Methods for allocating the Gross Value Added to regions/provinces vary since they are determined by the type of data available and the organisation of the national statistical system. In general there are three methods of regionalisation:

The Ascending Method which presupposes the collection and treatment of the elementary statistical units, taking into consideration local level units of economic activity (establishments) and institutional units (households and public administrations), and adding them together until reaching the desired regional level;

The Descending Method which consists of disaggregation of the national product on the basis of a regional indicator, resulting in the utilisation of apportioning units, i.e. if a global indicator of the approximate phenomenon we intend to measure is used. Hence, the national aggregate is distributed on the basis of an indicator that is approximate to the variable we intend to estimate. The method is known as descending because the aggregate is allocated to a region on the basis of a local or regional economic activity. However, the notion of local activity, in most cases, continues to require an accurate regional allocation. For example, Gross Value Added for the railway transportation can be allocated to regions according to the number of passengers and aggregate tonnage of cargo transported in a given period.

The Mixed Method, which consists in using simultaneously the ascending and descending methods. For the ascending method is rarely found in its pure form. There are always gaps in the data which have to be filled by using the descending approach. Similarly, many descending methods frequently include data from exhaustive sources, as do ascending estimates.Thus mixed methods are the norm.

However, the choice between ascending and descending methods depends above all on the statistical sources available.

Here, we shall use the descending method, where the main regional aggregate is a replica of the following aggregates in the national accounts: Production, Intermediate Consumption, Gross Value Added, and Gross Domestic Product (GDP), in the perspective that the output resulting from these estimates is the reflex of the National Accounts, compiled and published by the INE's Department of National Accounts.

One advantage of this method is the numerical consistency between the national and regional accounts. Such methods are cheaper to develop in that they make use of existing data, and they do not require new exhaustive records. This is the most recommended method in the situation where there is no information from local units of economic activity.

Based on these assumptions, an estimate was first made of the value of production of each of the sectors of activity based on a sample of 142 products regarded as representative of all economic activity,

To allocate the production by provinces, the reference point taken was the balance sheets for each year drawn up by the INE's Department of National Accounts for each of the 142 products, on the assumption that the sum of the production of all the provinces should be approximately equal to the value of the production on the balance sheets.

Decisions as to the criteria previously defined are strongly constrained by the available data and its quality. In general, the use of indicators of gross production or of sales in order to allocate the Gross Value Added to regions, starts from the principle that the intermediate consumption corresponds to the same proportion of production in all regions.

Based on these criteria, the value of production by provinces was estimated for 1997, 1998 and 1999. Once the value of production was obtained, the coefficients of intermediate consumption by activity were applied, on the assumption that these coefficients are the same for all provinces. Finally, as SCN93 recommends, the Value Added was estimated in residual terms, by the difference between production and intermediate consumption.