Witches' Brew: Media's Deliberate Misuse of Polling Data Changes Elections

I do not know who will win Tuesday's elections. But, I do know that the media deliberately misuses the "information" they get from polling data and treat it as news. And, I believe that that misuse can fundamentally change outcomes. With the exception of Fox News, which is essentially a media organ of the Republican party, the major driver of this deliberate abuse is to attract and retain viewers.

They did it in 2007-8. They are doing it again in 2010. They will again next cycle.

Why? Because to report the truth about the polling data would not serve the media's interests in gaining viewers. It would also force them to spend some money to do some real reporting.

First, 2007-8. Let me say that I report this as a supporter of the President and Secretary of State Clinton. Let me also say that neither had anything to do with what I am about to report. It was all done by the media, for the media.

If one recalls the reporting on the polling "news" about the Democratic candidates prior to the primaries, the numbers were said to be something like this: Clinton 40%, Obama 28%, Edwards 15%, Richardson 6%, and the rest at 3% or less. And, of course, as they changed this week or that, and Obama gained, or Edwards fell, or Richardson briefly cracked 10%, it created more news stories.

In fact, the only story is that there was no story. At all.

The real numbers from those polls were quite different. Undecided 80%. Clinton 8%, Obama 4.6%, Edwards 3%, Richardson 1.2%, and the rest less than 1%. Same ratios among the candidates, wildly different implications.

So, why did the media ignore the key fact about the primary race for president, that the overwhelming majority of potential primary voters were undecided for most of the pre-primary season? It is a question that answers itself: the real data are not really "news". What can the pundits say about the real information? "Well, people are just waiting to know the candidates better?" Or, the truth, "people are just not engaged enough to have an opinion?".

This is not a story that would attract and retain viewers. Indeed, it is not even a story.

The obvious remedy was for one of the candidates below 1% in either report to point out that most people were undecided. Candidate Joe Biden, for one, did. But it never got much traction because a) the news media would not report it, they had their "40/28/15/6/below 3" stories all neatly typed; and b) when one is way down in the polls, by any measure, the comment seems gratuitous when one makes it himself. If a scrupulous media outlet (a modern oxymoron if there ever was one) had reported it, it might have been different.

So, the media created the false impression that there were 3 candidates so far ahead of the rest of the pack that it would be foolish to fund the others. And, few did.

Moreover, whose surrogates got free air time? Clinton, Obama, Edwards. The rationale was that the others were too far down in the polls to be of interest to viewers. An interesting rationale if it bore a scintilla of relationship to the truth.

This polling misuse was done deliberately. When I confronted them with the truth, they acknowledged that they realized what they were doing, and continued to do it anyhow.

In my personal opinion the best candidate emerged anyhow, but that is not the point. It is that the media has its own agenda to grab and retain viewers, and thus the contest was jiggered to suit those needs. It also obviated the need to spend money on real reporting.

Next, 2010. This year. When the news media reports polling data on voters' preferences, how many times do they qualify the information by describing the inadequacy of their methods? My count is zero.

Why? The same reason. If they acknowledge the polls are flawed, their "news" value vanishes. If they are not "news", what in heaven's name could they yap about?

Yet, the methods are highly suspect. For one thing, most voters under 30 do not have landlines. There is -- thankfully -- no cell phone directory. Hence, even if they were sampling accurately in other ways, which they are not, the inability to include this cohort in their findings impugns the results' integrity.

Then, then there is the automated survey poll. I just answered one. I gave them as much misinformation as possible (I am not even voting in the state in which the phone rang) -- my own little 'rebellion' against the system. Had I not decided to do this for fun, I would have hung up. A key part of an accurate statistical analysis is how many calls have to be made to get your sample. If many more calls are required to get the numbers, how valid is the result for the population as a whole -- do the hanger-uppers have the same range of views of those who participate? Who knows? Do you ever hear that reported with the poll "information"?

Another flaw has been the questions asked, a well-known bias in public opinion polling. For example, for months the media all reported (they report the same polls) that 59% of the American people are unfavorable toward "Obamacare". That is the mantra continually referenced. Just today, Chris Matthews stated that Democrats post-election could not gain traction on health care because people are against it.

Well, Chris did not bother to look at the next question, and it turns out that 30% of that 59% are unfavorable because they wanted the reforms to go further! So, Matthews' conclusion was utterly false -- but it is the conventional wisdom that the media repeats. It is most likely that "Obamacare" is viewed not as a policy but as the 12 month, drawn out process, with compromises and deal-making galore. Do they want children with preexisting illnesses covered? Well, yes, that is a great idea. How about no dropping of your insurance when you become ill and actually need it? Not a bad idea either.

The absence of cell phone only users, mostly the 18-30-year-olds, from the polls is a methodological flaw. If one assumes that cohort's turnout to comprise 6% of the voting population, and they break 2:1 for the Democrats, that is an additional 2 points for those races. If it comprises 9%, then it is an additional 3 points for those races.

If these races are close, as many are, the youth vote could make a difference.

Will the media's reporting of polls that exclude the young voter, but never acknowledging it, reduce those voters' enthusiasm for voting?

Probably. The young, especially, have to feel that they can make a difference in order to rouse them to action. Thus, polling's flaw plus media's pursuit of its own agenda instead of its solemn duty to find the truth and report accurately produces a skewed outcome.

And, that could be the difference between a disastrous next 2 years and a 2012 campaign devoted solely to convincing people that the others are to blame for our demise versus a difficult but hopeful future that might just have been starting.