| Summary: | A new data set isolates the behaviour of the International Woodworkers of America and the British Columbia wood products industry 1963-79. Two non-nested models of wage and employment determination in a unionized industry are derived, specified, and estimated using the new data set. In one model (monopoly union model) the union chooses the wage unilaterally to maximize its ojective function subject to the industry labour demand function. The industry chooses employment subject to the union wage and an inefficient wage-employment package results. In the other model (cooperative union model) the union and industry bargain about wages and employment and reach, by some unspecified means, an efficient wage-employment package.
The estimated union objective function is increasing in real wages and employment and decreasing in the workers' real alternative wage. At the mean of the data the estimated elasticity of substitution between real wages and employment is 0.7 and the union is indifferent to a 1.5% decrease in employment and a 1% increase in real wages. Popular hypotheses about union behaviour (rent maximization and wage bill maximization) are rejected as are hypotheses that the union is indifferent to the alternative wage and the level of employment.
The estimated production technology shows that labour is substitutable with materials and capital, and materials and capital are complements. Industry cost functions are not concave in input prices, but input demand functions slope down. The estimated elasticity of the demand for labour is less than minus one in the monopoly union model so the union is operating on an elastic portion of the demand for labour function.
The cooperative union model is argued to be the appropriate
model since it predicts an efficient outcome. This preference for
the cooperative union model is supported by the data when the two
models are tested against one another. === Arts, Faculty of === Vancouver School of Economics === Graduate
|
|---|
