Wednesday, October 14, 2015

Did the Polling Agencies Take the Night of the First Democratic Presidential Debate Off?

Something is missing from the post debate analysis.

I watched the first Democratic Presidential debates of 2016 and thought both Bernie Sanders and Hillary Clinton did very well. Both exceeded my expectations, Hillary by a little larger margin and I think she won the debate. I posted on my Facebook page my prediction about how the major polling agencies scientific polls would look:
If I had to guess I would say 55% say that Hillary won, 40% say Sanders won, 5% say O'Malley won.
In past debates, the major polling agencies usually had results within an hour or two of the end of the debate. While I was waiting for those results, the online polls sprang up all over the place. To the Sanders folks' credit, they were all over those polls, posting links to them on the Sanders reddits and on Social media and driving up the results for him to ridiculous margins (Some had the results as high as an 85% victory margin for Sanders. It's possible he won the debate in the eyes of debate watchers but not by that much). But as we all know (or should know), online polls don't mean a whole lot and the ability of Sanders' supporters to manipulate these polls shows why.

On the other side of things, my fellow analysts and pundits were almost all proclaiming not just a Hillary Clinton victory but a victory by a large margin. While that felt nice to me as a Hillary supporter and perhaps means slightly more than an online poll for its effect on those who watch or read them, it's still not a scientific measurement of how regular people felt who saw the debate.

So I waited, and waited and waited and finally I realized that no real polls were forthcoming. This is all the more surprising to me because the host of the debate, CNN, has an impressive polling center that could easily have done this. 

Scientific polling is important because the pundits have been wrong about debate perceptions before. I remember watching a number of Presidential debates where the prevailing assumption among television punditry had been that one candidate won and when the polls came out it turned out the public thought another candidate had won. I think this happened once or twice with President Obama's debates with Mitt Romney and John McCain. Most of the time the punditry gets it right but not always.

Perhaps as they day or week progresses we will see some scientific polls released but I am not hopeful. It seems we will be left with meaningless online polls, what the talking heads believe and, egads, our own opinions.