Last week, Nate Silver flagged an intriguing question in the latest NBC/Wall Street Journal poll:
When you think about the current economic conditions, do you feel that this is a situation that Barack Obama has inherited or is this a situation his policies are mostly responsible for? (ASK ONLY OF RESPONDENTS WHO SAY INHERITED) And how much time would you say that Barack Obama has before his policies are mostly responsible for the country's economic conditions?
The question gets at an issue that has been on a lot of minds lately: Just how patient will Americans be with President Obama? Or, as Republican pollster Bill McInturff puts it, "when does gravity begin?" (McInturff conducts the NBC/WSJ poll along with Democrat Peter Hart). The same survey, like most reported in recent weeks, shows a huge gap between Obama's approval rating and attitudes about the economy. Specifically, a large majority (60%) approves of Obama's performance while virtually all (92%) are dissatisfied with the condition of the economy (70% say they are very dissatisfied).
As of right now, the NBC/WSJ survey tells us, Americans express patience. Only 8% hold Obama responsible for the downturn (as per the question above), and 85% say they expect the economic recession to continue for at least a year (49% say it will not be over for at least two years). As Nate Silver points out, the majority of voters say they will hold Obama accountable for the economy at about the same time they expect the recession to have ended.
But I want to emphasize the warning Bill McInturff hinted at in the MSBNC story: While Americans seem to be giving Obama a "long leash" on the economy, they are "notoriously impatient people."
There is also a technical issue: We should be wary of hypothetical, self-reported predictions of future patience from survey respondents. Surveys are good at taking "snapshots" of current attitudes, but they often fall short when asking voters to predict their future opinions.
Consider three examples:
In June 2007, CBS director of surveys Kathy Frankovic wrote a column reminding us of the pitfalls of the most famous hypothetical question: for whom would you vote "if the election were held today?" As she wrote those words, the leaders in the national surveys were Rudy Giuliani and Hillary Clinton. And in Iowa, Giuliani and John Edwards led among polls of likely caucus goers. Mike Huckabee, the ultimate winner of the Republican caucuses, was still mired in single digits even in Iowa. The point is not that these polls were "wrong," only that they were snapshots. They told us something about how voters felt at the time but fell short as predictions of future support.
Via email, Frankovic highlighted a better example. In 1998, just before and after the House of Representatives voted articles of impeachment against President Clinton, CBS News and the New York Times tracked responses this question (among others):
Given what you know right now, do you think it would be better for the country if Bill Clinton resigned from office, or do you think it would be better for the country if Bill Clinton finished his term as president?
In the week just prior to the House vote (December 13-15, 1998), 30% of Americans said they thought Clinton should resign, while just 67% said he should finish his term. The survey included another, more hypothetical question (emphasis added):
If the full House votes to send impeachment articles to the Senate for a trial, then do you think it would be better for the country if Bill Clinton resigned from office, or not?
Under that hypothetical scenario, 43% said they thought Clinton should resign.
How well did respondents predict their future attitudes just days before the hypothetical became real? Not very. On December 19, the House voted to impeach Clinton and hold a trial in the Senate. The CBS/Times pollsters went back into the field to conduct a second interview among those previously polled the week before. Just after the House vote, the percentage favoring Clinton's resignation held steady at 31% (and the panel-back design allowed the pollsters to confirm that roughly a third of those that had hypothetically supported Clinton's resignation in the event of an impeachment took the opposite position when the theoretical became real). Support for resignation remained essentially unchanged as they drew fresh new national samples and tracked again on January 3-4, 1999 (28% favored resignation) and January 10-11 (29%).
A third example comes via email from McInturff (who had no idea I was planning to write about the Clinton impeachment): "In 1998, we spent months and months asking people how would they feel if they learned the Monica story WAS true. That was interesting to track, but it had little bearing on what happened in August when they did, in fact, learn it was true.."
The point, again: Polls are best at providing snapshots of current attitudes. Results can mislead us when pollsters ask respondents to predict future attitudes and when we take their answers too literally.
I had a hunch that McInturff shared my skepticism, and he confirmed that intuition via email:
This is not a question we can ask the electorate confident they really know the answer. It would be interesting to track attitudinally, but not determinative. Voters do not "know" and will not accurately predict when, in aggregate, this economy becomes "his" in the public's mind, but all previous research suggests at some time it will.
McInturff suggests that the so-called "right direction wrong track" question" provides a better way to watch Obama's potential to sustain his current approval ratings. The latest NBC/WSJ poll showed showed a twenty-nine point increase since the election (from 12% to 41%) in the percentage of Americans saying the nation is "generally headed in the right direction" rather than "off on the wrong track" (a trend confirmed by others polls as illustrated by our right direction/wrong track chart). Most of that increase came from Democrats.
In Laura Meckler's Wall Street Journal poll story, McInturff explains the implications:
Typically, he said, presidential approval ratings hover about 20 percentage points above the national "right direction" number. With that figure at 41%, he said, Mr. Obama's approval rating, at 60%, is sustainable and high enough to allow for major victories.
The key question, as McInturff frames it, is whether the "right direction" number will remain at or near 40%, or whether it will recede back to the 25% measured by most surveys in late January. A 25% right direction number would provide "a major barrier" to Obama sustaining his current approval rating. If that number remains at or near 40%, however, Obama could maintain a job approval rating in the high 50 percent range. So, he says, "let's wait and see if [the right direction number] can be sustained at this high level given the very serious economic news."
PS: A related issue, also flagged by Nate Silver, is that voters typically "react slowly to changes in economic well-being." Last week, ABC's Gary Langer blogged the details, including this finding following the deep recession in 1990-1991:
Confidence in current economic conditions, as measured in the ongoing ABC News Consumer Comfort Index, didn't regain its pre-recession level until late 1994, more than three and a half years after that recession technically ended.