![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
The Swiss federal reform 2008 (NFA) has been a major undertaking in the recent history of the Swiss federation, with the re-allocation of policy tasks being a key component of the reform. So far, research has focused in particular on the decision-making process of the reform. This paper focuses on the fiscal impacts by asking whether the re-allocation of tasks has changed the vertical distribution of fiscal resources and whether these changes have increased fiscal autonomy of subnational governments. The paper shows that subnational government expenditure as well as non-earmarked federal government grants have increased slightly. At the same time, however, revenue has not increased correspondingly. This has imposed strong limitations on subnational fiscal autonomy and has led to a deterioration rather than an improvement of the situation of subnational governments in federal Switzerland. The paper concludes with suggestions for further research on constitutional change in federal states.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1 Although the NFA primarily regulated the relationship between the central government and the cantons, this paper generally refers to ‘subnational governments’, which encompasses both cantons and municipalities. The reason for this is twofold: First, it builds on literature on subnational fiscal autonomy or decentralization in general, which usually distinguishes between the federal government on the one hand and regional and local governments on the other. Second, empirically, cantons and municipalities should be considered together since the cantons have a great deal of freedom to regulate their relations with the municipalities. As a result, vertical relations between cantons and municipalities can vary significantly between cantons.
2 In the course of the NFA, a new constitutional article was introduced giving the federal government the possibility to declare inter-cantonal contracts to be generally applicable, meaning that individual cantons could – theoretically – be forced to join the contract (Vatter Citation2018, 191–192).
3 Exceptions are two evaluation reports of the Federal Council which looked at the development of the type of federal government grants (earmarked vs. non-earmarked) (Federal Council Citation2010, Citation2014). However, as argued in this paper, this is only one element of fiscal autonomy.
4 Essentially, my third indicator is what Hueglin and Fenna (Citation2015, 170) refer to as the ‘vertical fiscal imbalance’ – namely ‘the mismatch between available revenue and expenditure needs that remains at the subnational level after transfer’ – in distinction from what they call the ‘vertical fiscal gap’, which refers to the mismatch before transfers (Hueglin and Fenna Citation2015, 170; Shah Citation2007, 28). However, in literature, there exist a number of different definitions of the term. Sharma (Citation2012, 100) provides an overview of no less than 16 authors who all define ‘vertical fiscal imbalances’ slightly different (some refer to the mismatch between revenue and expenditure at the subnational level before transfers, others, in turn, refer to both levels of government when assessing the degree of imbalances). To avoid misunderstandings, I use the term ‘mismatch between revenue and expenditure after transfer’, which gives a more precise definition of indicator 3.
5 For all countries in the sample, GFS data is available since 1995, for Switzerland since 1990. A look at the GFS data for Switzerland before 1995 reveals that the value for 1995 was an outlier, which is why I choose 1996–2004 as time period for the analyses.
6 The donor pool comprises federal OECD countries for which data for expenditure and revenue on the different state levels is available for the period from 1996 to 2014.
7 The following explanations on the method are based on Abadie, Diamond, and Hainmueller (Citation2010) and Abadie, Diamond, and Hainmueller (Citation2015). A detailed description of the mathematical foundations can be found in Abadie, Diamond, and Hainmueller (Citation2010, 494–497) and Abadie, Diamond, and Hainmueller (Citation2015, 497–500). The estimations presented in this study are calculated by using the package ‘synth’ in R (Abadie, Diamond, and Hainmueller Citation2011).
8 The synthetic control method was also carried out including the two predictors GDP per capita and unemployment rate. The inclusion of these predictors did not change the findings. Given that Switzerland is an outlier among OECD countries for both variables, the model assigned weights of zero to both variables, meaning that they are irrelevant for the creation of the synthetic case.
9 For the mathematical foundations of the method, see Abadie, Diamond, and Hainmueller (Citation2015).
10 The effect is statistically significant in a times-series cross-sectional regression with the two dummies ‘Switzerland’ and ‘post NFA years’ (countries in the donor pool are weighted according to their weights for the construction of synthetic Switzerland): The coefficients are as follows (std. errors in brackets): post NFA years: 0.02 (0.16); Switzerland: 35.67 (1.41); post NFA years * Switzerland: 2.43 (0.54); Intercept 8.83 (0.41).
11 The negative effect on the ratio of revenue and expenditure in Switzerland is – at least as a tendency – confirmed when estimating a times-series cross-sectional regression with the two dummies ‘Switzerland’ and ‘post NFA years’ (countries in the donor pool are weighted according to their weights for the construction of synthetic Switzerland). The coefficients are as follows (std. errors in brackets): post NFA years: −0.001 (0.004); Switzerland: 0.758 (0.039); post NFA years * Switzerland: −0.027 (0.017); Intercept: 0.254 (0.010). Hence, the interaction term is only just not significant at the 90% level. However, it should be borne in mind that regression is particularly difficult in this case, as synthetic Switzerland consists largely of Austria, which means that practically only the annual figures of two cases are included in the analysis.
12 The comparison of expenditure and revenue is based on internationally comparable GFS data from the IMF. For comparisons within the country (e.g. between cantons), the Swiss Federal Finance Administration also reports expenditure and revenue of the different state levels on the basis of the so-called ‘FS-Modell’. However, due to a methodological change in the recording of expenditure and revenue in 2008, the figures before and after this year are not comparable.