The TERA CGE models: analysing labour migration in diverse regional economies in the EU Euan Phimister (University of Aberdeen, UK)
Dec 28, 2015
The TERA CGE models: analysing labour migration in diverse regional economies in the EU
Euan Phimister (University of Aberdeen, UK)
SAM & General Equilibrium models • Major Part of TERA project • This Presentation and Next Complementary • Aim - motivation, implementation, usefulness
Structure• Background • Case Study Areas• Modelling Approach SAMs and CGE Models• Model Structure • Using Models - labour migration.
Background
TERA: Economic development in remote rural areas
Aims:• Role territorial factors which influence development• review whether existing policies take account of factors• propose new policy interventions.
“The trends and choices that affect rural areas cannot be studied in isolation from what is going on in non-rural areas” (Saraceno, 1994)
Approach• Regional/Local • Modelling within region rural-urban linkages
6 Case Study areas. Reflect different • Economic and Institutional Context• Spatial Scale• Rural-urban relationship
OECD Rural Classification
Case Study Differences
Spatial Scale NUTS3 to NUTS 4/5
Population levels 110K – 400K
Rural Pop Densities - 9.5 - 125 persons/km squared (Finland) –(Italy)
Economic Size 0.5bn - 2bn euros/year
GDP per capita. Developed (UK) – less developed (Latvia)
Rural share of GDP – 5% (Greece) to 60%(Finland)
CZ FIN GR ITA LAT UK
Population
418 481 170 200 142 259 282 020 364 345 115 899
GDP (m Euros)2129.3 2982.8 1524.1 2953.6 592.2 2749.1
Rural Share (%)21.99 58.1 4.28 32.39 41.3 40.5
Urban Share (%)78.01 41.9 95.72 67.61 58.7 59.5
GDP Per Capita (Euros) 5088 16919 10711 10473 1625 23724
Rural GDP Per Capita 4482 14201 14345 6643 1127 15599
Urban GDP per Capita 5289 23047 10593 14470 2361 36731
CASE Study Areas: Summary Statistics
Modelling • Social Accounting Matrices (SAM) - all transactions given point in time• SAM - basis for Computable Equilibrium Model (CGE)
SAM Construction - each study area • Existing secondary sources, e.g. national input-output tables • Primary Data collection • Survey of Households and Business survey, interviews with key
informants
Computable Equilibrium Model (CGE)
Behaviour of representative agents in economy • Producers and Traders – maximise profits • Consumers – maximise their well-being (have demand curves) • Government collects taxes and makes transfers (tax rates and
transfers are exogenously set)
Model Closure rules – assumptions on how markets operate e.g. labour
All transactions in “economy” accounted for.
TERA-CGE Models • IFPRI Standard CGE Model (Lofgren et al) (
www.ifpri.org/pubs/microcom/micro5.htm )• Disaggregation of Accounts allows rural-urban analysis
Level of Disaggregation
CZ FIN GR ITA LAT UK
Activities/Industries 15 52 18 23 33 38
Of which Rural 8 23 9 12 17 19
Commodities 15 28 20 19 15 19
Factors of Productions 10 9 10 10 6 10
Of which Rural 5 5 5 5 3 6
Households 5 8 13 4 8 8
Of which Rural 3 4 6 2 4 4
1 Rest of World, 1 Government sector all case study areas
Production structure
Urban/Rural
Urban/Rural
Local/Regional
Local/Regional
Factors and Households
CGE Model Estimation• Each case study area• Data - SAM plus other literature estimates • Procedure - calibrate CGE models so each CGE
replicates Case study SAM
CGE Model Usefulness• Full Picture of case-study economic transactions• Controlled experiments – what if ?
Example Simple Scenario – labour migration • How different are the effects of large labour
inflow/outflows in case study areas?
Case Study Areas Evidence • Significant Growth Greek, Scottish Study areas• Decline Finnish and Latvian areas
Composition effect ? • Finland - out migration of highly educated people• Scotland- in migrants (skilled) but work in low skilled occupations
Scenario 1 + 10% change in total labour supply all areas
Scenario 2 a) -20% skilled labour category Czech R, Finland, Latvia b) +20% unskilled labour category Greece, Italy, UK
Key Assumptions Each case study area separate labour market Urban-rural labour market integrated within case study areaCapital fixed by sector, Government spending fixed
Scenario 1 +10% change in total labour supply all areas
Aggregate level (GDP) broadly similar effects across case study areas +10% positive impact 5-8% (-10% approximately same negative effect)
Components of GDP - Larger differences
Rural-urban decomposition +10% positive impact Rural GDP effects 2-9% Mostly Rural same or less than Urban effect (except Italy)Largest differences GR, UK
Rural-urban sectoral decomposition +10% positive impactMostly Rural sectoral effect same or less than Urban effect (except Italy)Largest differences GR, UK
Scenario 1 +10% change in total labour supply all areas
% Impact on Real Gross Domestic Product (GDP)
CZ FIN GR IT LAT UK
Private Cons. 6.7 4.3 4.2 8.9 3.9 2.6
Investment 16.2 43.6 15.0 14 3.5 25
Reg Exports 6.8 4.2 4.1 8.6 9.3 7.4
Reg Imports 8.4 12.7 6.4 7.4 4.9 8.1
GDP at Factor Cost 6.5 5.3 4.9 8.2 5.6 5.7
Scenario 1 +10% change in total labour supply all areas % Impact on Real Gross Domestic Product (GDP) at Factor Cost
CZ FIN GR IT LAT UK
Rural 6.4 5.2 1.9 9.0 5.2 3.3
R-primary 6.2 1.3 1.9 6.7 5.5 2.8
R-manufacturing 8.3 6.9 2.9 8.5 8.4 6.1
R-services 4.9 5.7 2.5 9.7 3.8 2.7
Urban 6.9 5.3 5 7.7 5.8 7.3
U-manufacturing 8.6 6.7 8.8 7.8 8.4 8.9
U-services 5.7 4.8 4.2 7.6 4.4 6.9
Scenario 2 a) -20% skilled labour category Czech R, Finland, Latvia b) +20% unskilled labour category Greece, Italy, UK
Areas losing skilled labour
Big differences in overall loss 5-12.6%
Urban areas worst hit
Broadly, impact by sector comparable Areas gaining unskilled labour
Some differences in overall gains 2-4%
No clear pattern whether Urban or rural areas gain most
Differential sectoral impact by rural-urban
Ratio Skilled: unskilled wages increases both losing & gaining areas
Scenario 2 a) -20% skilled labour category Czech R, Finland, Latvia b) +20% unskilled labour category Greece, Italy, UK
% changes in GDP at Factor Cost
-20% skilled +20% unskilled
CZ FIN LAT GR IT UK
Overall -12.6 -5.3 -6.9 4.3 4.8 1.8
Rural -12.2 -4.7 -5.2 2.6 5.2 1.3
R-primary -10.9 -0.7 -7.7 1.4 7.7 1.1
R-manufacturing -15.7 -5.0 -13.4 3.6 7.0 0.8
R-services -9.7 -6.1 -1.4 3.5 3.9 1.5
Urban -13.4 -6.1 -8.0 4.6 4.6 2.1
U-manufacturing -16.4 -5.1 -14.2 8.6 6.9 1.4
U-services -11.4 -6.4 -4.7 3.4 1.5 2.3
Summary & Conclusions
• Modelling Approach SAMs and CGE
• CGE Models capture Case Study areas differences (?)
• Labour migration General – Rural GDP case study area differences
Urban effect often bigger than Rural
Skills mix- differential losses and gains
• Example CGE - What if?
• Further simulations – tailored to specific circumstances of each case study area
• Range simulations envisaged – Tourism, Transport, Agric Policy
Scenario 1 -10% change in total labour supply all areas
% Impact on Real Gross Domestic Product (GDP)
CZ FIN GR IT LAT UK
Private Cons. -6.9 -4.5 -4.3 -9.0 -4.1 -3.1
Investment -17.1 -45.2 -16 -15 -1.4 -29
Reg Exports -7.0 -4.5 -4.3 -8.7 -9.7 -8.8
Reg Imports -8.8 -13 -6.8 -7.6 -5.7 -9.5
GDP at Factor Cost -6.7 -5.5 -5.1 -8.3 -5.9 -6.4