38th Congress of the European Regional Science Association: "Europe Quo Vadis? - Regional Questions at the Turn of the Century", 28 August - 1 September 1998, Vienna, Austria
In a big amount of economies (NUTS-I) the evolution of manufacturing production is analysed using Gross Domestic Product (GDP) and Gross Added Value (GAV) data from National Accounts. In Spain, the problem of using these data is that they are not available as soon as it would be desirable. In consequence, it is not possible to analyse the short term evolution of the industrial output through them. To solve these problems the Institute of Statistics of Spain (Instituto Nacional de Estadistica -INE-) constructs a monthly Industrial Production Index (IPI) from data belonging to a survey addressed to firms. At a regional level (NUTS-II), the difficulties to monitor the evolution of manufacturing production are even bigger due to the nearly absence of official data. During the last years, different public and private institutions have started to construct indices for some Spanish regions, but they do not use an homogeneus methodology and the indices are not directly comparable. In this paper, we summarize and extent the main results of previous studies about the possibility of using different indirect methods to analyse the short term evolution of regional industrial production. In concrete, two statistic and an econometric method are considered. First, we study the possibility of extending the methodology proposed by the Regional Institute of Statistics of Catalonia (Institut d'Estadistica de Catalunya -IEC-) to other Spanish Regions. Second, we analyse the relationships between electric energy consumption for industrial purposes and industrial production. Third, following Israilevich and Kuttner (1993), we apply a state-space model to obtain estimates of the industrial production indices using the Kalman Filter and the method of maximum likelihood. Next, to validate the indices obtained through these three methods we compare them with regional indices obtained by direct methods for the regions where they exist. Finally, we expose the main conclusions remarking the implications for public policy in relation with elaboration of regional statistics.