THE BLOG
10/16/2007 08:30 pm ET | Updated May 25, 2011

"Outlier"

Two weeks ago, a new ABC News/Washington Post poll showed
Hillary Clinton "surging:"

She has 53 percent support in the
latest ABC News/Washington Post poll, up 12 points from early last month, vs.
20 percent for Barack Obama (down seven points to his lowest of the year), and
a stable 13 percent for John Edwards.

In a blog
post
on that poll, I noted the dramatic interpretation of these results in
various corners and then looked at how the poll compared other results (based
on trial heats omitting Al Gore as an option). I wrote:

[T]he 53% result for new Post/ABC
poll is more of an "outlier" from the regression trend line than any
poll conducted this year (it's the purple dot at the extreme top right of plot
area). At 53%, the polls estimate of Clinton's
support falls a full ten percentage points higher than our current estimate of
the trend (42.5%) even without Gore in the race.

Since then, six new national polls have been released, and
all but one** show support for Clinton
above the 42.5% trend estimate we reported ten days ago. With the addition of
those new polls, the trend line below, based on trial heat results that do not
include Gore now shows an expanding Clinton lead, with her support increasing
roughly 3 percentage points since August.

AUS2TopzDems071016sml.png

The 53% result for Clinton on
the ABC/Post poll remains on the high end of support for Clinton compared to the trend line, but is no
longer "more of an outlier" than various other polls conducted earlier in the
year.

ABC's Gary Langer also sent a table this morning to point
out something I overlooked two weeks ago. I've reproduced it below (splitting it
into two parts to better fit this space). He first reproduced the results as
reported by the pollsters - the same results plotted in our chart above:

10-16%20recent%20dems.png

Notice that the ABC/Post result for undecided (2%) is lower by far than any of the other results in the table. That difference is not unusual, in that the ABC/Post poll typically shows a smaller "don't know" result than other polls. Langer then recalculated each
result as a percentage of those with a preference. Calculated this way, the
ABC/Post results look much closer to the average of all six poll and not at all
like an "outlier."

10-16%20without%20undecided.png

So point taken.

But let's also remember that the typically lower than
average undecided on the ABC/Post poll did not produce a similarly discordant
estimate for Clinton
in early September. Their September 4-9 survey put her
support at 41%, within a percentage point of our trend estimate at the time. So
we cannot explain away the apparently dramatic 12 point increase as an artifact
of a small undecided.

Finally, about my speculation twelve days ago:

Of course, we do not yet know
whether this poll is really a statistical outlier. Other polls have obviously
been showing a more gradual increase in her support recently, and it is still theoretically
possible that Clinton's
support suddenly lurched up ten points last week...

So we will wait and see. But I'll
wager that a month from now the real trend will not look nearly as dramatic as
the one suggested by yesterday's news.

A month has not yet passed and the trend line for September
might still change, but I stand by my wager.

**The Insider Advantage national
poll
of Democrats, not included in Langer's chart, looks far more deserving
of the "outlier" label if only because of its huge 23% undecided produced by
informing respondents that "no opinion" is an option.

Subscribe to the Politics email.
How will Trump’s administration impact you?