After weeks of polls showing that the Colorado’s governor’s race was either too close to call or leaning favorably toward incumbent Democratic Governor John Hickenlooper, a recent Quinnipiac poll showed Hickenlooper trailing Republican challenger Bob Beauprez by 10 points. Predictably, Hickenlooper’s supporters sprang into action, mainly by attacking the credibility of Quinnipiac’s polling.
But Democrats may, in fact, have cause to be concerned.
Other data – generated by the new platform Trendency – shows that Hickenlooper’s support may be weaker than what conventional polls are finding. In particular, we find that fewer of Hickenlooper’s supporters are consistent, definitive supporters and that the strength of his support has been eroding steadily since summer.
Typical polling – be it a stand-alone poll or a poll aggregator – only allows for the interpretation of a voter’s winner-take-all choice between available candidates. Trendency, on the other hand, gives voters the opportunity to demonstrate their uncertainty in a numerical way by allocating their percentage of support for each person in the race. This methodology enables us to track how Americans’ opinions are shifting from day-to-day, rather than simply taking a snapshot of a small sample of voters at one moment in time.
Among the advantages of this approach is the ability to measure the strength of a voter’s commitment to a particular candidate. In Kentucky, for example, we found that the strength of support for Democratic challenger Alison Grimes’ might be stronger than some polls are showing – if only because neither candidate has much room to cast his or her own opponent any more negatively than they already have.
In Colorado, we used a related analytical tool – which we call “thresholds” – to measure the confidence level of a voter’s support for a particular candidate.
While a traditional poll will treat two people who say they plan to vote for a candidate as exactly the same – even if one may be a consistent, definitive supporter while the other may be weak and wavering – this measure allows us to focus on and parse out different levels of support. The higher the threshold level, the stronger and more consistent the support among those voters.
In Colorado, this measure shows that fewer of Hickenlooper’s current supporters are confident and consistent in their choice. Where only 31 percent of Hickenloooper’s supporters are at the 90 threshold as of September 14, 42.5 percent of Beauprez’ supporters are at the same level. At the 75 confidence threshold, Hickenlooper looks slightly better, but overall Beauprez’ supporters are more decisively in his camp.
Share of Colorado voters showing “threshold” confidence levels of 75 and 90 in candidate support as of Sept. 14
Moreover, a look at trends in this threshold data over the past several months shows that Beauprez is gaining in his strength of support as Hickenlooper ebbs. One can speculate why this pattern has emerged, but whatever the reason, Beauprez has strengthened his base while Hickenlooper has seen his core support soften.
Share of Colorado voters showing a “threshold” confidence level of 50 or higher in candidate support
If Hickenlooper does win, it is likely to be with weak support, likely resulting from party affiliation and incumbency – but not with a clear mandate from the voters of Colorado.
Stefan Hankin is President of Lincoln Park Strategies, a public opinion research firm based in Washington, D.C.