Graham Firchlis has a blog

2010 election prognostication

Posted by grahamfirchlis on November 2, 2010

Pounding a stake in the sand with my crystal balls.

Last winter I ventured that the Dems would hold their own in the House, plus/minus a few seats, and keep their majority in the Senate intact with the chance to possibly pick up a seat. But that was when I still thought there was some semblance of decency left in a few congressional Republicans, and they would allow passage of a stream of small yet functional stimulus bills to help put the nation back to work.

Alas. It turns out there are no decent Republicans.

Worse, for reasons inexplicable to me apparently driven by delusion, irrational fear and a reptilian logic imbedded deep within the human brainstem, many voters have decided to blame the recently arrived and overwhelmed Democrats for thirty years of systematic Republican cruelty. That’s like punishing the firemen for water damage while continuing to befriend the criminal gang that stole everything you owned and set your house on fire. It just dosn’t make any sense.

Or maybe in their minds it is all about having “one of them” in the White House, who isn’t a servant – hard to tell. Apparently I have once again over-rated the common decency and good sense of otherwise apparently intelligent elected Repubicans and a significant cross-sectional share of the electorate, Right, Middle and Left alike. Just when you imagine that after 100,000 years people as a whole couldn’t possibly act more self-destructively….

But prospects likely aren’t as bad for the Democrats as the media and the pollsters and the Republicans want us to think, because the polls that all the current Republican whoop-and-holler relys on are based on a biased voter selection mechanism – the land-line telephone. Modern times, and with significant recent acceleration, the young and the technologically comfortable are moving away from landlines and going it more and more alone with cellular devices. (I’m 63, only modestly technologically adept, wouldn’t twitter if paid to do so, and haven’t had a landline in years.)

Yet pollsters still rely primarily on land-line calls, and even when the more adventurous polls do try to include cell phone users they almost always substantially under-represent them. This creates what is called “coverage error” and can seriously warp the poll outcome. Nate Silver, no less, has substantial concerns and no clear answers.

This particular “coverage error” matters for interpretation of the pollster’s accuracy because the young and the technologically capable of all ages tend to be more Liberal than average in both their thinking and their voting. The polls this year, as they have been for some time, are generally skewed towards recording opinions of Conservative/Reactionary land-line users and away from the intentions of their more Liberal cell-only counterparts. Compare, for instance, the results from two recent major polls. The traditional dominantly landline approach of McClatchy-Marist yielded results consistent with most other pollsters, giving Obama a net negative rating of -4% while showing a massive tilt towards Republican voter enthusiasm, 51% versus 28% for Democrats. A Newsweek poll that deliberately included a reasonable proportion of cell-phone only users provided quite different outcomes, with Obama having a net +14 approval and the “enthusiasm gap” gone entirely.

Hard numbers are difficult to come by, but my back-of-the-envelope pencil-metrics put the effect at 2% to 4% underestimation of Democratic-likely voters depending on the urban density – the better the cell phone connection pattern and the tighter the economic squeeze, the more people decide to dump their increasingly anachronistic landlines.

While pollsters do try to make corrections for this kind of bias based on prior responses, their correction algorithms are restricted by historical data which may no longer apply. This is especially true in subject areas with a rapid change in the demographic where adjustments must be made. The shift from landline to cell-only is one of those, with a rapid and increasing rate of conversion.

Currently, a quarter to a third of Americans are cell-only or essentially so and thus outside most pollster’s reach, and that percentage is increasing at an accelerating rate. When a quarter or more of your intended subjects are unavailable, a quarter-plus that differs significantly in makeup and attitude from the residual sample and that divergence is accelerating, any “correction” that might be applied falls more and more under the heading of Wild-Assed Guess.

The other aspect that has been misinterpreted is voter turnout. Anger will motivate, certainly, and angst will de-motivate, but the GOTV effort on the part of the Democrats still has a great deal of power left from two years ago. I’ve talked to a lot of folks who are just as active as in 2008 and some who are even more so this cycle, and the massive Obama-voter-apathy we’ve been told so much about doesn’t seem to me to be as big as is reported.

Add another 2% or so to the poll-predicted Democratic vote totals from this effect.

Combined, and hedging a bit as any decent prognosticator must do, I’ll predict that in general the Democrats as individual candidates will do somewhere from 3% to 5% better than the polls now suggest. Since there are a lot of races where Democrats trail within that range or are holding even, my analysis suggests that there will be a lot more Democrats in office come November 3rd than the MSM, the Republican Party and Nate Silver are predicting. I have Harry Reid being re-elected by 4,000+/- votes and the Democratic caucus holding the Senate 55 – 45, while the Democrats retain control of the House 222 – 213.

Or I could be wrong, and Republicans win everything still hanging in the balance. That would be bad, but from an immediate strategic sense perhaps won’t force much of a change from what already needed doing. Looking forward, every conceivable electoral outcome leaves those of us on the Left in about the same position we’ve been in for the last two years:

Wondering what we might do to make an impact for real Progressive change in the face of a deadlocked – and thus inherently Conservative – Congress, a systematically-obstructed, mostly-well-intentioned but agonizingly-deliberate-to-the-point-of-being-tentative president, massive corruption of the media, a Reactionary grip on the judiciary, all while climate change comes down upon us like all the hounds of Hell and the ascendant power within the Republican Party believes that an Apocalypse is a desirable event. Big picture, our challenge appears to be more systemic than any one general election.

Come Wednesday morning, bright and early, those of us not permanently gripped by chronic depression need to start a serious discussion about what to do going forward that can plausibly succeed. I look forward to getting past the election hype, and back to work on the really hard stuff.


One Response to “2010 election prognostication”

  1. […] Comments (RSS) « 2010 election prognostication […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: