Mass-production
pollster Public Policy Polling, which works for Democrats and leftist causes, recently
put out a poll that showed in hypothetical runoffs with Republicans major
Democrat gubernatorial candidate state Rep. John Bel Edwards hanging in
there and, in the case of frontrunner Sen. David
Vitter, decisively defeating him. Is this believable?
Probably
not, for a number of reasons, one of these being the quality of PPP surveys
regarding state-level elections. PPP produces a high number of these by using
less-rigorous methodology that serves to lower cost. As a result, their
products often are all over the map, sometimes pretty accurate and sometimes
wildly off, with the latter
represented by its final effort in last year’s Senate contest that significantly
over-predicted support for former Sen. Mary Landrieu. In fact, the 2014
cycle produced a great amount of inconsistency for it. As such, of the 21
outfits that have produced at least such 50 polls since 1998, it ranks in about
the middle for accuracy.
Further,
it seems that its recent trend towards increased chances of less accuracy has occurred
as it practices
fiddling with sampling frames to fit a preconceived notion of the
electorate that leans in the direction of favoring Democrat candidates in these
contests. Finally, keep in mind that this effort came at the behest of the political
action committee set up expressly to defeat Vitter and is run by the guy who
was formerly the head administrator of Louisiana’s Democrats and the campaign manager
of Vitter’s vanquished main 2010 Senate challenger.
All
in all, this has led to skepticism about these results even from a frequent user of
PPP, the Angry
Left web site Daily Kos. It sensibly notes the unreliability of the runoff
matchup numbers, in that Edwards has gotten free ride while the Republicans all
have attacked each other – perhaps indicative by Edwards having the highest
favorability/unfavorability numbers. The large
number of undecided voters found in it at this stage of the race also
raises doubts about the sample quality (perhaps influenced by a relatively
smaller sample size), where its automated system is less likely to get people
in a household who are registered voters or who really intend to vote (95
percent insisted they “definitely” will vote) and forces them to choose only
from among the top four candidates or for undecided also attracts a
less-representative sample.
And
there is a time-bound, hypothetical quality to it all: of Edwards secures a
place in the runoff, his liberal record specifically and tying him to national
Democrats generally publicized by his opponent will turn around favorability
numbers in a hurry. Is it credible that in a sample that claimed 55 percent of
it voted for the Republican presidential candidate in 2012 that Vitter is so loathed
that so many who would vote for a Republican in the general election would
abandon Vitter for a Democrat in the runoff? Nor does the poll jive with the others most recently, both in
terms of sample and in results.
Note
now that three distinct sets of polling have emerged for this contest. Most
show Vitter and Edwards at least about 15-20 points ahead of everybody else,
but those coming
from a particular pollster give Vitter less support and more to Public
Service Commissioner Scott Angelle
so that he rivals Vitter, and now comes this one showing Edwards beating Vitter
in a runoff and doing well against Angelle and the other main Republican Lt.
Gov. Jay Dardenne when no others show
those results. The electorate is not that volatile; some of these results
simply are not accurate.
That
doesn’t mean the PPP one must be inaccurate, but it does mean that it’s likely
that it is. Confirming this is a media poll released earlier this week that
does show a competitive race in an Edwards-Vitter runoff, giving the Democrat a
lead within the margin of error, not 12 points as the PPP poll asserts. Thus, it
probably was fashioned as a relevancy exercise on behalf of Democrats to keep
the troops from getting too discouraged in state elections this year by
creating a hypothetical path to victory for Edwards.
But
it also falls within the context that the political left always indicates who
it fears the most. This may be a strategy to push Republican voters into the
Dardenne and Angelle camps by making them think that Vitter is unelectable, because
liberal elites can live with those guys in office – even as they contradictorily
assume that fear of a Democrat elected will drive GOP supporters away from
Vitter in the general election, but not in the runoff. However, liberals know
Vitter will give their policies no quarter if elected governor, and they
desperately wish to avoid that. Thus, as in the case of this poll, desperate
people do desperate things.
No comments:
Post a Comment