1.1. Hard Realities

These changes bring with them at least three hard political realities. This is familiar ground but, again, it is important to be clear on what is at issue, because it is so often discussed in vague terms.

The first new reality is that the power of national governments, while it remains very real and very important, has declined noticeably, especially in governments' ability to regulate market activity and protect social welfare.¹

This brings us to the second hard political reality. In an increasingly borderless world, local communities everywhere are less protected by national government from the consequences of international economic competition and less likely to benefit from senior-government assistance than was the case in the 1970s and earlier. At the same time, each community is much more directly in competition with every other community than ever before.

In the past, regions that benefited less from market activity looked to the federal government for job creation and regional development programmes. Some of these initiatives have been taken out of the hands of government by the fact that they are deemed to be unfair trade practices in the North American Free Trade Agreement or by the World Trade Organization. Others have become victims of government cutbacks.

The third new reality, therefore, is that local communities—meaning municipalities of all sizes as well as metropolitan areas—have been thrown more than ever before upon their own resources. It has become the normal way of doing business for every municipality or metropolitan region to write its own economic development strategy and create an agency or agencies to implement it. Often such strategies take the form of joint ventures involving all the municipalities in a metropolitan region. Each municipality and each region has its own particular mix of resources, locational advantages and disadvantages, human capacities and shortcomings. As global market competition intensifies, it becomes more and more important for each community to assess its own potential strengths and design its economic development strategy accordingly.

The implications of these developments are momentous and have not been given the attention they deserve. There are at least two ways of interpreting what has happened. One is to take the view that the economic, technological, social and political changes constituting globalization have had the cumulative effect of de-centring the economy, so that a national government, which once was able to make economic and social policy on the premise that it was managing a single national economy, must now recognize that it is actually managing a series of discrete urban economies (Magnusson, 1996; Barnes and Ledebur, 1998; Clarke and Gaile, 1998).

An alternative view is that globalization has only made more evident what has always been the case. As early as 1969, Jacobs was arguing that cities are the real source of economic growth and ultimately of a society's wealth. In 1984, she made the case that national accounts are a misleading guide to economic policy, and have perverse consequences, because they mask the crucial differences between the economies of different cities, leading to national policies that favour the dominant city and harm the economies of the rest. Similarly, a colleague and I have made the case that immigration policies have too often been based on conditions in such rapidly growing centres as Toronto and Vancouver, and have thus failed in many other cities (Leo and Brown, 2000).

Whether we take the position that globalization has projected us into a new and different economic and political world, or join Jacobs in arguing that we have long paid a high price for our failure to appreciate the importance of city economies, the case for social and economic policies pitched to differences among urban-centred regions is compelling. But there is a considerable distance between agreeing on that proposition and knowing what to do about it.

1.2. An Unfettered Community?

In the twenty-first century, then, national governments are becoming less able to sustain the economies and the social safety nets of local communities while, at the same time, there are compelling reasons for cities to evolve economic development strategies and social supports specifically designed to deal with their own, unique set of problems and possibilities. Even as national governments weaken, the focus of economic development shifts ever more obviously from the nation to the urban-centred region.

Since the relative positions of cities and national governments have shifted markedly, it follows that it is time to take a fresh look at the role of each, and the relations between them. Both the urban literature and the Canadian federalism literature have arrived at approximately this same point by different routes.²

What are the consequences for governance? One answer comes from the charter city movement, based in Toronto. Its position is spelled out in a model framework for a city charter in which the affected city is declared to be “an autonomous and accountable order of government.” The model charter binds the province to an agreement to consult the city before taking actions that affect it, allows the city to negotiate directly with the federal government, and sketches out “a dispute resolution process to be used by City and provincial officials if any future disagreements arise over the meaning of the Charter” (Big City Mayors' Caucus, 2002). Such provisions would entail a revocation or a voluntary renunciation of the constitutional authority of provinces over municipal affairs. Clearly, the charter city concept is grounded in a demand for radical change in intergovernmental relations.

If the charter city argument ever approaches the threshold of political viability, it will encounter resistance, not only from provincial governments unwilling to relinquish a substantial share of power, but also from many who will question the democratic bona fides or the competence of municipal councils, and from such commentators as Castells, who argues that “local/regional autonomy reinforces territorially dominant elites and identities, while depriving those social groups who are either not represented in these autonomous government institutions or else are ghettoized and isolated” (1996).

More significantly, it will become enmeshed in the ongoing debate over the race to the bottom. Critics will point out that if autonomous communities were to be set free to fend for themselves in the wild west of an unfettered global economy, the casualties might well outnumber the successful contenders.

In fact, for every commentator making the case for city charters, there are probably several expressing dismay over the effects of government cutbacks and the downloading on low-income communities and on the integrity of the social safety net, and calling for the federal government to become more involved in the setting of standards and the financing of programmes. Greater centralization of power probably has more support than city charters would.

2.1. A Federalism Lesson for Students of Local Politics

The federalism lesson comes from Courchene, who rejects the hierarchical, either/or position, which he calls federalism as structure, and argues that the alternative to it, federalism as process, is a Canadian tradition.³

In a wide-ranging and perceptive essay, Courchene comments unfavourably on the Canadian preoccupation with constitutional change and with symmetry in the relations between provinces and the federal government that marked the period from the Meech Lake Accord through the collapse of the Charlottetown Agreement. He argues that Canadians have long practice in the regulation of the relations between federal and provincial governments by means that avoid the rigidities of constitutional provisions and circumvent the demands for symmetry that inevitably accompany constitutional discussions.

“Canadians,” he writes, “have displayed a rare genius in accommodating their political structures to emerging internal and external forces…. [T]hese innovations were the result of process, not structure, although in many cases they were tantamount to a de facto alteration of the division of powers in the federation” (1995: 11).

The suggestion is not that the constitutional division of powers is being in some way subverted, but that, within the division of powers, creative avenues of policy making are being found that involve co-operation between governments and that allow for policies which take account of the differences between the regions of the country. Courchene underlines the important role that structural arrangements play in the accommodation of difference by pointing out that asymmetry has always been a feature of Canadian government in constitutional provisions that vary from province to province. Quebec's civil code is an example.

However, a more significant role in the achievement of flexibility and the accommodation of difference has been played by federal-provincial administrative arrangements that allow for numerous differences in the treatment of different provinces, and do it within a nexus of negotiation and compromise unencumbered by the rigidity of constitutional provisions. Among the examples he cites are medicare, the Canada Assistance Plan, the decentralization of income and corporate taxation (with asymmetrical provisions for Quebec, Ontario and Alberta), negotiated and differential pension provisions for different provinces, and equalization.

Courchene argues that these federalism-as-process arrangements have, over time, worked in the direction of growing provincial self-determination. In many ways, his analysis parallels those of commentators who advocate a re-evaluation of the place of cities in national politics. He sees growing provincial self-determination as being related to the advance of globalization, which he prefers to call the new techno-economic paradigm (1995: 12). He stresses the enhanced importance of regional economies in a world of global trade and information flows, though in this part of the discussion, his focus is on world cities, with other urban-centred regions fading into the background as an ill-defined hinterland. This is a critically important omission, for reasons that will become clear when we take a look at what we can learn from the study of local political economy.

Thus, the study of Canadian federalism—which up to now has been synonymous with the study of federal-provincial relations—involves, in part, gaining an understanding of a constitutionally mandated division of powers. But it also offers an array of examples of voluntary arrangements short of constitutional change that can secure national objectives while taking account of regional differences.

Since these arrangements are worked out in a political setting through negotiation and compromise, instead of a constitutional/legal one, they have the added virtue of being flexible, and readily adaptable to changing circumstances. They are not a substitute for a constitutional assignment of responsibilities, nor do they relieve us of the need to consider constitutional changes when circumstances require them, but they offer a toolbox of flexible approaches to the accommodation of regional difference within a national framework.

If these tools are usable in a federal-provincial context, can they not also include cities? The federalism literature shows an awareness of the need to ask that question, especially in an increasingly borderless world, but since students of federalism have, in practice, been expert in federal and provincial, rather than local, governance, they show signs of being stymied by it. It seems reasonable, therefore, to look for answers in the literature on urban politics.

2.2. A Local Political Economy Lesson for Students of Federalism

If asymmetry is good governance, regional economies are growing in importance, and world cities need to be understood and treated differently from other communities, why should the same not be the case for all cities in a global economy? If the differences between provinces, separated, as they often are, by economically irrelevant boundaries, justify differential, flexible and asymmetrical arrangements with the federal government, the economically very real differences between cities provide an even better justification for such arrangements.

Courchene's treatment of cities is typical of current literature on federalism and the Canadian state. There is a growing recognition that cities have become more important to our understanding of the state, but a sense of puzzlement remains about what to do next, how to bring cities into the analysis. That puzzlement persists in a more recent article by Courchene (2005). Here he revives his earlier suggestion that all we need concern ourselves with are global city-regions, but he appears to redefine all major metropolitan areas in Canada as global.

However, as he acknowledges, the focus of the article is on Ontario, and he is explicit in stating that Toronto is the only Ontario city-region that need concern us. At the same time, he lists Halifax and Winnipeg among Canadian global city-regions, but does not address the question of what then becomes of the Ottawa/Gatineau metropolitan region, which is bigger, and arguably at least as important as either Winnipeg or Halifax (Courchene, 2005: 3, 4). By limiting the discussion to global city-regions, whatever the definition, Courchene barely scratches the surface of the role of cities and local communities in the twenty-first-century state. In this section, I offer some evidence and analysis of local governance and local political economy to flesh out the federalism literature's dawning awareness of the economic and political importance of cities. I begin by updating Courchene's 1995 analysis of the federal-provincial nexus.

Courchene's examples of federalism as process antedate the mid-1990s, and they are not exhaustive. A more recent example is the Social Union Framework Agreement (SUFA), a shared federal-provincial commitment to improve Canadian social programmes while also supporting the ability of Canadians to relocate from one part of the country to another and enhancing the accountability and transparency of social programming (Federal/Provincial/Territorial Ministerial Council on Social Policy Renewal, 2003). SUFA is modelled on the National Child Benefit, a programme built around the cooperation of the federal government, provincial and territorial governments, and First Nations, and which tries to ensure that parents on welfare will not suffer a financial penalty if they find employment (Federal/Provincial/Territorial Ministerial Council on Social Policy Renewal, 2003).

Another example of federalism as process is a series of federal-provincial and federal-territorial agreements whereby federal employment insurance funds are made available to support unemployed people taking advantage of provincially or territorially generated programmes designed to prepare them for employment (Lazar, 2002). These federal-provincial-territorial accords do much to address the differences between provinces within a flexible framework, but none of them offer any assurance that the differences between cities will be recognized in the implementation of federal government policies.

In a federal-provincial agreement, it remains in the purview of the provincial government to decide whether federal money will be used to address such problems as Winnipeg's decaying inner-city housing stock and deteriorating infrastructure, or whether to focus on the opportunities for employment and community development that can be gained from programmes to deal with these problems. On the face of it, it would appear that if we consider the differences among urban communities to be as important as those among provinces, federalism as process has failed Canada's cities.

A closer examination, however, shows that this is not the case and that, in fact, federalism as process for cities has been with us at least since the late 1970s. But we will not find it by confining our analysis to global city-regions. An early example of what I will call deep federalism-as-process was the Neighbourhood Improvement Programme (NIP), a federal government scheme aimed at the renovation of public facilities in declining neighbourhoods. The NIP became a community development tool through the simple expedient of a requirement that a plan for neighbourhood renewal be preceded by and based upon a public participation process in each targeted neighbourhood (Lyon and Newman, 1986; Jamieson and Smith, 1979). The NIP, therefore, was structured to try to ensure that it would be implemented in such a way as to respect the differences both between cities and between individual neighbourhoods.

A second example, unique to Winnipeg, was the Core Area Initiative (CAI), an 11-year, tri-level arrangement for the social, economic and physical renewal of Winnipeg's inner city, which was administered by a secretariat located in Winnipeg and responsible to all three levels of government. Such tri-level agreements have been all but institutionalized in Winnipeg, as the CAI was followed, respectively, by the Winnipeg Development Agreement (WDA) and the recently concluded Winnipeg Partnership Agreement (WPA) (Winnipeg Core Area Initiative, 1992; Canada, Manitoba, Winnipeg, 1998; Manitoba Intergovernmental Affairs and Trade, 2004).

The approach they represent has also migrated west, in the form of the Vancouver Agreement, a wide-ranging accord that drew in a large number of partners from all three levels of government, with efforts focussed on economic development, the health of residents and public safety in Vancouver's Downtown Eastside (Bradford, 2005; Bakvis and Juillet, 2004; Macleod Institute, 2004).

Another example of deep federalism as process was the urban development corporation, a recurring theme in federal-provincial-local relations in Canadian cities over the past quarter-century. Examples are Harbourfront in Toronto, Canada Harbour Place in Vancouver, Le vieux port de Montréal and de Québec and the Forks and North Portage corporations in Winnipeg. All of these projects were pitched to the specific circumstances of each city.

To be sure, the degree of genuine local involvement in decision making may have varied from case to case, since some were federal Crown corporations, but an indisputable case of deep federalism is that of the Winnipeg corporations, now merged into a single entity, the Forks/North Portage Partnership. The Forks and North Portage corporations, as well as their successor organization, were and are governed by boards, with equal representation from the three levels of government (Leo and Fenton, 1990: 185–206; Leo, 1995).

A student of traditional intergovernmental relations—in practice, federal-provincial relations—would not likely have known much about the NIP, the CAI and urban development corporations, because such matters were consigned to the sub-field known to the Canadian Political Science Association as “local-urban politics.” However, in a borderless world, as I have argued, it becomes increasingly evident that cities are the main economic engines and that each of them requires a different set of governance arrangements, one suited to a city's particular mix of economic, social and political circumstances (see section 1.1). Respecting local difference becomes a matter of national importance and perforce the study of the governance of cities becomes part of the study of federalism.

By exploring some of the mysteries of local-urban politics, we have learned that deep federalism as process has been with us for some time. It has not been as widely practiced as conventional (federal-provincial) federalism as process, and we do not have as many models to build upon, but we have enough to suggest that it is possible to pursue policies that are truly national and yet fully take into account the very significant differences between regions and communities.

But the literature on urban political economy has more to offer than just a few Canadian examples of deep federalism-as-process. It is global in scope and has gone some distance toward building an understanding of how the governance of local communities fits into the bigger picture of intergovernmental relations, and politics generally. Interestingly, two Canadian scholars—one of whom has not been associated with Canadian local-urban politics—have been in the forefront of this literature.

2.3. Unbundling Sovereignty

The NIP, the CAI and urban development corporations are individual instances of a larger phenomenon that David Elkins evocatively calls unbundling sovereignty. In his book, he challenges his readers to imagine a world beyond sovereignty, one of “non-territorial federalism.” In this scenario, governments are not pyramided in a neat hierarchy and spheres of jurisdiction are not exclusive. Governmental organizations, some with a territorial base and others without one, co-operate in carrying out some activities and go about others separately, as dictated by individual or mutual interest (Elkins, 1995).

Once we pursue this line of thought, and consider such examples as the NIP, the CAI and medicare, it becomes obvious that hierarchies of sovereignty never were as tidy, or jurisdictions as airtight, as many imagined them to be. The picture that emerges is a complex, multi-faceted one that even the most astute observers are still struggling to understand.

European commentators identify a number of concrete political and administrative changes that comprise the shift from what has traditionally been called intergovernmental relations in Canada, to what is now generally referred to as multi-level governance, or re-scaling.⁴

A second change is functional specialization, as special-purpose authorities, public-private partnerships or various other flexible arrangements involving public, private and third-sector bodies take over some of the functions that, in the past, might have been performed by relatively monolithic general-purpose governments.

A third change, implicit in the first two, is a shift from government to governance, following from a recognition that any given public good may be supplied, or any publicly necessary function performed, in a variety of ways—that it need not necessarily be added to the portfolio of a government department. In addition, the idea of governance embodies the belief that government departments have to learn to function in more collaborative and flexible ways in their dealings with superiors and subordinates (vertical administration), in co-ordination of their activities with those of other departments (horizontal administration), and in the development of joint administrative initiatives with other levels of government.

Finally, a change that is implicit in the idea of unbundled sovereignty, as well as deep federalism, is the notion that if we are to rethink sovereignty with any degree of seriousness, we need to do more than just rethink governance at global, national and regional scales. We must go a step further, and also look at local communities, as the NIP's creators did, right down to the neighbourhood level, and view them as potential participants in multi-level governance and in the mosaic of asymmetrical governance. Warren Magnusson (1996) has been particularly insightful in exploring this dimension of politics.

With the exception of the last one, these changes have been given a good deal of attention in a variety of literatures. Separate studies have been undertaken on governance versus government (reinventing government), horizontal and vertical administration, public-private partnerships, and so forth. There is also extensive, grand literature on globalization, global cities and regulation theory. But the attempt to draw them together, treat them each as parts of a larger whole, and try to develop an encompassing profile of twenty-first-century governance, is a recent phenomenon.

Two of the pioneers of this research, Warren Magnusson (1996) and Elkins (1995), are Canadian. It has become a cliché to call Canada the first post-modern state. (A Google search on July 5, 2005, using the terms “Canada” and “first post-modern,” produced 625 hits.) The seminal work of Magnusson and Elkins provides some support for the cliché. But it is easier to propose theories about unbundled sovereignty than it is to make a concrete reality of it. The real test of Canada is less in what its scholars say, than in what its decision makers and citizens do.

Political leaders and administrators must master the devil that notoriously lurks in the detail. How can any given programme be made responsive to both national priorities and local difference? What is the role of each level of government? How will an appropriate level of public involvement be secured, how will disputes be settled and tensions managed? Deep federalism, if it is to be accomplished, will reside not in theory, but in policy and implementation. In the final section of this paper, we undertake an overview of some of the findings of a series of studies of the political and administrative realities of deep federalism.

As we venture into the practicalities of multi-level governance in a borderless world, a definitional shift is in order. So far in this discussion, I have been using the word “community” to refer to municipalities and metropolitan areas, but Elkins's concept of unbundled sovereignty, as well as Magnusson's de-centring of the state, calls our attention to the fact that, just as there is a demonstrable need for differential governance in different cities, there are good reasons why neighbourhoods, or communities defined in other than spatial terms, need to be governed in distinct ways.

The need for differential governance of different neighbourhoods is easy to understand. The unique character of particular neighbourhoods, such as the Old South End in Halifax, Montreal's Milton-Parc neighbourhood, Toronto's Rosedale, Winnipeg's West Broadway neighbourhood and Vancouver's Downtown Eastside is obvious to anyone who has even a passing acquaintance with them. Each of these neighbourhoods has distinct characteristics that call for distinct governance arrangements. Everyone familiar with Toronto and Vancouver knows—indeed, takes for granted—that Rosedale and the Downtown Eastside are not governed in the same way.

But differential governance for different communities is not necessarily defined in spatial terms. One of the most successful ventures in deep federalism in a series of case studies I have conducted is Manitoba's provincial nominee programme for immigrants. A colleague and I concluded that the main key to its success was the provincial government's decision to devolve much of the responsibility for the programme's administration to non-governmental organizations representing such groups as Winnipeg's Jewish community, which sought to rescue Jewish people from Argentina's economic collapse, and the Mennonite communities of Winkler and Steinbach, where German-speaking immigrants were in demand (August and Leo, 2006). In those cases, the relevant communities were ethnic and religious, not spatial.

If we challenge ourselves to think in terms of unbundled sovereignty, and to emphasize process over hierarchy in our understanding of governance, our concept of community must extend beyond metropolitan areas and cities to neighbourhoods and other communities, defined according to the boundaries these communities implicitly draw by the way they understand themselves, not according to anybody's preconceived notion of how governance ought to look.

3.1. Local Capacity

In all seven communities, local service providers and other stakeholders were involved in the programmes under study. The researchers for this study, therefore, were well placed to gather information about the stakeholders and evaluate their abilities. In all but one of the seven communities, we found stakeholders who not only had the expected intimate understanding of the situation in their community, but were also well versed in the literature, knowledgeable regarding experiences in other communities relevant to their area of interest, and entirely capable of organizing themselves to study options, formulate priorities and implement them.

This was particularly conspicuous in the Winnipeg Infrastructure Renewal Demonstration Project, where local participants gave every indication of having a keener eye for policy problems and opportunities than policy makers in either of the senior governments. Welfare-to-work, or workfare, which, since the 1990s, has appeared in various North American jurisdictions in different forms, has been widely criticized as forcing welfare recipients to accept low-paid, possibly make-work jobs that do not offer useful job training or prospects for the future.

Before the issue had even appeared on the national agenda, in the last half of the 1980s, Winnipeg welfare officials were experimenting with welfare-to-work schemes that tried to improve on conventional workfare by operating on incentive rather than compulsion, by securing the completion of needed work (projects the City of Winnipeg could not afford to undertake), and offering reasonably well-paid jobs that provided useful job experience. A Dutch elm disease control programme that provided experience relevant to work in forest management, launched in 1986 with provincial and municipal funding, was sufficiently successful to encourage Winnipeg officials to lobby for tri-level funding for a programme in which welfare recipients were selected for jobs repairing roads, back lanes and sidewalks. Federal and provincial governments were asked to contribute the money they saved on welfare payments.

Winnipeg secured a two-year commitment for 1994–96 from the other two levels of government and, as a colleague and I show (Leo and Andres, 2006), the project achieved impressive success in its first year of operation before being cancelled by cutbacks. The federal government was responsible for the cancellation, despite the fact that it had actually spent a little less on the programme than it would have paid out in social assistance contributions. In other words, though the programme had saved the federal government money, it was cut back in an effort to save money.

Subsequently, the provincial government took over the welfare function and, as Andres and I show (2006), allowed the city's programme to expire while substituting precisely the kind of compulsory programme, offering low-paid, dead-end jobs, that has exposed workfare to widespread criticism. In this instance, therefore, Winnipeg policy makers gave every indication of being considerably more perspicacious than their provincial and federal counterparts.

It is not my argument, however, that local stakeholders and officials are more astute than their federal and provincial counterparts, only that they are perfectly capable of thinking for themselves, have access to much the same body of information and analysis and, in addition, have the advantage of being intimately familiar with the situation in their locality. Our studies showed that federal and provincial politicians and officials did not always perform well, nor did they always perform badly. The same was true of local politicians, officials and stakeholders.

In the first three-year term of the National Homelessness Initiative, all parties were handicapped by the fact that, though the problem to be addressed was homelessness, the creation of housing was not one of the items the federal government was prepared to fund. For reasons that I will consider in section 3.2, the federal government was prepared to fund emergency shelters and services to street people, but not housing renovation or construction (Leo and August, 2005; August and Leo, 2006; Leo and Friesen, 2006). As a result, the programme was doomed to be less than satisfactory in all three cities, despite a community planning process that was supposed to ensure that the programme (or at least its SCPI component) would be responsive to the particularities of each locality.

In Winnipeg, this fundamental flaw was exacerbated by the fact that federal officials handled the community planning process badly. In response, local stakeholders organized themselves and produced a much more intelligently conceived community plan—though one that was less clearly explained than it might have been. (In our study, we suggested that a fraction of the resources devoted to the federal government's bungled community planning exercise could have paid for editorial help to enable the stakeholders to produce a better-written account of their priorities.) In any case, in the first three years of Winnipeg's SCPI programme, the implementation process was rife with friction, but apparently not because of any significant lack of community capacity (Leo and August, 2005).

A more co-operative spirit prevailed in both Vancouver and Saint John and in both cities, as in Winnipeg, local stakeholders were able to produce and implement workable community plans. However, in Vancouver, as in Winnipeg, a significant number of stakeholders questioned the value of the planning process on the grounds that they already knew what was needed and were organized to provide it (Leo and Friesen, 2006; Leo and August, 2006). In all three cities the evidence left no question that local stakeholders understood both the local and national policy contexts and were perfectly capable of setting priorities and providing programming responsive to them.

The immigration and settlement case studies presented a somewhat different picture, because those programmes were a product of conventional federalism-as-process federal-provincial agreements, rather than direct interaction of the federal government with local communities. They did, however, represent an attempt at deep federalism, in the sense that the clear intention of the agreements was that immigration and settlement policy be tailored to the specific circumstances of different communities.

However, since the agreements were between federal and provincial governments, the question of whether that intention was translated into action rested very much upon the policies of provincial governments. The situation looked very different in each of the three cities covered in my study. In Manitoba, the primary objective of the agreement was to provide for an increase in the number of immigrants, both to address a number of critical labour shortages and to support population growth. To that end, the agreement allows Manitoba to nominate a specified number of immigrants each year for immigration to and settlement in the province.

Making the programme work was a learning process, especially since Manitoba is the provincial nominee pioneer, but a colleague and I concluded that the programme has proven impressively successful and that the main reason for its success is the provincial government's decision to work closely with community groups that had an interest in bringing in immigrants and to delegate a substantial share of the responsibility for administering the programme to them. In this case, the competence of local stakeholders is clearly in evidence (Leo and August, 2006).

In Vancouver, where there is no inclination to increase the already large number of immigrants, the focus of the federal-provincial agreement was on managing the settlement of those immigrants in a manner appropriate to local circumstances. However, unlike the situation in Winnipeg, the implementation of the agreement was fraught with conflict. Local service providers argued that the provincial government was diverting funds intended for settlement away from those areas where it was most needed. It was also alleged that the government was unreceptive to community participation in policy making and implementation, indeed, was ideologically averse to it. However, a colleague and I concluded that none of this reflected adversely on community service providers. On the contrary, we found evidence of a high degree of professionalism and a well-developed capacity to innovate as they carried on under often difficult circumstances (Leo and Enns, 2006).

In only one of our cases does the evidence suggest any lack of competence on the part of local service providers, for an interesting reason. In Saint John, as in Winnipeg, the primary objective was to bring in immigrants, but Saint John did not enjoy Winnipeg's success. A major reason—and one that could not be put down to lack of competence in the stakeholder community—was the predominance of people of European ancestry in the population. The relatively few immigrants who came found it difficult to fit in, and were often disinclined to stay.

But there was also a shortcoming in the local capacity in Saint John. Members of the stakeholder community proved unable to shake themselves free from a long-standing split within the community that rendered it incapable of pulling together a set of agreed-upon policy options and priorities. In that case, even with the best will in the world, it would have been difficult to find a way to ensure that federal government programming would benefit from community input, and ultimately to ensure the success of a programme premised on deep federalism.

Our findings suggest, however, that such cases are out of the ordinary. In six of the seven case studies, insofar as the federal government fell short in drawing on local wisdom to tailor priorities to local circumstances, it was not for want of appropriate local resources. In all those cases, local officials and community stakeholders came equipped with policy expertise, capable of providing usable policy advice to the federal government regarding options and priorities. Their input was likely to be as sophisticated as any available in Ottawa, and better informed regarding local circumstances.

3.2. Sources of Community Difference

There is no easy way to understand community difference, no simple set of generalizations that will allow us to say that a community of type A has characteristics B, C and D, while a community of type E has another set of readily definable characteristics. If there were, there would be no need for deep federalism. The federal government could develop a different policy model for each of a finite number of well-defined community types and administer everything from the centre. But there is nothing finite about community difference.

Despite that, our research has uncovered one variable that seems particularly robust, and our findings regarding this source of community difference confirm the results of earlier research (Leo, 1994; Leo and Brown, 2000; Leo and Anderson, 2005). A very basic reason why different cities need different policies is population growth rate. Cities with rapid population growth face a very different set of problems than cities that are growing slowly. This is obvious in both of the policy areas our research touched upon. An exploration of the importance of this source of difference in both policy areas helps us to better understand the need for deep federalism while providing insight into some of the problems posed by differences in growth rate.

Immigration is always a sensitive political issue in Canada, and it has been especially so in recent years, as immigration legislation has undergone a series of hotly contested revisions. Throughout these changes, the government has been under pressure to limit immigration, on the basis of fears that immigrants will place undue burdens on the social safety net and that they will take jobs from Canadians.

Whatever the merits of those arguments—the case against immigration is less than compelling—a point that has been frequently overlooked is that immigration has very different impacts on different communities. Much of the controversy surrounding immigration is centred in major metropolitan areas, especially such growth magnets as Toronto and Vancouver. In Toronto, much is made of fears that the city will attract large numbers of immigrants with limited skills, many of whom, it is feared, will end up a burden on the state, and perhaps become involved in criminal activity. In Vancouver, there has long been controversy over allegations that Asian immigrants are driving up the cost of housing.

If such arguments have any substance at all, they are relevant mainly for the few metropolitan areas in the country with booming economies and high housing costs. In our research, the clearest contrast with Vancouver and Toronto is Winnipeg, a slow-growth centre that is not even remotely in danger of becoming inundated by large numbers of any population. By the same token, the city is an ideal location for people, especially those with limited resources, who are looking for a stable community and a chance to make a future for themselves and their families: a large stock of affordable housing; some decent schooling at all levels, even in poorer neighbourhoods; and, for people from dozens of different countries in Latin America, the Caribbean, Africa, Southern and Eastern Europe and Asia, a supportive community environment.

Cost of Living and Housing: Comparison, 2001

The benefits Winnipeg offers immigrants are matched by the advantages their influx holds for a city suffering from labour shortages and badly in need of more residents in declining older neighbourhoods. The Manitoba government has been aware of the potential benefits of immigration in a slow-growth jurisdiction since at least the late 1970s (Leo and August, 2006), but a federal government response was slow in coming. Thus, until the late 1990s, Winnipeggers were treated to the spectacle of Torontonians bitterly complaining about immigrants while national policies denied Winnipeg the immigrants it needed. Immigration policy, therefore, provides an excellent example of the importance of deep federalism. A uniform national immigration policy is simply counter-productive, for reasons that may be intimately connected with urban growth rate. A slow-growth city like Winnipeg may be looking for more immigrants, even while rapidly growing cities are struggling to cope with the influx they already have.

Housing and homelessness. Similar observations can be made about growth and housing. Rapidly growing Vancouver, typical of cities in similar circumstances, suffers from runaway housing prices, prices high enough to pose serious problems for the middle class and to drive some poor people into the streets. Winnipeg and Saint John, meanwhile, have much more affordable housing. A Statistics Canada comparison of salaries and housing costs for Vancouver and Winnipeg gives some sense of the scale of that contrast.

With housing cost differentials that dwarf differences in income, it is small wonder that Winnipeg has less absolute homelessness—the social service term for life in the streets, under bridges, in parks or in shelters—than Vancouver and Toronto. In Vancouver, homeless censuses produced a total of 1049 in 2002 to 2112 in 2005. In Toronto, according to Hulchanski (2000), the average number of people using emergency shelters on any given night was 4900 in 2000, 4600 in 1999 and 2400 in 1992.

Meanwhile, a report titled “A community plan on homelessness and housing in Winnipeg” (2001), prepared by the Social Planning Council of Winnipeg and representing the views of 36 community groups involved in service delivery to homeless people, did not attempt a count of the absolutely homeless. Rather, in a carefully thought-out strategy for dealing with homelessness, the focus was not on street people but on what service providers call the relatively homeless: people who are paying far more than they can afford for housing, or are living in seriously inadequate shelter (Leo and August, 2005).

Why? In Winnipeg, as in other slow-growth centres, while the numbers of street people are not as overwhelming as those in Toronto and Vancouver, the numbers of people in desperate need of housing that is both affordable and conducive to stable family life is nevertheless very substantial, because low housing costs undermine the incentive for home maintenance. The result is relatively ready availability of a great deal of ramshackle housing, and a stakeholder consensus that the priority must be affordable housing.

The need for deep federalism here is at least as obvious as it is in the case of immigration, but in the National Homelessness Initiative, well-intentioned attempts to accommodate community difference went awry. A public outcry, which led to the formation of the NHI, originated primarily in Toronto and reached a fever peak in 1998. In the winter of 1995–96, three homeless men froze to death on Toronto's streets. Those tragedies, together with the highly visible suffering of large numbers of others, led to a demand for action by the Toronto Disaster Relief Committee, the Toronto Star, Mayor Mel Lastman, and others.

By 1998, there was widespread agreement that homelessness had reached intolerable proportions, especially in Canada's major cities. Therefore, by the time the federal government took official note of the problem, it was defined, in the public mind, as well as in the minds of government officials, as an urban crisis requiring an emergency response. Any notion of a flexible response to a variable problem was lost in the atmosphere of crisis.

As a result, NHI terms and conditions permitted the spending of federal funds on shelters, transitional housing, services to homeless people and even research, but did not allow the money to be spent on housing. The upshot was that despite $7.3 million being spent on affordable housing in Winnipeg from 2000 to 2003, all of it came from provincial and municipal sources, while by far the largest amount of money available, $23.5 million in federal funds, was excluded from use for the community's top priority (Leo and August, 2005). In this case, although the federal government took a stab at deep federalism by requiring a community planning process, it did not follow through with the necessary degree of flexibility in funding conditions.