Offshoring and offshore outsourcing may not be important sources of U.S. job loss

Is the U.S. losing a lot of jobs to offshore outsourcing?  Jacob Funk Kirkegaard has looked at the available data, and suspects not: Offshoring, Outsourcing, and Production Relocation - Labor-Market Effects in the OECD Countries and Developing Asia (Peterson Institute working paper, April 2007).

The bottom line:

...Yet, with the broadest and most valid available empirical evidence largely free of se­lection bias showing that just 1 in 25 US separations in mass layoffs can be attributed to either offshoring and offshore outsourcing, one cannot escape the conclusion that the heated public and political debate on the issue has been vastly overblown at least in the United States and that the direct employment effects are very limited. This is especially the case in the US economy as a whole, where very large numbers of jobs are constantly created and destroyed. The latest available data for 2005Q3show that more than 8 million jobs were created, while 7.4 million were lost.

Here are some more details:

Kirkegaard is interested in job loss from production relocation:

The focus here is "production relocation" - i.e., instances of (1) total or partial closure of an enter­prise's existing production units in one country with accompanying work force reductions, and (2) either the opening of affiliates abroad for the production of the same goods or services (offshoring) or the forg­ing of a subcontracting contract with a nonaffiliated firm for production of the same goods or services in their home country (outsourcing) and/or abroad (offshore outsourcing).

The data are terrible, with unattractive consequences:

Unfortu­nately, the near complete absence of valid empirical evidence on the phenomenon has allowed entrepre­neurs and consultants to frame the debate to promote their own interests, thus fueling public anxiety.

The best U.S. data - dating from 2004 - comes from the U.S. Bureau of Labor Statistics "Mass Layoff Statistics (MLS) Program:

...The MLS program collects from administrative unemploy­ment records and employer interviews data regarding the reasons for production establishment closings in the United States. As in most media monitoring estimates, the threshold for inclusion is layoffs of 50 or more employees. BLS data can therefore be expected to suffer from a similar downward bias with regard to the actual number of layoffs.

Since 2004, the BLS has published data on the number of jobs lost in mass layoffs associated with both domestic and overseas relocations of production by US businesses....

The BLS data show that:

Just under a million Americans lost their jobs as a result of unscheduled (i.e., excluding seasonal work and vacation periods) mass layoffs, and approximately 12 percent of these layoffs can be attributed to movement of work - i.e., domestic in-house relocation, outsourcing, offshoring, and offshore out­sourcing. This 12 percent number is relatively large, although not as big a reason for mass layoffs as, for instance, "contract completion" (~25 percent) or downsizing (~16 percent) and roughly on par with layoffs due to "bankruptcy" and "financial difficulty" combined.

It is possible, based on available data, to calculate the relative importance of the four types of "move­ment of work." However, given that detailed information regarding only 72 percent of the total num­ber of relevant separations is available..., estimating a precise "hard number of jobs lost" to either category of "movement of work"... is not appropriate. Instead, the far right column of table 1 estimates the share of total separations associated with movement of work attributable to each category.

...more than half (55 percent) of all separations associated with "movement of work"  is of the domestic in-house type and therefore have neither ownership nor international locational effects. They are therefore outside the scope of this working paper.

...the apparent employment impact of offshoring as a result of domestic outsourc­ing only is very limited: only 1.4 percent of all separations. This number seems very low and possibly indicates that in most instances, domestic outsourcing has already occurred or is carried out without any large-scale loss of jobs....  the combined employment effects of offshoring and offshore outsourcing represent just 4 percent of all separations from mass layoffs in the United States in 2004-05, with in-house offshoring representing three-quarters.

Caveat - the data is static and doesn't get at potential dynamic effects:

However, it is important to note that this conclusion, reached on the basis of even the US BLS data - which is at the top of the data validity ladder - is a "snapshot," i.e., merely states what the status is at a given point in time. It does not consider any dynamic employment effects arising out of even this limited level of offshoring and offshore outsourcing. The almost inevitable "technology spillovers" from offshored or offshore outsourced production to foreign domestic companies may lead to increased com­petitive pressure in the US (or other rich-country) home market as well. Such dynamic spillover effects may lead to adverse employment effects at competing home-country firms, if they as a result of increased foreign competition are forced to downsize or even go into bankruptcy. Due to the still very limited dura­tion of time-series, none of the empirical data on offshore outsourcing presented in this working paper can reasonably be utilized to estimate these effects.

The employment impacts of production relocation on the U.S. is only one of the Kirkegaard's topics:

Section I provides the necessary terminology and presents the most important empirical findings to date in the United States, EU-15, and Japan. Section II surveys different estimates for employment potentially affected by offshoring and offshore outsourcing, while presenting estimates from detailed US information technology (IT) occupations indicating very extensive skill bias. The potential impact of automation on the service sectors is also evaluated. Section III considers the impact on developing Asia of the findings in sections I and II and presents data from India pointing to the domestic origin of most employment growth. Section IV concludes the working paper.