Approaches to the measurement of efficiency in a dynamic context : with an application to the UK banking sector
In this thesis we put forward two approaches to the measurement of the cost efficiency of firms in which adjustment costs affect the optimal allocation of factor inputs. The two approaches we consider are, first, based o n direct calculation of an efficiency index and, second, the estimation of a parametric model. The implementation of both approaches require extensions to the existing theory. In developing the cost efficiency index approach we suggest that the Tornqvist index can be extended to be "exact and superlative" (Diewert (1976)), even when there are adjustment costs. However, the formulae for this index becomes dependent on the unknown adjustment cost function and this needs to be estimated in reduced form by econometric methods. We find an extension of the parametric model approach that incorporates adjustment costs to be preferable, since it gives a greater understanding of the structure of adjustment costs - not only their direct influence on the costs of firms, but also the extent to which they alter the optimal factor demands (a feature that in a static model might be interpreted as allocative inefficiency). The empirical analysis in this thesis applies both approaches to estimate efficiency differences between firms, and over time, in the UK retail banking and building society sectors. The parametric model we develop, incorporating adjustment costs, is particularly appropriate since these sectors have undergone considerable change over the period we study (1978 to 1987), in terms of both the level of output and optimal factor demands (with labour substituted for progressively more computer and information technology equipment). Furthermore, the parametric approach is easily adapted to deal with other particular characteristics of these sectors. In particular, we are able to extend our model to freely estimate equipment depreciation rates. This results in using an unbiased estimate of the user cost of equipment when estimating cost efficiency.