improvement in the educational skill of its workforce. Furthermore, the finance industry, a large
component of the services sector, suffered the largest disruptions in the aftermath of the 1997-98
crisis. As a result of this multitude of factors, TFP growth in services turned highly negative
after 1997. The longer-term prospective, however, may be more accurately measured by a focus
on the 1977-1996 pattern of a small positive growth rate, 0.5 percent per year. Uncertainty about
the outlook for productivity growth in services is the motivation for a more detailed examination
of that sector in the following section.
Finally, it is striking that the estimated rate of improvement in TFP for the aggregate
economy, 1.6 percent annually for the 1977-96 period in table 9, exceeds that of any individual
sector. In fact, the average annual rate of increase in value added-weighted TFP growth is only
0.5 percent for 1977-96 and it is zero for 1977-2004. The difference, 1.1 percent annually, is a
measure of the gains from resource reallocation. That is, the largest portion of the improvement
in overall TFP is the result of moving workers from the low-productivity agricultural sector to
industry and services.
A Disaggregate View of Services. The service sector has had the lowest rate of long-term
growth in both labor productivity and TFP, and it was the most severely affected by the financial
crisis. Labor productivity fell sharply after 1996, and the 2004 value is still 20 percent below the
1996 peak. The decline in TFP is an even-larger 25 percent. However, the sector includes a
very diverse set of industries, whose output is notoriously hard to measure, and it includes the
banking industry which was severely impacted by the 1997-98 financial crisis. The national
accounts division of NESDB supplied a more disaggregate breakdown of the sector’s value
added and capital stock that enabled a more detailed examination. The data are available for a
period of 1980-2003, but use the pre-2001 industrial classification, Revision 2 of ISIC.
Comparable employment data were obtained from the LFS. The identified industries include: (1)
transportation and communications; (2) wholesale and retail trade, (3) banking, insurance and
real estate, (4) ownership of dwellings, (5) public administration, and (6) other services.
The data for the first two industries should be of reasonable quality, and banking is of
particular interest. Ownership if dwellings in an unusual industry in which all of the valued
added is imputed capital income; and it is separately identified largely for the purpose of
excluding it from the productivity calculations. Public administration also raises measurement