Matthias Staat

Efficiency of hospitals in Germany: a DEA-bootstrap approach

Various attempts to assess the performance of German hospitals have generated a wide range of estimates regarding their efficiency. These attempts were based on different, often rather small data sets consisting of heterogeneous hospitals; the techniques applied range from simple benchmarking approaches to studies which employ Data Envelopment Analysis (DEA). Some studies report ‘dramatic differences in efficiency’ and propose savings potentials of 50%; others find an average efficiency in excess of 95% and characterize almost 75% of their observations as fully efficient. This study presents results for two datasets representative of two segments of the German hospital system. These segments comprise all hospitals that have one internal medicine and one surgery department; the hospitals are located in the old federal states of Germany. None of the hospitals provides tertiary care. DEA can be applied because all hospitals offer a comparable quality and range of services. The results were estimated with a DEA-bootstrapping procedure and suggest an average bias–corrected efficiency of around 80%.

The Effect of Sample Size on the Mean Efficiency in DEA: Comment

Zhang and Bartels (1998) show formallyhow DEA efficiency scores are affected by sample size. They demonstratethat comparing measures of structural inefficiency between samplesof different sizes leads to biased results. This note arguesthat this type of sample size bias has much wider implicationsthan suggested by their example. Models which implicitly restrictthe comparison set like some models for non-discretionary variableslead to biased efficiency scores as well. A reanalysis of theBanker and Morey (1986b) data shows that the efficiency scoresderived there are significantly influenced by the variation insample size implicit in their model.

Integrating categorical variables in Data Envelopment Analysis models: A simple solution technique

This paper introduces a novel method to incorporate categorical non-discretionary variables in Data Envelopment Analysis (DEA) models. While solutions to this problem have been introduced before, they have rarely been employed in applied work. We surmise that existing solution concepts pose problems for applied researchers and develop a simple and straightforward alternative based on indicator variables. We thereby provide a flexible tool for models with categorical variables that-unlike the approaches mentioned above-can be solved with standard DEA software irrespective of scale assumptions even if no option for non-discretionary variables is available. Furthermore, there is no need to split the data and run multiple DEA, one for each data set generated. The model is extensible to categorical discretionary variables and in addition to non-hierarchical data.

Product Performance Evaluation - A Super-Efficiency Model

This study introduces the concept of product performance from the perspective of customers. Product performance is measured as a ratio of outputs that customers obtain from a product relative to inputs that customers have to spend for purchasing and using the product. The output side is modelled by a set of customer-relevant parameters such as technical performance attributes but also non-functional benefits and brand strength; the input side reflects user costs. More than 60% of the cars in this study are rated as efficient and obtain the maximum efficiency value of unity. They form the efficient frontier of the compact car market representing a reference function for performance evaluation. Using a super-efficiency model, it is possible to differentiate the efficient products that are left with a score of 100% by standard efficiency models. Our approach is relevant for companies because implications for product design and market segmentation can be derived.

Measuring and Improving the Performance of Health Service Networks

Health maintenance organizations (HMOs) have intensified their efforts to establish network-like structures with service partners who are responsible for different functions along the health value chain. To calculate the potential value and cost benefits of service production within health care networks and to improve performance in such networks, the authors propose a two-step benchmarking approach. While the first step is concerned with measuring and comparing service provider performance, the second step relates to a contact program that disseminates the lessons learned during the benchmarking process. Across two empirical studies with general practitioners and specialty physicians, the authors identify in a first step tremendous overspendings and provide suggestions on cost reductions that could be achieved without threatening output levels. With regard to the second step, the authors find that detailing efforts based on the results of performance measurement helped physicians to improve their performance. Through detailing, the hub was able to inform network partners about the benchmarking results and to reveal performance gaps in their current resource utilization patterns. In addition, the authors show that managers of HMOs should seek out physicians with smaller practices and high-referral (i.e., risk-averse) physicians as targets for detailing, who are especially responsive to these initiatives.

Treating non-discretionary variables one way or the other: implications for efficiency scores and their interpretation

This paper explains the main DEA-techniques to model continuous and categorical non-discretionary variables as well as a related two-stage approach. The implications of using either alternative are demonstrated in practice using the pharmacy data from the original study by Banker and Morey (1986b) on categorical non-discretionary variables.

Health care in the CIS countries

The study analyses the technical efficiency of community hospitals in Ukraine during 1997–2001. Hospital cost amount to two-thirds of Ukrainian spending on health care. Data are available on the number of beds, physicians and nurses employed, surgical procedures performed, and admissions and patient days. We employ data envelopment analysis to calculate the efficiency of hospitals and to assess productivity changes over time. The scores calculated with an output-oriented model assuming constant returns to scale range from 150% to 110%. Average relative inefficiency of the hospitals is initially above 30% and later drops to 15% or below. The average productivity change is positive but below 1%; a Malmquist index decomposition reveals that negative technological progress is overcompensated by positive catching-up.

Bootstrapped efficiency estimates for a model for groups and hierarchies in DEA

Abstract We apply some recently introduced bootstrap techniques to derive bias corrected efficiency scores for a model for groups and hierarchies in DEA. The use of the bootstrap makes it possible to overcome some deficiencies of the original formulation of this model, which rests on rescaling individual efficiency scores using average efficiencies calculated from different subsets of the data. These average or structural efficiencies are differently biased and bias varies with sample size when standard DEA techniques are used. Bias correction makes it possible to identify the true differences in efficiency and thus to compare DMUs belonging to different groups via their rescaled individual efficiency scores on one common basis. Moreover, this type of bias problem is present in other DEA applications. Therefore, the method proposed to deal with it has many potential applications beyond the groups and hierarchies model.

Hedonic price function estimation in economics and marketing: revisiting Lancaster’s issue of “noncombinable” goods

Following Lancaster’s (J. Political Econ. 74(1):132–157, 1966; Variety, equity and efficiency, 1979) interpretation of his characteristics approach to consumer theory, this contribution focuses on theoretical and empirical arguments questioning the smoothness of traditional hedonic price estimation techniques. Lancaster argued strongly against “combinability”, i.e., that any efficient combination of characteristics is feasible and sensible. We therefore explicitly test the impact of convexity using a set of recent non-parametric estimators. The test is carried out on a sample of 114 digital cameras whose price evolution is followed over 6 months. The hypothesis of convexity is rejected using the Li (Econ. Rev. 15(3):261–274, 1996) test. The conclusions point out implications for economics and marketing.

Estimation of Labour Supply Functions Using Panel Data: A Survey

This survey aims at providing the reader with a thread through the literature on the topic of panel econometrics of labour supply, reporting also on the evaluation of the data used in these studies, and summarizing their substantive results. It documents the present trend away from models that take advantage of panel data almost exclusively in order to control for unobserved heterogeneity, towards fully dynamic models where wages become endogenous and consequently the concept of wage elasticity loses much of its appeal.