But in a new 10-page research note on the impact of small business on jobs, Goldman Sachs’ economist Kris Dawsey argues that’s not accurate.
In one paragraph, he explains why size doesn’t matter. Rather, age does.
…recent research indicates the conventional wisdom that small businesses are responsible for all or most net job growth is not correct. For example, Haltiwanger et al. (2010) find that after controlling for firm age, there is no systematic relationship between net job growth and firm size. They find that historically the most important contributor to whether a firm grows or not is its age, rather than its size. The process of young firms (which do tend to start out small) growing into larger firms is the true contributor to job growth. Others such as Hurst and Pugsley (2011) back up this finding, noting that the vast majority of small firms start small and do not grow significantly. Furthermore, small firms are disproportionately concentrated in areas of the economy that tend to have lower productivity growth, including doctors’ offices, small shopkeepers, restaurants, the building trades, etc.
That’s very intuitive. Check out Dawsey’s chart:
Unfortunately, that doesn’t address the fact that small business creation is lagging. The stats are very clear and they are undisputed.
But Dawsey explains:
…the underperformance of small business in the current business cycle can be largely traced back to the unique character of this business cycle relative to past cycles in terms of the composition of job losses. Construction employment declined more sharply relative to its peak―both on an absolute and proportional basis―than employment in any other major industry group. Furthermore, construction employment has regained only a small fraction of the jobs lost during the past recession. Because employment within the construction industry is disproportionately geared toward small firms, and because the share of employment in the construction industry among large firms is the lowest of any industry (Exhibit 3), the idiosyncratic poor performance of the construction sector goes a long way toward explaining the poor performance of small business employment more broadly. In other words, the question is not why has small business employment done so poorly since its peak, but rather why has construction employment done so poorly since its peak―a question that we can answer much more readily in light of the collapse of the housing bubble starting in 2006. Our rough calculations suggest that roughly all of the “missing jobs” in the small business sector (on a net basis) can be accounted for by the decline in employment among small firms in the construction industry.
Here’s Dawsey’s chart:
Despite all of this, Dawsey communicates the house view that job creation will continue.
“Although there are risks posed by the low rate of new firm creation in recent years, our outlook for job growth remains broadly positive,” he wrote. “Nonetheless, under our forecast, the labour market will take a number of years to normalize, with the unemployment rate likely falling back to a more normal level by around 2017.”
Next Friday, we’ll get the August jobs report. Goldman expects payrolls grew by 200,000.