1 The COHESION System of HERMIN Models: CSHM John Bradley (EMDS), Zuzana Gakova (DG REGIO), Philippe Monfort (DG REGIO), Gerhard Untiedt (GEFRA), Janusz Zaleski (WARR) Sixth European Conference on Evaluation of Cohesion Policy WARSAW, 30 NOVEMBER-1 DECEMBER 2009 NEW METHODS FOR COHESION POLICY EVALUATION: PROMOTING ACCOUNTABILITY AND LEARNING
2 HERMIN: Some Background Grew out of the HERMES model used to analyse the Single Market initiative Applied first to Ireland, and then to Greece, Portugal and Spain Extended to NMS/2004 over Extended to NMS/2007 after 2004 Extended to entire EU-27 in 2008 Models also developed for Turkey, and under construction for Croatia and Macedonia
3 HERMIN: Some Background HERMIN and then CSHM have been used for a long time by DG REGIO Evaluations and simulation of policy scenarios Close collaboration with the CSHM team
4 HERMIN: Theoretical Issues Detailed focus on supply side, with 5 production sectors: manufacturing, market services building & construction, agriculture and non-market services Output and factor demands modelled explicitly Standard treatment of demand side (consumption, trade) Uses simple autoregressive expectations, not forward-looking expectations (see Irish case study)
5 HERMIN: Incorporating cohesion policy mechanisms Explicit modelling of Cohesion Policy impacts of broad investment categories: infrastructure (PI) human resources (HR) R&D Distinction made between output effects and productivity effects of CP instruments International literature used to calibrate impacts Full accounting for CP instruments in government revenue and expenditure
6 CSHM/HERMIN: Country and Regional Coverage Original cohesion states: Greece, Ireland, Portugal, Spain Two macro regions: East Germany and Italian Mezzogiorno The ten new states joining in 2004 The two new states joining in 2007
7 CSHM/HERMIN: Supporting databases Data sources based on EUROSTAT data, mainly through the AMECO database provided by DG- ECFIN Augmented by national data sources, where necessary OMS (EL/IE/PT/ES): NMS 2004 & 2007: East Germany and Italian Mezzogiorno are specially constructed regional databases
8 CSHM/HERMIN Calibration and Testing Behavioural equations calibrated using time- series data, but imposed parameters sometimes needed Cohesion policy parameters derived from international literature, since little evidence available from EU cohesion states Models tested on standard policy shocks using projections to 2020 (remember: cohesion policy has long-term impacts!)
9 CSHM/HERMIN in action: Ex-post evaluation of Cohesion Policy Step 1: Use country model to derive a plausible projection out to 2020 time horizon, checking short-term projections against published ST forecasts Step 2: Extract cohesion policy shocks and carry out a without-CP simulation Step 3: Derive CP impacts by comparing Step 1 with Step 2
Cohesion Expenditure Impacts: Poland a) GDP Effect c) Net Trade Balance Effect b) Employment Effect
11 Comparing rates of return on Cohesion Policy (CP) expenditures The cumulative CP multiplier Cumulative % increase in level of GDP __________________________________ Cumulative CP injection as % of GDP
12 Time profile of Cumulative Multipliers Example of Poland: CP
13 Cohesion Expenditure Comparative Impacts Cumulative Multipliers by 2020
14 Cohesion Policy Sectoral impacts: Poland a) GDPFC effect b) OT effect
15 Cohesion Policy Sectoral impacts: Poland c) OM effect d) OB effect
16 Conclusions 1 Cohesion Policy Impacts Demand impacts of Cohesion Policy are confined to the implementation period Supply-side impacts of Cohesion Policy build up gradually, and endure after programmes end The rate of return on cohesion expenditure can vary due to structural differences (e.g., Greece versus Ireland; Lithuania versus Poland) Difficult to evaluate efficiency and effectiveness of Cohesion Policy national planning and implementation
17 Conclusions 2 Cohesion Policy Impacts Cohesion Expenditure increases demand for imports by beneficiary states during implementation Import demand remains higher after CP implementation compared to the no-CP scenario, but impacts on net trade balance become positive The orientation of net donor states to trade with beneficiary states affects the transmission of spillover impacts from recipient states to net donor states
18 CSHM/HERMIN: Ongoing Research Activities Adding in the net donor states and evaluating spillovers to net donors (see European Parliament publication, 2009) Adding in the candidate states (Turkey (2008), Croatia, Macedonia) Adding in the applicant and potential applicant states (Bosnia, Serbia, Montenegro, Kosovo, Albania (2009)) Refining the CSHM models and cohesion policy mechanisms
19 THANK YOU FOR YOUR ATTENTION Questions ?