Tuesday, June 24, 2008

BounceWatch update

UPDATED & BUMPED: The latest Gallup daily tracking poll shows no change, Obama 46, McCain 43:
This is the fifth straight day that neither candidate has held a statistically significant lead, although Obama has consistently polled a slightly higher number, as he has since the start of June. On this long-term basis, it seems clear that Obama has a significant, albeit slight, advantage in the race.
Which is to say, the 2%-3% lead is small, but because it has shown up consistently, it can be considered a statistical fact.

Still, the big story here is what has not happened. In the three weeks since the last Democratic primary, Obama has not surged ahead of McCain. The "bounce" Obama got from Hillary's concession peaked June 9 when he led by 7 points over McCain (48%-41%) and has since subsided. The general election campaign begins, then, with the two candidates in a virtual dead heat.

Today, the RealClearPolitics average has Barack Obama leading John McCain by 6.6 points. It is worthwhile to note, however, that the RCP average currently includes two recent polls of dubious merit:
Subtract these two outliers, and Obama's lead in the RCP average shrinks to 4 points. Of course, June polls aren't predictive, so it doesn't really matter at this point. The MSM, however, is reporting these results in such a way as to portray Obama as an unstoppable juggernaut in a transparent attempt to create a bandwagon effect for the Democrat.

UPDATE: The MSM's misrepresentation of poll results is nothing new, of course, and neither is my resentment of it. Two explanations are possible, and neither is flattering:
  • Editors and reporters don't understand polling, and thus ignorantly misrepresent the results; or
  • Editors and reporters are purposely misrepresenting the results of their own polls.
The most important cause for skepticism toward political poll results is the fact that we don't know whether any given poll has surveyed a truly representative sample of the electorate. This is especially true of polls taken far in advance of the election. To look at a June poll as predictive of the November result is to assume that nothing that happens in the next four months will affect voter preference. Yet, even as a "snapshot" -- a transitory impression of voter preference at this moment -- polls have profound shortcomings.

Voters fall into four basic categories: (a) loyal partisan Democrats, who always vote Democrat no matter what; (b) loyal partisan Republicans, who always vote Republican no matter what; (c) dedicated independents, who always vote, but have no strong partisan leanings; and (d) episodic voters, who do not always vote.

Given the fact that huge numbers of eligible voters don't vote, a pollster -- if his poll results are to be useful or credible -- must try to screen for "likely voters." This is a doggone difficult thing to do, but it must be attempted, because voters and non-voters differ significantly in their preferences. Non-voters are more likely to support liberal policies and Democratic candidates (a source of endless frustration to liberal Democrats). So a poll that doesn't properly screen for "likely voters" will always skew leftward (as was true of the Newsweek poll that surveyed "registered voters" rather than "likely voters").

This is probably why early polls have historically overstated support for Democratic presidential candidates. The closer you get to Election Day, the easier it becomes to determine who the "likely voters" are. Thus, the samples in early polls contain lots of liberal-leaning eligible voters who, in the end, won't actually bother to vote.

Whatever other biases one may attribute to MSM, they are definitely biased toward inflating the importance of these polls they commission at such tremendous expense. Skepticism is merited.

1 comment:

  1. Actually, the Newsweek poll purported to sample the opinion of "adults," not even "registered voters," much less "likely voters." As far as I know, "adults" includes non-citizens and convicted felons, who cannot vote at all. In addition, Newsweek is notoriously biased in favor of Democrats in general, and even more so with respect to the Messiah. In short, the Newsweek poll is a piece of garbage, not treated seriously by anyone who knows anything.

    Despite all the gloom and doom, McCain is actually doing quite well in the polls. He is outpolling the party "generics" by a ton, and is outpolling the Congressional Republicans by a good amount as well. In addition, after 2 straight GOP administrations, the Democratic presidential candidate, historically, should be way up in the polls (like Carter was in 1976,Dukakis in 1988, and Bush in 2000), but Barky is not.

    In some quarters, the urge to massage the data is overruling good judgment. Take, for example, the much celebrated and lauded "Pablano" of website FiveThirtyEight, and long time Obama supporter and "diarist" on Messiah-worshipping blog, Daily Kos. Up until quite recently, he had the race fairly even, with his actual numbers giving McCain the nod in popular and electoral votes, and only his "fudge" factor giving a slight edge to the Messiah. Within the last couple of weeks, however, Pablano-Boy has altered his method drastically, in a transparent effort to produce the kind of 300+ EV total prediction for the Messiah that has become an article of faith on such Messiah-worshipping sites as Daily Kos and MyDD.

    With the realization that, between the Federal funds and the RNC money, McCain is going to have quite enough cash, thank you, to rebut the horseshit propaganda of Mr. Charlatan added to the mix, this election is shaping up to be anything but the walkover that the brain-dead cultists over at Obama-nation expected.