Economics, Pethokoukis

Why Gallup’s chief economist says the big drop in the unemployment rate ‘should be discounted’


You can count pollster Gallup as another observer questioning Friday’s job report, in particular the big drop in the unemployment rate. In a blog post, Dennis Jacobe, Gallup’s chief economist, writes that he seriously doubts whether the economy created nearly 900,000 jobs in September, as measured by the Labor Department’s seasonally adjusted household survey.

While the payroll survey of businesses and government showed just 114,000 net new jobs created, the household survey showed a jobs boom, and it’s the latter which is used to calculate the unemployment rate. Jacobe:

The problem is that even though the Household survey tends to be very volatile, this decline seems to lack face-validity, particularly after the prior month’s numbers. The consensus estimate was that the government would report that the unemployment rate was unchanged at 8.1% in September. GDP growth was 1.3% in the second quarter and seems to be no better this quarter. The government’s Establishment survey shows there were 114,000 new jobs created in September — very close to the consensus of 113,000 — and not sufficient to lower the unemployment rate.

A quick comparison of the government’s seasonally adjusted and unadjusted employment data seems hard to reconcile with the weak economy. For example, the government shows the number of employed workers increasing by 775,000 in September from August on an unadjusted basis. This surge in hiring seems surprisingly large given the current economy, not to mention the even larger adjusted increase of 873,000. Similarly, the number of unemployed declined by 954,000 in September on an unadjusted basis. This is reduced to a smaller adjusted decline of 456,000 — but both numbers are also surprisingly large.

His bottom line: “The Household results should be discounted. … The obvious conclusion is that a new employment measure is needed.”

Jacobe’s preferred measure is one Gallup has created, Payroll to Population, or P2P. It is simply “the number of Americans employed full-time for an employer as a percentage of the U.S. population.” P2P is based on 30,000 phone interviews a month. And what did the P2P show for September? This: “The P2P deteriorated slightly to 45.1% in September from 45.3% in August, suggesting the real jobs situation was essentially unchanged last month.”

Here’s why I like this idea: The Gallup measure more or less reflects — accurately I think — what we see in other broad measures such as the labor force participation rate and the civilian-employment ratio. They show a depressed labor market, not one suddenly gaining huge momentum.


21 thoughts on “Why Gallup’s chief economist says the big drop in the unemployment rate ‘should be discounted’

  1. Since the release last week, people have been defending BLS and discounting conspiracy theories. Okay. Even IF the 50,000-phone call survey conducted came up with 775,000 new jobs (and someone didn’t “inadvertently” add a zero to the end of the figure), the top, top economists at BLS -according to Hilda Solis- should have known the survey was flawed based on their top, top knowledge of the state of the economy.

    Now, maybe they can’t “throw it out” but they certainly can weight it appropriately, right? And if not, then someone needs to look at how (or where) the survey was conducted.

    No one at BLS can point to any reasonable measure that would trigger a jump in hiring like this. No (bogus) “Jobs Bills” were passed. No tax incentives passed or extended. No Executive Order promising college tuitions will be paid by the taxpayers if only company owners would suddenly expand their payrolls.

    The whole thing defies logic- and when that happens, on the scale that it did and right before an election, how can one NOT look at this from a conspiratorial prism?

  2. Here’s the report I want to see followed monthly if possible, or quarterly, if that’s the best we can do:
    How much was collected in payroll taxes for the period
    How many companies filed
    How many employees were covered by those filings
    How much was collected in the self-employment tax
    How many people paid the self-employment tax
    How many Social Security numbers were associated with two jobs? Three or more jobs?
    How many people capped out of the Social Security tax
    This would give us a better handle on what the labor market is actually doing than all of those surveys and their adjustments.

    • I too would really be interested in a 1-year comparison of unemployment based on:
      the BLS survey
      the Gallup survey
      payroll-to-population derived unemployment from IRS
      payroll-to-population derived from any other payroll data as Popshobby mentioned (I think ADS does this too ..? I may have the acronym wrong!).

  3. Just a thought. The household survey polls only 60,000 or so and its statistical algos do the rest. And as we all know, the household survey looks like an EKG of a hummingbird.

    I would like to know if the questions asked of those polled in the household survey are always the same questions? Do these questions change from time to time? How might this tweak the outcome? And since the computer does a lot of the work, would there be an opportunity for just a small change in the wording of a question to rig a desired outcome?

    Just a thought.

    But without any doubt, the results do confirm the $4,000 or so reduction in household incomes over the past few years. No doubt that part time work (almost 600,000) doesn’t translate in full time income or full time spending. Part time work, better than no work at all, does little to buoy economic activity. It only provides the illusion of more jobs as it does little more than spread near-unemployment around, thereby reducing the statistical UR.

  4. This certainly is interesting. One point of caution: their numbers are volatile and subject to seasonal fluctuations. The last two Septembers fell from August, so we would need to do some more research before making a judgement call on this. Was this year’s decline more or less severe than normal?

  5. I’m just amazed that Jimmy P has discovered politics in the numbers. I’m sure this never happened under a Republican Administration. The question is WWRD ? What Would Reagan Do. If anyone on the board can answer that ? with confidence then please enlighten the rest of us “sheeples”

  6. Fast & Furious comes to the BLS.
    Oblabla, Holder, Solis, Clinton, Vilsack, etc…..all the same…Commies…ugh, pardon me; Progressives.

  7. Fake unemployment numbers, illegal foreign campaign contributions, everything Romney said in the debate was a lie, etc., etc., etc…..The corruption stops on November 6th!

  8. “think any one in your household will ever find a job?” If yes – put them down as gainfully employed. That is the ONLY way the household survey could show 837,000 new jobs. Or someone just added three zeroes after 837. Either one.

  9. Unemployment dropped by nearly 1 million in September? HMMMMM…… How about, “A bunch of out of work college kids on summer break went back to college and quit looking for jobs?” Just a thought.

  10. The answer was in the article, itself.

    For whatever reason, the Seasonal Adjustment from Aug to Sept is Huge. There’s quite likely to be a large “giveback” in the next number.

    Btw, be careful with any numbers that don’t take into account the large number of baby boomers retiring. I’ve read where that can skew the “employment to population” ratio as much as 3%.


    Contrary to what most reports have been saying, this report says almost all of the 873,000 new jobs were full time! Many experts explained the discrepancy between the BLS household survey and the employer survey as due to the part-time numbers. But if in fact the
    household survey is also measuring full-time jobs, is close to 1 million new FULL-TIME jobs in one month really believable? Why is the household survey new jobs count 8 times higher than the employer survey if they both are measuring full-time jobs?
    For statistical studies, when they most come under question is when they are INTERNALLY inconsistent.
    At the rate the BLS is claiming our economy is now “booming” we can match the Romney claim of 12 million new jobs over 4 years in just one year!
    I think you agree this number is not believable but that is indeed what the BLS data seems to be saying near the bottom of this table:

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>