Upload
jessica-gibson
View
216
Download
1
Tags:
Embed Size (px)
Citation preview
1
Quantitative assessment of Quantitative assessment of the impact of the NDP/NSRF 2007-13 the impact of the NDP/NSRF 2007-13
using a using a macroeconomic model for macroeconomic model for
the Czech Republicthe Czech Republic
Project 05/5Project 05/5Ministry of Regional Development – Czech RepublicMinistry of Regional Development – Czech Republic
Prague, 23 November 2006Prague, 23 November 2006Final presentationFinal presentation
GEFRAGEFRA EMDSEMDS IREASIREAS
2
Report: Table of contents, I1.1. Overview about the resultsOverview about the results
2.2. Modelling the Czech EconomyModelling the Czech Economy
3.3. The HERMIN modelling frameworkThe HERMIN modelling framework
4.4. The new HERMIN model for the Czech The new HERMIN model for the Czech RepublicRepublic
4.1 4.1 Some characteristics of the Czech EconomySome characteristics of the Czech Economy
4.2 4.2 Some characteristics model results (equations)Some characteristics model results (equations)
3
Report: Table of contents, II
5. 5. The NSRF EvaluationThe NSRF Evaluation5.1 5.1 How to analyse the Structural Fund interventions How to analyse the Structural Fund interventions
(demand and supply-side effects)(demand and supply-side effects)
5.2 5.2 Implementation into HERMINImplementation into HERMIN
5.3 5.3 Some detailed results Some detailed results
6. 6. The future of the Czech NSRF EvaluationThe future of the Czech NSRF Evaluation
A series of technical annexesA series of technical annexes
4
1. Overview of the results
• During the implementation phase 2007-15 (nine During the implementation phase 2007-15 (nine years instead of the formal seven year SF years instead of the formal seven year SF interventions) two separate mechanisms work:interventions) two separate mechanisms work:– The demand side effect (mainly through programmes of
public and accompanying private investment)– The supply-side effect which work through the build-up
of stocks of infrastructure, human capital and R&D.
• After the termination of the programmes in 2015 After the termination of the programmes in 2015 the demand side effects will vanish rapidly but the the demand side effects will vanish rapidly but the supply-side effects will remain for a longer period.supply-side effects will remain for a longer period.
5
1. Overview of the results, continued• When measuring the impact of the NSRF it is important When measuring the impact of the NSRF it is important
to distinguish betweento distinguish between– Level effect– Growth effect
• The The Level-effectLevel-effect compares the economic outcome compares the economic outcome relative to a baseline-scenario and can be expressed as relative to a baseline-scenario and can be expressed as a percentage or absolute difference from the baseline. a percentage or absolute difference from the baseline.
• The The Growth-effectGrowth-effect compares the different growth rates compares the different growth rates within one scenario with the growth rates in a baseline within one scenario with the growth rates in a baseline scenario.scenario.
• Even if there are Even if there are large level effectslarge level effects the the growth rates can growth rates can be very similarbe very similar. .
6
1. Overview of the results, continued
Cohesion policy 2007-2013: indicative financial allocations (million EUR, 2004 prices)1
Cohesion Fund Convergence Statistical Phasing out Phasing in
Regional competitiveness and employment
België/Belgique 579 1.268 173 2.019 Ceska Republika 7.830 15.149 373 346 23.697 Danmark 453 92 545 Deutschland 10.553 3.770 8.370 756 23.450 Eesti 1.019 1.992 47 3.058 Ellas 3.289 8.379 5.779 584 186 18.217 Espana 3.250 18.727 1.434 4.495 3.133 497 31.536 France 2.838 9.123 775 12.736 I reland 420 261 134 815 I talia 18.867 388 879 4.761 752 25.647 Kypros 193 363 24 581 Latvija 1.363 2.647 80 4.090 Lietuva 2.034 3.965 97 6.097 Luxembourg 45 13 58 Magyarorszag 7.589 12.654 1.865 343 22.451 Malta 252 495 14 761 Nederland 1.477 220 1.696 Österreich 159 0 914 228 1.301 Polska 19.562 39.486 650 59.698 Portugal 2.722 15.240 254 407 436 88 19.147 Slovenija 1.239 2.407 93 3.739 Slovensko 3.433 6.230 399 202 10.264 Suomi-Finland 491 935 107 1.532 Sverige 1.446 236 1.682 United Kingdom 2.436 158 883 5.349 642 9.468
Bulgaria 2.015 3.873 159 6.047 Romania 5.769 11.143 404 17.317
Not allocated 392 392
TOTAL
61.558 177.083 12.521 10.385 38.742 7.750 308.041
CONVERGENCE OBJECTIVE REGIONAL
COMPETITIVENESS AND EMPLOYMENT OBJECTIVE
EUROPEAN TERRITORIAL
COOPERATION OBJ ECTIVE
TOTAL
1 Figures result from the application of the methodology agreed by the European Council in December 2005. They are inclusive of all additional provisions decided by the European Council. Technical assistance at the initiative of the Commission is included in the national allocations. Amounts under the "European territorial cooperation" objective include the contribution of the ERDF to the financing of cross-border and sea-basin programmes on the external borders of the Union. Figures for Bulgaria and Romania are without prejudice to the date of accession of these countries
7
1. Overview of the results, continued
Revised expenditure profile of the NSRF (euro million per year)Revised expenditure profile of the NSRF (euro million per year)
And as % of the NSRF per yearAnd as % of the NSRF per year
8
1. Overviewof the results, continuedEffects of the NSPF on the real Czech GDPEffects of the NSPF on the real Czech GDP
Level effect compared to the baselineLevel effect compared to the baselineas %-differenceas %-difference
Growth rates effect, Growth rates effect, as % differenceas % difference
Moderate differences up to one Moderate differences up to one percent during the implementation percent during the implementation phase and a significant difference in phase and a significant difference in 20162016
Gradually increasing impact on the level of Gradually increasing impact on the level of GDP compared to the baseline-scenario GDP compared to the baseline-scenario (without NSRF), even in 2020 the level about (without NSRF), even in 2020 the level about 2% percent higher then without the NSPR2% percent higher then without the NSPR
9
1. Overview of the results, continued• In 2007 the initial effect on employment In 2007 the initial effect on employment
compared to the no-NSRF baseline is just over compared to the no-NSRF baseline is just over 20 thousand, given the initial low injection by 20 thousand, given the initial low injection by the SFthe SF
• In 2013 this has built up to 196.000 and is still In 2013 this has built up to 196.000 and is still 178.000 in 2015.178.000 in 2015.
• In 2020 total employment is still 29.000 higher. In 2020 total employment is still 29.000 higher.
• The NSRF introduces extra productivity gains The NSRF introduces extra productivity gains of around 1% over the baseline. This is notable of around 1% over the baseline. This is notable since even in the baseline there is are since even in the baseline there is are substantial increases in productivity.substantial increases in productivity.
10
2. Modelling the Czech Economy
• Very limited modeling research of the Czech Very limited modeling research of the Czech economy economy
• Only recently policy-making institutions (MoF Only recently policy-making institutions (MoF or the Central Bank) set-up a structural or the Central Bank) set-up a structural approach to economic modeling approach to economic modeling
• Not all material is in the public domain!Not all material is in the public domain!• Lack of long data series is a serious problemLack of long data series is a serious problem• Models developed have a main focus on:Models developed have a main focus on:
– Monetary and fiscal policy issues and – Long-term growth, structural change and
convergence prospects.
11
3. The HERMIN Modelling Approach A pictorial outline of the model
H E R M I N
[1 ] N a tio na la c c o unting
F ra m e w o rk
[3 ] E xpe nditure
[5 ] M o de l a sinte g ra te d s ys te m
[4 ] Inc o m e
[6 ] B e ha vio ura l a ndide ntity e qua tio ns
M a nufa c tur ing (T )
A gric ulture (A )
N on-m a rke t s e rv ic e s (G )
M a rke t s e rv ic e s (N )U tilite s
O the r m a rke t s e rv ic e s
O utput (G D P )
E xpe nditure (G D E )
Inc om e (G D I)
P riva te c ons um ption
P ublic c ons um ption
Inve s tm e nt
S toc k c ha nge s
E xports
Im ports
G e ne ra l G ove rnm e nt
H e a lth
E duc a tion
A gric ulture
F ore s try
F is hing
G D P
produc tiv ity
P S B R
O utput = E xpe nditure us e dto de te rm ine ne t tra de ba la nc e
(N T S = G D P - G D A )w he re G D A = C + G + I + D S
O utput = Inc om eus e d to de te rm ine
c orpora te profits(IY C = G D P - Y W I)
R e ve nue
E xpe nditure
B orrow ing re quire m e nt
D e bt a c c um ula tion
C orpora te s e c tor
H ous e hold s e c tor
B e ha vioura le qua tionsinf lue nc e dby the ory
Ide ntitie s a dding up,de f in itiona l & c los ure
P ublic s e c tor
P riva te s e c tor
C ons um ption func tion
W a ge ba rga in ing
F a c tor de m a nds
[2 ] O utput
B uild ing (B C )
12
Important economic relationships• OutputOutput: Exposed and sheltered sectors : Exposed and sheltered sectors
• Production inputsProduction inputs: Labour and capital: Labour and capital
• Output pricesOutput prices: Price taking and price making : Price taking and price making
• Wage determinationWage determination: Bargaining and sectoral : Bargaining and sectoral
transmissiontransmission
• Labour supplyLabour supply: natural growth, migration, participation: natural growth, migration, participation
• ConsumptionConsumption: Is there a l: Is there a liquidity constraint?iquidity constraint?
• TradeTrade: : Modelled as a net trade balance (exports minus Modelled as a net trade balance (exports minus
imports)imports)
• GovernmentGovernment: General government expenditure, revenue : General government expenditure, revenue
and balanceand balance
13
What determines sectoral output?
• Manufacturing: Local & external Manufacturing: Local & external demand; competitivenessdemand; competitiveness
• Market services: Local demandMarket services: Local demand
• Agriculture: ProductivityAgriculture: Productivity
• Government: Public employmentGovernment: Public employment
14
What determines production inputs?
• Output Output
• Cost of labour compared to cost of Cost of labour compared to cost of capitalcapital
• Technical progressTechnical progress
15
What determines output prices?
• Manufacturing: World prices, labour Manufacturing: World prices, labour costscosts
• Market services: Labour costsMarket services: Labour costs
• Building and construction: Labour costsBuilding and construction: Labour costs
• Agriculture: Common Agricultural Agriculture: Common Agricultural Policy (CAP)Policy (CAP)
16
What determines wage rates?
• Consumer pricesConsumer prices• Output pricesOutput prices• ProductivityProductivity• Unemployment rateUnemployment rate
17
What determines labour supply?
• Population growth, but more Population growth, but more specificallyspecifically
• Working age population growthWorking age population growth• Participation rateParticipation rate• Net migration flowsNet migration flows
18
What determines household consumption?
• Income, but more specificallyIncome, but more specifically
• Household income, but more Household income, but more specificallyspecifically
• Household disposable incomeHousehold disposable income
• Wealth??Wealth??
19
What determines trade?
• National data on exports and imports National data on exports and imports are availableare available
• Net regional trade balance (NTS) Net regional trade balance (NTS) determined as production (GDPM) less determined as production (GDPM) less regional absorption (GDA)regional absorption (GDA)
NTS = GDPM - GDANTS = GDPM - GDA
20
What determines general government revenue and expenditure?
• RevenueRevenue: Tax bases and tax rates: Tax bases and tax rates
• ExpenditureExpenditure: Policy decisions : Policy decisions (discretionary) as well as policy (discretionary) as well as policy rules (automatic)rules (automatic)
21
Model calibration
• ““Old” EUOld” EU models models: Simple : Simple econometrics econometrics
((OLSOLS)), using data for period 1980-2004, using data for period 1980-2004
• Czech national modelCzech national model: Curve-fitting and : Curve-fitting and
selective selective parameter parameter imposition, using imposition, using
data for period 1995-2004 data for period 1995-2004
• Other approachesOther approaches to calibration: to calibration:
(i) Extend data sample, using pre-1995 (i) Extend data sample, using pre-1995 (or even pre-1989) data?(or even pre-1989) data?
(ii) Panel data from group of countries (ii) Panel data from group of countries (imposing cross-country constraints)(imposing cross-country constraints)
22
HERMIN documentation
1.1. Data-based description of the Data-based description of the economyeconomy
2.2. Theoretical underpinnings of the Theoretical underpinnings of the HERMIN modelling frameworkHERMIN modelling framework
3.3. Calibration of the behavioural Calibration of the behavioural equationsequations
4.4. Model testing: simulations and shocksModel testing: simulations and shocks
5.5. Data appendix, model listing: public Data appendix, model listing: public domaindomain
6.6. Standardised software:Standardised software:
23
Uses of a HERMIN-type model1.1. Preparation of economicPreparation of economic forecasts forecasts2.2. General economicGeneral economic policy analysis policy analysis 3.3. NSRF-related policy analysisNSRF-related policy analysis4.4. DDesignesign/evaluation/evaluation of of nationalnational
industrial strategiesindustrial strategies
24
Preparation of economic forecasts
• Short-term national forecasts (12 months) are widely available
• Medium-term national forecasts (up to 5 years) are very rare
• Long-term economic forecasts of any kind are not usually available
25
4. The new HERMIN model for the Czech Republic
Shocking the model to test its properties
26
27
28
5. The NSRF Evaluation How to analyse the SF interventions
29
Construction phase vs Use phasefor cohesion policy
• During During constructionconstruction phase, there will phase, there will be large demand-side impacts. These be large demand-side impacts. These vanish after completion (i.e., after vanish after completion (i.e., after 2013 for next NSRF)2013 for next NSRF)
• Increased stocks of infrastructure, Increased stocks of infrastructure, human capital and R&D can generate human capital and R&D can generate long-tailed supply-side impactslong-tailed supply-side impacts
• The size of the supply-side impacts The size of the supply-side impacts depend on the appropriateness and depend on the appropriateness and effectiveness of the NDPeffectiveness of the NDP
30
Inserting a CSF/NDP into HERMIN
• The three aggregate elements of a The three aggregate elements of a CSFCSF
(a) Physical infrastructure(a) Physical infrastructure (b) Human resources(b) Human resources (c) Direct aid to productive sector(c) Direct aid to productive sector• Conventional Keynesian expenditure Conventional Keynesian expenditure
and income impacts (role of policy and income impacts (role of policy multiplier)multiplier)
• Additional externality effects onAdditional externality effects on output and productivityoutput and productivity
31
Implementation versus Use phasein evaluation
• During the During the implementationimplementation phase, phase, there will be large demand-side there will be large demand-side impacts. These vanish after completion impacts. These vanish after completion (i.e., after 2013/15 for next NSRF)(i.e., after 2013/15 for next NSRF)
• Increased stocks of infrastructure, Increased stocks of infrastructure, human capital and R&D can generate human capital and R&D can generate long-tailed supply-side impactslong-tailed supply-side impacts
• The size of the supply-side impacts The size of the supply-side impacts depend on the appropriateness and depend on the appropriateness and effectiveness of the NSRFeffectiveness of the NSRF
32
Physical infrastructure: PI• Demand-side impacts (implementation)Demand-side impacts (implementation): :
PI PI IG IG I (total investment) I (total investment) (Keynesian multiplier)(Keynesian multiplier) impact on GDPimpact on GDP
• Supply-side impacts (mainly post-Supply-side impacts (mainly post-implementation)implementation)::
PI PI increased stock of infrastructure increased stock of infrastructure (KPI) (KPI)
boost to output/productivityboost to output/productivity
33
Human resources: HC
• Demand-side impacts (implementation)Demand-side impacts (implementation): :
HC HC Income & Public expenditure Income & Public expenditure Keynesian multiplier Keynesian multiplier GDPGDP
• Supply-side impacts (mainly post-Supply-side impacts (mainly post-implementation)implementation)::
HC HC stock of human capital (KHC) stock of human capital (KHC) boost to output/productivityboost to output/productivity
34
Research and Development
• Demand-side impacts (implementation)Demand-side impacts (implementation): :
R&D R&D I (total investment) I (total investment) (Keynesian multiplier)(Keynesian multiplier) impact on GDPimpact on GDP
• Supply-side impacts (mainly post-Supply-side impacts (mainly post-implementation)implementation)::
R&D R&D increased stock of R&D (KRD) increased stock of R&D (KRD) boost to output/productivityboost to output/productivity
35
A serious methodological challenge
• Ex-ante impact analysis of “yet-to-Ex-ante impact analysis of “yet-to-be-implemented” NDPsbe-implemented” NDPs
• Is the NSRF appropriate? How Is the NSRF appropriate? How effective will be the effective will be the implementation?implementation?
• Strict monitoring and evaluation Strict monitoring and evaluation can help, but do not guarantee can help, but do not guarantee successsuccess
36
Infrastructure, human capital and R&D: interaction effects
• The links between infrastructure, The links between infrastructure, human capital and R&D are difficult human capital and R&D are difficult to measure.to measure.
• A parallel improvement in all three A parallel improvement in all three is probably necessaryis probably necessary
• But we cannot say much about the But we cannot say much about the optimum balance between them optimum balance between them within an NSRFwithin an NSRF
37
Where do the spillover measures come from?
• International literature on the impact of International literature on the impact of PI and HC on output and productivityPI and HC on output and productivity
• Comprehensive ESRI surveyComprehensive ESRI survey• Recent survey of HC impacts by IFSRecent survey of HC impacts by IFS• ProblemsProblems: : very little from cohesion countries!very little from cohesion countries! nothing from CEE countriesnothing from CEE countries no agreement on size of impacts!no agreement on size of impacts! no agreement on level versus growth no agreement on level versus growth
impacts!impacts!
38
Presenting model-based cohesion policy impacts
• Difficult to define an appropriate Difficult to define an appropriate counterfactual baseline scenario.counterfactual baseline scenario.
• Difficult to assign values to the Difficult to assign values to the spill-over (or externality) spill-over (or externality) elasticities to different countries in elasticities to different countries in he absence of empirical research.he absence of empirical research.
• Macro impacts are complex, and Macro impacts are complex, and GDP is an imperfect indicatorGDP is an imperfect indicator
39
5. The NSRF Evaluation
40
41
42
43
44
45
46
47
48
49
6. The future of the Czech NSRF EvaluationSumming up: Summing up:
• the NSRF interventions will have a large impact on the level of the NSRF interventions will have a large impact on the level of GDP during the implementation phase up to 4.5% of GDP until GDP during the implementation phase up to 4.5% of GDP until 2015 compared to a baseline scenario and 2015 compared to a baseline scenario and
• in the long-run the level of GDP will be around 1.5% higherin the long-run the level of GDP will be around 1.5% higher• there will be extra productivity gains and there will be extra productivity gains and • employment will be higher then in the non-NSRF scenario.employment will be higher then in the non-NSRF scenario.
Altogether the SF interventions will have a positive Altogether the SF interventions will have a positive impact on convergence, the size of which will depend impact on convergence, the size of which will depend
on the quality on the interventions.on the quality on the interventions.
50
6. The future of the Czech NSRF Evaluation
• These outcomes are based on a set of specific These outcomes are based on a set of specific assumptionsassumptions– Specific externalities for infrastructure, R&D and
human capital– These externalities are high when the quality of the
interventions are high– The likely impact of other issues like the Single
Market, enhanced FDI etc. are not explicitly modeled. – These linked to the Structural Funds will further
enhance convergence.
51
6. The future of the Czech NSRF EvaluationFuture developments: Future developments:
• In order to use the HERMIN CZ macro-model as an In order to use the HERMIN CZ macro-model as an instrument for policy evaluations, it has to updated on a instrument for policy evaluations, it has to updated on a regular basis. regular basis.
• During the implementation phase, and when the exact During the implementation phase, and when the exact content of the measures is known, a new and improved content of the measures is known, a new and improved classification of funding allocation into PI, HR and APS classification of funding allocation into PI, HR and APS should be done.should be done.
• In the absence of a comprehensive micro-economic In the absence of a comprehensive micro-economic evaluation, the spillover-elasticities have to be assumed evaluation, the spillover-elasticities have to be assumed on bases of the international literature.on bases of the international literature.
52
6. The future of the Czech NSRF EvaluationGains from these improvements: Gains from these improvements:
• As the NSRF is implemented can how we fine-tune As the NSRF is implemented can how we fine-tune the programme (at the Mid-term Evaluation stage) to the programme (at the Mid-term Evaluation stage) to optimise its impact on convergence. optimise its impact on convergence.
• To do this one must develop a micro-evaluation To do this one must develop a micro-evaluation alongside the macro evaluation of the type described alongside the macro evaluation of the type described by this report.by this report.
• The characteristics of micro and macro evaluation The characteristics of micro and macro evaluation techniques are set out in the following slide. techniques are set out in the following slide.
53
Micro (bottom-up)Micro (bottom-up) Macro (top-down)Macro (top-down)
Level of Level of disaggregationdisaggregation
High (individual High (individual projects)projects)
Low (sectoral Low (sectoral aggregates, whole aggregates, whole
economy)economy)
Use of theoryUse of theory Weak (judgemental, Weak (judgemental, CBA)CBA)
Strong Strong (macroeconomics)(macroeconomics)
Model calibrationModel calibration Judgemental/informalJudgemental/informal Scientific(?)/econometricScientific(?)/econometric
Policy impactsPolicy impacts Informal/implicit/Informal/implicit/ranking/some ranking/some quantificationquantification
Formal/explicit/Formal/explicit/quantifiedquantified
Treatment of Treatment of externalitiesexternalities
Limited or ignoredLimited or ignored Included/explicitly Included/explicitly modelledmodelled
54
How can we know if EU cohesion policy is successful?
Integrating micro and macro approaches to the evaluation of Structural Funds
(Bradley, Mitze, Morgenroth & Untiedt, March 2006)
Paper available on: www.gefra-muenster.de
55
56
57
Annex to the Presentation
58
Are HERMIN-based NDP impacts robust and reliable?
• Demand impactsDemand impacts: role of Keynesian : role of Keynesian multiplier (crowding out? Rational multiplier (crowding out? Rational expectations? QUEST versus HERMIN)expectations? QUEST versus HERMIN)
• Supply impactsSupply impacts: role of externality : role of externality (or spillover) effects(or spillover) effects
• Well-designed NSRF: large spillover Well-designed NSRF: large spillover effectseffects
• Poorly-designed NSRF: small spillover Poorly-designed NSRF: small spillover effectseffects
59
Can we give a “point” estimate of NSRF impacts?
• Almost certainly not!Almost certainly not!• But based on additional evidence But based on additional evidence
(CBA, qualitative review of NSRF, (CBA, qualitative review of NSRF, etc.), we can probably suggest etc.), we can probably suggest whether the spillovers are likely to be whether the spillovers are likely to be high or low.high or low.
• As experience accumulates, we can As experience accumulates, we can move towards more precise estimates move towards more precise estimates of the spillover parameters, and more of the spillover parameters, and more precise long-term NSRF impactsprecise long-term NSRF impacts
60
How can convergence become stronger and self-sustaining?
• There must be an initial “cluster” of activities, There must be an initial “cluster” of activities, supported by special incentives and local supported by special incentives and local specialised inputs (trained labour force): role of specialised inputs (trained labour force): role of industrial strategyindustrial strategy
• The increased local availability ofThe increased local availability of skilled workers skilled workers facilitates further growth of cluster: role of facilitates further growth of cluster: role of human human capitalcapital
• Spillovers of information and “learning-by-doing” Spillovers of information and “learning-by-doing” encourage further cluster growth: role of encourage further cluster growth: role of physical physical infrastructureinfrastructure
• National and regional policy stability, and National and regional policy stability, and consensual social partnership ensureconsensual social partnership ensuress efficiencyefficiency is is accompanied by accompanied by equityequity, and facilitates self-, and facilitates self-sustaining regional growthsustaining regional growth
61
Porter’s Diamond of competitive advantage