The unemployment rate dropped to 9.1% according to the July Employment Situation Report. The headline jobs number was +117,000 vs a Bloomberg consensus range of 25,000 to 125,000. Without the seasonal adjustment, things look substantially worse. 1.32 million people lost their jobs in the month of July. These job numbers are highly focused into a single industry that undergoes a rather pronounced phase transition between June and July: public schools.

The largest seasonal factor was government, which tends to shed only slightly fewer than the 1.22 million jobs it shed in July. This translated to only -37k after the seasonal adjustment. But by far the greatest share of these jobs are in the local government sector, in education.

The biggest gains on a seasonally adjusted basis were made in Education and health services (+38k). This is followed by Professional and business services (+34k). The biggest loser was Government (-37k), which was followed by Financial activities (-4k).

Table A15 provides alternative measures of labor underutilization. U-6, the broadest measure (frequently called the “real” unemployment rate), dropped from 16.2% to 16.1% (unadjusted it increased from 16.3% to 16.5%, though this also included a revision from 16.4%).

For those interested in the Birth/Death model, it subtracted 18k jobs to the unadjusted number. Once again, the seasonal adjustment factor is on total jobs.

This was somewhere between a good and mediocre report, depending on who’s estimates for sustainable employment rate growth you use. As a reminder, depending on who you listen to, we need between 100k and 175k jobs just to maintain the unemployment rate with our current demographics at a constant participation rate.

The unemployment rate crept up to 9.2% according to the June Employment Situation Report. The headline jobs number was +18,000 vs a Bloomberg consensus range of 65,000 to 160,000. This disappoints. Without the seasonal adjustment, things are mixed. The unadjusted jobs number is +376k, but the unadjusted unemployment rate is 9.3%.

The biggest losers (seasonally adjusted) were Government (-39k) and Financial Activities (-15k). The biggest winners were Leisure and Hospitality (+34k) and Durable Goods Manufacturing (+15k). Some of the categories hid larger moves based on their composition. Education and Health Services were a wash (0 jobs), but (private sector) Education was -17.4k while Health Services was +17.4k.

Table A15 provides alternative measures of labor underutilization. U-6, the broadest measure (frequently called the “real” unemployment rate), increased from 15.8% to 16.2% (unadjusted it increased from 15.4% to 16.4%).

For those interested in the Birth/Death model, it added 131k jobs to the unadjusted number. Once again, the seasonal adjustment factor is on total jobs. If you consider the Birth/Death model to be bogus (which, considering that it adds the same number of jobs in June during boom periods as during bust periods, isn’t the most crazy theory) you can’t just subtract the Birth/Death number from the headline number and get anything meaningful. Also, keep in mind that the model is significantly better than it was 2 years ago. At least now the revisions come more than once a year.

This was a terrible report. It’s unusual, but I don’t think the major scare headlines are missing much this time. As a reminder, depending on who you listen to, we need between 100k and 175k (up from 150k because I started listening to more people) to maintain our unemployment rate at a constant participation rate. At some point, demographics will kick in where we can maintain a constant unemployment rate with fewer (even negative) monthly jobs. If things continue at this rate, there is one clear solution to both the boomer retirement problem and increases in food and energy prices. At that point, we should eat and burn our elderly.

The Employment Situation Report was released today. The headline jobs number was +54,000 and the headline unemployment rate was 9.1%. On an unadjusted basis, these numbers were +682k and 8.7% respectively.

Table A-15 provides the alternative measures of labor underutilization. U-6, the broadest measure (frequently called the “real” unemployment rate), dropped from 15.9% to 15.8% (unadjusted it dropped from 15.5% to 15.4%).

March and April were both revised lower (fewer jobs added).

For those interested in the Birth/Death model, that added 206k jobs to the unadjusted number. If you consider the whole thing bogus, the correct adjustment is not to subtract it from the headline (seasonally adjusted) number. Be careful, some people do this. Note that it isn’t black magic to take the term out. You find the adjustment factor (ratio of total employment numbers, not monthly change), subtract the part you consider bogus from the unadjusted number, and reapply the adjustment factor. Now compare the adjusted number to the previous month’s adjusted number (you should take out the bogus component there too). That’s your revised headline number. Unfortunately, since the change in the model, it’s been difficult to determine how bogus the model is. Previously, it was a funny joke (March numbers during a recession were essentially the same as the March numbers during a nice expansion).

This was a mediocre report. It was the weakest report since September. However, because of the ADP report, it should have largely been expected. The biggest surprise was the lack of “surprise” and “unexpected” from headlines, though (CNN wins with “NASTY JOBS SURPRISE”). If you apply others’ post-ADP revisions to their expectations to the Bloomberg consensus, you’d get about 100k (down from 170k). With this extremely shoddy methodology, you’d still consider this a downside surprise, but you might also shrug your shoulders and say, “Meh.”

The unemployment rate fell in March to 8.8% according to the Employment Situation Report. 216,000 jobs were added on a seasonally adjusted basis, above the “sustainability” rates of both Bernanke and Reich. This was safely inside the rather wide Bloomberg consensus range of 150,000 to 310,000.

If you want to feel particularly good about this news, consider that this is a down-revision month; 925,000 jobs were added on an unadjusted basis. 117,000 of those are due to the adjustments from the Birth/Death model, but the rest correspond to the establishment survey.

Table A-15 provides alternative measures of labor utilization. U-6, sometimes referred to as the “actual” unemployment rate, fell from 15.9% to 15.7% on an adjusted basis and grew from 16.7% to 16.2% on an unadjusted basis.

This was a solid report (typical for a March). Given how long it’s been since the bottom in employment, though, I would like to see temporary help services somewhere other than the top of the list. The only categories that produced more growth were health care and professional and technical services. Of course, professional and technical services is more than 3x the size of temporary help (health care is more than 6x the size of temp), so on a percentage basis, temp workers are still growing faster. This was acceptable when we thought the jobs market was just turning around. Now, I’m finding fewer reasons to celebrate.

The unemployment rate fell in January to 9% according to the Employment Situation Report. However, only 36,000 jobs were added on a seasonally adjusted basis, significantly below Bernanke’s estimate of 100,000 jobs needed for stability (and therefore an even bigger gap from Reich’s estimate). This is also below the expectations illustrated by the Bloomberg consensus range of 55,000 to 200,000.

600,000 people are no longer unemployed, but the labor force was unchanged. That’s an interesting gap between the household survey (determines the unemployment rate) and the establishment survey (determines number of jobs added). Typically, changes in the labor force help explain the gap. This report is a bit unusual, though, and offers us a substantial population revision (Table C). This revision does not cause the BLS to change their official data for prior months, but they detail what the impacts would have been. The BLS points out that this policy has the unfortunate side effect of making historical comparisons less useful.

January is a negative employment month. On a seasonally-unadjusted basis 2,898,000 jobs were lost. Of those, 339k were due to a negative month of the Birth/Death model. This means that the BLS estimates that a lot of small businesses shuttered in January that were unaccounted for by the establishment survey.

Table A-15 provides alternative measures of labor utilization. U-6, sometimes referred to as the “actual” unemployment rate, fell from 16.7% to 16.1% on an adjusted basis and grew from 16.6% to 17.3% on an unadjusted basis.

This was a disappointing report, but one that was subject to some pretty substantial swings in some of the source data. The substantial changes in the data cushion the blow a bit, as things like the seasonal adjustments are crucial for interpreting the data, particularly this time of year. For actual working people, January is a very unfriendly month in any year. For those tracking trends, it’s merely mediocre this year.