Jeffrey D. Sadow is an associate professor of political science at Louisiana State University Shreveport. If you're an elected official, political operative or anyone else upset at his views, don't go bothering LSUS or LSU System officials about that because these are his own views solely.
This publishes five days weekly with the exception of 7 holidays. Also check out his Louisiana Legislature Log especially during legislative sessions (in "Louisiana Politics Blog Roll" below).
How badly Gannett News Service wants Democrat Rep. Charlie Melancon to wrest incumbent Republican Sen. David Vitter’s seat from him and/or at least to have something interesting to write about became evident in a puff piece about the contest.
In the parlance of the media, reporter Mike Hasten produced a “second-day” story the theme of which asserted that Melancon was “competitive” in the contest after looking selectively at some results from an internal poll conducted by the campaign. Several stories with more general information about it appeared last week at this time. This kind of story is different because, unlike when fresh news comes about and is reported with some general information, since it is some time after the fact – for whatever reason Hasten did not initially report on the release of the information – a new angle that does not concentrate on the general information must be found and presented.
That’s why the theme was “competitveness” – although, as previously noted, if the brightest thing that could be argued about the Melancon campaign’s status is he has a chance to win, that’s about as bleak a forecast as possible without actually conceding the race. What the story decided to emphasize and not, however, is as big a story.
A very curious aspect about it all on which no media have reported is it’s the releasing of a campaign’s internal poll information. This is seldom done both because it risks giving away potentially useful intelligence to the opposition and the public usually reacts more skeptically to the results when it is this kind of poll. Why then would the Melancon campaign release an internal poll that still showed him 10 percent down?
Because independent polls show Melancon getting trounced and uncompetitive, with the latest Rasmussen poll that came out after Melancon released his showed that over the previous month Melancon had closed the gap with Vitter a whopping one percent so that he was down only 23 and Vitter still had a healthy 57 percent of the vote. Thus, the campaign felt it had to try to dispel the growing notion that, for a number of reasons, Melancon was history less than eight months from the election’s outcome.
Thus, a selective release was arranged, to certain media outlets and/or interested others who might be expected to comment on the results in line with the narrative the campaign wished to create. Either it or Hasten found one such figure, Florida-based pollster Verne Kennedy who occasionally polls in Louisiana, willing to cooperate, as Kennedy said “it’s a 50-50 race.”
At least the story (as others previously had not) revealed it was a poll of “likely” voters, just as are Rasmussen’s. And as noted previously, several factors could influence the distribution of answers to explain the disparate results. For example, while Rasmussen’s polls don’t include any such questions, since this was commissioned by the Melancon campaign it probably contained some priming questions that typically guide the respondent towards a more favorable impression towards that candidate. Other aspects could be differences in the actual wording itself and in the sampling frames themselves (an issue brought up by Kennedy).
A final factor could be that the Rasmussen polls, being fairly short, are automated while the internal poll was conducted by live interviewers. The latter could be a problem depending upon the quality of the interviewers who, unless well-trained, have a tendency to emit signals that encourage answers favorable to the client. The former could be a problem if it allows unregistered voters to answer or discourages certain people from participating (because they could be more likely to hang up).
However, Rasmussen has been proven extraordinarily accurate both in national contests and in Louisiana, and in this particular case the automated method probably creates a more valid sampling frame. With enthusiasm much greater this election cycle for Republican candidates, if there is any response or “hangup” bias in Rasmussen’s method (none has been demonstrated), it probably better captures the fact that, as research shows, some proportion (around 10 percent) of “likely” voters never make it to the polls. If there is a “hangup” effect, it likely is filtering out disproportionately some of these people who support Melancon or do not support Vitter, given the enthusiasm gap.
This difference was noted in another story about how Melancon is caught between a rock and a hard place concerning the upcoming health care insurance vote in the House. Melancon, who has said he would vote against his party’s bill that is detested by a clear majority in Louisiana, only is mimicking Vitter on this but by doing so he could drain enthusiasm from potential supporters of him. The impact of enthusiasm was noted not only by a Louisiana-based pollster, Bernie Pinsonat, but also by a former Democrat operative who parachuted safely into academia, Bob Mann.
Again, even if Melancon really were down just 10 percent with, as his campaign manager hoped, “room to grow,” that might be realistic if political tides weren’t running so heavily against Democrats. But that isn’t the 2010 election cycle nationally and especially in Louisiana. Despite all it’s wishful thinking, there is no strategy that the Melancon campaign can employ that will win the election. Regardless of anything Melancon does, this is Vitter’s election to lose and he will lose it only if he blunders – even if that isn’t the narrative Melancon prefers and perhaps some in the media prove willing to disseminate.