Newsvine launched a distributed presidential polling widget called ElectionVine, which is a very clever idea with a lot of potential — except that it won’t ever provide any insight into who is going to be elected president in 2008. I’ve worked with folks at The Hotline, Hart Research Associates (who co-runs the NBC/WSJ poll) and many others who know A LOT about political polling, which means I know just enough to be dangerous.
Here’s why ElectionVine won’t predict any elections, even if it gets embedded across the entire blogosphere:
The 2008 presidential election is not going to be decided by Internet users, who remain still just a subset of the total U.S. population, and it certainly isn’t going to be decided by people who visit sites like blogs that embed third party widgets.
How many people over 65 do you think will ever see the ElectionVine widget? They are MUCH more like to vote than the people who will see it on Facebook.
Ask any professional pollster and they will tell you that regardless of how democratized the Internet has become, online polls are still unreliable predictors because they are not representative of the entire voting population, who actually decide elections. That’s why professional pollsters still use that old fashioned technology that is still in nearly every home — the land line telephone.
The ElectionVine widget also doesn’t ask for two critical pieces of information that election polling requires:
- Are you a registered voter?
- Are you a likely voter? (This isn’t a literal question — there are many models for determining this.)
That all said, ElectionVine is still a cool idea with well executed technology, and could be a very interesting data source, although not as a predictor of who will actually win the election or which candidate is actually ahead.
Instead, it can be used by sites to show the political affiliation of their users — an intensely social feature for the social web, especially in an election where politicians will try to use and abuse social media in every possible way. All that’s missing — and here’s an idea for you Mike — is national polling data to benchmark each site against. For example, you can find the latest national presidential polling data at PollingReport. Here’s the latest Gallup Poll on democratic vs. republican unnamed candidates (note the question is included, because pollsters know that how the question is asked affects the results):
![Gallup Democratic Primary](https://s3.amazonaws.com/publishing2-images/Gallup Democratic Primary.jpg)
![Gallup Republican Primary](https://s3.amazonaws.com/publishing2-images/Gallup Republican Primary.jpg)
Now let’s compare that to the ElectionVine results from TechCrunch (which should be easily embeddable without my having to do a screen capture):
![TechCrunch ElectionVine Results](https://s3.amazonaws.com/publishing2-images/TechCrunch ElectionVine Results.jpg)
Contrary to what TechCrunch readers would have you believe, Barak Obama and Ron Paul have NOT locked up the nomination. That said, the data is still REALLY interesting — it speaks volumes about the TechCrunch community and is a great way for TechCrunch readers to go beyond tech and share their views on politics.
The lesson here is that what’s missing in Web 2.0 is the merging of innovative technology with subject matter expertise. The ElectionVine widget should have been DEVELOPED in partnership with a professional polling company, but it’s certainly not too late to put such a partnership in place.
Who knows — in the hands of a professional pollster, this widget might actually be adapted so that it has some predictive power or other utility in the world of political polling.
(Mike: Ping me and I’ll give you some suggestions on who you might partner with.)
Of course, I’d be remiss if I didn’t put a political bias meter here on Publishing 2.0 — have at it: