Approach

The official agricultural censuses conducted by the national statistic offices of the Alpine countries since the end of the Second World War form the starting point for this project. These surveys took place in a more or less ten year cycle (1950, 1960, 1970, 1980, 1990, 2000). The results of AGRALP are based on the data on municipality level of these censuses (LAU 2, former NUTS level 5).
For the Alpine-wide analysis of the agrarian structures, central indicators were selected. Among others, indicators were considered from the following areas:

• farm structure (number of farms, farm abandonment rate, farms with full-time and part-time farming),
• land use (utilised agricultural area according to type of culture), and
• livestock (number of livestock by species, livestock density).

In order to compare the data collected from all the Alpine countries at national and international levels in a correct scientific manner, an essential step in the process is to harmonise the data. For this scope we analysed the national definitions of the indicators in order to determine the existing differences.

Example: Number of farms. In the Alpine countries they are registered according to different thresholds. Furthermore in some countries between the censuses these thresholds were changed. In most EU-member states farms are registered according to a definition of the EU counting farms with at least 1 ha utilised agricultural area (UAA). This limit is applied in AT/FR/CH/LI/SLO. In IT there do not exist any limit. In IT consequently only on farms with 1 ha and more UAA was looked at. It is therefore target-orientated to harmonise the number of farms according to the limit of the EU and if possible to balance the differences through relative calculations.

In 1999 in DE the limit of 1 ha UAA effective for farms since 1971 was raised to 2 ha UAA. For a correct comparison on a national level between 1980 and 2000 only farms with at least 2 ha UAA were investigated. Through this up rating in 1980 and 2000 a certain number of farms between 1 and 2 ha UAA is not registered within the German area of the Alpine Convention. Consequently these farms are "missing" within an transnational comparison. In the course of the accession into the EU even in Austria a rearrangement of the limit from 1 ha total area to 1 ha UAA took place. According to a statement of Statistik Austria Austrian-wide about 24,000 farms were not counted any more.

If a calculational equalisation is not possible due to the data situation respectively to large differences of the definitions, quantitative data on the differences found in comparisons is expressed as precisely as possible. Furthermore, there do exist further aspects, which require the data harmonisation. Recapitulating, an equalisation of data is important due to:

a) differing national definitions of the indicators taken in the main censuses. 
Example: Differing thresholds in the registration of farms (DE: 2 ha UAA, FR/AT/CH/LI/SLO : 1 ha UAA, IT: no threshold).

b) differing measurements of the registered characteristics of the farms. 
Example: Data recorded in ares and hectares.

c) changes in certain definitions between surveys. 
Example: Rise in threshold of registered farms in DE from 1 ha UAA to 2 ha, in AT from 1 ha total area to 1 ha UAA.

d) changes in the administrative organisation (municipality changes) / necessary adaptation of municipality data to the status of the year 2000. 
Example: In 1992 in IT new formation of the 3 Provinces, Verbano-Cusio-Ossola, Biella, and Lecco.

Special circumstances are given by the statistical secret. In FR aggregated data on municipality level are only publicated if there exist at least 3 farms in the municipality. Similar can be said for Bavaria where certain values are defined as not known or statistical secret.

The corrected data make a statistical analysis possible in a further step, as well as the cartographical presentation of the indicators. The cartographical presentation is carried out by means of the Geographical Information System (GIS); the statistical evaluations are carried out with the help of the statistic program, SPSS.

Integration of additional data

The development of agriculture is closely dependent on the ecological, social, and economical factors in which it is embedded and cannot, therefore, be examined without taking these factors into consideration. This includes the aspect of subsidies for agricultural sector, too. To understand agricultural development, it follows that the local conditions and the development of the rest of the economy must be included in the analyses. Therefore, the agricultural data is supplemented with demographical municipality data from population censuses, and with socio-economical, tourism, infrastructural (roads, railroads, industrial areas) and facility (hospitals, schools) data pertaining to the Alpine area. Information on compensatory allowances, agri-environmental measures and other forms of support is integrated. This allows a comprehensive interpretation of agrarian structure change and an investigation of the factors influencing it. Furthermore, linking the agri-structural data with the pan-alpine digital elevation model and the Corine Land Cover Data facilitates additional structural, regional, and spatial analyses.