Download Policy brief RURAL ECMOD - agrilife

RURAL ECMOD
Policy Brief
Policy context
The recent Common Agricultural Policy (CAP) reforms have been driven by a desire to
increase market orientation of EU agriculture, and to adapt to new societal demands. The
changing weight given to the various objectives of the CAP has been reflected in changes in
CAP policy instruments. Reflecting the need to address future challenges, the Commission
recently published “CAP towards 2020” (European Commission, 2010a), which emphasises
the multi-dimensional and complementary objectives of the CAP.
At present, the CAP is structured around two complementary Pillars. Pillar 1 provides
product and producer support mostly through decoupled income payments. Pillar 2 (the rural
development policy (RDP) Pillar) involves policy instruments aimed at improving
competitiveness of agriculture and forestry (Axis 1), improving the environment and the
countryside (Axis 2), and improving the quality of life in rural areas and encouraging
diversification of rural economic activity (Axis 3).
Against this background, the RURAL ECMOD research project aims to improve
understanding of the regional economy impacts of CAP policy instruments, and, in particular,
the impacts associated with switch away from an agriculture-centred focus, to an approach
aimed at the balanced and sustainable development of EU rural areas.
Methodological approach and study areas
The project adopts a dynamic Computable General Equilibrium (CGE) modelling approach
to the ex-ante assessment of various policy scenarios in six, specially selected, EU NUTS3
regions. Table 1 indicates the diversity in terms of population, income per capita, and
importance of agriculture across the six study areas.
Table 1: Case Study Regions (2005)
Arkadia
Potenza
Jihomoravsky
kraj*
1130.30
Aberdeen &
Aberdeenshire*
504.40
Guipúzcoa
RheintalBodenseegebiet**
273.20
Population
89.30
391.10
682.10
(thousands)
Per capita GDP (thousand euros)1
Total
14
12
9
30
26
Rural
11
12
8
22
25
Urban
21
16
9
37
27
Contribution of agriculture to rural areas (%)
Employment
37.5
11.5
2.9
0.8
1.0
Value added
12.5
6.6
2.6
2.8
0.8
Nature of CAP support in base year
% share of RDP
funds
47%
32%
34%
28%
30%
% share of Axis 3
8%
6%
9%
6%
2%
1
Derived from base year SAM for each case study region.
*Converted to Euros based on a 2005 exchange rate of 1 EUR = 0.67 £ and 1 EUR = 29.8 CZK respectively.
** Combined contribution of agriculture, forestry and fishing to employment, and value added.
27
26
27
0.1
0.2
80%
6%
1
The models are run over the period 2006 to 2020. In each period the models generate values
for all market transactions for all sectors, markets and economic actors in the local economy.
In particular, the direct and indirect effects, and displacement or spill over effects in factor
and product markets are captured by modelling all the sectors and markets in the local
economy simultaneously. Clearly in small open regional economies, imports and exports to
and from any region are important determinants of how any shocks to any sector are
transmitted to the rest of the regional economy. The models also allow imports to compete
with local products in regional markets, while exports provide alternative destination for
regional output.
Between periods, key model values are updated as required by the simulations, e.g. to allow
for adjustment in capital for each sector, or predicted population growth (Lofgren et al, 2002,
Thurlow, 2008).
Policy scenarios
Ex-ante policy impact analysis is based around seven policy scenarios (see Box 1). These are
compared to a baseline scenario consistent with the outcomes of the 2008 CAP Health Check
(European Commission, 2010b). The first four scenarios focus in the impacts of relatively
major changes in agricultural and rural policy in the six study areas, the last three assess the
impacts of changes in the relative weight given to different Axis 3 measures. As appropriate,
the policy changes are phased in over a period of time and the impacts monitored to 2020.
Box 1: THE RURAL ECMOD SCENARIOS
In each scenario all flows not mentioned in the specification follow the Baseline
specification.
Group 1: Changes in the distribution of Pillar 2 funds
Scenario 1 – “Agricultural” RDP: All RDP spending focussed on Axes 1 (competitiveness)
and 2 (environmental measures).
Scenario 2 – Diversification RDP: All RDP spending focussed on Axis 3 (economic
diversification and quality of life).
Group 2: Decrease in Pillar 1 funds
Scenario 3 – 30% Reduction of nominal Pillar 1 support.
Scenario 4 – Rebalancing Scenario: EU-wide flat-rate Single Payment Scheme introduced,
nominal non-SPS (e.g. Article 68) Pillar 1 funds decrease by 15%, nominal Pillar 2 funds
increase by 45%.
Group 3: Comparison of Axis 3 measures
Scenario 5 – Farm household diversification: All Axis 3 funds switched to 311
(Diversification into non-agricultural activities) targeting agricultural households.
Scenario 6 – Non-farm diversification: All Axis 3 funds switched to 312 (Support for
business creation and development) and 313 (Encouragement of tourism activities) both
targeting the non-farm rural households.
Scenario 7 – Rural Public infrastructure: All Axis 3 funds switched to 321 (Basic services
for the economy and rural population), 322 (Village renewal and development), 323
(Conservation and upgrading of the rural heritage) all of which target rural public
infrastructure.
2
Key Findings 1: The total impacts of the policy scenarios
Group 1 scenarios: Changes in the distribution of Pillar 2 funds
Figure 1 shows the aggregate GDP impacts of the first group of scenarios.
Figure 1: Average annual percentage change in total GDP arising from changes in the
distribution of Pillar 2 funds
0.25
% annual chnage in GDP
0.20
0.15
Arkadia
0.10
Potenza
0.05
Jihomoravsky kraj
0.00
-0.05
Scenario 1
Scenario 2
-0.10
-0.15
Aberdeen-shire
Guipuzcoa
Overall Impacts
Focusing Pillar 2 funds
away from agriculture
(scenario 2) typically
increases Regional GDP
very slightly.
Rheintal-Bodenseegebiet
-0.20
-0.25
The total (aggregate) effects of both scenarios are very small, with Jihomoravsky kraj
showing the largest GDP impact in both cases. Indeed, only in this region can the total effects
of the policies be viewed as non-negligible. The employment effects of the policy scenarios
are similar in magnitude and direction.
While the total impacts from a shift towards agriculture-related Pillar 2 spend (Scenario 1)
gives rise to negative or zero effects across all study areas, the total effects of a shift in Pillar
2 funds towards Axis 3 measures (Scenario 2) gives rise to positive effects in five of the six
study areas (Arkadia being the exception). However, again the total effects are extremely
small.
Differences in the magnitude and, in Arkadia’s case, direction of total impacts are due to the
unique structure of each economy and the nature of sectoral and spatial spill over effects as
discussed further below.
Group 2 scenarios: Decrease in Pillar 1 funds
Figure 2 shows the aggregate GDP impacts of the second group of scenarios. Again the total
GDP effects are marginal. In the case of Scenario 3, three of the six study areas have a zero
total impact and only the Jihomoravsky kraj impact can be considered non-negligible.
The variable direction of impacts under Scenario 4 is perhaps not surprising given the fact
that the switch to an EU-wide flat-rate Single Payment Scheme would increase Pillar 1 funds
in some study areas and decrease it in others. However in this scenario, as well as in the
others analysed, there are significant underlying adjustments in the distribution of GDP and
3
employment between sectors and across rural-urban areas of the study cases which are hidden
by the total effects shown in Figure 2.
Figure 2: Average annual percentage change in total GDP arising from a decrease in Pillar 1
funds
0.20
% annual change in GDP
0.15
Arkadia
Potenza
0.10
Jihomoravsky kraj
Aberdeen-shire
0.05
Overall Impacts
Significant Reductions in
Pillar 1 funds (scenario 3)
have negligible overall GDP
effects (but may have
sectoral effects e.g. on
agriculture).
Guipuzcoa
Rheintal-Bodenseegebiet
0.00
Scenario 3
Scenario 4
-0.05
Overall Impacts
A Flat rate Single
Payment Scheme
(scenario 4) increases
Pillar 1 flows in some
areas.
Group 3 scenarios: Comparison of Axis 3 measures
Figure 3 shows the aggregate GDP impacts of the third group of scenarios.
Figure 3: Average annual percentage change in total GDP arising from a redistribution of
Axis 3 funds across measures
0.60
% annual change in GDP
0.50
Arkadia
0.40
Potenza
0.30
Jihomoravsky kraj
0.20
Aberdeen-shire
0.10
Guipuzcoa
0.00
-0.10
Scenario 5
Scenario 6
Scenario 7
Rheintal-Bodenseegebiet
Overall Impacts
Investing in Rural Public
Infrastructure (scenario 7)
will have negligible effects
on Rural GDP unless it
significantly increases in
tourist demand and
population increase.
-0.20
Scenario 5 and 6 (which switch funds within Axis 3 towards agricultural and non-agricultural
labour respectively), have the lowest total GDP impacts of all scenarios. This suggests
changing the distribution of axis 3 funds within a study area from its initial distribution to
either agricultural or non-agricultural diversification, has no effect.
Scenario 7 (rural public infrastructure) provides an illustration of the impacts arising from a
direct impact of increased investment in rural public infrastructure at the expense of
investment in marketed productive rural industries. As with the other Group 3 scenarios, the
magnitudes of the results are very small in five of the areas. The results for the Aberdeen and
4
Aberdeenshire study area are much stronger due to the attempt to try to capture (albeit
crudely) the impact of the extra service provided by the public sector investment on the
attractiveness of rural Aberdeenshire as a place of residence and tourism. It follows that the
results for this study area thus provide an indication of how much in-migration would have to
increase in order to obtain a significantly positive impact on GDP (in this particular case, the
0.5% increase in GDP shown in Figure 3 was associated with a 0.1% annual increase in
population and 1% increase in tourist demand).
Key Findings 2: Sectoral spill over effects
Each scenario represents a different combination of positive and/or negative shocks to
agriculture and non-agricultural rural industries. The associated direct effects of these depend
on the implementation of RD policy which varies widely across study areas. The indirect and
spill-over effects occur through the changing structure of input demand, changing product
and factor prices. The overall impact of these is ambiguous. Hence, for example in scenario
2, the direct impact of moving funds from Axis 1 and 2 to Axis 3, decreases agricultural
investment and (partially coupled) payments to farm households, while the increase in nonagricultural rural investment (associated with Axis 3) increases output in some sectors.
Hence, the direct impact on agriculture in this scenario reduces agricultural GDP for all
regions, while the sectoral spill over effects of these to the rural secondary and tertiary sectors
are region dependent.
Table 2 illustrates that the sign of the overall sectoral spill over effects differs across study
areas. The Table also shows that although the overall impact on secondary and tertiary rural
GDP is typically positive (except for Potenza and Arkadia), the pathways through the shock
differ across regions, with the pattern of changes in wages and prices quite distinct.
Employment effects follow GDP in terms of direction (see Table 2) and magnitude.
Table 2: Direction of sectoral GDP, Employment, Wage and Price effects, Scenario 2
(Diversification RDP)
Arkadia
Potenza
Jihomoravsky
kraj
Aberdeen &
Aberdeenshire
+
-
+
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+
+
+
+
+
(Semi) Skilled
Labour
-
+
-
+
+
Unskilled Labour
+
-
-
-
-
-
+
-
+
-
-
+
-
0
-
+
-
GDP
Agriculture
Rural secondary
Rural tertiary
Guipúzcoa
RheintalBodenseegebiet
Employment
Rural secondary
Rural tertiary
Wages
Prices
Total manufacturing
Total services
Sector Spill over effects
The pattern of price and wage changes induced
5
by moving Pillar 2 funds away from agriculture
is very different across regions.
Key Findings 3: Rural-Urban spill over effects
Both the magnitude and direction of effects on urban areas from agricultural and rural
policies is study-area-specific. Figure 4 shows, as an example, the spatial impacts of Scenario
1 (Agricultural RDP). The spill over effects of this scenario on urban areas is negative in the
rural and more agriculturally-dependent regions, variable in intermediate regions
(Jihomoravsky kraj and Aberdeen and Aberdeenshire) while in the urban regions, there are no
discernable spill over effects. As in the case of the sectoral impacts, the results reflect
differing characteristics of the regions, including, amongst other factors, the spatial
distribution of agri-businesses within the region and spatial patterns of labour and capital
ownership.
Figure 4: Average annual percentage change in GDP by study area, Scenario 1("Agricultural"
RDP)
0.20
% annual change in GDP
0.10
0.00
-0.10
Total GDP
Rural GDP
-0.20
Urban GDP
Spatial Spill over effects
Concentrating Pillar 2 on
agriculture can lead to both
positive and negative spill-over
effects on wider rural and urban
GDP.
-0.30
-0.40
-0.50
Consistent with Scenario 1, the other scenarios showed region-specific rural-urban spill over
effects.
Concerning the rural GDP, scenario 1 (Agricultural GDP) has clear positive effects in regions
where agriculture constitutes an important share in the economy (Arkadia, Potenza), while
opposite effects are shown in the diversification policy (scenario 2). On the other hand,
diversification strategy has beneficial effects on the rural GDP of regions which are already
well advanced in diversification.
Key Findings 4: Farm household income effects
Data availability allowed the modelling of a specific Farm Household group in five out of the
six regions. Table 3 shows that the impact of the simulated changes on farm households
varied both in terms of direction and magnitude.
6
Table 3: Direction and Magnitude of Farm Household Income Effects
Arkadia
Potenza
Jihomoravsky kraj
Group 1: Changes in the distribution of Pillar 2 funds
Scenario 1
+
+
+
Scenario 2
Group 2: Decrease in Pillar 1 funds
Scenario 3
+
Scenario 4
+
0
Group 3: Comparison of Axis 3 measures
Scenario 5
-/+1
Scenario 6
-/+
+
Scenario 7
+
Min/Max % Change
-8.5/ 0.3
-25.6/6.8. -0.02/0.02
1
Impact for Small and Large farm Household respectively.
Aberdeen &
Aberdeen
-shire
Guipúzcoa
RhientalBodensee-Gebiet
+
+
-
n/a
n/a
+
+
n/a
n/a
+
-
+
-
n/a
n/a
n/a
-10.8/ 5.3
-10.3/2.7
.
Farm Household Income Effects
Typically, increased RDP Farm Diversification
investment alone does not compensate Farm
Household income derived from r reductions in Pillar
1 or agricultural Pillar 2 support.
As expected, Scenario 1 is associated with an increased in farm household income, while in
Scenario 2 farm household income fell in all regions except Aberdeen and Aberdeenshire
suggesting that the returns to the increased investment in farm diversification in this scenario
are insufficient to counteract income falls derived from agriculture. With the exception of
Jihomoravsky kraj where the impact is very small, the decrease in Pillar 1 support in Scenario
3 reduces farm income. Scenario 4 typically increases farm household income, except in the
case of Arkadia where Pillar 1 support decreases substantially.. With the exception of
Jihomoravsky kraj, the redistribution of Axis 3 funds away from measures tied to farm
households reduces farm incomes.
There is some evidence that in areas with low levels of pluriactivity (Arkadia, Potenza,
Aberdeen and Aberdeenshire, and Guipúzcoa), the negative effects on farm household
income derived from reducing agricultural support is more pronounced. However, further
research is required before this result can be validated.
Policy implications
Dependence on CAP support

The importance of CAP support to rural areas varies widely across the EU in ways that
are not reflected by the sectoral importance of agriculture in the economy.

In areas where farm household income is an explicit objective of the CAP, support
associated with agricultural production remains an important determinant of farm
household income. Therefore, it appears difficult to compensate for a reduction in
agriculture-related support through measures aimed at on-farm diversification.
7

The role of RDP in stimulating overall rural development appears limited in many areas.
Where the objective is overall rural development, the nature of RDP policies need to be
re-evaluated as there may be more effective policy measures for supporting the wider
rural economy.
Territorial differences

The diversity of results across study areas reinforces the menu-driven nature of the
RDP where member states are able to tailor the policy to specific regional needs; an
obvious example being the more or less beneficial effect of diversification measures
on rural GDP depending on the degree of diversification already achieved in the
region concerned.

Horizontal policies or measures that are implemented not considering regional
differences, will inevitably fail to take into account territorial factors that mediate
policy impacts such as the degree of labour market integration or the spatial
distribution of upstream and downstream firms within a region..

The results confirm that changes in the CAP can have impacts for urban as well as
rural areas which need to be taken into account in policy design.
Improving policy evaluation

There is need for the development of indicators to reflect territorial factors that are
important in determining the policy outcome. These include the size and integration
of labour markets, the extent of sectoral integration within the rural economy and
agri-food chain, the distribution of agriculture-related businesses in a region and
patterns of factor ownership.

The results suggest that focussing on total effects may mask negative income and
employment effects at the sectoral level or at a sub-regional (rural versus urban) level.
More detailed analysis of policy impacts is therefore required.

The experience from the project suggests that better data on the sectors that are
benefiting from Pillar 2 measures would improve policy evaluation.
Further research

Further research on the economy-wide impacts of improving rural public
infrastructure on the wider economy. The integration of Cohesion policy related
support would also give a better account of these impacts.

As the number of study areas was limited, the extension of the modelling approach
would enhance understanding of the transferability of the projects findings.

Further research is needed on the way Axis 2 measures are modelled, especially
taking into account that in many EU countries/regions these measures represent a very
significant part of RDP expenditure. In this context, the evaluation of the impacts of
8
RDP environmental public goods measures on rural development presents a
considerable challenge for both researchers and policy makers.
Further Details of the methodological approach
This section provides some additional information on the methods used in the project. Further
details are available from the Project Coordinator on request.
SAM construction and model calibration
The starting point to build the dynamic CGE models is the construction of a Social
Accounting Matrix (SAM) which accounts for all flows in the regional economy at a point in
time. The SAM structure reflects the structure of the underlying model and consists of a set
of accounts covering production activities, commodity balances, flows to and from factors of
production, households and other institutions such as government and the rest of
economy/world. There are a number of key elements in the RURAL-ECMOD SAM accounts
and CGE models which facilitate the simulation of the policy scenarios. The most important
of these are the disaggregation of agricultural sector by farm size and the rural-urban
disaggregation of activities and households which allows the models to account for the spatial
impacts of policy shocks within the study regions.
The models are then “calibrated” to the SAMs, i.e. the initial solution of the models recreates
the SAM values. In this process, both the values from the SAM and information on demand,
production and trade elasticities are used to initialize the model parameters. In addition,
certain assumptions concerning the overall rate of growth of certain key exogenous
parameters including total factor productivity and labour supply are also required.
Simulating RDP Policies
The scenarios are simulated as follows. First, paths for capital stock are generated in the base
run. The changes in the distribution of RDP funds implied by the scenarios are then
calculated and, under the assumption (case study dependent) that these affected certain key
sectors, changes in investment were imposed exogenously. To operationalize this approach,
RDP spending in each region is mapped into investments in specific SAM sectors within the
models. This process requires a range of auxiliary assumptions including how National RDP
schemes mapped to the EU RDP measures, how spend on each measure mapped into
economic sectors, and the commodity composition of sectoral investment.
The exception to this is the simulation of changes in Axis 3 investment in public
infrastructure (Scenario 7). Here the investment is assumed to be non-productive, i.e. only the
extra commodity investment demand is considered. In addition, other impacts have been
considered in certain regions, exogenous changes in tourism demand and increase in
population/labour supply.
Selection of the study areas
The six RURAL ECMOD study areas are NUTS 3 regions with a range of different structural
characteristics. The regions were selected in a two step process, firstly drawing on the
Diversification typology of the TERA-SIAP project (Weingarten et al., 2009), and the OECD
typology (European Commission, 2009). At the second stage cluster analysis was used, a
9
statistical method which using given criteria, groups similar regions into relatively
homogenous groups. The criteria used in this analysis captured differences in Population,
Agricultural Productivity, Farm Structure, Employment Rate, Importance of Food, and
Tourism Sectors, Structure of RDP spending, Importance of the Pillar 1 payments to
regional agriculture.
Table 4 shows how the six conducted case studies fit into TERA-SIAP/OECD territorial
frameworks.
Table 4: Rural Classification of Case Study Areas
OECD Regions
Importance of
agriculture below
average
average importance Economy dependent
of agriculture
on Agr, Forest & Fish
TERASIAP
Types
Rural
Peripheral
Low pluriactivity
Rural
Accessible
Intermediate
Open
Intermediate
Closed
Urban
Open
Urban
Closed
GR252
Arkadia
Avg pluriactivity
High pluriactivity
Low pluriactivity
ITF51
Potenza
Avg pluriactivity
High pluriactivity
CZ064
Jihomoravsky
kraj
Low pluriactivity
UKM50
Aberdeenshire
ES212
Guipuzcoa
Avg pluriactivity
High pluriactivity
AT342
RheintalBodenseege
biet
10
References
European Commission (2010) The CAP towards 2020: meeting the food, natural resource
and territorial challenges of the future, Communication from the Commission to the Council,
the European Parliament, the European Economic and Social Committee and the Committee
of the Regions.
European Commission (2010b) Overview of the CAP Health Check and the European
Economic Recovery Plan - Modification of the RDPs - Some facts and figures"
http://ec.europa.eu/agriculture/healthcheck/index_en.htm accessed 24 January 2011
Lofgren, H, R.L Harris, S Robinson (2002) A Standard Computable General Equilibrum
Model (CGE) in GAMS, Microcomputers in Polciy Research 5, IFPRI, Washington.
http://www.ifpri.org/pubs/microcom/micro5.htm. accessed 24 January 2011
Thurlow J (2008). A Recursive Dynamic CGE Model and Microsimulation Poverty Module
for
South
Africa.
Washington:
IFPRI.
Available
online
www.tips.org.za/files/2008/Thurlow_J_SA_CGE_and_microsimulation_model_Jan08.pdf
accessed 24 January 2011
Weingarten, P., Neumeier, S., Copus, A., Psaltopoulos, D., Skuras, D. and Balamou, E.
(2009). Building a Typology of European Rural Areas for the Spatial Impact Assessment of
Policies : Final Report, Seville: JRS-IPTS.
Contacts
Project Coordinator:
Demetrios Psaltopoulos, University of Patras, Greece
Email: [email protected]
Project Partners:
Demetrios Psaltopoulos ([email protected]) and Dimitris Skuras
([email protected]), University of Patras, Greece
Deborah Roberts ([email protected]) and Euan Phimister ([email protected]),
University of Aberdeen, United Kingdom.
Tomas Ratinger ([email protected]) and Zuzana Bednarikova
([email protected]), UZEI, Prague, Czech Republic
Project Sponsors:
European Commission, Directorate General Joint Research Centre, Institute for Prospective
Technological Studies, SUSTAG action, AGRILIFE unit, Seville.
Fabien Santini ([email protected]), Maria Espinosa Goded
([email protected]) and Sergio Gomez y Paloma ([email protected])
Contract 151408 – 2009 A08 - GR.
11