Last Tuesday, a CNN/USAToday/Gallup poll showed Bush’s approval rating slipping to the critical point of 50%. CNN ran three stories on it, including an interview with Bill Schneider headlined “Bush is sinking; Clark is surging.” Schneider opined that the cause of the slippage “is all men…men are getting very angry with Bush because they supported the tax cut policy, but they don’t see the payoff in terms of the job generation that was promised.” Holy panic attack, Batman!
At first glance, the poll seemed suspect. Forty-eight percent of the respondents identified themselves as either Democrats or “lean Democrat.” Andrew Sullivan noted, “It seems pretty loaded.” Well, no, it isn’t. The results have nothing to do with any bias at Gallup, and everything to do with the method Gallup uses to conducts its poll — a method that is practiced among many polling outfits. The Gallup poll treats party preference as a variable, not a constant. In other words, party preference is not a fixed characteristic, but can shift with changes in the political landscape. According to Frank Newport of Gallup, “party preference fluctuates with the circumstances. If President Bush’s popularity rises, more people will likely identify themselves as either Republican or leaning Republican. If his popularity drops, more people will likely identify themselves as Democrat or leaning Democrat.” Since Bush has been having trouble with Iraq and the unemployment rate isn’t dropping, the argument goes, more people are inclined to side with the Democrats.
Other polls, like those done by John Zogby, treat party preference as a constant. That is, they assume that party identification does not change much, at least over the short term. Thus, they “weight” their samples so that they have a percentage of Democrats, Republicans, and Independents roughly equal to the percentage in the actual population. Not all polling outfits weight the same way. According to Zogby, his poll weights “based on party registration, party turnout in previous elections, and the current landscape.” There is considerable debate among professional pollsters and academics over these methods, and the debate is not settled over which method is more accurate.
While the Gallup Poll may indeed have bad news for Bush, it has more interesting implications for Democrat-come-lately Wesley Clark. The Gallup Poll shows that Clark is leading the race with 21%; his closest rival, Howard Dean, is at 13%. However, those who lean Democrat are not actually registered Democrats, and they constitute just over 30% of those who expressed a preference for a Democratic candidate. Indeed, the influx of leaning Democrats is boosting Clark’s poll numbers. Clark draws about 21% of those who lean Democrat, almost twice as many as Dean. Given that the Gallup poll coincided almost exactly with Clark entering the race, the General is more flavor of the week than one who is deeply admired among the party faithful. This does not bode well for Clark because in the first three weeks of primary voting, most states have closed primaries in which only registered Democrats can vote.
While the Gallup poll is registering false strength for Clark, it (along with new polls by FoxNews and Newsweek) is registering some weakness for Bush. So how weak is Bush? Not very.
There seems to be three factors which are currently driving the president’s numbers lower. The first is Iraq. An absolute avalanche of negative press coverage combined with Bush’s request for $87 billion to reconstruct Iraq (foreign aid is never popular with the public) has taken its toll. Yet neither of these problems is intractable. Internet stories about troops reporting back home on how well things are going in Iraq are becoming legion. Furthermore, the public can come to accept the new expenditure on Iraq if the Bush camp repeatedly emphasizes how it is an essential part of the War on Terrorism. The President needs to put far more effort into that and into highlighting a lot of the good news from Iraq.
Second, if Bill Schneider is right that it is a “jobs issue,” then a dropping jobless rate will turn Bush’s numbers around. We are likely on the cusp of that happening. According to Don Luskin of TrendMacrolytics, “job growth requires economic growth, and with the tax cuts we’re getting more economic growth.” While increased productivity is slowing job growth right now, “increased productivity is a precursor to the next round of job creation,” says Luskin.
Finally, some of the Democrats’ attacks are likely hitting their mark. Unfortunately, Bush is in a position where he can’t really fire back. Professor George C. Edwards III, a presidential scholar at Texas A&M University, states: “If the President responds to the criticism of the Democratic candidates, he will draw more attention to the criticism. At this point, most incumbent presidents act as though they are above the political fray, going about the people’s business.” It can be grinding to watch, as there is so much Bush could say in response. But not to worry: Once the dust settles and a Democratic nominee emerges by early next year, the presidential gloves will come off.
Ultimately, the current opinion polls are not necessarily inaccurate, or the product of liberal bias in the media. But neither are they much for Bush backers to worry about.