2 years, 10 months ago
(H/T Drudge Report)
This post will require a bit of set up, so a little patience, please.
Historically, political polling originated as a means for the public to gauge which candidate was winning at various stages of a political race. These early polls were crude efforts at asking gathered crowds which candidate they preferred with predictably poor results. The consensus seems to be that modern polling started with George Gallup in 1935 as an effort to apply a more scientific approach with random sampling involving personal interviews with voters in 300 cities. Because of the cost involved, newspapers and politicians did not conduct their own polls but relied upon the polling companies to inform them of public opinion. In other words, the purpose of the polls, according to Gallup himself, was to help politicians figure out what the public wanted in order to be more responsive representatives.
This approach continued until a major shift to the use of telephone polling in the 1950’s. The drastically reduced cost of telephone polling (which has its own inherent bias problems) enabled newspapers and politicians to conduct their own polls. It did not take long until the purpose of polling in the hands of the media and politicians evolved into an attempt to influence the attitudes and opinions of the public.
With this background in mind, consider the explosion of political opinion polls. The list of polling organizations– private, public and partisan– is long indeed. With this plethora of polling groups there is a similar variety of methods used to garner opinion: general adult population, registered voters, likely voters, sampling by partisan affiliation, automated polling, land-line polling, cell phone polling, internet polling, personal survey, and so on. The decision by the pollster on which methods to employ, the framing of the questions, the order of the questions, the choices in the demographics and sampling size… all of these and more factors can greatly affect the results of any given poll.
So last week we had a poll released by NBC/Wall Street Journal that purported to show President Obama with a six percentage point lead over Romney in the general election. Shortly after the poll was released, several bloggers did the necessary work to examine the choices that the pollsters made and how that affected the results. Here is just one of those take-downs from Hot Air:
The 2008 national exit poll sample, taken when Hopenchange fever was at its zenith, was 39D/32R/29I, or D+7. This one, after three years of Obamanomics dreck, is somehow D+11 if you include leaners and D+12(!) if you don’t. Anyone feel like taking these results seriously?
In other words, the pollsters decided to radically over-sample Democrats in their poll for no, apparent good reason. By comparison, the national party registration for Democrats and Republicans is now about even or slightly favoring Republicans. Add to that the recent Gallup poll that found a huge lead for Republicans in voter enthusiasm which would drive proportionately more Republicans to the polls and you can only conclude that this is a poll designed to influence public opinion rather than report it.
And this is not the only, such skewed poll. If you look at most of the polling being featured at Real Clear Politics (a purportedly neutral website for conservative and liberal news/opinion), most of the polls feature a similar, biased party sample, or sample only the general population or registered voters rather than likely voters. All of these polls show Obama and Romney either at a dead heat or Obama with a small lead.
Add to this a report in The Weekly Standard from last week that the Obama Campaign spent over $2.6 million on polling just in the month of June (compared with just $460,000 in April) and a pattern seems to be emerging: Obama is losing the voters and losing them badly. Unless he takes some dramatic action or events intercede to change public opinion, he is headed toward a sizable, electoral collapse.
Despite everything that most pundits– conservative and liberal— are telling you, I believe that Obama is headed for a resounding defeat. The spending by Obama’s campaign on polling is the result of panic. They cannot believe the numbers they are getting and they are polling and re-polling for any and every conceivable angle that would turn opinion in favor of El Presidente. Nothing is working, so far.
The hyper-skewed polling by NBC/WSJ is also telling. Clearly this poll was designed to push public opinion in the direction of Obama by giving the impression that he is pulling away from Romney in voter preference. This is not surprising in itself. What is shocking is that NBC/WSJ had to resort to a sampling that favors Democrats by 12 points over Republicans in order to get the numbers that they wanted.
Contrast this with the daily tracking polls of likely voters from Rasmussen Reports that had Romney ahead of Obama by 5% last week. For an incumbent like Obama to be trailing the challenger at this point in the election– before the public has focused on the race and before the convention and debates– is an indication that Obama is getting all the support he is going to get. He is already maxed out and there is nowhere to go but down.
Obama’s only hope right now is to conceal his tenuous position from the public in order to avoid the “Bandwagon Effect.” This is a well-known phenomenon in polling in which voters who have not yet decided or are not strongly committed to either candidate are strongly influenced by polling which indicates that one of the candidates is pulling far ahead in the race. These voters want to be part of the winning team, so to speak, and throw their vote in with the majority. So long as Obama and his allies in the Statist Media can manipulate the polling to maintain the appearance of viability, they can hope for a Romney implosion or some, other intervening event to save the election.
Along, then, comes an article in The Hill that attempts to do this very thing. Using the ridiculously skewed demographics, the article makes use of the NBC/WSJ poll as follows:
Despite voters’ worries about the economy, they continue to give Obama the edge on personal popularity. In a recent NBC News/Wall Street Journal poll, Obama led Romney by 20 points on the question of whether voters liked each of them on an individual level. Two thirds said they liked Obama, whether or not they disagreed with his policies, while just 47 percent said the same about Romney.
Just 35 percent of voters held a positive opinion of Romney overall, with 40 percent negative, while 49 percent had positive opinions of Obama and 43 percent felt negatively about him. Both candidates saw slight upticks in their negative numbers since the ad onslaught began.
The article also mentions the Real Clear Politics average of polls for the idea that the candidates are locked in a statistical tie, a tie that is only made possible by the inclusion of polls that use unreliable or skewed data to give Obama a lead.
But even these partisan polls can only mask the reality for a limited amount of time. If Obama’s numbers continue to slide and it becomes too obvious to skew the numbers in order to make the race look competitive, expect to see even Leftist polling groups grudgingly showing a Romney lead. My personal prediction is that by September, absent an unlikely implosion by Romney, we will start to see a snow-balling of public opinion where the public has finally tuned in to the race and begins breaking for Romney. Once that starts to happen the Bandwagon Effect will take hold and the bottom for Obama will fall out with only the hard-core, 35% of Democrats voting for him. At that point we could be looking at a rout similar to the 2010 Elections, thus ending our national experiment in Mass Insanity.