by Paul Goldman
Dick Morris, as you probably know, has made a living as the Obama-bashing former Clinton aide turned Republican/Conservative campaign stalwart. Recently, Morris has been getting more than a little help from his friends in actively promoting his poll showing Romney winning by 7% percentage. It is the “it” poll on the conservative Internet right now.
Mr. Morris is quite the self-promoter, and one has to admit to his unique ability to morph from lefty liberal Manhattan Democratic consultant to his current perch as the go-to guy on conservative television talk shows. No small feat. Moreover, Morris is a very knowledgeable and highly competent political guy, one has to concede this as well. Dick has game.
But it’s very lame relative to his new poll which he promotes as the only honest one in the business right now, he basically accuses the “mainstream” polls as having an anti-Romney bias right now.
n slightly more than two months, we will know the truth or perhaps not, anything can happen in this presidential contest.
However, in terms of what we know for sure today, we can be absolutely certain that Mr. Morris has cleverly, and grossly, manipulated his poll. Remember, there are no laws to be broken here: anyone can be a pollster or hire someone to do a poll. So it is “profession” loosely defined. As for Mr. Morris’ poll, it is, as I say, cleverly designed.
Manipulating a poll to get better numbers for Romney isn’t all that hard in this climate. For example, in 2008, the national exit polls of the electorate found 39% identifying as Democrats, 32% as Republicans. In the Morris poll, it is 33% DEM, 31% Republican. Or in pollster jargon, it goes from a +7 DEM to + 2 DEM, or a shift of 5 percentage points.
That’s huge: but is it wrong? No one knows. In 2004, the exit polls said the electorate had as many Democrats and Republicans. The point being: Given the incredible partisanship right now, where there is almost no cross over voting, the 5-percentage-point shift in the Morris poll basically explains why his poll has Romney up +7, while other polls, whose turnout model is more like 2008, would have a statistically significant difference in result.
Likewise, the Morris poll reflects a significantly lower turnout among minority voters than in 2008.
Right now, there is no need to do anything but get pollsters to spend their time trying to determine how best to project the most accurate turnout model for this coming November. It is all about turnout right now, not persuasion. The folks left to persuade appear to be largely unhappy with both major party candidates. This may change: but these folks are also the hardest to reach in terms of getting them new information and in terms of believing their vote really matters in the final analysis.
But those who say they are likely to vote but undecided will vote in decent numbers: and depending on the percentage of Democrats and Republicans who vote, these non-aligned folks could easily decide the presidential election.
Logic suggests that Mr. Morris has decided to lay down a marker for being the most pro-Romney pollster out there: his + 2 DEM electorate with underrepresented minority voters strikes me as at the far edge of the boundary of responsible polling. It is the most pro-Romney one could get without stepping over the line.
A +2 Democratic electorate in 2008 would have resulted in a narrow Obama victory. In 2012, a movement of independent voters toward Romney within a + 2 Democratic electorate raises all kinds of possible election night outcomes given the Electoral College.But + 7 for Romney doesn’t compute, even with the Morris assumptions. However, a win is a win.
The bottom line: Morris is manipulating the top line numbers by playing with the internals of his electorate, skewing it as far pro-Romney as he can without stepping over the line. Morris also cleverly timed his poll to allow results to be released as the RNC convenes. You got grin at his play here, like I say, the boy knows how to game the system.