There has been a huge amount of conversation going on over the last couple of weeks, at least in Conservative circles, on the validity of the polling that is being reported on and commented on in US media for the Presidential race.
The contention is that the polling data is flawed based on bad, faulty or intentionally biased metrics and that the media – rather than be a watchdog and either call pollsters on these perceived flaws or research, report and debunk Conservative claims – are, instead, merely reporting polling numbers as if they were gospel.
Given most national polls are favoring President Obama, Conservative concerns are understandable. If bad poll results discourage Conservatives and Republicans from turning out to vote, the President might win a close race.
Conservatives contend if polling were being conducted professionally and in an unbiased fashion, Romney’s numbers would be better, perhaps significantly so. The contention is that due to media bias, reflected in both reporting and results desired by media company polling, the media is intentionally reporting misleading results derived from bad sampling in order to help Barack Obama win.
The question is, of course, is there any proof?
Most of us don’t subscribe to pollsters results to have access to metrics and details. The rest of us might not understand what we were reading if we did or the explanations from those who do. What is needed is something easily grasped which demonstrates both a media bias and the involvement of pollsters, at least media pollsters, in the mix.
The Washington Post may have provided exactly that.
Today’s Nashville Business Journal ran a story that has the Washington Post reporting that this year, the “…National Election Pool — a joint venture of the major television networks and The Associated Press…” is dropping Exit Polling data from 19 states, including Tennessee. This is a break in a 20 year practice of including exit polling data from all 50 states.
The bias comes in when you look at what states are being excluded and the graphic that the Post includes in their own piece.
Of the 19 states excluded from Exit polling, 16 of them are considered firmly for Romney. This represents 135 of the 170 electoral votes the Post reports as solidly for Romney. Only 5 states considered solidly in Romney’s camp will be included in exit polling. Texas, the GOP’s biggest prize, will not be included.
Just three states – Delaware, Hawaii and Rhode Island – are excluded from the solidly Obama column. They represent 11 of the 196 electoral votes the Post attributes to Obama. There will be 10 states from the Obama camp with exit poll data reported on election night. The three largest prizes – California, Illinois and New York will be included.
Exit poll reports will come from 16 undecided states, 10 states in the Obama camp and just 5 states from the Romney camp. What viewers, listeners and readers will hear is that state after state has decided to re-elect Barack Obama. Again, 16 of the remaining 19 states from which there will be no exit poll reporting would favor Mitt Romney.
Further, as the majority of Obama’s states are in the Northeast and many of the excluded Romney states are in the South, exit poll reporting will be of Obama sweeping states in the Eastern and Central Time Zones with no news at all from Romney states. This will have interesting implications for voters in Central, Mountain and Pacific Zones.
The media will be able to accurately and truthfully report the results of the exit polling as favoring Obama. But they will not likely report that their own media based consortium has skewed the participants in that polling in favor of the President.
I cannot find any way to twist this to make it fair. I cannot spin this as impartial and unbiased. I’m open to explanations. Please tell me what I’m missing and how I’m wrong.