US job creation stalls in March
Or does it?
The US economy lost a bit of momentum in March, if you gauge it by job creation, falling well short of the needed 200,000-plus net new jobs that are needed just to keep up with population growth, and well below any level necessary to make a dent in the unemployment rate.
According to the Department of Labor's Bureau of Labor Statistics, which issues a monthly report on the "jobs situation", as it calls it, the American economy created a middling 120,000 net new jobs in March. Total non-farm payroll growth was revised upwards a bit for February to 240,000 (instead of the 227,000 reported a month ago) and downwards for January to 275,000 (from 284,000 reported two months ago.)
Because the official labor pool – meaning the estimated number of people actually looking for work – shrank a bit, the unemployment rate dropped one-tenth of a point to 8.2 per cent.
However, as everyone knows full well, the unemployment rate is far from uniform around the country and does not really reflect the number of people who can't find work. If you stop looking for work long enough, you drop out of the statistics – it is generally believed that unemployment is at least twice as bad as stated in official US government reports.
The good news is that governments at the local, state, and federal level have stopped slashing their payrolls, and collectively cut only 1,000 jobs in March. Manufacturers added 37,000 jobs, and those engaged in healthcare services added 26,000 jobs. Financial services firms added 15,000 workers, and those running various professional and business services operations added 31,000 net new employees. After defying gravity thanks in part of the warmer-than-usual weather in the States for the past several months, the retail trade looked down and shed 34,000 workers.
As usual, the BLS doesn't do the proper job and actually count workers by their jobs, rather than the industries their companies are engaged in. So it is not a "jobs report" so much as an "industry employment report".
As El Reg has said countless times since the Great Recession was under way, it would be far more useful for those with and without jobs to know what areas for which get training. And while we're at it, wouldn't it be nice as part of the monthly surveys of 160,000 employers that the jobs report is based on if Uncle Sam asked them in what areas they were looking to hire and can't find qualified candidates.
Big data is not just about taking whatever you have and dicing and slicing it. It is about gathering the right information in the first place.
And moreover, rather than doing surveys of employers, the US government should be getting unemployment data from individual states, which administer unemployment insurance benefits, as well as actual payroll data and running it through a giant wonking parallel cluster and crunching the actual employment and unemployment data each month.
Big data is actually about gathering up the, er, big data. The household and establishment surveys should continue and be used as a control.
Don't even get me started on the black magic of the seasonal adjustment of the data that the BLS does. If you look at the raw, unadjusted data, the US added 811,000 net new jobs outside of the farming sector in March, added 902,000 workers in February, and added 236,000 workers in January. If the scale of those numbers against the adjusted figures quoted by bureaucrats doesn't raise eyebrows, I don't know what will.
Give me the raw data every time and then show me why – and how – you are adjusting the data. Economies don't grade on a curve. But the Obama administration, like the Bush administration before it, is graded on a curve.
Anyway, in lieu of actual job counts, we have to make do with the industry-sliced "establishment data" the BLS provides, and drill down into it a bit to get a sense of how the IT sector is doing. As usual, El Reg is using the raw data here.
Within the manufacturing sector, the computer and electronic product makers who participate in the BLS survey every month said they added 2,000 workers in March, to just over 1.1 million workers. This number does not change very month over the years. Within this group of companies, those making computer equipment added a mere 400 employees and had 163,200 people on the payrolls, while those making communications equipment held steady at 110,300 workers. Semiconductor and electronics component makers added a mere 100 jobs, to 387,200 workers.
In the so-called "information" sector – which includes broadcasting, telecom, publishing in its many forms, and data processing – employment fell by 15,000 to 2.63 million workers. Telecommunications companies cut 5,500 people from the payrolls (with 838,100 workers in total), and those in the data processing and hosting rackets added a mere 400 workers, to 241,000.
Within the professional and business services sector, companies engaged in computer systems design cut 1,600 workers and employ 1.57 million workers. Management and technical consulting firms, which are often involved in IT projects (but certainly not exclusively) added 4,900 workers and employ 1.11 million people. ®
Sponsored: Hyper-scale data management