For September, the survey based on business payroll reported an increase of 114,000 jobs. The increase based on household surveys was up by a startling 873,000. From last September, payroll surveys show an increase of 1.8 million jobs while the household surveys show an increase of 2.9 million jobs. Those are huge discrepancies.
Meanwhile, the New York Times reported on how startups now require fewer employees in the article "When Job Creation Engines Stop at Just One," by Catherine Rampell.
For more than a decade, start-ups have been getting leaner and meaner. In 1999, the typical new business had 7.7 employees; its counterpart in 2011 had 4.7, according to an analysis of Labor Department data by E. J. Reedy at the Kauffman Foundation, a research organization focused on entrepreneurship.
Rampell goes on to report how entrepreneurs are more reliant on contractors than ever before, contracting employees for specific tasks or projects rather than hiring them and then assigning tasks. As needs change through the life of the company, it is easier for entrepreneurs to assign more or less work to certain kinds of contractors than it is to lay-off and hire different kinds of employees.
This aligns with what I've seen in my 15 years of working with project teams inside of Fortune 100 companies and startups alike. When I began, a typical project might outsource about 10% of its tasks to contractors or vendors who weren't actually employees of the firm. Today, it is not unusual to encounter projects in which 40% or more of the work is done by people outside the firm. An outside vendor might manage clinical trials for a pharmaceutical company or create custom parts for the development of a new device. Based on my experience, I would believe claims that outsourcing on product development has tripled from about 10% to 30% in the last decade. It could even be more.
When a job is outsourced from inside the company to outside, a few things can happen. One, that job can simply be reported on the payroll of a new American company. Two, it can now be reported on the payroll of a foreign company. Three, it can now be reported on the payroll of a less established company, perhaps even assigned directly to an independent contractor who doesn't have a payroll..
If the job goes to a new American company, it will be captured in both the household surveys and the payroll surveys of businesses. If the job goes to a foreign company, it doesn't appear in either survey and unemployment goes up. If the job goes to an independent contractor or just-established company, it might show up in the household survey but not in the payroll survey. As more jobs are outsourced, unemployment would go up and the gap between jobs reported on payroll surveys and household surveys would grow.
It seems clear that the modern economy is becoming more entrepreneurial. One way that shows up is that the ratio of business opportunities to employment opportunities is going up. It may be easier than ever to find projects or tasks even as it is more difficult than ever to find traditional employment.
This explanation of the gap between households and payroll surveys is just a theory, but it seems to accord with my experience with large, established companies and startups. It seems to accord with what Catherine Rampell is reporting as well. Further, if you add to these situations the many internet based business that might be run out of homes and off of platforms like eBay, as mentioned by Matthew Yglesias, it might just fully account for the fact that households have more jobs to report than do the traditional companies that government survey takers may contact.
If there is truth behind this, there are a lot of policy implications, most beyond the scope of this post. It's worth mentioning just a couple. Universal healthcare that lowers risk for contractors and entrepreneurs alike makes more sense than health insurance provided by an employer, for one thing. Even more importantly, we need to change our education and job training programs to prepare people to become entrepreneurs, contractors, and independent agents rather as well as employees.
It might be that the techniques perfected to monitor work in an economy of traditional employment just miss too much of what is going on in an entrepreneurial rather than an information economy. And it might just be that the economy is doing better than we thought, even if it less often promises the apparent stability of a "real job."
Our models, from schooling to labor market reporting, should be updated. The way we depict and prepare for the economy would ideally be as dynamic and responsive as the economy. For now, that doesn't seem to be the case.