Bibliothèque

Conclusion to Intergovernmental Relations in Federal Countries

CONCLUSION Richard Simeon INTRODUCTION he eight countries represented in this volume reflect the enormous variety and diversity that exist in federal systems. They include two of the world’s oldest and most established federations – the United States and Canada – and two of its youngest – Russia (in its modern form) and South Africa (which has not formally designated itself as federal). They include four parliamentary federations – Canada, Australia, South Africa, and Germany – and four presidential/congressional federations – the United States, Argentina, Brazil, and Russia. They include four of the wealthiest federations, and four in the low to middle income range. Despite these differences, intergovernmental relations are ubiquitous and pervasive in all of them. This flows from the inevitable fact of interdependence among their constituent governments, a result of the complexities of the contemporary policy agenda and the impossibility, even when the inspiration originally was to create water-tight compartments, of drawing clear and separate lines of responsibility. But the institutions and processes that these countries have developed to manage the intergovernmental relationship vary widely in several dimensions. They vary from country to country, and within countries between different time periods and between different policy areas. The word “fluid” to describe relations among governments crops up repeatedly in the preceding chapters. In this conclusion, I will summarize the findings and conclusions of these case studies, and explore the ways in which each country responds to some common questions. What are the major dimensions along which the theory and practice of intergovernmental relations can vary? What explains variations across space and time? And with this in mind, what lessons might practitioners and scholars learn from others’ experience as they contemplate reform in their own countries? The importance of doing this is underlined by the fact that, in differing degrees, every one of these authors finds their system wanting in at least some dimensions. VARYING PATTERNS IN INTERGOVERNMENTAL RELATIONS There are many ways in which one can describe, organize, and classify systems of intergovernmental relations. No single system of classification can capture all the detail and nuance in these analyses. But the following dimensions appear to capture the major aspects of variation. • Institutionalization or formalization To what extent are the institutions of intergovernmental relations built into formal governing structures? To what extent is the machinery of intergovernmental relations mandated by the constitution or by legislation? To what extent are the operations of the institutions themselves governed by explicit procedures and formal decision rules? Or are the institutions fluid and ad hoc, developing and changing according to the political needs of the participating governments? It is important to distinguish here between the constitutional provisions that shape the overall structure and character of the federal system and the more focused institutions that have been developed to foster co-ordination in intergovernmental relations. Two groups of federations do not build intergovernmental structures formally into their constitutional system. First are the older federations – the United States, Canada and Australia. All were born in an era of limited government, so that their founders saw little need for formal mechanisms to manage interdependence. In an era of complex, all pervasive governance, interdependencies and spillovers grow exponentially, with the attendant risks of contradiction and duplication, requiring the development of extensive mechanisms of intergovernmental relations. From their outset, these federations emphasized a dualist, separated or divided model of federalism, in which each government would be responsible for both law-making and implementation of a defined list of responsibilities. They did not anticipate the overlapping and interdependence that defines modern government, and so did not build formal intergovernmental arrangements into their constitutional systems. This is not to say that their constitutions were silent on some critical determinants of how the intergovernmental relationship would work. In Canada, federal and provincial powers were set out in two separate lists, and the division of powers contains two important residual clauses –the “peace, order and good government”clause for the federal government, and the “property and civil rights”clause for the provinces. In addition, the “disallowance,”“declaratory,”and “reservation” powers all suggested an intergovernmental relationship in which the provinces would be subordinate to overriding federal power. Moreover, federalism was accompanied by a Westminster style parliamentary system, the foundation of which is the accountability of each executive to its own legislature. This would clash directly with a system in which governments instead became accountable to each other, and bound by their collective decisions. Thus in Canada even the most minimal institutionalization of the process – such as a commitment to annual meetings of First Ministers – has not been put in place, despite many proposals to do so. Indeed, constitutional entrenchment was included in three Canadian constitutional agreements, in 1971, 1987, and 1992, but all three failed (for other reasons), leaving the ideas in limbo. A federal-provincial agreement in 1985 contained a five-year commitment to annual conferences, but it was not renewed, despite strong provincial pressure to do so. Indeed, the trend in recent years has been towards less frequent, and less formal “First Ministers’Meetings”rather than “Conferences.” Australia is another federation joined to a Westminster style parliamentary system. Here too the model is a dualist one, with separate lists of powers, and with each order of government equipped with a full set of powers both to legislate and to implement its legislation. As in Canada, therefore, intergovernmental institutions must respect the overriding principle of accountability of executives to their respective legislatures. Rather than being part of the fundamental constitutional design, intergovernmental mechanisms are “add-ons” responding to the reality of interdependence, but with little or no legal or constitutional status. Both have created a wide variety of intergovernmental mechanisms, but at the same time, both have resisted suggestions to entrench these in the constitution, and have avoided giving them a statutory base. They remain ad hoc and fluid. Since they have no decision-making power, there is no need for the establishment of formal decision rules. Both Canada and Australia have recently concluded intergovernmental agreements and accords, complete with firm commitments and dispute-settlement mechanisms, but these remain informal – they too are not enshrined in legislation, and they are not judicially enforceable. The United States is a presidential/congressional rather than a parliamentary federation. Its original design, with equal state representation in a Senate appointed by the states appears to envisage Congress as the primary site for managing intergovernmental relationships, but this role diminished following the constitutional amendment to require election of Senators. The United States’ constitution also envisions a dualist pattern, with each level of government responsible for both legislation and implementation. As Kincaid shows, many other aspects of the U.S. constitution have implications for intergovernmental relations – such as the “full faith and credit” clause – and the constitution does envision co-operation as well as competition between state and federal governments. But here, too is a reluctance to institutionalize formal intergovernmental institutions. While there are a number of permanent bodies, such as the National Governors’ Association, the more common intergovernmental relationships are informal and ad hoc, varying widely among policy fields – “picket fence federalism.” The German parliamentary model – in sharp contrast – is one of “shared” and “integrated” federalism, emphasizing not the distinct status and roles of the different orders of government, but of their collective responsibility for both legislation and implementation. Most national legislation is implemented and delivered by Land governments. This model requires that the constitution and legislation spell out a complex set of institutions and rules to govern how they will operate in order to manage the relationship. Intergovernmental relations as a result are at the institutionalized and formalized end of the continuum. South Africa is an interesting blend of Westminster style parliamentary government, and German style co-operative governance. Chapter Three of the 1996 constitution sets out broad principles for co-operative government closely related to the German principle of Bundestreue. Like Germany, South Africa has a long list of concurrent powers, and provides for provincial implementation of national laws. Not surprisingly, therefore, it looks to a formalization and institutionalization of the intergovernmental relationship; and the constitution requires that national legislation to establish it be passed. Instead, however, the legislation has not yet been devised, and the South African model of multi-sphere governance has instead developed a more informal, ad hoc set of institutions (the Budget Council, MinMECs, and the like) which operate along lines similar to those in Canada and Australia. It remains to be seen whether these will be sufficient to meet South African needs, or whether legislation formalizing the institutions and practices will turn out to be necessary. In Argentina and Brazil, the two other presidential/congressional federal regimes, the constitution also profoundly affects the operation of intergovernmental relations. They differ from the United States primarily in the extent to which state governors, often with close relationships with members of Congress, are powerful actors at the national level. The nexus of governors, Congress and the presidency are the key focus of intergovernmental relations. Brazil is also distinctive in that its constitution spells out in detail many aspects of jurisdiction and finances. In neither country, however, does law or the constitution establish the mechanisms for ongoing co-ordination at the administrative level. In Brazil, Ferreira Costa notes that the “operational rules … of the political and administrative institutions that should be responsible for co-ordinating or promoting intergovernmental co-operation are left completely informal,”resulting in a “patchwork pattern of intergovernmental relations.” As for Argentina, Alberto Fohrig argues that there are few “institutional routines” for mediating among governments. Economic and political uncertainty, combined with “institutional instability,”contribute to a complex and fluid set of relationships between levels of government, strongly influenced by personal and partisan considerations. He laments the lack of effective standards for governing intergovernmental relations, even as he describes a wide-variety of ever-changing mechanisms. Russia is also a federation in flux. Its early years were characterized by a complex jockeying among the units of the federation, governed by no common view of the federation, and no clear rules. The current Russian President has been seeking to formalize the relationships, through changes in the constitution of the second chamber, establishment of a State Council which will brings heads of the constituent governments together quarterly, and by appointment of seven regional “plenipotentiary presidential representatives,”whose role is to encourage co-ordination. • The balance between “executive” and “legislative” intergovernmentalism In most federations, the bulk of intergovernmental affairs are carried out within the executives, in relationships between First Ministers (Presidents and Prime Ministers, Governors and Premiers), members of their cabinets, and officials in the bureaucracy. This reflects the general tendency towards executive dominance in complex modern governments, but is especially true in the Westminster style federations, where the concentration of power in the hands of executives is especially marked. “Executive federalism”is the term universally applied to the Canadian model; Australia follows a similar pattern. In some federations, legislatures do play a significant role. They can do so in two ways. First, the second chamber in the legislature may represent regional governments and populations in the national parliament and thus provide a forum for representing and advocating regional interests in the national legislature. Second, elected legislators may play a role in overseeing the executive’s conduct of intergovernmental relations. The first of these elements suggests that in federal systems with strong second chambers representing provinces – that is, strong elements of “intra-state federalism”– the second chamber will be the primary arena that will shape the intergovernmental relationship. The evidence in the previous chapters suggests that this is seldom the case. In Canada, the appointment of senators by the federal government has meant that the Senate plays virtually no role in mediating between the two orders of government. In Australia, another Westminster system, the Senate is more a partisan body than it is a representative of state interests. In the United States, the Senate, with its equal representation of the states, does provide a counterweight to the strengths of the larger states. But it is predominantly a nationally oriented legislative body that can no longer be seen primarily as a guardian of state interests. The second chamber in Germany, the Bundesrat, is indeed a powerful body that speaks directly to the interests of the Länder in national policy making. It is perhaps the strongest example of intra-state federalism in this group of countries. Yet even here, the fact is that the Bundesrat represents and is composed of members of the executive of Land governments. It is, therefore, more to be seen as an insertion of executive federalism into the national legislative process than it is a forum for elected legislators. The National Council of the Provinces (NCOP) in South Africa is modelled strongly on the German example. Its members are nominated representatives of the provincial governments. They play an important role in national legislation. On matters directly affecting the provinces, they vote according to the “mandates” they receive from the provincial legislatures, which are required to consider and debate national legislation. But here too, party interests tend to predominate, and provincial legislatures have developed little capacity to debate national laws. The result is that the NCOP has, at least so far, played little role in managing the national-provincial relationship. As with the other Westminster-type parliamentary systems – Canada and Australia – the most important contacts take place at the executive level. Both the German and South African cases also show that a strong provincial presence in the second chamber does not diminish the need for on-going executive mechanisms, such as the German bi-annual Conference of Minister-Presidents and numerous conferences of specialized ministers, and the numerous “MinMEC’s” and the regular meetings of Premiers and the President in South Africa. The Argentinean case is especially interesting. Here provincial governors appear to exert enormous influence over the conduct of representatives in the national Congress. Governors are politically secure compared to Members of Congress. Governors play an important role in nominations for Congress and often instruct “their” congressional representatives on how to vote. Hence policy success at the national level tends to require a consensus among provincial governors. The President must expend many of his resources to win state support for national initiatives. This pattern parallels that in Brazil, where again the national Congress is strongly penetrated by provincial interests. Commentators in both these countries suggest that the colonization of the center by the peripheries undermines the capacity of the national government to set national priorities and act decisively as a national government. There is an interesting contrast here with the debate in Canada. There, it is conventional wisdom that the single biggest institutional failure of federalism is the inability of its Senate to represent provincial concerns in national parliamentary institutions. Many proposals have been put forward to reform the Canadian Senate in order to strengthen its capacity as a forum for intra-state federalism. The Brazilian and Argentine cases suggest, on the contrary, that too much provincial influence over the center can have paralysing effects. The second dimension of legislative federalism assumes that most intergovernmental relations do take place among executives, but then asks how well elected members of federal and provincial legislatures are able to monitor, scrutinize, oversee, and debate how their governments are performing. This has been an important issue in Canada since the popular revolt against closed-door intergovernmental meetings on the constitution a decade ago. Reformers argue that there is too little parliamentary debate on the positions their governments will take into discussions, or on their results; that there are no legislative committees that oversee intergovernmental affairs, no regular reporting back to parliament, and so on. Judging from the studies presented here, such concerns are not as prominent in other federations. In Australia, Cheryl Saunders and Roger Wilkins note that neither the Senate nor the national House of Representatives plays a significant role in managing intergovernmental disputes, but they do not suggest that this is considered a problem. Members of the Lower House in South Africa seem unconcerned with federalism issues. In Germany, members of the lower house, the Bundestag, do become involved in the mediating committees necessary to resolve differences between it and the Bundesrat. In the Congressional systems, however, Congress plays a somewhat larger role, as it does in the legislative process as a whole. As we have seen, in Argentina Members of Congress frequently act on behalf of the state governors and in Brazil, the provinces have to some extent colonized the national legislature. In the United States, the equal representation of states in the Senate, and the localized bases of power of members of the House of Representatives might suggest a central role for Congress in intergovernmental relations. But, Kincaid concludes that “there is no political capital in intergovernmental relations.” Both Houses have committees on IGR, but they are of little importance. Congressional dependence on support from state and local governments drastically declined after the electoral reforms of the sixties. Congress has facilitated, rather than blocked, the growth of “coercive federalism.” The chief question for reform here is whether a greater and more effective legislative presence in the intergovernmental arena would help mitigate the “democratic deficit” often associated with executive federalism. • The balance of power among governments Is the relationship among governments a partnership among equals; or does it look more like a hierarchy of superior and inferior governments? Clearly either relationship requires intergovernmental relations, but hierarchy and equality are likely to lead to very different dynamics. In Canada, the predominant impression is one of relative equality between the two orders of government. The provinces, armed with extensive powers, bureaucratic and fiscal resources, and political clout show little deference to federal leadership. Indeed, they are zealous in defending provincial turf against real or perceived federal intrusions. The Prime Minister continues to chair the – increasingly infrequent – First Ministers’ Conferences. But other Ministerial Councils are frequently co-chaired by federal and provincial ministers. Provinces acting on their own have a welldeveloped network of intergovernmental mechanisms such as the Annual Premiers’ Conference, related ministerial councils and task forces, and regional groupings of premiers, in which Ottawa does not participate. These “Provincial/Territorial” mechanisms provide an opportunity to develop common strategies vis-à-vis the federal government on contentious issues, to forge compromises and provinces with varying resources and interests, and to share information in managing common problems. The absence of such processes in Brazil may contribute to the intense competition and “fiscal wars” noted by Ferreira Costa. However, the relative power balance between national and sub-national governments varies over time. The term most often used to describe intergovernmental relations in Canada in the 1940s and 1950s was “co-operative federalism,” a model that suggested that the federal government provided the policy and fiscal leadership, to which most (but not all) provinces deferred. By the 1960s, first with the rise of a more nationalist Quebec, and then with other provinces becoming more capable and self-confident, this acceptance of federal initiative sharply declined. The relationship became at once more equal, and more competitive. In contrast to Canada the gap in power, status and visibility between Washington and the state governments in the U.S. is very large. Kincaid notes that IGR in the country is at once “cooperative, conflictual, competitive, collusive and coercive,”but leaves little doubt that Washington is the dominant government. In the aftermath of the collapse of communism, many Russian sub-national units strongly challenged central dominance. The “treaties”that they signed certainly shows the absence of hierarchy, and the initial dynamic seemed strongly centrifugal as some powerful regions carved out an almost independent position in the face of disarray and weakness at the center. But under President Putin, the intergovernmental machinery has come under much stronger central control. The machinery of Australian intergovernmental relations is much like the Canadian, but overall, the states appear to exercise less autonomous power or influence over the centre than their Canadian counterparts. In Germany, Schnapauff sees a gradual extension of federal powers under concurrency and framework legislation, with the Länder remaining responsible for administration. It is difficult, however, to sort out relative powers in a system that places so much emphasis on co-operation. Despite some sections of its constitution that suggest a relationship of equality, South Africa is a much more top down federalist system than the other cases here. The centre has wide powers both to trump provincial legislation, and to intervene in provincial administration. It has almost total control of public revenues. Provinces have not developed a strong sense of themselves as independent political actors. It appears that many intergovernmental meetings involve the centre enlisting the provinces in central government priorities. In Argentina, Fohrig notes a recent trend towards decentralization, as “provincial governors have adopted an increasingly active role as autonomous power brokers.”“A process of gradual territorial pluralization of political power is underway.”Whether the recent economic and political crisis will reverse or accentuate this trend remains to be seen. The balance of power in Brazil seems to have varied greatly depending on the relative political strength of state governors and the national President. While Ferreira Costa describes the country as “one of the most decentralized in the world,” and notes that the most recent democratic constitution pushed political and fiscal decentralization further, he also concludes that President Cardoso has been able to engineer “a precarious political balance.” • The balance between co-operation and conflict Are intergovernmental relations characterized by a sense of shared and common values and purposes, which emphasizes the need for cooperation and for consensus? Or is the relationship a more competitive one, reflecting sharply different preferences and struggles for power between the central government and the states and provinces, and among the units themselves? Intergovernmental tensions over policy priorities and public finance are inevitable in any multilevel system. But there are some major differences in the relative levels of co-operation or conflict among these cases. The German model is predicated on the idea that the centre and the Länder act together. Hence co-operative governance and loyalty to the federation, Bundestreue, lie at the heart of the system. Decision-making at all levels stresses the need for cooperation and consensus. Most intergovernmental mechanisms require unanimity, and then become binding. In South Africa, too, the integrated pattern of governance emphasizes the need for cooperation. Chapter Three of the constitution, entitled “Cooperative government” calls on the three spheres to “co-operate with one another in mutual trust and good faith.” This injunction is strongly reinforced by the party system: the ruling African National Congress governs in seven of the nine provinces, and its national executive exercises considerable control over recruitment of provincial leaders. The underlying logic in the Westminster system is one of adversarial, competitive, majoritarian, winner-take-all politics. Despite frequent and eloquent calls for greater harmony and co-operation among the constituent governments, this pattern tends to be reflected in the conduct of intergovernmental relations in Canada and Australia. This is especially so when issues rise to the senior political levels, and thus engage questions of overall strategy, power and status. In both countries – and in others such as the United States – co-operation is greater lower in the administration, where officials are more likely to share common professional values and similar clienteles. With its strong regional and linguistic divisions, expressed through powerful provincial governments, the Canadian pattern appears to me more competitive and adversarial than the Australian. In the highly complex American system, Kincaid suggests that the intergovernmental relationship combines elements of co-operation and conflict, competition and collusion. He finds much resentment among the states against the “coercive federalism” of “un-funded mandates” and the like. But intergovernmental tensions do not dominate the political field to the extent that they do in Canada. In Argentina, national-state conflict depends greatly on whether or not there is divided government at the centre, with one party controlling the Congress and another the Presidency. A more systemic reason for intergovernmental competition, Fohrig suggests, is that as a result of the electoral system, and the distribution of population, the Presidency tends to represent urban interests and the Congress more rural and peripheral areas. Finally, in its brief history, the Russian Federation has witnessed high levels of conflict, as new institutions jockey for power and influence in the new political order. • The Settlement of Disputes In all federal systems disagreements are likely to arise, whether of fundamental issues of jurisdiction, or on more detailed matters of the administration of intergovernmental agreements, fiscal arrangements and the like. In all these countries, the courts are the final arbiter or umpire of the intergovernmental relationship. In several of them, such as the U.S., Canada, Australia, Germany and South Africa, judicial decisions have greatly affected the actual distribution of power. Yet in most federations, appeals to the courts are relatively uncommon. In South Africa indeed, Chapter Three on co-operative government specifically enjoins governments to avoid “legal proceedings against each other,” using the courts only as a last resort. Instead most countries rely on the informal processes of intergovernmental relations to work out differences In the United States and South Africa, the predominance of the central government is so strong that most differences will be resolved in favour of the exercise of national authority (although in the U.S. some recent judicial decisions –“the federalism five”– have restored a measure of state autonomy). In other cases, the absence of authoritative dispute settlement procedures means that resolution is primarily a political matter, and disagreements may be protracted. Where governments conclude formal agreements among themselves, there is a strong incentive to include specific dispute settlement mechanisms. This is the case, for example, with the Canadian Agreement on Internal Trade. The more recent Social Union Framework Agreement included the commitment to develop such a mechanism, but governments have been reluctant to bind their own hands, and thus little progress has been made. Germany, with its emphasis on consensus decision-making, and its unique integration of parliamentary and executive federalism in the Bundesrat constitutes perhaps the most fully developed system for dispute resolution. It was the model for the South African NCOP; and inspired many proposals for a House of the Provinces or House of the Federation to replace the Senate in Canada. • Intergovernmental institutions as decision-making bodies To what extent do intergovernmental bodies act as authoritative decision-makers in federal systems? The alternatives here range along a continuum. At one end intergovernmental deliberations are primarily about exchanging information and ideas, they provide a forum for discussion. In the middle are processes that emphasize bargaining, negotiation, and persuasion, but with the governments remaining responsible to their own legislatures and electorates for the actions they take. At the other extreme are intergovernmental institutions that can make formal decisions, binding on all the partners. This issue arises especially in the context of the parliamentary federations of Canada and Australia. In these countries there is strong resistance to assigning binding powers to any nonelected intergovernmental body. The resulting accountability of governments to each other would potentially undermine the accountability of each to its legislature. Nonetheless, as Cameron and Saunders and Wilkins demonstrate, in both countries intergovernmental accords and agreements in a wide variety of shared policy areas have been negotiated. Typically, however, they are careful to spell out that they to not infringe on the legislative powers of either order of government and are not judicially enforceable. In integrated federations, like Germany, on the other hand, where concurrency is common, where the Länder administer federal laws, and where homogeneity and consistency are highly valued, binding intergovernmental agreements are the norm. In Russia, binding agreements – at least until 1999 – took the form of bilateral treaties between the centre and a large number of constituent units, although according to Alexei Avtonomov their “legal-constitutional status” is unclear. The ad hoc administrative bodies in the presidential/congressional federations also do not appear to act as formal decision-making bodies. Intergovernmental arrangements developed in the national legislative process will be expressed in legislation or constitutional amendments. • The place of local government Are local municipal governments considered as a third order of government to be included in the broad structure of intergovernmental relations, or is the focus primarily on the state/provincial relationship with the centre? Are provincial-local relationships generally placed in a separate category; and do central government-local government relations play an important role? Most of the papers in this compendium focus on the relationship between national and state/provincial governments. A few federations (in this group Germany, South Africa and Brazil) give local government a specific space within the constitutional order. In others, such as Canada, Australia, and the United States, local governments are constitutional creatures of the state or provincial authorities. In the U.S. however, relatively strong linkages between cities and Washington, sometimes by-passing the states, have made local governments significant actors in the federal system. In Australia, representatives of local governments were included in the Council of Australian Governments, established in 1993. In Canada, on the other hand, provinces have tenaciously guarded their jurisdiction over municipal government, and have strongly discouraged the development of direct federal-local linkages. Local governments may be about to play a greater role in South African intergovernmental relations, as a recent restructuring and empowerment of local governments has enhanced their status. Some South Africans believe that with a dominant center, and stronger local governments, it is the provinces whose significance in the system might well decline. Given the importance of urban areas for economic growth and cultural dynamism, it is perhaps unfortunate that the study of intergovernmental relations and federalism and the study of local governance have developed as largely separated fields of study. Multilevel governance is increasingly a matter of the interaction of four levels: local, state/provincial, national and supranational. Another such gap in the papers presented here is the place of indigenous peoples in federal systems. In Canada, the movement towards “Aboriginal self-government” has been strongly shaped by the logic and values of federalism – even as it is also in tension with the practice of federalism. The same is true in the United States. Australia, Russia and Brazil, among the countries represented here, also have important aboriginal minorities, and the institutions and practices we associate with federalism may also prove useful in those countries. EXPLANATIONS As the preceding summary shows, the patterns of intergovernmental relations vary widely across countries. IGR is ubiquitous, but it takes many forms. What explains some of the differences? Some of the possible answers have been implied in the preceding paragraphs. • Institutional design of the larger framework The first set of explanations lies in the broader institutional structure or framework, within which federalism and IGR are embedded. Intergovernmental relations differ considerably between the parliamentary federations, and the Presidential/Congressional systems described here. In the former, intergovernmental relations take place largely in the form of relationships between the executives of the two orders of government. They emphasize “inter-state” federalism. In the latter, the major forces affecting the relationship between the center and the provinces are to be found in politics within Congress and in its relationship with the president. As the American, Argentinean and Brazilian cases show there are many specific intergovernmental mechanisms, but they deal mainly with relatively low-key administrative matters. The emphasis is on “intra-state” federalism. A second major feature of the broader institutional framework affecting IGR is whether or not states or provinces are represented in the second chamber of parliament. If they are not, as in Australia or Canada, then the burden not only of policy co-ordination, but also of broader political accommodation falls on the institutions of “executive federalism,” or “federal-provincial diplomacy.” In Germany a strong and effective Bundesrat integrates both legislative and executive federalism. In Argentina and Brazil, locally oriented national legislatures, often with strong ties to state governors, profoundly shape the pattern of intergovernmental relations. The electoral and party systems constitute a third institutional factor. If the electoral system, as it interacts with the underlying pattern of social cleavages, leads to strong representation of regional minorities in the national legislature and in coalition cabinets, then the task of accommodation is more likely to take place within the national legislature. This appears to be the case in Brazil and Argentina. As the case of Canada shows, however, a regionalized party system combined with the concentration of power in a single party cabinet undermines the integrative capacity of the national parliament. With an integrated party system in which national parties are able to win support across all or most regions, and in which national and provincial/state parties are closely linked, with considerable mobility of leadership from one level to another, much of the accommodation between center and regions will take place in the course of party politics and in the national political arena. This seems to be the case both in Germany and the United States. But where party systems are regionally divided as they are in Canada, or where different parties operate at national and provincial levels, then again the burden of national accommodation falls to executive institutions. In both Argentina and Brazil the party system and the electoral system have major effects on representation in Congress and on its relationship with the President. • The design of the federal system The next set of factors shaping the intergovernmental relationship has to do with how the federal arrangements themselves are designed. Most important here is the distinction between dualist or separated systems, in which each order of government is responsible for legislation and implementation in a specified list of powers, versus systems built on the principles of concurrency and shared responsibility. In the former (Canada, the U.S., and Australia) the emergence of de facto concurrency in many areas makes IGR necessary, but the mechanisms through which to conduct them tend to be later additions to the institutional framework. In shared systems like Germany or South Africa, intergovernmental relations are integral to the original design, and are thus built into it from the beginning. The division of powers (whether weighted to the centre or to the states/provinces, whether concurrent or separated, whether symmetrical or asymmetrical), fiscal arrangements and the distribution of resources and rules about paramountcy will also have major effects on the dynamics of the intergovernmental relationship. • Federal societies The effect of institutional arrangements depends greatly on the nature of the society in which they are embedded. In societies that are relatively homogeneous, and in which national identities are predominant (as in Germany, the United States, Australia, and, in some respects South Africa), primary political influence is likely to lie with the central government. In such societies intergovernmental relations are likely to be relatively free of conflict and public controversy. Where regional and territorially based ethnic or linguistic divisions are deeply entrenched, IGR is much more likely to become the arena in which competing visions of the country are played out, to embody a strong competition for power among competing orders of government, and to encompass a broader range of policy disputes, including, in Canada, the nature of the constitution itself. In the first group of countries, IGR, even if complex, is mainly about administrative co-ordination, a problem in public administration. Not so in Canada, Russia, or Argentina. Similarly, as the Argentine and Brazilian cases demonstrate, stark discrepancies in wealth between richer and poorer states and province can also increase conflict, not only between state and national governments, but also among states themselves. • The policy agenda These studies also show that the character of intergovernmental relations at any given time is influenced by the contemporary policy agenda. To what extent do the issues being dealt with by the system divide the country on territorial lines? Are the central issues ones that lie primarily within the jurisdiction of one level of government or another? Economic and social crises such as that Argentina is undergoing can have dramatic effects on IGR, as did the debt and deficit crises that faced countries like Canada and the United States just a few years ago. Globalization in its many forms is also having an effect on the conduct of intergovernmental relations in these countries – but whether its effects will lead to less or more cooperation, less or more centralization remain unclear. Finally, the values, commitments and ambitions of political leaders at every level can have important effects on the conduct of IGR. This has certainly been the case in Canada and Australia, two countries where the First Minister has enormous ability to shape the relationship. Patterns of IGR, then are largely a result of such factors external to IGR itself. In none of these countries are the institutions of IGR so strongly entrenched that they have a strong independent effect on the behaviour of political actors. They are reactive and responsive, rather than determinative of the character of the federation. This is not so say, however, that policy makers in each of these countries can ignore the task of improving their machinery for co-operation. ASSESSMENT How well do the varying patterns of intergovernmental relations described in the preceding papers perform? Two dimensions of performance are important. The first test is the ability to meet substantive policy challenges; the second is to meet expectations about democratic politics. • Policy Effectiveness To what extent does the intergovernmental process facilitate effective coordination among governments? To what extent does it minimize the costs of contradiction and duplication, or limit the transaction costs of achieving common goals? In other words, to what extent does it avoid the “joint decision trap?” This is not to suggest that the test of effectiveness is the ability to arrive at common policies, standards and norms in every policy area. It does suggest the necessity of ensuring that the results of intergovernmental relations are sensitive both to the need for variation in relation to regional preferences that is one of the central values of federalism, and to uniformity with respect to values common to the whole country. Each country will, of course come to its own conception of the appropriate balance. In some of these countries there appears to be a generally settled view as to the nature of the federation – Germany, the United States and Australia stand out in this regard. While disagreement remains on specific issues, their systems of intergovernmental relations are also relatively well established and stable. In Canada there are deeper disagreements about the nature of the federation – centralist or decentralist; symmetrical or asymmetrical – that make IGR more contentious and unsettled. This is also the case in Argentina and Brazil, though for somewhat different reasons. Settled arrangements have yet to emerge in the newest federations. In South Africa, there is still uncertainty about the utility and viability of the provincial administrations, and a debate about the relative roles of provincial and local governments. In Russia, centrifugal trends in the Yeltsin years have been replaced by centralizing trends under President Putin. The overall assessments of the effectiveness of IGR mechanisms and processes in our sample of countries reflect this portrait. Saunders and Wilkins conclude that in Australia “the blend of constitutional and cooperative governance has facilitated a dynamic and flexible federal structure.” In Germany, Klaus-Dieter Schnapauff states that “co-operative federalism promotes the functioning, performance and stability of the German Republic.” Kincaid sums up American IGR as “organized confusion.” But it works because of the absence of deep regional cleavages, the openness of the system to political forces at all levels, and because of formal and informal norms that constrain destructive behaviour. David Cameron, writing on the Canadian case, provides a mixed assessment – considerable recent progress towards a higher degree of collaboration, together with some doubt about the longer-term viability of this model, and some strongly expressed concerns about its policy implications from some affected groups. Fohrig is very dubious about the consequences of the Argentinean pattern IGR. The complex frameworks for decision-making and the lack of institutional means to mediate among stakeholders result in policy that “lacks insight” and is “short-lived and inconsistent.” He calls for “a new vision of intergovernmental relations” that can be the starting point for a path toward “the creation of a medium and long term development strategy.” In Brazil, Ferreira Costa, depicts a similar pattern of complex bargaining between President, Congress and the states, but he does suggest that movement and co-ordination are possible when the President is able to mobilize his own Congressional coalition and block opposing coalitions. Olivier concludes his chapter with a long list of proposed reforms of South African IGR structures in order to equip that country with the means to address more effectively its extensive policy agenda. Alexei Avtonomov makes similar points even more strongly for the Russian Federation. In both cases, along with those of Brazil and Argentina, serious questions can be asked about how intergovernmental relations can assist, rather than hamper, them in meeting their challenging policy agendas of socio-economic transformation and development. • Democracy and transparency The second set of criteria for assessing or evaluating intergovernmental mechanisms has to do with democracy. How open, transparent, responsive, and accessible are the mechanisms established for the conduct of IGR? The more important these processes are in the political life of a society the more it is necessary to subject them to the same democratic tests as are applied to other institutions. The question of the possible “democratic deficit” is common to all systems of multilevel government – indeed it was invented to address problems in the European Union. Among the countries represented in this collection, the question appears most pressing in Canada. This is partly because of the executive dominance in the system, and the closed character of its executive federalism, which has received much criticism in recent years. It is also because, as an “add on” to Canada’s institutional framework, the mechanisms of IGR do not fit easily into traditional conceptions of responsible government. Cameron sees the reconciliation of democracy and IGR as perhaps the most pressing concern for Canada. Similar questions might be raised about Australia, and South Africa but they appear not to be a major issue of controversy in those countries. Germany, with its bureaucratic system, and its high degree of “interconnectedness and intertwining,” raises questions about the dangers of blurring lines of responsibility, the avoidance of political responsibility, and the removal of the line between government and opposition. Since IGR in the United States is primarily a matter of administrative co-ordination, rather than debate about fundamental issues, as in Canada, again a “democratic deficit” in IGR does not loom large. In Argentina, Brazil, and Russia democracy with respect to IGR can only be seen in the larger context of democratization in those countries. CONCLUSION Clearly, there is no single model of “good” intergovernmental relations. To a great degree the patterns described in each of these cases responds to the particular circumstances that face each country. It is impossible to transplant institutions holus bolus from one setting to another and probably dangerous even if one could. It is nonetheless true that the experience of other countries described in this set of studies suggests many possibilities for cross-national learning, and provides a useful starting point for assessing and improving one’s own institutions and practices.