site-info | contact | login | mail deze pagina
 
    Department of Economics and Econometrics

  Regional Science Groningen

Paul Elhorst

 
 
Home Publications Curriculum Vitae Software Spatial Econometrics
+31 50 3633893/7018
  • Home Staff REG foundation IRIOS Projects Contact
  • Data to estimate spatial panels 

    Download the demo data set on cigarette demand taken from Baltagi's book "Econometric Analysis of Panel Data"(2001). This data set consists of 29 years. Some demo files use a data set of only six years, which can be downloaded here.

    Download the corresponding spatial weight matrix of 46 U.S. states.

    Matlab Software to estimate spatial panels

    *** Updated versions (April 11, 2011) based on:

    Elhorst JP (2010) Spatial Panel Data Models. In Fischer MM, Getis A (Eds.) Handbook of Applied Spatial Analysis, pp. 377-407. Springer: Berlin Heidelberg New York. http://www.springerlink.com/content/u8086626076458v0/

    Elhorst JP (2010) Matlab Software for Spatial Panels. Presented at the IVth World Conference of the Spatial Econometrics Association (SEA), Chicago, June 9-12, 2010. Revised version finished in Spring 2011.

    Download working paper version of the book chapter here and of the revised version of the paper presented in Chicago here. The 2010 updates now also include the bias correction procedure proposed by Lee and Yu (2010) if the spatial panel data model contain spatial and/or time-period fixed effects, the direct and indirect effects estimates of the explanatory variables proposed by LeSage and Pace (2009), and a selection framework to determine which spatial panel data model best describes the data proposed by Elhorst (2010). Note: sar_panel_FE and sem_panel_FE have been improved on 11/4/2011 together with the revision of the Matlab paper.

    Download sar_panel_FE, f_sarpanel, f2_sarpanel to estimate static panel data model (N regions*T time periods) with a spatially lagged dependent variable and fixed effects: 0=pooled model without fixed effects (default, x may contain an intercept), 1=spatial fixed effects, 2=time period fixed effects, 3=spatial and time period fixed effects. The file direct_indirect_effects_estimates is needed to calculate the direct and indirect effects estimates of the explanatory variables. Alternatively, you may use panel_effects_sar.m and/or panel_effects_sdm.m.

    You may also need sar_parse, sar_eigs, sar_lndet and hessian developed by J.LeSage.

    Download sem_panel_FE, f_sempanel, f2_sempanel to estimate static panel data model (N regions*T time periods) with spatial error autocorrelation and fixed effects: 0=pooled model without fixed effects (default, x may contain an intercept), 1=spatial fixed effects, 2=time period fixed effects, 3=spatial and time period fixed effects.

    You may also need sem_parse, sem_eigs, sem_lndet and hessian developed by J.LeSage.

    Download sar_panel_RE, f_resar, f2_resar to estimate static panel data model (N regions*T time periods) with a spatially lagged dependent variable and spatial random effects.

    Download sem_panel_RE, f_respat, f2_respat to estimate static panel data model (N regions*T time periods) with spatial error autocorrelation and spatial random effects.

    Download file to print the output of these panel data models.

    Download demoLMsarsem_panel to test for a spatially lagged dependent variable or spatial autocorrelation in a spatial panel data model using (robust) LM tests based on spatial panel data. You may also need demean. Note: Demofile has been improved on 11/4/2011 together with the revision of the Matlab paper.

    Download demopanelscompare of the different panel data models, and to test for the joint significance of spatial fixed or random effects as well as to compare spatial fixed and random effects models using Hausman's specification test. Note: Demofile has been improved on 11/4/2011 together with the revision of the Matlab paper.

    Also download a Notepad file Matlab-paper-results which gives the results when running the file demopanelscompare.

    End updated versions (April 11, 2011) ***

    Code Regional Science and Urban Economics Paper 2012: Download the Matlab-file Statregion to determine the stationariy region of a second-order spatial lag model or second-order polynomial in two spatial weights matrices. Code for third-order spatial lag model is also provided. Reference: Elhorst J.P., Lacombe D.J., Piras G. (2012) On Model Specification and Parameter Space Definitions in Higher Order Spatial Econometrics Models. Regional Science & Urban Economics 42: 211-220.

    Code Journal of Regional Science paper 2011: Download the Matlab-file les_sar_cost differences and a rar-file with additional Matlab-files needed to run this file to estimate a Linear Expenditure System with endogenous interaction effects and cost differences among jurisdictions based on cross-section data. Reference: Allers M.A., Elhorst J.P. (2011) A simultaneous equations model of fiscal policy interactions. Journal of Regional Science. 51: 271-291.

    Code Journal of Regional Science paper 2009: Download sarregime_panel, f_sar2_panel, f2_sar2_panel, prt-spreg and demo-file to estimate spatial panel data model (N regions*T time periods) with two regimes (two spatial lags with different coefficients) and with spatial and/or time-period fixed effects. Reference: Elhorst J.P., Fréret S. (2009) Evidence of political yardstick competition in France using a two-regime spatial Durbin model with fixed effects. Journal of Regional Science 49: 931-951.

    Code Economics Letters paper 2008: Download serialspatial, f_serialspat, f2_serialspat and demo-file to estimate pooled regression model (N regions*T time periods) with serial and spatial error correlation. Reference: Elhorst J.P. (2008) Serial and spatial autocorrelation. Economics Letters 100: 422-424. http://dx.doi.org/10.1016/j.econlet.2008.03.009.

    Code International Tax and Public Finance paper 2005: Download matlab-file to estimate spatial econometric model with two regimes in the spatially lagged dependent variable (Note: Use this file for cross-section data, for spatial panel data see Journal of Regional Science paper 2009). You also need f_sarpaul, f2_sarpaul, f_sarpar2 and f2_sarpar2. Reference: Allers M.A., Elhorst J.P. (2005) Tax mimicking and yardstick competition among local governments in the Netherlands. International Tax and Public Finance 12: 493-513.

    Code Geographical Analysis paper 2005: Download matlab-file to estimate dynamic panel data model (N regions*T time periods) including a serially lagged dependent variable, regional fixed effects and spatial error autocorrelation. Reference: Elhorst J.P. (2005) Unconditional maximum likelihood estimation of linear and log-linear dynamic models for spatial panels. Geographical Analysis 37: 85-106.

    Code Regional Science & Urban Economics paper 2010: Matlab-file to estimate dynamic panel data model including a serially lagged dependent variable, a spatially lagged dependent variable and regional fixed effects is in preparation.

    Go to top of page

    Mail to ... webmaster
    Last update: January 19, 2012