Introduction

Federal countries, like all other OECD countries, have been under strain during the last 15 years or so because of soaring budget deficits and globalization pressures. This has led to a number of reforms or reform attempts reorganizing the way policy competencies are distributed, revenues are shared, policy strategies are developed, and democratic accountability is incorporated. Despite the obvious differences in the outlook of reforms, efficiency considerations based on new public management ideas as well as on the economic theory of federalism, seem to have inspired policy-makers about the ways and means to revise federal governance and public service delivery. Although most political scientists are aware that new public management has brought fundamental changes to the delivery of public services and the organization of public bureaucracy in general, there are almost no studies from a comparative perspective that endeavour to discuss the influence of this framework and how its main objective, efficiency, has changed the working of federalism, though some case studies of individual countries deal with the influence of new public management on federal organization¹ in general and intergovernmental relations² in particular.³ The pertinence of efficiency as a point of reference in recent federal reforms is the point of interest in this article.

Usually, one distinguishes between different types of federalism, in particular between a 'dual type of federalism' or 'separate jurisdictions' (Wheare, 1967, 139) on the one hand, and 'shared jurisdictions' (Elazar, 1962) on the other. Börzel and Hosli (2003) speak — and we will use this expression — of the 'power separation' and the 'power-sharing' type. Most of the time, this distinction is based on the way to distribute competencies formally in policy areas and can, therefore, be seen as a distinction concerning the federal organization. In the power separation model, the federal government has extensive powers to determine its strategies in revenue extraction and spending, but the member states also have the means to thwart the federal government's strategy. This is due to the principle that both types of governments have constitutionally guaranteed jurisdictional authority in a large number of policy matters. The accent is put in this system on autonomy and variation rather than on harmonization, solidarity and standardization. The power-sharing model is based on a functional division of labour between governments. While most of the decisions are taken at the federal level where both member states and the federal government have their say, implementation is almost completely in the hands of member states. Such a construction calls for a high degree of interdependence in policy-making and, hence, consultation and bargaining.

Often, different types of intergovernmental relations have been established on the basis of these distinct types of federal organization. The more 'adversarial' style of coordination in power separation types of federalism resembles 'intergovernmentalism': 'Coordination is unfettered by any mechanisms of joint action, unilateralism is common, and "partisan mutual adjustment" is a regular mode of interaction' (Painter, 2001a, 139). The 'cooperative' style in systems of power sharing by contrast results in 'the evolution of interlocking, collaborative forms of coordination, such as networks and joint decision systems. Governments must surrender autonomy (...)' (ibid.).

It is interesting to ask — and this is the second point of interest in this article — if in case one can demonstrate the importance of efficiency considerations in federal reforms, this has led to any convergence in these two types of federal organization and intergovernmental relations.

In order to answer these two questions (pertinence of efficiency in federal reforms and convergence), we will refer to the reform experiences of five industrially advanced federal countries (i.e., Australia, Austria, Canada, Germany, and Switzerland) that represent the two types of federalism but have also 'within-group variance': Australia and Canada are usually seen as power separation federations with adversarial intergovernmental relations while Austria, Germany, and Switzerland, the 'Continental' federations, belong to the power-sharing type. In each group we can distinguish countries on still another dimension, that is between more centralized federations in terms of the extraction and distribution of state revenues (Austria, Australia, Germany) and more decentralized federations (Canada, Switzerland).⁴

I will proceed as follows: An overview of reform experiences in these federal countries is provided in the next section. In the third section, I will analyse the 'x-inefficiency' we can expect in federal countries — that is, to what extent the existing federal organization may produce results that are in conflict with efficiency, thereby opening up the doors for discussions over better ways to improve efficiency. Finally, we will discuss the actual reforms and analyse to what extent efficiency has played a role in these reforms and whether we can find convergence among the power-sharing and power separation countries. The conclusion gives an answer to the two starting questions and finishes with some remarks on the role of interests and efficiency considerations.

Recent Reforms in Federal Countries

Each of the five federal countries under review has launched discussions on its federal organization at one point of time or another during the last 15 years. Table 1 gives an overview of key reform periods — that is, the period when discussions took place, whether revisions were decided upon ('successful') or not ('unsuccessful'), and what modality of reform had been chosen (only a more informal reform of intergovernmental arrangements or including formal constitutional revision).

Table 1 - Overview of periods, outcomes and modalities of reform in five federal countries.

Full table

In three countries — Australia, Canada and Switzerland — discussions started as early as in the first half of the 1990s; major revisions had already been undertaken during the 1990s in Australia and Canada while Switzerland needed 10 years to adopt the new federal laws by a referendum in 2005. Implementation laws were only adopted in 2007. In Germany and Austria, it took much longer to embark on serious reforms. One explanation for the quicker progress in the Anglo-Saxon power separation type is that reforms did not need constitutional revisions because they addressed the interaction between territorial governments rather than constitutional rights on competencies and revenues, while it was the contrary in the European power-sharing type.

It is only in Austria that a final decision has not yet been taken. A constitutional committee, the 'Austria-Konvent', proposed reforms in the context of a general revision of the Constitution, but parliamentary discussions are still continuing. Reforms in the other four federations were successful in the sense that significant changes in the makeup of intergovernmental relations were decided upon.

In Australia, intergovernmental relations became collaborative instead of adversarial (Painter, 1998, 6) with the effect that policy areas, especially relevant for competition, could be harmonized to a certain degree, and the overall planning of policy areas was facilitated.

In Canada, two agreements took place, one in relation to the harmonization of trade policy (Agreement of Internal Trade; AIT) in 1994 and the other concerning social policy (Social Union Framework Agreement; SUFA) in 1999. Both agreements established a collaborative style of intergovernmental relations with specified rules of conduct, which overcame the adversarial relations that had evolved since the 1960s (Brown, 2002).

In Switzerland, both the financial relationships and the competence distribution were revised with some implications for the coordination between territorial governments. The main thrust was to decentralize, to demarcate competencies, and to strengthen both delegation to cantons and horizontal coordination between cantons. A first step was a general revision of the Constitution in 2001, and another was the adoption of the 'new financial equalization' package ('Neuer Finanzausgleich'; NFA) in 2005 by a referendum (Dafflon, 2004; Larpin, 2006).

In Germany, a parliamentary committee, composed of representatives from the federal chamber and the people's chamber, developed plans to reduce the number of concurrent areas and the number of laws where the member states had veto-rights at the federal level. Financial relationships were only attacked after this reform had been adopted. Decentralization, efficiency, and demarcation were the key words of the first reform process. A first attempt failed in 2004, and only after a grand coalition came into power in 2005, both the Bundestag and the Bundesrat adopted the reforms (Sturm, 2005). The reform of financial relations is still debated.

The question is: Why do we see such a number of reform attempts since the 1990s? Were the reasons similar in countries? The most obvious explanation would be to assume that economic and social pressures are at the base of federal reforms. Such pressures can induce a re-assessment of existing policies and organizational arrangements. Frames like new public management or the economic theory of federalism can then help to focus on efficient means to cope with such pressures leading to a reconsideration of existing strategies and organizational arrangements.

Analysing the reform discussions in the five federal countries⁵ demonstrates that a general sense of crisis existed everywhere, bringing territorial actors together and launching discussions on existing organizational arrangements and intergovernmental arrangements. In Australia and Canada, the feeling of a crisis came above all because of a deteriorating trade balance and the recognition that the fragmented competencies in the trade and competition policy prevented, to a significant extent, the countries' adaptation to the globalization of the economy (Brown, 2002). Globalization and the perception that major economic and social reforms were necessary for adaptation also played an important role in Germany (Adelberger, 2001). Intergovernmental relations became the focus of reform because it was realized that the 'joint decision trap' of German 'cooperative federalism' was a major obstacle in launching such reforms (Scharpf, 1988). In the three power-sharing types — Austria, Germany, and Switzerland — it was budget deficits, however, that played the most prominent role in reform discussions.

One can assume that political elites on all territorial levels and in all political parties perceived these pressures as a serious challenge that needed adequate political answers. At the same time, it became clear, whatever the type of federalism, that existing federal arrangements both produced results that were detrimental to the adaptation to such challenges and restricted possible political solutions. In this context, the notion of 'efficiency' could become an 'eye-opener', helping policy-makers to see the failures of existing federal arrangements and at the same time becoming a 'tool' to design federal reforms.

Federal Organization and X-Inefficiency

The objective of efficiency could become so important, first because of the 'neo-liberal turn' in policy making during the 1980s with its distrust of the utility and efficiency of a large public sector and the rise of new public management theory, which delivered the causal theory needed to develop the design and the implementation tools for more efficiency. The concept expanded rapidly in the OECD area and not the least in the five federal countries discussed here.⁶

New public management 'aimed at fostering a performance-oriented culture in a less centralised public sector' (OECD, 1995, 8). A 'clearer focus' on 'results in terms of efficiency, effectiveness and quality of service' was needed (ibid.). Efficiency, which is the key notion in this frame, means in general cost-effective policy outcomes. An important means to achieve efficiency and hence a better performance is by establishing productivity targets and explicit standards of control. Corollaries are the creation of competitive environments within and among public sector organizations (OECD, 1995), less hierarchy and centralization in favour of a more flexible and innovative administration. The separation of political decision-making from the direct management of public services has often been seen as the main mechanism for achieving more efficiency (Osborne and Gaebler, 1992, 10; McLaughlin and Osborne, 2002). The relationship between the political 'principal' on the one hand and the administrative 'agent' on the other is to be organized above all by 'contractual mechanisms' (Carroll and Steane, 2002, 196).⁷

Although new public management concepts have been discussed almost exclusively in the context of public administration reforms, we believe that they have also influenced discussions about federal organization and intergovernmental relations. The reason for this is simply that we find regular complaints in all federal states about inefficiencies linked to the federal organization of such services.⁸

In this section we demonstrate in a general fashion what kind of 'x-inefficiencies'⁹ the federal organization of public services is likely to produce. The degree of 'x-inefficiency' is an indicator for the likelihood that the objective of efficiency and concomitant new public management ideas could infiltrate discussions on the reform of federal organization and intergovernmental relations. We will very briefly discuss three efficiency problems that are linked to the distribution of competencies with regard to decision-making and implementation, to the distribution of competencies in revenue distribution, and to the organization of strategic capacities.

Distribution of competencies

Policies in federal countries are often poorly coordinated because of unclear or concurrent competencies, which result in the duplication of tasks, negative spillover, and, hence, the inefficient use of public money. Given the fact that in most federal states concurrent policy areas (i.e. areas where authority is shared between territorial powers) have increased during the second half of the last century (Ter-Minassian, 1997; Watts, 1999), the discussion of the territorial distribution of competency rights is widespread. The unclear demarcation of competencies and blurring of responsibilities are clearly in conflict with basic ideas in new public management and would induce demands for a clear demarcation and disentanglement of authority in the various concurrent policy domains. In case that this is not feasible or not desirable, a solution in terms of a clear separation of the kind of authority should be envisaged where the federal government would develop the strategic outlines of the policy domain and member states would obtain the generous rights and leeway to implement such policies. Targets, contracts, and control measure would help to increase efficiency under such conditions.

Distribution of revenues

Another efficiency problem in federal states has been the transfer of more and more rights with regard to the extraction of revenues to the central level, resulting in what is often called a 'vertical fiscal imbalance' (VFI) (Ahmad and Craig, 1997). This means states must carry out the majority of administrative and public service delivery tasks in the federal system but are lacking adequate financial resources of their own to pay for these tasks. The federal government then transfers money in various ways to compensate the states for their efforts. In addition, most central governments are also responsible for the organization and — in part, the financing — of equalization measures that equilibrate forces within a federal country. Such a revenue structure can cause 'rent-seeking' or 'over-fishing' of resources and irresponsible behaviour because the central government can bail states out (Walsh, 1992; Ahmad and Craig, 1997).

The central government can abuse this situation in its favour by using money with conditions attached to interfere with the politics of the state, thereby blurring responsibilities. Such tendencies violate the basic principles of new public management, above all the principle of 'empowering agents', introducing clear responsibilities and accountability into the system. A solution to this efficiency problem according to the 'economic theory of federalism' would be to decentralize resources and give states enough tax autonomy to pay for services on their own. In addition, both taxing and spending rights should be concentrated on the same territorial level in order to raise accountable spending behaviour and to avoid negative spillover effects. This is the so-called 'connectivity principle' or the principle of 'fiscal equivalence' (Olson, 1969).

Strategic capacities

Depending on the different types of federal countries (power-sharing or power separation), a country's strategic capacity to adapt may be jeopardized. In power-sharing types of federal countries like Germany and Austria, and — to a lesser degree — Switzerland, 'joint decision traps' (Scharpf, 1988) may produce policies with the lowest common denominator, high transaction costs, and lengthy procedures. In power separation types of federal countries like Australia and Canada, there is the danger of policies of 'thrust and riposte' (Painter, 1991), failing coordination, and hence of a policy of duplication or of negative spillovers. In both cases, the capacity of federal countries to adapt to challenges strategically is reduced. Strategies often — in both types — develop and diffuse by indirect means, for example, by the 'power of the purse', which again raises the problem of clear responsibilities. In many ways, the lack of such strategic capacities becomes a serious disadvantage at a time of a 'fast-changing policy environment' when 'the need to strengthen the coherence of government action has never been greater' (OECD, 1995, 73).

New public management would foster a strengthening of strategic capacities at the centre. The main shift in thinking is that the central government no longer tries to interfere directly on the state level, but that policies formulated on the central level guarantee coherence, empowering member states in the execution of commonly agreed strategies and an intensive process of monitoring outcomes.

To what extent then did efficiency as an objective and concomitant concepts derived from new public management and economic theory of federalism influence the design of new solutions? Do we find converging trends in the set up of federations?

The Role of Efficiency in Federal Reforms

In order to determine the status of efficiency as an objective in reforms and likely converging trends, we need to understand the 'causal stories' that were behind federal reforms in countries. There are quite a number of different causal factors that one can discern in this process, but there is a clear distinction in causal patterns between power-sharing countries (Austria, Germany, Switzerland) on the one hand, and the two federal countries with a Westminster democracy and power separation (Australia and Canada) on the other (Bannon, 1999; Galligan, 1992). In both groups, efficiency and concomitant reform concepts had an important role to play, but in different ways.

If we take the power separation countries first, we see two causal paths that led to a reform of intergovernmental relations, and only to a minor degree to reforms of the federal organization. Both paths are determined by the specific federal organization of the Anglo-Saxon countries.

First, the impetus for reforms was due to the severe austerity policies the federal governments adopted in both countries in the 1980s. Federal governments contributed in several ways to the expenditures of member states, in particular in the welfare field. With increasing budget pressure, federal governments began to reduce their subsidies in substantial ways, thereby obliging member states to fill in the financing gap. Above all, in a federal state with considerable VFIs like Australia, this contributed to increasing dissatisfaction of member states. But the member states in Canada felt the problems in a very similar way. In both countries, this gave rise to initiatives of member states to revise the existing subsidy structures. In Australia, member states asked for more tax autonomy, and in Canada, they attempted to convince the federal government to stick to its previous commitments. In both cases, the confrontation led to the set up of discussions on the highest political level in order to reach an agreement.

The second impetus came from problems in maintaining economic competitiveness. In both countries, competencies in the field of competition were fragmented between the federal and the member state levels. The harmonization of laws in the field of competition became the main imperative for the federal government. Given the fragmentation of competencies, this could only be achieved if the member states were willing to join forces. Again, these driving forces led to the set up of negotiations and discussion at the highest level of politics. In this way, there was a bottom-up and a top-down motivation in Anglo-Saxon federations to change existing patterns of intergovernmental relations from the more adversarial style of the past to a more collaborative style of policy-making that would help to reach agreements on the subjects under discussion. Efficiency was not a factor in this part of the causal story, but it entered the arena of policy design once states decided to start building up collaborative relations. On the one hand, policy-makers were quite aware that collaborative relations should not become a new source of inefficient decision making with long periods of negotiation and unfulfilled promises during the implementation of agreed upon policies. Since the beginning, it was stressed that collaborative federalism should be organized in an efficient way that would allow the best possible outcome. On the other hand, a more collaborative style was also used to organize concurrent areas like environmental, health, and education policies in an efficient way. Duplication of efforts and waste of resources should be avoided. New public management and the economic theory of federalism helped to think about new ways to organize such concurrent areas.¹⁰

This demonstrates that, once the decision was taken to work differently in intergovernmental relations, discussions on the efficiency and effectiveness of concurrent policy areas were put on the agenda. It is here that new public management was used to inform policy-makers on likely solutions in order to organize intergovernmental relations, to guarantee an efficient handling of policies, and to overcome inefficiencies in concurrent areas.

In the power-sharing federal countries the causal story is different simply because cooperative relationships existed and became the problem instead of the solution. While the Anglo-Saxon federations started to organize collaboration in an efficiency-enhancing way, the European federations started to disorganize cooperation and used efficiency and its concomitant concepts to find alternative ways of organizing federalism.

One can discern different pressures for change in Europe. Owing to new opportunities emerging in the context of globalization, the richer member states in above all Germany and Switzerland developed rising ambitions to develop economic and trade policies of their own. These states increasingly felt the VFIs, rigid and generous equalization procedures, and centralized decision procedures as a hindrance in regional development, and they put forward demands for more decentralization and a relaxation of equalization obligations. In this case, the demands were simply to strengthen member states by decentralizing competencies and resources. Efficiency was not part of this discourse but the demands were of course compatible to the ideas of empowerment in the new public management framework and could profit from this discourse of enhancing the efficiency of the organization of governments.

The more important causal pressure for all three power-sharing countries, however, was the influence of budget deficits. Deficit problems were very high on the political agenda in Germany and Austria because of the severe stability criteria used within the European Monetary Union, and in Switzerland because of a rigid budget policy, which was fixed in the Constitution. The burden of excessive deficit spending drew attention to the waste of resources caused by the duplication of policy activities, additional costs of failing coordination, costs of overregulation and blurred competencies, the inefficient use of subsidies, etc. It is because of the discussion on budget deficits that we find an opening to the efficiency objective and arguments from new public management and the economic theory of federalism.

The overview demonstrates, first, that various causal paths explain the reform process in a country; second, that the kind of problems made policy-makers sensible to the objective of efficiency and allowed the entry of new public management and economic theory of federalism ideas in reform discussions; third, that the role efficiency and reform concepts linked to efficiency played was different in power separation and power-sharing countries. In the first group, efficiency considerations were used to organize the new forms of collaboration and to organize major concurrent policy areas in an efficiency-enhancing way. In power-sharing countries, attempts were made to overcome inefficiencies caused by established practices of 'cooperative federalism' by referring to concepts of new public management and economic theory of federalism. In sum, efficiency became a framework that all federal countries grasped at.

In the following section, we describe in more detail how the federal countries — with regard to the distribution of competencies and revenue distribution — have attempted to reform their organization and relations on the base of the efficiency objective.

Distribution of competencies

If we first take a look at reforms concerning the distribution of competencies in policy areas, we notice that the power-sharing countries direct their attention to the question of disentangling the many concurrent areas of competence that had been installed over time.

In Germany, reforms were both interest- and efficiency-driven. On the one hand, it has already been said that the pressure of the stronger member states demanding fewer equalization obligations and a decentralization of competencies has played an important role in debates of reform. But the federal government was also genuinely interested in becoming more independent from the influence of member states in decision-making on the federal level. In order to achieve this objective, the federal government was prepared to abandon some of its competencies. These converging interests helped to make decentralization and territorial separation of authority a reform topic.

On the other hand, German reforms were clearly driven by efficiency considerations. When the economic situation deteriorated in the beginning of this century, the federal structure — the veto-power of opposition parties in the Bundesrat, the disincentives for economic efficiency, and the incrementalism caused by the entanglement of competencies — was seen more and more as the culprit (Thaysen, 2004). This led, together with stronger demands from parliaments to render federal decision-making more transparent (see below), to the creation of a 'Federal Committee' ('Kommision von Bundestag und Bundesrat zur Modernisierung der bundesstaatlichen Ordnung') with the task to prepare a constitutional reform of federal arrangements. In the task description of this Committee it is said that a 'modernization of the existing order' should be achieved in order to make the federal government and the member state governments more capable to act and to make responsibilities clearer 'in order to fulfil public tasks more efficiently'.

In deciding to disentangle at least some policy areas, the major source of inefficiency, the 'joint decision trap', was set out of work in these areas and therefore it might have become easier to decide more quickly and unfettered by various territorial interests that in the past have led to policies of the lowest common denominator instead of policies raising performance. Disentanglement as an objective corresponds to the recommendations of new public management demanding clear responsibilities. However, Germany did not follow the advice of new public management to organize the remaining concurrent areas more in terms of the principal–agent or delegation model (i.e., by clearly attributing strategic prerogatives to the federal level and operational autonomy to the member state level). On the contrary, the existing framework setting legislation, which allowed the federal government to give general guidance in diverse matters, was abolished. In fact, there are no reforms envisaged that would change the logic of the joint decision trap in the remaining concurrent areas.

In Austria, a preparatory constitutional committee, the 'Austria-Konvent', set up in 2003, was supposed to deal with constitutional questions in general and with inefficiencies in federal arrangements in particular (Konrad, 2005). The revision of competency distribution was particularly stressed. In 2005, the recommendations of this committee were published with a number of reform items for the organization of federalism. Discussions in Parliament are still continuing, and it is unclear if the two-thirds majority on these (and other) questions can be found. The views of experts are in general pessimistic (Hämmerle, 2005). One explanation for this is the relative disinterestedness of political parties, which remained outside of most discussions in the 'Austria-Konvent'.

Discussions in the 'Austria-Konvent' itself were less interest-driven than in Germany, though of course interests always played a role as a constraining element for acceptable solutions. There were a number of member states that were claiming more powers, but in terms of a stronger co-decision-making power on the federal level than in terms of more decentralized competencies. On the contrary, several times during discussions in the 'Konvent' member states refused the decentralization of competencies because of excessive administrative and financial burdens this would put on them.

The real preoccupation of actors, however, was inefficiency. The financial situation of the country was at stake and new public management with its recipes was used to attack this problem. There was unanimity that the existing distribution of competencies was inefficient and that transparency was lacking. It was said that there were too many different matters and too many — often overlapping — competencies. The objectives were, therefore, to reduce the number of competence areas, to make decisions on competence distribution more flexible, and to simplify the demarcation of areas. The 'Konvent' made in its report concrete propositions how to reallocate competencies to the different territorial tiers.

Switzerland has been extremely successful with its reform project. One reason for this is, without any doubt, that there was a widely shared consensus about the need for fundamental reforms. Another reason is the careful agenda-setting of the federal government trying from the beginning onwards to integrate the different points of view of the cantons (Larpin, 2006). No canton should consider itself a loser of reforms. Finally, Switzerland's success must be explained by a strong role of economic experts who, since the beginning, were involved in the reform design process and accompanied the further development of the reform project. Because of these elements, one can speak of a rather technocratic reform spirit.

The fundamental reforms that should overcome the process of 'creeping centralization' in the form of financial aids of the federal government, blurred competency boundaries, and inefficiencies in the distribution of revenues was the new fiscal arrangement (NFA), which entailed a complete revision of both competencies and financial arrangements in Swiss federalism.

The main objective of the reform was to achieve a clear-cut demarcation of responsibilities. In the end, a reduction of 40% of such joint tasks was agreed upon. They were most of the time transferred to the cantons. Only a small number of policy areas now remains under the sole responsibility of the federal government. However, in those areas where interdependencies seem to be justified and where non-coordination would be counter-effective because of possible negative externalities, spill-over effects or free-riding behaviour, a new category of competence distribution, was introduced, forcing cantons to organize themselves by setting up inter-cantonal cooperation and financing. It was even agreed upon that inter-cantonal bodies that regulated such areas could obtain legislative rights in order to install more flexibility and to avoid otherwise a more centralist intervention by the federal government. In other interdependent policy areas, competencies were more clearly redistributed along the lines of principal–agent ideas: the federal government obtained the competency to 'decide' (strategic rights), while the cantons were free to use their operational discretion to execute federal laws.

In the power separation federal countries, it was not primarily inefficiencies that were the driving force but, as stipulated above, problems linked to the globalization process. Inefficiencies in concurrent areas were, however, recognized and led to a number of reforms. For example, the introduction of the 'Canada Health and Social Transfer' (CHST) arrangement in 1996 was designed as a measure of more economic efficiency in Canada. The previous cost-sharing programmes had been incentives for the provinces to spend as much as possible and were, hence, inefficient. Australia used the occasion to build up a stronger coordination in policy areas like health, environment, and education, which should help to overcome existing duplication of efforts and waste of resources.

Disentanglement was not seen as the major solution for problems, though it appeared occasionally on the agenda both in Canada and Australia. Labour force training in Canada, for example, was, according to the will of the federal government, either to become subject to a form of co-management between territorial powers or should be devolved. The provinces had the right to choose. In environmental policy, the federal government gave most competencies to the provinces because they had better, more direct and local information. In Australia, disentanglement was discussed without a major revision of existing practices.

More important in both countries was the objective to overcome fragmented competencies — above all in the field of competition — by making them subject to negotiation and decision-making on the federal level. Evidently, such a move could not be made without the consent of the member states, and this may be the major reason that a serious attempt was made to overcome the adversarial working of intergovernmental relations. In contrast to the European countries, Australia and Canada had to create new forms of coordination, and it is in this context that new public management reappears — this time as a framework that indicates ways to best to organize such cooperation.

The AIT in Canada, for example, was the first attempt to introduce a stable, rule-based system of cooperation aiming to reduce uncertainty and ad hoc behaviour in intergovernmental relations and improving the economic efficiency of economic policies (Brown, 2002). To make 'collaborative federalism' work, the federal government had to agree to limit its own authority in these areas and accept the co-decision of the provinces in social policy matters (Cameron and Simeon, 2002). The provinces also insisted on new arrangements for dispute settlement that could minimize the degree of conflict in the intergovernmental arena. Such settlements should be 'simple, efficient, effective and allow for a maximum flexibility of governments'.

While new public management ideas informed policy-makers on the best way to work together, the arrangements on social policies, formalized within the SUFA, endeavoured to reorganize social policies in an efficiency-enhancing way by means of the search for best practices, the use of performance indicators, and the use of monitoring. Information sharing, the development of indicators, and joint planning on the basis of these data were means to raise the performance in these areas (Théret, 2001). While in this way, a common and coherent policy should be developed, the provinces were attributed sufficient discretion to determine how to implement these objectives. This can be seen as a device to set up a principal–agent form of governance: the executing member states participate in the development of general aims where the federal government takes the lead, common objectives are formulated, monitoring procedures installed, and control takes place.

Australia has set up very similar devices in the areas of competition, environmental, health, and education policies. In addition, often independent regulatory agencies composed of experts and stakeholders were set up to avoid political interests entering the process of decision-making too early and so that a more expert-based policy with integration of stakeholders could take place (Selway, 2001; Saunders, 2002).

In both countries, we find, therefore, the will of all actors to coordinate action and to overcome partisan divide. Efficiency has been an important framework to arrive at this point, and it helped in general to develop a problem-orientation and 'whole-of-government' perspective (Galligan, 1995).

In sum, in the power separation federal countries the question of disentanglement played a role. There is no doubt about that. But the main task was to overcome fragmented competencies. This was done first by developing new 'collaborative relationships', which were organized based on recommendations derived directly from the new public management framework. Joint decision-making should not fall into 'joint decision traps', like it had in Germany. New public management could help to avoid such traps because it served to overcome the narrow-minded interests of actors. Second, it could do this by reorganizing concurrent policy areas and injecting a more 'rational' handling of ways to decide on basic objectives and the execution of these objectives.

One should, however, make clear that these reforms were acceptable for both the federal government and most member states because abandoning competencies on the side of member states was compensated for by co-decision-making powers and, often, additional money by the federal government was offered for the implementation of agreed upon objectives.

Revenue distribution

Revenue questions are generally the most difficult to settle. Propositions of the economic theory of federalism, that is, to grant member states more financial autonomy and to introduce the connectivity principle, met considerable resistance by federal governments and even by some member states. What has often been achieved, however, was a reduction of the 'power of the purse' of federal governments. Concerning this point, we do not find significant differences between power-sharing and power separation countries. It is rather a difference between more centralized (Australia, Germany, Austria) and more decentralized (Canada, Switzerland) federal countries.

This is not to say that there have not been discussions or attempts at all to change the revenue distribution structure in a more fundamental way, but most of the time interests have blocked such attempts. In Germany, it was the richer member states that were pushing to change the fetters of a rigid and egalitarian equalization system. In the discussions on the first federal reform such financial topics were, however, discarded as a major financial settlement had just been concluded and it seemed unfeasible, given the resistance of the poorer member states to reopen the negotiations on financial issues. In the meantime, however, a new parliamentary committee has been set up that should tackle matters of revenue distribution. Most observers remain sceptical if these discussions will achieve a compromise between the various actors.

In Australia, the member states were attempting to reduce the VFIs that granted the federal government considerable powers in concurrent policy areas. Although the topic appeared on the agenda at the end of the 1980s, it was quickly abandoned when the new prime minister, who took office in 1992, defended vigorously the financial prerogatives of the federal government (Keating and Wanna, 2000).

In Austria, in the 'Konvent', we find a strong group of actors that favoured the connectivity principle, but another discourse, the imperative of harmonization in the light of globalization and the European Union, prevailed. The decentralization of some revenue property rights, necessary for the introduction of the connectivity principle, was seen as in conflict with the necessity to overcome a disparate development of tax policies at the level of the states.

Although in Switzerland the states already had considerable tax autonomy, the connectivity principle was discussed as an option to overcome existing inefficiencies of the financial arrangements. Finally, the principle was introduced in those policy fields, which will be organized according to the principle of horizontal cooperation after the implementation of the reform. In Canada, the question was not one of more tax autonomy but how to keep the federal government in the financing of welfare expenditures.

These developments demonstrate that the interest struggle between the federal government and member states immediately took the objective of 'empowerment' by stronger tax autonomy hostage. It is typical that such interest conflicts were more stubborn in the more centralized federal nations Austria, Germany, and Australia, where the federal government has traditionally had a very strong position in revenue distribution and where VFIs are high. In these federal countries, it seemed difficult to change from an 'interest view' to an 'efficiency view' in fiscal matters. Moreover, some member states were often ready to accept the financial aid of the federal government in favour of less financial and administrative burdens. In the more decentralized federal nations (Switzerland and Canada), this was different. In both Canada and Switzerland, there was already considerable tax autonomy, and at least in Switzerland, this made it possible to launch a more depoliticized debate on the advantages of the connectivity principle.

However, in all countries there were changes in revenue distribution, which did not envisage a general strengthening of financial capacities of member states but did attempt to overcome some of the inefficiencies in the system of subsidizing expenditures of member states by federal government money. It is here that efficiency ideas were used for reforms.

The general trend was to replace 'conditional grants' by block grants that come as a lump sum and are for the most part at the discretion of lower governments to spend. This trend fit into efficiency considerations and new public management thoughts, as it meant a reduction of the administrative burden that was linked to the use of conditional grants. It recognized the operational autonomy of the states and improved their capacity to act flexibly with regard to local problems. Block grants can also be seen as strengthening the position of the states vis-à-vis the central government as the degree of intervention decreases.

With the exception of Austria, all countries at least intended to reduce conditional grants: Germany wanted to use block grants in the long run to finance the policy areas which were planned to be decentralized; Canada introduced block grants as the main instrument in the CHST; and in Switzerland block grants were an essential part of the NFA. Also, Australia strengthened the role of block grants. In addition, matching grants were discredited everywhere and fewer grants were used because of windfall-profits and difficulties of poorer states to match the federal money.

Finally, in Switzerland, Australia, and Canada federal government money — either in the form of conditional or as block grants — became more and more subject to the establishment of 'contracts' and regular monitoring procedures. There is a difference though in the enforcing character of monitoring procedures. Australia, for example, introduced in some regulatory areas independent committees, the judgment of which has consequences for the continuation of funding for member states. In Switzerland, there is no legal possibility to enforce sanctions. Another important element introduced in Switzerland, Australia, and Canada is that grant programmes are now limited in time with usually an evaluation in the end in order to give no perennial rights for subsidies. All these elements are used in order to raise efficiency in the use of public money.

Conclusions

Without any doubt, the objective to increase the efficiency of the federal organization and intergovernmental relations has influenced recent reform attempts in federal countries; however, this has been to different degrees and with different success. Reducing 'x-inefficiencies' has been on the minds of all federal policy-makers, but this has not led to the emergence of one type of federal organization or intergovernmental relations in federal countries.

More specifically, it was demonstrated that though both groups of federal countries — the power separation type with adversarial intergovernmental relations and the power-sharing type with cooperative intergovernmental relations — were plagued by similar challenges, the existing structure provoked different answers. The Anglo-Saxon countries with established adversarial intergovernmental relations were required to look for more joint decision-making procedures and intensified cooperation, while the Continental federal countries with a strong institutionalization of cooperation felt obliged to revise and reduce their extensive cooperation mechanisms. Efficiency considerations helped in both instances to reorganize. Where does this leave us concerning the question of converging intergovernmental relations and/or federal organization?

Apparently, all federal countries are now using the coordination of intergovernmental relations as an important mechanism to find answers to the challenges of globalization and budget deficits. There is, however, a distinction between 'collaborative federalism' in Australia and Canada (Painter, 1998; Lazar, 2000a; Saunders, 2002) and 'cooperative federalism' (Braun, 2005). The first one has been more clearly founded on ideas of efficiency than the second one, which has emerged much earlier, and, hence, appears to avoid rigid institutionalized and immobile structures that often contribute to the joint decision trap in cooperative federal countries. Therefore, though cooperation becomes the dominant principle of organizing intergovernmental relations in both types, this seems not to have resulted in the same practices of decision-making.

Second, in all federal countries there is a common trend to turn the tide and avoid further centralization of competencies by achieving a disentanglement of competencies in concurrent areas, though this trend has clearly been more remarkable in the cooperative federal countries. Switzerland has been the most progressive in this respect. Although Australia has a large number of concurrent areas, it has scarcely used this option and Canada had already decentralized a large number of areas.

Third, there are commonalities in the way Australia and Canada make concurrent areas more efficiency enhancing on the one hand, and Switzerland on the other. Although we find similar propositions in Austria, they are unlikely to be accepted, and Germany has not chosen this path of reform. The tendency to use monitoring for the execution of federal programmes and the use of federal money or principal–agent arrangements has been designed and implemented in Switzerland, Australia, and Canada. With regard to revenue distribution, these three countries have started to organize fiscal relations in a similar fashion. There has been a strengthening of the decision-making powers of member states concerning the resources transferred by the federal government as well as the introduction of efficiency-enhancing procedures in federal government transfers to member states.

In sum, efficiency and its corollary ideas, new public management and economic theory of federalism, has influenced reform design in federal countries in many ways. This has contributed to similar conceptions of how to organize concurrent areas and federal subsidies. Disentanglement and the decentralization of competencies has become accepted as a general principle, even in the most centralized federal countries. Intergovernmental relations, finally, have become more similar because of the intensification of collaboration in power separation countries. The ways to organize intergovernmental relations have, however, been distinct with efficiency as an important means of re-organizing intergovernmental relations in power separation countries while it has not become a point of reference in power-sharing countries.

Efficiency considerations were — this has become clear in the presentation of the various reforms in federal countries — often confronted with egoistic interests of territorial actors. In fact, one should not think that efficiency, which has served in many ways as a 'focal point' (Weingast, 1995) of discussion, as a point of reference, which was shared by almost all actors, could shift the interest struggle in the territorial distribution of powers to the background. Our description of reforms demonstrates that, first, reforms could also be inspired by actors' interests and, second, that interests acted as constraints or facilitators of reforms based on new public management ideas:

Interests as stimulators of reform were important in Canada, where member states united to integrate the federal government into a cooperative game on social policies. They were a pertinent factor in Australia where member states strived for more financial independence from the federal government, and they have been at the base of German reforms where the rich member states embraced ideas of competitive federalism to remove themselves from the bonds of federal equalization procedures and gain more autonomy. Such interests of member states were undercurrents in the maelstrom of federal reforms. In each case, they were accompanied, however, by the more encompassing efficiency considerations.

Interests have also been facilitators for efficiency reforms in federalism. One could not explain the success of German reforms if one did not include the idea that the federal government had a strong self-interest in granting decentralization and disentanglement in order to free itself from the burden of joint decision-making in the federal chamber. Of course, the same holds true for Australia and Canada where the federal government was drawn into collaborative agreements with motives of its own — that is, to harmonize legislation in important policy fields and to gain (more) access to policy domains managed by member states.

Interests have, however, most often been constraints on the adoption of reform agreements. One cannot imagine that Australia would have succeeded in setting up its collaborative intergovernmental relations and new public management reforms in concurrent areas without the 'power of the purse'—that is, without the federal government injecting additional money to member states in order to buy their good will (Braun, 2005). The poorer German member states successfully blocked discussions on fiscal reforms during the first phase of reform. Some of the Austrian member states were not willing to pay the price for more decentralization of competencies and rejected such proposals. The successful reform process in Switzerland needed careful management by the federal government and more than 10 years of negotiation in order to integrate all the different views (Larpin, 2006).

Interests are therefore undeniably involved in the reform process. They play a role, but one can state that interests were often subsumed to and/or accompanied by 'efficiency' considerations during the reform process. In fact, the search for greater efficiency seems to have been the preponderant motive in stimulating and designing reforms. The result is a federal organization that becomes, at least in part, more similar between countries and which can 'cross bridges' as similar reform designs in a power-sharing federal country like Switzerland (and Austria at least in terms of propositions) and in power separation federal countries like Australia and Canada demonstrate. This does not mean, of course, that we can expect one single type of federalism in the future. There are components in the organization of concurrent areas that begin to look similar while other features, such as the organization of cooperation in intergovernmental relations, remain distinct.

The findings in this article are conclusive enough to suggest that efficiency has become a 'prime mover' in federal reforms today, and this both in power separation and power-sharing countries.

Notes

¹ Federal organization refers to the formal and informal distribution of competencies in policy fields between the federal and member state governments.

² Intergovernmental relations denote the formal and informal ways of the federal and member state governments to deliberate and decide on policy issues.

³ See for example Painter (1996) for Australia; Wollmann (2001) for Germany; Peters (2001) for the USA; Cameron and Simeon (2002) for Canada; a comparison between Canada and Australia is done by Brown (2002).

⁴ The choice of countries was initially reflected by these considerations of variance: each cell in the typology of power sharing/power separation and centralized/decentralized should have at least one federal country. A second condition was that the countries should have undergone recently fundamental institutional reform or reform attempts. This was for example not the case in the United States though there have been incremental changes in the States that, by the way, confirm the general trend towards more efficiency considerations though not towards more decentralization (Posner, 2007). Belgium and Spain were not taken into account, as they are (still) difficult to define in terms of power separation and power sharing, with Belgium having become a federal country only in 1993 and with Spain still in a transition period from a central to a federal state.

⁵ The empirical basis of our analysis is secondary literature in countries on the one hand, and documentation of the reform process on the other. In Germany, the proceedings of the discussions in the parliamentary committee on federal reforms are published on the website of the Bundestag (www.bundesrat.de); the same holds for the proceedings and propositions of the Austria-Konvent (www.konvent.gv.at); in Switzerland, the various reform papers produced during the process were published. Dafflon (2004) gives an excellent overview of the discussion. In Australia and Canada, there have been no formal parliamentary committees. Various secondary literature has served as the main background for our discussion, most notably (Galligan, 1995; Painter, 1998; Keating and Wanna, 2000) for Australia and (Lazar, 2000b; Théret, 2001; Cameron and Simeon, 2002; Boadway, 2004) for Canada. Brown (2002) has been very useful for both cases. Additional empirical information concerning SUFA was taken from the evaluation report 2003.

⁶ See above all Pollit and Bouckaert (2000, 2003) for a general comparison; in particular for Austria: Hammerschmid (2004); for Canada also Brown (2002); for Germany, Wollmann (2001); for Switzerland, Mooser and Kettiger (2004); for Australia, Davis and Rhodes (2000), Carroll and Steane (2002).

⁷ Concerning these key points, the OECD finds, after 20 years of experience with new public management, that today OECD governments have indeed become 'generally more transparent, accessible and customer-aware, more devolved, more efficient, and more performance-focused' (OECD-Observer, 2005, 6).

⁸ Which, of course, have also been prominently discussed by the 'economic theory of federalism', which can be seen as another intellectual source of inspiring reforms of the federal organization that are supposed to raise efficiency (Olson, 1969; Tullock, 1969; Oates, 1972).

⁹ X-inefficiency signifies the gap between the inefficiencies caused by the federal organization and intergovernmental relations on the one hand and an optimal organization of public services proposed by the new public management framework on the other.

¹⁰ See for Australia, Davis and Rhodes (2000), Keating and Wanna (2000), Painter (2001b) and for Canada, Battle (1998), Government of Canada (1999).

References

1. Adelberger, K. (2001) 'Federalism and its discontents: fiscal and legislative power-sharing in Germany, 1948–99', Regional and Federal Studies 11(2): 43–68.
2. Battle, K. (1998) 'Transformation: Canadian social policy since 1985', Social Policy and Administration 32(4): 321–340. | Article |
3. Boadway, R. (2004) 'Should the Canadian federation be rebalanced?', Working Paper IIGR, Queens University.
4. Börzel, T.A. and Hosli, M.O. (2003) 'Brussels between Bern and Berlin: comparative federalism meets the European union', Governance: An International Journal of Policy and Administration 16(2): 179–202.
5. Braun, D. (2005) 'Between market-preserving federalism and intergovernmental coordination: the case of Australia', Swiss Political Science Review 12(2): 1–36.
6. Brown, D.M. (2002) Market Rules. Economic Union Reform and Intergovernmental Policy-Making in Australia and Canada, Montreal-Kingston: McGill-Queens University Press.
7. Cameron, D.M. and Simeon, R. (2002) 'Intergovernmental relations in Canada: the emergence of collaborative federalism', Publius 32(2): 49–71.
8. Carroll, P. and Steane, P. (2002) 'Australia, the new public management and the new millennium', in K. McLaughlin, S.P. Osborne and E. Ferlie (eds.) New Public Management. Current Trends and Future Prospects, London: Routledge, pp. 195–209.
9. Dafflon, B. (2004) 'Federal-cantonal equalisation in Switzerland: an overview of the present system and reform in progress', Working Paper No. 356, BENEFRI Centre d'Etudes en Economie du Secteur Publique, Fribourg.
10. Davis, G. and Rhodes, R. (2000) 'From hierarchy to contracts and back again: reforming the Australian public service', in M. Keating, J. Wanna and P. Weller (eds.) Institutions on the Edge? Capacity for Governance, St. Leonards NSW: Allen & Unwin, pp. 74–98.
11. Elazar, D.J. (1962) The American Partnership, Chicago: Chicago University Press.
12. Galligan, B. (1995) A Federal Republic: Australiás Constitutional System of Government, Cambridge: Cambridge University Press.
13. Government of Canada (1999) A Framework to Improve the Social Union for Canadians. An Agreement between the Government of Canada and the Governments of Provinces and Territories, Government of Canada.
14. Hämmerle, W. (2005) 'Föderalismus in Österreich. Es bleibt beim Prinzip ohne Substanz', in A. Khol, G. Ofner, G.R. Burkert-Dottolo and S. Karner (eds.) Österreichisches Jahrbuch für Politik 2004, Wien-München, pp. 669–682.
15. Hammerschmid, G. (2004) 'Verwaltungsreform in Österreich. Stand und Perspektiven vor dem Hintergrund internationaler Erfahrungen', Forum Parlament 2(2): 71–77.
16. Keating, M. and Wanna, J. (2000) 'Remaking federalism', in M. Keating, J. Wanna and P. Weller (eds.) Institutions on the Edge? Capacity for Governance, pp. 126–155.
17. Konrad, C. (2005) 'Dann bleibt es eben so: Föderalismus und Kompetenzverteilung als Themen des Österreich-Konvents', Österreichische Zeitschrift für Politikwissenschaft 34(4): 351–366.
18. Larpin, B. (2006) Réformes contemporaines du fédéralisme: le cas suisse à la lumière de l'expérience allemande, Lausanne: Institut d'Etudes Politiques et Internationales, University of Lausanne.
19. Lazar, H. (2000a) 'In search of a new mission statement for Canadian fiscal federalism', Ideas, Theories and Concepts 3–39.
20. Lazar, H. (2000b) 'The social union framework agreement and the future of fiscal federalism', Ideas, Theories and Concepts 99–128.
21. Mooser, C. and Kettiger, D. (2004) '10 Jahre Wirkungsorientierte Verwaltungsfuhrung in der Schweiz: Entwicklungen', Ergebnisse und Perspektiven, www.pumaconsult.ch.Bern.
22. Oates, W.E. (1972) Fiscal Federalism, New York: Harcourt Brace Jovanovich.
23. OECD (1995) Governance in Transition. Public Management Reform in OECD Countries, Paris: OECD.
24. Olson, M. (1969) 'The principle of "Fiscal Equivalence": the division of responsibilities among different levels of government', American Economic Review 59: 479–487.
25. Painter, M. (1996) 'Economic policy, market liberalism and the "end of Australian politics". [Revised version of the Presidential Address to the Australian Political Studies Association. Conference (1995: Melbourne)]', Australian Journal of Political Science 31(3): 287–299. | Article |
26. Painter, M. (1998) Collaborative Federalism: Economic Reform in Australia in the 1990s, Cambridge: Cambridge University Press.
27. Painter, M. (2001a) 'Multi-level governance and the emergence of collaborative federal institutions in Australia', Policy & Politics 29(2): 137–150. | Article |
28. Painter, M. (2001b) 'When adversaries collaborate. Conditional co-operation in Australia's arm's lenght federal polity', in U. Wachendorfer-Schmidt (ed.) Federalism and Political Performance, London: Routledge, pp. 130–145.
29. Peters, B.G. (2001) 'Administrative reform and political power in the United States', Policy & Politics 29(2): 171–179. | Article |
30. Pollit, C. and Bouckaert, G. (2000) Public Management Reform. A Comparative Analysis, Oxford: Oxford University Press.
31. Pollit, C. and Bouckaert, G. (2003) 'Evaluating public management reforms: an international perspective', in H. Wollman (ed.) Evaluation in Public-Sector Reform, Cheltenham: Edward Edgar Publishing, pp. 12–35.
32. Posner, P. (2007) 'The politics of coercive federalism', Publius 37(3): 390–412.
33. Saunders, C. (2002) 'Collaborative federalism', Australian Journal of Public Administration 61(2): 69–77. | Article |
34. Scharpf, F.W. (1988) 'The joint-decision trap: lessons from German federalism and European integration', Public Administration 66(3): 239–278. | Article |
35. Selway, B.M.B.M. (2001) 'The federation: what makes it work and what should we be thinking about for the future', Australian Journal of Public Administration 60(4): 116–122. | Article |
36. Sturm, R. (2005) 'Föderalismusreform: Kein Erkenntnisproblem, warum aber ein Gestaltungs- und Entscheidungsproblem?' Politische Vierteljahresschrift 46(2): 195–203. | Article |
37. Thaysen, U. (2004) 'Föderalismusreform in Deutschland: auch aus der Sicht der Nachbarn zum Erfolg verpflichtet', Forum Parlament 2(1): 28–33.
38. Théret, B. (2001) 'La solidarité sociale dans le pacte fédéral canadien. Histoire d'une crise et de son dénouement', Critique internationale (11): 145–160.
39. Tullock, G. (1969) 'Federalism: problems of scale', Public Choice 6(1969): 19–29. | Article |
40. Weingast, B.R. (1995) 'A rational choice perspective on the role of ideas: shared belief systems and state sovereignty in international cooperation', Politics & Society 23(4): 449–464. | Article |
41. Wheare, K.C. (1967) Federal Government, London, New York: Oxford University Press.
42. Wollmann, H. (2001) 'Germany's trajectory of public sector modernisation: continuities and discontinuities', Policy & Politics 29(2): 151–170. | Article |

About the Author

Dietmar Braun is Professor of Comparative Political Science at the Institut d'Etudes Politiques et Internationales of the University of Lausanne (Switzerland). He has studied political science at the University of Amsterdam (PhD in 1988) before he worked as a research fellow at the Max Planck Institute for Societal Research in Germany (1989–1991) and as an assistant professor at the Institute for Political Science at the University of Heidelberg (1991–1996) where, in 1996, he presented his habilitation thesis. In October 1996, he became full professor in Lausanne. He was editor of the Swiss Political Science Review between 2001 and 2004. Since the 1st of March 2004 he is a member of the National Research Council of the National Science Foundation in Switzerland. Dietmar Braun has published on labour market policies, higher education and research policies, federalism, as well as on topics in political theory. Recent monographies are: Delegation in Contemporary Democracies (London: Routledge) (with Fabrizio Gilardi) (2006) and Fiscal Policies in Federal States (Aldershot: Ashgate, 2003).