The Value of Daily Tracking
It’s not unusual for politicians to react to polls. I’ve certainly seen it many times over the years, particularly when elected representatives or candidates are confronted with poll results they don’t like.
White House Press Secretary Robert Gibbs was asked Tuesday about Gallup Daily poll results showing a drop in President Obama’s job approval rating. According to news reports, Gibbs responded with some critical remarks and unusual analogies.
I’m certain Gibbs didn’t intend to impugn the value of presidential job approval polls in general. It appears he was reacting more to the fact that the president’s approval numbers are not stable, but, in fact, in a period of some change. More specifically, Gibbs was reacting to our report Monday highlighting the fact that, while there was a short-term positive uptick in Obama’s job approval ratings after his Afghanistan speech last week, his ratings through the weekend fell back.
But this type of movement is the nature of the beast. Gibbs said that if Gallup were his EKG, he would visit his doctor. Well, I think the doctor might ask him what’s going on in his life that would cause his EKG to be fluctuating so much. There is, in fact, a lot going on at the moment -- the healthcare bill, the jobs summit, the Copenhagen Climate Conference, and Afghanistan.
We live in a representative democracy. Our politicians are accountable to the people. Certainly the accountability that matters most is on Election Day. But keeping tabs on the people’s views of their elected representatives between elections is vitally important – and something in which the people of the country are demonstrably interested. We at Gallup are fortunate to have the capability to interview random samples of Americans on a daily basis. This helps us closely monitor the ways in which presidential actions are being received by the national constituency.
Of course, it's not just Gallup that finds this important. I’m sure the White House was just as interested as we were in how the president’s major speech at West Point last week played to the American public. Our tracking helped provide the answer.
Obama is set to travel to Oslo, Norway, to receive the Nobel Peace Prize. The White House is probably just as interested as we are in how the American public is going to react to this event. Our tracking will give us the answer -- both in the short-term and in the long-term. (And I’m sure the Obama campaign in 2008 paid a great deal of attention to their own tracking polls measuring how his candidacy was doing as the events of the campaign rocketed across the news each day.)
The foundation of the tracking poll is state-of-the-art, sound scientific sampling methodology, as is the same with all of our Gallup polling. Gallup reports presidential job approval using a three-day rolling average. Users are, of course, free to make whatever use they would like of the daily tracking information. The same pertains to their use of daily stock market reports, daily Nielsen ratings of television shows, or any other frequent measure.
For those interested in trends over a longer period than the three-day average reported daily, Gallup aggregates job approval on a weekly basis and reports weekly demographic breakouts.
Doing the Math on Obama's Job Approval Rating
A couple of readers have asked about the math in my colleague Jeff Jones' recent analysis of race differences in President Obama's job approval rating. In the story, Jeff points out that Obama's overall approval rating is 49%, while his approval among whites is 39% and among nonwhites is 73%.
The readers have apparently taken the sample sizes for whites and nonwhites given in the "Survey Methods" box at the end of story, multiplied these Ns by the group's respective approval ratings and calculated an average. The results give an average approval rating lower than the actual 49%.
The sample sizes given at the end of the article are the actual unweighted Ns for each group. The reported results, as is the case for almost all national polls done by Gallup and others, are based on weighted results, which take into account sampling variations as compared to U.S. census estimates. Because nonwhites are usually underrepresented in national samples, the final results weighted by U.S. census parameters will give the nonwhites a higher weighted representation and the whites a lower weighted representation. The weighted N for whites is 2,606, and for nonwhites 1,014. Doing the math on these weighted Ns yields the 49% overall approval rating.
Pres. Obama's Approval Slide Finds Whites Down to 39%
Obama last week fell below 50% approval in Gallup Daily tracking for the first time in his presidency, both in daily three-day rolling averages and in Gallup Daily tracking results aggregated weekly.
In his first full week in office (Jan. 26-Feb. 1), an average of 66% of Americans approved of the job Obama was doing, including 61% of non-Hispanic whites and 80% of nonwhites. In the most recent week, spanning Nov. 16-22 interviewing, his approval rating averaged 49% overall, 39% among whites, and 73% among nonwhites. Thus, since the beginning of his presidency, his support has dropped 22 points among whites, compared with a 7-point loss among nonwhites.
Given the 17-point drop in his approval rating among all U.S. adults, it follows that Obama's support has declined among all major demographic and attitudinal subgroups, with one notable exception -- blacks.
Blacks' support for Obama has averaged 93% during his time in office, and has been at or above 90% nearly every week during his presidency. Thus, part of the reason Obama's support among nonwhites has not dropped as much as his support among other groups is because of his consistent support from blacks. (With Hispanics' approval rating down five points, greater declines among Asians, Native Americans, and those of mixed races account for his total seven-point drop among nonwhites.)
The accompanying table shows how Obama's approval rating has changed by subgroup from his first full week in office to the most recent week. The only subgroup showing a greater change than whites is Republicans, down 24 points during this time. Independents' approval of Obama has declined nearly as much (down 18 points), whereas support among Democrats is down only 6 points.
Obama's strongest support comes from blacks, Democrats, and liberals -- all of whom give him approval ratings above 80%. He maintains solid support of more than 60% from nonwhites, Hispanics, and young adults.
A Closer Look at Race and Party
One reason Obama may have maintained support among blacks is their overwhelming affiliation with the Democratic Party. This is not a sufficient explanation, though, because Obama's approval rating has dropped among Democrats even as it has held steady among blacks.
In fact, it appears as though Obama's relatively small loss in support among Democrats has come exclusively from white Democrats. In late January/early February, Obama averaged 87% approval among white Democrats and 90% approval among nonwhite Democrats. Now, his approval rating among white Democrats is 76%, down 11 points, but is essentially the same (if not a little higher) at 92% among nonwhite Democrats.
Bottom Line
Obama won the Democratic nomination and the presidency with strong support from blacks and other racial minorities. In fact, according to exit polls and Gallup's final pre-election estimates, he won the election despite losing by double digits to John McCain among white voters.
Those patterns of support seem to have persisted into his presidency, with his support among whites starting out lower and dropping faster than his support among nonwhites. And though he maintains widespread loyalty among Democrats, the small loss in support he has seen from his fellow partisans seems to be exclusively from white Democrats.
It is important to note that this pattern is not unique to Obama. For example, Bill Clinton averaged 55% job approval during his presidency, including 52% among whites but a much higher 76% among nonwhites and 82% among blacks.
Survey Methods
Results are based on telephone interviews with 3,611 national adults, aged 18 and older, conducted Nov. 16-22, 2009, as part of Gallup Daily tracking. For results based on the total sample of national adults, one can say with 95% confidence that the maximum margin of error is ±2 percentage points.
For results based on the sample of 2,879 non-Hispanic whites, the maximum margin of error is ±2 percentage points.
For results based on the sample of 732 nonwhites, the maximum margin of error is ±5 percentage points.
Interviews are conducted with respondents on land-line telephones and cellular phones.
In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.
Sources: Gallup Report, Gallup.com
.jpg)
No comments:
Post a Comment