Conciliación espacial-sectorial de trimestralización de seriesuna aplicación a las contabilidades provinciales
- Fernández Arufe, Josefa E. (dir.)
- Rojo García, José Luis (dir.)
- Moyano Pesquera, Pedro Benito (coord.)
- Somarriba Arechavala, Noelia (coord.)
Publisher: Asociación Española de Economía Aplicada, ASEPELT
ISBN: 84-96477-93-2
Year of publication: 2007
Volume Title: Área VII : Métodos cuantitativos
Volume: 7
Pages: 188-212
Congress: ASEPELT España. Reunión anual (21. 2007. Valladolid)
Type: Conference paper
Abstract
In a previous paper, we present a bayesian hierarchical normal-gamma model, allowing for benchmarking of time series, using indicators. The model has been compared with classical ones (AR(1)-Chow-Lin, among others), and its practical interest showed by benchmarking annual data from the Spanish Regional Accounts (I.N.E.) The disaggregation of time series linked by accounting constraints (contemporaneous and/or temporal ones) involves the conciliation of individual disaggregations. In recent years there have been relevant contributions to this topic but, as far as we know, the bayesian approach has not been used. Classical approaches are well described in the literature, and recent proposals point to the conciliation of SA series from concilied original ones. In this paper, we deal with the simultaneous benchmarking of a two-way classified table of time series, by using a bayesian hierarchical normal-gamma model that superimposes the individual and the global indicators models and by settling the appropriate random structure. Authors derive the joint optimal solution and, finally, the method is used to benchmark the province's Spanish Regional Accounts (Spanish Statistics Institute) with the Quarterly Regional Accounts (Regional Statistics Institute of Castilla y León)