Please e-mail me if you do so, telling
me the title (as I have other work on the web).
This is the pre-copy-editing text of the article appearing as: "Social
Democracy Going Down or Down Under: Institutions, Internationalized Capital and
Indebted States" in Comparative Politics 30:3, April 1988, pp.
253-272
A PDF version is available.
A note on the notes: click on the note number to see the note. Click on "back" to return to the text.
SOCIAL DEMOCRACY GOING DOWN OR DOWN UNDER: INSTITUTIONS, INTERNATIONALIZED CAPITAL AND INDEBTED STATES
Herman
Schwartz
Government and Foreign Affairs
PO
Box 400787
University of Virginia
Charlottesville VA 22904-4787
434 924 3192 (x3359 fax)
e-mail: hms2f@virginia.edu
Author bio sketch: Herman Schwartz is Associate Professor of Government and Foreign Affairs at the University of Virginia, and author of In the Dominions of Debt and States vs. Markets. His current research compares contemporary public sector reorganization in Australia, Canada, Denmark, New Zealand, and Sweden. See http://www.people.virginia.edu/~hms2f for other published work on similar topics.
Abstract:
Internationalization of goods producing capital and rising public debt have undermined the institutional preconditions -- a unified labor movement, corporatism and a universal welfare state -- for the class compromises constituting European style social democracy. Australia's labor movement has faced these problems longer and so developed different, more appropriate institutions. Australian wage setting institutions use juridically set high minimum wages to help overcome tensions between unions while protecting the wage share of national income. Its largely privatized and residual welfare system is more resilient in the face of large public debts than universal systems. These institutions provide positive and negative lessons for European labor movements confronting internationalization and debt. Australia's experience suggests that Fritz Scharpf's call for 'socialism within one class' is a plausible response to these changes.
Introduction
Most analyses of social democratic strategy have taken for granted two crucial preconditions for the institutional structures that both reflected and sustained social democratic political power after World War II: internationally immobile capital in the goods producing sector of the economy, and a state with sufficient revenue flows to provide a high social wage. In the 1970s and 1980s both these preconditions for successful social compacts over wages and investment and for funding welfare and incomes policy eroded.
The 1970s and 1980s witnessed a rapid internationalization of both financial and productive capital. International bond placements tripled over the 1980s and reached $1.6 trillion in 1996. By 1988 foreign direct investment flows were running at three times their average level in the early 1980s and six times average flows in the 1970s. These flows disproportionately affected the small European social democracies as firms positioned themselves for the emerging Single European Market.(1)
Many European states responded to more trade, and more volatile trade, by expanding the public sector spending to boost domestic demand and shelter the economy. By the early 1980s these policies had created fiscal deficits that cumulated into substantial public debts, and, among the weaker economies, substantial foreign public debts. Despite efforts to meet the Maastricht criteria, deficits and debt levels remain quite high.(2))
If the preconditions for the institutional underpinnings of social democracy are gone, does this mean an end to social democratic policy everywhere? One of the most important insights of the new institutionalist literature is that institutions need to be considered in their historical and environmental contexts. Institutions well suited for one environment may be ill-suited for another. Most of today's social democracies developed institutions for maintaining successful social compacts over wages and investment and for funding welfare under conditions of capital immobility and clean fiscal balance sheets. Other institutions might be better suited for an environment of mobile capital and indebted states. As it happens, Australian labor has had over a century's experience with the kind of environment European social democrats now confront, and constructed different institutions in response to this different environment.
Francis Castles has already elaborated in some detail what he characterizes as the Australian model of domestic defense cum social democracy: judicially set high minimum wages, protected markets to assure low unemployment, and a residual welfare state for those few outside the labor market. He argues that Australia and some similar welfare states constitute a distinct fourth type of welfare state, in contrast to authors like G sta Esping-Andersen who usually sort it with the 'liberal' cases.(3) I agree with Castles that Australia has social democracy 'by other means.' But I think Castles' argument suffers in two respects. First, despite his pains in Australian Public Policy and elsewhere to show that Australia delivers welfare outcomes similar to Sweden's, Castles still argues that Australia should try to generate institutions more like Sweden's. For Castles and others, Sweden remains the normative and analytic yardstick against which all welfare states are judged. But the Swedish model has foundered as the two hidden preconditions mentioned above have eroded.
Second, Castles' argument about Australia is strongest when he argues about what Australia's welfare state is, but weakest about why Australia has this kind of welfare state. As Rob Watts has pointed out, Castles' argument in both The Working Class and Welfare and Australian Public Policy is strongly functionalist.(4) For Castles, because Australia has a strong labor movement, it should also have some kind of welfare state; because small European states facing volatile international economic environments evolved the strategies for domestic defense Peter Katzenstein has described, so too must have Australia.(5) Castles' arguments about origins provide necessary but not sufficient reasons for understanding the Australian model's nature. He locates the origins of the wage earners' welfare state in labor's early political mobilization, the absence of a feudal past, a strong state, and the apparent absence of concurrent ideas of universal welfare. All of these are correct, but equally so they are only necessary and not sufficient conditions for the emergence of the welfare state Castles describes. Sufficient reasons can only be found by examining actors' motives. These motives were shaped by the extremely steep discount rates on the benefits of future wages and social benefits that capital mobility in the goods producing sectors and large public debts created. Australian "workerism," in other words, had a rational basis and involved a long term perspective.
By going back to first principles, particularly the questions of how capital mobility and debt affect labor's policy preferences, I hope to provide a stronger argument for why the kind of welfare state Castles correctly describes emerged, and by doing so illuminate the problems European social democrats now face. In particular I wish to address two issues highlighted by recent revisionist literature on the 'Swedish model,' namely the degree to which the construction of centralized collective bargaining and universal welfare both reflected cross-class coalitions rather than unilateral labor strength. Both revisions suggest that much of the literature on Sweden is hagiographic, and thus highlight the importance of examining the effects of capital mobility and debt on labor strategies. Both revisions support Castles' claims about a distinct Australian model and suggest that Australia, paradoxically, may represent a case where a strong labor movement got much of what it wanted, given the constraints it faced.
The first section looks at the classic arguments about European social democracy, focusing on two institutional efforts to maximize and maintain working class strength: corporatist collective bargaining and universal welfare states. It shows how these arguments presumed capital immobility and fiscally sound states. Then it discusses recent revisionist accounts of Scandinavian social democracy to show how different levels of capital mobility and debt shape labor strategies. Both of these revisions suggest that Castles argues correctly for an Australian model of social democracy, suggest that Australian labor was neither as weak as nor as 'laborist' as has been thought, and illuminate Australian labor's motives.
The second section describes and analyses Australian institutions as reasonable adaptations to high capital mobility and high public debt. This analysis shows that European style corporatism is merely one modality for institutionalizing class compromise via a system of collective bargaining. Australian institutions function like European corporatist institutions: they aggregate labor and capital, set limits on industrial conflict, and maintain economically and morally feasible wage relativities among different groups of workers. But reflecting Australian labor's distrust of capital's commitment to future investment -- and thus its extremely high discount on the value of future wages and social benefits, these institutions focus on delivering high current direct wages, not on social and future wages as in Europe. Similarly the privatization of welfare reflects labor's response to high levels of foreign held public debt and repeated demonstrations that property rights in debt overrode property rights in welfare. Labor's distrust of the state's ability to fund future claims reinforced the orientation to the present created by capital mobility. Both discouraged the creation of an extensive, and thus expensive welfare system, in favor of a means tested, tax funded, residual welfare system.
The final section advances some conclusions about the utility of Australian model. First it briefly examines the consequences of rising capital mobility and public debt on Sweden and Denmark. In both inter-union tensions over wage relativities, increased migration by goods producing capital, and rising public debt have led to a disintegration of centralized bargaining and a weakening of universal welfare provision. In contrast, from 1983 to 1996 the Australian labor movement was relatively successful in pursuing Fritz Scharpf's 'socialism in one class,' securing inter-union solidarity, stable incomes and employment, while also slightly expanding the social wage.(6) Although national 'models' are not easily transferable, this suggests that Australian institutions have some adaptive advantages in dealing with greater capital mobility and fiscal stress.
Social Democracy from a European Perspective
For the sake of broadness and simplicity I will assimilate the conventional wisdom on European social democratic strategy with Adam Przeworski's delineation of three crucial, interlocked, 'either-or' decisions about class compromise.(7) Social democratic success is a function of how well these decisions are transformed into two institutional pillars. First, strategically, labor movements accepted democratic political institutions and property rights rather than seeking revolutionary transformation of the system. This choice reflected pessimism about workers' ability to organize to a level sufficient to secure revolutionary change, but their ability to organize sufficiently to negotiate with capital and secure a modicum of economic security for the worst off.(8) This decision then entailed two further decisions, one economic, the other political. The economic wing of the movement confronted a choice between pushing for worker control of the production process (as distinct from property rights over productive capital) or accepting existing relations of production and bargaining only over wages. This was a choice about the level of militancy.(9) The political wing of the movement confronted a choice between remaining a minority working class party or expanding its potential electorate outside the organized working class and compromising with non-working class parties in order to win elections and govern successfully. This ultimately was a choice about redistribution.
Most social democratic movements opted to accept managerial control inside existing property relations and a pursue a 'big tent' political strategy. Given of electoral democracy and continued capitalist social relations, social democratic movements then tried to maximize working class power by institutionalizing centralized collective bargaining and politically acceptable welfare programs. Both mitigated market pressures that weakened working class power by providing institutions to resolve collective action problems. Centralization eased fears that lack of inter-union solidarity might weaken labor's negotiating power at the same time that it dissuaded capital from bargaining with labor for fear that a compromise might not hold. A welfare state eased pressures on individuals to (re-)enter or remain in the labor market who might underbid other workers for access to wage employment, weakening labor's negotiating position. Each in turn.
Collective Bargaining and Corporatism
The classic vision of social democratic collective bargaining strategy involved a compromise over wages with capitalists that is best described by Przeworski and Wallerstein, and which is largely based on a particular image of Scandinavia.(10) Present and future employment and thus future wages could only be assured if capitalists continued to invest. So union movements regulated their wage demands -- their militancy -- in order to assure capitalists of what capitalists felt was an adequate level of profits, and thus induce capitalists to continue investing and creating future jobs. In more recent times, unions also used wage restraint to control the trade off between employment and inflation.
This collective bargaining strategy had two preconditions. The first, centralized bargaining units, is well known, but the second, immobility for goods producing capital is not. Each in turn. Bargaining worked best when both actors operated through centralized actors, particularly when they had state enforced representational monopolies. Central actors could assure the other party of their compliance with bargains in a way that individuals (or many individual organizations) acting self-interestedly in response to market signals could not. So successful social democratic bargaining worked best with some kind of centralized, organized representation.(11) The literature on social democracy sees the emergence of union federations as a precursor for corporatism, particularly as these federations seem to have spurred parallel organizing by employers. This analysis of corporatism's first precondition has two problems.
The first problem concerns the formation of central bargaining agents for workers. The classic vision, including Castles', presumed that labor had few internal conflicts of interest, and saw the major organizational problem as a prisoner's dilemma, e.g. preventing individual defection during strikes.(12) But Peter Swenson's revisionist treatments of the emergence and maintenance of centralized bargaining in Scandinavia suggests a major cleavage in the labor movement around the issue of regulating wage relativities across unions, relativities which in turn are sensitive to unions' differing exposure and vulnerability to world market competition.(13) Swenson argues that intra-labor solidarity may be more an outcome of corporatist bargaining than a precursor, because by externalizing control over wage relativities to corporatist institutions unions can control intra-class conflicts of interest at the same time that they protect capital's and labor's future interests by containing inter-class conflicts. Corporatism helped labor maintain unity, not the reverse.
Still, Swenson's analysis raises a troubling question. While Swenson overturns the notion that labor's strength in Sweden alone formed the basis for social democratic advances, there is no denying that many unions were strong and militant. Given high levels of labor militancy, why didn't Swedish (and pari passu Danish) productive capital simply exit? Why did they expend so much energy building encompassing organizations, which after all, much of collective action theory suggests is an unlikely outcome? The answer lies in the second, overlooked precondition for organized collective bargaining: immobility in goods producing capital.
Capital Immobility
The origins and heyday of corporatist bargaining in Europe run from the 1900s to the 1960s, when internationalization of goods producing capital was low. The distinction between goods producing capital and financial capital is important, because the current debate tends to confuse the two, despite quite distinct effects and differing degrees of mobility.(14) Financial capital has always been mobile, except during the Great Depression. The contemporary increase in international financial flows means that they now approximate those of the pre-World War I period when their volume and speed are scaled against output and turnover times in the real economy. In 1914 Britain's overseas assets were nearly twice its GDP, outflows amounted to 10 percent of GDP, and telegraphy linked active financial markets globally. Increasing financial capital mobility allows states to finance larger fiscal deficits than before, while subjecting impecunious states to market pressures via lower bond ratings and higher interest rates, just as in the late 1800s.
The real change in capital mobility these days has occurred with industrial capital. Most pre-World War II European industrial firms were small family held firms and thus rooted firmly in their soil of origin. In Sweden, for example, 95 percent of firms in 1939 had fewer than 250 employees, and only Alfa-Laval could be considered a real multinational firm.(15) Internationally mobile goods producing capital largely flowed into resource development. For example, the Australian Agricultural Company, a London chartered and capitalized firm, operated mines and wool farms in Australia and Argentina, and actively allocated capital across these and other potential production sites.
Industrial capital's relative immobility during the formative period of European social democracy gave it an incentive to compromise in order to secure their future stream of profits. Much as workers' future employment and wages rested on investment by capital, immobile capital needed some cooperation from workers in order to generate future income. Some of this derived from local effective demand created by local wages, but workers' propensity to strike was more important. Currently employed workers without any assurance of long term employment were likely to pursue a highly militant strategy and strike in pursuit of wages today rather than worrying about employment tomorrow - much as Australian (and Swedish) workers did. Unlike corporately owned firms, family firms' own consumption could not be divorced from the stream of income their own firm generated. Without a viable long run exit strategy, immobile productive capital had to opt for voice within corporatist institutions; the absence of exit created the same incentives for compromise as workers' reliance on capitalist investment for future wages.(16)
The Welfare State and the Fisc
Social democratic efforts to build a welfare state economically and politically complemented efforts at centralized bargaining. Social democratic strategy viewed the welfare state as a means to clear marginal workers to remove themselves from the labor market by approximating a lifetime wage.(17) Old age pensions, health insurance, sickness pay, and free education removed market compulsions to sell labor power, strengthening working class solidarity and bargaining power in primary labor markets. Just as most analyses equate corporatism with successful social democratic efforts in the labor market, they take the universal, tax financed, and institutional welfare states of Scandinavia as the epitome of efforts to decommodify the reproduction of labor, and, as Esping-Andersen puts it, to make people somewhat indifferent between welfare and work.
This kind of welfare strategy must balance two interlocked but competing political considerations, however.(18) Social democratic parties needed non-blue collar votes to win elections. European social democrats have generally tried to put white collar groups under the welfare state's umbrella in order to build a glacis of middle-class interests around core working class interests in welfare.
But a universal welfare state relies on high levels of taxation to generate large inter- and intra-class and inter-generational transfers. (Similarly an incomes policy based on tax concessions presumes fiscal maneuvering room.) These taxes can create political opposition among both white and blue collar groups if they are disproportionate to the perceived benefits of the welfare state or alternate ways of reducing the risk of illness, accident, etc. In that case social democratic efforts to "manufacture its own class base" through welfare reformism may backfire and generate an opposing coalition.(19) So a fiscally sound, debt free state is a hidden prerequisite for a decommodification strategy just as much as immobile capital is for corporatist bargaining.
If substantial state revenue is already hypothecated to foreign or domestic creditors, then tax pressures will already be high and without any visible benefit for citizens. Furthermore when recessions shrink state revenues, debt service usually takes precedence over welfare spending. Welfare transfers are a legally secured claim to a stream of income, that is, essentially a property right. When implicit property rights in welfare commitments conflict with the explicit property rights constituted by public debt, the former almost always loses. Fiscal ministries generally administer the public debt and their primary concern is creditworthiness, particularly vis a vis foreign creditors. As social democratic governments by their very nature are committed to reassuring frightened capitalists, a commitment to service debt follows from a commitment to explicit property rights in general. So indebted states always find it difficult to live up to their welfare commitments at precisely the moment when unemployment and its associated pathologies create pressures to boost welfare spending. The larger the public debt, the more likely it is that public welfare will be scaled back; consider recent increases in the eligibility age and deindexation of old age pensions in the US, Canada and Sweden.
Just as Swenson's revision helps highlight the importance of capital immobility and weakens the argument that labor alone constructed corporatism, Peter Baldwin's revision undercuts the notion that a strong labor movement acting strategically created the universal welfare state. This revision highlights the relation between debt and labor's choice of financing systems for welfare. Baldwin argues that Scandinavian universality, with its broad coverage, differentiated payouts based on contributions, increased overall tax burden and reduced vertical income redistribution, flowed out of a political compromise between labor's desire not to fund a contributory pension, and non-labor's desire not to pay for a pension they could not receive.(20) Baldwin undercuts the idea that a hegemonic labor movement necessarily would have constructed a universal welfare state. In turn this implies that Australian labor's choice of a residual welfare state did not reflect the failure of vision or bad timing that Castles proffers as explanations. Instead, Australian labor seems to have gotten some of it wanted -- social protection for ex-workers, paid for by direct taxes on the wealthy -- while, as later sections will show, being constrained by debt in regard to the scope of that protection.
What then are Australian wage setting and welfare institutions and how do they reflect high capital mobility and debt?
The Australian Model: What and Why
As Castles has argued, while Australia has neither European style corporatism nor a universal welfare state, it does have centralized bargaining institutions and a well developed system for delivering welfare to workers. Below I will argue that these institutions help maintain intra-labor solidarity by externalizing parts of Swenson's two trilemmas, and that both represent adaptations to the external economic environment Australian labor confronted. But as these Australian institutions are largely terra incognita for most readers, I will briefly sketch each before I analyze them. How collective bargaining is institutionalized in Australia
From roughly the turn of the century, courts of conciliation and arbitration located at both the federal and state level have set wages through judicial proceedings. Only labor and employer organizations may initiate proceedings; individuals have no standing. The courts hand down 'awards' setting wages for specific occupations along with relativities among occupations, regulating conditions of work, and setting the duration of the award. From 1907, in differing guises, the federal court has enforced a high minimum wage, and has also maintained a rough 10 to 7 parity between the wages of skilled and unskilled workers.
Awards generally bind workers and employers who have not participated in either of the organizations presenting the suit, but who work in the industry or trade governed by the award. Nominally strikes and lockouts are prohibited while an award is in force, and arbitration in the event of a strike/lockout often has been compulsory. Unions and employer organizations violating the strike/lock-out provisions risk deregistration. De facto, the award system compels capital and labor to organize in order to participate in wage setting; de jure, registration before the court creates a monopoly of representation. Consequently, "Australian unions are a part of the state, but in the sense of local governments with entrenched traditions of autonomy," and constituted on an occupational rather than territorial basis.(21)
From the 1930s to 1990s roughly 85 to 90 percent of the labor force has been covered by either state or federal wage awards, with the federal court's rulings tending to set benchmarks for the state tribunals. Until recently, free wage bargaining as in the US occurred only in a narrow range of occupations and only literally at the margins, when workers with market power attempt to acquire (often illegal) wages above and beyond those the court awards. Unions and firms can bargain outside the system and have their awards 'registered' in the courts, and during the early 1990s the Labor Party reformed the award system to formalize and liberalize the possibilities for enterprise bargaining.
As with European collective bargaining institutions, Australian institutions emerged in order to structure class conflict after a series of brutal turn of the century battles. But European institutions are merely one modality for embodying a particular class and/or cross-class compromise. Australia's courts present a different modality but with the same essential consequences. Australia's conciliation courts created and legitimated organizations for interest representation, set the parameters for class conflict, decommodified market wages to a strong degree, and, in concert with trade protection, facilitated vertical redistribution of wealth.
The courts' recognition and regulation of unions did the same thing the Danish September 1899 and Swedish December 1906 accords did. The courts reconstituted and reconcentrated the unions shattered in the 1890s battles, and moreover gave unions monopolies of representation comparable to those enjoyed in later European corporatism. From around 5 percent of workers, in the mid-1890s, union membership jumped to 28 percent by 1911, and to 60 percent by 1951. By 1921 more than half of all unionized workers were in unions of more than 10,000 members and over 80 percent of unions operated in more than one state. After the Court began to award a national minimum wage in 1921, the share of workers covered by federal awards jumped from 14 percent in 1921 to 59 percent in 1926/29, and by 1927 unions organized the Australian Council of Trade Unions (ACTU) to argue these cases.(22)
The federal court's ability to extend its rulings on an industry-wide basis also compelled capital to organize industry-wide and national organizations. The first nation-wide federation, the Central Council of Employers of Australia formed in 1903 to influence the federal bill constituting an arbitration court. But the Central Council persistently split along the fissure between traded and non-traded firms. Protected, domestic market oriented manufacturers, organized into Chambers of Manufacturing, vied with export oriented firms and farms, organized into various employers' federations, for control over any national employers' federation.(23)
Three principles animating the courts' regulation of wages tended to decommodify primary incomes strongly: the living wage, comparative wage justice, and indexation. Beginning with a federal court decision in one industry in 1907, the courts directly set minimum wages and consistently argued that this minimum wage was a social right based on human needs, unaffected by considerations of profitability.(24) The world's first elected, majority national Labor government in 1910 confirmed and expanded the Court's minimum wage setting power to the national level. The new national minimum, or "Basic Wage," diffused into the rulings of most state level courts. By 1920 75 percent of workers were covered by the basic wage, either from federal or state level awards.
The Court's 1907 decision also fixed relativities for skilled and unskilled work in a 10 to 7 ratio, establishing the norm of "comparative wage justice." Over time this ratio tended to compress skilled-unskilled differentials. Comparative wage justice required that the same wage be paid for a given task regardless of where it was performed, and whether or not an employer was capable of paying the wage.(25) Although major tensions continued between unions over wage relativities and between unions representing workers in domestic and export market oriented sectors, the externalization of control over wage relativities into arbitration tended to promote labor solidarity by removing one of the corners of Swenson's external trilemma. Comparative wage justice froze relativities in place, allowing labor to concentrate on full employment and maximizing its share of national income. Although there have been periods of conflict over relativities, most notably in the 1970s, recourse to the arbitration system has allowed unions to avoid the kinds of wage-chasing battles which split the union movement in Sweden, for example, in both the 1970s and 1980s.
Finally, after 1922 automatic quarterly indexation adjusted the basic wage, but not margins for skill, in response to inflation. Indexation has been the most contested of these principles, but over time it has facilitated both an incomes policy and macro-economic regulation.(26) Arbitration's principles of wage compression, comparable wage, and the irrelevance of profit levels strongly parallel elements of the Rehn-Meidner model.
So Australia's peculiar wage setting system provided the essentials of European corporatism in a different form. In Australia the state had an over-large role: by enforcing decisions judicially it inhibited the emergence of intra-organizational methods for maintaining discipline, like centralized control over strike/lock-out funds. In northern Europe the organizations participating in corporatist bargaining grew out of the soil of civil society; in Australia the state planted and watered a similar set of organizations. The Australian state thus helped overcome some of the imbalance against labor inhering to Offe and Wiesenthal's two logics of collective action.(27) This system also determined the distribution of income, by making a living wage a social right, and by imposing the equivalent of Sweden's solidarity wage on labor via the basic wage and its associated relativities.
Before turning to labor's motives for institutionalizing arbitration, a quick look at overt welfare is useful, because the to a large degree the choice of arbitration also determined the choice of welfare instruments.
How Welfare is Institutionalized in Australia
Australia has some of the lowest levels of social security spending as a percentage of GDP among the industrial OECD countries, fairly low replacement rates, and widespread use of waiting days before flat rate benefits can be accessed. All this leads Esping-Andersen to label Australia as the least decommodifying welfare state of the 20 countries he pigeon-holes in Three Worlds of Welfare Capitalism.(28) Castles has correctly taken Esping-Andersen to task for this interpretation while remaining critical of Australia's welfare system.(29) For Castles the provision of work at high minimum wages obviates the need for all but a residual welfare system to handle those outside the arena of work. The Conciliation system compels extensive provision of many benefits. Sickness benefits, for example, are provided by employers, not the state.(30) Thus Australians have a social right to work at high wages and with good benefits, rather than Scandinavia's social right to an extensive set of public benefits and services. The Australian state then uses tax revenues to provide a means tested safety net for those who do not work. The major exceptions to this residual system all have sufficiently broad social bases to generate the continued political support Esping-Andersen sees as desirable. Family and old age benefits all reach over half the population, while health insurance in principle is universal.
Castles and his collaborators are certainly correct that Australia's welfare state delivers better outcomes than Esping-Andersen might predict on the basis of his categorization of Australia. But for Castles the fact that a strong labor movement -- one with unionization rates well above European levels before 1914, and with a political wing capable of winning not just elections but also absolute majorities -- accepted such a "weak" welfare state remains puzzling. Castles correctly notes that Australian workers saw their primary source of security in policies to generate full employment at high minimum wages and with privately provided welfare, and that until World War II constitutional problems limited the federal government's ability to levy the kind of income taxes needed to finance anything more than a residual system. However Castles' argument needs to be supplemented with an appreciation of the potential for capital flight and the problems associated with having a highly indebted state, because as Baldwin observes, it is not at all clear that a labor movement strong enough to impose its will would necessarily have opted for a universal welfare system. Rather, Australia seems to present us with a case of a labor movement getting what it asked for under the constraint of high public debt.
Explaining Australian Collective Bargaining and Welfare Institutions
Why did Australian workers (and domestic manufacturers) fairly consistently support arbitration, whose outcomes have been high minimum wages, wage compression, and the privatization of welfare provision as their particular way of institutionalizing class conflict and maximizing labor's power? Labor prefers all this because of the demonstrated international mobility of some goods producing capital in Australia and the Australian state's weak fiscal situation, which the international mobility of financial capital exacerbates.
Australian labor opted for wages today because mobile capital could and did opt not invest in Australia, thus threatening tomorrow's wages. Wage compression similarly reflected union efforts to disperse prosperity in the export sector -- the major site of mobile capital before 1945 -- into the non-traded sector and thus prop up employment. Labor similarly cannot trust the state to maintain an elaborate welfare system in the face of creditor demands.
Australia relied and relies on imported capital for development. 1870 to 1914 foreign investment supplied about 40 percent of gross domestic capital formation (GCDF) and 19.1 percent 1920 to 1939. But these averages conceal important subperiod trends bearing on the development of attitudes to foreign investors and the likelihood of continued investment. Foreign investment occurred in surges whose abatement triggered economic collapses. Net foreign investment fell 72 percent 1891 to 1894 from its 1888-1891 average level, driving GDCF 1891 to 1897 down to half its average levels in the five preceding years. Similarly 1930 to 1937 gross domestic capital formation fell to 40 percent of its average level 1922 to 1929 as foreign investment fell by 80 percent.(31) During the 1970s direct foreign investment fell about 25 percent from its 1950s/1960s levels, but then recovered slightly in the 1980s to account for over 10 percent of GDCF annually.(32)
The erratic pattern of foreign investment demonstrated the risks of relying on foreign capital for growth and future wages. contributed to the 'workerist' orientation of the Australian labor movement. Foreign investment typically went into labor intensive activities. High foreign investment thus created tight labor markets, but the obverse was also true. Falling foreign investment and its negative multiplier effects tended to create high unemployment as in the 1890s and 1930s. While reliable unemployment statistics are not available, the 1890s witnessed considerable net outmigration, a sure sign of high unemployment in Australia, and anecdotal evidence suggests unemployment rates in the 20 to 30 percent range. Australia had the second highest unemployment level in the industrial world 1930 to 1934, averaging 23 percent versus 19 percent, for example, in the US.() Demonstrated capital mobility made workers discount the probability of long-term investment. Wage restraint now in hope of getting investment and wages later looked nonsensical, particularly as foreign investors habitually remitted profits overseas and reinvested with a global perspective. Strategically, trying to maximize current wages in order to ride out fluctuations in foreign investment and thus employment appeared better. Conciliation forced capital to pay high current wages, including provision of welfare services publicly provided elsewhere.
Here the connection with trade protection is both politically and economically salient. The critical economic and political problem in the Australian system is assuring high levels of employment. Low unemployment means less intra-labor competition which might drive down wages and also lower burdens on the fisc. High levels of tariff protection -- rates of effective protection reached 50 percent by the 1930s -- transformed manufacturing into a non-traded good, allowing manufacturing employment to be much higher than it might otherwise have been, and creating about 10 percent of jobs at that time.(34)
Politically, conciliation and arbitration were linked in a 1907 deal between domestic oriented manufacturers and labor that constituted Australia's 'Saltsj”baden.' By linking conciliation with tariff protection for domestic market oriented manufacturing, labor took wages out of competition for those domestic firms most likely to be hurt by high wages. In turn this boosted overall employment by anchoring part of the workforce in a sector with stable investment and employment, and dominated by small and immobile capitals. It also broadened the political base for conciliation by pushing some traded firms into the non-traded sector.
The link between conciliation and protection also helped bridge different levels of militancy and trade preferences between craft unions, primarily in domestic manufacturing, and industrial unions in the export sector. Industrial unions, who were most exposed to the risks of foreign capital flight, had a natural tendency towards (extreme) militancy, and favored free trade. Craft unions leaned in the opposite direction. The conciliation system contained industrial militancy while still delivering fairly high current wages, particularly for the unskilled workers that populated the export oriented industrial unions.(35)
Australia thus had the inverse of the coalition Swenson describes for Sweden. Non-traded capital and labor united with traded labor to control the profit share and wage demands in the foreign-owned traded sector, while redistributing income into the non-traded sector. The non-traded sector deliberately raised wage and input costs for the traded sector in order to squeeze economic rents out of traded sector production.
The amalgam of foreign capital and medium and large farmers that constituted export oriented capital lost from this alliance. Goods producing foreign capital -- usually in extractive activity -- usually accepted this income loss, since Australia's position as the low cost producer of most minerals and wool meant that they could still make profits, even though the conciliation system shifted rents to domestic market producers and workers. Multinational capital in manufacturing responded with a very cautious attitude towards expansion and took advantage of the broad domestic market. Farmers meanwhile formed their own party, vehemently anti-Labor, in 1919.(36)
Australia's residual 'welfare state' also reflects labor's experiences with mobile capital and an indebted state. Roughly half of Australia's foreign debt was public debt. During the 1890s and 1930s these governments privileged debt service over welfare and other state spending. Debt service absorbed roughly one fifth of all government spending by the end of the 1920s, just about equaling 'welfare' spending. Reliance on an indebted state to provide a social wage looked just as foolish as wage restraint looked in the face of mobile foreign capital. The state simply did not have the revenue flow to assure current, let alone future payments.(37) Non-Labor governments looked for economies in welfare to fund debt service, and their welfare proposals in the 1930s were almost always for contributory programs; Labor party governments attempting to repudiate debt were dismissed by the Crown. As Watts shows, when the Australian Labor Party (ALP) regained power in the 1940s, fiscal considerations strongly motivated their continued orientation towards generating full employment and expanding the residual welfare state rather than creating a universal welfare state.(38) (Debt concerns also constrained New Zealand's Labour government when it began building a welfare state in 1935.)(39) As with wage bargaining, Australian labor's preoccupation with the present and steep discounting of future benefits seems reasonable.
So too does its reliance on publicly mandated (via arbitration) but privately supplied welfare. "Privatization" of the welfare state secured workers' savings (which in Scandinavia took the form of state-controlled deferred wages that could be diverted to debt service) from foreign creditors. For example, from 1920 the basic wage award made provision of six days of sickness pay obligatory for employers.(40) The 1985 Accord renewal provided for an employer contribution to individual pensions in lieu of pay increases that eventually will reach 12 percent of wages. Given a steep discount on the future value of state supplied benefits, and an orientation towards current wages in the face of mobile capital, it makes sense that Australian labor opted to redistribute profits immediately through mandated benefits and through direct bargaining with employers. This made them less vulnerable to changes in government. Privately controlled pensions are less subject to the political erosion of public pensions now seen in Scandinavia. In Sweden, for example, the state has changed the old age pension from a defined benefit to a defined contribution program, while also legislating that cost of living adjustments will be foregone when the budget is in deficit.(41)
The centrality of state institutions in the construction of interest organizations and for setting the pattern of welfare expenditure provides a key to understanding the limits to class compromise in Australia. The non-traded sector in Australia sought to transfer wealth from the traded sector, not to constrain wage demands as in Scandinavia. Or to put it somewhat differently, in Scandinavia unions in the traded sector wanted to hold down wages in the non-traded sector, while in Australia non-traded unions sought to pull their wage levels up to those in the traded sector, and this could only be done via the state. The state's centrality is an indicator of how politically strong labor actually was, in contrast to the relative weakness of labor revealed by revisionist accounts of Scandinavia. Australia's institutions survived even though Labor rarely ran the government; Sweden's required social democratic government to persist.
Conclusions
Can European social democracies learn from Australia's experience? I consider two connected issues here: first, the degree to which the salience of immobile resources in Australia's economy renders comparison pointless; second, the degree to which Australian institutions permitted easier adaptation to a world with more mobile capital and more public debt. The first issue is not as problematic as it might seems; because (relatively) immobile labor constitutes a resource, too. Second, if Australian labor has generated specific institutional responses to mobile good producing capital and indebted states, has this been a wholly successful response? This has several subsidiary aspects: do these institutions help maintain the existing compromise with capital?; do they help foster intra-labor unity?; do they permit adaptation to a changing world market? Below I argue that Australian wage and welfare institutions were better adapted to the 'new' environment and thus could accomplish these tasks without sparking the wage chasing battles, the collapse of central bargaining, and the tax pressures that the traditional social democracies experienced. Before addressing these issues a brief discussion of Sweden and Denmark is in order.
Sweden and Denmark in the Australian Mirror
In the 1980s the Swedish and Danish labor movements ran into trouble trying to cope with the kind of problems with which Australian labor has long familiarity. In both countries fights over wage relativities in the 1970s and 1980s have weakened the left and thus undermined either one of the essential preconditions for corporatist bargaining or one of its major goals.(42) As 1970s incomes policies failed in both countries, their current account and fiscal deficits rose to record highs. The failure of incomes policy, i.e. of corporatist bargaining to control wage demands in the non-traded sector, motivated export sector unions and firms to seek new ways of controlling the non-traded sector. These primarily took the form of efforts to reorganize the public sector so as to disconnect public and private sector wages and impose wage discipline on the public sector.(43)
In Sweden the basic bargain between capital and labor began to crumble in the 1980s as Swedish firms migrated to Europe and America and as efforts to maintain solidarity and prevent wage-chasing among unions prevented strong unions from claiming the wages they wanted and firms from getting the kind of contract they needed. By 1990 the ratio of overseas production to exports shifted decisively in favor of overseas production, reflecting a tenfold increase in outward FDI flows by Swedish firms and a tripling of overseas production. Mobile Swedish employers forced decentralization of the industrial relations system.(44) The decline of institutions supporting centralized bargaining and solidarity wages once more confronted unions with the problem of trying to reconcile a conflicting set of goals: wage leveling inside and outside their union and full employment. Simultaneously, very large deficits in the early 1980s and early 1990s public debt pressures had cumulated into a foreign public debt position equal to 30 percent of GDP. In response, both conservative and social democratic governments cut welfare to try to restore budget balance.
In Denmark foreign debt reached 33 percent of GDP by 1982, and 40 percent by 1989, forcing cuts in what had become the OECD's largest public sector in proportion to population. The central state deindexed the whole range of public benefits. Collective bargaining became more decentralized during the 1980s, reflecting pressures to expand wage differentials compressed by incomes policies in the 1970s and increase flexibility.(45) Danish manufacturing, dominated by small family owned firms, remains relatively immobile. The institutional structure surrounding collective bargaining increasingly resembles that which emerged in Australia in the 1980s in terms of the location of bargaining, which decentralized into 'cartels' of unions, rather than national unions, as in Sweden.
Lessons from Australia?
Does Australia's enormous comparative advantage in resource production mean there are no transferable lessons? Clearly resource-based production allowed labor to bargain for a share of resource rents, and Australian institutions reflect this fact. Australian labor's high current wage strategy was partly successful due to the immobility of Australia's natural resources, which did attract an erratic stream of foreign capital. However the Scandinavian economies were themselves resource based. Social democratic strategies in general find may more fertile ground in extraction economies because of concentrated populations of homogenous workers and because rents are so visible. Castles has argued that Finland -- still a highly resource based economy -- is institutionally closest to the Australasian model.(46)
But Australia's resources are not unique, and mobile goods producing capital can and did go elsewhere. Capital also chases another relatively immobile factor: skilled labor. European workers are particularly immobile compared with Australians or Americans. Plausibly individual skills and the skills/expertise embedded in firms as organizations constitute the same kind of draw for capital that extensive grazing land and ores did in Australia. Current trends towards differentiated batch production, the debureaucratization of large corporations, and rising real wages for skilled labor all suggest that skilled labor is both in high demand and an increasingly important factor in competitive struggles among capitals.(47) Aggressive efforts by unions and social democratic parties to create the widest possible skill base should create the kind of resource base that otherwise footloose local firms might find irresistible. This suggests that strategies combining skills upgrading, job sharing, and institutions that redistribute wage gains to those parts of the non-traded sector that make use of skills up-grading to increase productivity may be best for social democratic movements. This strategy would also would help control inflation arising from skills shortages in the traded sector.
If Australia's resource base is replicable in this sense, then Australian institutions are worth examining because they enabled Australian labor to remain unified, to adapt better to the current global economic environment than did Scandinavian institutions, and to maintain the existing class compromise with capital. During the 1980s the ALP expanded employment and the social wage while using arbitration both to ameliorate tensions between the traded and non-traded wings of the labor movement and to upgrade skills across the economy. Australian institutions did not determine the nature of the strategy the ALP and ACTU used in the 1980s, but it they facilitated the adoption and execution of that strategy. In particular, Australian unions' reasonably successful experience in the 1970s with an arbitration imposed incomes policy that externalized wage relativity problems made them receptive to arbitration led restructuring during the 1980s.(48)
Restructuring took place through eight "Accords" between the ACTU and ALP. The Accords permitted a decline in real wages and the growth of the non-farm profit share from 14 percent to 16 percent of GDP. As in Scandinavia real wage decline largely fell on public sector workers, but it was acceptable to both wings of the labor movement because it also generated above average growth and a large decline in open unemployment. Despite labor force entry at three times the OECD average, Australia created jobs at roughly twice the OECD average rate until the 1990 recession. In contrast Sweden and Denmark experienced below average GDP growth and job creation over the decade, despite falling real wages.(49)
The ALP and ACTU also used the arbitration system to restructure work processes in ways that most Australian management was unable to do.(50) Unions whose non-traded status was a function of trade protection recognized that labor could no longer live by extracting rents from the export sector; their non-tradability was tenuous. These unions wanted restructuring to create alternate sources of growth and ease the inevitable transition to tradability by upgrading the skills base in all of Australian manufacturing, and by rewarding increased skills with increased margins. Arbitration allowed Australian labor to try to re-engineer production and upgrade skills on a national basis without generating the kinds of extensive relativities battles Sweden experienced. Juridically set minimum wages in the context of a broad skills base would tend to maximize full employment and lifetime wages. Given that labor can rely on neither corporatist structures nor keynesian policies to resolve its trilemmas, homogenization of the work force's skills and wages has helped reduce the tensions that accompany efforts to impose labor movement wide wage restraint in order to attract capital.
This restraint is politically plausible because arbitration still captures and redistributes economy-wide productivity gains to all workers, making changes in margins and relativities tolerable to the ACTU's component unions.(51) Under the ALP, centrally imposed skills upgrading was the quid quo pro for the partial decentralization of wage bargaining (so-called enterprise bargaining) desired by both firms and workers in the metals sector.(52) These metals firms dominate the domesticly oriented, non-traded sector, so their loyalty to the ALP program is also essential.
The Accord also promised an increased social wage in return for wage restraint; essentially an incomes policy within one class. Consonant with past institutional arrangements and its recurrent debt problems, Labor increased the social wage by simultaneously increasing means testing and raising benefits for those at the bottom, particularly by increasing family benefits to the point where it ranks third among the OECD countries for such benefits for low income strata. It also expanded public health care. Overall federal social spending increased roughly twice as fast as the general budget. Because so much Australian poverty now involves single (female) parents, this targeting is both redistributive and appropriate. Workers (male and female) who get the basic wage continue to receive 'welfare' via artificially high wages that implicitly redistribute income away from better paid workers and salary earners.
All of this is consonant with the Fritz Scharpf's argument that only 'socialism in one class' can restore full employment. Scharpf counsels that a unified labor movement do this by imposing wage restraint on itself: "[the un-]employment problem...[can only] be solved through a redistribution of existing work opportunities and working incomes at the expense of the great majority of those who are presently employed."(53) In Sweden and Denmark this has proven difficult to accomplish. In Australia the arbitration apparatus made it possible to enforce this bargain made 'within one class.'
Not all of the institutional features of the Australian model can or should be transferred to Europe. In particular, Europeans would probably find two things troubling. First, regardless of how Australians feel about it -- and many see means tested welfare as simply part of the social wage -- Europeans would probably regard means tested welfare as stigmatizing unless the test were set quite high. Attacking the middle class's right to feed at the public trough is also politically irrational even if it is fiscally rational.
Second, arbitration continues to weaken the ability of employers and labor unions to undertake collective action; they are passively aggregated by the system and do not have the edifying experience of organizing themselves. Arbitration reflects the power of a state that constructed and dominates civil society. In Europe the principle of free collective bargaining is deeply entrenched and defended by unions and employers' organizations, which themselves are products of civil society, not the state. While state imposed wage settlements were common in Denmark and in Sweden in the 1980s, unions and employers both found these hard to swallow.
Despite this the essential features of the Australian system are worth noting and perhaps imitating, because even if they are not optimally suited to a world of mobile capital and indebted states they may be better suited. Given everyone's inability to influence macro-economic outcomes in an environment of internationalized capital, arbitration has provided the Australian working class with a measure of security and income stability. Targeted benefits have allowed the state to reduce overall tax pressure while benefitting those whose labor market outcomes place them at the bottom of society. State enforced wage leveling reduces the size of that bottom, while freeing unions to pursue other goals, such as economic democracy, in negotiations with firms which themselves will be trying to maximize their own operational flexibility.
Of course all of this rests on a particular class compromise between unions and capital in the non-traded sector. The 1996 Liberal Party electoral victory led to proposals for sweeping changes to arbitration. So far, firms in the metals sector have tried to preserve the current arbitration structure and prevent wholesale change along the lines of New Zealand's Employment Contracts Act, which individualizes wage bargains. So far it appears that the Liberals will have to continue on the path set by the ALP in the 1980s -- the encapsulation of enterprise bargaining with-in the arbitration system. If so, this will provide strong evidence both of the durability of the class compromise that Australian wage and welfare institutions reflect and of their utility in a world of increasingly mobile capital.
NOTES:
Author's Note Thanks to Francis Castles, John Echeverri-Gent, Ira Katznelson, Carol Mershon, Steve Rhoads, Peter Swenson, David Waldner, and Rob Watts for comments. University of Virginia Summer Humanities and Bankard Fellowships helped fund this article. All errors remain mine.
(5) Peter Katzenstein, Small States in World Markets (Ithaca: Cornell University Press, 1985).
(9) Przeworski and Wallerstein, p. 217.
(10) Przeworski and Wallerstein; Przeworski, ch. 5.
(11) Przeworski and Wallerstein, p. 220.
(17) G sta Esping-Andersen, Politics Against Markets (Princeton: Princeton University Press, 1985).
(20) Peter Baldwin, Politics of Social Solidarity (Cambridge: Cambridge University Press, 1990).
(26) Plowman, p. 28; Anderson, pp. 211-221.
(27) Offe and Wiesenthal, "Two Logics of Collective Action."
(28) Esping-Andersen, Three Worlds of Welfare Capitalism, pp. 52, 70-71, and ch. 2.
(38) Rob Watts, Foundations of the National Welfare State (Sydney: Allen and Unwin, 1987).
(40) Castles, "On Sickness Days and Social Policy," pp. 16-17.