Small-area estimation of income statistics Coro Chasco Yrigoyen Ana López García Título Instituto Lawrence R. Klein Department of Applied Economics Universidad Autónoma de Madrid C.Chasco & A.López. Auton.Univ.Madrid
Small-area estimation methods Direct method Sectorial employment matrices Huge information database, not always very precise Cannot reflect the underground economy nor non-wages earnings Indirect estimations Relation between income and socio-economic variables Variables available for all the small-area units Don’t require excessively ardous proceedings C.Chasco & A.López. Auton.Univ.Madrid
Availability of statistical data Official data: INE data Supplies provincial income data in the Regional Accounts, for the period of 1995-2000 Socio-economic indicators Related to income Exogenous variables in a multivariate regression model Available for both the provincial and the small-area level C.Chasco & A.López. Auton.Univ.Madrid
The process Aggregate –ecological- level (provinces) Dissagregate –microdata- level (municipalities) Ecological inference or spatial prediction C.Chasco & A.López. Auton.Univ.Madrid
Spatial extrapolation problems MAUP: some statistics can vary in quantity or even in sign! Spatial effects can also be different in both aggregate and disaggregate level. Economic macro and micro theories are not always coincident. C.Chasco & A.López. Auton.Univ.Madrid
Spatial extrapolation solutions Added information about income distribution in different geographic levels. Good selection of explicative variables Exploratory spatial data analysis can detect spatial spillovers and spatial regimes Spatial econometrics models can specify, estimate and test spatial effects correctly C.Chasco & A.López. Auton.Univ.Madrid
Spatial extrapolation process The basis: theoretical principles Selection and manipulation of explicative data Exploratory spatial data analysis Econometric estimation of the ecological model Spatial extrapolation of small-area data Analysis and validation (?) C.Chasco & A.López. Auton.Univ.Madrid
Theoretical principles Explicative variables of household income in small areas 1. Educational qualifications 5. Locacional factors 6. Historical factors 7. Sex and age 8. Public sector investment 2. Professional category 3. Economic activity 4. Employment Household income = wages + non-wages earnings – – taxes – social security contributions C.Chasco & A.López. Auton.Univ.Madrid
Selection of household income explicative variables 1. Consumption/saving indicators Telephone lines Cars Total electricity use Bank offices Average price for a house/m2 2. Employment Unemployment rates Occupation rates Activity rates Professional category 3. Economic activity Retailing outlets Tourism outlets Industry outlets 4. Others Educational qualifications Distance to service centres C.Chasco & A.López. Auton.Univ.Madrid
Exploratory spatial data analysis Periphery Core Periphery “Wealthy” Spatial autocorrelation Spatial heterogeneity: spatial regimes “Poor” C.Chasco & A.López. Auton.Univ.Madrid
Confirmatory spatial analysis Estimation and testing of spatial effects in spatial regression models New specifications New estimation methods OLS C.Chasco & A.López. Auton.Univ.Madrid
C.Chasco & A.López. Auton.Univ.Madrid Spatial model Spatial lag C.Chasco & A.López. Auton.Univ.Madrid
Spatial extrapolation C.Chasco & A.López. Auton.Univ.Madrid
Analysis and validation C.Chasco & A.López. Auton.Univ.Madrid