14 May 2012

Reducing Unemployment: Manufacturing vs. Services

The graph above answers a question I've wondered about for a bit:
By how much would the services and manufacturing sectors each have to grow employment in order to reduce the current rate of unemployment by 1%?
The answer is 1.3% growth in employment in the services sector and 12.9% in the manufacturing sector. (Data is for April, 2012, and can be found here for employment and unemployment, and here for employment in the service sand manufacturing sectors).

This graph explains that while the current uptick in manufacturing employment is worth noting and welcoming, growth in manufacturing employment is not going to be the primary long-term solution to bringing down unemployment. That said, even though manufacturing and services are distinct categories of economic accounts, they are of course inter-related within the broader economy. However, claims that special treatment for manufacturing will reduce unemployment have a high hill to climb in terms of the simple math that follows from the small portion of the economy that manufacturing employment currently comprises. Productivity gains make that hill even steeper.

And of course, do not forget that from the perspective of employment rather than economic sectors, all jobs are service jobs.


  1. "And of course, do not forget that from the perspective of employment rather than economic sectors, all jobs are service jobs. "

    I don't understand the point of this statement, which you've taken to repeating over and over recently. It certainly is true that if you redefine manufacturing jobs as service jobs, then they become service jobs. And if I redefined my cat as a dog, it would have puppies.

  2. The BLS stats show a wide variance by educational level http://www.bls.gov/news.release/empsit.t04.htm
    Low skilled/educ. jobs are needed to push the unemployment rate down given these numbers. I'll bet that the better paying jobs for this group is in traditional manufacturing.

  3. Dr. Pielke:

    My understanding is that primary industry jobs (mining, manufacturing etc.) have a multiplier. I've heard values from 2 to 5 additional jobs for every primary job. Is this included?

  4. More folks in the US are employed in the services sector than what’s defined as the manufacturing sector. Ergo, policies aimed solely at boosting manufacturing jobs will produce a lower decrease in unemployment numbers than would policies aimed at boosting jobs overall. That’s math simple enough that even I can understand it, but, as may be your objective, many policy-makers may not. Given the low workforces participation numbers we now experience, our larger problem is attracting more folks to notion of employment by expanding the number of good-paying jobs available regardless of how we define the activity in which the job-provider is classified.

    As a right-wing retard with libertarian tendencies, may I suggest that our federal tax code -- that thingy through which the federal government collects funds and provides incentives -- is much too complex to do so fairly or efficiently. Factor in tariffs and other trade-related factors, and one starts to appreciate the mess associated with doing business globally by a US-based enterprise.

    Relying on the distant past of my entrepreneurial endeavors, I note that all businesses require capital and labor, whether they are cottage industries producing tweed or industrial workhorses producing aircraft. Why not set depreciation rules that benefit both extremes, whether it’s Harris tweed or a Dreamliner? Immediate expensing versus depreciations should be driven by how the business finances its capital, not how Washington might like to collect it: the amount collected is the same, it’s the schedule over which deductions are taken that’s at issue.