A new and noteworthy paper by Claudia Canals from Columbia University seeks to decompose the widening US wage gap: What Explains the Widening Wage Gap? Outsourcing vs. Technology (PDF).
The approach taken is an accounting decomposition: using a two-step methodology based on Haskel and Slaughter (2002) to decompose the change in factor prices or wages into their different sources:
In particular, we derive a way of measuring technological change (first step) and a theory that links wages to technological change (second step). We then link the changes in technology from the first step with the change in wages from the second step.
The results are interesting:
..we quantify that outsourcing accounts for 28 percent of the actual widening in the wage gap in the 1980-1999 period, biased technological change explain 15 percent of the widening, and total biased technological change explain up to 58 percent of it.
The methodology is innovative, overcoming some of the problems with previous approaches. Her paper also goes beyond most in this field by including not only manufacturing but major service industries as well.
Unfortunately the data is quite dated, only going up to 1999. Canals used the BEA's Historic Input-Output Accounts, which are based on the 1987 Standard Industry Classification, and are not consistent with the newer NAICS-based estimates for 1997-2004. Let's hope she does another paper using the newer data, to see whether outsourcing effect has grown since the 1990s.
I am also looking forward to reading her forthcoming paper, Why previous methods of tackling the effect of outsourcing into labor demand have failed?