Decentralization as a constraint to Leviathan: a panel cointegration analysis

Résumé du document

Basing fiscal federalism on the government is indeed a study to be reconsidered. This paper reflects on the effects of fiscal federalism. The Leviathan hypothesis (twisted speculation), suggests a negative relationship between fiscal decentralization and government growth. This relationship has been recently enriched by theoretical and empirical research and highlights the importance of distinguishing between grants and self-owned resources to gauge the effects of fiscal decentralization on the size of public sectors. Moreover, several controllable variables such as economic, demographic and political factors have been proved to be empirically relevant. This paper is an improvement on this literature by distinguishing between long term and short term relationships. The methods adopted in this paper are appropriate for panel cointegration techniques.

Extraits

[...] and Marchand, M. (1997), “Fiscal Competition and the Pattern of Public Spending”, Journal of Public Economics 66, 33-53. Hadri K. (2000), “Testing for Stationarity in Heterogeneous Panel Econometrics Journal 148-161. Harris R. and Sollis R. (2003), Applied Time Series Modelling and Forecasting, John Wiley, Chichester. Im K.S., Pesaran, M. H. and Shin, Y. (2003), “Testing for Unit Roots in Heterogeneous Panels”, Journal of Econometrics 115, 53-74. [...]

[...] Several reasons lead us to focus on this nexus. First, the size of the public sector is the dimension on which fiscal decentralization exerts its most direct and immediate impact. This should allow for greater precision in the estimates, a necessary condition to discriminate the long run from the short run nature of the relationship. Secondly, many of the countries that have recently engaged in a process of fiscal decentralization are going through either a transitional dynamics from non market to market based structures of the economy (e.g. [...]

[...] This is not the case if the estimation is done using a standard panel data approach. In those circumstances, there is a reverse of the results with respect to grant and sub-national revenue raising with grants appearing to push up expenditure and sub-national raising of revenue placing a dampener, underlining the approach (but possibly also underlining the difference between different groups of countries with the group examined here being more homogenous than those considered by Rodden). The difference between these estimation techniques is in their approach to dealing with serial correlation: the former taking a non- parametric approach and the latter a parametric approach where lagged first differences are estimated. [...]

[...] Inman, R.P. and Rubinfeld D.L. (1997), “Rethinking Federalism”, Journal of Economic Perspectives 11, 43-64. International Monetary Fund (2001), Government Finance Statistics, Washington, DC. Joulfaian, D. and Marlow, M. (1990), “Government Size and Decentralization: Evidence from Disaggregated Southern Economic Journal 56, 1094- 1102. Kao C. H. (1999), “Spurious Regression and Residual Based Tests of Cointegration in Panel Journal of Econometrics 90, 1-44. McCoskey S. [...]

[...] The results of the estimation of the ECM are given in table 3. It is noteworthy that there was little difference in the results, no matter whether the residuals came from the long-run relationship estimated using the FMOLS or the DOL methods of finding the long-run relationship. Table 3 here What is clear from these results is that there are very few effects on the growth of the sector in the short run other than a very significant and quite quick error-correction of around In order to clarify matters, only a direct discussion will be made of the significant effects. [...]