Where are the Household survey Bulls?

Usually, when we have a lousy NFP number, the Household survey Bulls come out, touting the superiority of that metric. 

Their silence this morning was rather deafening.

The reason is fairly apparent from the BLS data: The June-July change in employment shows 34,000 less employed than last month. At the same time, those describing themselves as Unemployed in the Household survey rose 248,000.

I’m not sure which is worse: the hypocrisy or the ignorance.

Well, Household survey apologists:  Come out and defend yourselves . . .

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. DH commented on Aug 4

    Sometimes I likes your site, other times I realize that everything seems to fit your world view while you criticize other for thinking everything fits their world view.

  2. Jdamon commented on Aug 4

    DH, that was actually a very insightful post.

    I enjoy reading the site as I am by nature an optomist on the markets (why this is, since I am a very pessimistic person by nature, I’ll never know). That being said, I find it troubling when everyone has the same opinion (perma-bears or perma-bulls). I love the back and forth exchange of ideas that we sometimes get here.

    That being said, economy clearly slowing, no reason for FED to fight high oil with rate hikes (doesn’t work). So they pause with language that alerts the market that they will be watching the numbers very closely. I give that a 80% chance of happening. If they raise, it will be the economy’s death blow – full bore recession for sure. If BB is secretly a Democrat, that is what he will do. The GOP will go down in flames in Nov.

  3. JoshK commented on Aug 4

    I’ve always believed that the ADP numbers are much more reliable, up or down. I’m sure the bls couldn’t calculate it’s own employee count correctly.

  4. Trend Watcher commented on Aug 4

    Barry: Right on target. The silence is in the same league as your recent Ed Yardeni comments. If the data doesn’t agree with the theory, then the data must be wrong.

    I personally believe both Household and Establishment TRENDS (as opposed to individual data points) are worth looking at and I tend to discount any individual month data point, looking instead at some multi month moving average or year over year change and looking back at least 10 years for employment data. It’s hard to fully understand today’s numbers without having any idea of how things looked during the late 90’s when employment was booming for comparison.

    Also, there’s a lot more underneath these headline numbers as you have pointed out numerous times in this blog

  5. 23 commented on Aug 4

    “no reason for FED to fight high oil with rate hikes (doesn’t work)”

    I would much rather have the Fed in control of the economy’s throttle than the foreign interests that control most of the world’s oil.

  6. Fed commented on Aug 4

    Barry ,
    You’re so right……
    and so humble

  7. Mike_in_Fl commented on Aug 4

    First time poster, though I’ve been reading for a while. Why do people say that Fed rate hikes can’t impact oil prices when they readily accept that Fed rate hikes can affect inflation overall? Why are Fed rate hikes supposedly the cure for every other kind of inflation (inflation in the price of haircuts, milk, a head of lettuce, shoes, molybdenum, etc.) but not oil? I’ve seen a few pundits say on TV: “Well, the Fed shouldn’t use interest rates to target oil prices because it can’t impact them.” But as far as I’m concerned, that’s a bunch of BS. The Fed can hike rates, slow money supply growth, and slow the economy — something that restrains ALL prices, including oil prices, by restraining demand. Yes, they can’t prevent SHORT-TERM blips due to hurricanes, a shooting war in the Middle East, etc. But they can prevent those from fueling LONG-TERM inflation by refusing to accomodate the increased oil prices. Needless to say, for three years now, the Fed has NOT done its job in that regard (choosing instead to target — in Barry’s words — inflation, ex-inflation).

  8. Bob_in_MA commented on Aug 4

    Barry, you make a good point.

    I will take a stab at the reason behind the bad numbers in the household survey: there has been a big fall-off in casual work in residential construction. New home construction is down 15% from last year but according to the establishment data, employment in residential construction is UP.

    Small builders use mostly “contract” workers and it’s most likely small builders who’ve cut back the most.

    A lot of these workers never appear in the offical data (I’d be willing to bet some of the people who worked on our house have never paid taxes in the life.)

    That would also explain why in a period of quickly expanding home construction a few years ago the household survey showed much greater growth in the number of jobs.

  9. HHs commented on Aug 4

    DH ……. so sorry , you
    can’t have any opposing thoughts or opinions here —— you’ll get murdered —– wrongly or rightly

  10. Mark commented on Aug 4


    CNBC ran with your SBUX picture again today and one of the houses used it as one of the reasons to raise its rating on the stock! Way to go!

    Time to short is what that tells me.

  11. Cherry commented on Aug 4

    My view is, the late 90’s “hiring” boom was made up. It had to be. The NFP would come out with rediculously high totals in payroll additions, sometimes over 300000 every few months and 200000-300000 a “sequence” of months.

    Simple put, that seems about a impossiblity other than the fairytale economy to make people “feel good”. It didn’t begin with Bush(though his BS has spun it to a whole new level lol).

  12. JEl commented on Aug 4


    good insight

    the bubble economy of the 90’s brought great #’s but at a cost that we live with today …. but it was a fraud………. we can complain about today’s numbers , yet job growth is a disaster today as we still work off that scam …. just another sidebar to the tech stocks unravelling the last 6 years

  13. phil commented on Aug 4

    where’s diva with her chart of the 30 yr. T bond muaahaha

  14. szara commented on Aug 4

    I really disagree with the paranoia that is often expressed about the BLS and the employment numbers. There are limitations to what the BLS does, and you often have to read the fine print, but belief that the books are cooked, either under Bush or Clinton, is way off base.

    I think that Bob in MA may be on to something in suspecting that the divergence between household based emp and payroll emp is in the casual work in residential construction. It is very true that the there is a large fringe of construction labor, more often in remodeling than in new construction, that is not going to show up in payroll numbers.

    If you look at table A5 in the Employment situation report, you see that all the employment loss per the household survey is in the self-employed segment, which would support Bob’s idea.

  15. TDM commented on Aug 4

    I will defend the household survey.

    The household survey is a more accurate measurement of employment, since it directly measures employment. It is less precise because of the smaller survey size. However, comparing survey employment to last month is no less precise than comparing it to last year since each month is an independent measurement. The payroll survey is a less accurate measure of employment since it measures something different, payroll. The payroll survey will become less accurate over time because of compounding birth/death and other errors. Unemployment claims are a very precise, but inaccurate measure of employment.

    The BLS CPS figures show that employment is strong, averaging an increase in employment of 190k per month over the past 3 years. With over 210k per month over the last 3 months, I am still bullish on jobs. One month does not a trend make.

    Increase 3yr, 1yr, 3mo
    Survey 6851, 2218, 641
    Payroll 5532, 1737, 337

  16. Cherry commented on Aug 4

    Yeah, but historically, it isn’t that “strong”. You are confusing being labelled with a “job” vs. actually having a job.

    I figure the unemployment rate went up because the Gov. finally got around to cleaning out those “self-employed”(aka really unemployed, maybe in building as Bob believes) people and put them on the unemployed list. Under the Kennedy-Clinton econ model, unemployment was 5.3%, thus only a .5 departure from the Bush overhauled model, the closest in months. We are nearing full employment…….finally. But still are not their yet, such as is, wages still aren’t going to start signifigently going up yet………..and they won’t because your 1 month “blip” is going to turn into a Fall flop. Augest will be even worse and show more decline.

    Bleak days and bitter nights.

Read this next.

Posted Under