Important legal notice
AGRILIFE- Agriculture and life sciences in the economy
SUSTAG - Sustainable Agriculture
Modelling Rural Economies (MoRE)


This research aims at attempting to capture the local rural / urban effect of policies. It covers an extension of the previous RURAL EC MOD project by adding further case studies and ensuring higher consistency between them, as well as examines the possibility to derive from CAPRI-related work (IONUTS2) a more exhaustive approach than case studies to the representation of local / regional rural economies (for the purpose of modelling agricultural and regional policies at NUTS3 level).

Task 1) More detailed case studies better representing the diversity of regional economies throughout the EU-27 ('Expert' SAMs)

Building on the previous RURAL EC MOD project, the first element of the present project is to build SAMs and CGE models for more case studies (12) in addition to the existing 7 realised previously.

A preliminary work will be carried out to improve the 'representativeness' of case studies (in terms of total number and their representation of different clusters of NUTS-3 areas, identified on the base of several indicators: rurality, peripherality, economic diversification, CAP policy mix,…), which will be taken into account in the choice of the new case studies and in the concluding steps. This task will be outsourced early in the process and will rely on the main results indicated in the last section of the RURAL EC MOD final report.

Case studies depart from existing official broader level (in general national, NUTS2 in few cases eg Spain) SAMs (or input-output tables): an 'expert' rural-urban ('rurban') SAM is built for each NUTS3 region using the full set of statistical and expert information locally available (in particular FADN, FSS micro data, regional or national detailed economic activity and regional accounts for employment and gross value added, household and labour force surveys).

One important improvement of the previous RURAL EC MOD project aims at ensuring the structural uniformity of SAMs (same accounts, same level of disaggregation, with some limited flexibility – e.g. small/large farms separation; agriculture disaggregation) in order to increase the comparability of results. Rural / urban definition will be harmonised and will strictly follow the LAU classification agreed between AGRI and REGIO.

Concerning the models, the IFPRI-based dynamic recursive non linear models will continue to be used (following a Monte Carlo approach). Consistency will be sought throughout case studies as regards model parameters and closure rules. Comparison of such complex 'black-box' models with more simple ones such as static linear CGE models or SAM multiplier models will continue being carried out.  

Analysis of ex-ante impact of several policy instruments and/or options (policy mixes) will be carried out in a territorial impact assessment framework (differentiating per type of region the reaction of different types of responses to the policy measures).

For this purpose further investigation will be dedicated to the concrete modelling of RDP measures in the framework of the new RD policy (improving the modelling of (ex) axis 1 and 3 measures; further attention towards (ex) axis 2 measures, if possible new measures in RDP 2013-2020 or cohesion policies instruments of relevance for the rural economy) as well as on pillar 1. In contrast with the former projects, simulations will rely more on average policy measure than on the specific mix of each region (in order to facilitate comparison of impacts).

Several specific key methodological issues will be addressed in the course of the project (e.g. how to best model ex axis 2 type measures?...) and they will be identified in detail from the beginning and during the project. For this purpose, an advisory board will be set with high-level academics.

Task 2) Towards an exhaustive representation of EU rural economy ('automatic' SAMs)

The feasibility of deriving by mechanical procedures from IONUTS2 full set of SAMS at NUTS-2 level database a  full set of NUTS3 level rural economies (and accordingly of urban ones) will be assessed.

As such, the mechanical procedure will have to rely on simple data available throughout the EU-27 for all rural areas (defined at lower level (LAU-1 and/or LAU-2), e.g. share of rural / urban population), the feasibility of such approach will be investigated, taking into account the availability of data at EU level.

Beyond the feasibility of such approach, its coexistence with more detailed case studies, relying on a larger set of regional statistics, will allow assessing to what extent an exhaustive and mechanical approach allows to capture the differences of economic structures at local level.

A two-step comparative analysis of the 'expert' SAM (case study) and the 'automatic' SAM (ex ante and ex post) will be carried out following a somewhat related work of Cardenete and Sancho (2004) . In the first place we will calculate some proximity/distance indicators to assess the degree of similarity between the qualitative SAM and the automatic SAM. Among those, the index of similarity (Le Masné, 1990), the Pearson correlation coefficient, or Theil's U can be calculated to indicate the closeness between SAMs. Then, the comparison will be based on the examination of implications of using both SAMs in the RURAL ECMOD CGE model. Simulation results obtained via the qualitative SAM will be compared with those obtained through the use of the automatic SAM, in terms of relevance and magnitude of impact estimates. Pearson correlation and weighted correlation coefficients can be calculated to evaluate the degree of similarity of impact patterns produced by the use of both SAMs. The comparison will allow obtaining more insights with respect to the influence of the SAM creation process on simulation results

Cardenete M. A. and Sancho F., Sensitivity of CGE simulation results to competing SAM updates, The Review of Regional Studies, Vol. 34, No. 1, 2004, pp.37-56

Expected results: end 2013


Sébastien Mary, Fabien Santini and Sergio Gomez y Paloma
Joint Research Centre