Content provided in partnership with
Business Services Industry
Trade affected workers in the service sector: 1987 and 1990
American Journal of Economics and Sociology, The, April, 1995 by Bartholomew Armah
I
Introduction
At the urging of the United States government, services have become prominent on the agenda of negotiators of multilateral trade liberalization. The primary example of their province was furnished by the Uruguay Round multilateral trade negotiations. Underlying the U.S. initiatives is the presumption that US comparative advantage has shifted to service transactions (Bhagwati, 1987). This shift raises an obvious question regarding the distribution of trade-related gains and losses among various groups in the country. An analysis of the gains and losses is of paramount importance since policymakers have rarely been confronted with the possibility of trade-related adjustment pressures in services. To date, however, few empirical studies(1) have examined the domestic impact of liberalized service trade.
Using input output analysis, this study identifies the characteristics of workers most likely to bear the brunt of trade-related employment displacements in the service industry.(2) The demographic profile of three categories of service industries are examined: a) industries which experienced positive net trade-related employment in 1987 and 1990 with a positive trend, b) industries which generated positive net trade-related employment in 1987 and 1990 with a negative trend, c) industries which suffered net trade-related employment losses in 1987 and 1990, albeit at a declining rate.(3)
Table 1
A. The Model. Starting from the balance equation in the input-output system that states that for any sector, domestic output is the sum of intermediate goods demand AX, and final demand F, minus imports:
X = AX + F - M [1]
where:
X is an n x 1 column vector of final output,
A is an n x n square matrix of input-output coefficients
F is an n x 1 column vector of final demand including exports (Ex) and consumption (C).
M is imports
To determine the effects of changes in the levels of final demand on domestic inputs, a domestic requirements coefficients matrix was created by adjusting the 228-sector input-output table for imports:
Ad = (I - m) * A [2]
where, Ad is the matrix of domestic coefficients, and m is a diagonal matrix with each element defined as the import share of a commodity in the domestic market i.e. the ratio of total imports of a commodity to the sum of total intermediate use of that commodity plus domestic final demand for that commodity. Substituting Ad for A in equation [1]:
X = AdX + F [3]
The total (direct and indirect) domestic inputs required to satisfy a unit increase in final demand can be computed by pre-multiplying the vector of final demands by the inverse of the direct domestic requirements matrix:
X = [(1 - Ad).sup.-1]*F [4]
B. Employment effects of Increased Exports and Imports. For each year, the change in exports ([Delta]Ex) and imports ([Delta]M) per sector is estimated by respectively multiplying the vector of weighted sectoral export ([Ex.sub.1]) and import ([M.sub.1]) shares by the corresponding vector of sectoral changes in export (EXP) and import (IMP) levels between 1987 and 1990.
[Delta]Ex = [Ex.sub.1]*EXP [5]
[Delta]M = [M.sub.1]*. IMP [6]
where:
[Ex.sub.1] is a n by n diagonal matrix of weighted sectoral export shares
[M.sub.1] is a n by n diagonal matrix of weighted sectoral import shares
EXP is a n by n diagonal matrix of sectoral changes in export levels between 1987 and 1990
IMP is a n by n diagonal matrix of sectoral changes in import levels between 1987 and 1990
Net trade-related employment ([J.sub.n]) for each reference year is determined by pre-multiplying the inverted square matrix in [4] by the labor output coefficient matrix L, - whose (1,n) element shows the amount of wage and salaried labor required to produce a unit of output in sector 'n' - and post-multiplying the resulting row vector by the change in net exports ([Delta]Ex - [Delta]M):
[J.sub.n] = L * [(I - Ad).sup.-1] * ([Delta]Ex - [Delta]M)[ 7]
This yields the total number of potential jobs that can be created directly and indirectly as a result of a change in net exports.
C. Limitations and Assumptions of the Model. Although the model takes into account the direct and indirect employment effects of changes in net employment, the impact of trade on real income through changes in prices and wages is not captured. The estimates provided above indicate the increase or decrease in demand for workers resulting from a change in imports and exports. They do not reveal the final change in equilibrium employment which are determined by labor supply and other factors affecting labor demands. Since it is conceivable that there may be an insufficient number of workers with the skills requisite to fill the positions created by trade, the estimates should be interpreted as "employment opportunities" rather than employment per se.
Input-output analysis also assumes a given technology with no substitution possibilities, hence, changes in exports and imports are assumed to have a proportionate impact on employment opportunities. The latter assumption may not hold if the periods of analysis are characterized by extremely large price and wage fluctuations. This qualification is important since market reactions through changes in wages and prices are not incorporated in the model. Nonetheless, the use of input-output tables for short to medium term analysis is justified on the grounds that technical coefficients tend to be stable over long periods (Blair and Wyckoff, 1989).
