Exit Poll Echoes

First, the Social Science Resource Council (SSRC) has released a summary of the exit poll controversy named A Review of Recent Controversies Concerning the 2004 Presidential Election Exit Polls (pdf). It's informative, only 18 pages, and does not take sides - an excellent summary. It's a great way to avoid the monstrosity (pdf, 77 pages) that is Edison/Mitofsky's report on the same matter.

For those that want a super-brief summary, here is my take:

• The leaked exit poll data did not mean what the blogosphere thought it meant.
  
• Aside from that, the leaked exit poll data was flawed anyway.
  
• Edison/Mitofsky's response to the flaw was to say, "Well, it could have been because of different shyness levels among question respondents...".
  
• Edison/Mitofsky did not address fraud as a possibility.
  
• Fraud remains a viable hypothesis (nothing more).
  
• The hypothesis could be tested with the release of the raw exit poll data.
  
• Edison/Mitofsky is refusing to do so, on the grounds that doing so could reduce the effectiveness of future exit polling - this is a valid point, but it's arguable that it's valid enough to counterbalance the need to test the fraud hypothesis.
  

I personally support the release of the raw exit poll data. I wouldn't know what the hell to do with it, and I cringe at the thought of some of the self-professed statistical "experts" going through it, but I believe it would be worth it and would increase the public trust in our elections.

Second, many of you may remember a raging controversy among the progressive polling crowd, regarding the "weighting of Democrats versus Republicans". Polls would routinely come out showing Bush ahead of Kerry, and the criticism was constantly that the polling outfits simply asked too many Republicans what they thought, and not enough Democrats - which would inflate Bush's support and deflate Kerry's.

The problem is that some of the polling outfits would not release the party breakdown of their polls. That is changing. Mystery Pollster received this notice from Gallup:

As far as I know, Gallup has no history over the last 70 years of routinely posting the party ID composition of each survey we conduct, just as we routinely don't report ideology and a lot of other measures regularly asked in each survey. As noted, we send the party ID composition percentages to anyone who is interested (actually, we really don't get that many requests for them). But since this seems to be an area in which there is perhaps bourgeoning interest, we'll probably start posting them on our website for each survey, along with rolling trends and some explanations of how Gallup measures party ID and what it's significance is.

And this notice from Pew:

Given the evolution of the dialogue on the subject - for which MysteryPollster deserves a lot of credit -- and the greater understanding among political observers regarding the perils of weighting party ID to an arbitrary parameter (clearly illustrated by the party ID distribution on Election Day 2004), we will begin posting party ID and its trend in our toplines in future survey releases.

What's the irony about all this? Progressives were angry that the pollsters were not weighting their polls to historical Democrat/Republican percentages, where Democrats outnumbered Republicans. And then on Election Day, the breakdowns ended up being much more Republican than they were historically. So, before Election Day, the "Bush-biased" polls, while still biased towards Bush, were closer to the truth than we thought, anyway.

At any rate, the extra transparency is a good thing.