A model for measuring the accuracy in spatial price statistics using scanner data

Abstract

Given the crucial role of Spatial Price Indices (SPIs) for comparing standard of living across a country, there is a need to assess their accuracy on a regular basis. However, despite the importance attached to SPIs, with few exceptions reliability measures are not computed and published both at an international and national level. Focussing on SPIs, this paper aims at suggesting a measure for assessing SPIs accuracy through the adoption of a new approach based on the Jackknife replication technique. To illustrate the potentialities of the suggested approach an empirical application is provided using the 2018 Italian National Institute of Statistics - Istat scanner data on the ten provinces of Toscana for selected groups of products. Our results demonstrate a large spatial heterogeneity between the SPIs of the Toscana provinces and their standard errors