Essays on the study of production efficiency
MetadataShow full item record
In this dissertation, we present a test of the Stole-Zwiebel hypothesis that individual employees have bargaining power in negotiating compensation and that firms respond by overemploying labor and by adopting inefficient technologies. Because the hypothesis generates a production function that includes the market wage, rendering problematic the use of standard duality-based estimation procedures to test their predictions of allocative and technical inefficiencies, we develop and implement a two-step procedure to test their predictions: the first step determines whether there are allocative inefficiencies in the use of input factors, and the second determines whether any extant allocative inefficiency follows the pattern implied by the Stole-Zwiebel hypothesis. Secondly, this dissertation proposes a way of estimating a profit system with technical and allocative inefficiencies in order to identify the source of profit inefficiency. To overcome the drawback that the existence of unobserved technical inefficiency interactive in a nonlinear form has hindered panel data estimation of the system, homogeneity in technology is introduced to separate technical inefficiency from the profit frontier. Because introducing homogeneity in technologies with multiple outputs necessitates choosing a numeraire output and a different numeraire yields different result, a rule of thumb based on careful investigation of the data is inevitable. Thirdly, this dissertation addresses two questions regarding the relation between allocative inefficiency, substitutability between inputs, and the choice of a quantity index. The first question is how the degree of substitutability between input factors affects the extent of allocative inefficiency and the second question is how well the degree of allocative inefficiency is captured by the Fisher and Tornqvist indices. We showed that allocative inefficiency increases with input substitutability between input factors and that both the Fisher and the Tornqvist indices increase with allocative inefficiency. However, we conclude that the Fisher index captures increased allocative inefficiency more accurately than does the Tornqvist index.