Analysis of documentary evidence and representative in-depth interviews
Pillar 2 measures have always been considered as a means for reconfiguring regional structures (Shortall, 2008) representing a change from an agri-centric policy view to a broader multi-sectoral one (Scott, 2004). The current approach is mainly focused on the effects of Pillar 2 and is a mixed-method case study intended to provide an understanding of the impacts on predominantly rural regions. The approach tries to explain how Pillar 2 interacts with the structure and performance of the local rural economy (Yin, 1994) instead of identifying effects on rural employment. Applied in all five EU regions, the method followed a coordinated two-stage data gathering process, an investigation of secondary data offering a contextual framework for the overall study, and in-depth semistructured interviews with representatives of different interest groups.
First, a regional profile was developed to provide the context in which key informants operate and to inform the process of analysis. Then, key informants were identified and interviewed to explore their perspectives on policy issues. Participants in the interview process — drawn mainly from policy makers, business managers, regional NGO officers and LEADER group managers — were invited to respond to and interact with a set of pre-drafted thematic questions. Finally, analysis explored patterns within the multiple data sources, in order to provide support for explanations of the casual relationships (Midmore et al., 2008).
Positive Mathematical Programming (PMP). PMP was applied to identify and measure policy-induced changes at the individual farm level and then upscaled to regional level. The methodology was common to all case regions, and FADN data were used. Regional models allowed the assessment of the main effects of two different policy scenarios: full decoupling, and full decoupling plus price variations. A special sub-model, implemented within the PMP model, captured labour allocation inside the farm with respect to new production plans induced by CAP reform (Arfini et al., 2003; Heckelei, 2002; Judez et al., 2001; Paris and Arfini, l995).
Input-Output Analysis (I-O). This approach was selected to assess impacts on output, household income, and employment on the whole regional economy of the selected regions. I-O analysis constitutes a tool which can be used to show how industries are linked together through supplying inputs for the output of an economy. Thus, building a regional I-O table provides a clear picture of the structure of the economy, and the existing relationships amongst various regional sectors can be identified. First, regional input-output tables were constructed using the accurate and widely adopted Flegg-Weber technique (Flegg et al., 1997). Second, following the GRIT (Generation of Regional Input-Output Tables) approach (Jensen, 1990), these non-survey regional input-output tables were hybridized with the addition of survey data on key rural economic transactions. The application of the model allowed the estimation of various I-O linkage coefficients (multipliers) for each region, i.e. involving Chenery-Watanabe direct linkages, Rasmussen-Hirschman linkages (output, income and employment multipliers), Mattas-Shrestha I-O elasticities (output, income and employment elasticities), and Papadas and Dahl supply-driven multipliers (Chenery and Watanabe, 1958; Hirschman, 1958; Mattas and Shrestha, 1991; Papadas and Dahl, 1999; Rasmussen, 1956). This facilitated the identification of the most important economic sectors (as regards their potential to enhance regional employment, income and output levels), the estimation of indirect and total economic impacts, and ultimately comparable results among the selected regions. Results from the PMP model were fed into the I-O model to observe the indirect and induced changes for the whole economy (Mattas et al., 2005; Miller and Blair, 1985).