April 29, 2007 - May 5, 2007
A new Diageo/Hotline national survey (release, results) of 801 registered voters (conducted 4/26 through 4/30) finds:
- 35% approve of the job Bush is doing as president; 62% disapprove.
- General election match-ups are as follows (significant leads marked with *)
Giuliani 47%, Clinton 43%
Giuliani 39%, Obama 48% *
Giuliani 41%, Edwards 47%
McCain 45%, Clinton 45%
McCain 37%, Obama 48% *
McCain 37%, Edwards 48% *
The only instant reaction survey done immediately after last
night's MSNBC/Politco Republican debate came from automated polltaker SurveyUSA.
Their release tells us, without equivocation, that Rudy Giuliani "convincingly
won" the debate, because when asked, 30% of respondents picked him as the
winner compared to 17% for McCain and single digits for other candidates. We linked
to it, as did many
Mickey Kaus noticed, and asked a
Most commentators I've seen or heard thought Rudy Giuliani did badly in Thursday
night's debate (except my mother, who said he "makes a very good
impression.") I didn't think Giuliani did well either. So how did he
decisively win that Survey USA poll cited on The Corner? Part of the
answer is that Survey USA polled only California debate watchers.
A mere 45% of whom were Republican. 53% of whom were pro-choice.
That's not the national Republican primary electorate.
No, it's not, but an even bigger issue is the fundamental
weakness of "instant reaction" polls of this sort in judging winners and
A more general problem with using a survey to declare a
debate "winner" is agreeing on the criteria. What does it mean to "win?" Is it
enough to make a "good impression," or is the winner the one who picks up the most
new supporters? Is it about the impression made on those watching the entire
debate or on the much larger numbers who catch snippets in the news coverage
that follows? Is any coverage good coverage (think Gravel)? Do we only care
about Republican primary voters and caucus goers, or is it still important to
"win" a larger constituency. I could go on, but the various commentators seem
to apply different criteria.
What SurveyUSA did was straightforward enough. They called a
random sample of households in California
last night, and asked those who answered the phone whether they watched the
debate and, if so (13% said yes), asked who "won the debate?" It is hard to argue
with their presentation of the results. Giuliani was certainly the runaway
choice as "winner" among those who told SurveyUSA they watched it.
The problem with that approach, however, is that debates usually
serve to reinforce existing impressions. If you tuned in to last night's debate
feeling like a supporter of one of the candidates, the odds are pretty good
that if pressed, you would say that your candidate won. If most of the California voters that
watched the debate were already
Giuliani supporters, then most will likely tell us that their candidate won.
To test which candidate changed the most minds, we really
need to do what pollsters call a "panel-back" survey. They start with a very
large random sample of all adults or voters (perhaps as many as 2000) and call
several days before the debate to ask about vote preference and views of the
candidates. Immediately after the debate, they call back and ask the same
respondents (or as many as they can reach), who "won?" More important, they also
repeat the same vote and favorable ratings questions asked during the first
interview. Thus, with all the data in hand, the pollster can see which
candidate (if any) was judged a "winner" most often by those who were initially
undecided or supporting other candidates. They can easily see which candidates actually
improved their standing as a result of the debate.
Both the Gallup Organization and CBS News take exactly that
approach to polling around debates and presidential addresses, something I
wrote about many
times during the 2004 campaign (see especially this
Unfortunately, the SurveyUSA poll provides only a quick
snapshot of the impressions among those who happened to be home in California last night,
but we know nothing about how respondents felt about the candidates before the debate. However, we can at
least compare the "who won" response among likely Republican primary voters to
those interviewed by SurveyUSA and other pollsters in recent weeks, as in the
First, notice that the percentage who judged Giuliani the
winner is slightly smaller among likely Republican primary voters (30%, as
above) than among all debate viewers (33%, omitted from table). But notice that
the percentage judging Giuliani the winner is smaller than Giuliani's support
in other recent polls by SurveyUSA
(43%), the Mellman
Group (36%), the Field Poll
(36%) and PPIC
(33%). Probably more important, 40% told SurveyUSA they thought Romney or one
of the candidates other than Giuliani and
McCain "won" the debate last night. That is much higher than the lower tier
candidates received on the other recent surveys.
The comparison is far from conclusive. The sample sizes are
all small, and we have no way of knowing whether the sample of self-described
debate watchers was skewed to supporters of particular candidates. However, the
data we have suggests that if anything, the real winners were probably the
lower tier candidates who got a boost in exposure and recognition.
Also, put me down as skeptical that 13% of California's adult population watched the
debate last night. After last weeks' Democratic debate, Nielsen reported that 2.26
million Americans watched the debate. If we assume those were mostly adults,
it amounts to roughly one percent of the roughly 219
million adults in television households nationally. Perhaps this week's audience
was bigger in California,
but I doubt it was thirteen times
bigger. The 13% number is more likely is a combination of three factors: Debate
viewers were disproportionate among those who were (a) at home last night and
(b) willing to complete the interview. Also, consider the potential for (c)
measurement error -- some probably said they watched when then did not.
Update: Jay Leve,
the founder and CEO of SurveyUSA, responds in the comments section.
SurveyUSA thanks Mark Blumenthal
for his observations. SurveyUSA will use the "panel back" approach
before and after the next nationally televised debate. Those who want to
suggest other ways that SurveyUSA could improve speech-reaction and
debate-reaction polls are invited to do so.
Leve is obviously reading, so if you have a suggestion,
please leave a comment. And please bear with us if our software tells you your
comment is "waiting for approval." We have not not changed our comment policy, but have been working on some site upgrades that are behaving strangely. We will try to approve any such comments promptly.
Update II (5/6): I
had not seen it until just now, but TNR's
John Judis used the SurveyUSA data noted the same way I did on Friday and
reached a similar conclusion:
If one assumes that "who
won" tabulations are going to roughly resemble voters preferences, any
sharp divergence becomes significant. By that count, Romney was the clear
winner last night.
Gallop Gallup Poll analysis looking at what's behind Americans' views of Sen. Hillary Clinton (conducted 4/23 through 4/26 of 1,0007 adults) finds:
- 52%/45% now have unfavorable view of Clinton; 58%/40% favorable in February 2007.
- Among those with a positive opinion of Clinton (45%), her strength/stamina is cited as the number one reason, followed by intelligence.
- Among those with a negative opinion of Clinton (52%), her liberal/political views are cited as the number one reason, followed by distrust.
New Rasmussen Reports automated surveys find:
- Among 800 likely voters, 22% believe Bush "knew about the 9/11 attacks in advance;" 55% believe he did not (conducted 4/30 through 5/1).
- Among 1,000 adults, 40% said Congress subpoenaed Condaleezza Rice to "embarrass the administration;" 39% said to gather information (conducted 4/28 through 4/29).
- Among 15,000 adults interviewed from 4/1 through 4/30, a "new low" of 31.0% now say they belong to the Republican party (down from 37.3% in November 2004).
Having posted a set of tag clouds last
week on the Democrats comments during last weeks' MSNBC debate, we
naturally wanted to do the same for last night's Republican debate. So we once
again turned to Janet Harris (president of the media analysis firm, Upstream Analysis), who kindly sent
over another set created using the free site TagCrowd
and the transcript of the debate provided by the Federal
For those who missed last week's post, each graphic below displays
the 50 words used most often by each candidate during his answers to last
night's debate, with the type size varying according to its frequency of usage.
The larger the type size, the more often each candidate used that word. The
clouds omit common words like "and," "of," "the" and
(relevant to Giuliani) "new."
A few minor changes from last week, mostly by popular
demand: The clouds now also display the count for each word (in parentheses). As
you will see the, size of each displayed word is partly a function of the
number of words spoken by each candidate. In other words, the scale is not constant
across candidates. Also each Cloud now automatically groups similar words
(e.g., learn, learns, learned, learning are all grouped under
"learning"), but for some reason it doesn't do this for
America/American. Janet also excluded references to the moderator
As she did for the Democratic tag clouds, Janet created a PDF version suitable for printing.
Keep in mind with that each candidate answered different
questions, so the words they used varied accordingly. With that huge caveat,
here are a few observations, mostly from Janet:
- You can
clearly see areas of emphasis for some candidates: New
York for Rudy Giuliani (the program filtered out "New"), the
border fence for Duncan Hunter, jobs and faith for Mike Huckabee, and for
John McCain, references to president and Iraq,
war, weapons and security.
Romney, like John Edwards last week, shows clear emphasis of simple
language and man-of-the-people rhetoric: America(n), values, nation,
Romney to the cloud of negativity from Ron Paul: bad, critical, fight,
interfere, ought, poor, rid, secrecy, war.
was most prominent for Romney and Giuliani (as well as
"pro-choice" for Romney), that is probably a function of those
candidates being asked about that issue more than the others.
Last week, your comments, questions and interpretations were
as interesting as the clouds themselves. Once again, please, have at it...
Correction: Representative Tom Tancredo originally labeled as Governor Tancredo.
Update: Hotline On Call has posted the number of questions each Republican recieved and the total amount of time for which they spoke.
A new SurveyUSA automated survey of 2,400 adults in California (conducted immediately following the 5/3 Presidential Debate) finds:
- Of those who both watched and are likely to vote in the California Republican primary, 30% thought former Mayor Rudy Giuliani won, 17% thought Sen. John McCain won, and 13% though former Gov. Mitt Romney won.
Hotline On Call's Marc Ambinder posted
his thoughts on the "Branding
Study" of the Democratic candidates conducted by two South Carolina advertising agencies, Chernoff
Newman and MarketSearch. "We're sure," Ambinder wrote, "Mark Blumenthal will
weigh in." Well, since he asked...
I started to write something on the notion of "brand analysis"
of presidential candidates, but wanted to first kick the tires on the survey
behind the study before thinking about the results. Unfortunately, what I found
raises some doubts about how representative the sample is of true "likely
Democratic primary voters" in South
First, consider the racial composition of the sample. African-Americans
comprised 47% of Democratic primary voters in the 2004 South Carolina primary, according to the
poll. Two surveys also conducted in April report an even bigger percentage
of African-Americans among likely Democratic primary voters in South Carolina: 53% on
the poll by Democratic firm Garin-Hart-Yang
and 56% on the survey from the bi-partisan team of Hamilton-Beattie
(D) and Ayres-McHenry (R). In However, African-Americans make up only
30% of Democratic primary voters on the CNMS survey. That's a pretty
big discrepancy, especially given the importance of African-Americans vote in
Second, consider geographic composition. The CNMS report breaks
out the contribution of the six largest counties, leaving 51.3% in the "other"
category of generally smaller and more rural counties. Yet according to the turnout
statistics I obtained from the South Carolina Democratic Party, those smaller
counties contributed 58.5% of the vote cast in the 2004 Democratic presidential
primary. My assumption is that rural Democrats in South Carolina tend to be more
African-American than those in bigger more urban counties (as indicated by
in the Garin-Hart survey).
Third, the authors of the CNMS questionnaire made a very
unconventional choice in terms of question order. They asked the vote horse-race
question at the end of the survey, after
their full battery of candidate attribute questions. Notice that the two other
polls linked to above follow the standard procedure of media and campaign
pollsters, asking the vote after the favorable ratings but before specific
questions that may "prime" attitudes and skew vote preference.
All of which brings me to the horse-race results for this
survey, which are something of an outlier. They show Hillary Clinton leading
Barack Obama by 17 percentage points, far more than the three surveys conducted
over the same period, which show Clinton
leading by margins of only 3 to 7 points.
So while the issue of "brand attributes" is worth exploring further, and
something I'll try to come back to it tomorrow, I have some grave questions about how
representative this survey of actual Democratic primary voters or how
accurately their horse-race question measures existing vote preference.
**The discrepancy in the racial composition may not explain
why the CNMS survey shows Clinton
with a bigger lead. Unfortunately, of the four recent South Carolina surveys, only Garin-Hart-Yang released tabulations by race. Those show a big difference in
for John Edwards -- he wins 35% of white voters but only 9% of African Americans
- but not as much net difference in the Obama-Clinton contest. Clinton leads Obama by 4
points (23% to 19%) among white voters and by 5 points (40% to 35%) among
African-Americans. Nationally, the vote-by race tabulations in the most the Pew Research
show essentially the same pattern.
emailed the Chernoff Newman and MarketSearch team for comment and received
the following reply from MarketSearch regarding the differences in racial
When we drew our sample of registered voters we selected those who had
voted in one or more elections starting with the November election of 2004 as
well as those who are now under 21 and registered since the last presidential
election. As we noted earlier, we also excluded anyone who did not indicate
they were "very likely" or "somewhat likely" to vote in the
upcoming primary. Our intention was to select a sample that would have the most
familiarity with candidates and thus be in the best position to address our
questions about branding. I don't have the exact figures at my disposal right
now, but these steps may have resulted in a lower percentage of
African Americans in the final sample.
And regarding placement of the vote horse-race question:
We did this deliberately since our
interest in this study is not the horserace but the branding issue. Our
intention was to do a follow through from brand image to the person for whom
the respondent would vote. We are not positioning this study as a head-to-head
comparison of those whose interest is the horserace.
Additional results from the recent Working Californians statewide survey of 400 likely Republican primary voters in California (conducted 4/19 through 4/12 by The Mellman Group (D) using registration-list based sampling) finds:
- Former Mayor Rudy Giuliani leads Sen. John McCain (36% to 21%) in a statewide primary; former Gov. Mitt Romney trails at 9%.
- 67% of California Republicans have a favorable opinion of Giuliani (17% unfavorable), 66% have a favorable opinion of Bush (29% unfavorable), and 58% have a favorable opinion of McCain (31% unfavorable).
A new Strategic Vision (R) statewide survey of 800 likely voters in New Jersey (conducted 4/25 through 4/27) finds:
- 24% approve of Bush's overall job performance; 67% disapprove.
- Among Democrats, Sen. Hillary Clinton (at 40%) leads Sen. Barack Obama (23%) and Sen. John Edwards (12%) in a statewide primary.
- Among Republicans, former Mayor Rudy Giuliani leads Sen. John McCain (49% to 15%) in a statewide primary; former Sen. Fred Thompson, former Gov. Mitt Romney, and former Speaker Newt Gingrich trail with less than 10% each.
A new Quinnipiac University national survey of 1,166 registered voters (conducted 4/25 through 5/1) finds:
- 35% approve of way Bush is handling his job as president; 60% disapprove.
- Among 499 Democrats, Sen. Hillary Clinton (at
38% 32%) leads Sen. Barack Obama (18%), former V.P. Al Gore (14%), and former Sen. John Edwards (12%) in a national primary.
- Among 469 Republicans, former Mayor Rudy Giuliani (at 27%) leads Sen. John McCain (19%) and former Sen. Fred Thompson (14%) in a national primary; former Gov. Mitt Romney and former Speaker Newt Gingrich trail at 8% each.
- In general election match-ups, Gore trails both Giuliani and McCain, Clinton trails Giuliani but runs within the margin of sampling error of McCain, and Obama runs within the margin of error of both Giuliani and McCain.
New SurveyUSA automated surveys of statewide general election match-ups of 500 registered voters in ten states (conducted 4/13 through 4/15) finds:
Three new analyses from Pew Research Center find:
- Among 2020 adults (analysis, results), 56% think mothers are doing a worse job today than they did 20 to 30 years ago; 9% say better; 29% say about the same (conducted 2/16 through 3/14).
- 41% agree that the government should "help more needy people even if debt increases," "gaurantee food and shelter for all," and "take care of people who can't care for themselves;" 13% disagree with all three statements. In 1994, 29% agreed and 24% disagreed.
- Andrew Kohut answers whether or not a Republican "agent of change" can win in 2008.
One of the more intriguing things included in last week's post
on recent polling in California is the shift by the state's venerable Field Poll to voter registration
list sampling for their pre-election surveys.
The debate over the use of sampling from lists, rather than
randomly generated telephone numbers continues to rage among pollsters. We had
post by Democratic pollster Amy Simon late last year which addressed the key
issues. The gist is that most media pollsters continue to rely on "random digit
dial" (RDD) samples, while campaign pollsters have largely shifted to sampling
from registered voter lists. The media pollsters argue that RDD samples are
superior because the randomly generated telephone numbers allow coverage of all
telephone households, including unlisted numbers. The campaign pollsters argue
that computerized lists have improved in recent years, and that the improved
ability to select actual registered voters with some history of voting makes
for a more accurate measurement of likely voters.
The shift to "registration-list based sampling" (RBS) by the
Field Poll is remarkable given its history and influence among other pollsters.
As every Field release reminds us, the Field Poll has "operated continuously"
since Mervin Field first founded it in 1947. Field has been on of the most
active and respected members of the American Association for Public Opinion
Research (AAPOR), winning its award
for exceptionally distinguished achievement in 1979. A shift in methodology at
such a firm does not come easily.
After Mark DiCamillo, Field's current director, emailed me
with a correction to the item I posted last week, I asked him about the change.
Here is his response:
The Field Poll switched over to using RBS based samples last
year, and used them throughout/t the 2006 gubernatorial election cycle. There
are obviously pluses and minuses to the change, but long term we think this is
the way to go, since the cell phone only household segment missed by RDD
samples willonly get larger over time. We've found that by using the RBS samples,we can incorporate cell phone only voters since they tend to be thenewer registrants and when providing a phone contact on their voterregistration card, they provide their cell phone number. So, we've developed some special procedures for calling these listings. In addition, prior to the survey we send the entire voter sample to Survey Sampling and they identify for us which numbers on the voter list are cell phone listings, so we know in advance which numbers are cell phones. So, far we've been successful in bringing in a majority of the cell phone voters into the sample.
BTW-- Our final pre-election poll estimates in last yea's cycle were extremely accurate, in that not only did we show Schwarzenegger ahead by 16 pts (he won by 17), but our measures for each of the down ballot candidate races were all on the money as were our proposition measures, so we are encouraged by our first experiences with the change-over to these samples.
Two things are especially intriguing here: First, Field's
methodology page explains that the voter list samples they purchase from the firm Voter
Contact Services currently provide "a telephone number for about 85% of the
voters listed." If that statistic is accurate and if the voter list is
otherwise up to date, it would mean that the aggregate coverage provided by
Field's RBS sample may not be that much worse than the current coverage of random
digit samples nationally. According to the most recent report
by the National Health Interview Survey, the proportion of U.S. households with traditional
landline phone service had fallen to 87.5% as of the first six months of last
Similarly intriguing are the "special procedures" the Field Poll
has developed to interview voters on their cell phones. Hopefully, they will
have more to say about these innovations soon.
A new Chernoff Newman/MarketSearch statewide survey (release, brand analysis, results) of 400 registered, likely Democratic primary voters in South Carolina (conducted 4/9 through 4/16) finds:
- In a South Carolina Democratic primary, Sen. Hillary Clinton leads Sen. Barack Obama 38% to 21%, while Edwards and Gore trail at 17% at 10% respectively. In a three-way race, Clinton (at 43%) leads Obama and Edwards (both at 25%).
- The survey tests the "branding" of the Democratic candidates with extensive questions on their familiarity, reputation, personality, performance, and connectivity.
Who is up and who is down in recent presidential nomination polling? What are the current trends? The question is not as trivial to answer as it might seem. If we look at different polls, we can find some bouncing up while others bounce down. Commentators often reach different conclusions because the are comparing different polls. None of the recent polls have the order of finish significantly different--- all have Clinton and Giuliani in first place and Obama and McCain in second, with Edwards and Gore together and Romney, Fred Thompson and Gingrich mixed together. But the gaps between the candidates, and who has moved up or down since the last poll varies quite a bit across polls.
The goal of my kind of analysis is to avoid the trap of focusing on only one or a couple of polls. My approach is quite skeptical of the evidence provided by any single poll, but quite confident in the information from all the polls taken together. The problem is how to combine the polls to get a good estimate of what is "really" happening, and not to be deceived by the random variation from poll to poll. So let's see how the presidential nomination races are shaping up when we take all the polls seriously.
Regular readers know that my "standard" trend estimate is the blue line in the charts. This is a line that is calculated to go through the "middle" of the data, with an average error of zero, meaning the points below the line balance the points above the line. "Old Blue", as I affectionately call this line, is deliberately conservative in the sense that it takes quite a bit of new polling data to convince it to change trend direction. The reason for this is that we know there is quite a bit of noise in the polls (just look at the spread of points around the line!) so when a new poll comes in high it might mean an upturn in support, but it is just as likely that it simply reflects random noise and the next poll is as likely to come in low. If we allow the trend estimate to chase each new data point too much, we'll just plot random noise rather than the best estimate of the trend in support. Experience with these and other data (such as presidential approval) has shown that Old Blue is seldom misled about new trends, though it does take a while (about a dozen polls) to notice changing trends.
While it is good to avoid responding too much to a single poll, it is also true that Old Blue may stick to a trend longer than it should. A more sensitive estimator would notice a change in direction quicker, and would jump on the new trend while it is still news-- and before others notice it. "Ready Red" is the answer to this. The red line in the charts is twice as sensitive to change as is Old Blue. As a result it will pick up changes in momentum more quickly, letting us spot new trends early. Unfortunately, it will also sometimes be misled and will think it sees a new trend when in fact none exists-- just a few polls that happen to be "down" or "up" but which really don't represent any significant shift.
Of course you can adjust the sensitivity of the trend estimator to anything between Ready Red and Old Blue (or outside them too, for that matter) to see how much difference the sensitivity makes. There is no perfect way to choose a "best" estimator. I've settled on the more conservative Blue estimator as my standard because I find the hasty red estimator has often jumped the gun on presidential approval trends, which more data has subsequently shown were not really changing. But that doesn't mean we shouldn't examine the more sensitive trend estimate-- it tells us a lot, even if we have to be a bit cautious. (I'd rather be right but slow. Others prefer to be quick, and adjust to mistakes as necessary. Both approaches have their virtues.)
The chart above shows the data and the trend estimates for the top Democrats. Republicans are in the chart below. In addition to Old Blue and Ready Red, there are a number of gray trend lines (81 in each figure). These show the estimated trends for levels of sensitivity from quite a bit MORE sensitive than Ready Red to MORE conservative than Old Blue. If you can see the gray lines, this means that at least for some levels of sensitivity the estimated trend differs from either Blue or Red. If you cannot see the gray lines, or only barely, this means that the estimated trend hardly depends on the amount of sensitivity and the many gray lines all lie under Red or Blue in the plot, and so are covered up. This typically happens when there is a smooth, steady trend with no bends in it.
So enough statistics, let's look at the politics.
In the Democratic race, the Old Blue estimator says that Clinton has been flat since January, after a bit of decline in 2006 and a slight rebound late in 2006. Not much action.
But if we look at Ready Red, the Clinton campaign appears to be falling off in recent polling, declining by about 3 or 4 points since her peak in January. That isn't a large drop, but it does suggest that the stable picture of Old Blue may be masking some short term decay.
The Obama campaign is similarly interesting in comparison of the two trend estimates. Old Blue sees a sharp rise in early 2007 with a slower but still upward trend recently. Red sees more of a plateau in recent polls, with some indecisive and quite small up and down bounces. If I believe Red, I say Obama has stalled. If I believe Blue, I say he has slowed but is still moving up a little. (If I'm really crazy, I say the last little uptick in Ready Red suggests Obama is about to move up again, but that would be giving an awful lot of weight to the very last polls on a sensitive estimator. I'm not that crazy.)
The two trends are in pretty close agreement for Gore, but with Red suggesting a slight downturn at the very end, while Blue says the trend remains up a bit. Again, the difference is driven by only the polls at the very end and I'm not willing to bet much on them.
The Edwards campaign could take heart in Red's somewhat higher rate of climb in support compared to Blue. Both agree Edwards has been moving up, but Red sees the upturn as sharper and ending at a higher level. The best that can be said here for Red is that this trend has been supported by more polls than is the last little change of Obama or for Gore.
As for the numbers, the estimates are not far apart regardless of which estimator we pick.
Clinton: 35.7 (Blue)/34.1 (Red)
On the Republican side, Giuliani has enjoyed a long and sustained rise based on Old Blue, but suffers a recent downturn if Ready Red is to be believed. If the sensitive estimate is right, there has been over a five point decline in Giuliani's recent standing since early February. If Blue is right, then don't be hasty and Giuliani has continued to gain, though at a slower rate than in late 2006.
Red and Blue agree that 2007 has been a bad time for the McCain campaign. After a flat 2006, McCain has dropped over five points in both estimators. Sensitive Red thinks there may be a chance of a recent reversal of that slide, but Old Blue remains entirely unconvinced that McCain's fortunes are improving.
Blue and Red also agree that Gingrich has suffered a bit of a recent decline (more or less coinciding with talk of a possible Fred Thompson candidacy.) This is a nice example of both estimators reaching the same conclusion, even with late trends. Red sees Gingrich slightly worse off than does Blue, but the difference is slight.
Likewise, both estimators are largely in agreement that Romney has sustained his tortoise-like slow but steady increase. Despite some campaign gaffes, both trends remain up, with Red being a little more bullish than Blue.
Fred Thompson lacks enough data to provide a fair assessment of the estimators, but who can ignore him at this point. The only rational approach would be to be conservative in the face of very limited data for a trend estimate. By that account, Old Blue says the sudden possibility of a Thompson campaign has generated 10 points of support, but with no evidence of a trend either way since polling on Thompson began. The Red estimator jumps around-- there just isn't enough data for a sensitive trend.
The current estimates for each trend are:
Giuliani: 34.8 (Blue)/30.4 (Red)
We can check the sensitivity of these estimates to the amount of smoothing used to estimate the trend. Here I use 81 separate estimates of the current standing of each candidate, with the smoothing ranging from MORE sensitive than Red to MORE conservative than Blue. This is a wider range than I think anyone would reasonably want. The most sensitive end produces trends that jump around way more than anyone could believe, and the most conservative fit is basically just a straight line with hardly any change at all. But somewhere between these limits of silliness are a range of reasonable estimates. If the bottom line estimate for a candidate is pretty compact, then the amount of smoothing doesn't matter. If the estimates are spread out, then we at least know that sensitivity matters and we should be cautious. The summary of the data are presented below.
The top half of the plot shows that the estimates for most candidates are in fact within a fairly small range regardless of how sensitive the estimates happen to be. Gore and Edwards are quite close, with some overlapping estimates of support. But Obama and Clinton are clearly distinct from each other and from Edwards and Gore. Similarly, Gingrich, Thompson and Romney show similar estimates and considerable overlap, while McCain is clearly above them and Giuliani clearly ahead of McCain.
Giuliani stands out among all the candidates in demonstrating more dependence on the sensitivity of the estimator. His box is more spread out than those of other candidates in the top half of the plot. In the lower half, which plots the distribution of all the estimates, Giuliani shows a bi-modal distribution. If we pick a more sensitive estimator, Giuliani support falls in the lower "hump" of the distribution, while less sensitive estimators suggest a stronger standing, producing the right hump. This difference is not trivial-- the more sensitive estimator says Giuliani is at about 30% support, while the more conservative one says 35%. No other candidate shows as large a discrepancy. This is due to the rather substantial downturn that Ready Red sees in Giuliani's polling over the last three months, but which Old Blue is still reluctant to accept. Which is right? Well, that's the whole point here: If you are a bit more daring, believe what Red has to say. If you like to buy municipal bonds, go with Blue.
One final way to look at sensitivity is to plot the estimated support for each candidate against the degree of smoothing used for each of the 81 estimates. Low values are less smoothing and more sensitive trends, while high degree of smoothing are more conservative and less sensitive.
The good news from my point of view is that most of the lines do not demonstrate a strong relationship between amount of smoothing and the estimated support. While there is a little movement, it isn't sharp for anyone. The Giuliani line is the one showing the greatest variation across degree of smoothing, as I already noted. The upshot of this is that while I constantly worry about how much my estimates are affected by my preference for Old Blue, the data show that mostly it doesn't matter a lot, and certainly not within a reasonable range of smoothing. (In the plot above, Old Blue is a degree of smoothing of .7, while Ready Red is at .35.)
It is good to compare Old Blue and Ready Red-- both offer helpful insights into the nomination race. Your acceptance of the risk of being too slow to recognize change versus the risk of chasing phantom blips should help you decide which to give more credence.
Cross-posted at Political Arithmetik.
A new SurveyUSA automated survey of likely gubernatorial primary voters in Kentucky (conducted 4/28 through 4/30) finds:
- Among 612 Democrats, businessman Bruce Lunsford edges out former Lt. Gov. Steve Beshear in a Democratic gubernatorial primary.
- Among 421 Republicans, Gov. Ernie Fletcher leads former Rep. Anne Northup 46% to 34% in a Republican primary.
Three new American Research Group statewide surveys of 600 likely Republican primary voters and 600 likely Democratic primary voters in Iowa, New Hampshire, and South Carolina (conducted 4/27 through 4/30) finds:
- Among Democrats, former Sen. John Edwards edges out Sen. Hillary Clinton (27% to 23%) in Iowa, while Sen. Barack Obama trails at 19%. Clinton leads Edwards (37% to 26%) in New Hampshire, Obama trails at 14%. Clinton leads Obama (36% to 24%) in South Carolina, Edwards trails at 18%.
- Among Republicans, Sen. John McCain leads former Mayor Rudy Giuliani in Iowa (26% to 19%) and South Carolina (36% to 23%); in New Hampshire McCain edges out former Gov. Mitt Romney (29% to 24%), while Giuliani trails at 17%.
Ann Kornblut's 2,800 word, front-page must-read
profile in yesterday's Washington
Post of Clinton
pollster and "chief strategist" Mark Penn has been stirring up quite a bit of
critical commentary on the left side of the blogosphere. Although Media Matters takes the
piece to task for speculating without "any evidence" that "American voters
‘suffer from Clinton fatigue,'" most are critical
of Penn. Greg
Sargent questions what Kornblut described as Penn's "deep roots in the
national security wing of the Democratic party." Both Mark
Schmitt and Matthew
Iglesias take exception to what Kornblut described as Penn's "undisputed
brilliance" as a Pollster.
Others focus on Penn's corporate conflicts. "In the 1980s," Matt Stoller writes, "[C]entrists
like Penn . . . were often on retainer to tobacco, telecom and pharma because
it was good business to have influential consultants on their payroll." Similarly,
Sirota asks, "could Mark Penn and the Clinton
team be any more of a walking advertisement for corruption, insiderism and
Here are the two key paragraphs on that subject from Kornblut's
The job [worldwide chief executive of the public
relations firm Burson-Marsteller] is the latest iteration of the lucrative
corporate work that Penn and Schoen began in the 1980s, at the same time they
were making their names as political pollsters, and that put them in the
company of a new generation of business-minded Democratic consultants.
Among their clients over the years were ATT, Eli
Lilly, Texaco and Microsoft. Their specialty was corporate research and
positioning -- figuring out, for example, how AT&T could outflank
competitor MCI by targeting uncommitted customers, the business equivalent of
seeking out swing voters. While some Democratic rivals criticized the crossover
work, suggesting that Penn had sold out or worse, the polling firm expanded
rapidly, with Penn and Schoen adapting corporate models to the political sphere
and vice versa.
I have two thoughts to add, although readers should remember that that until
last fall, I worked for 20 years as a Democratic campaign pollster and thus
technically qualified as a "Democratic rival" (although in terms of clients, I was
certainly not in Penn's league).
First, yes, as Mark Schmitt writes,
"life is full of conflicts." Penn and Schoen are certainly not the first Democratic
consultants to take on corporate clients, nor will they be the last. And yes, as
Schmitt puts it, "everyone in Washington
has at least two jobs," or at least it sometimes seems that way. However, Penn
and Schoen have displayed a thirst for corporate work, often in conflict with the
policy agendas of their political clients, that has long set the bar among Democratic
pollsters. My employers and partners over the years had corporate lines they (variously)
refused to cross -- tobacco, pharma, big oil, aggressively anti-union -- both out
of ideological principle and to avoid putting their valued political clients in
a tough spot. A quick glance at the Penn,
Schoen, Berland client list shows they not only crossed some of those
lines, but did so with enthusiasm.
One personal irony is that while I have never met Mark Penn, my one or two encounters
with his former partner Doug Schoen were the result of a mutual corporate client,
America Online, Inc. For eight years, I conducted customer satisfaction surveys
for AOL -- something that provided a big chunk of my annual income -- while
Schoen did research focused more on AOL's advertising and corporate image. I
was fortunate that my work for AOL never posed a direct conflict with my
political clients, at least, none that I was aware of. But it certainly could
have. Consider a hypothetical example: What if I had been polling for an
Attorney General who had filed a lawsuit against the company and wanted to
highlight that action as a message in a campaign? It would be absurd to argue
that my advice to that candidate would be unaffected, even subconsciously, by my
regular stream of income from the corporate client.
These sorts of ethical questions are difficult, and I assume that my
colleagues in the consulting world find the harsh criticism from the "Netroots"
highly annoying. But these are important issues, and the bloggers are right to
ask tough questions. Many campaign pollsters chose to avoid lucrative corporate
projects to avoid creating conflicts for their valued political clients. As
such, it seems entirely fair to hold candidates accountable for the apparent
conflicts of interest of their influential consultants.
A new Rasmussen Reports automated survey of likely primary voters (conducted 4/23 through 4/26) finds:
- Among 765 Democrats, Sen. Barack Obama edges out Sen. Hillary Clinton (32% to 30%) in a national primary; former Sen. John Edwards trails at 17%.
- Among 602 Republicans, former Mayor Rudy Giuliani (at 30%) leads both Sen. John McCain and former Sen. Fred Thompson (both at 14%) in a national primary; former Gov. Mitt Romney trails at 11%).
Here is something of an exclusive (for the moment - and we'll
spare you the flashing red light): Our friends at the Cook Political Report
have shared advanced results of the latest Cook/RT Strategies survey on 2008 presidential
primary preference (conducted over the last three days, April 27-30, among
1,000 adults nationwide).
Their results for Democrats
show a tightening national race in the last month. Among 389 registered voters that
identify with or lean to the Democrats, the survey shows Clinton leading with 32%, followed by Obama
at 24%, Edwards at 15%, Gore at 11% and all other candidates in the low single
digits. Without Gore in the race, they show Clinton leading Obama by ten points (36% to 26%)
trailed by Edwards at 18%.
On the previous
survey (of 355 Democrats) in late March - which did not include Al Gore as
a potential candidate - they showed Clinton with a 24 point lead over Obama
(41% to 17%), who ran two points behind Edwards (at 19%). The nine-point increase
in Obama's vote over the last month on this survey (from 17% to 26%) is
statistically significant despite the small sample sizes, although the five
point decline for Clinton
(from 41% to 36%) is not.
For those watching, four national surveys (NBC/Wall Street Journal, USA
have shown a similar narrowing, while four others (ABC/Washington Post, the Pew Research
Center, CBS News
News/Opinion Dynamics) have not.
The survey also indicates a possible tightening of the Republican contest. Among 319 registered
voters that identify with or lean to the Republicans, Rudy Giuliani now leads
by seven points (28% to 21%), followed by Mitt Romney (11%), Fred Thompson
(10%), Newt Gingrich (6%) and all others at 2% or less.
That result represents a drop in Guiliani's lead over McCain
from 17 points (34% to 17%) a month ago (among 290 Republicans), a decline that
looks right on the edge of statistical significance given the relatively
small sample sizes.
Watch for Charlie Cook's upcoming column on NationalJournal.com
and CookPolitical.com for more details.