U.S. Had to Export Jobs for Demographic Reasons

Recent analysis found that the U.S. trade balance deficit for manufactured goods was the equivalent of almost 29 million jobs over the past 19 years.  The analysis referred to these as exported jobs.

What if some of the manufacturing jobs had been retained?  How would U.S. employment have been affected?  Would we have anything like the high unemployment rates experienced over the past two years?

Labor Participation Rate

First let’s look at the labor participation rate over the past 60 years.

Over the last 25-30 years the institution of two wage-earner households became entrenched and the labor participation rate rose above 65% as of 1985.  In 2010 the participation rate fell back below 65% as a result of severe employment dislocations produced by The Great Recession.

The following graph focuses on the years starting with 1992.  Two arbitrary areas are identified by the author, a “peak” participation rate and a “healthy” rate.  These arbitrary definitions may be subject to debate.

These two definitions are used to examine counterfactual participation rates in the following discussion.

Manufacturing Jobs Have a High Multiplier

In the previous article a graph was presented which showed the number of manufacturing jobs “exported” each year starting with 1992.  What was not discussed in the previous article is that jobs have multiplier factors.  If one person receives a wage much of that money is spent and that creates additional jobs.  For each person working in a particular job, other people are employed in support industries.  The total of all these additional jobs is what comprises the jobs multiplier.  The mathematics can be reviewed here.

The multiplier varies for different types of jobs and is determined by collecting experimental data in the actual economy.  Manufacturing jobs have high jobs multipliers.  A table of jobs multipliers is shown in the following table:

A number of data sources have been reviewed for the value of the manufacturing job multiplier in the U.S.  These are shown in the following table.

Active links from the table:

Economic Policy Institute
Content First
Washington State.gov

A jobs multiplier of 3.0 means that each job represents a total of three jobs:  the original job plus two others that exist only because that original job is there.  One estimate for all employment is an average multiplier of 1.9.  For manufacturing the multiplier is close to 3.  For every manufacturing job gained or lost directly, two other jobs, on average, are gained or lost.

For the rest of this article we will use the multiplier of 2.9.  Using that value, the number of jobs lost each year since 1992 because of the trade deficit for manufactured goods  has been calculated and is shown in the following graph.  The lowest line represents the number of direct manufacturing jobs exported.  The second line is the number of dependent jobs lost and the top line is the total number of jobs, direct and indirect, that were exported.

What if All “Jobs Exported” Had Been Retained?

If all the direct and indirect jobs that have been ‘exported” had been retained, what would the unemployment picture have been forth past 18 years?  This is displayed in the following graph.

There are two shaded areas in the graph.  The blue area encompasses all the commonly accepted values for NAIRU.  The salmon area defines where labor shortages would occur, the lower the percentage the greater the shortage.

 The red line shows the counterfactual case calculated for the “exported jobs” being retained and the labor participation rate staying at the peak (67%).  The green line shows the counterfactual unemployment rate provided the participation rate actually experienced is used.

A reasonable conclusion from this graph is that, to retain the number of jobs estimated, labor would have to be imported or the labor participation rate would have to rise above the all-time peak.  Without some combination of those two factors there would have been labor shortages 14 of the last 16 years (1994-2007)

Note:  NAIRU refers to a Milton Freeman defined term:  Non-Accelerating Inflation Rate of Unemployment

 What if “Jobs Exported” Had Been Reduced by Half?

The next counterfactual situation examined is the case where half of the “exported jobs” were retained each year.  The following graph shows the official unemployment rate average each year (blue), with two counterfactual unemployment rates.

 The red line shows the counterfactual case calculated for ½ of the “exported jobs” being retained and the labor participation rate staying at the peak (67%).  The green line shows the counterfactual unemployment rate provided the participation rate actually experienced is used.

What if 25% of “Jobs Exported” Had Been Retained?

The following graph shows official and counterfactual unemployment rates if 25% of the “exported jobs” had been retained.

 Even if 75% of the “exported jobs” had occurred, the U.S. would still have had labor shortages almost in seven of the past 19 years.  The unemployment rate would have been below 4% for1997-2000  and 2005-2007.  If the labor particpation rate had remained at peak (67%) throughout this time period, there would have been labor shortages only in the last four years of the 2oth century.


There are many assumptions to be questioned in this analysis.  A few are listed here:

  • Would our trade surplus in services have been less (thereby reducing service employment) if there had been no deficit for goods?
  • Would FDI (foreign direct investment) in the U.S. have been less?  See Elliott Morss.  FDI in the U.S. creates jobs.
  • Would the FDI by the U.S. have been reduced to offset any loss of FDI in the U.S.?
  • Is the multiplier for manufacturing jobs accurate for the type of manufacturing that was “exported”?
  • Are the assumptions about peak and healthy participation ratios really appropriate?

I expect there are many more open issues here and I expect that readers will eagerly point them out to me.


The U.S. has been on a consumption binge.  The examination discussed here indicates that the country has been living beyond its means to produce what it consumes.  That doesn’t even consider that we have to import more than half of the energy used.  If the production of even part of the production of goods that have been imported over the lat 14 years had been retained domestically, there would not have been enough labor available to fill the jobs that would have been required.

So the U.S. has been living beyond its means in three ways:

  1. Beyond its means to provide energy used;
  2. Beyond its means to pay; and 
  3. Beyond its means to produce what it consumes.

Further work is in progress to examine:

  • How much more labor force participation would have been needed to prevent labor shortages from occuring; or
  • How much labor would have had to have been imported; or
  • What combinations of the two would have done the trick.

There is one additional area being studied:  How does the undocumented worker influx that has occurred over the past two decades interact with these labor requirement scenarios.

More articles on this topic will be forthcoming in the next few weeks.

Related Articles

Exporting Jobs  by John Lounsbury

USA Trade Deficit Exports 1.3 million Jobs  by Steven Hansen

Non-Farm Business Productivty Improves in 2010 – Is This Bad News?  by Steven Hansen

6 replies on “U.S. Had to Export Jobs for Demographic Reasons”

  1. I am not convinced. Is it necessary, for example, that “downstream” activities from a manufacturing job are gone when the manufacturing job is overseas? Transportation, retail, warehousing, etc., may in fact be greater. Thus the multiplier is gutted.

    If the jobs were retained, there may well have been (would have to be) lower consumption, obviously. It would be interesting to compute the trade deficit in terms of number of “jobs borrowed” in order to purchase the imported product.

    No doubt, had you seen highly paid manufacturing jobs retained, you would see more factory workers than barristas. You would also have seen higher productivity growth, since manufacturing can be more amenable to efficiencies and labor would be shifted to its more productive uses. You would have seen higher inflation to the degree it was not kept down by offshoring to low-wage countries.

    At the bottom, however, it seems to me the statement that we consumed more than we could have produced is the same as saying we borrowed a lot. And we did borrow a lot. The debt bubble is now our millstone, and it remains to be seen whether we can swim the tide with it.

Comments are closed.