Job Growth Stalls In March, IT A Mixed Bag
April 9, 2012 Timothy Prickett Morgan
Is the economy doing better or worse? Is the monthly jobs report that comes out of the Bureau of Labor Statistics any kind of real indicator, or is it massaged so much with seasonal adjustments and other stat tricks that it is more like an impressionist painting of the economy than a picture?
I don’t know, but I sometimes have my doubts about what is going on with the unemployment rate and job counts that the Department of Labor puts out after it surveys 160,000 businesses and 60,000 homes each month to find out who is employing and who is working. But it is the only data we have to work with. If we were really serious about Big Data, as everyone is yammering about all the time, the Federal government would take all of the monthly payroll data it has from the tax bills and actually count the number of people who got paid, how much, and what jobs they were doing. Instead, Uncle Sam polls a sample of businesses and organizes then by industry to figure out what the country’s non-farm payroll looks like and sounds some poly-razz-ma-tazz on the data to seasonally adjust it. It then calls households and asks them if they are working and uses this data to calculate the unemployment rate.
I don’t like seeing cooked data without also being able to see the raw data. I suspect, as techies, you don’t either. So before I even get into the March jobs report, I want to show you some charts I built using BLS data from the past decade. I used line charts instead of area charts to show the number of net new jobs created or destroyed each month for the past decade, and I know that is technically not correct, but it is too hard to visualize as an area chart. In any event, here is what the actual pulse of employment looks like, if you want to think of net job creation or destruction is the pulse:
I have been watching the monthly report with great interest since the Great Recession started in December 2007 and have been picking at it to try to come up with a proxy for the IT industry by pulling out the stats for the computer and semiconductor makers, the data center operators, the telcos, and the systems designers and management consultants who all have some part in the IT racket. The only problem with this is that it ignores the actual IT departments that exist in every industry, so I think it is a poor proxy.
But that is not what I want to talk about right this second. I have never seen the raw data published in such a manner, and what it shows you is what you already know: a slew of jobs are killed off every summer, and something like three slews is killed off every winter around the holidays. Those are the icicles hanging down in the chart below the X axis. The humps above the X axis are when we are creating jobs in the in-between periods, and in a good year, this humps have more area under them than the icicles hanging down do, and we get net new jobs for the year. If you do the math over the past decade, the non-farm workforce has actually grown by 3.4 million jobs to 132 million workers. The only problem is that we needed something on the order of 24 million additional jobs, more or less, to keep up with population growth.
Here’s what the data looks like if you shift to the seasonally adjusted monthly job creation or destruction numbers from the BLS:
This chart, among other things, takes into account seasonal retail and construction jobs, breaking off the icicles in essence. Judge for yourself if this makes any sense, but doesn’t that seasonally adjusted data make the Great Recession just look awful? And don’t numbers like that just freak people out and cause them to do things like stop spending and stop hiring? But be careful about letting Excel pick the mix and max on your axes.
If you make the adjusted data display on the same scales as the raw data I just showed you above, the Great Recession just doesn’t look that bad. Take a look:
You’ll also notice that we have not filled in that gap and I would go so far as to predict that we never will. (Sorry. Those are people that we will all be either taking care of individually as family members or through tax-sponsored government programs.) And the other thing to remember is that a tiny little swing in those curves affects millions of people’s lives directly and tens of millions more indirectly.
For fun, and to try to get some sort of IT proxy, I plotted the actual raw job counts for the entire American economy (on the right scale, in red) against the employment for the computer systems design and related services sub-sector of the economy. Take a gander:
During the last baby recession, which ran from 2001 through 2004 as far as I can tell, those designing computer systems actually shed jobs more intensely than the economy at large. But somewhere around 2004, it shifted and this sub-sector started adding jobs at twice the pace as the economy at large. And when the Great Recession hit and the economy shed on the order of 14 million jobs, computer system design firms more or less held steady after taking an initial hit and then resumed the climb while the overall economy has saw-toothed its way on a slightly upward trajectory.
Pick your profession carefully.
Now, on to the March jobs report, which you can see here. Using the adjusted figures, the U.S. economy added 121,000 net new private sector jobs and shed 1,000 net jobs in the local, state, and Federal government. The BLS revised its job counts for January, saying the economy added 275,000 workers, not the 284,000 it previously thought, and ditto for February, which is now pegged at 240,000 new jobs, up a bit from the 227,000 originally reported. Based on the household survey, the unemployment rated ticked down one-tenth of a point to 8.2 per cent, and that was because the official labor pool–meaning the number of people looking for work–shrank as some people stopped looking for work (in some cases, stopped again).
Using the raw data, within the professional services sector, companies engaged in computer systems design increased their payrolls cut 1,600 people and now have 1.57 million people on the payrolls. Those in the management and technical services sub-sector, which often has IT aspects, added 4,900 people to the paycheck troughs, now totaling 1.11 million people across the 50 states. Telecom companies employed 838,100 workers at the end of March, down 5,500 compared to February, and data processing and hosting firms added 400 workers and employ 241,000 people. Data centers may get lots of headlined, but you can’t really expect for giant banks of computers that essentially replacing the filing and sorting and processing and calculating functions of human beings to in themselves generate a lot of jobs. (Well, you can expect it if you want, but that’s silly.) Computer and communication system makers employed 1.1 million people as March ended, up 2,000 people since the prior month.