Political Animal


September 06, 2012 12:58 PM Obama’s “Premonition”

By Ed Kilgore

Sudeep Reddy of the Wall Street Journal’s Real Time Economics blog offers a small but important factoid to consider in anticipation of the president’s acceptance speech in Charlotte tonight: he’ll know the August Jobs Reports numbers the rest of us will wait until 8:30 AM EDT on Friday to hear:

Mr. Obama doesn’t have to wait until the formal release to see the numbers. Under a decades-long practice, a select group of U.S. officials learns the contents of each month’s jobs report on the Thursday evening before its release. The Bureau of Labor Statistics delivers the information sometime Thursday afternoon to the White House Council of Economic Advisers, which analyzes the data and prepares a memo for the president. (The CEA chairman or director of the National Economic Council often informs the president in person.)

Huh. I did not know that. Reddy suggests this could affect the final speech-tweaking, as it seems to have done in 2004, when George W. Bush faced an analogous situation:

Assuming the president gets a briefing on Thursday before his speech, what might he do with it? He isn’t exactly getting on stage, for one of the most important speeches of his career, to recite Labor Department data. But he conceivably could tweak his adjectives in describing the economy if the employment report is a surprise in either direction….
Eight years ago, President George W. Bush faced a somewhat similar scenario. His speech at the Republican convention in September 2004 came the night before a monthly jobs report, following a year of choppy data (and after what was routinely described as a jobless recovery). The prior jobs report showed that payrolls rose just 32,000 in July, so he would have had good reason to worry about more trouble in the labor market. But his convention speech offered plenty of upbeat commentary about the economy, which means he either knew the next round of data or didn’t care. The report released the morning after his speech showed a substantially better performance, a gain of 144,000 jobs in August.

It’s probably not something to obsess about during Obama’s speech. Maybe Reddy’s wrong about the advance notice (though it certainly sounds plausible). The consensus predictions last I heard was for a report indicating 130,000 net new jobs in August (though there is considerable variation in what different economists expect) , a level that’s neither great nor bad; if the actual numbers are in this range, they’d probably have very little effect on presidential rhetoric even if Obama is otherwise inclined to act on his “premonition” for Friday. And as the whole hep world is saying today, the main burden of his speech is to project a post-election vision and governing agenda, not to reframe his economic record (Bill Clinton’s already done a very good job of that), much less talk about micro-trends.

Still, it’s intriguing to know that the president may head to the podium with a bit of knowledge about something the business and political communities will be spending a good part of Friday morning discussing, at least if the numbers don’t almost exactly meet expectations. If you get bored during the speech, you can certainly look for hints, at the risk of buying into a snails-eye-view of the presidential election.

Ed Kilgore is a contributing writer to the Washington Monthly. He is managing editor for The Democratic Strategist and a senior fellow at the Progressive Policy Institute. Find him on Twitter: @ed_kilgore.


  • Pete on September 06, 2012 1:31 PM:

    According to my BLS and CEA contacts, POTUS does indeed get the numbers early. So does the Fed Chairman.

  • pattyp on September 06, 2012 1:37 PM:

    I'm a little confused now. (Not an uncommon occurrence.) This morning I read about the ADP monthly jobs report that was released today showing 201K new private sector jobs. I understand the ADP report is not official, but would the BLS report be substantially different? If so, why?

  • Art Hackett on September 06, 2012 1:50 PM:

    The ADP is private sector only. If there have been big government layoffs, they could offset gains in the private sector. ADP is in the business of managing company payrolls. They know exactly how many checks they cut, how many are new to the system, and how many dropped off. They massage those numbers to represent the total private sector economy using methods similar to those the BLS uses to arrive at their numbers from a survey of employers. Sometimes the two numbers are off from each other. Interestingly, last month they were exactly the same (if you count the private jobs only from the BLS report)
    The other number out tomorrow is the unemployment rate which is arrived at through a different methodology. Unfortunately for the President, strong job growth has a tendency to draw discouraged workers into the job market which can cause the unemployment rate to nudge up. If that happens, expect Republicans and the TV networks to headline.

  • stinger on September 06, 2012 2:00 PM:

    The BLS, as part of the Executive Branch, would naturally report to its boss before releasing important information publicly. SOP in the private sector as well as in government. Why would this be news?

  • pattyp on September 06, 2012 2:12 PM:

    Thanks, Art. I actually know who ADP is since I've had jobs processing payroll. LOL. I figured that much about the net job gain/loss, but still, wouldn't the BLS private sector numbers be at least somewhat similar to ADP's?

  • Oscar Leroy on September 06, 2012 2:44 PM:

    "130,000 net new jobs in August, a level thatís neither great nor bad"

    That's bad. It barely matches the increase in workers for that month, let alone chips away at the numbers of long-term unemployed.

  • Pete on September 06, 2012 3:13 PM:

    Also, the ADP report is based on a subsample of ADP clients, so it doesn't capture the entire universe. The BLS sample is larger, and also includes non-ADP firms. It's also benchmarked annually to unemployment insurance payments, which basically captures everyone out there working. The BLS is the offical measure, and the gold standard.

    According to ADP the correlation between the two measures of job growth is 0.95, so it's pretty good. But ADP had a big miss in June; it way overshoot the BLS numbers (+172,000 for ADP, and +64,000 for BLS).

  • dweb on September 06, 2012 3:14 PM:

    Rush today already anticipating this tonight and today is claiming that it is proof that the Obama team is cooking the job figures to their advantage and that his followers should not be downhearted if the numbers show an improvement in jobs.

    Think about that....more jobs and Rush's followers are gonna be real unhappy.....and so will Rush. His wet dreams are drifting away.

  • Doug on September 06, 2012 4:43 PM:

    If I understand correcly, 130,000 new jobs a month roughly, very roughly, covers those entering the job market, but does nothing to reduce the number of unemployed.
    Therefor, I'd expect that a bad numbers report for August MIGHT cause President Obama to include again call for providing Federal funds to rehire/call back to work those teachers and members of police and fire departments that have been laid off due to budget constraints in the various states. However, even if the current Congress DID pass such legislation before the election, and I'm NOT holding my breath on that, I expect that it would still take 2-3 months before such actions would start showing up in unemployment figures as many of those rehired/recalled would be paying off bills already run up before spending on any non-necessary items.

  • dnno1 on September 07, 2012 2:05 PM:

    Just one comment. ADP's numbers are based on the size of their database of an average of 400,000 payrolls covering some 23 million employees. That's a pretty large sample size (much larger than the one used by the BLS). Another thing to note is that the BLS methodology admits that the confidence interval for the month to month changes in non-farm employment for their establishment survey is 100,000. That means that there is 90% confidence that the true change in employment for the month of August was between -40,000 and 196,000, which would mean that it is very well possible that the economy added almost 200,000 jobs (on the flip side it may have lost 40,000 jobs). Additionally, the BLS also admits that they have an inability to capture employment generated by new firms (something that ADP may not have a problem with). I think that both reports should be looked at side by side and used to rationalize what may actually be going on taking into context the error.