the report on employment situation for September, delayed two and a half weeks due to the shutdown, was finally released by the Bureau of Labor Statistics on Tuesday, and as many have noted, it wasn't worth the wait...showing weak job creation, barely enough to keep up with the increase in the working age population, the survey data for this summary was gathered during the week of the 9th of September, well before the prospect of a shutdown and debt ceiling crisis was even mentioned it the media, so it's unlikely this recent government dysfunction influenced employment decisions that early in the month....for the employment impact of the shutdown, we'll have to wait for the October report, which will not only be distorted by the late surveys, but will be also complicated by a method that classifies the furloughed workers as employed in the establishment survey, but counts them as “unemployed, on temporary layoff” in the household survey...
the establishment survey for September indicated payroll employment increased by a seasonally adjusted 148,000 to 136,290,000, while the change in payroll employment for August was revised from a net gain of 169,000 jobs to an increase of 193,000 and July's job creation was revised downward from a gain of 104,000 jobs to just 89,000...the unadjusted data for September showed an increase of 612,000 jobs, which was due in part to the 882,600 seasonal increase in employment at local school districts, which was adjusted down to a net payroll jobs gain of just 9,500, and the 468,000 seasonal decrease in jobs in leisure and hospitality, as most temporary summer jobs wound down...our adjacent FRED bar graph shows the seasonally adjusted change in payroll jobs monthly since the beginning of 2008; you might note that payroll jobs gains over the 3rd quarter of this year are the weakest since the 2nd quarter of 2012..
although 148,000 additional jobs was weak compared to most months of the recovery, at least for once they weren't mostly just retail and fast food jobs; the professional and business services category saw an increase of 32,000 payroll jobs, of which 20,200 were in temporary help services; a net of 23,400 jobs were added in the transportation and warehousing sector, with 17,900 of those working in ground passenger transportation services, such as bus drivers; and state governments added a seasonally adjusted 22,000 jobs, 19,800 of which were in state level education; another 20,800 net jobs were added in retail trade, with food and beverage stores adding 7,700 and clothing stores shedding 7,200 employees, while wholesale trade added 16,100 jobs, with 8,700 of those involved in wholesale non-durable goods...in addition, 20,000 jobs were added in construction, of which 10,900 were created by specialty trade contractors, with 6,800 of those non-residential...the only significant seasonally adjusted job losses were in the leisure and hospitality category, where the net employment loss was 13,000 more than a normal September; 7,100 of those were in food services, reversing the trend we've seen in recent months wherein most of the net job gains were in food service or retail...
our FRED bar graph below shows some of these seasonally adjusted trends over the last 13 months; in each monthly grouping of 8 bars, the monthly change in manufacturing employment for that month is indicated in blue, the change in monthly construction employment is in red, the change in retail employment is in dark green, the monthly change in government jobs is in yellow, and the change in employment in professional and business services, which includes everything from accounting services to garbage collection, is in grey; also included are the BLS employment subcategories of jobs in bars and restaurants, the major component of the leisure and hospitality category, in light green, and the two major subcategories of the education and health services category, health care jobs in orange, and jobs in education shown in violet...it's clear from this that a significant portion of the jobs created over the past year are in the two categories shaded green, retail sales and food services, and one other area of major jobs creation - professional and business services in grey - is indicative of mostly clerical and temporary help, while creation of higher paying jobs in manufacturing - blue - and education - violet - have lagged...(a net of just 100 new education jobs were created in September, hence the violet bar is invisible in the last group below)
one additional graphic we’ll add here to put it all in perspective comes from Robert Oak of the Economic Populist; the pie graph below shows the percentage of payroll jobs by industry as of September, making it clear where the largest proportion of jobs in our current economy are at...in his graphic coverage of this report, Robert pairs this graph with a similar pie graph from 2008, in an effort to show that the wedges for the decent paying jobs in manufacturing and construction have both shrunk by 1.1% of the total each since the recession started...our purpose here is just to show the percentages of each by way of general overview... a clarifying footnote we should add is that the large green government wedge - 17% of all payroll jobs - includes jobs in education at the state and local levels, while the CES category of education and health services is mostly jobs in health care; just one-sixth of that are those in educational services, which includes jobs at private colleges, occupational training services and the like, while nearly half of the government jobs are in education at the state and local levels.. 
the establishment survey also details seasonally adjusted data on average weekly hours by industry sector and average hourly and weekly earnings by industry sector...in September, the average workweek for all private payroll employees remained at 34.5 hours, the same level as it was at in August and In September a year ago; the average workweek for nonsupervisory employees was also unchanged from August and a year ago at 33.7 hours, while the manufacturing the workweek was at 40.8 hours, the same as August, but up .2 hours from the year ago 40.6 hours...the average hourly earnings on all private nonfarm payrolls rose 3 cents to $24.09, after rising 5 cents last month...for production and nonsupervisory employees, the average hourly increase was 4 cents to $20.24, the same increase as in August....
next, in looking at the monthly employment data that's extrapolated from the household survey, we'd do well to recall these national numbers are derived from a survey of just 60,000 households, and they have a margin of error of +/- 300,000 in the monthly change in the number unemployed and an 90% accuracy range of +/- 0.2 percentage points in the unemployment rate; in September, the seasonally adjusted extrapolation from this survey indicated that 144,303,000 of us were employed, 133,000 more than in August, and 11,255,000 of us were unemployed, 61,000 less that August, giving us a civilian labor force of 155,559,000 and thus resulting in the widely reported September unemployment rate of 7.2%, down 0.042% from the 7.3% unemployment rate in August....in addition, the number of us "not in the labor force" and hence not counted in this survey's calculation rose by 136,000 to 90,609,000, a new record, as the civilian non-institutionalized population rose by 209,000...
there was no improvement in the two metrics computed from this survey that we watch; our adjacent FRED graph shows that the labor force participation rate, which is the percentage of us working or actively looking for work, remains at its lowest level since the 70s (when women didn't need to work) at 63.2%, while the employment population ratio, or the equivalent of an employment rate (graphed in blue), remains at 58.6%, stuck roughly at the same level that it's tracked over the past 4 years, as job creation has barely kept up with increases of the population…
recognizing the larger margin of error in the smaller demographic samplings, the employment summary reports that the unemployment rates for adult men at 7.1 percent, adult women at 6.2 percent, teenagers at 21.4 percent, whites at 6.3 percent, blacks at 12.9 percent, and Hispanics at 9.0 percent showed little or no change in September; however, with the sample size available to them, they do show a decline of 1.3% in the teenage unemployment rate from the 22.7% teenage unemployment reported in August; that breaks out to an unemployment rate of 19.3% for white teenagers and an unemployment rate of 35.1% for black teenagers....the seasonally adjusted unemployment rate also fell by a full percent for those without a high school diploma, from 11.3% in August to 10.3% in September, while it rose from 3.5% to 3.7% with a bachelor's degree or more education...
the number of us employed part time "for economic reasons" in September rose by 15,000 to 7,926,000; these are those who reported that either their workplace cut their hours or they could only find part time work; nonetheless, U-6, the broadest measure of labor underutilization, still fell from 13.7% in August to 13.6% in September...those who were working part time who didn't indicate that they wanted full time work fell by a seasonally adjusted 372,000 to 18,967,000; oddly, the unadjusted data shows the opposite; that those who are working part time voluntarily rose by 1,472,000 to 18,848,000...of course, some of those part time workers may be holding two jobs; the number of multiple jobholders was up by 56,000 in September to 3,393,000; note that the establishment survey does not distinguish between full and part time jobs and includes no data on the individuals holding the reported payroll jobs...
the number of us officially unemployed more than 27 weeks fell by 144,000 in September to 4,146,000; this doesn't necessarily mean those who are no longer in this metric found work; they may have simply not looked for work during the reference period and hence weren't counted...those of us unemployed 5 weeks or less, ie, the newly unemployed, rose by 33,000 from August's level to 2,596,000...among those of us not officially in the labor force and hence not counted in any of these metrics, an additional 5,775,000 reported that they still want a job; of those, 2,302,000 are categorized as "marginally attached to the labor force" because they've looked for work sometime during the last year, but not during the 30 day period covered by the September household survey...852,000 of those are further characterized as "discouraged workers, because they say that they havent looked for work because they believe there are no jobs available to them...the number of "discouraged workers" has increased by 50,000 from September a year ago...
Inconsistent Adjustments and Reports
lastly, we have to again point out that there still is a major incongruity in the seasonal adjustments between the two surveys; you may recall that last month we noted that the seasonal adjustments subtracted more than 200,000 jobs from the August establishment survey but added nearly 500,000 to the August household survey; although that was disturbing, we were advised that it would likely be at least partially reversed with this release; that didn't happen; instead, the seasonal adjustment continued to subtract hundreds from the the number of payroll jobs reported, while it only subtracted a trivial 9,000 jobs from the household survey count of those employed; to recap, taking the latest revisions into account: last month the unadjusted payroll data showed a gain of 411,000 jobs in August; the seasonal adjustment lowered that to 193,000 jobs gained, while this month's unadjusted establishment survey data showed an increase of 612,000 jobs, which was lowered by the seasonal adjustment to the headline 148,000 job gain...conversely, the household survey showed a 604,000 drop in the count of the employed in August, which was raised to a loss of just 115,000 in the employed, while in this September household survey we've just looked at, unadjusted employment rose 142,000 to 144,651,000, and the seasonally adjustment lowered that to an increase of 133,000 employed...so, over the course of the past two months, the seasonal adjustment subtracted 682,000 from the payroll jobs recorded by the establishment survey, while the adjustment on the household survey added 480,000 to the count of those employed...while we wouldn't expect the seasonal adjustments in the two surveys to move in lockstep, this much of a divergence just doesn't make sense...
while we’re talking about data that doesn't make sense, we should also point out that the Job Openings and Labor Turnover Summary for August, originally scheduled for release on October 8th, was also released this week; you may recall that this report, which generally lags the broader employment summary by a month, includes data on hiring and total separations for the month from a survey of establishments, in addition to the data on job openings...in August, JOLTS showed 4,488,000 hires, down 9,000 from July, and 4,376,000 total separations, up 103,000 from July; the difference between hires and separations, which includes both layoffs and quits, was 112,000, and that should be equal to the total net job creation that the broader employment summary shows...except that it isnt; as we noted earlier, the change in payroll employment for August was revised from a net gain of 169,000 jobs to an increase of 193,000 with the release of the employment summary...so one or both reports are off by a total of 81,000 jobs...the job category details don't match as well; you might recall that according to the August jobs report, retail trade saw the most job creation with 44,000 net additional jobs; JOLTS shows 672,000 retail hires in August, and 669,000 retail separations, for a net creation of just 3,000 retail jobs....similarly, the August employment summary indicated net job creation of 22,000 in manufacturing; August JOLTS shows 229,000 manufacturing separations and 231,000 manufacturing hires for a net of just 2,000 new manufacturing jobs...hopefully, future revisions to one or both reports will correct this divergence, but right now there's no way of telling which is most correct...
August Trade Deficit Virtually Unchanged from July at Levels Above 2nd Quarter
in addition to the jobs reports, the August report on our International Trade in Goods and Services (pdf), originally scheduled for release on October 8th, was released by the Commerce Dept on Thursday morning...it showed a slight widening of our trade deficit, as our exports fell a bit and our imports were virtually unchanged; total August exports of $189.2 billion, up from $189.1 billion in July and imports of $228.0 billion, essentially the same as July's number, resulted in a goods and services deficit of $38.8 billion, up from the revised trade deficit of $38.6 billion recorded in July...exports of goods fell $0.3 billion to $132.4 billion, while exports of services increased $0.1 billion to $56.8 billion; and imports of goods decreased $0.1 billion to $190.7 billion, while imports of services increased $0.2 billion over July to $37.4 billion....since the trade deficit in July represented a 13.3% increase over June's figures, that August was unchanged from that means we're now two thirds of the way to seeing a significant hit from net trade when 3rd quarter GDP figures are released...our FRED bar graph below shows the monthly change in exports in blue, imports in red, and the balance of trade in brown over the past two years…although our net trade is always a deficit, it is the change over the quarter shown in brown that influences the GDP calculation…in the 2nd quarter of this year, the increase in exports matched the increase in imports, resulting in no change to GDP from our international trade..
major changes in our August goods exports from July's include a $1.3 billion decrease in our exports of industrial supplies and materials, headed up by a $880 million decrease in exports of non-monetary gold down to $2.47 billion, a $527 million decrease in exports of "other" petroleum products, to $4.85 billion, and a $410 million decrease in exports of organic chemicals to $2.83 billion, which was slightly offset by a $283 million increase in exports of fuel oil...August also saw a $0.4 billion decrease in exports of foods, feeds, and beverages; a $187 million decrease in wheat exports and a $147 million decrease in soybean exports accounted for the lions share of that....on the other side of the export ledger, August saw a $0.7 billion increase in exports of automotive vehicles, parts, and engines, a $0.3 billion increase in exports of consumer goods, topped by a $424 million increase in exports of gem diamonds diminished by a $147 million decrease in exports of pharmaceuticals, a $0.2 billion increase in exports of capital goods, topped by a $418 million increase in exports of civilian aircraft, which counted as $5.26 billion of our August exports, and which were offset by decreases of $232 million in telecommunications equipment and $222 million of computer accessories...
the major month over month changes in our imports of goods included a $1.0 billion increase in imports of capital goods, which included a $642 million increase in imports of computers to $5.56 billion, and a $251 million increase in imports of 'other" industrial machines, which was offset in part by a $168 million decrease in imports of semiconductors and a $167 million decrease in imports of engines for civilian aircraft...August also saw an $0.8 billion decrease in imports of consumer goods, which included $667 less pharmaceuticals, $297 million less artwork and collectibles, and $169 million more cotton apparel imports, and a $0.2 billion decrease in industrial supplies and materials, which notably included $594 million less imports of crude oil, and a $233 million increase in imports of automotive vehicles, parts, and engines...and although the major import category of foods, feeds and beverages was virtually unchanged in August, we did see a $159 million increase in imports of fish and shellfish, offset by a $138 million decrease in imports of edible oils and oilseeds...in a special category tracked separately without seasonal adjustment, our advanced technology products exports were $26.7 billion in August while imports were $32.6 billion, resulting in a high-tech goods deficit of $5.9 billion…
our bilateral deficits in August goods trade, which are also not seasonally adjusted, generally showed modest improvement; our goods deficit with China fell to $29.9 billion, from $30.1 billion in July; our goods trade with the EU shrunk to $9.8 billion, from $13.9 billion, and our goods deficit with OPEC fell slightly from $7.4 billion to $7.3 billion....other larger bilateral goods deficits in August were with Japan at $6.4 billion, Germany at $5.4 billion, Mexico at $4.9 billion, Saudi Arabia at $3.6 billion, Canada at $2.3 billion, Ireland at $1.9 billion, Korea at $1.7 billion, India at $1.6 billion, and Venezuela at $1.5 billion; meanwhile, we ran small surpluses in August goods trade with Hong Kong at $3.7 billion, Brazil at $1.7 billion, Australia at $1.4 billion, and Singapore at $1.1 billion...the graph below from Bill McBride tracks our total trade deficit in blue for each month since January 1998, and then breaks that into an oil component in black and an “everything else” component in red…note the graphs are shown as a negative amount from the top "$0” bar, ie, the further down, the worse it is…clearly, roughly half our trade deficit has been in oil since the onset of the recession (the pale blue vertical bar); the price of crude averaged $100.26 a barrel in August, up from $97.07 in July, and up from $94.48 a year ago…
(the above is my weekly commentary that accompanied my sunday morning links emailing, which in turn was mostly selected from my weekly blog post on the global glass onion…if you’d be interested in getting my weekly emailing of selected links that accompanies these commentaries, most coming from the aforementioned GGO posts, contact me…)