Abstract
Using the Colombian experience, this paper introduces new empirical evidence about the relationship between fiscal decentralization and regional income disparities. The study has made some advances in the empirical analysis of this relationship. First, a panel data approach is introduced to catch the dynamics of adjustment involved in a fiscal decentralization policy. Secondly, the analysis is based on a country experience rather than a cross-country analysis, so the effects of fiscal decentralization are estimated more objectively than previous research that exhibits cultural, historical, and institutional variation. Finally, other limitations observed in previous work, such as the absence of spatial dependence and sensitivity of the conclusions to the measures of fiscal decentralization used, are addressed in this paper.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
During the 1990s, Colombia was immersed in an intense process of structural reforms. An economic liberalization process introduced trade reforms and opening to direct foreign investment. There were also labor and social security system reforms, monetary and financial reforms giving autonomy to the Central Bank, and deregulation and privatization of the economy.
The source of the fiscal data is the Ministry of Finance. The subnational fiscal database was produced by the Direccion de Apoyo Fiscal, based on the fiscal information of the Contraloria General de la República and Banco de la República. For the municipal level, the database contains information for 804 municipalities that represent 89% of the national population of 2000. For subnational revenues, we take the current incomes, which include tax and non-tax incomes and capital income. That is, we just exclude the value of transfers from another level of government. We do not limit the subnational revenues to local taxes income because these governments have autonomy over two other sources: non-tax and capital income.
These indicators have been built by extracting the national transfers from the national government’s expenditures, and from the subnational governments’ revenues. In this way, both the national government’s expenditures and subnational government’s revenues are net of transfers.
These six territorial entities accounted for approximately 64% of the national GDP, and 53% of the national population.
For a detailed presentation of the Bogotá’s experience, see Gutiérrez and Guzmán (2000).
This measure of income inequality was introduced to run our panel data model at the province level. Aggregating the measure for each department by year, we obtain an estimate of the national income inequality. This aggregated measure follows a trend similar to the traditional measures of inequalities such as Theil index, Gini coefficient, or Sigma convergence. For instance, the correlation coefficient between our measure and those measures are 0.8, 0.97 and 0.95, respectively.
What had been singular in Colombia until the 1960s was that urban growth was quite uniform among the four main cities. However, beginning in the 1960s, urban growth paralleled the experience of other Latin American countries with a pattern of one dominant city. At the end of the period 1960–2000, Bogotá had the major participation in the national GDP; in 1960, this region contributed 15% of the national GDP, whereas, by 2000, this share increased to 22%.
Data have been taken from different sources. The GDP and population information are provided by the Colombian Statistical Agency (DANE), the fiscal information were taken from the two sources mentioned in Section 3, and the export and import data were provided by the Banco de la República. For the purpose of this estimation, the subnational level is composed of municipalities, districts, and departments.
Moran’s index is the most used measure of spatial dependence. When the weight matrix is normalized such that the row elements sum up to 1, the I statistic is \(I = {e^{\prime } We} \mathord{\left/ {\vphantom {{e^{\prime } We} {e^{\prime } e}}} \right. \kern-\nulldelimiterspace} {e^{\prime } e}\), where e is a vector of OLS residuals and W is the standardized weight matrix. The null hypothesis is the absence of spatial dependence (Anselin 1987).
The formula of the LM test for spatial error is \({\text{LM}}_{{{\text{ERR}}}} = \frac{{{\left( {{e^{\prime } W^{*} e} \mathord{\left/ {\vphantom {{e^{\prime } W^{*} e} {{\left( {{e^{\prime } e} \mathord{\left/ {\vphantom {{e^{\prime } e} N}} \right. \kern-\nulldelimiterspace} N} \right)}}}} \right. \kern-\nulldelimiterspace} {{\left( {{e^{\prime } e} \mathord{\left/ {\vphantom {{e^{\prime } e} N}} \right. \kern-\nulldelimiterspace} N} \right)}}} \right)}^{2} }}{{tr{\left( {W*^{2} + W*{}^{\prime }W*} \right)}}}\) and for spatial lag is \({\text{LM}}_{{{\text{LAG}}}} = \frac{{{\left( {{e^{\prime } W*y} \mathord{\left/ {\vphantom {{e^{\prime } W*y} {{\left( {{e^{\prime } e} \mathord{\left/ {\vphantom {{e^{\prime } e} N}} \right. \kern-\nulldelimiterspace} N} \right)}}}} \right. \kern-\nulldelimiterspace} {{\left( {{e^{\prime } e} \mathord{\left/ {\vphantom {{e^{\prime } e} N}} \right. \kern-\nulldelimiterspace} N} \right)}}} \right)}^{2} }}{{{\left[ {{\left( {W*X\widehat{\beta }} \right)}\prime \left( {I_{N} - X{\left( {X^{\prime } X} \right)}^{{ - 1}} } \right.{{\left( {W*X\widehat{\beta }} \right)}} \mathord{\left/ {\vphantom {{{\left( {W*X\widehat{\beta }} \right)}} {\widehat{\sigma }^{2} }}} \right. \kern-\nulldelimiterspace} {\widehat{\sigma }^{2} }} \right]} + tr{\left( {W*^{2} + W*\prime W*} \right)}}}\), where e is the estimated error, \(\widehat{\beta }\) and \(\widehat{\sigma }^{2} \)are the OLS estimations of the estimators and variance, and \(W* = I_{T} \otimes W\). These statistics have an asymptotic χ 2(1) (Anselin et al. 2004).
A weight matrix is a measure for spatial dependence based on the notion of binary contiguity between spatial units. This notion implies that the underlying structure of neighbors is expressed by 0–1 values. They are considered to be contiguous, and a value of 1 is assigned, when two spatial units have a common border of non-zero length (Anselin 1987).
Detailed information about this process can be found in Echavarria et al. (2002).
References
Ahmad E, Baer K (1997) Colombia. In: Ter-Minassian T (ed) Fiscal Federalism in the theory and practice. International Monetary Fund, Washington, District of Columbia
Alesina A, Carrasquilla A, Echavarría JJ (2000) Decentralization in Colombia. Working paper series 15. Fedesarrollo, Bogotá, Colombia
Amstrong H, Taylor J (2000) Regional economics and policy, 3rd edn. Blackwell, Oxford
Anselin L (1987) Spatial econometrics: methods and models. Kluwer, Dordrecht, The Netherlands
Anselin L, Jayet H, Le Gallo J (2004) Spatial econometrics and panel data models. In: Matyas L, Sevestre P (eds) The econometrics of panel data, 3rd edn, Kluwer, Dordrecht (in press)
Baltagi BH (2001) Econometric analysis of panel data, 2nd edn. Wiley, Chichester, New York
Barón JD, Meisel A (2003) La descentralización y las disparidades económicas regionales en Colombia en la década de 1990. Documentos de trabajo sobre economía regional, 36. Banco de la República, Cartagena, Colombia
Bhargava A, Franzini L, Narendranatham W (1982) Serial correlation and the fixed effects models. Rev Econ Stud 49(4):533–549
Birchenall J, Murcia G (1996) Convergencia regional: una revisión del caso Colombiano. Desarro Soc 40:273–308
Bonet J, Meisel A (1999) La convergencia regional en Colombia: una visión de largo plazo, 1926–1995. Coyunt Econ (Bogota) 29(1):69–106
Cerquera D, Jaramillo P, Salazar N (2000) La educación en Colombia: evolución y diagnostico. Boletines de divulgación económica, 6. Unidad de Análisis Macroeconómico, Departamento Nacional de Planeación, Bogotá
Consejo Nacional de Política Social—CONPES (2001) Distribución del situado fiscal y de la participación en los ingresos corrientes de la nación de la vigencia 2001. Documento CONPES social, 51. Departamento Nacional de Planeación, Bogotá
Consejo Nacional de Política Social—CONPES (2002) Distribución del sistema general de participaciones Vigencia 2002. Documento CONPES social, 57. Departamento Nacional de Planeación, Bogotá
Correa P, Steiner R (1999) Decentralization in Colombia: recent changes and main challenges. In: Callahan CM, Gunter FR (eds) Colombia: an open economy? JAI, Lehigh University
Davoodi H, Zou H (1998) Fiscal decentralization and economic growth: a cross-country study. J Urban Econ 43(2):244–257
Echavarria JJ, Renteria C, Steiner R (2002) Decentralization and bailouts in Colombia. Working paper series, Fedesarrollo, Bogotá, Colombia
Gil C, Arzoz PP, Garate MR (2002) Decentralization and regional economic disparities. Proceedings of the 42nd European Regional Science Association (ERSA) Congress, University of Dormund, Germany, 27–31 August 2002
Gutierrez JA, Guzman C (2000) Evolución y estado de las finanzas publicas de Bogotá en la década de los noventa y perspectivas. Estudios de Economía y Cuidad, 14, Secretaria de Hacienda-Alcaldía Mayor de Bogotá
Iregui AM, Ramos J, Saavedra LA (2001) Análisis de la descentralización fiscal en Colombia. Borradores de economía, 175. Banco de la República, Bogotá
Junguito R, Melo L, Misas M (1995) La descentralización fiscal y la política macroeconómica. Borradores de economía, 31. Banco de la Republica, Bogotá
Karl CR (2000) Incidencia del gasto publico en salud 1990–1999, boletines de divulgación económica, 5. Unidad de Análisis Macroeconómico, Departamento Nacional de Planeacion, Bogotá
Kim E, Hong SW, Ha SJ (2003) Impacts of national development and decentralization policies on regional income disparity in Korea. Ann Reg Sci 37:79–91
Meisel A (1993) Polarización o convergencia? A propósito de Cárdenas, Pontón y Trujillo. Coyunt Econ 23(2):153–160
Mora J, Salazar B (1994) Fábula y trama en el relato de la convergencia. Boletín socioeconómico, 27. CIDSE-Universidad del Valle, Cali, Colombia
Qiao B, Martinez-Vasquez J, Xu Y (2002) Growth and equity tradeoff in decentralization policy: China’s experience. Working paper, 02-16. Andrew Young School of Policy Studies, Georgia State University, Atlanta
Rocha R, Vivas A (1998) Crecimiento regional en Colombia: persiste la desigualdad? Revista de Economía del Rosario 1(1):67–108
Rodriguez-Pose A, Gill N (2003) Is there a global link between regional disparities and devolution? Research papers in environmental and spatial analysis, 79. London School of Economics
Sanchez F, Smart M, Zapata JG (2002) Intergovernmental transfers and municipal finance in Colombia (mimeo), Bogotá
Shah A (1997) Fiscal federalism and macroeconomic governance: for better or for worse. The World Bank, Washington, District of Columbia
Shankar R, Shah A (2001) Bridging the economic divide within nations: a scorecard on the performance of regional development policies in reducing regional income disparities. The World Bank, Washington, District of Columbia
Tsui K (1996) Economic reform and interprovincial inequalities in China. J Dev Econ 50:353–368
Wiesner E (2003) Fiscal federalism in Latin America—from entitlements to markets. Inter-American Development Bank, Washington, District of Columbia
Wooldridge JM (2002) Econometric analysis of cross section and panel data. MIT, Cambridge, Massachusetts
Acknowledgements
The author wants to thank Geoffrey Hewings, Werner Baer, Stephen Parente, Maria Teresa Ramirez, Ana Maria Iregui, Armando Galvis, Javier Perez, and Adolfo Meisel for their comments. He also thanks the Banco de la Republica de Colombia for its financial assistance. The author takes responsibility for any errors.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bonet, J. Fiscal decentralization and regional income disparities: evidence from the Colombian experience. Ann Reg Sci 40, 661–676 (2006). https://doi.org/10.1007/s00168-006-0060-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00168-006-0060-z