Essays on: Application of Cross Sectional Efficiency Analysis

This thesis consists of three different essays, all of which have arisen from a practical problem. In the first essay, “Technical efficiency and ownership: The case of booking centers in the Swedish Taxi Market” the purpose is to examine the question of competition on an equal basis between privately and publicly owned booking centers in the Swedish taxi market. One way to answer this question is to study technical efficiency, and decomposing technical efficiency into managerial and organizational efficiency. In the empirical investigation, no evidence was found of a relation between amount of technical efficiency and ownership type. The result of the Mann-Whitney test indicates that there is no statistically significant difference in technical efficiency between the privately owned and the publicly owned booking centers. In the second essay, “Identification of References: An Explorative Approach” the focus is on implementing the result from a efficiency study in the investigated industry, by providing inefficient units with information on relevant reference units. The aim of the study is to provide guidelines on which properties one can expect of a reference unit, and on methods for detecting reference units for an inefficient unit. Relevant reference units make it possible for inefficient units to, on site, study production that is more efficient than its own. This makes it possible to adopt more efficient ways to organize production. In the study, we argue that to be a relevant reference unit the unit has to be the existing and efficient unit that are most similar, indicated by the Euclidian distance, to the inefficient unit. To identify reference units, three methods are explored: Dominance, non-zero Intensity variables and the Sphere. Strengths and weaknesses of each method are explored in the study. In the third essay, “The Effect of Omitted Variables on Optimal Scale: An Application on Swedish Employment Offices”, we investigate the effects of omitting/adding variables on both the source of scale inefficiency and on the input-output size of scale efficient units. Both a non-parametric and a parametric distance function approach are used to compute measures of scale efficiency and scale elasticity respectively. This information is used to classify units in one of the following groups; scale efficient, scale inefficiency due to increasing returns to scale and, finally, scale inefficiency due to decreasing returns to scale. We then investigate if, as an effect of omitting/adding variables, changes exist between these groups and changes within the group of scale efficient units, with respect to input and output size. Changes between the groups imply that policy recommendations, concerning means to reducing inefficiency, change. Changes within the scale efficient group, with respect to input and output size, imply a change in the optimal scale of production. In the empirical application, we show that omission of variables will influence the source of scale inefficiency. In general, when the two quality indicators are omitted, scale inefficiency is to a larger extent explained by increasing returns to scale. Further, we show that optimal scale is larger, if the quality indicators are omitted.