Generating legitimacy for labor market and welfare state reform-the role of policy advice in Germany, the Netherlands, and Sweden.

Eichhorst, Werner ; Wintermann, Ole

1 Introduction (1)

Policy advice can help inform policy makers on societal problems, their causes and potential solutions, thus contributing to appropriate institutional reforms and effective societal problem-solving capacities. The relationship between scientific research and politics, however, is a delicate one, with the effective supply and transfer of policy advice depending on institutional prerequisites in both the science sector and the political system so that policy-relevant information can be generated and provided which can influence the choice and implementation of appropriate policies.

This paper first lays out some theoretical considerations on the potential of policy advice with special reference to the area of labor market and welfare state reform, emphasizing the role of policy advice in the generation of legitimacy or - at least-acceptance of often unpopular decisions on institutional reforms.(2) We then present empirical evidence on the role of policy advice provided by research institutes, expert committees and other think tanks in social and labor market policy reform in three countries: Germany, the Netherlands, and Sweden. All three are developed welfare states with strong consensus requirements stemming from minority or coalition governments and a strong position of social partners. All faced the need for institutional change but reacted in different ways. Part of this variation can be explained by the role of policy advice. We therefore analyze the structure of policy advice and its actual function in recent labor market and welfare state reforms. In our analysis we focus particularly on the relationship between 'independent' expertise, social partner bodies and government. The paper shows to what extent the structure of policy advice in Germany inhibits the realization of its full potential regarding the design and legitimization of effective policies and why this is different in Sweden and the Netherlands.

2 The Potential of Policy Advice in Labor Market and Welfare State Reform

Welfare state and labor market reforms aiming at institutions that are consistent with sustainable economic activity and social policies often imply cutbacks on social policy programs, budget consolidation and increased flexibility of the labor market. In general, these issues are unpopular since they imply distributional effects with shortterm losses to be experienced by powerful societal actors and social groups, whereas positive effects may take time (Pierson 1994). Therefore, welfare state and labor market reforms are risky and difficult in political terms and can only be adopted and implemented with sufficient legitimacy so that immediate opposition and allocation of blame is avoided (Weaver 1986, Pierson 1994). Otherwise, political actors may suffer from loss of political support. Status quo orientation of important segments of the electorate stabilizes existing institutions and forms barriers to reforms, thus contributing to strong 'path dependence' (Pierson 2000).

Science is fundamentally different from politics as it does not deal with acquiring or defending power in electoral campaigns but is autonomous and mainly oriented towards the academic discourse. It focuses on the identification of causal relationships between different factors, with economic and social science research into the labor market and the welfare state mainly exploring the effects of institutions on labor market and social outcomes.

In order to analyze the role of policy advice in welfare state and labor market reforms, it is useful to differentiate between the concepts of 'puzzling' and 'powering' (Heclo 1974, Hemerijck/Schludi 2000). Puzzling points at the process of identifying problems and possible solutions, whereas powering means the struggle for political support needed to safeguard the acceptance of reforms. In principle, policy advice from science can provide valuable input for both the puzzling and the powering phase in policy-making. Regarding puzzling, research can help detect economic or social problems and the main causal factors responsible for them. This, in turn, can help identify potential remedies and effective policy solutions. Scientific policy advice can inform policy-makers about the probable effects of maintaining institutional status quo as opposed to different reform scenarios. Through exploring the preconditions of institutional change, it can also help formulate policy reform strategies. Political advice, however, can be most effective if it is based on a sufficiently broad consensus among experts. Often, this is not the case as researchers frequently apply diverging theoretical frameworks, and research findings are often ambiguous. A virtual monopoly in policy advice or a unified analytical framework on certain issues may therefore raise the effectiveness of policy advice. This, however, may be problematic if 'monopoly providers' of policy advice lose track of the scientific debate or if consensus is generated by ignoring new findings or competing approaches. Hence, the appropriateness of advice crucially depends on policyoriented researchers taking part in the academic discourse, with their work being inspired, but also discussed and evaluated by other, more 'academic' researchers.

With respect to powering, political actors can benefit from policy advice to the extent that it helps legitimizing decisions in political and substantial terms. Political actors can use policy advice in an opportunistic way to justify decisions taken for other reasons with selective reference to experts' statements. But they can also use policy advice to legitimize more far-reaching reforms that are painful for major parts of their constituency at least in the short run. Policy advice can be used to bind hands and avert demands to water down reforms and avoid blame. The extent of political and substantial legitimacy to be gained from policy advice depends, in turn, upon the extent of consensus among experts (Dyson 2005). Policy advice can play a crucial role if government faces high consensus requirements, i.e. government formed by coalitions, in situations of minority government or in political systems with social partner involvement. In such settings, policy advice can provide potential focal points for compromise and legitimize policy decisions.

The logic of policy-making, however, is not only dominated by the aim of problem solving but also, and maybe to the major part, by the goal of acquiring or defending political power (Lompe 2006). It may be the case that implementing an effective policy raises the chance of political success, but substantial labor market and welfare state reforms are often controversial and risky in political terms since they imply losses to be experienced by major groups in the electorate with positive effects resulting only in the long run. Policy advice aiming at relevance with regard to political decisions cannot remain completely 'academic'. It has to take into account the institutional restrictions of the status quo, the institutional incentives of the political system and the necessity of political actors to gain sufficient political support and legitimization for decisions.

Therefore, in order to become effective, policy advice has to be organized in a way that facilitates the provision of expertise on policy reforms that is aware of the institutional status quo and the political economy of reforms. (3) Since this moves beyond the analysis of policies, it may require the creation of a segment of policy advisors either through research institutes specializing in more applied research and policy analysis or temporary or permanent expert committees. However, to achieve broader and more sustainable legitimacy of policy reforms or more ambitious reform strategies, the creation of a "matter-of-fact" public discourse that allows for the adoption of pragmatic problem solutions seems crucial. This, in turn, may be facilitated by longstanding structures of policy advice that are highly reputable in both science and the public as opposed to expert committees created ad hoc upon initiative by government in a more "Machiavellian" style in order to generate short-term acceptance of policy proposals.

Policy advice may play a specific role in a corporatist setting with strong social partner involvement in the formulation and administration of labor market and social policies. Here, institutional infrastructure favoring bi-or tripartite talks can help overcome political deadlock and exploit policy complementarities in particular if institutional settings favor the convergence of policy concepts and broad political exchange (Ebbinghaus/Hassel 2000). But social Therefore, the structure of policy advice will be related to its effectiveness. In the next sections we will analyze the provision of policy advice and assess its role in recent welfare state and labor market reforms in Germany, the Netherlands, and Sweden. We particularly focus on the relation between the structure of policy advice and the potential for effective influence on the adaptation of social and labor market policies. In doing so, we aim at empirical evidence on the capacity of policy advice to legitimize reform policies and further societal problem-solving capacities.

3 Germany: Multiple Forms of Policy Advice, Limited Effectiveness

The German landscape of policy advice in economic and labor market policy is both rich and highly differentiated (Gellner 1995, Thunert 2001, Cassel 2004, Eichhorst/ Wintermann 2006). We can distinguish five types of providers of policy advice: 1. public research institutes, 2. social partners' think tanks, 3. private think tanks and research institutes, 4. permanent expert committees, and 5. temporary committees with either corporatist or noncorporatist composition.

Regarding the first group, six leading economic research institutes can rely on stable basic funding from Federal and Land governments: the German Institute for Economic Research (DIW, Deutsches Institut fur Wirtschaftsforschung), the Ifo Institute for Economic Research (ifo Institut fur Wirtschaftsforschung), the Centre for European Economic Research (ZEW, Zentrum fur Europaische Wirtschaftsforschung), the Rhenish-Westphalian Institute for Economic Research (RWI, Rheinisch-Westfalisches Institut fur Wirtschaftsforschung), the Kiel Institute for World Economics (IfW, Institut fur Weltwirtschaft) and the Halle Institute for Economic Research (IWH, Institut fur Wirtschaftsforschung Halle). All of them cover the whole range of economic research and specialize to a significant extent in economic analysis of labor market and social policies. They are clearly part of the academic sector since incentives for scientists are set in such a way that academic achievement is more appreciated than particular effort in policy advice or applied work. These institutes are evaluated in regular intervals with the criteria for assessment mainly referring to academic output, i.e. publications in refereed journals. In case the scientific output is found to be insufficient, an institute will lose its public funding (4).

In the domain of policy advice, the main output is the joint economic forecast published on a semi-annual basis. Apart from projects funded through research grants from science foundations, they prepare reports commissioned by Federal or Land Ministries or other institutions. Since the research institutes benefit from basic funding they can also carry out autonomous research and provide genuine input to the public debate.

The social sciences feature less prominently in policy advice provided by public research. However, there are notable exceptions: the Max Planck Institute for the Study of Societies (MPIfG, Max Planck Institut fur Gesellschaftsforschung) and the Social Science Research Center at Berlin (WZB, Wissenschaftszentrum Berlin fur Sozialforschung). Although they mainly focus on basic social science research and contributions to the scientific community, both institutes played a major role in policy advice through their directors being members of temporary expert commissions.

In contrast to these institutes funded through Federal and Land Governments, the Institute for Employment Research (IAB, Institut fur Arbeitsmarkt-und Berufsforschung) is affiliated with and funded by the Federal Employment Agency, the central body responsible for implementing unemployment insurance and active labor market policy in Germany. It has the legal mandate to evaluate the effectiveness of labor market policy schemes. Through this particular position it is closer to the Federal Ministry of Economics and Labor than other institutes, thus benefiting from more direct interaction with policy-makers, although tendered research has grown recently so that other researchers could enter the field of labor market policy evaluation.

Public research institutes are complemented by the social partners' 'advocacy' think tanks. On the one hand, German employers mainly from the metal and electronic industry fund the Cologne Institute for Business Research (IW, Institut der Deutschen Wirtschaft), whereas the Institute of Social and Economic Research (WSI, Wirtschafts-und Sozialwissenschaftliches Institut) is part of the Hans Boeckler Foundation, which is funded through trade unions officials' advisory board compensations. Both IW and WSI provide analytical support for policies favoring their respective stakeholders by carrying out applied research that generates evidence in favor of political tendencies in accordance with the general orientation of their principals. In comparison with the publicly funded institutes, both WSI and IW are peculiar in their orientation towards the media and easily accessible publication formats. They are not subject to regular evaluation based on academic criteria.

In Germany, major private enterprises devote part of their wealth and revenue to think tanks that form the third category of policy advisors. The German Post AG supports the Institute for the Study of Labor (IZA, Forschungsinstitut zur Zukunft der Arbeit), an independent research institute specialized in labor economics and applied labor market research. A different strategy is pursued by the Bertelsmann Foundation, a private think tank organized as an operative enterprise foundation benefiting from part of the profit generated by the Bertelsmann AG, Germany's largest media company (Welzel 2006). In contrast to research institutes, the Bertelsmann Foundation is less oriented towards contributing to the scientific debate on its own, but puts more emphasis on the public dissemination of analytical and policy papers and on establishing networks bringing together experts from the academic sphere and policy-makers. The Bertelsmann Foundation aims at furthering reform-oriented action through facilitating the transfer of expertise and policy concepts to actors and the media.

The fourth major category of policy advisors are permanent expert committees specializing in economic and labor market analysis. The most important one is the German Council of Economic Advisors ("Sachverstandigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung") established by legal provisions dating back to 1963 (Stratling 2006). The Council consists of five university professors of economics nominated by the Federal Ministry although an informal rule stipulates that two of them can only be appointed with consent by German employers or trade unions respectively. The Council has a high reputation for its independent scientific analysis of the German economy and the labor market. The main task of the Council is the preparation of an annual report on the German economy and economic policy issues which is published in late autumn. The government is obliged to reply to this analysis through the official Annual Economic Report presented by the Minister of Economics and Labor.

Temporary expert committees finally complement the advisory landscape (Siefken 2003). It is useful to distinguish between corporatist and non-corporatist, i.e. more pluralist committees appointed by the government. Regarding labor market policies and reforms the most relevant governmentinitiated expert commissions were the Benchmarking Committee of the Alliance for Jobs, Vocational Training and Competitiveness (1998-2001) and the Hartz Commission (February-August 2002). In a way similar to the Council of Economic Advisors, the Benchmarking Committee was composed by three independent social scientists, not economists -the director of the MPIfG, the head of the research section on labor markets at the WZB, and a university professor -on the one hand, and the heads of IW, the employers' institute, and the trade unions' Hans Boeckler Foundation on the other. Through this combination of both independent experts and scientists affiliated with the social partners, the Benchmarking Committee was supposed to provide analytic input for the tripartite Alliance for Jobs created in late 1998.

In contrast, the Hartz Commission was a more pluralist expert commission. Peter Hartz, then member of the board of Volkswagen and responsible for personnel affairs, headed a commission created in spring 2002 that was asked to present a reform proposal for labor market policies. The Hartz Commission was formed by fewer scientists, only a minority of representatives from the social partners and some active politicians from the Land and municipal level, whereas a prominent role was reserved for entrepreneurs and business consultants, i.e. professional 'change agents' (Schmid 2003). The model of pluralist expert groups was repeated with the 'Rurup Commission' created in late 2002 in order to formulate reform proposals for the social security system. It was headed by a leading member of the Council of Economic Advisors. In contrast to the permanent councils, temporary expert groups have a limited mandate as regards the scope of their task and they are asked to not only provide analysis but also policy recommendations, and their activities end with the presentation of a final report.

Compared to the rich infrastructure of policy advice in Germany, its actual role in recent German labor market and welfare state reforms is less impressive. First, what is most notable is the relatively minor influence exerted by the public research institutes and the permanent expert committees such as the Council of Economic Advisors when it comes to concrete decision-making. Although both the institutes and the Council continuously work on labor market issues and publish their findings regularly, the actual shortterm impact of their analyses and reform proposals is limited. Research output, the semi-annual forecasts as well as the Council's annual report receive a certain amount of public attention and are referred to in a selective way by political actors, but they rather provide background information for ministerial officials.

A clear short-term influence on actual decisions can hardly be identified although economic research can shape the public debate and agenda in the long run. One reason for this may lie in the fact that neither the analytical nor the conceptual work by the researchers reflects the logic of political decision-making which faces significant institutional and politico-economic restrictions. Hence it is difficult for policy-makers to draft institutional reforms that might be feasible both in political and institutional terms based on this type of analysis.

Therefore, over the last years the most relevant attempts at labor market reforms were based on the work of temporary expert commissions created explicitly to provide analytical and conceptual input for the government. In comparison to the research institutes and the advisory councils, these commissions adopted a more policy-oriented perspective in that they took political and institutional restrictions into account, thus reflecting the conditions for the realization of their proposals. The most recent commissions also differed in that they did not mainly unite economists but experts with different professional backgrounds. This holds in particular for both the Benchmarking Committee and the Hartz Commission.

The Benchmarking Committee aimed at establishing a set of empirical data on German labor market performance as well as on the major institutional factors influencing it. This Committee was also expected to assess the strengths and weaknesses of the German employment system so that areas could be identified where reforms were needed most urgently. Policy proposals should be derived from 'best practices' in other countries. Through this analytical work by the Benchmarking Committee, the founders and coordinators of the Alliance for Jobs in the Federal Chancellery expected to reach a joint assessment by trade unions and employers taking part in the Alliance for Jobs on the German employment situation and on potential remedies to improve labor market performance.

In that way both the Alliance for Jobs and the Benchmarking Committee were designed in accordance with the Dutch model (Visser/Hemerijck 1997, Ebbinghaus/ Hassel 2000), which at that time was perceived as an effective regime of cooperative welfare state reform. Building upon futile attempts by the Concerted Action ("Konzertierte Aktion") of the seventies and by the Kohl government in the mid-nineties, the creation of the Alliance for Jobs and the Benchmarking Committee under the Red-Green coalition was an attempt to establish an institutionalized forum for wider and more general discussions between the social partners.

However, scientific support of the Alliance for Jobs could not succeed in helping establish a political agenda as both government and social partners were unwilling to engage in a "joint perception of unpleasant facts" and a substantial discussion on structural reforms (Streeck/Hassel 2003, Siegel 2003). On the one hand this was due to strategic deficiencies and the institutional weakness of the German Federal Government that was neither willing nor capable of inducing employers and trade unions to serious joint reflection and negotiations on labor market reforms within the framework of the Alliance for Jobs. Government on an unconditional basis made major concessions to both sides. On the other hand the work of the Benchmarking Committee was hampered by half-hearted support from the social partners' think tanks that provided counter-evidence supporting their divergent perceptions of German problems at the same time. During its short existence it could not achieve sufficient public reputation, which would have been necessary to save some influence beyond the gridlocked framework of the Alliance.

Hence it was not possible to formulate a joint view on the labor market and the institutional adaptations to be implemented. Trade unions, employers, and government did not share beliefs concerning the economic situation, the basic explanatory framework and policy options. The main contentious issue was the diverging perception of the need for institutional adaptations in the welfare state, the role of wage moderation and macroeconomic policies (Dyson 2005).

This was highlighted in the debate on the Committee's proposal to subsidize low-wage earners through reduced social security contributions. Whereas advocates of this approach expected significant job growth favoring the low-skilled, critics from the trade unions feared undercutting of established minimum wages. Employers favored cuts of benefit levels, and government was afraid of the high fiscal cost of such a scheme in the short run (Heinze/Streeck 2003).

The final report by the Benchmarking Committee was completed in 2001. It provided a comprehensive and comparative assessment of the German labor market (Eichhorst/Profit/Thode 2001) but was published at a time when the Alliance for Jobs was in a severe deadlock. Some of the findings advocating institutional adaptation of the German employment system were rejected by the trade unions, whereas the employers claimed part of the arguments to be supportive of their point of view. Hence, it could not form the basis for further joint work by the Alliance for Jobs, which was dominated by the social partners' interest in defending established political positions and safeguarding direct access to the government. Faced with persistent deadlock of the Alliance for Jobs, the advisory infrastructure also lost its relevance (Streeck 2003, Streeck/Hassel 2003, Schmid 2003). (5)

The situation changed completely with the de facto termination of the Alliance for Jobs in early 2002 and the shift to a different temporary expert commission, the Hartz Commission. In direct response to the "placement scandal" in the Federal Employment Agency in early 2002, the Hartz Commission was created upon initiative by the Federal Chancellery and asked to formulate proposals for the reform of labor market policy and administration in Germany (Sell 2005, Schmid 2003). In terms of politics it was the attempt to limit the role of the social partners in German labor market policy, which were made responsible for ineffective policy design and implementation. The Federal Employment Agency, one of Germany's largest public administrative bodies, was heavily criticized for its bureaucratic inefficiency. This was attributed to a lack of competition and pervasive influence of the social partners advocating ineffective measures favorable to their clientele but detrimental to cost efficiency and the effective operation of labor market policy (Trampusch 2002, Streeck/Trampusch 2005).

Although labor market policies had attracted some criticism in the past, reforming both the administrative setting and the repertoire of instruments was assumed to be virtually impossible and not considered a feasible option before. Most economic policy advisors and the Benchmarking Committee had urged reforms in labor market policy in the past, but this did not enter the political agenda until the placement scandal opened up the window of opportunity for a policy change and a shift in the set of actors, which was a necessary precondition to implement a different labor market policy regime.

The reform initiative was triggered by the Federal Chancellery, thus marginalizing not only the Ministry of Labor and Social Affairs which was deeply entrenched in corporatist labor market policy controlled by trade unions and employers, but also the established advisory councils, the Alliance for Jobs and the Benchmarking Committee. Through bringing in a pluralist ad hoc expert group headed by a charismatic labor director of a large firm who had proven his ability to implement innovative models for enterprise-level personnel policy in the past, the work of the Hartz Commission marked a break with past labor market policies and corporatist arrangements. The Hartz Commission, under close supervision by the Chancellery and the media, formulated a package of reform proposals addressing the administration and the set of active labor market policy programs as well as some adjacent policy areas, thus expanding its mandate to cover also employment policy in a broader sense.

Reaching consensus among all members of the Commission was not only the personal achievement of Hartz. This was also the result of purposeful action neutralizing potential opposition through omitting issues that would have been controversial such as cuts in unemployment benefits or dismissal protection. Consensus was furthered by package deals, e.g. liberalization of temporary work agencies in exchange for coverage by equal pay or collective agreements. Innovations in terminology and substance stem from the involvement of entrepreneurs and consultants and from policy learning based on the perception of national and foreign 'best practices'. Finally, Hartz' early disclosure of the major modules of his reform concept via the media helped neutralize potential opposition from within the commission (Fleckenstein 2004, Schmid 2003).

Together with the strategic use of the Hartz Commission's work in the Social Democrats' electoral campaign of summer 2002 based on the general statement to fully implement the Hartz reforms in order to halve unemployment over the years to come, the peculiar and ambiguous character of the proposals contributed to the stunning career of the Hartz concept. It dominated the political agenda of Chancellor Schroder's second term. In the tight schedule of the run-up to the elections, the government committed itself credibly to a comprehensive implementation of the Hartz proposals, thus attempting to bind hands through "government by commission", leaving little room for interest group intervention, in particular trade union opposition (Dyson 2005).

As with the Benchmarking Committee before, the Bertelsmann Foundation supported the Hartz Commission by providing a collection of 'best practices' of labor market policies at the local level and in other countries and by funding fact-finding missions to some countries (Schmid 2003, Fleckenstein 2004). Finally, the Hartz reforms introduced an innovative element in that the Federal Ministry of Economics and Labor devoted significant resources to careful and comprehensive evaluation, an issue neglected in the past. The main beneficiaries of the evaluation grants are the IAB and the public research institutes, whereas the internal restructuring of the Federal Employment Agency was organized with the help of business consultants that had participated in the Hartz Commission.

To take advantage of the reform momentum gained through the Hartz Commission and to further exploit the successful model of a pluralist expert commission as agenda setter for a complex policy area susceptible to deadlock, the government installed the Rurup Commission in autumn 2002. But it was not possible to repeat the Hartz experience. On the one hand, this can be explained by the fact that no window of opportunity was open for a general reconstruction of the pension and health care system. On the other hand, the members of the Rurup Commission could not agree on a unified reform proposal on health care funding. Finally, interest groups were much more influential in this phase, thus drawing lessons from being taken by surprise by the Hartz Commission and the government's determination to implement the proposals without much debate.

The Rurup proposals were also met with stronger resistance from the Social Democratic Party not under pressure to adopt them under the tight schedule of an electoral campaign. The government, however, continued its more active agenda setting through the Agenda 2010, substantially specifying some of the issues raised by Hartz and Rurup (Dyson 2005) which had not been regulated so far, such as reforms in dismissal protection or the determination of the benefit level for the long-term unemployed. This sequence of government-initiated reforms was characterized by an erosion of political support and legitimacy that finally resulted in the defeat of the Red-Green coalition in early elections in 2005. This was most pronounced with respect to the Fourth Hartz Reform Act that modified benefits for the long-term unemployed. Hence, initial acceptance of the Hartz package withered away as more concrete and "cruel" decisions had to be made (Eichhorst/Sesselmeier 2006). Postponing the statutory retirement age from 65 to 67, originally proposed by the Rurup Commission in 2003, but rejected by the government at the time, resurfaced on the political agenda in spring 2006 and was adopted in fall 2006.

The experience of the Alliance for Jobs, the Hartz and Rurup Commissions and the Agenda 2010 show new forms of governance repelling the role of the social partners and their affiliated think tanks and a growing role of temporary expert committees with a pluralistic composition capable of providing pragmatic advice that is of direct use to the policy-makers and helps government-driven agenda setting beyond established party programs. The rise of pluralistic commissions, entrepreneurs and consultants as policy advisors also means that established bodies such as the Council of Economic Advisors but also the public and the social partners' research institutes are challenged in their position as prominent providers of policy advice and risk losing influence in the political arena.

In order to counter this threat, the Council of Economic Advisors, but also the public research institutes and the think tanks associated with the social partners, modified their publication portfolio and intensified interaction with the public. As regards the Council of Economic Advisors, the recent annual reports have become more accessible and more oriented towards concrete reform proposals although, in formal terms, the Council shall not recommend selected policies, e.g. the report issued in autumn 2002 presented a package of labor market reforms. The academic orientation of the public research institutes is now being supplemented by attempts at gaining more attention through shorter policy papers, workshops with policy makers and more intense media contact.

The need to adapt is even stronger for the social partners' institutes that are affected negatively by, first, the longterm decline of social partnership and organizational membership and, second, the stronger and more autonomous agenda setting by the government. Apart from providing analytical support for their stakeholders, both IW and WSI now increase activities addressing the media and the wider public through short and concise newsletters providing easily accessible information and pointed arguments. "Chancen fur alle", a medium-term campaign funded by the German metal and electronic industry, is notable. Its main activity is the dissemination of information advocating further reforms of the labor market and the welfare state in the direction of a "new social market economy" with a clear liberal orientation, the content for the media mainly being provided by IW. However, the employers' support is camouflaged in that the initiative prefers to appear as independent and non-partisan.

Recent developments of policy advice in Germany show different modes of bridging the wide gap between scientific knowledge and policy-making. While the growing importance of non-academic think tanks and pragmatic advice provided by temporary expert commissions but also by business consultants may help overcome the divide between academic research and policy-makers, this might also result in blending highly selective scientific arguments with subtle forms of lobbying in favor of particular economic interests. This may raise some concerns regarding the scientific accuracy and quality of the most influential consulting activities in recent years. It certainly is problematic to give too much room to actors that are less transparent and not controlled by public discourse or scientific standards. While this might help overcome some of the traditional blockages in German corporatist policy-making and thinking, it might also mean a higher degree of selectivity concerning the 'puzzling' phase of policy-making processes with arguments and proposals receiving higher attention if they are orchestrated by more powerful campaigning. In the long run, this might put the legitimacy of policy advice itself at risk.

Analyzing policy advice in Germany, we can see that despite of or even due to the wide variety of policy advice and different approaches in recent years the results regarding its effectiveness are mixed at best. Although legitimizing reforms through policy advice is of particular importance given the precarious resources of German governments, this was only achieved in a particular situation where a window of opportunity could be exploited by government action based on the work of a temporary expert committee like the Hartz Commission. This facilitated more autonomous agenda setting by the government. Bypassing the social partners through a pluralist committee was also a clear break with futile attempts at bringing about reforms through corporatist agreements. Despite recent attempts to bridge the gap between science and politics, policy advice still suffers from a strict divide between economics and social science and between academic research and policy-oriented work taking institutional and politico-economic restrictions into account. Hence, policy advice was hardly able to realize its full potential in raising capacities for institutional reform through viable policy proposals and legitimizing further labor market and welfare state reforms.

4 The Netherlands: Social Partnership Controlled by Independent Policy Advice

In the Netherlands, policy advice played a prominent role in labor market and welfare state reform. This is mainly due to specific conditions regarding the supply of policy advice and the discussion of social policy and welfare state issues by social partner bodies (Visser/Hemerijck 1997, Hemerijck 2003, den Butter/Mosch 2003, Andeweg/Irwin 2002).

As regards the science segment, independent but pragmatic and problem-oriented advice has a strong position in the Netherlands. Located right at the heart of policy advice in economic and labor market policy is the Netherlands Bureau for Economic Policy Analysis (CPB, "Centraal Planbureau"), which was established through legislation after the Second World War. Although it is mainly funded by the Ministry of Economic Affairs, it is regarded as an independent policy-oriented research institute. Its major activity is the macro-economic modeling of the Dutch economy and the potential effects of policy reforms. Due to a long experience in developing and fine-tuning a macroeconomic model of the Dutch economy, it has a virtual monopoly on this kind of analysis.

Nevertheless, the CPB's work is up to date with economic research as it interacts intensely with the scientific community. Reputation in science reinforces reputation and credibility in the political sphere which is also strengthened by CPB's mainly empirical approach refraining from direct policy proposals and normative statements. CPB produces the "Central Economic Plan" in spring which forecasts the development of the Dutch economy and some other economies in the current year, which then is updated in the autumn "Macroeconomic Outlook" with a projection on the following year. This forms the basis for the government's budget proposal presented at the same time. CPB forecasting also influences wage bargaining between the social partners. As CPB's major strength lies in its modeling capacities, it can operate as a sort of "court of audit" ex ante and thus help take the economic consequences of policy decisions into account, avoid mistakes and strengthen long-term orientation in politics.

In addition, CPB evaluates the probable economic and fiscal outcomes of policy proposals that are part of the electoral manifestos presented by political parties. Although referral to CPB is voluntary, this is high on the public agenda so that political parties cannot refuse to have their proposals evaluated by CPB (Seils 2005). Since all manifestos are assessed using the same macroeconomic model and parties even refer to CPB before formulating their proposals, this leads to convergence in parties' economic policies.

CPB effectively influences coalition negotiations through providing data and simulations of budgetary room to maneuver and an ex ante evaluation of potential policies, this being implemented by the Central Economic Commission of the Ministry of Economics. It helped stabilize budget policy in the Netherlands through the "Zalm norm" of 1994, a limit on real public expenditure that was based on cautious projections by CPB.

Apart from forecasting, CPB carries out mandated research and policy analysis for the government and other bodies as regards potential effects of policy proposal, this, however, in competition with other research institutes, which also means that CPB has to be competitive and up to date regarding the latest developments in economics. Since CPB maintains strong networks with political actors, ministries, the social partners and other researchers, it can be described as a provider of "commonly understood facts" (Hemerijck 2003).

The Netherlands Scientific Council for Government Policy (WRR, Wetenschappelijke Raad voor het Regeringsbeleid), established by law in 1976, is an important independent and interdisciplinary think tank that is highly influential in raising public and political awareness of longterm issues relevant to the Dutch economy. E.g. in 1990, the WRR presented a report entitled "A working perspective" (Hemerijck 2003) urging a "paradigm shift" in favor of employment growth and labor market integration instead of lowering open unemployment through reduction of labor supply, i.e. expansion of disability pensions. By making the inactive/active ratio a major governance benchmark, it facilitated long-term reorientation towards more inclusive and activating labor market policies.

Policy advice in the Netherlands cannot be understood properly without taking social partnership into account. The more "societal engineering" perspective of CPB and WRR is complemented by corporatist bodies organizing the link between state and society, i.e. trade unions and employers.

First, the Labor Foundation (STAR, Stichting van de Arbeid), established in 1945 under private law and mainly funded through the social partners, operates as a national consultative body. The peak associations of employers and trade unions form STAR. Within this framework, the social partners can enter pragmatic discussions on a wide range of issues well beyond collective bargaining. They can adopt joint opinions on social or economic subjects and respond to government and parliament requests. The government meets with STAR twice a year.

The bipartite Labor Foundation is supplemented by the tripartite Social and Economic Council of the Netherlands (SER, Sociaal-Economische Raad) created by public law in 1950. It is seen as the main advisory body of the Dutch government although it is financed by industry. It brings together employers' peak associations, the peak associations of the Dutch trade unions and a third group of 'crown representatives', i.e. independent experts such as university professors in economics, social science or law, the president of the Dutch Central Bank and the director of CPB. The government, upon proposal by the Council, appoints the head of SER. In the meetings of the Council, but also in the sessions of its tripartite working parties, ministerial officials take part as observers, thus facilitating transfer of information. In SER, the independent members provide input from science and foster compromise.

Unanimous statements issued by SER are highly influential with regard to advising the government on economic issues and concrete policy decisions, divided opinions are less effective, but also signal less political support to the government (Hemerijck 2003). As of 1995 mandatory consultation by government was abolished. Incentives to issue unanimous opinions have become stronger within SER since then.

SER and STAR form the core of the Dutch 'consultation economy'. The corporatist bodies promote exchange and convergence of views on economic problems and possible action to be taken and thus make coordination of policies across policy areas possible. At the same time, social partner bodies rely on analytical work by CPB, which reinforces convergence as regards policy assessment and proposals. This could not be achieved by the Benchmarking Committee in the German Alliance for Jobs. Another important factor that distinguishes the Netherlands from Germany is the stronger role of the government as an agenda setter and its potential and sometimes actual threat of intervention (Hemerijck 2003, Seils 2005).

With regard to policy-making, STAR played an important role in reforms in the Netherlands, in particular in the sequence of welfare state and labor reforms triggered by the change in government and the famous Wassenaar agreement in 1982, which was signed at STAR, with both the coalition agreement and the social partners' approach mirroring the analytical findings of CPB (Visser/Hemerijck 1997, Seils 2005). Further steps often relied on joint policy recommendations by the social partners in the 'shadow of hierarchy' and were based on CPB work (Hemerijck 2003). This holds for the agreement "A New Course" in 1993, advocating a more activating labor market policy, and the bipartite agreement on wage restraints in exchange for postponed cuts in disability benefits in 2003. After a phase of public unrest, it was replaced by a genuine tripartite agreement in 2004, when government backed down on a more stringent reversal of early exit from the labor market (Seils 2005).

As regards the administration of social security and labor market policy, the role of social partners was curtailed after criticism from the crown members of SER, the Dutch Audit Office and an all-party parliamentary enquiry commission, the Buurmeijer commission, that presented its report in 1993 and attacked corporatist administration of disability pensions which favored reduction of labor supply at the expense of the public. This led to fundamental reorganization of social security and labor market policy administration. Further steps restricting the use of disability were advocated by the Donner Commission which presented its report in 2001. It was supported by SER unanimously.

However, if the social partners are able to achieve substantial compromise, they can effectively influence legislation as done with the adoption of the 'flexicurity' approach to labor market regulation which was furthered by a STAR document on flexibility and security in 1996 (Hemerijck 2003). The most recent economic difficulties and more intense distributive conflicts in the Netherlands question the effectiveness of the Dutch arrangement of 'smooth' policy advice and policy-making as the government seems to be less willing to listen to both independent policy advisors and the social partners. But it would be premature to claim that it is not operative anymore (Seils 2005, Hemerijck 2003).

Policy advice in the Netherlands is organized through a dense network of researchers, the social partners, and political actors, with the formal setting also allowing for continuous informal exchange of information, analytical findings from different parts of science, but also political arguments, thus helping achieve a high level of trust (den Butter/Mosch 2003). A pragmatic approach towards policymaking is also furthered by the interdisciplinary composition of Dutch labor market research. The Dutch system of policy advice can be seen as an arrangement where CPB and WRR, independent think tanks with a high reputation, evaluate policy options and point at long-term challenges, thus improving the governance capacity of the Netherlands. At the same time they influence the social partners' STAR and SER, which are effective bodies for policy design and negotiations in that CPB and WRR help limit externalization policies favored by the social partners. Such reforms had to be initiated by the government with analytical support from CPB and WRR since it could not rely on support from the social partners (Hemerijck 2003, Seils 2005). They were rather legitimized with reference to CPB and WRR, but nevertheless Dutch capacity for institutional reform is also enhanced by institutional infrastructure for social partners' deliberation in the government's shadow of hierarchy and based on findings of independent think tanks.

5 Sweden: Policy Advice by Institutionalized Expert Commissions

Sweden has an institutional arrangement that facilitates the consideration of scientific findings and policy advice in actual policy-making when it comes to welfare state and labor market reforms. Part of the impressive reform sequence in the nineties can be attributed to an effective system of policy advice (Palme/Wennemo 1998, Benner/Vad 2000, Lindbeck 2002, Jochem 2003, Wintermann 2005).

Two main features play a crucial role: regarding the policy-making process, minority governments have always been quite common in Sweden. A government without a stable majority in parliament, with the prime minister being elected despite the lack of an absolute majority, is more dependent upon sufficient societal support and credible justifications for their decisions. Minority governments have to search for legitimacy for each policy decision. This is a strong incentive for problem-oriented and highly pragmatic policy-making. In such a context, policy advice can be helpful in establishing a basis for consensus and in legitimizing deviation from party programs that might be necessary to build an issue-oriented coalition. Therefore, minority governments can be assumed to be more open to policy advice and thus be more able to take long-term considerations arising from advisory bodies into account in order to legitimize their political action.

This relates to the second prerequisite for effective policy advice in Swedish social and employment policy making: the crucial role played by temporary expert committees which differ in their position, their mandates, objectives and operation from advisory councils in other countries (Jann et al. 2005). Expert commissions in Sweden are not primarily involved in the discussion of concrete bills but deal with general and long-term challenges to the Swedish economy and welfare state. The joint opinion of a commission is expressed in a highly renowned series of reports, the "Statens Offentliga Utredningar" (SOU). These reports shape both public and political debate and also trigger further analysis. Since the public also discusses the findings of SOU reports and is quite aware of the major conclusions, policy-makers are well advised to take the expert commissions' work into account. Upon assignment by the government, the Swedish commissions bring members of parliament and experts together, in particular researchers, representatives from the public administration and the social partners with some experts having voting rights only.

External experts and assistants support the commissions' work; the speaker of a commission is usually not a member of parliament. Regarding the role of academic experts, the Swedish commissions benefit from the fact that both experienced and young researchers are involved, the latter having the opportunity to acquire some reputation through their contributions to the commission. The participation of both researchers and members of parliament, and the funding through the governments' budget, point at the close interaction between scientists and political actors. The incorporation of politicians as well as researchers into Swedish commissions is a notable feature. It is fundamentally different from committees that consist of either experts or politicians only. The joint discussions of politicians and experts facilitate the transfer of arguments with respect to both scientific expertise and political considerations.

Regarding Swedish welfare state reform over the nineties, the SOU played a major role in providing policy orientation and informing political actors on viable policy options. At the beginning of the nineties, commission reports triggered a debate on the Swedish welfare state and the need for institutional adaptation (SOU 1990:44, SOU 1994:20, SOU 1996:113, Lindbeck et al. 1994). This is particularly true for preparing the ground for reforms in pensions and disability benefits. Initial analyses addressed demographic changes and their consequences for the economy and the public budget. Further commission work resulted in the formulation of a reform proposal for the Swedish old-age pension regime, which was implemented with broad support from the major political parties in 1994 against opposition from the trade unions (Lachman et al. 1995, Palme/Wennemo 1998). The Swedish budgeting procedure was recalibrated after further commission work in 1997 and 2000.

In general, we can see that commissions contributed to a rather centralized budgeting and policy-making process involving the social partners much less than in prior phases of corporatist politics in Sweden. In particular, expert commissions helped formulate a strategy to consolidate the public budget and to reform welfare state schemes, such as old-age pensions, with positive long-term effects on the economy. The Swedish budgeting procedure is now mainly based on the economic framework and forecasting and takes potential effects on the economic activity into account while limiting interest group influence that would have made longterm budget consolidation more difficult (Molander 2001).

Commission work, finally, is supported by research institutes working in the areas examined, while at the same time some institutes were even created upon request of a commission. This holds for the CEFOS (Centrum for forskning om offentlig sector) that specializes in studies on public sector efficiency. Other research institutes are also closely related with the political sphere in that their heads are assigned by the government and the main budgets are allocated by the ministries, among them the Institute for Future Studies (Institutet for Framtidsstudier) and the Swedish Institute for Social Research (SOFI, Institutet for social forskning) at the University of Stockholm, the major institute for welfare state research. In addition, the decentralized National Institute for Working Life (Arbetslivsinstitutet) focuses on working life.

The evaluation of active labor market policies and more general labor market research are the main task of the Institute for Labor Market Policy Evaluation (IFAU, Institutet for Arbetsmarknadspolitisk Utvardering). Rather skeptical evaluation reports by IFAU influenced the reorientation in Swedish labor market policy in the late nineties away from massive spending and increases in participant inflow in the earlier years of the decade to cuts in expenditure and lower participation (Calmfors/ Forslund/Hemstrom 2001, Jochem 2003). Research councils dealing with different topics such as the Swedish Council for Working Life and Social Research (FAS, Forskningsrad for arbetsliv och socialvetenskap) play an important role in funding social science research but also in advising the government.

The essential role of the commissions in the longterm adjustment of the Swedish welfare state was emphasized again in the late nineties when the reforms over the last decade were reexamined through commission analyses (SOU 2000:3, 2001:57). Intense public debate, partially fuelled by commission experts participating in series of debates in newspapers such as "Dagens Nyheter" or "Svenska Dagbladet" (Jahn 2003), contributed to high public awareness of underlying problems and viable policy options so that policy makers can hardly ignore the commissions' work. Hence, the commissions' opinions often form focal points for problem-oriented solutions beyond party tactics. The high reputation and public standing of the expert commissions helped legitimize this shift, which removed budgeting and welfare state reform to a certain degree from party competition and helped adopt a more objective or 'technocratic' approach. It is in that respect similar to the Dutch modeling and forecasting approach. Through these mechanisms-mixed expert commissions including both researchers with different disciplinary backgrounds and political actors, fundamental and balanced reports, and high public awareness-the findings of the Swedish expert committees have a higher chance of being taken into account in policy-making processes.

The Swedish commissions have proven to be of particular importance with regard to the long-term consolidation of the budget and in promoting structural changes of the welfare state. In recent years they were clearly more important in these policy areas than advice given by the social partners. The erosion of highly centralized Swedish corporatism since the early nineties regarding social partner influence on the formulation of social and labor market policies, their participation in expert commissions (SOU 1999:121, Jochem 2003), and the transformation of collective bargaining (Wintermann 2005, Jochem 2003) is striking in this respect.

The most important welfare state reforms were not designed in cooperation with the social partners but mainly pushed by subsequent (often minority) governments in a unilateral way with more emphasis on budget consolidation and structural adjustment facing deep economic crisis in the early nineties and access to the EU when pursuing a national full employment policy was no longer possible. The erosion of corporatist patterns, however, dates back to the midseventies when trade unions tried to establish wage earner funds and was highlighted by the breakdown of centralized bargaining caused by the withdrawal of the employers in the early nineties (Meidner/Hedborg 1984, Jahn 1994). Particularly trade unions lost influence as was highlighted in the 1994 pension reform, when they could not influence the political agenda to a significant extent but only postpone the implementation of the new law. However, these reforms paved the way for economic recovery in a fundamentally different economic environment.

Therefore, it seems fair to argue that the institutionalization of the Swedish commissions as an essential part of the policy process with respect to both puzzling and powering helps legitimize effective policies since a high degree of open and pragmatic public debate provides a high level reputation to the commissions' reports that form the nucleus for generally acceptable policy decisions. The commissions' work also benefits from the existence of a whole range of interdisciplinary research institutes combining economics and social science with a clear empirical and pragmatic approach to societal problems and their potential solution.

6 Conclusion: Legitimizing Welfare State Reforms through Effective Policy Advice

The comparative analysis of policy advice in labor market and welfare state reforms in Germany, the Netherlands and Sweden shows different institutional structures of advisory bodies and divergent experience with the practical relevance of policy advice. As regards Germany, we can see a highly complex arrangement of public and private research institutes, social partner think tanks and both permanent and temporary expert committees. In spite of that, there is a deep divide between science-mainly economics-and policy makers. However, during the Schroder government we could see growing importance of pluralist expert committees with limited mandate and fixed-term duration, the major example being the Hartz Commission, which was used to legitimize substantial changes initiated by the red-green government after the breakdown of corporatist concertation.

Pluralist expert committees appointed by government seem to be able to bridge the gap between science and politics best and to take institutional and politico-economic restrictions into account, thus generating legitimacy or facilitating acceptance for partial welfare state and labor market reforms. However, Germany lacks a structure of policy advice that can generate conceptual convergence, pragmatic compromise and legitimization for reforms compatible with the demands of long-term institutional adjustment. This is different in the Netherlands and Sweden. Hence, ad hoc advice by temporary committees does not seem to be conducive to legitimizing more long-term solutions to problems based on a "matter-of-fact" approach, which is facilitated by highly reputable advisory bodies found in the Netherlands or Sweden.

The Dutch experience provides a contrasting example as think tanks like CPB and WRR evaluate policies and potential reform options and analyze long-term challenges to the economy and the welfare state, whereas social partner institutions provide the institutional infrastructure for close interaction between trade unions, employers and government, thus facilitating convergence. Independent and interdisciplinary think tanks inform the tripartite talks and point at negative effects of externalization strategies by the social partners. In Sweden, we can see a system that was no longer characterized by a dominant role of corporatism in recent years but by government-induced reforms that could rely on extensive analysis by mixed commissions bringing researchers from different disciplines and policy makers together. Commission reports intensely debated in public can form the basis for policy reforms that are also acceptable to the parliamentary opposition whose consent is often necessary in order to get a majority in parliament. This setting could help legitimize fundamental welfare state reforms in Sweden.

Our analysis shows how the institutional structure of policy advice can help further societal problem-solving capacities based on a careful joint assessment of facts and feasible options with regard to welfare state and labor market reforms. It is fair to say that highly reputable and less contested expert committees and research institutes providing balanced policy-oriented advice are most influential and conducive to furthering labor market and welfare state reforms in corporatist settings. In combination with a shadow of hierarchy thrown by government they also facilitate social partner consensus. Hence, an appropriate supply of policy advice can help ensure sufficient political and substantial legitimacy for institutional reforms and increase societal problem-solving capacities. If government is weak for institutional reasons and policy advice is rather fragmented and less policy-oriented, as in the German case, policy advice cannot realize its full potential. "Machiavellian" ad hoc committees cannot make up for this deficit. Further research is needed to analyze the development of policy advice in Germany after the shift from the Schroder government to the Grand Coalition.

References

Andeweg, Rudy/Irwin, Galen, 2002: Governance and Politics of the Netherlands, New York.

Benner, Mats/Vad, Torben Bungaard, 2000: Sweden and Denmark: Defending the Welfare State. In: Scharpf, Fritz W./Schmidt, Vivien A. (Eds.): Welfare and Work in the Open Economy, Vol. 2, Oxford, pp. 399-466.

Calmfors, Lars/Forslund, Anders/Hemstrom, Maria, 2001: Does active labour market policy work? Lessons from the Swedish experiences. In: Swedish Economic Policy Review, 8 (2), 61-131.

Cassel, Susanne, 2004: Politikberatung und Politikerberatung, 2nd ed. Bern.

Den Butter, Frank A.G./Mosch, Robert H.J., 2003: The Dutch miracle: institutions, networks and trust. In: Journal of institutional and theoretical economics, 159 (2), pp. 362-391.

Dyson, Kenneth, 2005: Autoritative and Reflexive Stategies for Binding Hands: Europeanization, Government by Commission, and Economic Reform. In: German Politics, 14 (2), pp. 224-247.

Ebbinghaus, Bernhard/Hassel, Anke, 2000: Striking Deals: The Role of Concertation in the Reform of the Welfare State. In: Journal of European Public Policy, 7 (1), pp. 44-62.

Eichhorst, Werner/Profit, Stefan/Thode, Eric, 2001: Benchmarking Deutschland: Arbeitsmarkt und Beschaftigung, Berlin.

Eichhorst, Werner/Sesselmeier, Werner, 2006: Die Akzeptanz von Arbeitsmarktreformen am Beispiel von Hartz IV, Bonn.

Eichhorst, Werner/Wintermann, Ole, 2006: Politikberatung in der Arbeitsmarkt-und Beschafitgungspolitik. In: Falk, Svenja/Rehfeld, Dieter/Rommele, Andrea/Thunert, Martin (Eds.): Handbuch Politikberatung, Wiesbaden, pp. 419-436.

Fleckenstein, Timo, 2004: Policy-Lernen in der Arbeitsmarktpolitik: Das Beispiel der Hartz-Kommission. In: Zeitschrift fur Sozialreform, 50 (6), pp. 646-675.

Gellner, Winand, 1995: Ideenagenturen fur Politik und Offentlichkeit, Opladen.

Glynn, Steven/Cunningham, Paul/Flanagan, Kieron, 2003: Typifying Scientific Advisory Structures and Scientific Advice Production Methodologies (TSAS)-Final Report, Manchester.

Lachman, Desmond et al., 1995: Challenges to the Swedish Welfare State. IMF Occasional Paper 130, Washington.

Heclo, Hugh, 1974: Modern Social Policy in Britain and Sweden: From Relief to Income Maintenance, New Haven.

Heinze, Rolf G./Streeck, Wolfgang, 2003: Optionen fur den Einstieg in den Arbeitsmarkt oder: Ein Lehrstuck fur einen gescheiterten Politikwechsel. In: Vierteljahrshefte zur Wirtschaftsforschung, 72 (1), pp. 25-35.

Hemerijck, Anton C./Schludi, Martin, 2000: Sequences of Policy Failures and Effective Policy Responses. In: Scharpf, Fritz W./Schmidt, Vivien A. (Eds.): Welfare and Work in the Open Economy, Vol. 1, Oxford, pp. 125-228.

Hemerijck, Anton C., 2003: A Paradoxical Miracle: the Politics of Coalition Government and Social Concertation in Dutch Welfare Reform. In: Jochem, Sven/Siegel, Nico A. (Eds.): Konzertierung, Verhandlungsdemokratie und Reformpolitik im Wohlfahrtsstaat-Das Modell Deutschland im Vergleich, Opladen, pp. 232-270.

Jahn, Detlef, 1994: Das Schwedische Modell-Zukunftsoder Auslaufmodell?, in: Rising Hintz, Gunilla (Ed.), Das Schwedische Modell-Zukunfts-oder Auslaufmodell?, Deutsch-Schwedische Gesprache, Marburg, pp. 21-38.

Jahn, Detlef, 2003: Das politische System Schwedens. In: Ismayr, Wolfgang (Ed.): Die politischen Systeme Westeuropas, 3rd ed., Opladen, pp. 93-130.

Jann, Werner, 1989: Parlament und Gesetzgebung: Aketure und Ressourcen der parlamentarischen Gesetzgebung im internationalen Vergleich, Habilitationsschrift, Hochschule fur Verwaltungswissenschaft, Speyer.

Jann, Werner et al., 2005: Regierungsorganisation als Institutionenpolitik: Ein westeuropaischer Vergleich, Forschungspapiere "Regierungsorganisation in Westeu ropa" 01/2005, Potsdam.

Jann, Werner/Tiessen, Jan, 2005: Gesetzgebung in Schweden, mimeo, Berlin.

Jochem, Sven, 2003: Das schwedische Modell im Wandel. In: Jochem, Sven/Siegel, Nico A. (Eds.): Konzertierung, Verhandlungsdemokratie und Reformpolitik im Wohlfahrtsstaat - Das Modell Deutschland im Vergleich, Opladen, pp. 271-310.

Lindbeck, Assar, 2002: Changing Tides for the Welfare State, CESifo working Paper 645 (3), Munich.

Lindbeck, Assar et al., 1994: Turning Sweden Around, Cambridge, Mass.

Lompe, Klaus, 2006: Traditionelle Modelle der Politikberatung. In: Falk, Svenja/Rehfeld, Dieter/Rommele, Andrea/Thunert, Martin (Eds.): Handbuch Politikberatung, Wiesbaden, pp. 25-34.

Meidner, Rehn/Hedborg, Anna, 1984: Modell Schweden - Erfahrungen einer Wohlfahrtsgesellschaft, Frankfurt.

Molander, Per, 2001: Budgeting Procedures and Democratic Ideals. In: Journal of Public Policy, 21 (1), pp. 23-52.

Palme, Joakim/Wennemo, Irene, 1998: Swedish Social Security in the 1990s: Reform and Retrenchment, Valfardsprojektet, Printing Works of the Cabinet Office and Ministries, Stockholm.

Pierson, Paul, 1994: Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment, Cambridge.

Pierson, Paul, 2000: Increasing returns, path dependence, and the study of politics. In: American Political Science Review, 94 (2), pp. 251-267.

Scharpf, Fritz W., 1994: Games real actors could play. In: Journal of Theoretical Politics, 6 (1), pp. 27-53.

Seils, Eric, 2005: Das Hollandische Wunder: Korporatismus und Konsens oder konfliktgeladene haushalts-und finanzpolitische Anpassung? In: Zeitschrift fur Sozialreform, 51 (2), pp. 194-220.

Sell, Stefan, 2005: Vom Vermittlungsskandal der Bundesanstalt fur Arbeit zu Hartz IV: Tiefen und Untiefen rotgruner Arbeitsmarktpolitik in einer Mediengesellschaft. In: Haubner, Dominik/Mezger, Erika/Schwengel, Hermann (Eds.): Agendasetting und Reformpolitik. Strategische Kommunikation zwischen verschiedenen politischen Welten. Marburg, pp. 285-310.

Siefken, Sven T., 2003: Expertengremien der Bundesregierung-Fakten, Fiktionen, Forschungsbedarf. In: Zeitschrift fur Parlamentsfragen, 34 (3), pp. 483-504.

Siegel, Nico A., 2003: Die politische Okonomie der Konzertierung in Deutschland. Das Beispiel Bundnis fur Arbeit. In: Jochem, Sven/Siegel, Nico A. (Eds.): Konzertierung, Verhandlungsdemokratie und Reformpolitik im

Wohlfahrtsstaat. Das Modell Deutschland im Vergleich, Opladen, pp. 148-193.

Statens offentliga utredningar, 1990: Makt och demokrati, SOU 1990:40, Stockholm.

Statens offentliga utredningar, 1994: Reformerat pensionssystem - Betankande av Pensionsarbetsgruppen, SOU 1994:20, Stockholm.

Statens offentliga utredningar, 1996: En allman och aktiv forsakring vid sjukdom och rehabilitering, SOU 1996:113, Stockholm.

Statens offentliga utredningar, 1999: Avkorporativisering och lobbyism, SOU 1999:121, Stockholm.

Statens offentliga utredningar, 2000: Valfard vid vagskal, SOU 2000:3, Stockholm.

Statens offentliga utredningar, 2001: Valfardens finansering och fordelning, SOU 2001:57, Stockholm.

Stratling, Ansgar, 2006: Der Sachverstandigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung. In: Falk, Svenja/Rehfeld, Dieter/Rommele, Andrea/Thunert, Martin (Eds.): Handbuch Politikberatung, Wiesbaden, pp 353-362.

Streeck, Wolfgang, 2003: No longer the Century of Corporatism. Das Ende des "Bundnisses fur Arbeit", MPIfG Working Paper 03/4, Koln.

Streeck, Wolfgang/Hassel, Anke, 2003: The Crumbling Pillars of Social Partnership. In: West European Politics, 26 (4), pp. 101-124.

Streeck, Wolfgang/Trampusch, Christine, 2005: Economic Reform and the Political Economy of the German Welfare State. In: German Politics, 14 (2), pp. 174-195.

Thunert, Martin, 2001: Politikberatung in der Bundesrepublik Deutschland seit 1949, in: Willems, Ulrich (Ed.): Demokratie und Politik in der Bundesrepublik, Opladen, pp. 223-242.

Trampusch, Christine, 2002: Die Bundesanstalt fur Arbeit und das Zusammenwirken von Staat und Verbanden in der Arbeitsmarktpolitik von 1952 bis 2001. MPIfG Working Paper 02/5, May 2002.

Visser, Jelle/Hemerijck, Anton, 1997: 'A Dutch Miracle'. Job Growth, Welfare Reform and Corporatism in the Netherlands, Amsterdam.

Weaver, R. Kent, 1986: The Politics of Blame Avoidance. In: Journal of Public Policy, 6 (4), pp. 371-398.

Welzel, Carolin, 2006: Politikberatung durch Stiftungen. In: Falk, Svenja/Rehfeld, Dieter/Rommele, Andrea/ Thunert, Martin (Eds.): Handbuch Politikberatung, Wiesbaden, pp 275-289.

Wintermann, Ole, 2005: Vom Retrenchment zur Krisenreaktionsfahigkeit--Ein empirischer Vergleich der Wohlfahrtsstaaten Schweden und Deutschland 1990-2000, Wiesbaden.

(1) We are grateful to Anton Hemerijck, Eric Seils and Thore Wintermann for helpful information on recent developments in policy advice in the Netherlands and Sweden. We are particularly indebted to Sven Jochem, Julia von Blumenthal and an anonymous referee for valuable comments on an earlier version of this paper.

(2) In this respect we focus on the role of policy advice in strengthening legitimacy of political action with regard to policy reforms. We do not address the delicate issue of legitimizing expert influence and the ambiguous role of experts in parliamentary democracy (cf. the paper by Julia von Blumenthal in this volume). However, there might a certain difference between the generation of legitimacy in a more sustainable, long-term view and more short-term, ad hoc acceptance of policy decisions.

(3) For an attempt at typifying scientific advisory bodies and systems see Glynn/Cunningham/Flanagan, 2003. partner organizations also provide policy advice on their own or support think tanks embedded in their respective environment. This may neutralize independent policy advice and hamper basic compromise on economic issues and favor joint strategies of externalizing the cost of labor market adjustment. Stalemate in social partner negotiations can be overcome by a 'shadow of hierarchy' thrown by a government that is able and willing to act unilaterally (Scharpf 1994). A strong position of non-corporatist policy advice with a high reputation in the political sphere, science, and the public can balance policy advice from the social partners.

(4) This happened to the Hamburg Institute of International Economics (HWWA, Hamburgisches Welt-Wirtschafts-Archiv) which was one of the leading institutes, but lost public funding for its research sections after a critical evaluation by the Science Council. It was replaced by ZEW. HWWA's research activities are now to a significant part taken over by a public-private partnership under the name of Hamburgisches Welt- WirtschaftsInstitut (HWWI).

(5) It is notable, finally, that the analytical work of the Alliance for Jobs was not funded by the government or the social partners but by the private Bertelsmann Foundation, the lack of public funding being a potential sign of the low priority attributed to this endeavor.

Werner Eichhorst Institute for the Study of Labor, IZA, Bonn

Ole Wintermann Bertelsmann Foundation, Gutersloh