The (non)impact of revenue decentralization on fiscal deficits: some evidence from OECD countries†

†The views expressed in this article are those of the author and should not be attributed to the International Monetary Fund.

Abstract

Recent studies examining the relation between fiscal decentralization and consolidated government fiscal balances generally have not taken proper account of the extent of the independent taxing powers available to sub-national governments and thus have overstated the degree of effective revenue decentralization. Results from a panel regression study of 19 OECD member countries suggest that when the measure of fiscal decentralization is limited to the revenues over which sub-national governments have full autonomy, its impact on fiscal balances is not statistically significant. Accordingly, when accurately measured revenue decentralization appears to have had no negative impact on fiscal discipline.

Notes

†The views expressed in this article are those of the author and should not be attributed to the International Monetary Fund.

¹ Specifically, the OECD used a classification of sub-national tax revenues ranging from (a) where the sub-national government can set both the tax rate and the tax base, to (e) where the central government sets both the base and the tax rate. Tax sharing schemes are divided into four categories from (d.1) where the sub-national government can determine the revenue split, to (d.4) where the national government can decide the revenue split. The countries included in the study were Austria, Belgium, Czech Republic, Denmark, Finland, Germany, Hungary, Iceland, Japan, Mexico, Netherlands, New Zealand, Norway, Poland, Portugal, Spain, Sweden, Switzerland and the United Kingdom.

² Data on the share of sub-national government tax revenue in total tax revenue is from the annual OECD annual publication Revenue Statistics. The data on the revenues from the tax base and tax rates over which sub-national governments have full autonomy it is constructed from OECD (1999), as described in the text. The sample period was ended in 2000 because of fiscal reforms in the late 1990s that changed the degree of autonomy over revenues in several of the countries, increasing it in some and reducing it in others (Joumand and Kongsrund, 2004).

Fig. 1 Total and discretionary revenue of sub-national governments, 1982 to 2000 (In percent of national and sub-national tax revenue)

³ Data for the growth of real GDP per capita and government size are from the Penn World Table (version 6.1); fiscal balance and inflation data is from the OECD's main economic indicators database.

⁴ Country dummy variable in OLS estimates of the pooled data with a common constant were highly significant, also indicating the existence of country-specific effects. The choice of the fixed country effects was also based on the Haussman (1979) specification test (chi-square) which consistently favored fixed over random effects for each equation.

⁵ In fact, the correlation coefficient for GOV, OPEN and INF are very low (0.1–0.27), suggesting that multicollinearity is not a problem in the regression results.

Table 1. Panel least squares results for overall fiscal balance (fixed effects)^a

⁶ In developed countries, tax average collection lags – which measure the moment they are actually paid to the fiscal authority – vary from 1 month in some cases and for particular sources of taxation (e.g. taxes that are withheld at source) to 6 and 10 months in other cases (e.g. indirect taxes). In developing countries, in contrast, the share of revenue generated by taxes withheld at source is small and taxes (e.g. import duties and excises) are often levied at specific rates). In such conditions, an increase in the inflation will bring a fall in conventional tax revenue. See Agénor and Montiel (1999), chapter 5 for a discussion of the issues, including of the ‘Olivera-Tanzi effect’.