Title: On the recent extension of price and production statistics to health and education
Abstract: Abstract The paper examines the recent extension of price and production statistics to health, social assistance services and education. In the first part of the paper it is emphasised that using net prices instead of gross prices, as for the European harmonised consumer price index, is inaccurate. The index is thus partly hybrid and inconsistent. The second part of the paper explains why, even if quality is a significant issue, production measures should not rely on the user's changes of condition. It should rather be based on supplied quality. The paper then describes and discusses two complementary ways to estimate quality changes. Notes 1. The Combined Activity Index is based on a panel of French public hospitals. Each particular therapy is classified according to the national list of Diagnosis Related Groups and corresponds to a standard level of activity. The total score influences the budget of hospitals. 2. There are two conceptions of the measure of the average cost. One consists in dividing the global cost by the number of students. It gives the average recorded cost that varies from year to year, depending on the number of students. The second is computed through the number of hours of training planned for each grade, taking into account the average cost for each hour of teaching. The number of students does not therefore influence the latter computed value. In the perspective of the paper the second conception is more relevant (see also below). 3. Vanoli 2003 Vanoli, A. 2003. Mesurer le ‘volume’ des biens d'équipement à l'aune de leurs performances. Courrier des statistiques, 108(December): 33–9. [Google Scholar] illustrates this kind of perspective in the case of equipment goods. 4. However in this perspective, when benchmarking relies on costs by pathology, it is crucial to adjust medical costs by pathology with the variation of provided quality. It would facilitate avoidance of the danger noted by Henriet [2002 Henriet, D. 2002. Tarification par pathologie: enjeux et perspectives de l'expérimentation en France. Dossiers solidarité et santé, : 21–8. hors série (July), La documentation française [Google Scholar]: 23]: ‘if no care is taken or no control is done for quality, this price system [cost by pathology] may lead to a decrease in care quality, because the implicit struggle on price drives the hospitals to give up quality’. 5. Unfortunately data give the number or duration of stays. 6. Usually the global expenditures of public administrations is utilised [Beudaert, 2004 Beudaert, M. 2004. “Mesure des performances dans le domaine de l'éducation”. In 10ème colloque de Comptabilité nationale Paris 22 January [Google Scholar]]. This indicator could be improved to converge with the approach implemented in health. It is preferable to multiply the official number of hours for each grade by the cost of an hour. This perspective would take into account the volume of education transferred and avoid potential colinearity between parameters. Managerial costs could be added, on a conventional allocation basis. 7. This computation is like splitting the change of the global index into two parts that can be multiplied. The first part is a Paasche's index (coefficients based on current year) calculated on prices. The second part is a Laspeyre's index (coefficients use the reference year) based on volumes.