![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
Despite a large number of empirical studies on the flypaper effect, it remains disputed whether the effect exists and to what extent it is asymmetrical. The flypaper effect suggests that intergovernmental grants tend to result in higher increases in public expenditures than a similar increase in citizens’ private income would have led to. An asymmetrical effect exists when the fiscal response differs depending on whether grants are increased or decreased. By considering political institutions that moderate the effect of intergovernmental grants, this article offers a theoretical explanation that accounts for the mixed empirical evidence. The local response to intergovernmental grants is tested using a reform of the Danish intergovernmental grant scheme in 2007. In line with the expectation, the article finds a strong asymmetrical effect, but more surprisingly, this effect is found both when subnational budget institutions are centralised and when they are fragmented.
Notes
1. Or only in small tax reductions.
2. In this situation, taxes would normally be raised to compensate for the grant reduction.
3. Explanations focus on factors such as budget-maximising public servants (Filimon, Romer, and Rosenthal Citation1982; Niskanen Citation1968) and dead weight loss by the extraction of taxes (Bailey and Connolly Citation1998, 342; Hines and Thaler Citation1995, 221).
4. A few other studies incorporate political factors such as the composition of and bargaining in the local council in the models (Rattsø and Tovmo Citation2002; Tovmo and Falch Citation2002) but do not focus on budget institutions.
5. In analysing expenditures, this article follows among others Gamkhar (Citation2000) and Heyndels (Citation2001). See Stine (Citation1994) and Goodspeed (Citation1998) for examples using tax response to analyse the asymmetric effects.
6. This behaviour can only be sustained in the short run by using savings or taking on debt or in the long run by adjusting taxes. In Denmark it was at the time to a considerable extent possible for municipalities to do all of these things.
7. In addition, a number of smaller pools exist that are allocated on a discretionary basis.
8. Ideally one would have variation across municipalities on the degree of fragmentation in the budget process. Unfortunately, specific data on the fragmentation are not available.
9. The analyses use the structure of the ‘new’ municipalities, which means that the pre-reform measures are adjusted to the ‘new’ municipalities. This procedure is done by adding the values for the ‘old’ municipalities belonging to a ‘new’ municipality. Only a few ‘old’ municipalities were split into two ‘new’ municipalities. For these, the ‘old’ municipality has been split according to the share (based on inhabitant) of the old municipality belonging to each of the two ‘new’ municipalities.
10. Random effects regression was also applied. These are not shown, since the Hausman tests of differences between fixed effects and random effects models indicate that fixed effects should be used.
11. For three municipalities (Copenhagen, Frederiksberg and Bornholm), the exact same data as for all other municipalities could not be obtained on the independent variable ‘grant’ as well as on the variable ‘liquidity’. However, relying on data from Statistics Denmark, measures on these two variables were calculated that resemble the measures used for all other municipalities. Substantively similar results are obtained with and without the three municipalities. To maximise comparability between units, the results are reported excluding the three municipalities.
12. Tests of influential observations indicate one outlier (Læsø) with potentially major influence on the results. An exception in the law regulating the reform implied that Læsø received a much larger grant increase than any other municipality. Since Læsø also raised expenditures radically, this municipality will influence the results significantly. To avoid this, Læsø was excluded from the analysis.
13. The coefficient is 0.016 with a SE of 0.042. Results are not shown.
14. The coefficient is 0.027 with a SE of 0.069. Results are not shown.
Additional information
Notes on contributors
Marie Kjaergaard
Marie Kjærgaard is a PhD fellow at the Department of Political Science and Government, Aarhus University. Her dissertation investigates the role of politics in intergovernmental fiscal relations.