A couple weeks ago, two of the nation’s most respected polls caused quite a kerfuffle.
It started with the ABC/Washington Post poll, which found President Obama’s approval rating tumbling from 50 percent approval, 46 percent disapproval in early February to a mirror-image 46 percent approval, 50 percent disapproval in mid-March.
Thousands of words were written explaining the president’s somewhat mysterious decline. Some ascribed it to gas prices, others to GOP criticism. Still others complained about the White House message. “President Obama’s approval rating has taken a hit in a new poll that shows a growing frustration with the president’s handling of the economy,” opined the Los Angeles Times. “The … drop coincides with a spike in gas prices and an increase in criticism from Obama’s Republican rivals on the issue, even as the economy has shown noticeable signs of growth.”
Democrats were nervous. An improving economic situation, along with a series of polls showing the president at or near 50 percent approval, had created a sense of optimism that the latest figures from these respected polls appeared to dash.
A new narrative was being written, or at least the earlier one revised. Webs of meaning were spun and pundits could not stop talking about the altered political dynamic.
Whoops. It may have all been a false alarm. Other respected polls conducted at virtually the same time found no such drop — indeed, they appeared to reveal small gains.
The higher, earlier number in the ABC/Post time series was +4 (that is, four points more approve than disapprove), recorded Feb. 1-4. At the same time, Gallup showed approval and disapproval closer to equal, at 48 percent to 46. What happened in the Gallup data over the same time period, during which the ABC poll found a “significant” drop? It got a touch better for the president, with approval going up one point and disapproval falling three, so that +2 became +6 in Gallup, while +4 was turning into -4 in the ABC data.
Pew was in the field the same days as the CBS/Times poll in February but came out with +4 instead of +7. And the Pew trend? While the CBS/Times data went from +7 to -6, Pew moved from +4 to +9.
Following all the numbers is difficult, but suffice it to say that while two polls showed the president’s approval rating falling, two others said it was improving slightly. Of course, it so happens that the two more negative portraits were painted by polls associated with the two most politically influential newspapers in the country, both of which are also affiliated with TV networks, guaranteeing that the coverage of those two surveys dominated the discussion. In relative terms, data from Gallup and Pew tended to get lost in the shuffle.
Bringing the story up to date, in just the last four days Gallup went from -3 to +3.
Buried in this too-complex compilation of facts are some lessons: Polls bounce around. Even seemingly substantial variations often prove to be nothing more than noise. Second, no one poll, indeed no two polls, enjoy a monopoly on truth. When they are all saying the same thing and pointing in the same direction, confidence is usually (but not always) warranted. (Note Nevada last cycle, when every poll but one — ours, I am happy to say — pointed to a Reid defeat.) When different polls lead to different conclusions, caution is well-warranted.
Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the Majority Leader of the Senate and the Democratic Whip in the House.