Grading PolitiFact: Michele Bachmann and the promise

It was "dumb" for President Obama and his aides to promise that unemployment would not surpass 8 percent if the stimulus act passed, a top House Democrat said Tuesday.
--The Hill

The issue:

The fact checkers:

Robert Farley:  writer, researcher
Martha Hamilton:  editor


Analysis:

Recently PolitiFact has deliberately drawn attention to the fact that Michele Bachmann (R-Min.) was the only political figure with an appreciate number of ratings who had failed to have one other than "False" or "Pants on Fire."  This claim rates "Barely True" according to PolitiFact, so that record evidently falls by the wayside.

The claim, of course, surrounds the Obama administration's sales job on the stimulus package.  That bill was the first major legislation to pass after Obama took office.

First, the Bachmann version:

Two years ago, when Barack Obama became our president, unemployment was 7.8 percent and our national debt stood at what seemed like a staggering $10.6 trillion dollars. We wondered whether the president would cut spending, reduce the deficit and implement real job-creating policies. Unfortunately, the president's strategy for recovery was to spend a trillion dollars on a failed stimulus program, fueled by borrowed money.

The White House promised us that all the spending would keep unemployment under 8 percent. Well not only did that plan fail to deliver, but within three months the national jobless rate spiked to 9.4 percent. It hasn't been lower for 20 straight months. While the government grew, we lost more than 2 million jobs.

Let me show you a chart: Here are unemployment rates over the past ten years. In October of 2001, our national unemployment rate was at 5.3 percent. In 2008 it was at 6.6 percent. But just eight months after President Obama promised lower unemployment, that rate spiked to a staggering 10.1 percent. Today, unemployment is at 9.4 percent with about 400,000 new claims every week.

NPR (yellow highlights added, denoting the portion quoted by PolitiFact)

Now the PolitiFact version:

On several occasions as the stimulus was being considered, President Obama warned that if "dramatic action" were not taken, the recession could linger for years, and "the unemployment rate could reach double digits." But we could find no instance of anyone in the administration making a public pledge along the lines of "if we pass the stimulus, we promise unemployment will stay below 8 percent."

I respectfully recommend dropping "promise" from the phrase of the type for which PolitiFact searched.  After all, PolitiFact lists dozens of "promises" from Obama made where neither "promises" nor a parallel term occurs..  We use "promise" to describe a statement such as "if we pass the stimulus, unemployment will stay below 8 percent."

PolitiFact continues:

(T)he claim has its roots in a Jan. 9, 2009, report called "The Job Impact of the American Recovery and Reinvestment Plan" from Christina Romer, then chairwoman of the president's Council of Economic Advisers, and Jared Bernstein, the vice president's top economic adviser.

A chart from that report, indeed, carried the gist of the perceived promise:

click image to enlarge

PolitiFact may be correct that the report (with the accompanying chart) by itself accounts for the perception of a promise from the Obama administration.  If that's the case, then the quotation from Rep. Barney Frank (D-Mass.) probably indicates Frank's dismay with the administration's poor attempt to control its message.  The chart, minus what passes for the "fine print" of the surrounding report, does send a fairly unambiguous message that the "Recovery Plan" will prevent unemployment from exceeding 8 percent.

Based on the evidence before me, it's not quite fair for Bachmann to count this as a promise that unemployment would not exceed 8 percent.  However, her method is similar to the way PolitiFact picks out promises to subject to the "Obameter" and other similar ratings.  Granted, the anticipated effect of legislation is one step removed from the implementation of legislation.  Of the two, an intent to legislate is far easier to fulfill.  In spite of that, the similarity is worth highlighting.

One other thing worth noting:  The chart shows immediate improvement in the employment picture with passage of the "Recovery Plan."  Romer later defended the efficacy of the plan with the claim that it was too early to measure the effects.

The chart ended up as a ball and chain for the Obama administration.

We return to the PolitiFact narrative:

(W)hat we saw from the administration in January 2009 was a projection, not a promise. And it was a projection that came with heavy disclaimers.

The chart was ineffective in communicating disclaimers.   And one of the disclaimers PolitiFact emphasizes hardly appears relevant:

There's also a footnote that goes with the chart that states: "Forecasts of the unemployment rate without the recovery plan vary substantially. Some private forecasters anticipate unemployment rates as high as 11% in the absence of action."

The fact check does not concern administration estimates of unemployment in the absence of action.  The footnote exists, right enough, but it doesn't contribute anything to the fact check.

Nearing conclusion time:

(T)here is an inherent uncertainty in economic forecasting. And how can you ever prove that if the unemployment rate got to X percent, it would or would not have gotten a point or two higher if not for the stimulus? The implication of Bachmann's comment is that rising unemployment rates prove the stimulus didn't work. Most economists don't agree -- and argue that without the stimulus, unemployment would have been worse -- but it's difficult to empirically prove one way or the other.

PolitiFact is partly correct about a flaw in Bachmann's argument.  Rising unemployment does not necessarily mean the failure of the stimulus bill.  But certainly the prediction was poor and that rightly reflects on the administration.  After all, the state of the economy is worse than what the chart predicted in the absence of the stimulus bill.  It's fair to hit the administration regarding its poor predictions, and the fact that it's hard to empirically prove that the stimulus worked or failed isn't much help in rescuing the administration's credibility.

And the wrap:

The White House claimed the stimulus would improve the employment picture and applied that to the baseline for projected unemployment going forward. As Krugman said, the baseline projection turned out to be way off (most economists at the time underestimated the severity of rising unemployment). Bachmann's claim suggests Obama was offering some sort of guarantee the stimulus would keep the unemployment rate below 8 percent. The administration never characterized it that way and included plenty of disclaimers saying the predictions had "significant margins of error" and a higher degree of uncertainty due to a recession that is "unusual both in its fundamental causes and its severity." In short, it was an economic projection with warnings of a high margin for error, not a take-it-to-the-bank pledge of an upper limit on unemployment. 

So we find Bachmann's claim Barely True.

I think I agree with the sense of the final finding.  Bachmann's statement had a basis in fact but tended to mislead without additional explanation.  The final ruling surprised me a bit considering the content of the story.  PolitiFact often labels statements with a foundation in the truth "False" rather than "Barely True."

PolitiFact deserves the most pointed criticism for its flirtation with hypocrisy in identifying campaign "promises" using methods akin to those that helped identify this "promise" from the White House.

The grades:

Robert Farley:  C
Martha Hamilton:  C

I don't hold the writer and editor responsible for the stench of hypocrisy in this case.  Though they may have participated at some point.

Unlike PolitiFact, I will remind readers that totaling my grades for individual PolitiFact staffers tells you little more than how I have graded them.  I evaluate stories that I expect will exhibit problems, so selection bias colors the grade averages.  Don't waste time playing statistics with the grades if it's not to learn something about me.