Suitably Flip

Jobs

Here comes the number of the month.  In the wake of Wednesday's upside-surprising ADP report that convinced no one to raise their expectations for today's payroll data, we're left with a consensus of 120,000 jobs added in September, with unemployment ticking back up to 8.2%.

A deviation similar to last month's ADP-actual gap would imply something closer to 60,000 jobs, but I'd be surprised if we're meaningfully under 100,000.  8.2% sounds about right, but as always, it'll be at the mercy of the cratering participation rate.  If it continues to tank, we could see the rate hold at 8.1% or even "improve" further.  If it normalizes somewhat, we could have a multiple tick jump.

Data at 8:30 am...

Update:  Wow.  114,000 new jobs (slightly below expectations, and comfortably below the pace needed to keep up with population growth).  Unemployment down to 7.8%.

This may partially explain the declining rate:

The change in total nonfarm payroll employment for July was revised from +141,000 to +181,000, and the change for August was revised from +96,000 to +142,000.

The revision largely dwells in a change of heart about whether government jobs were lost (as previously reported) or gained (tada!) over the last two months.

Rick Santelli's prediction that the Labor Department would, come hell or high water, manage to get the headline rate under 8% by election day has been validated.

Update:  Ed notes that for every job created in September, exactly 3 people abandoned the workforce entirely.

However, the number of unemployed dropped 456,000 last month, while only 114,000 jobs got added.  That either means that 342,000 people left the US, or they left the work force in one way or another.

Update:  Yeah, this doesn't smell right.  The household survey (the part used to calculate the unemployment rate, not official payroll growth, which comes from the establishment survey) shows a whopping 873,000 jobs added in September (seasonally adjusted).

How whopping?  It's the best month of the millennium to date.

Monthly change (thousands) in seasonally adjusted employed, per household survey

In fact, it's the best month since 1983 (excluding Januarys, which usually show crazy numbers, due to annual revisions, which is why they're removed from the chart).

That's just not remotely plausible.  In the last 29 years, we've had 22 quarters of growth exceeding 5%.  And never did the household job creation rate hit the ostensible peak we just experienced, with growth hovering in the 1-2% range.

We've either got a massively massaged seasonal adjustment in place, a drastic change in household survey methodology, or the number is real, the economy is booming, and ADP undercounted by 700,000.

Update:  If it's a change in methodology, it may be in how they count (or adjust for) the number of part-time jobs.

The number of people with part-time jobs who wanted full-time work rose 7.5 percent to 8.6 million, the most since February 2009.

That skyrocketing of 600,000 is not too far from the discrepancy between the household (+873k) and establishment (+114k) surveys.

Labor Secretary Hilda Solis was on CNBC a few minutes ago defending the household survey and its attempts to capture various flavors of self-employed that the establishment survey misses.

Maybe DOL is taking a cue from Obama's welfare work requirement tinkering, and redefining self-employment to include things like motivational reading and making your bed.

Update:  That about sums it up.

So BLS isn't "cooking the books" they just changed the recipe. suitablyflip.com/suitably_flip/…

— DrewM (@DrewMTips) October 5, 2012

Update:  Here's another chart for you to ponder.  This is the monthly disrepancy between the job growth as reported in the household and establishment surveys (positive numbers indicate household > establishment).

This month's gap of 759,000 just isn't something you tend to see, outside of the months surrounding massive terror attacks and financial meltdowns.

In fact, while it's a volatile series, it's been less so in recent years, without a monthly divergence nearly as large as this one since the official end of the recession in 2009.

Monthly discrepancy in payroll growth (thousands) between household and establishment employment surveys

Comments

Jobs

Here comes the number of the month.  In the wake of Wednesday's upside-surprising ADP report that convinced no one to raise their expectations for today's payroll data, we're left with a consensus of 120,000 jobs added in September, with unemployment ticking back up to 8.2%.

A deviation similar to last month's ADP-actual gap would imply something closer to 60,000 jobs, but I'd be surprised if we're meaningfully under 100,000.  8.2% sounds about right, but as always, it'll be at the mercy of the cratering participation rate.  If it continues to tank, we could see the rate hold at 8.1% or even "improve" further.  If it normalizes somewhat, we could have a multiple tick jump.

Data at 8:30 am...

Update:  Wow.  114,000 new jobs (slightly below expectations, and comfortably below the pace needed to keep up with population growth).  Unemployment down to 7.8%.

This may partially explain the declining rate:

The change in total nonfarm payroll employment for July was revised from +141,000 to +181,000, and the change for August was revised from +96,000 to +142,000.

The revision largely dwells in a change of heart about whether government jobs were lost (as previously reported) or gained (tada!) over the last two months.

Rick Santelli's prediction that the Labor Department would, come hell or high water, manage to get the headline rate under 8% by election day has been validated.

Update:  Ed notes that for every job created in September, exactly 3 people abandoned the workforce entirely.

However, the number of unemployed dropped 456,000 last month, while only 114,000 jobs got added.  That either means that 342,000 people left the US, or they left the work force in one way or another.

Update:  Yeah, this doesn't smell right.  The household survey (the part used to calculate the unemployment rate, not official payroll growth, which comes from the establishment survey) shows a whopping 873,000 jobs added in September (seasonally adjusted).

How whopping?  It's the best month of the millennium to date.

Monthly change (thousands) in seasonally adjusted employed, per household survey

In fact, it's the best month since 1983 (excluding Januarys, which usually show crazy numbers, due to annual revisions, which is why they're removed from the chart).

That's just not remotely plausible.  In the last 29 years, we've had 22 quarters of growth exceeding 5%.  And never did the household job creation rate hit the ostensible peak we just experienced, with growth hovering in the 1-2% range.

We've either got a massively massaged seasonal adjustment in place, a drastic change in household survey methodology, or the number is real, the economy is booming, and ADP undercounted by 700,000.

Update:  If it's a change in methodology, it may be in how they count (or adjust for) the number of part-time jobs.

The number of people with part-time jobs who wanted full-time work rose 7.5 percent to 8.6 million, the most since February 2009.

That skyrocketing of 600,000 is not too far from the discrepancy between the household (+873k) and establishment (+114k) surveys.

Labor Secretary Hilda Solis was on CNBC a few minutes ago defending the household survey and its attempts to capture various flavors of self-employed that the establishment survey misses.

Maybe DOL is taking a cue from Obama's welfare work requirement tinkering, and redefining self-employment to include things like motivational reading and making your bed.

Update:  That about sums it up.

So BLS isn't "cooking the books" they just changed the recipe. suitablyflip.com/suitably_flip/…

— DrewM (@DrewMTips) October 5, 2012

Update:  Here's another chart for you to ponder.  This is the monthly disrepancy between the job growth as reported in the household and establishment surveys (positive numbers indicate household > establishment).

This month's gap of 759,000 just isn't something you tend to see, outside of the months surrounding massive terror attacks and financial meltdowns.

In fact, while it's a volatile series, it's been less so in recent years, without a monthly divergence nearly as large as this one since the official end of the recession in 2009.

Monthly discrepancy in payroll growth (thousands) between household and establishment employment surveys