC. But in the end Bonds beat Silverman by 4 points and Mara by 7.
10) A week before the Democratic primary for Senate in Colorado in 2010, a poll came out finding Andrew Romanoff leading incumbent Senator Michael Bennet 48/45. We came back the next week and showed something completely different with Bennet leading by a relatively healthy 49/43 spread. Bennet won 54/46 and went on to win the general election as well, where we were generally the only pollster to show him tied or with the lead.
11) There were 21 polls of the Montana Senate race last year in the Huffington Post poll database. 11 of them showed Republican Denny Rehberg in the lead, as did the FiveThirtyEight forecast which found Rehberg favored by 2 points. We released 5 out of the 10 Montana Senate polls last year that didn't have Rehberg winning. Every single one of them showed Jon Tester up, by an average of 3 points. He won 49/45.
12) The outcome of the West Virginia Governor's race was described as 'highly uncertain' in 2011 because only one firm had polled the race. That was us and we found a toss up contest with Democrat Earl Ray Tomblin leading Republican Bill Maloney just 47/46. Tomblin won by a similar margin of 50/47.
13) There were 50 polls of the Presidential race in Colorado released in the last 3 months before the election last year. Only 6 out of those polls showed Barack Obama winning the state by 5 points or more- 3 out of those 6 were by us. Expanding the analysis to polls that showed Obama winning the state by 4 points or more, only 12 out of 50 polls conducted in the state fell into that category and we accounted for 4 of them. The final Huffington Post estimate found him leading by 1.7 points. Our final poll found him winning by 6 and he won by 5.4%. Good thing we weren't paying attention to the averages.
14) It had been established by a poll showing Republican Jane Corwin up by just 5 points that the 2011 special election in NY-26 was going to be closer than expected, but still no one expected Democrats to win a district they hadn't even been able to win in 2008 so soon after the 2010 landslide. But when we went in the field we found Democrat Kathy Hochul leading by 4 points, and she went on to win the election by 5.
15) Polling in July of 2008 had found Elizabeth Dole leading Kay Hagan by anywhere from 8 to 15 points in the North Carolina Senate race. So when our August poll found Hagan leading 42/39, the Dole campaign hit back hard, characterizing it as 'a bunch of junk.' Hagan led every single poll we conducted the rest of the campaign, most other pollsters started to find the same thing and Hagan ultimately won by 9 points.
Have we also gotten special elections and primaries wrong over the years? Of course. Those races are much harder to poll than even year general elections, and that's why you don't see many pollsters or forecasters taking a crack at them. But the simple reality is that if we relied on 'copying' other pollsters' work for our accuracy, we wouldn't have had all of these successes on things no one else was looking at. In all of these cases our methodology worked.