When looking at the weak hiring so far during this recovery, we need to consider some long-term themes, some of which predate the recession. One theme pertains to jobs created by small businesses.
After the dot-com bubble burst, employment growth was rather tepid, once you exclude the effects of the housing bubble. Since the trough after the recession, in December 2002, total employment has expanded by about 7.8 million, or about 6%. However, the housing bubble directly accounted for 16% of those job gains, in industries including construction, credit, securitization, real estate and architecture, according to data from the Bureau of Labor Statistics.
Now consider that the growth in consumer spending during that period was financed in large part by increasing amounts of consumer debt, much of it mortgage-related. Real household income was essentially stagnant (according to data from the Census Bureau), while household debt as a percentage of disposable income increased from about 97% in 2000 to 133% in 2007, according to data from the Federal Reserve. If you add in the increase in the number of jobs in retail, all of that accounts for 24% of jobs created during this period. In other words, of those 7.8 million jobs created, over 1.8 million of them were directly or indirectly related to the housing bubble.
Since the 6% job growth during the 2002 to 2007 period was actually a bit less than the growth of the adult population during this period (6.6%), the unemployment rate fell during the mid-2000s because a smaller percentage of the population was in the labor force. Those who were "not in the labor force" grew by 7.5%, or 5.5 million people, while the participation rate fell in tandem. Without those 1.8 million jobs created by the housing bubble and the fact that more people were not part of the labor force, unemployment during that period would have been much higher than was reported. In fact, jobs directly created by the housing bubble -- not even including the multiplier effects of spending by people employed in these jobs -- boosted employment by roughly 1.3 percentage points.
Thus, the housing bubble really did mask a more fundamental issue in job creation that predates the recession. Without another such boom, what could be the job engine that will drive the economy going forward? The Kauffmann Foundation (which bills itself as "The Foundation of Entrepreneurship") recently published a paper, "Starting Smaller; Staying Smaller: America's Slow Leak in Job Creation" (you can find a .pdf link here) that discusses why job creation may continue to be tepid.
Small Business, Fewer Jobs
Remember that old adage, "Small businesses are the engine of job creation"? Well, that hasn't been quite as true in the past decade as it used to be. Since this trend has been evident for some time and became more pronounced beginning around 2001, we can't blame it on the recession or lack of credit. Without even focusing on small businesses that fail, fewer new businesses were created, and those that were created ended up producing and retaining fewer jobs than in earlier decades.
The Kauffman researchers note that the cohort of new businesses created in 2009, for example, is on track to create 1 million fewer jobs in those businesses' first five to 10 years than historical averages, in part because of "changing industrial dynamics and long-term structural change," along with other factors yet to be explained. Look at how historical trends have changed: In the 1980s, start-ups created 3.5% of total U.S. jobs (not just jobs created, but total U.S. jobs) annually, on average, but in the 2000s, that number was only 2.6% of total U.S. jobs.
Part of this was due to fewer new companies being created or new establishments of existing companies being opened, especially recently. The number of new employer businesses – those that actually have employees, rather than just sole proprietors – plummeted by 27% since 2006. And the number of employees that those new businesses have when first created has fallen. In the 1990s, new businesses opened their doors with 7.5 employees on average. In 2010, that figure was only 4.9 new jobs per establishment.
That means that new businesses created about 4.65 to 7.0 million jobs in the period of 1997 to 2000, depending on whether one uses data from the Bureau of Labor Statistics or the Census Bureau, respectively, but in 2009, the number of new jobs created by start-ups were 2.5 million (BLS) to 4.5 million (Census). These data measure employment at both new companies that have formed and at new establishments by existing businesses, such as might open a new store or office. This approach offers a more expanded focus than just looking entirely at new firms being created.
Those trends have been on a consistent downward trend for a whole decade – well before the recession began – and represent about 2.2 million to 2.5 million fewer jobs created annually by start-ups from the beginning of the decade to the end of the 2000s. (Of course, some of these jobs don't last; about roughly 45% or so of newly formed businesses in 2004 survived after five years, vs. about a 50% survival rate for businesses created in earlier years.)
Furthermore, new businesses opened over the past decade, and which have survived their early years, retained fewer of their existing employees over time than historical norms. Businesses opened in 2001 retained 80% of their staff five years on, but businesses created in 2004 retained only 70% of their employees after five years. That's equivalent to 300,000 fewer jobs, just on the basis of retention, in addition to the fewer number of jobs created when businesses form.
The Kauffman researchers summed up their findings: "The relatively restrained job creation of new businesses in recent years will, therefore, have a substantial negative impact on longer-term employment levels."
While the exact reasons for this longer-term trend remain unclear, this study goes a long way toward explaining why fewer jobs were created during this recovery. Weak confidence means this trend is almost certainly unlikely to reverse soon. The housing boom had masked this trend, in my opinion, and only now are we seeing the full effects of this theme, which has been at work for over a decade. Perhaps this is an issue policymakers should have explored long before now. We sorely need to know why it took a get-rich-quick scheme in the form of a housing bubble to create jobs in the past decade instead of good old-fashioned entrepreneurship.