Table 4.5 should be interpreted with some caution. If overall revenues exceed overall public expenditures or vice versa, a vertical imbalance cannot be adequately expressed by calculating the ratios of total (regional and local) revenues and total (regional and local) expenditures. Yet, if we assume that regions aspire to balanced budgets, total revenue and expenditure levels for any given period of time should be of comparable magnitude.
The table displays VFI of considerable size. The level of VFI is high in Belgium, but in Switzerland the combined revenue resources of the cantons and municipalities exceed the already comparatively high level of regional and local expenditures. The level of VFI seems low in Germany as well, albeit that the German regions are primarily financed by means of so-called shared tax revenues. Conversely, the limited fiscal autonomy of the British regions (and municipalities) is in tune with the embryonic or only partial regional- ization of their state. In spite of the significant policy autonomy of the Belgian and Spanish regions, they have remained largely dependent upon central grants.
most mobile tax base; (2) generate unevenly distributed revenuesbecause they are linked to tax bases with an unevenly distributed regional basis – for instance, a tax on oil in a federation with only one or a few oil-producing regions; and (3) are highly elastic, that is, are most sensitive to changes in income. These taxes are best suited for central stabilization objectives because their decentralization could spark acute regional budget deficits (and subsequent deficit spending) in the event of a sharp economic downturn (Ter-Minissian 1997: 9).
Usually, the following tax bases are ranked in decreasing orderofmobility:
capital or corporate income; personal income; sales or value added (VAT);
and finally property. Put differently, capital and income taxes are best suited for centralization; VAT and sales taxes are suited for decentralization either in part or in their entirety; property taxes are ideal regional (or often local) taxes. On the basis of the elasticityof tax revenues, income and cor- porate taxes qualify best for centralization whereas revenues of property or natural resources are less dependent upon fluctuations in the economic cycle.
The suitability of certain taxes for decentralization is also influenced by thecostof their administration. For instance, on the basis of the mobility of the tax base, sales taxes and VAT are equally well suited for decentralization.
However, for efficiency reasons VAT is best levied centrally, whereas sales taxes can be more easily decentralized. This is so because sales taxes are single-stage taxes, that is, they are levied only once, when the consumer buys goods that are subject to taxation. In contrast, VAT is a ‘multi-stage tax’, charged at different stages of the production process. Therefore, the admin- istration of VAT requires that the information concerning the taxation of goods at each phase of the production and distribution process is fully and efficiently provided. This criterion is more easily fulfilled if only one tax agency is preoccupied with the tax.
Furthermore, the factors that determine whether a tax is suitable for decentralization vary with the sociocultural context. For instance, in Belgium, levels of interregional personal mobility are low because of the presence of an important linguistic fault line (Murphy 1988). Citizens who live in Flanders (and thus are Dutch-speaking) are not likely to move to Wallonia and (until recently also) to Brussels (where most citizens speak French).
Walloons are not likely to relocate to Flanders. Consequently, if personal income taxes were regionalized, Belgians would not be easily persuaded to move to the region with the lowest personal income tax rates (i.e. assuming that the tax is levied on the basis of where a person lives, not works). The cost of acquiring a profound knowledge of the other national language may be too high even if it would lead to tax credits. In this respect, the Belgian context is quite different from the American, where citizens are more easily persuaded to ‘vote with their feet’ (Tiebout 1956) by moving to the region which offers them the best tax/service ratio.
5.2. Why not everyone favours a decentralized tax structure:
Keynesians versus monetarists
In the past decades Western policy-makers in federal or non-federal states alike have relied on two major philosophies for governing the economy. We can simplify them as Keynesianism and monetarism (Barrat Brown 1990;
Braun et al. 2003).
Keynesianism prescribes that, in the event of an economic downturn, governments should be allowed to raise their expenditure levels and lower taxes, seeking to stimulate aggregate demand(consumption). Public investment combined with lower taxes should increase consumption levels and eventu- ally lower unemployment rates. Keynesian strategies initially cause deficit spending and require governments to borrow money on the market.
However, in the medium to long term, the economic growth which these policies generate should reduce public debts overall.
In contrast with Keynesianism, monetarist or neo-classical strategies put the objective of deficit reduction and inflation control first. They focus on thesupply-side of economics, that is, the creation of an environment that stimulates entrepreneurs to invest and create employment. Keynesian policies are not seen as an appropriate response to economic crises; they rather aggravate them. Like Keynesians, monetarists favour tax reductions, but they should be accompanied by cut-backs in public expenditures levels.
In addition, governments should stimulate wage moderation and labour flexibility, enabling private firms to increase profits and thus generate additional employment.
At present the monetarist option finds stronger resonance among the group of economic policy-makers in the Western world. For instance, in the years following Germany’s unification, rising social security expenses resulting from record unemployment figures pushed up public budget deficits to unprecedented heights. Should German governments still adhere to Keynesian economic policies, as they were in the 1960s, these budgetary shortfalls would not be considered as inherently problematic. Keynesianism lost much of its attractiveness in the 1970s, when increasing debt levels and a towering inflation were not offset by economic growth and higher employ- ment levels. In fact (in theory), the EU’s Growth and Stability Pact does not allow the German public debt to exceed certain levels. Therefore, the German Social Democratic Schröder government has approved some painful structural reforms in labour and social policy instead of leading public expenditure to much higher levels.
Whether or not a state is federal affects its ability to pursue the Keynesian or monetarist philosophies.
First, problems may arise when not all governments of a federation stick to similar economic philosophies. In theory, a federal government which favours a monetarist policy may coincide with one or several regional gov- ernments which prefer Keynesian demand strategies or vice versa. Central
governments which pursue monetarist policies may wish to establish some rules that not only force their own budgets but also that of the regional gov- ernments into equilibrium. Similarly, they may wish to constrain regional borrowing policies that lead to regional deficit spending, for instance, by applying the so-called ‘golden rule’. This rule requires that borrowing should be used to finance public investment projects (Ter-Minassian 1997). Regional governments which prefer a less restrictive economic philosophy may object to such regulatory interference from above. In some federal states, central governments may not even be authorized to prescribe such rules in the first place. Conversely, central governments which prefer a Keynesian economic policy may seek to strengthen their policy capacity by coordinating public investment projects and increasing the overall size of redistributive pro- grammes. Such strategies may be vetoed by regional governments which defend more orthodox economic policies and consider the centre’s ambi- tions to increase redistribution and to coordinate policies as a means to cur- tail their political autonomy.
Second, a central government’s choice for a Keynesian or monetarist approach affects its view on whether it should constrain regional economic and fiscal autonomy. Public economists who adhere to a monetarist philo- sophy believe that public economies are best served with limited governments.
Therefore, their main concern is to prevent central and regional governments from overspending, and to keep the level of central redistributive payments within certain limits. In this sense, monetarists do not dispute that the central level is more suitable for redistribution, but they prefer not to see too much of it. Central governments which support a supply-side philosophy may also not seek to intervene in regional taxation all that much. We would even expect them to defend regional tax autonomy. Regional tax autonomy would strengthen links between the regional governments and their electors and gen- erate regional policies that best suit local market conditions. In contrast, by sep- arating taxation from spending autonomy, regional citizens would lose an important link between the regional programmes from which they benefit and the price (taxes) which they are willing to pay for them. In the view of some public economists such a lack of transparency inevitably leads to overspending or inefficiencies. Proponents of fiscal decentralization point at the so-called fly- paper effectwhich results from a lack of regional fiscal autonomy. The flow of central grants may not be used efficiently: too much of the money may go into extending regional public expenditure programmes, too little will be used as a
‘tax subsidy’. Money sticks with the regional governments which receive it, instead of flowing to the citizens for which it is meant (Ahmad and Craig 1997:
82–3). If regional voters were able to decide for themselves on the use of these expenditures, they might opt for tax cuts, instead of more programme spend- ing. In this sense, the flypaper effect is the consequence of distorting the link between the paymaster (central government) and the corresponding electorate (citizens living in the region).
Conversely, central governments which adhere to a Keynesian strategy are keener on centralizing tax-raising powers. Part of the central tax revenues can be used to co-finance regional expenditures. Regions may lose some of their autonomy, not only because they are dependent upon federal income for financing a part of their expenditure programmes but also because cen- tral governments may specify how they want their money to be spent.
Proponents of a strong redistributive component argue that the long-term benefits of redistribution offset its short-term disadvantages. Furthermore, they argue that allocative efficiency would be undermined if regions were to engage in a downward spiral of competition, seeking to attract private investment by minimizing social expenditures and reducing regional tax rates. Federal states with considerable patterns of interregional migration may witness a movement of high-skilled labour to regions that provide the lowest income tax rates. Conversely, the unemployed may flock to regions offering the most generous social benefits. Facing considerable budget deficits these ‘welfare magnets’ may be forced to cut back social expenditures and bring their tax levels down to that of the economically stronger regions.
For this reason, the financing of these redistributive programmes which operate as an indirect form of territorial equalization (because they are largely based on individual rather than regional need) should be kept under central control (Ter-Minissian 1997).
Prevailing economic philosophies may change in time. At present, concepts such as ‘market-preserving’ or ‘competitive’ federalism which proclaim limited government, orthodox budgetary policies and inter- regional competition coupled with regional tax autonomy find consider- able resonance among policy-makers (for academics supporting such forms of federalism, see in particular Weingast 1995). Such philosophies square better with the prevailing monetarist ideas of our time. For instance, in Germany, the relative fashion of concepts such as ‘competi- tive federalism’ has been exemplified by statements from some political leaders representing the resource-richest Southern regions, in particular Bavaria, Baden-Württemberg and Hesse (for an overview and discussion, see Ziblatt 2002; Jeffery 2002; Renzsch 2002). However, while German fed- eralism may be moving into a slightly more competitive direction, much more ‘competition’ is needed to entrust its regions with the level of fiscal autonomy that the Swiss (or beyond the scope of this analysis also US or Canadian) regions have enjoyed for several generations. In this sense, the current tendency in most West European federal states other than Switzerland may be to decentralize fiscal powers and to foster some degree of interregional competition, but the platform upon which these policies develop may vary substantially from one state to another. Furthermore, existing policies may be hard to reverse, when vested-interest groups or political parties can utilize veto points to prevent structural fiscal reform (Tsebelis 2002; Pierson 2004).
Overall, if we take the assignment of tax powers between the federal and regional policy levels as a rough indicator for which economic philosophy has left its strongest mark in Western Europe, it is still the Keynesian, not the monetarist. In fact, with the partial exception of Switzerland and the Basque Country, personal income, corporate income and the most important sales taxes are controlled by central governments. Consequently important vertical fiscal imbalances have emerged, making the regional governments partially dependent upon the allocation of these surplus resources.
6. Vertical and horizontal fiscal imbalances and