World Politics

Introduction

Throughout Europe a revolution is transforming the welfare state. Inspired by EU leadership, nations from Sweden to Spain are seeking to make the welfare state more "active" in order to rescue those on the dole from the social isolation that comes from years and even generations of unemployment and poverty. Employers, viewed as frontline warriors in this battle against unemployment, are pressed into service by governments eager to bring new economic relevance to the types of jobs and training programs that have often had a make-work quality in the past.

Many view this shift in social provision as simultaneously representing radical change within welfare regimes and an impressive amount of formal conceptual convergence among regime types. Radical transformation has occurred to the extent that new reforms are changing the essential goals of social provision: the aim of "employability" for all individuals may be replacing the older goal of full employment (with the marginally employed relegated to safety-net programs).¹ At least conceptually, these reforms also seem to represent an impressive amount of policy convergence in that they borrow from various regimes, thus bringing countries to create new policy forms that break with their historical approaches to social provision.² [End Page 39]

Given this conceptual congruence, the most significant differences among the various national welfare reform efforts may well be at the point of implementation; this variation poses intriguing puzzles for scholars. For example, while the British and Danish social plans are similar on paper, they have inspired very different take-up rates among employers. Danish employers have had virtually no prior involvement in implementing social programs, yet they have participated in far greater numbers in the programmatic delivery of services than their British counterparts, who, by contrast, have traditionally played a substantial role in providing private employment-based social benefits. While Scandinavian employers have historically supported the welfare state more than managers in liberal countries, Danish firms' current involvement in programmatic service delivery nonetheless represents a fundamental break with the past.³

This article's central empirical objective is to evaluate why firms participate in the new active programs and, thereby, to account for why Danish employers have participated at higher rates than their British counterparts. The empirical analysis draws from interviews with 107 randomly selected British and Danish corporations, from which we can isolate the causal variables that bring firms both to participate in active labor-market and social programs and to explore business attitudes toward the newest generation of welfare reform.

More broadly, this article uses the case of active labor-market and social policy to explain the perpetuation of welfare regime trajectories at moments when social policy forms appear to be both digressing from their regime traditions and converging across countries. Because the most significant divergence among plans seems to appear at the point of implementation, this article examines whether the processes of operationalizing and implementing the concept of active social reforms contribute to the perpetuation of regime trajectories. My hypothesis is that even at moments of policy convergence, regime distinctions should persist in the subtleties of programmatic design and in employers' expectations about state administrative capacities that result from cumulative experiences with past policies. These legacies, embedded in regimes, should have important feedback effects on employers' participation in the programs and influence both firms' reasons for joining the effort and the comparative take-up rates. [End Page 40]

The programmatic designs of Danish and British active labor-market plans have diverged in a manner predicted by differences in their regimes. In keeping with social democratic traditions, the Danish plans cover a more universal population, link the programs to broader training efforts, and arguably contribute to a major collective goal—an encompassing labor market. Befitting a liberal welfare state, the British plans target a more narrow population, do little to expand skills, and continue in the punitive track of former Prime Minister Thatcher's social interventions.⁴

These programmatic differences have had profound feedback effects on firms' willingness to participate in these programs; indeed, Danish firms participated more than their British counterparts precisely because of the social democratic tint to welfare reform. While 68 percent of Danish employers participated in the active labor-market and social programs, only 40 percent of British firms did so (many of which ultimately reneged on their promise to create jobs for the unemployed). The interviews reveal that Danish employers viewed program constituents less negatively, anticipated greater material benefit from participating, and evaluated the programs' administration more highly than did British firms. While Danish firms participated in order to gain access to a more highly skilled labor supply, British companies joined the effort largely for political reasons (to appease the new government) or to gain access to cheap labor.

Regression analyses, by demonstrating that different causal variables significantly influenced the likelihood of participation in the two countries, further support the theory that British and Danish companies participated for divergent reasons. The British firms most likely to participate were large, had a high percentage of sales to the public sector (in short, they were vulnerable to political pressures from the government), and had many unskilled blue-collar workers (thus, they could benefit from an influx of cheap labor). By contrast, the Danish firms most likely to participate were smaller and had blue-collar workers at all skills levels (thus, these firms were well suited to benefit from the training offered by the programs).

This article aims to contribute to our understanding of how differences in implementation contribute to regime perpetuation and cross-national diversity, an understudied area in welfare state theory that is of growing salience in this era of private sector solutions in which employers' [End Page 41] and in implementation is critical to policy success. Regime theory has provided important analytic constructs with which to ponder how policy legacies shape new legislative initiatives and macrotrends in social spending.⁵ Scholars have shown how regime-driven policy legacies alter the programmatic realization of conceptually similar social reforms across countries and influence the politics of retrenchment.⁶ These analytic tools, however, have not been utilized to explain cross-national differences in the implementation of social programs. My aim is to demonstrate that the policy feedback effects that shape national trajectories in welfare state spending also influence employers' implementation of social programs.

In an age of neoliberalism, questions about regime perpetuation have profound implications for the fate of social democracy. The Danish experiment, at heart, is an effort to locate a new equilibrium between growth and equity appropriate in the postindustrial economy. The falling productivity growth rates that have accompanied deindustrialization have created a "trilemma" for nations seeking simultaneously to achieve wage equity, to realize full employment, and to maintain their fiscal solvency.⁷ In asking firms to reemploy the socially excluded, Danish policy entrepreneurs hope to solve the trilemma of the postindustrial economy and to fix the decommodification traps of the Scandinavian welfare state. These bureaucrats would match unemployed persons' skills levels to employers' job requirements and would subsidize (by up to 50 percent) firms' employment of disabled workers to perform unproductive tasks within the companies. These "win-win" arrangements would expand employment, reduce government social expenditures (by getting people off the welfare rolls), allow companies to fill (with low-wage workers) unproductive positions that were sacrificed in the push for global competition, and maintain wage equality (because the new low-wage positions are subsidized by the state).

The ultimate success of the Scandinavian model depends on the state's ability to build pro-social-policy coalitions of broad majorities; in effect, they need to construct a new labor-management accord. This article provides an intimate picture of this process of consensus building in the Scandinavian welfare state. In the case of active labor-market [End Page 42] policy (ALMP), Danish employers have expectations about the legitimate scope of government that their British peers do not share. Their social democratic government (in substantial agreement with the other parties) has demonstrated powerful leadership in shaping employers' interests in the social policy area; these efforts have been successful in great part due to the pragmatism and efficacy of the Danish welfare state. These embedded policy legacies may keep the Scandinavian model alive, creating continuities in the implementation processes and structuring patterns of business/government relations, even at moments when neoliberal sentiments are changing the policy tools and location of service delivery.

Employers and Active Labor-Market Policy

New risks that have challenged contemporary welfare states have initiated this process of social policy transformation. Globalization and structural unemployment may have created pressures for welfare states to retrench or at least to cap social spending increases.⁸ The falling productivity growth rates that have accompanied deindustrialization have created a trilemma for nations seeking simultaneously to achieve wage equity, to realize full employment, and to maintain their fiscal solvency. Different solutions to the welfare trilemma have different implications for the future of inequality, growth, and social justice.⁹ Demographic and family-structure changes have shrunk the worker/dependent ratio, added to concerns about budget deficits, altered the caregiving structure, and increased the need for working-age individuals to contribute to the labor force. The existing structures of the welfare state put into place during the Golden Age may be ill equipped to handle the challenges of the postindustrial era.¹⁰

These emergent risks have inspired new conceptions of the welfare state, most prominently, active labor-market and social policies designed to reintegrate the long-term unemployed back into the core economy, to relieve budgetary pressures on social spending, and to end social exclusion. With rules about the duration of benefits ("sticks") and [End Page 43] access to training or to subsidized jobs ("carrots"), these policies seek to redirect attention toward the socially excluded and toward skills development.¹¹ While Fordist-era social plans met the initial training needs of the core labor force and channeled other individuals into safety-net programs, the new policies emphasize the importance of renewed skills development in meeting the challenges of rapidly changing technology in the postindustrial economy.¹² ALMP programs have been described as "realignments" rather than as "path dependencies" in scaling back social rights as well as in shifting the goals of social intervention from income maintenance to employment promotion, from full employment to individual employability, and from a state guarantee of income to a commitment to invest in social capital.¹³ These policies typically involve employers more so than did prior interventions, both because big, universal state programs are more difficult to fund and because private sector training and subsidized jobs are considered more likely to lead to permanent employment than are public options.¹⁴

Since the November 1997 Luxembourg Employment Summit, the European Union has pushed this active model, which has resulted in substantial similarity among national plans.¹⁵ Yet cross-national variation becomes apparent at the point of implementation, which suggests that this somewhat-neglected aspect of welfare state studies is increasingly responsible for such differences.¹⁶ For instance, Denmark outspends Britain on active labor-market policies and more successfully obtains its employers' participation in the programs.¹⁷ These comparative [End Page 44] spending figures may not surprise us, given Denmark's historic leadership in public spending on all labor-market measures. Yet given their virtual lack of involvement in previous social programs, it seems paradoxical that Danish employers would suddenly take this lead in implementing active labor-market measures. Granted, the employer-provided benefits still represent only a small portion of the overall social reforms and Danish employers historically have politically supported social policy to a greater extent than British firms. Nonetheless, the use of Danish employers to implement active social programs—a core component of active social policy—diverges profoundly from that nation's tradition of governmental solutions to social problems.¹⁸ Denmark's initiative in the move away from decommodification is also surprising and has caused critics to fear an attack on the rights of citizenship; nonetheless, the Danish lead in social intervention remains unchallenged, even in this pan-European move to a liberal reform.¹⁹

Several possibilities might explain these puzzling developments. First, Danish firms may have economic structural characteristics that make their participation in the social programs more tolerable (or even beneficial) than it is for their British counterparts. Although managers generally resist regulations, companies still may support policies that carry special benefit for them, such as policies that either improve their market position vis-à-vis their competitors or appease their workers.²⁰ In an analysis of employers' preferences, we might anticipate that the same structural characteristics would shape firms' preferences in the two countries, and in accounting for the cross-national divergence, we might anticipate that Denmark has a greater number of firms with the salient structural characteristics than does Britain.

Scholarly literature identifies eight economic characteristics that may influence a company's preferences for social policies (see Table 1). A firm's size might be directly proportional to its participation in social programs because larger firms are more likely to have preexisting training programs or to have economies of scale in political action.²¹ Firms with lower profits (as a percentage of total sales) might be more likely to participate in order to obtain access to cheap labor (although it is [End Page 45]

Click for larger view

Table 1 Hypotheses

also possible that more profitable firms could more easily devote resources to social initiatives). Firms with higher percentages of exports might participate in order to gain a source of cheap labor. The presence of unions within a firm might compel its management to take more socially responsible positions, although unions might also view these programs as threatening to more standard employment.²² Firms with a greater percentage of public sector sales may be more likely to support government policies, either to curry favor or to control their environment.²³ Companies with a higher percentage of unskilled blue-collar [End Page 46] workers (or their unskilled, service sector worker equivalents) might participate more in order to gain cheap labor. Firms with a higher percentage of blue-collar workers at all skills levels—that is, not white-collar workers—might participate in order to secure higher skills for their workforce.²⁴ The terms "unskilled" and "skilled" workers designate both manufacturing and service sector equivalents. Firms with lower average wages might participate to gain cheap labor.

Theories of corporate preference that rely upon individual firms' characteristics suffer in that they fail to take into account institutional differences across settings. In particular, they neglect the differences in public policy regimes and their divergent impacts on employers' incentives to engage in the new social reforms.

Thus, a second explanation for the cross-national variations in employer take-up rates of the new programs (and my core theoretical argument) is that the Danish and British policies create different incentives for employers to participate, or have different "feedback effects." This argument draws from two somewhat separate theoretical traditions; taken together, they imply that the essential qualities of regimes embedded in programmatic design have feedback effects on employers' willingness to help implement social programs.

Comparative welfare regime theorists study how policy legacies shape new social reforms. Each of the three (or four) welfare regime types embraces a characteristic set of policy tools, beneficiaries, ideologies, and links between social protections and production.²⁵ Even when nations are confronted with common exogenous pressures (such as globalization), their particular regime's policy legacies influence new legislative reforms and perpetuate distinctive welfare state trajectories.²⁶ Existing policies mold the construction of constituencies for new welfare state programs and create "policy ratchet" or "lock-in" effects that define solutions to new problems. Differences in the links between social protections and economic production embedded in the regimes have bearing on the programs' benefits to the social partners. Regimes have ideological legacies or interpretive effects: partly related to long-term patterns of party control, citizens over time come to expect certain things from their welfare states and from government.²⁷ Thus, one finds mutations [End Page 47] in the programmatic realization of the ALMP concept that vary in accordance with these regime distinctions and that mirror the values, assumptions, deep logics, patterns of discourse, and institutional supports of each country's welfare regime.²⁸ The new first- and second-order changes do not necessarily entail a rejection of the fundamental goals and values at the heart of the welfare state.²⁹

The study of policy feedback effects on political behavior (found largely in the American politics subfield) suggests that social policy characteristics influence private sector actors' engagement with the state; policy characteristics have an impact on group formation because some societal actors are advantaged by the policies more than others.³⁰ Programmatic characteristics, in addition to having resources effects (in distributing goods), also have interpretive effects (in conveying meaning about the policies to the general public). They influence the mass public's perceptions of viable policy alternatives. Consequently, prior policy choices tend to become locked in, both because changing the status quo has learning and coordination costs and because policies foster their own sets of norms and values.³¹ Furthermore, these characteristics have ripple effects in shaping constituents' broader views of government and their willingness to participate in democratic processes.³² Policy feedbacks also shape employers' preferences and, as this article seeks to demonstrate, managers' capacities to participate in implementation.³³

Taking employers' involvement in implementation as a subset of political participation, I hypothesize that Danish and British firms' engagement [End Page 48] will be conditioned by policy effects from the current reforms and by legacies from past experiences with state programs. Regime-driven programmatic distinctions analogous to those perpetuating national regime trajectories in the legislative arena might influence the implementation processes through "policy feedback effects."³⁴ To begin with, regimes construct target constituencies differently: social democratic programs are usually universally available, while liberal regimes are more likely to use means-tested eligibility standards. The social construction of target populations influences both the choice of policy tools and the arguments used to justify the policies; thus, policies that benefit "good" and "powerful" populations are more likely to appeal to the collective good than are those that benefit the "less deserving."³⁵ While active social policies generally target the unemployed, we might expect Danish programs to be more universally available than British ones; consequently, Danish employers should view program recipients more favorably than should their British counterparts.

In addition, at the programmatic level, links between social protection and economic production may influence employers' calculations of the material costs and benefits of participation. Because social protections and economic production are coupled more tightly in coordinated market economies and Scandinavian welfare states than in liberal systems, Danish employers should anticipate greater material benefits from participation than should their British counterparts. Danish firms should participate in order to upgrade workers' skills while British firms should participate in order to obtain either cheap labor or public relations recognition. The preexisting literacy differential in the Danish and British workforces (22 percent of British citizens are in the bottom quintal in skills, yet only 10 percent of Danish workers fall into this category) enhance these differences in expectations about the programs' capacity to develop skills.³⁶

Finally, regime differences in ideology may affect how employers assess the welfare state's efficacy and influence their willingness to participate. Over time, employers accept state social policies that they view as inevitable (especially under long-term, left party domination);³⁷ [End Page 49] however, experience with the government agencies that administer the policies also shapes participants' attitudes toward the state, because the agencies represent a microcosm of government.³⁸ These cumulative experiences form distinctive regime legacies and are observable in cross-national distinctions in beliefs about governmental capacities and legitimacy.

We can expect administrative legacies to enhance Danish firms' greater willingness (than their British peers) to participate in government programs because universal welfare states (such as those found in Scandinavia, where citizens feel that they are treated equally) generate higher levels of trust in government than do liberal regimes.³⁹ In addition, Denmark has a tradition of strong local government dating back to at least the nineteenth century; Danish municipalities have long exerted enormous control over social programs and thus have provided citizens with a greater sense of self-determination.⁴⁰ By contrast, the British party system theoretically allows the ruling party complete control over legislative initiatives; consequently, regime change often entails brave new initiatives that have little cross-partisan appeal, minimal societal input, and not much local control.⁴¹ Accordingly, while the Danish programs have been implemented with considerable local community input and initiative, the British New Deal has been seen very much as a national brainchild of the Blair administration. One might anticipate that a Danish local government official would feel ownership over an activation project in a manner not shared by her British counterpart. Thus, due to their past and current experiences in administering agencies, Danish firms should regard the administration of the plans more highly than should British firms.

Ideological differences over the scope and efficacy of government may also influence the strategies used by government actors to obtain employer participation. While social democratic regimes value norms of social solidarity, liberal regimes refer to market externalities to legitimate government interventions. Thus, we should expect the social democratic Danish state to make different types of appeals to employers than should the liberal British state. [End Page 50]

The institutional processes of business-government relations offer a third explanation for the differences in the take-up rates by Danish and British firms. Especially under conditions of bounded rationality, the institutional settings in which companies receive information and deliberate on political matters influence employers' perceptions of their social interests.⁴² It seems plausible that Danish firms' institutional traits may make them more likely to enlist than British firms, or that the institutional context of policy implementation differs across regimes.

These relevant institutional characteristics may pertain to internal company structures for processing public policy; for example, U.S. firms that have a government affairs office have been shown to be more supportive of government intervention than those that do not.⁴³ While most European firms lack government affairs offices, the size of a professional human resources department may be an important determinant of employers' support for social programs.⁴⁴ Alternatively, the relevant institutional characteristics may pertain to the external networks for firms; thus, a company's membership in an employers' association may largely determine its position on social policy. In either case, we might anticipate that the same institutional characteristics would shape firms' preferences in the two countries and that a greater number of Danish firms possess these characteristics.

It is also possible that the benefits of membership in an employers' association differ in the two countries, because corporatist employers' associations found in Denmark offer greater incentives for collective action toward shared social goals than pluralist British organizations. When employers are organized into corporatist groups, they may be more likely to favor social policies; corporatist organizations also increase the state's capacity to build support for and to secure compliance with its initiatives.⁴⁵ Tripartite negotiations through the system of Danish corporatism have created a legacy of social partnership in overseeing labor-market programs (for insured workers); therefore, although firms have had no prior involvement with the social realm (for the uninsured), the corporatist channels might be replicated to bring them into [End Page 51]

Click for larger view

Table 2 Programmatic Details of the Danish and British Plans

this new area.⁴⁶ Thus, membership in a business association might be a significant determinant of preference in corporatist Denmark but not in pluralist Britain.

Programmatic Variations in Danish and British Social Policies

The Danish and British plans appear to be remarkably similar in the time limits they set on the receipt of passive benefits (six months for youth and one year for adults), in the recipients' compulsory participation, and in their outreach to firms.⁴⁷ With ALMP, Denmark greatly reduced access to unemployment insurance, largely by imposing time limits and obligations to work. In his "fuzzy membership scores," Kvist moved Denmark from .98 in the accessibility of unemployment benefits in 1990 to a .53 ranking in 1998, similar to Britain's score of .52.⁴⁸ Both countries asked firms to offer subsidized jobs and training. While training historically had been provided by the state in Denmark, active policies encouraged firm-based training; in Britain, training choices had been considered the responsibility of the individual since the rejection of more activist state measures in the 1970s, but the Blair reforms reassigned the government some responsibility for ensuring training.⁴⁹ State subsidies paid to employers for hiring the unemployed have been similar—about 50 percent of wages in Denmark and 54 percent in Britain.⁵⁰

Beyond these similarities, the Danish and British active labor-market and social plans diverge as anticipated (see Table 2). First, in keeping with social democratic values, the Danish plans serve a more universal constituency by offering jobs to the unemployed disabled for whom no rehabilitation is possible. With these protected jobs, the state has effectively developed subsidized minimum-wage employment that helps [End Page 53] firms hire for unproductive tasks within the company and retain disabled workers otherwise headed for early retirement.⁵¹ Alternatively, reflecting liberal regime assumptions, the British plans have served a more targeted population—long-term unemployed adults and youth—and have created only a small pilot program to address the needs of the disabled.

Second, the plans' benefits differ, especially in how much they contribute to skills development. Denmark spends more both on ALMP (1.58 percent of its 2000 GDP, compared with .37 percent of Britain's 2000 GDP) and on training for unemployed adults (.67 percent of its GDP, compared with Britain's .04 percent).⁵² The Danish plans for the long-term unemployed are closely linked to general state training programs for low-skilled workers; for example, job rotation schemes allow firms to use state subsidies to hire the long-term unemployed while their own employees receive skills training.⁵³ By contrast, British programs, not surprisingly, have significantly less connection to skills development because public training resources have been limited since Thatcher's time and labor-market policy has been conceived of as a residual service for the unemployed.⁵⁴

Third, the two countries, in keeping with their respective regime types, differ in their state ideological appeals to employers and plan administrators. Danish government policy entrepreneurs urged employers to satisfy their own long-term self-interests by working to advance collective economic and social interests; they linked a new economic formula for ending unemployment called the "encompassing labor market" (an economy that can be expanded to bring the socially excluded into its fold) to the collective social goal of ending social exclusion. The government also urged firms to identify unproductive tasks within the production process that people with subsidized protected jobs (available to individuals with reduced skills) could handle, and the state would then pay one-third to one-half of their wages. These new low-wage positions offered companies value by helping them with otherwise [End Page 54] difficult-to-complete tasks, reduced state spending (because the companies paid one-half to two-thirds of the wage), moved the long-term unemployed into the core economy (and thus reduced social exclusion), and maintained wage equality because the wages were subsidized by the state. Thus, the Danish experiment aimed to create more efficient social benefits, to revise social protections to meet the needs of the new economy, and to ameliorate long-term unemployment and social exclusion.⁵⁵ British state officials used a very different ideology—the language of markets—to sell their programs and advocated "demand-led" strategies to tailor the programs to private-market needs. With this liberal welfare state strategy, firms were told that job seekers would be made "job ready," and personnel promised to tailor the skills of the potential workers to the needs of individual sectors.⁵⁶

Danish and British bureaucrats used different mechanisms to mobilize employers to join the effort. The Danish Social Ministry decided not to spend resources on a big media campaign; instead, it reached out to bring firms into the social policy project with two diverse national strategies: expanding corporatist ties with the peak associations to the social arena and mobilizing firms directly with the creation of the "National Network of Firms." Initial efforts to involve the organizations in social policy deviated from the status quo, because government had responsibility over beneficiaries of social assistance. The subsequent outreach to individual firms marked an even greater deviation, because it was seen as undermining traditional corporatist channels. An industry respondent explained, "DA and DI didn't like the firms getting involved with the ministry; they were worried that there would be anarchy. Individual firms could give all sorts of opinions, with one firm saying one thing and another firm saying another. But the associations speak with one voice for all business."⁵⁷

By contrast, the Blair administration did organize an expensive media campaign and held breakfast meetings at 10 Downing Street with CEOs of major firms to advertise the New Deal. It also created the New Deal Task Force to solicit employer input on program design and to encourage private voluntary participation. Although the administration made some overtures to the Confederation of British Industry and other groups, the business mobilization efforts largely worked outside of the formal employers' associations. As one high-ranking official at No. 10 put it, "We tried to work with business, but they just weren't organized." [End Page 55]

Denmark and Britain spend about the same on their public employment services (1.2-1.3 percent of GDP), but the two countries' bureaucrats created different types of administrative innovations to further their social projects.⁵⁸ The Danes merged social and labor-market policy into a new employment ministry with control over both the insured workers (previously the domain of the old labor ministry) and the uninsured workers (previously the domain of the old social ministry). They have also constructed new institutions very much along the lines of labor-market policy for the socially excluded. A recently enacted law now requires each municipality to form a corporatist committee of representatives from business and labor to oversee social policy in the community. Dansk Arbejdgiversforening (DA) is responsible for selecting company participants to sit on these local social coordination committees. In addition, many municipalities have hired "firm consultants" to visit firms and identify potential protected jobs.

The British government reorganized the old job services and sought to raise confidence in the programs by transferring many tasks (especially those related to preparing New Dealers for work) to private intermediaries. A smaller pilot project, the Large Organizational Unit of the Employment Service, was developed to create account managers within the public sphere to handle vacancies within firms.

Policy Feedbacks on Employers' Views of Active Social Programs

Interviews with British and Danish employers confirm that differences in both programmatic design and patterns of implementation have had an impact on employers' willingness to participate. First, as anticipated, employers in the two countries perceived the program constituents quite differently. While 53 percent of the British firms declined to participate due to negative views of the unemployed, only 20 percent of the Danish companies offered this as a reason for nonparticipation (see Table 3).

Many Danish companies viewed the unemployed as a reasonable source of labor even though—as one manager explained—it might "take a bit longer to make sure that these workers can do OK." Negative views of the unemployed propelled some firms to participate, out of their irritation at seeing "people on unemployment sit around and smoke cigarettes and play the guitar." Yet 53 percent of British firms viewed the socially excluded as "unemployable," reporting that "most [End Page 56]

Click for larger view

Table 3 Employers' Responses to the Danish and British Plans

people with good attitudes are not on the New Deal programs" and that "managers don't want these people . . . the hard core that don't want to work."⁵⁹ Although many firms ultimately had good experiences with workers, their initial reservations reflected the social construction of the target group. [End Page 57]

Second, British and Danish firms estimated the material benefits of the plans quite differently, especially in the plans' contributions to skills development. While 53 percent of British firms reported that they did not participate due to higher skills needs, only 35 percent of the Danes offered this as a reason for nonparticipation and, in fact, many Danish employers offered skills development as an important reason for participation. Many Danish companies sent unskilled workers to an "industrial operator" course and participated in job rotation schemes to this end, hiring the subsidized, long-term unemployed while their employees sought training.

British respondents felt, by contrast, that the New Deal programs did little to address the pressing skills gap problem and repeatedly complained about the insufficient investment in education and training. One manager exhorted government to deliver "basic competencies" and another claimed the Blair government was "going through the motions but [could not] address the fundamentals."⁶⁰ Thus, managers shared the academic critique of the New Deal—that the programs aimed to eradicate unemployment without actually enhancing skills. Because the two countries' plans had a perceived differential impact on skills development, British and Danish firms viewed the programs' contribution to the labor supply differently; 31 percent of Danish firms offered labor shortages as a reason for participating, compared with only 22 percent of British firms.⁶¹ Danish firms believed that participating in the programs provided tangible benefits; accordingly, they were more likely to view the subsidies as easing their participation (38 percent) than were the British firms (10 percent). Although the value of the subsidies was very similar in both countries, British firms did not believe that the subsidies compensated for the hassles of participation; Danish firms, by contrast, viewed the support as allowing them to pursue their social responsibility.⁶²

Consistent with the programs' tenuous link to skills, many more British employers (31 percent) than Danish ones (9 percent) attributed their participation to the political benefits of responding to government pressures. One British manager reported that the local employment center acknowledged the project's political salience in urging, "You don't need to take anyone, but would you just sign up?" Another recalled, [End Page 58] "The incentive was largely political. It was something that the government was trying to promote and as we have a number of relations with the government in our business, we thought that it was a sensible thing to do." Yet another firm reported that its close ties to the Ministry of Defense made it feel "obliged to support a new and very key program for the Blair government." One firm signed up for the New Deal as part of its application to a local board to secure permits to build an airport. By contrast, almost no Danish firms listed appeasement of the then-ruling social democratic party as a motive for participation; in fact, even since the right gained power in 2001, there has been little change in the programs.⁶³

Related to this pattern of top-down political pressure, 31 percent of British firms identified their CEO as the driving force behind their firm's involvement in the New Deal, whereas only 9 percent of the Danish firms made a similar claim regarding their involvement in their nation's programs. Blair invited many of the sample's CEOs to Number 10 Downing Street to hear presentations about the New Deal. One respondent recalled, "The first anyone in the company heard about the CEO's interest [in participating in the New Deal] was from reading about it in the press." Another British manager blamed the government for misunderstanding corporate chains of command: "The government doesn't talk to the right people either. They come and talk to the chairman or the CEO, but they also need to talk to the people who will have to make it happen."⁶⁴

By contrast, Danish employers were more likely to attribute their participation to social responsibility. While managers from both countries agreed that firms should assume social responsibility, only about one-fourth of the British firms mentioned it as a motivational factor in participation, compared with about one-half of the Danish managers surveyed. Yet some Danish firms objected to the programs' erosion of state responsibility: "The company pays so much tax, it should be relieved of the obligation to take care of these people." Many British respondents refused to participate because their firms were moving away from the conception of "paternalistic" responsibility for their workers' social lives and from policies that undermine the principle of "the survival of the fittest."⁶⁵

Third, British firms expressed greater ideological skepticism about state capacities than did Danish firms, a sentiment partly shaped by [End Page 59] their experiences with the plans' administration. While 40 percent of Danish firms reported vigorous outreach by the local employment services, only 27 percent of the British firms were approached. Of firms needing to contact their municipality or labor-market agency, 74 percent of the Danes found their municipality or agency to be responsive, compared with only 47 percent of the British firms. British employers also felt that the programs differed little from previous government efforts to reduce long-term unemployment, describing the programs as merely "old wine in new bottles." While 50 percent of the Danish employers believed that the government was asking business to engage in the social arena in a new way, only 13 percent of the British firms felt similarly. One British manager reported that "job centers were slow to react to the company call for vacancies," while another complained of "absolutely no follow-up" in referrals once the program was implemented. Many felt that the job center screened candidates inadequately: "We saw a lot of young people who didn't know why they were there, didn't know what the job was, hadn't had any Gateway, and hadn't received any help in the interviews." Some New Deal pilot projects, including the Large Organization Unit and some private sector intermediaries, earned greater praise.⁶⁶

Danish firms criticized the programs less than did their British counterparts. While some Danish employers felt that the clientele sent by the municipality lacked skills suitable to their needs and described state efforts as "heavy-handed," others found local governments to be "very well-prepared and helpful" or the job exchanges for insured workers (AF) to be great in helping firms "to figure out the bureaucracy." Some managers recalled municipal social workers' Herculean efforts to bring recipients to the workplace. The creation of a company consultant (virksomheders konsulent) position within the local government seemed especially important in easing company involvement with active social policy. In some municipalities, firm consultants held regular sessions at the large companies to evaluate problematic employees' needs and to give advice about state programs available to meet those needs. Danish employers also rated governmental mobilization strategies more highly than did their British counterparts, as many firms cited local social coordination committees and regional networks of firms as crucial to their interest in the active programs.⁶⁷ [End Page 60]

Determinants of Firms' Participation in Active Social Programs

Empirical tests of the causal determinants of employers' participation in the two countries evaluate the policy feedback hypothesis: for the hypothesis to hold true, different company characteristics should determine participation in Denmark and Britain because the two nations' plans offer divergent benefits. Structured, one-to-two-hour interviews with human resources officers or CEOs at fifty-two British firms and at fifty-five Danish ones (out of a randomly selected sample of two hundred firms) provide some of the data.⁶⁸

Two separate but parallel firm-level comparisons (comparing firms within a country but not between countries) made it possible both to isolate significant causal factors driving firm behaviors in each country and to compare significant variables across countries. The cases were chosen to match countries at comparable stages of active labor-market policy formulation and with similar employment conditions, but with different welfare regimes. Both Denmark and Britain developed active social policies early, as their countries' respective initiatives predated the November 1997 Employment Summit in Luxembourg and had been in place long enough to evaluate employers' responses. In both countries, governments controlled by left-center parties made strong commitments to the programs. Despite differences in size, skills levels, and production strategies, the two countries had the highest rates of employment among EU nations. In both countries, manufacturing and services constitute 25 percent and 71 percent of the economy, respectively, although Denmark has slightly more social service workers and Britain has more producer services employees. While we might expect British skills needs to be less than Danish needs, skills development and training have been prominent topics in Britain.⁶⁹

The dependent variable was operationalized as a firm's participation in the active labor-market and social programs identified in two countries' [End Page 61] National Action Plans.⁷⁰ The dependent variable measuring firms' participation in the programs consisted of a scale, moving from a position entailing full participation to a position ideologically opposed to involvement. Tables 4 and 5 report values of the dependent variable.

Click for larger view

Table 4 Firm Participation in Active Social Policies in Denmark

The economic independent variables were operationalized as follows. The size variable identified the firm's total revenues (in 100 million kroner for Denmark or 100 million pounds for Britain), the profits variable measured the annual net profit divided by total sales, and both were taken from Amadeus Company Reports. The exports variable, which measured the percentage of total sales devoted to exports, was derived from firm-provided data. The unions measure was a dummy variable that reported the presence of unions. The variable that estimated public sector sales represented the share of total revenues taken up by sales to the public sector, as reported by the firm. The variables that measured unskilled blue-collar workers and white-collar workers were derived from data provided by the company and identified the percentage of these types of employees in the workforce. Amadeus Company Reports provided the estimate of average wages (in increments of ten thousand). The human resources institutional variable estimated the number of workers in a firm's HR department and the association variable specified membership in a peak or sector employers' association, as reported by the company. An OLS regression method was appropriate for estimating the hypotheses because of the interval-level quality of the dependent variable.⁷¹ [End Page 62] Hours of interviewing indicated that the firms truly were distributed along an ordinal scale with interval qualities.

Click for larger view

Table 5 Firm Participation in Active Social Policies in Britain

At the time of the survey, 68 percent of the Danish firms were engaged in subsidized jobs and training programs for the long-term unemployed, while 40 percent of British firms were involved to some extent in parallel programs. The findings confirm the importance of regime-driven, policy feedback effects on firms' participation in the active social initiatives. The regression analyses demonstrate that the economic and institutional independent variables performed quite differently for the two countries and suggest that, as anticipated, Danish and British firms participated for different reasons. The institutional context of the regimes also contributed to their diverse patterns of engagement, as the association variable had very different effects in the Danish and British studies.

Why did this variation occur in the performance of the variables? First, in the Danish study, there are three significant economic control variables: size, average wages, and white-collarworkers. A decrease in revenue size from 200 to 100 million kroner produces a .0085-point increase in participation (a substantively small effect), while a 10,000 kroner (about $1200) decrease in average wages produces a .074-point increase, and a 10 percent decrease in white-collar workers produces a .15-point increase. In addition, the association institutional variable has a significant impact on the dependent variable: joining an association produces just over a two-and-a-half-point increase in participation (see Table 8). [End Page 63]

Click for larger view

Table 6 Descriptive Statistics (Denmark)

The regression analysis of the British study finds that three economic independent variables—size, public sector sales,and unskilled blue-collar workers—have a statistically significant effect on the dependent variable. A 10 percent increase in public sector sales produces a .4-point increase in participation and a 10 percent increase in unskilled blue-collar workers produces a .15-point increase, but an increase in size from 100 to 200 million pounds in revenue produces only a .0075-point increase. The data also support the importance of institutional factors to companies' engagement. Firms with larger human resources departments were significantly more likely to participate than those with fewer HR personnel: a 10-person increase in HR staff produces a .03-point increase in participation. Oddly, membership in an association was negatively associated with participation.

The regression analyses, thereby, confirm the policy-feedbacks hypotheses for each country. In Denmark, firms with blue-collar workers of all skills levels are significantly more likely to participate in order to obtain both training benefits and access to a new labor pool. That these firms are not currying political favor (but rather are trying to meet labor needs) is confirmed by their less visible status: participating firms are significantly smaller than nonparticipating companies and, although this figure is not significant, these firms actually have fewer sales to the public sector. [End Page 64]

Click for larger view

Table 7 Descriptive Statistics (Britain)

By contrast, British companies with the very least skilled blue-collar workers (and their service sector equivalents) are significantly more likely to participate as expected, in keeping with their desire to gain access to cheap labor or to gain political benefits from the Blair administration. Since many employers reported signing up for political reasons, it makes sense that larger (more visible) firms and those with higher public sector sales would participate the most. That the average wages variable was not statistically significant in the British study also reflects that a sizable proportion of participating firms engaged for political reasons. Large, visible participating firms in the media and defense industries, for example, had fairly highly skilled and well-paid workers, which dampened the performance of the average wages variable.

The findings also support the contrast between corporatist and pluralist groups: association membership brings firms to participate in Denmark but works against participation in Britain, which is somewhat surprising in that one would expect association membership to have no effect. Yet joining a trade association in Britain constituted something of an alternative to developing a large, professional human resources department. Firms with developed HR departments tended to be much better informed about the programs, in part because the alternative—membership in the association—offered so little information about active social policy. [End Page 65]

Click for larger view

Table 8 OLS Regression Analyses of Danish and British Firms

Conclusion

Governments across Europe have converged in embracing active labor-market policies to recast social protections in the postindustrial age. Consequently, country differences in the implementation of active programs may well represent the greatest source of cross-national variation in future social programs and may be a major factor in the relative successes of diverse regimes in adapting social policy to the new economic environment. For example, despite shared conceptions of active social policy and comparable subsidies for the programs, Denmark and Great [End Page 66] Britain have diverged in the programmatic details of their plans and in the rates of implementation by employers. Denmark devised an active social policy to maximize the economic contribution of the socially excluded and to enhance the economic benefits derived by firms, while the British plans largely neglected skills development. In addition, the Danish plan set out to change the rules of the welfare game by redrawing the labor market, deviating from the social democratic tradition of state-provided service in seeking business implementation of the programs, and renegotiating business-government boundaries in the social policy area. Despite the plan's deviation from the status quo, Danish firms have participated in the social programs in record numbers; they have offered training and work trials for the long-term unemployed, created permanent special jobs for individuals with reduced working capacities, and generally responded to the government's cry for help in ending social exclusion.

This article suggests that the comparatively positive response by Danish employers reflects the policy feedback effects on implementation. The plans' programmatic details offered Danish employers greater incentives to participate and the legacies embedded in the implementation process gave them greater confidence in their government's capacities than it did their British counterparts. Danish firms with a blue-collar, low-paid workforce were significantly more likely to participate in order to expand the labor supply and to enhance their flexibility with even the least-skilled workers. In comparison, despite historically providing many job-based benefits, British employers were more wary about the new reforms, although the lowest-skilled firms and those with greater sales to the public sector were significantly more likely to participate in order to gain access to cheap labor and to win public relations kudos, respectively. Accordingly, the Danish state managed to muster considerable coalitional support for rather profound change, while the British government ultimately achieved little movement from the status quo and failed to inspire much enthusiasm among those asked to implement the programs.

These findings have implications for how regimes are perpetuated, even as their historical forms change. At points of radical restructuring, these regimes may stay true to their underlying assumptions about social protection; these assumptions persist in the programmatic design of policies and have an impact on the processes of implementation. The findings demonstrate the limits of ideas as a causal variable: while new conceptions may motivate broad temporal shifts in welfare state provision, at points of convergence, reform ideas may not be particularly [End Page 67] salient to continuing cross-national distinctions. Processes of implementation thus may bring new ideas to resemble older conceptions of social provision and may thwart policy change. Despite significant legislative changes, institutional legacies at the implementation level are distinctively shading the new reform ideas, preserving cross-national comparisons in policy outcomes, and subverting policy convergence. New wine in old bottles may be consumed in much the same way.

This study's findings have policy implications for experiments in welfare state restructuring. Welfare state regimes' capacity to come to terms with social exclusion depends in part on social protection systems' ability to respond to changing economic conditions, to cure earlier welfare traps, and to be implemented successfully. The Danish experiment has attempted to develop a welfare strategy consistent with postindustrial competition and its social costs: long-term unemployment has been linked conceptually to addressing projected labor shortages and enhancing human capital development. Programs seek to save money for the state (by moving beneficiaries off of social assistance) and to serve a real economic need for firms (by assigning subsidized workers with reduced capacities to complete unproductive tasks in the companies and by allowing regular workers to gain additional skills training while the long-term unemployed take their place). Low-wage, low-skills jobs are allocated by ability rather than by more normatively suspect criteria such as race, class, gender, or even prior skill levels. Programs are designed to assist the shift to postindustrial manufacturing without sacrificing employment. Social exclusion has been viewed as a collective problem with far-reaching ramifications and Danish firms have been very much engaged in the dialogue about this collective concern.

Danish implementation processes, characterized by local-level discretion and pragmatism, reveal an understudied aspect of the Scandinavian welfare regime and may be a neglected source of strength in the Danish social system. The social democratic facility for class compromise has been much discussed, especially with regard to employers' role in constructing and expanding the welfare state.⁷² Yet stories about the pragmatism of the social democratic model are only imperfectly integrated with our conceptions of the Scandinavian welfare regime. While regime theory emphasizes universality, centralized state control, and equality, the Danish social tradition accords local bureaucrats considerable [End Page 68] discretion to fashion practical solutions to community needs. Many scholars and policy pundits perceive local variation as a weakness of American social provision, yet this feature has different policy implications under Danish social democracy.

This new brand of welfare state restructuring has its potential drawbacks. Strategies for integrating the unemployed may be at odds with firms' efforts to improve productivity and the working lives of their own employees, and may deflect from core competencies. Questions about these programs' budgetary implications persist and the strategy's scope and ability to effectively address the needs of the truly disadvantaged are unclear. Just as previous social experiments resulted in a preponderance of makeshift jobs in the public sector, the private sector-subsidized jobs being created today could evolve into a system of second-tier employment. Yet if designed, implemented, and administered correctly, the programs may improve firms' productivity while also taking some financial burdens off of the welfare state. In breaking with tradition, Danish policy entrepreneurs paradoxically tried to preserve, not to undermine, the social democratic welfare regime, and it remains to be seen whether the social democratic regime can maintain the balance between growth and equity that was its hallmark during the golden age.