In this post, Robert Reich makes the point that most of the recent job growth has come from two areas, the health care sector and from employment related to the military. Michael Mandel of BusinessWeek Online said recently in a cover story for the magazine that growth in health care is "propping up the economy." Kevin Drum challenged this assertion as "statistical trickery" to which Michael Mandel responded here. Kevin Drum summarizes the debate and responds with:
The Healthcare Boom Revisited, by Kevin Drum, Political Animal: ....Michael Mandel, who wrote the Business Week cover story... [T]he takeaway from his story was that the American economy is being kept afloat by jobs in the healthcare industry. If you take those jobs away, private sector job growth elsewhere in the economy has been zero for the past five years. I objected that this was "statistical trickery" that could be done during any economic cycle, since if you remove the highest-growth industry you can make any economy look bad. Mandel emails to say that's not right:
This is a very unusual period where employment gains are so highly concentrated. Let's look at the previous business cycle, for example. Employment peaked in June 1990. Five years later, private sector employment had grown by 6.5 million.
The single biggest contributor to that growth was health services...but it only accounted for 25% of the private employment gains from 1990 to 1995.
....To put it a different way, private employment grew at a 1.4% annual rate from June 1990 to June 1995. Take out health services, and the annual growth rate of the rest of the private sector fell a bit, to 1.1%. Not that big a difference
Point taken. In the previous cycle, measured five years from the employment peak, the biggest industry (which was healthcare back then too) had contributed a lot of jobs, but not all the jobs. The non-healthcare economy really does look unusually anemic this time around.
On the other hand, it's also worth looking at a chart Mandel posted elsewhere on his blog. As you can see, it shows very healthy non-healthcare job growth for the past three years. There's no question that ... (measuring from the employment peak for both the 1990 cycle and the 2001 cycle) overall job growth has been exceptionally weak this time around; and there's equally no question that healthcare has been the principal standout. On the other hand, since 2003 non-healthcare industries have accounted for about 80% of all new private sector jobs. I'm not sure this really makes the case that healthcare is the main industry keeping our economy afloat.
I'm not sure I understand the original objection. If you are trying to explain job growth and taking one industry out has a substantial effect on the job growth figures, that's the point isn't it?
It isn't about making the economy look good or bad, it's an attempt to explain the source of the change in employment in recent years. Over this time period there is little or no growth in private non-health employment, the level of employment at the end of the sample is very near the level at the beginning. Thus, according to the evidence presented, all of the job creation is from the health care sector. Stated another way, if the health care sector had experienced negative growth, overall growth would be negative and in that sense health care is "propping up the economy" as Michael Mandel claims.
However, to acknowledge a slight variation on Kevin Drum's point, you could also take out the biggest negative growth industry and, if it has a large impact on the growth figures, talk about how booming the economy would be had that industry not been in decline. I think it's better to identify the largest sources of change - positive or negative - as a starting point in explaining variations in job growth. This is essentially he same type of argument made about core versus overall inflation, core can be a better measure of the underlying trend if there are a few outliers skewing the picture.