becoming harder to maintain. During the period of Atlantic Fordism which emerged after 1945, it was the national scale that was primary in both economic management and political organization. This was reflected in the dominance of the 'Keynesian welfare national state' (or KWNS) as the principal institutional complex in and through which the market-mediated processes of capital accumulation and social reproduction were regularized in advanced capitalist societiesvi. This implied in turn that most cities operated primarily as sites of capital accumulation and social reproduction within a national context and oriented their actions within that context; and also implied that local states were primarily relays of policies settled at national level. Likewise, during this period of KWNS dominance, the main social movements were producer organizations, political parties, and pressure groups concerned with national economic growth and with social redistribution among the citizens of a given national state (cf. Offe 1985; Hirsch and Roth 1987). Various economic, political, socio-cultural pressures and forces (see below) have undermined this primacy of the national level. Indeed, it seems that there is currently no primary scale on which the global economy is being effectively instituted, organized and regularized -- whether global, triadic, national, regional, or local. The current period of 'after-Fordism' seems to have no privileged scale of organization on which the global economy is (or can be) managed. Instead there is a more complex nesting and weaving of different spatial scales as attempts are made to re-articulate them in the search for new spatial and scalar fixes. Moreover, as there is no single, self-enclosed and circumscribed spatial scale that can be taken-for-granted, 'the geographical boundaries of social relations have become direct objects of socio-political contestation' (Brenner 1997a: 24).
The current restructuring of economic and political relations (within which 'globalization' is only one of several processes) is associated with what can usefully be termed the 'relativization of scale' (Collinge 1996; cf. Brenner 1997b). This has obvious implications for the state, cities, and new social movements in advanced capitalist societies. This point can be developed by re-examining the nature of the Keynesian welfare national state. This had four key features in relation to economic and social reproduction in Atlantic Fordism. First, in its role of helping to provide the external and internal conditions for capital accumulation, the KWNS was Keynesian in so far as it aimed to secure full employment in what was treated as a relatively closed national economy and also aimed to do so primarily through demand-side management. The KWNS attempted to adjust effective demand to the supply-driven needs of Fordist mass production with its dependence on economies of scale and on full utilization of relatively inflexible means of production. Second, to secure the conditions for social reproduction, the KWNS was oriented to welfare in so far as it tried to regulate collective bargaining within limits consistent with full employment levels of growth, to generalize norms of mass consumption beyond those directly employed in Fordist sectors so that all national citizens might share the fruits of economic growth (and thereby contribute to effective domestic demand), and to promote forms of collective consumption favourable to the Fordist mode of growth. Its economic and social policies were closely linked to an expanding understanding and a progressive institutionalization of economic and social welfare rights attached to individual citizens of the national territorial state. Third, the KWNS was also national in so far as the national state had the primary responsibility for developing and guiding Keynesian welfare policies on different scales. In this context, local and regional states acted mainly as relays for policies framed at the national level; and the various postwar international regimes linked to Atlantic Fordism were aimed at stabilizing national economies and national states. And, fourth, the KWNS was statist in so far as state institutions (on different levels) were the chief complement to market forces in the operation of the 'mixed economy' and had a dominant role in shaping the institutions of civil society.
This said, the concrete forms of the KWNS and the specific modalities in which its functions were performed varied from case to case. One can distinguish among KWNS regimes in terms of their typical forms of economic and social intervention -- liberal social market regimes, tripartite social democratic regimes, dirigiste regimes with strong states and a relatively fragmented labour force, and more corporatist conservative regimes in which social welfare is partly organized on occupational or status lines and also tends to conserve rather than weaken inequalities. There is already an extensive literature on this topic. My chief concern here is the general nature of still emerging forms of economic and social policy and their associated institutions rather than with the past.
The KWNS experienced a crisis in the 1970s and 1980s. This had a variety of general economic, political, and socio-cultural causes. There were also more specific, conjunctural factors that affected the timing, forms, and incidence of the crisis in particular cases.
Economically, the KWNS was undermined by the increasing opening of national economies and their resulting interpenetration through a variable mixture of extraversion, inward investment, and expanding international social division of labour. These processes are often subsumed, of course, under the current