YOU MAY PRINT ONE COPY FOR FAIR USE

Author's Note:  Thanks to Karen Anderson, Leslie Eliason, and Jonas Pontusson for comments on the version delivered at the 1999 Society for the Advancement of Scandinavian Studies Annual Meeting; and to Asbjørn Sonne Nørgaard and David Waldner.  All errors remain mine.

Abstract:

The 1980s and 1990s saw economic or employment "miracles’ in Denmark, Australia and most famously the Netherlands. This paper analyzes the dynamics and substance of Danish industrial policy, collective bargaining and fiscal retrenchment to see how much of favorable outcomes in export performance, employment performance, and fiscal balance can be attributed to pluck (intentional, strategic remediation of dysfunctional institutions to make them conform with the external environment), luck (environmental change that makes formerly dysfunctional institutions suddenly functional) or just being stuck (endogenous, not entirely strategic change that leaves institutions in conformity with the environment). It addresses these issues because they are fundamental to the economic viability of the welfare state. It addresses these issues to remedy biases in the literature towards Sweden-as-model, towards pessimism about the welfare state’s survivability and towards privileging intentional action. The analysis finds that stuck (endogenous dynamics) probably explains as much as pluck (strategic choice), suggesting only limited transferability for policy lessons from the "miracles."

The Danish "Miracle": Luck, Pluck or Stuck?

Introduction

The 1980s and 1990s saw a number of economic or employment "miracles’ in the OECD, of which the Dutch miracle is the best publicized. However both Australia in the 1980s and Denmark in the 1990s also experienced employment miracles. The stark contrast of rising employment in Holland, Denmark and Australia with rising unemployment in France, Germany and Sweden created a small literature looking for transferrable policy solutions. The first three were particularly attractive to analysts uneasy with the British or New Zealand neo-liberal policy route to employment gains, because they generated relatively good employment outcomes without sacrificing essential features of their welfare state.

Most of the literature on these economic miracles assumes that actors’ intentional policy behavior changed dysfunctional local institutions in ways that created desirable economic outcomes. However deliberate policy choices that bring local institutional structures into better conformity with the environment are not the only possible cause for desirable economic outcomes. Positive outcomes can also occur if the external environment changes in ways that make what were dysfunctional and unchanged institutional structures more functional in the context of the new environment. Similarly, dynamics which are endogenous to a given but dysfunctional institutional structure can create unintended changes which accidentally make institutional structures more functional in the context of a changed environment. This paper analyzes the Danish case to see how much of the Danish miracle can be attributed to pluck (intentional change), luck (environmental change) or just being stuck (endogenous, not entirely strategic change).

This analysis thus addresses three biases in the literature. First, parochially, the literature on welfare states and social democratic welfare states in particular has always taken Sweden as the epitome of both. This literature saw Denmark as weak version of the generic ‘scandinavian" (i.e. Swedish) model, whose incomplete acquisition of Swedish institutions and policies left it burdened by higher unemployment, and fiscal and current account deficits. The 1980s seemed to confirm this vision, for Denmark entered the 1980s on a fast train to macro-economic hell, albeit in the first class coach, while Sweden seemingly pursued a successful third way between outright socialism and neo-liberalism. But in the 1990s Denmark and Sweden traded places. Sweden endured rising unemployment, and current account and fiscal deficits, while the Danish economy absorbed labor market entrants and generated current account and, for a while, fiscal surpluses. So an analysis of Denmark on its own terms usefully corrects the implicit Swedo-centrism of most writing on Scandinavia.

Second, precisely because the generic ‘scandinavian social democratic model" was so closely identified with Sweden, Sweden’s difficulties call into question the degree to which social democratic welfare states can survive in what is taken to be a new global economic environment. If pluck in any degree underlies Danish prosperity in the 1990s, this suggests transferable policy lessons for those wishing to avoid neo-liberal policy choices. In fact the policy choices made in the Netherlands, Australia and Denmark are remarkably similar, suggesting potential policy choices different from both (failing?) Sweden and the standard neo-liberal package.

Finally, this analysis corrects for the implicit privileging of intentional action in causal explanations by explicitly considering the possibility that while actors considered their actions intentional, their policy choices might have been essentially endogenous outcomes of specific institutional structures and particular strategic responses to environmental conditions. If markets, like any environment, select for and reward specific institutional structures and behaviors, then some actors will always appear to have made the "correct" strategic response to their environment, even if they chose their strategy somewhat randomly. But this may not necessarily be the "optimal" response or strategy. Furthermore because actors do respond to their environment, that environment is always changing, eroding the degree to which any prior "best" response to a given environment fits the current environment. At any given time, stochastic changes rather than intentional action may create what looks like an "optimal" or "best" response to a given environment. But in this situation causality will be located in the system (in the environment created by other actors’ behaviors), not in the choices of specific actor(s) who are usually studied in isolation.

This paper will test luck, pluck and being stuck as competing explanations for Danish ‘success." The answers to these questions illuminate the narrow trading places question and also broader questions about whether any welfare state can survive or thrive in a world of increasingly disembedded liberalism and increased economic integration and competition. To foreshadow the findings, Danish capacity to preserve the welfare state in the face of severe macro-economic constraints without generating popular dissatisfaction is only partly accidental. Intentional reform whose potential was created by Danish political and social institutions did ameliorate Danish macro-economic problems. While this makes it hard to adjudicate adequately between pluck and stuck arguments, the Danish case suggests that neither environmental change nor endogenous dynamics are sufficient or necessary conditions for a crisis of a welfare state composed of publicly funded services and transfers. Sufficient conditions for economic crisis and a crisis of the welfare state rest in domestic policy choices and institutions.

The Structure of the Analysis

Roughly speaking this paper presents nine different explanations for Danish success in the 1990s, by applying luck, pluck and being stuck to the issues of current account balance, employment, and fiscal balance. These three policy areas all have immediate consequences for the economic sustainability of a welfare state based on tax funded services and transfers. For many analysts (see e.g. Martin 1996 and Huber and Stephens 1998) unemployment is the critical problem, as it raises expenditures, decreases revenues and erodes social solidarity. Fiscal deficits also clearly erode the long term sustainability of this kind of welfare state as deficits cumulate into debt and as interest payments begin to crowd out services and transfers. Finally current account deficits and public foreign debt are simply an external and more pernicious version of fiscal deficits. Foreign debt usually carries higher real interest rates than domestic debt and cannot be monetized away. Current account deficits are also a proxy for competitiveness, and so are often also associated with higher unemployment as imports crowd out local production or as competitors displace exports from third party markets. In short, these three issue areas represent core economic preconditions for a tax, service, and transfer welfare state. However I will also touch upon narrowly political issues of sustainability while discussing fiscal balance. I will also devote more space to the question of public sector behavior and fiscal deficits because these touch directly on the question of welfare state sustainability.

Luck explanations start from the usual assumptions about institutional stickiness and inertia, and argue that while Danish institutions have been relatively constant over the post war period, the external environment has changed in ways that favor either or both of Danish production and public sector institutions. Pluck arguments in contrast assume that while institutions are relatively sticky, actors (again in either production or the public sectors) consciously remodeled their own or others’ institutions to make them comport better with the external environment. Finally stuck arguments suggest that endogenous dynamics made Danish institutions change in ways that comported well with a new environment. All nine sorts of explanations thus address both Danish problems before the 1980s and Danish success afterward.

The structure of the luck arguments is transparent: dysfunctional institutions promoting excessive wage gains, a weak capacity to export and high (imported) consumption in the 1960s and 1970s came into their own in the 1980s and 1990s when world markets shifted in favor of differentiated quality production and when declining global interest rates made past deficits less burdensome. Danish SMEs were well suited to fragmented markets that put premia on good design. Danish public sector institutions didn"t need to be fixed once their interest cost decreased.

The pluck arguments are more complex. Briefly, with regard to the private sector, central and corporatist actors successfully resolved problems plaguing the provision of collective goods for an economy characterized by SMEs and high levels of long term unemployment. On the enterprise side an innovative industrial policy provided R&D and other collective goods to businesses which would otherwise be too small to fund these items themselves. This was formalized into a cluster strategy in the early 1990s. On the labor market side a well timed shift to an active labor market policy first eased tight and potentially inflationary labor markets in the early 1990s and then firmed up softening demand for labor in the mid to late 1990s. Meanwhile collective bargaining patterns shifted power downward from national organizations and upward from shop stewards towards sectoral associations (so called cartels). This kept wage increases below productivity increases. On the public sector side, central actors with a reform / reorganization agenda used two interlocking attributes of the public sector’s institutional terrain -- intergovernmental corporatism and administrative corporatism -- to generate dynamics in the public sector that moderated the growth of the public economy, thus allowing a return to fiscal stability.

Stuck arguments would suggest that Danish institutions evolved incrementally according to logics of appropriateness held by actors in those institutions, and that the institutional outcomes were either better than prior configurations or at least less dysfunctional than those into which the competition stumbled (March and Olson 1989). Actors’ conscious policy choices were conditioned by embedded notions about the social purpose of their activity and what could be attained given the institutional landscape in Denmark. In that sense they were not perfectly free choices, but rather conditioned by accidental or incidental qualities of those organizations. Because the external environment surrounding Danish production and public sector institutions was also not characterized by optimal organizations, Danish organizations merely had to be less dysfunctional than their global competitors in order to look "good." Note that ‘stuck" arguments are thus not arguments for convergence towards any optimal organizational form, nor do they offer much guidance about policy transferability. As Alchian and others have argued, markets are like ecologies (Alchian 1950). Firms display a multitude of strategies -- expressed as organizational structures -- which can be well or ill suited to their environments. Competitive pressures force firms to adapt their strategies (organizational structures) but they do not enforce conformity. Ecologies with multiple niches permit multiple successful strategies, and both successful and unsuccessful strategies change the environment. Moreover, competitive pressure on any given organization can be diffuse, if it is in a ecological niche (market) with few competitors. Pressures on public sector organizations are even more diffuse, since they have quasi-parasitic sources of revenue (or put differently, something approximating a monopoly in the provision of regulated and common pool services).

Luck, Pluck or Stuck: Private Sector Exports and the Trade Balance

Luck

A luck perspective would argue that Danish failure to avoid current account deficits in the 1960s through 1980s derives from a mismatch between an industrial sector dominated by small and medium sized enterprises (SMEs) and an environment favoring long production runs of standardized goods; Danish 'success' in running current account surpluses in the 1990s is nothing more than a fortuitous change in world markets favoring the differentiated goods at which SMEs excel. Secondarily, Denmark’s bad luck in joining the European Union meant the general deceleration of European growth after the 1970s disadvantaged Danish exports, of which two thirds go to post-Single Market EU markets.

The arguments about fordism in the post war period are reasonably well known. Postwar reconstruction and government regulation of the economy favored production of standardized goods. Production technologies mixing assembly lines and dedicated capital goods with unionized semiskilled labor gave management strong incentives to minimize design and work rule changes that might interrupt long continuous production runs. This environment and strategy favored large firms making standardized goods. In contrast, the break up of fordist mass markets in the 1980s and 1990s should favor smaller, more agile firms producing design intensive goods with skilled labor and general purpose machinery.

Denmark’s industrial sector is both small in relation to the economy and small in terms of structure (see Tables 1-3 below). Only 20 percent of private employment is in firms with than 500 workers, below the level predicted by a regression over OECD18 economies taking into account size of the economy and export dependence. On the other hand, the share of employment in enterprises with fewer than 20 employees is the second highest in the OECD, after Italy. Manufacturing specializes in consumer non-durables (about 5 percent of exports), design intensive goods, intermediate metal inputs for other firms, specialized machinery and chemicals (about 10 percent of exports) and the processing of agricultural outputs (about 20 percent of exports). Most of the first four products are characterized by short production runs, high levels of differentiation and low levels of R&D, while most of the latter were in the 1980s typically fordist in having long production runs of undifferentiated goods made by continuous flow processes. Danish service exports (about 20 percent of exports) are also produced by large firms with scale advantages.
 
 
 

Table 1: EMPLOYMENT IN PRIVATE SECTOR BY ENTERPRISE SIZE, 1991, % of total
Enterprise size by # person employed	1-19	20-99	100-499	500+
Denmark	39.8	22.8	17.3	20.1
Belgium	25.2	20.8	19.1	34.9
Finland (1989)	26.3	18.0	17.0	38.6
Germany	25.9	18.7	18.2	37.2
Spain	38.8	24.3	15.5	21.4
OECD, Economic Survey -- Denmark, 1994, Paris: OECD, 1994, p. 62.

Table 2: SIGNIFICANCE OF LARGE SCALE ENTERPRISE IN THE ECONOMY, 1982
	Ratio of value added by 10 largest Swedish firms to whole manu-facturing sector in:	Average sales by 10 largest firms in 1982, millions of 1982 SEK	Share of manufacturing value added by 10 largest firms	Share of value added done in domestic market for 10 largest firms
Denmark	1.3 : 1	3000	13.9	81.3
Finland	1.5 : 1	5200	23.3	84.9
Norway	1:1	4600	22.1	72.2
Sweden	n/a	23300	32.9	53.3
Lars Oxelheim,"The Largest Nordic Manufacturing Companies," in Economic Growth in a Nordic Perspective, pp. 185-193, 198.

Table 3: TECHNOLOGY AND INVESTMENT INTENSITY, 1990
Technology Level (OECD scoring)		Distribution of Value Added, %	Investment intensity (Gross investment / Value Added)
High	Denmark	12.4	0.12
	World	20.0	0.08
Medium	Denmark	27.8	0.15
	World	30.1	0.10
Low	Denmark	59.8	0.15
	World	49.9	0.09
OECD, Economic Survey: Denmark, 1994, Paris: OECD, 1994, p. 66.

 

This industrial structure arguably creates some problems for the Danish economy. Despite the predominance of SMEs, a substantial share of Danish exports are scale intensive goods like food, wood products, metals with low technology content, and transportation equipment. Denmark’s firms are also heavily concentrated in low technology, low growth sectors. In theory at least, this should predispose them to slower growth. Small firms generally lack the funds to enter new markets and to generate significant amounts of R&D; in Denmark the 20 percent of workers in firms with more than 500 employees were associated with 60 percent of R&D spending in 1995 while the 27 percent in firms under 100 workers were associated with only 10 percent of R&D spending (OECD 1999, p. 145). Finally the absence of a well developed shares market in Denmark might make it hard for firms to attract new capital. The Danish share market amounted to only 8 percent of GDP in 1980 in contrast to an average of 25 in comparable economies like Sweden, Australia and New Zealand.

These problems should make it difficult for Danish firms to export or to substitute local production for imports. And indeed, Danish exports by and large did not grow as fast as export markets in the 1970s, when the fordist model (while in crisis) created conditions conducive for scale intensive production. Danish export performance in the 1970s was below what a constant market shares analysis would have predicted, partly because Denmark exported goods with below average growth rates and partly because it exported those goods to countries with below average growth rates (Horwitz, 1984, p. 271). Although devaluation in the late 1970s helped Danish exports grow more rapidly 1979-82 than the OECD group, the trend was declining growth through 1988, revival through 1992 and then a decline again through 1997 (See Figure 1 [click] which compares Danish, Swedish and small OECD  export performance and  Figure 5  for trade balance). Over the 1980s, manufacturing exports underperformed export market growth by about 11 percent (OECD 1994, p. 70).

Danish export performance surged from 1986 through 1992, a period characterized by stagnant domestic demand and then rising demand in eastern Europe ( Figure 1 ). But logically if there had been an environmental shift to post-fordist demand structures in the 1980s, Danish export performance would have been consistently high from the mid 1980s on, rather than plummeting precipitously after 1992 and then recovering slightly in 1997. Despite this relative decline, Denmark generated merchandise trade surpluses through 1999, rather than the deficits that characterized the period before 1986.

Pluck

What about pluck? The Danish government and private sector actors generated a series of industrial policies addressing the structural deficiencies of SMEs in the 1980 and 1990s. Both kinds of actors addressed the under provision of collective goods. While there was at least initially no effort at a comprehensive program, by the late 1980s various elements added up to something like a comprehensive program, spending around DKK 1.3 billion in all (Christiansen and Sidenius, 1998, p. 254-255). In 1983 the existing Technology Board (Teknologistyrelsen) put forward three new programs for technological development after the OECD pointed out the typical weaknesses of Danish SMEs (Annerstedt, 1989). The Board helped create networks to rapidly diffuse knowledge about new technologies, new managerial strategies, quality control and new financing arrangements. The Academy for Technical Sciences and two other national technical institutes helped develop dispersed, locally integrated engineering and consultancy services through the Danish Technological Services Network. R&D expenditures by Danish firms rose steadily from 1.13 percent of GDP in 1986 to 2.02 percent by 1995 (OECD, 1996 p. 159, and OECD 1999, p. 145.)

On the financing side, the Teknologistyrelsen allocated around DKK 2 billion for the acquisition and dispersal of foreign source technologies. The Danish National Bank and private actor used DKK 0.5 billion to capitalize Dansk Udviklingsfinansiering A/s to provide venture capital in the absence of a deep shares market (Christiansen 1989, p. 42). Two other existing public sources of funds were also redirected towards venture funding and the promotion of public-private cooperative R&D. The government now guarantees up to 50 percent of venture type loans from banks (OECD 1994, p. 88). The Danish state absorbed the exchange rate risks of overseas borrowing through earmarked loans to manufacturing. The stock of outstanding loans amounted to 8 percent of GDP by 1983 (OECD 1984, p. 38). The shares market itself was deregulated in the late 1980s, and by 1996 total capitalization of shares handled on the exchange equaled 41 percent of GDP. Some of these programs were eliminated in a push to balance the budget in 1990. But the self-financing public-private networks lived on. Finally, the government elaborated a strategy based on seven local clusters of firms (called ‘strongholds’). This included an aggressive effort to develop tradable services. Note that all this is exactly what the governments in Taiwan, Québec, and Australia did in the 1980s to overcome the same kinds of development limitations created by SMEs in their own economies.

On the private sector side, individual and collective actors also responded to SME’s problems. Danish firms seem to have compensated for some of their weaknesses by aggressively substituting capital for expensive Danish labor at all levels of technology (see Table 3). This probably allowed them to increase their share of low technology markets. During the early 1990s the union controlled Lønmodtagernes Dyrtidsfond cooperated with Denmark’s largest food processor, Mejeri Danmark, to consolidate the food processing industry, by buying up, e.g. Tulip, a slaughtering operation, and launching common brands in major European markets. Tulip itself had formed in 1990 from the merger of four smaller food processors. As Nielsen (1991) has argued, this constituted a "negotiated" private industrial policy, in which organized capital and labor cooperated to fend off foreign takeovers, promote more R&D and generate external financing for firms.

Stuck:

Disproving a ‘stuck" argument is easier than proving one. A stuck argument ultimately rests on the degree to which actors’ behavior flowed from a given logic of appropriateness rather than a strategic appreciation of and response to their environment. Most private sector actors certainly could be accused of simply acting out of a pre-existing logic of appropriateness, especially Mejeri Danmark’s drive to consolidate food processing. Food processing firms everywhere consolidated in the 1980s. Similarly the preference for cooperative efforts and the diffusion of knowledge through extension systems was characteristic of Danish agricultural policy in the 19th century and carries over into all industrial policy today.

Some aspects of government efforts, by contrast, suggest a true strategic vision and moreover one ideologically and normatively at odds with traditional Danish practices. Annerstedt (1989) makes the strongest argument that "industrial modernists’ pushed for a new approach after OECD analyses highlighted SMEs’ technological deficiencies rather than simply looking at Denmark’s relative unit labor costs (RULCs). These modernists stepped into the vacuum created by dissensus between government and business organizations, and did their work outside the usual policy networks and arenas (Annerstedt, 1989, p. 136). Their program also diverged somewhat from the usual "help to self help" orientation of Danish government policy in its commitment to making firms change, and strongly in the intensity of government involvement and directiveness. Not surprisingly, funding for core elements of this program was cut in a 1989 budget battle in which the social liberal party Venstre defended small industry’s right to be left alone against the more interventionist, big-firm oriented Conservative party (Nielsen, 1991, p. 308).

Evaluation:

The weight of evidence here mostly favors a stuck argument. Luck should have produced a secular rather than cyclical or erratic pattern to Danish export success. The industrial policy of the 1980s clearly helped boost Danish export performance for a while, suggesting pluck matters. But if pluck explained everything, then Danish export shares would have expanded gradually after the initiation of the aggressive industrial policy of the 1980s. However the termination of much financial support for industrial policy in 1990 correlates well with declining export performance, and this termination also corresponds to political changes that restored policy making to patterns "appropriate" with normal Danish policy making. Finally Danish export growth also seems sensitive to absolute and relative labor costs and not just to problems with firms’ size. The abrupt slide in Danish export growth from 1992, and the corresponding decline in the trade surplus from 6.0 percent in 1992 to 3 percent in 1997, indicates that the elimination of the trade deficit in the later 1980s and early 1990s has more to do with the unusually high level of demand created by German reunification, and perhaps by falling RULCs 1986 to 1988 (about which more later), than with a favorable structural change in Danish industry or a determined effort to change structures. Score this two-thirds stuck, one third pluck.

Luck, Pluck or Stuck: Private Sector Collective Bargaining and Employment

Luck

The luck argument in labor markets is very short: markets worked in the normal fashion. Rising unemployment from 1987 on led quite normally to wage moderation and a rising share of income going to capital just as the international economy began to pick up. The wage share of business sector value added fell from around 75 percent in 1980 to 66 percent in 1985, recovered to 70 percent in 1987, and then fell again to 61 percent in 1994 (OECD 1996, p.102). This in turn permitted the modest export expansion noted above and a downturn in unemployment despite a stable employment to population ratio in the mid to late 1990s (Figure 2  and  Figure 3). There is a .41 correlation between unemployment rates and capital income in the business sector from 1988 to 1997; the correlation between RULCs and capital income is similar. Danish wage restraint can be explained easily by reference to its centralized bargaining structure and the usual Calmfors-Driffell model.

The luck argument has two problems. The first concerns timing. There is an anomalous period in the mid 1980s when unemployment fell from 10.5 to 7.7 percent 1983 to 1987 (local definitions) while RULCs rose from 81.3 to 103.4 (1990 = 100) and while labor recovered 4 percentage points of value added which a luck argument cannot explain. Similarly just as Danish RULCs began to edge upwards, in the late 1990s, unemployment actually declined rapidly from 10.1 percent in 1995 to an estimated 6.2 percent in 1999. Second, Danish collective bargaining practices changed considerably in the late 1980s, away from a centralized system, but it is precisely at this point that relative unit labor costs stabilized and labor gave up 9 percentage points of value added to business ( Figure 4 ).

Pluck

How did Danish labor market actors get relative unit labor costs under control and boost employment? Here we can rely on the excellent study by Jesper Due, et al. (1994), who argue that Danish collective bargaining arrangements experienced simultaneous decentralization and centralization after 1989. Centralized negotiations among a small number of new 'cartels' set broad frameworks in which specific local negotiation over wages and conditions could then occur. The creation of cartels simultaneously removed power in wage negotiations downward from union and employer confederations and upward from stewards and firms.

Prior to these changes Danish collective bargaining generated two different kinds of contracts. Centrally bargained standard or normal wage contracts set industry wide, essentially non-negotiable wage increases and conditions. Minimum wage contracts set a floor beneath second tier, local plant level bargaining that occurred without central supervision. From 1983 on employer associations and some unions introduced a third variation into this system by creating minimum-pay contracts in which central negotiations simply set a floor beneath wages and conditions within five large sectoral groupings, but then all increases were bargained locally at the plant level. Organizational changes paralleled and mediated this change, with distinct, sector . specific bargaining "cartels’ emerging on each side. Table 4 shows the relative coverage of different bargaining arrangements
 

Table 4: Wage bargaining structures in the DA/LO area (percentage of employees covered)
year	1989	1993	1995	1997
Normal (i.e. centrally set) wage increase	34	16	16	16
Minimum wage (local bargaining up to central maximum)	32	13	12	17
Minimum pay (central minimum, local top-up)	30	67	61	46
Fully local bargaining	4	4	12	21
Source: OECD, Economic Survey: Denmark, 1999, p. 66.

Minimum pay contracts remedied the tendency for centrally set wages to generate across the board increases in pay for sectors that were unable to generate commensurate productivity increases, as well as the difficulties centrally set wages and conditions created for employers seeking to induce skills formation and introduce multi-task work practices (Pontusson and Swenson, 1996). Rather than having wage increases imposed externally on employers, whether via central negotiations or automatic cost of living increases, wages would now increase predominantly from local conditions. Arguably this big change in the structure of private sector collective bargaining created significant stability in RULCs (which measure wage growth in relation to productivity) from 1986 on, as  Figure 4  shows. Productivity has grown faster than compensation in Denmark since 1986 (OECD 1996, p.102). Wage restraint then generated rising employment, large trade surpluses ( Figure 5 ), but not better export performance for Denmark. (However it might also have averted even worse deterioration in export growth). In contrast, during the same period Sweden resorted to large devaluations to push relative wage costs down and thus boost export performance at the expense of imported consumption.

Finally well timed changes in labor market policy after 1994 helped activate workers who otherwise would have slid into long term unemployment, reversing a secular decline in the employment to population ratio in the mid 1990s ( Figure 3 ). Expanded early retirement and paid leave schemes absorbed an additional two percent of the labor force, bringing their total to about eight percent, and helping bring open unemployment down from twelve percent in 1992 to six percent in 1998.

Stuck:

As with industrial policy and exports, evidence that actors consciously chose to change collective bargaining structures in the direction described above provides enticing evidence for a pluck argument. The outcomes Due et al. describe did not emerge spontaneously but rather as the outcome of political struggles within employer organizations and unions as well as between them.

But a comparison with other countries that started out with similar collective bargaining structures and similar problems shows quite similar responses to those problems in pursuit of successful wage restraint and employment growth. The evolution of Danish collective bargaining in the 1980s/90s mirrors events and outcomes both in Australia and in the Netherlands (Schwartz, 2000; Visser and Hemerijck, 1997). Both of these countries had relatively centralized collective bargaining systems in which the state generalized wage gains and cost of living increases across sectors through processes similar to the "concatenation" found in Denmark. Dutch state mediators or Australian Arbitration courts intervened recurrently in bargaining and this frequent resort to legislated or juridically imposed settlements meant labor market actors conducted their conflicts under the shadow of hierarchy. Consequently organized actors sought to reestablish their autonomy in the 1980s by behaving responsibly and using state institutions to punish or discipline potential defectors, rather than suffering indiscriminate state sanctions.

Employers in all three countries sought one firm-one contract type bargains from one industry-one organization type actors, and all three bargaining systems saw rising proportions of purely locally negotiated labor contracts. The progressive decentralization of the Australian system towards so-called enterprise bargaining during the early to mid 1990s strongly resembles events in Denmark, particularly in its formal link between second tier (local) wage gains and productivity gains and the central negotiation of wage floors. The Dutch system also decentralized in the 1980s although wage restraint was exchanged for part time rather than full time work.

Moreover in all three countries decentralization followed a long established logic of appropriateness present in the metals industry, and was primarily driven by actors located in the metals industry. Decentralization essentially made national bargaining structures conform with structures latent in metals. Metals employers and unions had long standing traditions of taking wages partly out of competition and then letting employers and workers adjust local wages to local conditions (Due, et al., 1994; Thornthwaite and Sheldon, 1996).

Evaluation:

A luck argument is completely unpersuasive in light of both intentional action, the evidence on the timing of changes, and the similarity of change across several countries. Given that actors deliberately changed their behavior, can a stuck argument be persuasive? The congruence between Danish changes and equally successful change elsewhere, and the similar origin of these changes in established bargaining patterns in the metals industries suggests that a stuck argument cannot be completely dismissed. While Australia, the Netherlands and Denmark all had centralized bargaining systems they also all had incomplete centralization of both business and labor organizations. Perhaps in this kind of structure endogenous change allows politically dominant firms to impose their own, established preferences on the others. Score this one two-thirds pluck, one-third stuck.

Luck, Pluck or Stuck: the Public Sector and Fiscal Balance

Luck

The public sector luck argument is also fairly simple. The single fastest growing source of public sector deficits was rising interest expenditures on old deficits from the 1970s and 1980s (see Table 4, Table 5, line 4 and  Figure 6 ). As real interest rates rose through the early 1980s, the cost of running deficits climbed rapidly ( Figure 7 ) . In the mid 1980s, despite a narrowing of the interest rate differential between Denmark and Germany due to Danish ERM/EMS entry, real interest rates hovered over 7 percent; debt service on gross public debt amounting to 72 percent of GDP (roughly half of which was foreign owned) absorbed 8.8 percent of GDP. But as interest rates declined to less than 5 percent by 1997, the cost of servicing old debt declined to 6.7 percent by the late 1990s, permitting a falling total fiscal deficit (Table 5, line 5). The timing of shifts in the total deficit comports somewhat with a luck argument. Slightly rising real interest rates in the mid 1990s created renewed deficits after a period of falling interest rates and surpluses in the late 1980s; falling interest rates after 1995 are associated with renewed surpluses. The problem with the luck argument is that it fails to explain the origins of public sector deficits in the first place, how and why surpluses replaced deficits during the high interest rate environment of the early to mid 1980s, and why deficits reemerged when real interest rates were actually somewhat lower than int e 1980s. Part of the answer surely is rising taxes in the 1980s, which helped make rising interest rates affordable. As well, the expansion of transfers to persons (Table 5 line 6) nicely mirrors the changing interest rate environment. Once interest rates fell, transfers expanded from 19.6 percent in 1989 to 24.1 percent in 1995, more than absorbing the slack created by falling interest rates.
 

Table 4: INDEX OF REAL GENERAL GOVERNMENT EXPENDITURE IN DENMARK 1980-1992

(1980 = 100)	1980	1982	1984	1986	1988	1990	1991	1992
Consumption Expenditure	100	106	105	109	112	110	111	112
Transfers, excluding interest	100	109	112	113	127	134	140	147
Interest	100	153	262	262	239	229	233	219
Total	100	111	120	122	128	129	132	135
Consumption,  as % of total	51	49	44	44	45	44	43	43
Total as percentage of GDP,  excluding capital outlays.	56.9	61.7	61.4	56.6	59.8	59.3	59.9	61.1

Source: Munk Christiansen and Goul Andersen, 1992.

Table 5: FISCAL INDICATORS (% of GDP)
General Government account	1980	1986	1987	1988	1989	1990	1991	1992	1993	1994	1995
1. Current Receipts	52.9	59.1	59.7	59.5	59.1	56.9	56.7	58.2	59.9	60.3	59.5
2. Non-interest Expenditure	52.3	46.9	49.0	51.1	52.2	51.1	51.4	54.2	56.0	56.6	54.4
3. Primary Budget Balance	0.7	12.2	10.7	8.4	7.0	5.8	5.3	4.0	3.9	3.7	5.1
4. Net Interest Expenditure	3.9	8.8	8.3	7.9	7.5	7.3	7.3	6.8	7.8	7.1	6.7
5. Budget Balance	-3.3	3.4	2.4	0.4	-0.5	-1.5	-2.0	-2.9	-3.9	-3.5	-1.6
6. Transfers (exc. Interest)	18.4	17.7	18.4	19.6	20.7	20.5	21.4	22.0	23.0	24.7	24.1
7. Consumption	26.7	23.9	25.2	25.6	25.6	25.2	25.2	25.6	26.3	25.6	25.2
Source: OECD, Economic Survey: Denmark, Paris: OECD, various years

Pluck

Public sector 'pluck' arguments rest at the intersection of public choice theories and organizational behavior under the shadow of hierarchy (Scharpf, 1997). Public choice theories (PCT) provide a more robust explanation for the emergence of public sector deficits and the expansion of transfers than the bad luck of rising interest rates. The reaction of Danish public sector organizations operating under the shadow of a hierarchy threatening their institutional power in the pursuit fiscal restraint also provides a robust explanation for the containment of budget growth and thus those deficits.

PCT argues that producers seeking budget growth and client groups seeking increased services and transfers should cause welfare spending to rise inexorably and make budgets 'sticky downwards'. Danish public spending grew 63 percent in real terms in the 1970s; public employment grew eleven percentage points to 32 percent of the labor force (L. Nørby Johansen, 1986, pp. 307-311). Increased spending was not matched with increased taxes in the 1970s, leading to a widening fiscal deficit.

But the 1980s present a puzzle for PCT predictions about pervasive state incapacity to control entrenched public sector producers. While Danish fiscal restraint in the 1980s was implemented in ways which did not upset delicate institutional balances among welfare service producers, well-organized service producers inside the welfare state did not manage to boost their budgets. Quite the contrary: while interest expenditure expanded in the 1980s, discretionary spending, particularly on consumption, stayed remarkably flat. As a share of GDP, consumption spending actually fell from 1980 through 1986 and only recovered to 1980 levels in 1996. In Thatcherite Britain, by contrast, public consumption has grown more or less continuously since 1985 (OECD, 1999, p. 56). However Danish politicians were more than willing to share the mid 1980s and mid 1990s budget surpluses with core constituents as transfers.

How did this happen? Put simply, institutional actors cannot maximize all their desires at once. Critically, public sector producers can only (ab)use their position if the asymmetries that give them their privileged position stay intact. When producer organizations' ability to legitimately control an activity is threatened, these institutional actors will defend long term interests and sacrifice short term budget maximization. Established rights, powers and procedures are more important than short run monetary gains, because institutionalized power involves high sunk costs and eliminates uncertainty about the future.

A pluck argument has to assert that central politicians in Denmark used fiscal crisis to politicize the asymmetries privileging producer groups, and thus restrain those actors. This parallels labor market events both in the Netherlands after Wassenaar and in Australia under the Accord. In both cases the central state used its ability to deny the extension of collectively bargained contracts to unorganized sectors to extract wage restraint from collective actors. What is unusual in Denmark is the extension of this pattern to a public sector in which producers groups exercise managerial control. Danish central politicians used two institutional features to constrain consumption spending. The first is the localization of public consumption, production, and financing, which creates regularized inter-governmental negotiations over budgets, and which we will call governmental corporatism. The second is the domination of public service sector administration by professional producer groups rather than professional managers, which we will call administrative corporatism.

Governmental corporatism: Localization's consequences

Public consumption, the largest part of Denmark’s welfare state, is extremely localized in terms of provision and financing. Considerable local flexibility in welfare provision increases all local politicians' responsiveness to citizen and producer group demands. On the funding side, Denmark has one of the highest ratios of local to central employees in the OECD, and local income taxes cover a substantial share of local expenditures (Christiansen 1995). Sweden aside, Danish localities raise a greater proportion of total OECD governments’ revenue than in any other unitary state, roughly 31 percent as compared to an unweighted average of 6 percent (OECD, 1992: 194).

In Denmark local decisions over spending immediately thus affect the center's capacity to maintain fiscal balance, since the center. Superficially the central government has few instruments to implement national policy priorities. However, local autonomy is constrained three ways. Central transfers to localities make up a large part of total local outlays, allowing the center to exert influence on local budgets. Second, the center mandates and regulates most local responsibilities, as well as inter-municipal redistribution of revenues. Finally, nationwide collective agreements constrain the organization of welfare service. Even if local unions and professional organizations develop localized demands, they closely coordinate these activities with their national organizations.

This standoff puts the central and local government into yearly budget negotiations which emerged from the consolidation of local autonomy after the 1970 Local Government Reform. Meetings between the Finance Ministry and the Association of Local Governments (ALG here, but in Danish, KL) determined the level and content of central transfers to localities. During the 1970s the central government desired considerable public sector growth to absorb labor, so there was little conflict (Nannestad 1991). But fiscal stress in the 1980s motivated Danish central government to try to control local spending. The center then waged a long and quite successful budget battle with localities, using changes in the structure of central funding and regulation, as well as financial penalties, to change local politicians’ spending behavior.

The Finance ministry changed central funding from a system of straight reimbursement for local service expenditures (e.g. health) to a system of block grants, which it then systematically cut (Schou, 1988). Consequently, while in 1982 central government grants provided 35.6 percent of funding for local government activities, by 1985 they provided only 28 percent (Eliason, 1992, p. 596). The center also punished overspenders by sequestering funds. At the same time the center progressively loosened control over what local governments did with central funds.

The institutionalization of center-local budget negotiations formalized a peculiar public sector corporatism in which governments, not labor market actors, received representation and a say in national policy making. Normatively, this increased the center's capacity to translate its tight fiscal policy into local budgetary restraint, because it recast Denmark's widely cherished local autonomy as operational rather than fiscal autonomy. But norms aside, why did local politicians refrain from spending more? Governmental corporatism worked because it changed the incentives for local politicians in ways that motivated them to control producer demands in their own locality, and thus do the center's dirty work in restraining welfare state consumption spending. While local politicians naturally like to buy votes through expanded services, they also must be wary of raising local taxes. The center’s policies increased the share of local funding for local services, deliberately heightening local politicians' sensitivity about tax increases. Simultaneously, however, block grants allowed local politicians more freedom to shift money to the most pressing -- i.e. vote generating -- local problems. Local politicians could use block grants to bridge the gap between service demands and tax resistance.

Like Ulysses, local politicians used ALG's negotiations with the center to tie themselves to the mast when facing local Sirens: the negotiations enabled them to hear professional groups' and citizens' demands for more resources and services, and yet effectively claim that they could not meet all their demands. The annual budget deal with the central government precluded a general free for all at the public trough, while the concentration of central transfers into block grants gave them some flexibility to help different groups at the margin.

On the other hand, as the 1990s expansion of transfer payments shows, once budget surpluses reemerged, central politicians who were not tied to the mast could not resist the temptation to buy votes and popularity. A positive political pay off for improving welfare policies is a given in Danish politics, and national politicians responded in very visible ways, introducing universal child allowances and a 30 percent increase in student grants, for example. In contrast, national politicians could never fully reap the political benefits of any improved locally managed welfare services.

The somewhat unusual public sector governmental corporatism and public sector structures described above are not unique to Denmark, because all Scandinavian localities levy taxes and spend the bulk of welfare expenditures. But while all the Scandinavian countries have developed variations on governmental corporatism as a crucial instrument for public budget control, it has developed most fully in Denmark. For example, Sweden had this kind of institutionalized bargaining up until 1976, but the central government ended it when localities did not hold their end of the bargain (Lotz, 1990). Ironically, Sweden is now shifting towards structures that more closely resemble Denmark’s.

Administrative Corporatism: something different?

The second decisive institutional feature of the Danish welfare state which central politicians could consciously use to produce fiscal restraint is administrative corporatism. Corporatism clearly is a system for sustaining and stabilizing collective action. Corporatism can help subordinate narrow interests to broader social goals, but it also can entrench the interests of specific organized groups. As above, governmental corporatism produced some capacity to contain public expenditures while maintaining the welfare state. But even if local politicians had incentives to curb welfare service budget growth, why did strongly organized and often highly professionalized service producers go along?

Here the second peculiarity of Danish corporatism emerges. In all welfare states professional street-level bureaucrats exercise considerable influence on policy implementation (Lipsky, 1980; Wilson, 1989: 148-58, 168-71). However, Danish professionals are integrated into the management and governance structures of the welfare state to a much greater degree than elsewhere. Professional unions of course negotiate pay and work conditions with the ALG and the central state and lobby political parties and local officials on issues of interest. But professionals (and semiprofessionals) are also usually also given a privileged position in the management of welfare services.

Thus, day care centers are run by trained day care teachers themselves, and typically many local bureaucrats dealing with the day care centers have started as day care teachers. In primary and secondary schools teachers have traditionally held a strong position on the boards (Christensen, 1995). Principals and even the local administrative chief of the local school department are also trained at teachers' colleges, and they have all taught before getting their present job. And Denmark is perhaps the only country in the Western world where nurses (and doctors) take part in all levels of hospital management. Organizations representing professionals are thus simultaneously unions and professional associations. Due to the lack of powerful centralized labor organizations, doctors negotiate as doctors, nurses as nurses, etc. on a number of issues broadly affecting member interests. All are conscious not only of working conditions and pay, but also institutional privileges and professional turf.

What, aside from professional norms, has constrained Denmark's professional groups and local government from abusing their position through budget maximization and other forms of opportunism with guile? One answer might be: nothing has. Certainly in the health and education sectors effective professional organizations defended work norms and institutional prerogatives against changes in the 1980s. Moreover, as Christiansen (1995) shows, cutting budgets and shifting resources was virtually impossible to do short of freezing budgets and letting inflation take its toll. But the lack of control over spending is only partly real.

The dual functions of professional producer organizations means they necessarily pursue more than one goal at the same time. Short term budget gains - preferably in the form of improved working conditions and pay increases - are one of these goals. But long term institutional privileges and management prerogatives are the prerequisites for the professionals' capacity to protect their turf and to negotiate budget increases also in the future. Administrative corporatism, like all corporatism, institutionalizes 'iterated games.' This extends the time horizon of those enjoying a privileged position within these administrative networks. However, the fact that the organizations pursue more and potentially conflicting goals can be turned into a resource for policy makers who want to control spending.

Actors will participate in corporatist administrative structures not only when the institution constitutes an immediate positive-sum game, but also when the long term net gains are maximized by staying within the institution even if means immediate economic sacrifices. Furthermore, rational actors will remain in corporatist forum even when they incur losses, if they there by can avoid even worse outcomes. This is particularly apparent in negotiations over pay and conditions. Danish administrative corporatism is not organized around central labor market organizations worried about macro-economic outcomes, but rather organized around professional organizations which have a strong presence at all levels of government. These organizations are as much concerned with professional prerogatives as with pay packets and working conditions narrowly defined. Meanwhile ALG is bargaining as a representative of 'shareholders' (i.e. taxpayers and voters) and the central government, because management per se has been colonized by the professional organizations via the para-public committees that actually control policy implementation.

Because professional organizations control policy execution, it focused the local and central states' attention on gross budgets and general policy mandates rather than on detailed budgetary control and regulation of service delivery. For their part, professional organizations acceded to stagnant budgets to preserve their work place autonomy, much as local politicians acceded to stagnant block budgets so long as they had discretion to use those blocks in vote maximizing ways. In the 1980s, for example, nurses traded away budget and pay increases for more control over hospital administration and over the allocation of work inside hospitals, effectively expanding trained nurses' professional turf.

Stuck:

A stuck argument starts from the same observations as a pluck argument, but adds the following spin. Unlike industrial policy, no one’s behavior deviated from long established practices except for the shift from reimbursement to block budgeting. It’s hard to see how any other dynamic could have merged from this situation.

The Danish combination of professional group corporatism and a highly fragmented parliament made changing welfare institutions or the entire welfare system quite difficult. In general, any central government run by minority coalitions would find it hard to articulate a coherent vision for change, despite the adverse budgetary consequences of producer control. Furthermore the center found it even harder to control itself when budget constraints eased in the mid-1980s. While the corporatist institutions discussed above allowed them to impose fiscal restraint on the localities and welfare service producers, nothing prevented the center from improving highly popular social transfer schemes. If the central government had controlled welfare services, the costs of running these programs might have increased too.

In turn this also shows that in the absence of a central vision for change, the well entrenched Danish moral consensus favoring the welfare state strongly conditioned efforts at change (Goul Anderson 1997). Change to the welfare state had to be sold as welfare preserving cost containment, not as a gutting and rebuilding. Well-positioned welfare service producers, who stood to lose from the few reforms which were introduced, could always challenge the wisdom of the reforms by arguing that they attacked not their privileges but the quality and scope of the welfare state. This reveals limits to change in a hegemonic welfare state. It is hard to generate or maintain popular dissatisfaction with a welfare state only by raising efficiency issues. But expanding the attack to include policy effectiveness risks undermining the legitimacy of the welfare state as well, and thus causes supporters of the welfare state to rally in its defense. In Denmark at least, entrenched producer interests would probably be capable of encouraging large segments of the population to stand up in their defense, and inflicting electoral losses on politicians. But on the other hand, the center could use identification of privileges running counter to the moral base of the welfare state to threaten producers’ core strengths. In this dynamic producers could only justify their continued control if they delivered reasonable quality services at a reasonable price. They could not allow service quality to deteriorate along with funding because this would make an attack on their professional privileges even more tempting.

So whether from calculation or belief, few political actors risked commonplace, neo-liberal style assaults on welfare. Such tactics would gain neither immediate votes nor free up enough resources to buy votes later. Budget restraint and some efficiency gains could only be implemented by working within the various logics of appropriateness that already existed in Denmark. These logics dictated that local and central government challenge professionals’ institutional privileges to get cost containment. Denmark’s entrenched producer groups were well positioned to resist assaults on professional autonomy and prerogatives, particularly from minority governments, but it was less risky for them to accommodate budget restraint than to put their privileged position at stake. Thus while on their own terms the threatened reforms had negligible impact, the politicization of management structures increased policy makers' capacity to curb budget increases. As detailed budget control probably impedes efficiency anyway, this structure probably is superior in terms of producing acceptable outputs and cost containment from public sector organizations (Saltman and von Otter, 1992; Schwartz, 1994). The logic of appropriateness in government-professional group relations let the governments set the budget and producer groups spend it, subject of course to their own internal norms about service delivery. This may then create political problems if producer norms involve abuse of the consumers of services.

Evaluation

While luck -- falling interest rates -- eased fiscal problems in Denmark, it can be ruled out as the determining influence on fiscal balance. Instead, deliberate efforts to restrain some kinds of spending secured fiscal stability by the 1990s. The question is, how deliberate were these efforts? Since the strategies and outcomes all followed quite closely with routine Danish logics of appropriateness about the provision of welfare services, and since the critical actors pursing fiscal balance -- central politicians -- themselves contributed to fiscal imbalance by expanding transfer payments, it seems reasonable that pluck was not the overwhelming cause for success here. Let’s score this 20 percent luck, 30 percent pluck and 50 percent stuck.
 
 

Table 6: WHAT SHARE OF OUTCOME COME FROM LUCK OR PLUCK?
	Exports	Employment	Fiscal Balance
Luck	0 %	0 %	20%
Pluck	33%	66%	30%
Stuck	66 %	33%	50%

Luck, pluck, or stuck?

We’ve (admittedly arbitrarily) scored the three issue areas we’ve examined in Table 6.

On balance this suggests that while deliberate policy rather than circumstance helped ameliorate dynamics which were undermining the Danish welfare state in the late 1970s and early 1980s, actors operated within a very small margin of freedom in choosing their actions. This is somewhat comforting because it suggests that at the margin, correct (!) policies can make a difference, assuming of course a correct policy can be found and implemented. Because policy emerges out of existing institutions which are already mal-adapted to the external environment, the probability that actors will generate "correct" policy is bound to be low, although some authors (e.g. Genschel 1997) have considerable faith in muddling and incremental adaptation as solutions.

Narrowly, the economic crisis in the 1980s in Denmark was not a sufficient cause of crisis for the welfare state, and it is likely that the same will be true of the Swedish state as it confronts a period of economic hardship and management problems similar to those Denmark faced in the 1980s. The Danish evidence suggests not only that welfare state preserving responses exist, but that precisely the kinds of things Lindbeck et al. (1994) see as pathologies are also sources for political responses that can remedy basic economic problems with the welfare state

More broadly, this analysis of the Danish miracle suggests that it, like that in Australia and the Netherlands, does not present fully transferable policy lessons. In all three countries policy responses to the kinds of problems analyzed could only partially be characterized as fully strategic responses to each country’s maladaptation to the international economic environment. Even without invoking the language of path dependence it is clear that the inability of actors to generate fully strategic responses to institutional maladaptation to their environment makes these miracles in at least one sense: human agency probably did not have much to do with them.

Sources:

Albæk, Erik, et. al., eds., Negotiated Adaptation, Aarhus: mss.

Alchian, Armen (1950), "Uncertainty, Evolution, and Economic Theory," Journal of Political Economy 58, June, pp. 211-221.

Annerstedt, Jan (1989), ""Industrial Modernists’ and "National Programmers’ for High Technology," pp. 125-158 in Jørgen Lingaard Pedersen, ed., Technology Policy in Denmark, Copenhagen: Fredriksberg Bogtrykkeri.

Christensen, Jørgen Gronnegaard (1995), "Education Policy," in Albæk, Erik, et. al., eds., Negotiated Adaptation, Aarhus: mss.

Christiansen, Peter Munk (1989), "The Danish Policy on Technology," pp. 31-60 in Jørgen Lingaard Pedersen, ed., Technology Policy in Denmark, Copenhagen: Fredriksberg Bogtrykkeri.

Christiansen, Peter Munk, and Nils Christian Sidenius (1988), "Forsknings- og Teknologipolitik i Danmark," Politica 20:3, pp. 246-268

Christiansen, Peter Munk (1995), "Public Expenditures: Is the Welfare State Manageable?" in Erik Albaek, et al., eds., Negotiated Adaptation, Aarhus: mss.

Due, Jesper et al. (1994), The Survival of the Danish Model, Copenhagen: DJØF Publishing.

Eliason, Leslie (1992), "Reading the Cards on the Table," Scandinavian Studies 64:4, pp. 544-581.

Genschel, Philip (1997), "Dynamics of Inertia: Institutional Persistence and Change in Health Care and Telecommunications," Governance 10:1, January, pp. 43-66.

Golden, Miriam and Michael Wallerstein (1997), "Fragmentation of the Bargaining Society," Comparative Political Studies 30:6, December 1997, pp. 699-732.

Goul Anderson, J. (1994), ASamfundsøkonomi, interesser, og politisk adfærd@, pp. 15-136 in E. Petersen, et. al., Livskvalitet og holninger I det variable nichesamfund, Aarhus: Psychologisk Institut.

Horwitz, Eva (1984), "Export Performance of the Nordic Economies 1965-1982: A Constant Market Shares Analysis," pp. 259-299 in Economic Growth in a Nordic Perspective, Copenhagen, DØR.

Lindbeck, Assar et al. (1994), Turning Sweden Around Cambridge: MIT Press (translation of Nya Villkor för Ekonomi och Politik, Ekonomikommissions förslag SOU 1993.)

Lipsky, Michael (1980), Street-Level Bureaucracy: Dilemmas of the Individual in Public Services, New York: Russell Sage Foundation.

Lotz, Jørgen (1990), "Controlling Local Government Expenditures: The Experience of Five European Countries," pp. 249-262 in Public Finance with Several Levels of Government, ed. by Rémy Prod’homme, Brussels: Proceedings of the 46th Congress of the International Institute of Public Finance.

March, James and Johan Olsen (1989), Rediscovering Institutions, New York: Free Press.

Martin, Andrew (1996), "What does Globalization have to do with the Erosion of Welfare States," Harvard Univeristy, unpublished mss.

Nannestad, Peter (1991), Danish Design or British Disease: Danish Economic Crisis Policy 1974-79 in Comparative Perspective, Aarhus University Press.

Nielsen, Klaus, and Ove K. Pedersen (1988), "The Negotiated Economy: Ideal and History," Scandinavian Political Studies 11:2, pp. 79-101.

Nielsen, Klaus (1991), "Learning to Manage the Supply Side," in Bob Jessop et al., the Politics of Flexibility: Restructuring State and Industry in Britain, Germany, and Scandinavia, Aldershot, England: E. Elgar.

Nørby Johansen, Lars (1986), "Denmark" in Peter Flora, Growth to Limits. Berlin: Walter de Gruyter.

OECD (1984), Economic Survey: Denmark, 1984, Paris: OECD. Also: 1994, 1996, 1997, 1999.

OECD (1992), Revenue Statistics of Member Countries, Paris: OECD.

Oxelheim, Lars (1984), "The Largest Nordic Manufacturing Companies," pp. 184-201 in Economic Growth in a Nordic Perspective, Copenhagen: DØR.

Pontusson, Jonas, and Peter Swenson (1996), "Labor Markets, Production Strategies and Wage Bargaining Institutions," Comparative Political Studies 29, pp. 223-250

Premfors, Rune (1996), "Reshaping the State: Swedish Experiences in Comparative Perspective," Lund: xerox.

Rothstein, Bo (1992), Den korporativa staten: Interesseorganisationer og statsforvaltning I svensk politik, Stockholm: Norstedts.

Saltman, Richard and Casten von Otter (1992), Planned Markets and Public Competition: Strategic Reform in Northern European Health Systems. Buckingham; Philadelphia: Open University Press.

Scharpf, Fritz (1991 [1987]), Crisis and Choice in European Social Democracy, Ithaca: Cornell University Press.

Scharpf, Fritz (1997), Games Real Actors Play, Boulder CO: Westview Press.

Schou, Bent "Udgiftsstyring eller Fornyelse?" in Karl-Henrik Bentzon, Fra Vækst til Omstilling: Modernisering af den Offentlige Sektor, Copenhagen: Frederiksberg Bogtrykkeri, 1988.

Schwartz, Herman (1994), "Public Choice Theory and Public Choices: Bureaucrats and State Reorganization in Australia, Denmark, New Zealand, and Sweden," Administration & Society 26:1, pp. 48-77.

Schwartz, Herman (2000), "Internationalization and Two Liberal Welfare States: Australia and New Zealand," in Fritz Scharpf and Vivien Schmidt, eds., National Employment and Social Policy in a Global Economy, volume II, Oxford: Oxford University Press, forthcoming.

Thornthwaite L. and P. Sheldon, (1996), "The MTIA, Bargaining Structures and the Accord," Journal of Industrial Relations 38:2, pp. 171-195.

Visser, Jelle and Anton Hemerijck (1997), The Dutch Miracle, Amsterdam, University of Amsterdam Press.

Wilson, James Q. (1989), Bureaucracy: What Government Agencies Do and Why They Do It, New York: Basic Books.

return to my home page

You can mail me comments.

Last modified: October 26, 1999