Of Landslides & Preference Cascades

by JASmius

Mitt Romney is going to win....the popular vote.  Mitt Romney will probably win the Electoral College.  Either way, contrary to the burgeoning wishcasting on the Right, he is not going to win in a landslide.

Not to be a killjoy, but that's the reality.  We would do well to remember the incontrovertible reasons why.

1) Polarization.  This is still 50/50 nation.  When I think of landslides, I think of Johnson-Goldwater in 1964, Nixon-McGovern in 1972, Reagan-Mondale in 1984.  60%-40% popular vote blowouts.  Electoral Vote totals near, at, or over the 500 mark.  The days when Democrats weren't radical, extreme, hyperpartisan zealots and were open to bipartisan voting in significant numbers.  Those days ended a long, long time ago.  Since 1984, no popular vote winner has exceeded 54%, and that was Bush41 four years later.  Barack Obama's 53% was the largest total since then, and that was the biggest Donk wave year of my lifetime (if you exclude my first four days outside the birth canal).  Unless the term is to be redefined into meaninglessness, landslides are no longer possible on the national level.

2) The polls don't show it.  I'm not talking about the RCP average (which today has Romney up 1).  RCP includes all the polls, all but one (Rasmussen) of which are Donk-oversampled garbage.  They even include officially partisan surveys like PPP, which they never used to do.  I'm talking about Rasmussen, the gold standard of polls over the last three presidential election cycles.  And as of today, Rasmussen has Romney up 2.  Two points is not a landslide in the making.

However, I am constrained to point out that even Rasmussen has succumbed to the temptation to oversample Dems recently (moving their sample weighting from D+3 to D+5).  Which brings me to Dave In Florida's Poll Analysis, a blog that does what I used to have time for: drill to the core of multiple polls and adjust the samples to more realistic proportions.  Per Dave's analysis yesterday, here is The RCP average under the following turnout scenarios:

O+0.42% - Current RCP Average
O+0.81% - Average using the 2008 turnout model
R+1.94% - Average using the D+3 turnout model
R+4.14% - Average using the 2010 turnout model
R+4.20% - Average using the 2004 turnout model
R+5.88% - Average using the Rasmussen Party ID turnout model

Even if the turnout mirrored the Rasmussen Party ID (R+2.6), which is highly unlikely, Governor Romney's popular vote margin would still be a point and a half less than Obama's was four years ago.

Or, as I said above, Mitt Romney will win the popular vote; but he will not, because he cannot, win in a landslide.

We're as vulnrerable to drowning in our own existential cloture bubble as is the Left.  We would do well to maintain the grip on reality that they heaved overboard long ago.

And part of that reality is the very real likelihood of voter fraud on an unimaginable scale.

I don't have the poll-spelunking time I used to, but I have created a spreadsheet that takes the state by state popular vote numbers from 2008 and adjusts them equally by the difference between the 2008 national popular vote margin and the current Rasmussen margin as well as splitting the difference between the Rasmussen Party ID and 2008 turnout models (which yields a very realistic D+2.2 sample), then projects the popular vote result assuming undecides break 2-1 for the challenger (Romney), and finally adjusts the margin two points toward Obama to account for the ACORN factor.

As of today, that projection indicates Mitt winning 50.8% to 47.8% nationally.  But the Electoral College result is Romney 266, Obama 263, with Colorado a flat-footed tie.  That would make the Centennial state the Florida of 2012.  Given that Dems run the show there, I'm less than sanguine about Mitt's chances of averting having the election recounted away from him.

But leave aside those hypothetical specifics.  The bigger point is the popular vote theshold Governor Romney needs to clear to be reasonably assured of going over 270 Electoral Votes.  If he doesn't win by at least three points nationally, the likelihood is that Team Messiah will cheat themselves over the top in enough swing states to steal the election.

All of the above stipulated, there does appear to be a preference cascade underway.  If I had to prophesy, I would soothsay that we're looking at a scaled-down version of 1980, where the race was neck and neck until the GOP challenger pulled away at the end.

Bottom line: This is a pipedream; but this will be more than sufficient.


[cross-posted @ Hard Starboard]