Giancarlo Lutero, Paola Pianura and Edoardo Pizzoli WYE CITY GROUP On statistical on rural development and agriculture household income Rural Areas Definition for Monitoring Income Policies: The Mediterranean Case Study Rome, june 2009 – FAO Head-Quarters
Outlines The Mediterranean region: political subdivisions and data available Rural-Urban classifications The Panel model Results Concluding remarks and future developments WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters
The Mediterranean Region WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters
The Mediterranean Region WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Political subdivisions: 24 countries; 8 members of European Union (EU), 2 city-states (Gibraltar, Monaco) and 3 countries with a limited political status: Gibraltar under the sovereignty of the United Kingdom, North Cyprus recognised only from Turkey and Palestinian Territory occupied by Israel Economic differences among countries:
The Mediterranean Region WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters
The Mediterranean Region WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Data available: Dishomogeneous in different countries (different variables and frequency) Sources (United Nations, World Bank, FAO, EUROSTAT, CIA and national statistical offices) Missing data for southern Mediterranean countries, Balkan countries and city states Annual Frequency Sample
The Mediterranean Region WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters List of variables: VariableDefinition gdppcGross Domestic Product (GDP) per-capita (current US$) gcf_pcGross capital formation (% of GDP) electric_powerElectric power consumption (kWh per-capita) energy_use_kgEnergy use (kg of oil equivalent per-capita) agricultural_laAgricultural land (% of surface area) for_densityForest density (forest area over surface area) primary_completPrimary completion rate, total (% of relevant age group) mobile_and_fixeMobile and fixed-line telephone subscribers (per 100 people) internet_usersInternet users (per 100 people)
The Mediterranean Region WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Summary statistics: VariableMeanMedianMinimumMaximum gdppc12,899.26, ,670.0 Electric_power3,477.23, ,944.6 Energy_use__kg1,987.11, ,551.1 pop_density1, ,769.2 for_density gcf_pc248, , ,664.91,477,000 Primary_complet Mobile_and_fixe Internet_users agricultural_la VariableStandard DeviationC.V.SkewnessEx. kurtosis gdppc14, Electric_power2, Energy_use__kg1, pop_density3, for_density gcf_pc290, Primary_complet Mobile_and_fixe Internet_users agricultural_la
Rural-Urban Classifications WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Several territorial classification variables calculated on available data Criteria: 1.Single indicator (population density is the default indicator) 2.Two combined indicators (population and agricultural density) 3.Multivariate clustering (two or three clusters) Warning: no political or administrative area subdivision is purely urban or rural (i.e. distance of probability)
Rural-Urban Classifications WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters List of classification variables VariableDefinition Rural_urban2Composite indicator 2*: real continuous number between 0 (purely urban) and 1 (purely rural) Rural_urban3Composite indicator 3**: real continuous number between 0 (purely urban) and 1 (purely rural) Agr_forAgricultural and forest land (% of surface area) Rural_urban21Binary variable: 1= Composite indicator 2*>0.5 (rural); 0=otherwise (urban) Clus12Cluster analysis 1: 1=rural, 0=urban Clus22Cluster analysis 2: 1=rural, 0=urban Clus23Cluster analysis 2: 2=rural, 1=intermediate, 0=urban Clus32Cluster analysis 3: 1=rural, 0=urban Pop150Binary variable: 1=Pop_density<150 (rural), 0=otherwise (urban) Pop200Binary variable: 1=Pop_density<200 (rural), 0=otherwise (urban) Pop250Binary variable: 1=Pop_density<250 (rural), 0=otherwise (urban) Pop_densityPopulation density (total population over surface area)
The Panel Model WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Fixed effects estimation: Random effects estimation:
Results WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters The best starting model: Fixed-Effects Estimates. 192 observations. 24 cross-sectional units. Time-series length = 8. Dependent variable: gdppc CoefficientStd. Errort-ratiop-value const < *** gcf_pc < *** *** indicates significance at the 1 percent level Mean of dependent variable = Standard deviation of dep. var. = Sum of squared residuals = e+008 Standard error of the regression = Unadjusted R 2 = Adjusted R 2 = Degrees of freedom = 167 Durbin-Watson statistic = Log-likelihood = Akaike information criterion = Schwarz Bayesian criterion = Hannan-Quinn criterion = Test for differing group intercepts: Null hypothesis: The groups have a common intercept Test statistic: F(23, 167) = with p-value = P(F(23, 167) > ) = e-089
Results WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Fitted and Actual Plot by Observation Number (best Fixed effects model)
Results WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters The random effects estimation: Selected Models in Order of Efficiency (from left to right) VariablesModel 3Model 4Model 12Model 10Model 8 Common constant 1.388e+04** (5384) 2.888e+04** (2434) (1171) 3902 (3432) 2060 (3083) Electric_power2.471** (0.4425) 1.361** (0.2552) 1.317** (0.3098) 2.206** (0.4633) 2.152** (0.4770) Gcf_pc ** ( ) ** ( ) ** ( ) ** ( ) ** ( ) Primary_complet-1410* (716.7) -1385** (628.6) -1456** (626.7) -1594** (724.3) -1475** (730.9) rural_urban e+04** (6704) rural_urban e+04** (2277) pop_density7.602** (0.7281) pop ** (3168) clus * (2822)
Results WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters The best final model: Random-Effects (GLS) Estimates. 168 observations. 21 cross-sectional units. Time-series length = 8. Dependent variable: gdppc CoefficientStd. Errort-ratiop-value const ** Electric_power < *** gcf_pc < *** Primary_complet * rural_urban *** * indicates significance at the 10 percent level ** indicates significance at the 5 percent level *** indicates significance at the 1 percent level Mean of dependent variable = Standard deviation of dep. var. = Sum of squared residuals = e+009 Standard error of the regression = 'Within' variance = e+006 'Between' variance = e+007 theta used for quasi-demeaning = Akaike information criterion = Schwarz Bayesian criterion = Hannan-Quinn criterion = Breusch-Pagan test - Null hypothesis: Variance of the unit-specific error = 0 Asymptotic test statistic: Chi-square(1) = with p-value = e-099 Hausman test - Null hypothesis: GLS estimates are consistent Asymptotic test statistic: Chi-square(4) = with p-value =
Results WYE CITY GROUP Rome, june 2009 – FAO Head-Quarters Fitted and Actual Plot by Observation Number (best Random effects model)
Concluding remarks and future developments Roma, 23 giugno 2009 Results highlight a cross-sectional heterogeneity among the Mediterranean countries but the diagnostic analysis and fitting show that a common model for the available data is a satisfactory solution Several rural-urban classification variables are significant in this panel data approach A composite indicator, such as a combination of population density with agricultural density (i.e. rural_urban3 in this paper), undoubtedly improve per- capita income explanation WYE CITY GROUP
Roma, 23 giugno 2009 References Agresti, A. (2002) Categorical Data Analysis, John Wiley & Sons, 2nd edition Baltagi B. (2008) Econometric Analysis of Panel Data, John Wiley & Sons, 4th edition FAO (2007) Rural Development and Poverty Reduction: is Agriculture still the key?, ESA Working Paper No , Rome Pizzoli E. and Xiaoning G. (2007a) How to Best Classify Rural and Urban?, Fourth International Conference on Agriculture Statistics (ICAS-4), Beijing, UNECE, FAO, OECD and World Bank (2005) Rural Household’s Livelihood and Well-Being: Statistics on Rural Development and Agriculture Household Income, Handbook, UN, New York,