JCMS 2008 Volume 46. Number 4. pp. 743–764
Government Learning in German and British European Policies* NILS C. BANDELOW University of Braunschweig
Abstract Why do some governments change their policies toward European integration while other national positions endure? Governments may change their policies even while the governmental parties remain the same. This article focuses on the explanatory capacities of policy learning, and thereby develops some contra-intuitive theses concerning the impact of political systems. Under some circumstances even the existence of many veto-players with opposite belief systems may increase the probability of learning.
Introduction Why do some governments change their policies toward European integration while other national positions endure? Changes of governmental programmes typically occur for two reasons. Firstly, due to a power shift between actors and, secondly, based on collective learning. Comparative politics usually focuses on power shift (Lijphart, 1999; Tsebelis, 2002; Zohlnhöfer, 2003). This article focuses on the explanatory capacities of policy learning and thereby develops some counter-intuitive theses concerning the impacts of political systems. * Thanks are due to Regine Paul, Klaus Schubert, Ulrich Widmaier and to the participants of section 13 on policy-learning of the ECPR General Conference in Budapest 2005. I also thank the anonymous referees for their extensive and insightful comments on an earlier draft of this article. Responsibility for remaining errors and omissions rests entirely with the author. The research reported in this article was generously funded by the DFG (BA 1912/1-1). © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
744
NILS C. BANDELOW
Policy learning is defined here as any enduring change in the perception of policy-related problems, beliefs and attitudes of government. Furthermore, it is considered sensible to apply the concept of policy learning only for changes based on experience and information, but not for enforced changes (Maier et al., 2003). Although this narrow definition differs from broader concepts like policy transfer, it enables us to confront policy change by learning (‘puzzling’) with external factors such as power shift (‘powering’) (Heclo, 1974, pp. 305–6). Collective government learning depends on different factors: the probability of government learning is not solely determined by the availability and type of information related to the respective policy, but is also affected by the political system (Sabatier and Weible, 2007). Thus, in this article, the correlation between power concentration within the political system and the government and the probability of collective learning will be explored. In section I, the article develops a theoretically derived analytical framework for testing the influence of political systems on collective government learning. The underlying hypotheses will then be applied to the European policy of the Kohl (section II) and Thatcher/Major governments (section III). Germany serves as an example of a consensus democracy whereas the UK can be viewed as a typical majority democracy (Lijphart, 1999). In both countries the governments, the party composition unaltered, were confronted with challenges of European integration policies. A change in government as a possible cause of political changes can be excluded during the timeframe of the research. European policy differs from other policies for different reasons. First of all, being a part of foreign policy, it lies within the responsibility of the executive branch of government. Hence, legislative veto players play distinct roles. Moreover, the general absence of party competition in German European policy seems to have some influence on the probability of learning. The described policy background for the two countries allows for an examination of the hypotheses on government learning. The case studies are based on literature analyses, background interviews and the analysis of documents and statements of both governments. The 29 background interviews were used to identify the relevant documents and test some rival interpretations that can be found in the literature. References always name the function that the interviewee had when he or she acquired the information. Additionally some biographies and autobiographies of leading actors are used to obtain information about goals and arguments. Essential information is backed up with different sources or by other literature whenever possible. © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
745
The article offers an explanation for the change of Germany’s beliefs and attitudes in central issues regarding European integration. At the same time, it explains why in the UK’s majoritarian democracy, change in some policies is less likely than in a consensus democracy. Resuming the case studies on a more abstract level, there are three independent variables that explain the probability of collective governmental learning: (1) the level of power concentration within the political system; (2) the range of belief systems of the individual governmental actors; and (3) the interaction orientation of the individual governmental actors. I. Collective Government Learning: Theoretical Considerations The concept of collective governmental learning implies that cognitive processes can provoke changes in governmental programmes. Governments can react to experiences of success and failure of previous programmes; they can take into consideration changed international conditions or take scientific advice. Policy analysis and organization research have developed various concepts and theoretical frameworks, dealing with the type of, reasons for, and possible outcome of, learning processes (for an overview see Maier et al., 2003). Both disciplines apply a distinction between first order learning, second order learning and (in some theoretical frameworks) other learning forms like deutero learning (Argyris and Schön, 1978). First order learning addresses the change of policy instruments while second order learning refers to the change of overall goals; the change or stability of core beliefs being the main difference between both concepts.As far as the general values, attitudes and perceptions of the policy problem and of the suitable solution remain the same, first order learning can be attested to. Yet, if the changes of general goals are more than instrumental and if these changes are based on experience or new information, these can be classified as second order learning. Finally, deutero learning is ‘learning to learn’. It does not change policy beliefs but political institutions that frame the policy process. This distinction between several types of learning is far from undisputable. The different frameworks still do not consequently refer to each other (Bandelow, 2006, pp. 759–60). Eventually, any application has to clarify at first how the idea of ‘general values’ or ‘core beliefs’ will be adopted. Empirically, core beliefs are more general and used to build political coalitions. In European integration policy, core beliefs name the general attitudes towards monetary integration and general ideas about the future structure of Europe. Secondary or instrumental beliefs can be seen in attitudes towards exchange rates within a monetary union or timetables for the integration process. © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
746
NILS C. BANDELOW
The following analysis will transfer hypotheses from sociological organization research to political science. In particular, actor-oriented approaches of organization research can be applied to governments, defined as core executives. In contrast to institutional definitions of governments, the term core executive refers to all actors that form the executive policy-making centre. Core executives usually do not include all members of the cabinet and can include other leading politicians like party figures or MPs (Rhodes, 1995, p. 12; Smith, 1999). Hence, new belief systems of the core executive can be based on processes that Wiesenthal (1995) described as unconventional organization learning. Wiesenthal assumes that learning processes can be started by members with divergent core beliefs (dissident behaviour). This sort of learning expects own resources for opposition actors. Transferring this idea to core executives, the following thesis emerges: changes of collective governmental core beliefs by ‘dissident’ actors are more likely, the lower the concentration of power in a political system. This thesis is counter-intuitive viewed through the lens of rational choice institutionalism. Frameworks like the veto player theory argue that higher concentrations of power increase the probability of policy change (Tsebelis, 2002). Rational choice institutionalism focuses on the realization of given preferences and is looking for power structures that hamper innovations that are pursued by governments. By contrast, policy learning focuses on the change of preferences in response to new information. While higher concentration of power allows for the realization of given preferences, it simultaneously limits the flow of information and thereby contributes to stable policy preferences. That is why in Karl Deutsch’s well-chosen definition, power is named as ‘the ability to afford not to learn’ (Deutsch, 1966, p. 111). Thus, the more power single actors hold, the less they will use new information to change their preferences. Similar hypotheses can be concluded from other theoretical approaches of organization research. Assuming, for instance, that new ideas are introduced and enforced by innovative project groups within organizations, especially ‘overlapping’ groups may cause changes in collective core beliefs (Wilkesmann, 1999, pp. 166–7). A transfer of these ideas to core executives supports the hypotheses stated with regard to dissident actors: consensus democracies with many veto players seem to fulfil better the preconditions for organizational learning than majority systems. This assumption relies on the fact that majority systems are confronted with difficulties including the governmental leader without producing internal power imbalances within a project group. Hence, sociological organization research can explain cases of political stability in political systems with few veto players and policy change within consensus democracies. This approach may explain why German European © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
747
integration policies featured diverse forms of change by collective learning which was less likely in the UK. II. Collective Government Learning in Consensus Democracies: The Example of German European Policies between 1982 and 1998 During the era of Chancellor Kohl, Germany developed a European policy network that was typical of its consensus democracy. While usually only one veto player has to be faced in Britain (i.e. the party in power, Tsebelis, 2002), there are several veto players and other influential actors in the German political system. First of all, the three coalition parties (CDU, CSU and FDP) had to agree on every main proposal. In contrast to British political parties, German parties have at least three variably independent power bases: the government, the parliamentary party and the party itself (Helms 1996, 2005; Schmidt, 2007, p. 171). These are not necessarily represented by three leading politicians. Even if the Chancellor is the formal leader of his party (this is neither constitutionally binding nor bound by precedent as in Britain), he has to cope with other powerful actors within his party. The CDU has a peripheral structure that mirrors the federalism of the political system (Schmid, 1990). The German Bundestag is much more independent than the British House of Commons and the Chairman of the Chancellor’s party is not supposed to restrict his role to that of a chief whip. Conflicts between government, the Chancellor’s party and the parliament therefore do not only result from temporary splits, but are fundamental attributes of the ‘semi-sovereign’ state (Katzenstein, 1987; Paterson, 2005). Additionally, having direct as well as indirect authority, the Ministerpräsident of the federal states importantly impact on European policies. They can use the veto rights of the Bundesrat and have a right to participate in the formulation of German proposals on EU decisions (new article 23 of the Basic Law). Leading politicians of the federal states also play major roles in their respective parties. Germany lacks an ex-ante coordination system of EU policies (Derlien, 2000; Bulmer et al., 2001, p. 269), which renders a hierarchical definition of the German policies by the head of the government more difficult. As a result, the German core executive is characterized by a very low degree of power concentration. Although party competition is a central mechanism within the German political system, it does not traditionally play an important role in EU policies (Bulmer and Paterson, 1987; Katzenstein, 1997; Kohler-Koch, 1998). Hence, the individual actors of the German government are not necessarily competitive, rather a ‘solidaristic’ interaction orientation may be expected (for the © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
748
NILS C. BANDELOW
definition of this concept see Scharpf, 1997, pp. 84–9). This is particularly the case for coalition partners interested in joint success. The absence of competitive interaction orientation should not detract from the fact that there have been different core beliefs and goals within the German government. If EU policy is classified on the basis of the given range of policy goals of important individual actors, one has to distinguish between economic and political policies. The development of German attitudes towards the economic and monetary union (EMU) serves as an example of a field with a low level of concentration of power, a solidaristic interaction orientation of the individual actors and a large range between the belief systems within the core executive. By contrast, the development of German proposals towards political integration and enlargement of the EU demonstrate developments in an area featuring a small range of individual belief systems within the core executive. During the negotiations on European monetary integration, the main conflict within the German core executive was between ‘monetaristic’ and ‘economistic’ belief systems (Moravcsik, 1998, pp. 320–1; Karama, 2001, pp. 189–95). On the one hand, leading actors like Foreign Secretary Genscher supported monetary integration for political reasons. From this monetaristic view, monetary integration was a means of securing peace in Europe and of supporting political and economic integration. On the other hand, actors like Stoltenberg, the Minister of Finance, used an economistic perspective evaluating the steps of monetary integration according to their respective economic effects (Gaddum, 1994, pp. 287–8; Dyson and Featherstone, 1999, pp. 20, 48, 278). Germany’s currency policy is usually and accurately described as an example of an about-turn. Until 1987, the national autonomy of the Bundesbank contributed to the foundations of the German political culture (Kaltenthaler, 1998, p. 27). Thus, the German government categorically rejected any step toward a monetary union. For example, it refused the proposal to authorize the private ECU in 1985 (Gaddum, 1994, pp. 316–42). However, within a very short time Germany’s attitudes changed dramatically. The starting point for this change was Genscher’s speech on 24 March 1987, leading to an about-turn within only one year. By June 1988, the German government was supportive of the appointment of the Delors Committee and also favoured the rapid creation of EMU (Dyson and Featherstone, 1999, pp. 308–421). The change in the late 1980s is an example of second order change (see Hall, 1993), as the core belief of the German government changed: the government agreed on the idea of a monetary union even including partners like Italy that were far from reaching the ‘economistic’ preconditions. Yet, the question about the nature of change remains. Was it caused by policy learning © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
749
based on new information or experience, or was there a power shift based on external events or on political strategy? There are indeed several attempts to explain the German change with regard to EMU by referring to non-learning reasons. Explaining the political U-turn with the changing international context at the end of the cold war and German unification, political scientists put the change in the German strategy down to a package deal with the French government. Yet, this interpretation contradicts the real course of the relevant events. Neither the idea of package deals nor a power shift within the core executive is adequately able to account for the change: a power-shift-explanation had to assume that economistic actors from the German finance ministry and the Bundesbank would have lost power for reasons other than new information or new arguments. This was not the case. On the contrary, the former opponents of an early economic integration were even included in the formulation of the new German proposals. This new policy did not only reflect the old monetaristic belief system; rather the German proposal for the later Stability and Growth Pact combined the main elements of both belief systems. This was only possible because both sides learned on the basis of different information: economists interpreted the progress of the single market as a fulfilment of the economic conditions for a monetary union (interviews in the German Ministry of Finance and Bundesbank). The experiences with the European Monetary System (EMS) were considered even more important, exchange rates within the system having stabilized after initial fluctuations. Between 1987 and 1992 interventions were superfluous and the EMS had proved itself as a favourable framework for economic growth. Yet, the EMS could not be continued in its hitherto form, since the liberalization of EC capital markets would no longer allow for a protection of weak currencies in future crises. These findings were later confirmed in the withdrawal of sterling from the Exchange Rate Mechanism (ERM) (section III) and substantially contributed to the change in the German proposals towards the EMU (interview German Foreign Office). Beside the specific experiences with the Single European Market, the EMS and the economic recovery, external and internal advisers also influenced the shift within Germany’s collective attitude towards EMU. Thus, politicians from other European Community states, using the contact through the European People’s Party, formulated a common position of all EPP members on EMU. Additionally, independent advisers such as the chairman of the Deutsche Bank, Herrhausen, influenced leading politicians using economic and political arguments (interview German Chancellor’s Office). To summarize, the sudden and sustained change of German collective attitudes appears to be based on second order learning. Collective government learning seized both economic and political information. The peculiarities of the © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
750
NILS C. BANDELOW
selected case substantially favoured government learning. Firstly, only the variety of the belief systems of actors within the core executive enabled the government to use all economic and political information. Accordingly, the political learning by opposition actors can be classified as ‘dissidence’, as described by Wiesenthal (1995). Secondly, the absence of party competition had a significant impact on EU policy: despite their opposing belief systems, relevant actors were interested in an arrangement that took the beliefs of other actors into consideration (solidaristic interaction orientation). As a result, collective government learning seems to be favoured by low power concentration and a solidaristic interaction orientation. The second order collective learning of the German core executive in the case of monetary integration neatly fits into the definition of unconventional organization learning. A sudden and lasting change in attitude is only likely if actors with different belief systems already belong to a core executive. If the actors share the same core beliefs it is to be expected that only first order learning occurs. An example of this form of learning is the development of German attitudes towards enlargement of the European Union and deepening of political integration after the Maastricht Treaty came into force on 1 November 1993. The German proposals for European political integration and EU enlargement were not particularly contentious within the core executive at the beginning of the 1980s. When Germany was to formulate proposals for the creation of a Single European Market, arguments about secondary questions within the government dominated. Causes of conflict resulted especially from the resistance of many departments to losing national veto power in their respective policy area. Yet, the basic approval of the political integration process was not questioned by any member of the German core executive. The whole government equally approved of the consolidation, enlargement and democratization of the European Community (Mazzucelli, 1997, p. 137). Despite the enduring character of the core of this collective belief system, secondary changes by first order learning can be observed between 1982 and 1998. Since the 1980s, the process of European integration has been characterized by simultaneous deepening and enlargement. The Single European Act (SEA), the Maastricht Treaty and the Amsterdam Treaty were all agreed by the Kohl government. At the same time, the Community has been enlarged significantly. Germany’s strategy of combining deepening and enlargement was not supported by all Member States to the same extent and it became less enforceable. In this context, despite keeping their substantial core belief regarding political integration, German officials strategically separated the goals of deepening and enlargement and developed new instruments in order to reach them. © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
751
The proposal of variable integration constitutes a typical example of first order learning with regard to the German core executive. Various models for ‘step by step’ integration (Willy Brandt), ‘differentiated integration’ or ‘variable geometry’ (Delors) have already been discussed since 1974. Yet, these proposals did not prominently impact on the German core executive until 1988. According to the project group thesis of organization research, small groups within the German core executive developed their own proposals. The first example, a proposal for ‘concentric circles’ presented by two civil servants from the German Chancellor’s Office (Mertes and Prill) in 1988, was actually supported by leading actors of the core executive – despite official denials to the contrary (Giering, 1997, pp. 13–18; Mertes, 2001, p. 4). The proposal anticipated the later German position on ‘opt out’ solutions at the Maastricht negotiations. The presentation by a project group enabled the core executive to wait for external reactions, before deciding on an official statement. Retrospectively, the proposal for a ‘Europe of concentric circles’ was the first step towards a steady paradigm change within the German government, which gave up the idea of a homogeneous community. The second step was the proposal for ‘asymmetrical integration’. Like the Mertes and Prill proposal, this plan was compiled by a project group and presented in 1994 as the unofficial idea of individual actors. The responsibility was taken by Schäuble and Lamers (Deubner, 1995). With Schäuble, then chairman of the parliamentary party, the group included one of the most powerful members of the German core executive. The proposal to build a ‘core Europe’ provoked sharp criticism, mainly due to the fact that the paper excluded a founding member of the European Communities, Italy, from the European ‘core’. As a result of the criticism, the proposal was excluded from official statements, though the German government later presented the main ideas of the paper in a modified version at the negotiations for the Amsterdam Treaty (Interview German MP). These examples illustrate that in Germany’s European policy, new proposals can be promoted by actors beyond the official government. These external presentations are favoured by the low level of power concentration within the German political system. Another important finding is that veto players within the German political system did not block the learning process initiated by ‘dissident’ actors or project groups within the core executive. On the contrary: the example of the monetary integration policy demonstrates that the different core beliefs of actors within the core executive actually significantly favoured the second order learning process. Another main condition for both types of government learning in Germany was the specific tabooing of the European policy as a subject of party competition. This absence of competition contributed to a solidaristic interaction orientation of © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
752
NILS C. BANDELOW
the actors involved. Interestingly, the specific dilemma of the German political system caused by the contradiction between party competition and consensus democracy (Lehmbruch, 2000) was not relevant in the case described. In the following, the example of British European policies illustrates how power concentration and especially competitive interaction orientation of the actors involved can hamper collective government learning. III. Obstacles for Collective Government Learning in Majority Democracies: The Example of British European Policies between 1979 and 1997 The aims of the British European policy traditionally differ from those of most other EU Member States. Within both main parties and also other political elites (including prominent journalists) eurosceptic beliefs and attitudes are widespread (see for example Buller, 2000; Bulmer, 1992; George 1998; Diez, 1999; Gowland and Turner, 2000; Young, 2000). Although the respective roles of the leading parties as (weak) supporters or (strong) opponents of European integration have changed over time, European policy has always been a central theme of competition between and within British parties. Party competition and the absence of any tradition of consensus democracy led to competitive interaction orientations of all actors involved in the formulation of British proposals towards European integration. The British political system indirectly influences the actors’ interaction orientation and also directly affects the composition of the core executive. The (usually) high level of power concentration results in an ex-ante coordination of every statement with the position of the Prime Minister (Bulmer and Burch, 2005, p. 868; Kassim, 2000). This task is performed by an institutional triangle of the Foreign Office (European Communities Department), the UK Permanent Representation in Brussels and especially the European Secretariat of the Cabinet Office (interview UK Cabinet Office). The ex-ante coordination aims at reducing the influence of ‘dissident actors’ within the core executive. Therefore, in theory, a high level of power concentration within the British core executive can be assumed for European policy. This assumption, however, has to be qualified by a precise analysis as will be shown in the following. European policy did not play a significant role in internal conflicts within the first Thatcher government. The aim of other Member States to transfer further national competencies to Brussels was not yet foreseeable in 1979. During the 18 years of the Conservative Party’s period in office, the British core executive was confronted with a large amount of new information and © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
753
proposals for further integration, as it had to react to proposals for economic, political and monetary integration. By doing so, the British actors had to take a lot of information into account. For example, in the area of monetary integration this information contained the respective relevant data of the ERM (Exchange Rate Mechanism). Additionally, the British core executive was continuously confronted with changing strategies of other Member States, the European Commission and other actors. The regime change in Eastern Europe in 1989 and the German unification of 1990 led to further changes that were relevant for European integration policy. Other challenges concerned special problems – for example, the British Budget Question (BBQ) and European policies to react to the BSE crisis. The European Community policies of the Thatcher government were first of all concentrated on the BBQ. During the conflict about the British EC contribution, a tactical change of the British government could be observed. Margaret Thatcher and Foreign Secretary Geoffrey Howe prepared a British proposal that can be interpreted as a kind of learning at first sight (Thatcher, 1993, pp. 541–3). Despite the proposal being the most pro-European statement of the whole Thatcher era, it lacked elements of political integration (Thatcher, 1984). The proposal was not based on new information. On the contrary: Thatcher generally excluded the civil service and the parliament, right until the publication of the proposal in the JCMS made the idea known to the British public (interview Foreign Office). Unsurprisingly, the British proposal did not reach its aims: whilst it aimed to shape the political debate, it became no more than an agenda setter for the further intergovernmental debate. Hence, the proposal indicated a short-dated tactical change; whereas the core of the British European policy remained unaltered as Thatcher was not prepared to give up any political sovereignty. In the context of the agreement of Fontainebleau, the British government had to react to demands for the creation of a single market accompanied by measures of political integration. The single market as such was endorsed by the British government during the whole negotiations between 1984 and 1987. It also supported the evolution of the CFSP and the strengthening of the Court of Justice. However, in particular the approval of an expansion of majority decisions in the Council was contentious. Thatcher and her Chancellor, Nigel Lawson, on the one hand rejected any expansion, whereas, on the other hand, Howe and the Head of the European Secretariat of the Cabinet Office, David Williamson, were prepared to compromise. At this time, the Prime Minister had strengthened her internal power; nonetheless Howe and Williamson succeeded in this struggle. Eventually, the British government agreed to the SEA including the accompanying measures of political integration. Different explanations for this result are offered by the protagonists and © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
754
NILS C. BANDELOW
observers (Cockfield, 1994, pp. 135–6; Thatcher, 1993, p. 729). The best explanation for this contradiction is that Thatcher was not able to see the importance of political integration within the Single European Act (SEA). Eurosceptics later blamed Williamson for not having informed the Prime Minister adequately and having misled Thatcher about the scope of the agreement. It is impossible to prove the substance of this story. But even Thatcher’s opponents stressed the problem of information: ‘[Thatcher] was surrounded by those sharing her own prejudices and whose views of the world outside were as through a mirror darkly. As she became more and more dictatorial, officials and others increasingly became afraid of telling her the plain, unvarnished truth’ (Cockfield, 1994, p. 140). In fact, the British government did not use the chance to deliver its own proposals for political integration without giving up its veto rights. It did not use information on strategies of other Member States to find compromises suitable for the belief system of a majority of governments. The lack of information about the SEA at the head of the British government appears astonishing for further reasons: the Commission’s SEA proposal was prepared by the British (Conservative) Commissioner Cockfield. Though Thatcher expected him to report to her and to follow her instructions, the Commissioner had no accountability to the UK government and preferred to interpret his role as it was formally fixed. He discussed some ideas with other members of the British government than Thatcher, in fact with more proEuropean actors of the Foreign Office who shared most of Cockfield’s beliefs. They did not see any way to win the Prime Minister’s support and therefore did not report all details to Thatcher (interview British Cabinet Office; Thatcher, 1993, p. 547; Cockfield, 1994, pp. 55–7; Howe, 1995, pp. 457–535). The interruption of the information flow within the British core executive illustrates an often neglected effect of the UK’s majoritarian democracy: dissidents usually have no chance of reaching their political goals, except when they strategically withhold information from the Prime Minister. These limitations in the information flow importantly decrease the probability of collective learning. Between 1983 and 1989, the British political system worked as a typical majoritarian system. Hence, changes in the collective governmental belief system could only be provoked by changes in the Prime Minister’s belief system. Therefore, the core of the euro-political beliefs remained constant – in spite of the essential change of external challenges. During this period, learning was restrained by power concentration. After the adoption of the SEA, Britain’s membership of the ERM was at the centre of British European policy. ERM membership was as controversial in the core executive as political integration. While Thatcher was determined to oppose membership, Chancellor Lawson and Foreign Secretary Howe © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
755
were in favour of joining. Despite rising internal conflicts, Thatcher was able to resist her internal opponents in the first years with the exemption of one year. Between March 1987 and March 1988, Chancellor Lawson followed an unofficial strategy of ‘shadowing the Deutschmark’. Thatcher never agreed to this strategy and afterwards claimed not to have known about it. After being informed by journalists in an interview, Thatcher stopped Lawson’s policy, which had set an upper limit value of the pound of 3 DM, making Britain a member of the ERM in all but name (interview British Treasury, see also Lawson, 1992, pp. 494–5; Thatcher, 1993, pp. 668–710). The UK finally joined the ERM officially in October 1990, a fact which can be interpreted as a major political change. Yet, it cannot be explained by government learning. Nobody within the British core executive changed his or her belief system: Thatcher opposed the ERM as she did before. Until 1989 she was able to resist Howe and Lawson who wanted to join the ERM. After Howe (June 1989) and Lawson (October 1989) were forced to leave the cabinet, the new Foreign Secretary Hurd and Chancellor Major adopted the policy positions of their predecessors. Nothing but the loss of internal power finally forced Thatcher to give way to the demands of her internal opponents. There were no policy-related arguments or new information that could have justified the time of Britain’s entry and the official exchange rate. The exchange rate was set by Thatcher without consulting any economic advice. The ERM membership was the result of a sudden loss of power by the Prime Minister that finally ended with her replacement (interview British Treasury, see also Lawson, 1992, pp. 929–33; Thatcher, 1993, p. 712). The successive formulation of proposals for the later Maastricht Treaty did not face any change of British core beliefs either. The UK positioned itself strongly against both EMU and the political union. In all phases of negotiation, actors within the core executive rejected the Social Charter and the strengthening of the European Parliament. Although the UK rejected nearly every section of the later Maastricht Treaty, the British government barely developed alternative proposals in the negotiation process. Only two counter drafts to the EMU seem to indicate collective learning at first glance. The first draft of ‘Competing Currencies’ was developed by Middleton (Treasury) and presented by Chancellor Lawson in 1989 (Dyson and Featherstone, 1999, p. 565). Not only was the proposal rejected by the other Member States, even the British core executive never took it seriously. It was always considered an obvious attempt to prevent the EMU without any chance of realization (Major, 2000, p. 139). For the same reason, the second British proposal ‘Hard ECU’ from June 1990 cannot be interpreted as second order learning (Dyson and Featherstone, 1999, pp. 612–43; Forster, 1999, pp. 52–4). In contrast to the German drafts for variable integration, the British monetary proposals © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
756
NILS C. BANDELOW
never aimed to test the reactions of other Member States in order to develop feasible compromises. They were used without any clear goal and did not change the collective attitudes of the British core executive. Hence, there was no lasting second order learning of the British core executive at the preparation of the Maastricht Treaty. Even if the monetary proposals were not lasting, they may be interpreted as first order (tactical) collective learning, but in no way as second order learning. While learning, as previously shown, did not have a lasting influence on the British European policy, a power shift significantly changed the governmental tactics. Immediately before the Maastricht government conferences, John Major became the new head of the British government. Major did personally not support the EMU, but in contrast to Thatcher, he strove to improve the British reputation within the European Union. He was especially interested in courting the German Christian Democrats and started his premiership with a visit to Bonn where he promised, to bring ‘Britain where we belong, at the heart of Europe’ (Turner, 2000, p. 172). But Major’s role as the new Conservative leader had not been established by the beginning of the Maastricht conferences and he was unable to change the composition of the core executive. During the government conference for the EMU, the UK was represented by the eurosceptic Chancellor Norman Lamont. In addition, followers of Thatcher exercised pressure on the government. Hence, the preparation of British positions was dominated substantially by internal struggles for power within the government. The weakening of the radical eurosceptics by the fall of Thatcher at least caused a tactical change in that the British government abandoned the strategy of complete blockade of the Maastricht Treaty and the EMU. Agreement of an opt-out solution in Maastricht can be interpreted as a result of a short-term tactical change caused by the election of the new Prime Minister (interview British Cabinet Office). How little the British European policy reacted to new information becomes similarly obvious during the episode that led to the ejection of sterling from the ERM only two years after its entry. While Thatcher’s absolute power blocked all information that could have supported Britain’s entry by 1989, the new conflict between the eurosceptic Chancellor Lamont and the more moderate Major blocked collective learning to react to the crises of the pound within the ERM. In particular the high interest rates in Germany put increasing pressure on sterling. There are conflicting accounts by the actors involved on the precise reaction of the government to this pressure in autumn 1992 (Heseltine, 2000, pp. 431–3; Lamont 2000, pp. 250, 521; Seldon, 1997, p. 465). All reports indicate that the government was not prepared for the final result – the involuntary exit of sterling on ‘Black Wednesday’ (16 September 1992) from the ERM. During this episode, © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
757
Major’s eroding power led to a blockade of any new proposal by one of the competing factions within the British core executive (Howe, 1995, p. 461; Turner, 2000, pp. 26–7). ‘Black Wednesday’ was the beginning of a special situation that characterized the Major government prospectively. This episode between 1992 and 1997 shows that the British majoritarian political system can indeed imply veto players as competing factions within the parliamentary party of the government. They not only stopped the implementation of each governmental proposal, but also blocked collective learning. A precondition for the emergence of intra-parliamentary veto players is a lack of internal authority on the part of the Prime Minister. This was the case during the Major government. Before Major became leader of the Conservative Party and Prime Minister in November 1990, he had been part of the government for only three years. Not having been a leading figure within the party before, Major owed his quick promotion to Thatcher alone. Major himself only had a very small number of personal supporters within the parliamentary party. Although Major continued most of Thatcher’s eurosceptic strategies and at least prevented British entry to the EMU and to the Social Charter, the number of his enemies within the Conservative Party rose for several reasons (interview British MP; Peele, 1997, p. 100; Gowland and Turner, 2000, pp. 194–5). One of these reasons was the loss of seats held by pro-Europeans in elections, which increased the eurosceptic ‘veto’ position against Major’s policy objectives, such as the ratification of European treaties. Even more damaging for the Prime Minister was the threat that he could be brought down by his own parliamentary party, like his two predecessors. As a result of the frequent challenges to the party leadership between 1975 and 1977 and changed election rules within the Conservative Party, even minorities within the parliamentary party could bring down the Prime Minister. As a result, the competing factions blocked each other, so that a change of political programme was prevented and the government was unable to react to the new challenges and new information with strategic changes. Thus, this episode can be classified as idiosyncratic. At the beginning of the period of office of both Prime Ministers, strategic change appeared possible. However, actual changes were usually based on power shifts within the core executive and rarely on collective learning. Interestingly enough, the absence of power concentration within the British majority democracy provoked change, because it enhanced the probability of a power shift. On the other hand, the often underestimated power of the backbenchers increased to such an extent that the competing factions could establish themselves as actual veto players. These veto players, contrary to the German consensus democracy, did not contribute to collective learning. Acting with a competitive interaction © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
758
NILS C. BANDELOW
orientation, they hindered not only the conversion of proposals into policies, but also the change of proposals by collective learning. This review of the development of British positions on central issues of European integration shows first of all that dissident behaviour may be significant in the British core executive. In certain periods dissident actors were able to gain influence and therefore change certain policies. Yet, the probability for collective learning processes within the core executive provoked by dissident actors is very low. Project groups usually play an important role at the beginning of a new Prime Minister’s term of office. Sub-committees of the cabinet were used especially to prepare new policy proposals (Stuart, 1998, p. 372). During the period of the strongest concentration of power (1983–89), the subcommittees only endorsed Thatcher’s goals. A significant problem of these sub-committees was their total domination by the Prime Minister, who determined the composition, the agenda and the results. Hence, whilst British majority democracy seems to be generally open to policy changes, these changes are seldom based on collective learning as in the case of Germany. Conclusion: Veto Players, Belief System Range, Interaction Orientation and Collective Learning The case studies presented demonstrate that government learning is possible. Therefore, the probability of change not only depends on the probability of the realization of government proposals, but also on the probability of changes to the government proposals themselves. The results further show that political systems may influence the probability of collective government learning. To specify this argument, the two case studies can each be split up and clustered in four types of political situations and probabilities of collective learning. The types can be arranged on a continuum of the probability of political change by collective government learning (Figure 1). Thus, the European policy of the Major government (1990–97) can be interpreted as an example of blockades of collective learning in a majority democracy. Apparently, intra-parliamentary veto players can appear in the UK. The British political culture of conflict contributes to the veto players’ influence, reducing the probability of collective learning. Indeed, in this period, individual learning on the part of the governmental leader was more likely than during the Thatcher era. In contrast to Thatcher, Major steadily changed his personal attitudes towards European policy (Turner, 2000, p. 172). For example, he changed his beliefs and attitudes according to experiences with the reactions © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
759
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
Figure 1: Continuum of Collective Government Learning
competitive veto
one veto player:
solidaristic veto
solidaristic veto players
players: no
only first order
players with similar
with different core
core beliefs:
beliefs: second order
continuous first order
learning (D monetary
learning (D political
integration)
collective learning learning (UK 1979– (UK 1990–97)
90)
integration)
unlikely
collective learning
likely
Source: Author’s own data.
of other governments during the BSE crisis, however, Major’s individual learning did not lead to collective government learning. Generalizing this finding, it could be assumed that majority systems in periods with low power concentration lead to the lowest probability of collective government learning. The Thatcher government serves as an example of a ‘normally’ functioning majority democracy. Between 1983 and 1989 in particular, there were actually no blockades against majority rule. Interestingly, the example of European policy (in contrast to other policies) indicates that strong power concentration does not necessarily lead to a change in broad policy. A comparison of individual learning by the powerful Prime Minister Thatcher and her far less powerful successor Major confirms Deutsch’s definition of power as the ‘ability to talk instead of listen’ (Deutsch, 1966, p. 111). Political learning in this episode has only occurred as tactical learning (first order learning). The high level of power concentration reduced the circulation and political implementation of new information, and thus hampered proposals for changes in policy by collective learning. While the British European policy produces two systematic cases by a chronological division, for German European policy two issue-related cases can be distinguished. Both German cases, typically for consensus democracies, differ from the British examples in their rather solidaristic interaction orientation. However, generally, a solidaristic interaction orientation is rather © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
760
NILS C. BANDELOW
unusual for Germany. In most policy areas the party competition causes a competitive interaction orientation (Lehmbruch, 2000). Both German cases are characterized by a low level of power concentration. Most decisions needed the consent of several veto players. The German cases can be distinguished by the respective range of the veto players’ belief systems. Regarding issues of political integration, German actors featured only a small range of policy goals. This example leads to the assumption that the similarity of the individual core beliefs within the core executive renders second order learning unlikely. Yet, the case study provided examples for first order learning. In contrast to the British case, German first order learning resulted from collective processes. The actors readily exchanged information; therefore, different actors contributed to the development of new policy proposals. While Thatcher was not prepared to consider new policy-related information when she reached the height of her power, the German political system secured the possibility of collective learning during all periods. German attitudes towards monetary integration were much more contentious. From the case study, one can assume that even this wide range of beliefs promoted second order learning. As with political integration, the German core executive showed a high level of collective learning. The existence of actors with different beliefs within the core executive even before 1988 caused a process of unconventional organizational learning. The different belief systems of the individual actors involved favoured the inclusion of both economic information (for example, experiences with the EMS and the single market) and political information (for example, information on changes to other Member States’ goals and proposals) in a collective learning process. Therefore, the German policies toward monetary integration were the only example of a change in governmental collective core beliefs. The case examples presented here raise further questions. First of all, the four cases delineated do not fulfil the strict methodological rules of positivistic causal-explanatory methods: three (dichotomous) explanatory variables, as they are presented here, would have required at least eight cases to prove their separate influence on the dependent variable. In order to reduce this problem, similar case studies can be explored. For example, European integration caused a paradigm change of British environmental policy (Jordan, 2003). This change was not based on collective learning within the British core executive but caused by the external force of the EU. There are similar findings for other policy areas (Bulmer and Burch, 2005, pp. 868–72). They confirm the argument that European integration caused some major changes to UK government and policies. Yet, until the end of the Conservative era in 1997, European pressure only caused minor learning whereas, more generally, EU business was incorporated into the traditional pattern of the © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
761
Westminster model (Bulmer and Burch, 2005, p. 876). However, the period of the Blair government appeals for the inclusion of another aspect. In contrast to the Conservatives, Blair took an active approach to Europeanization. He used external pressure to enforce domestic reforms and thereby developed a pro-active use of EU arenas (Bulmer and Burch, 2005, p. 865). Further research, eventually, might focus on the effect of party government on the probability of learning in different institutional settings. Correspondence: Nils C. Bandelow University of Braunschweig Bienroder Weg 97 D-38106 Braunschweig, Germany email
[email protected]
References Argyris, C. and Schön, D.A. (1978) Organizational Learning. A Theory of Action Perspective (Reading, MA: Addison-Wesley). Bandelow, N. (2006) ‘Advocacy Coalitions, Policy-Oriented Learning and LongTerm Change in Genetic Engineering Policy: An Interpretist View’. German Policy Studies, Vol. 3, No. 4, pp. 747–805. Buller, J. (2000) National Statecraft and European Integration. The Conservative Government and the European Union, 1979–97 (London: Pinter). Bulmer, S.J. (1992) ‘Britain and European Integration: Of Sovereignty, Slow Adaptation, and Semi Detachment’. In George, S. (ed.) Britain and the European Community. The Politics of Semi-Detachment (Oxford: Clarendon Press), pp. 1–29. Bulmer, S.J. and Burch, M. (2005) ‘The Europeanization of UK Government: From Quiet Revolution to Explicit Step Change?’ Public Administration, Vol. 83, No. 4, pp. 861–90. Bulmer, S.J. and Paterson, W. (1987) The Federal Republic of Germany and the European Community (London: Allen & Unwin). Bulmer, S.J., Maurer, A. and Paterson, W. (2001) ‘Das Entscheidungs- und Koordinationssystem deutscher Europapolitik: Hindernis für eine neue Politik?’ In Schneider, H., Jopp, M. and Schmalz, U. (eds) Eine neue deutsche Europapolitik? (Bonn: Europa Union Verlag). Cockfield, F.A. (1994) The European Union. Creating the Single Market (London: Wiley Chancery Law). Derlien, H.U. (2000) ‘Germany. Failing Successfully?’ In Kassim, H., Peters, B.G. and Wright, V. (eds) The National Co-ordination of EU Policy. The Domestic Level (Oxford: Oxford UP), pp. 54–78. Deubner, C. (1995) Deutsche Europapolitik: Von Maastricht nach Kerneuropa? (Baden-Baden: Nomos). © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
762
NILS C. BANDELOW
Deutsch, K.W. (1966) The Nerves of Government. Models of Political Communication and Control (New York: The Free Press). Diez, T. (1999) Die EU lesen. Diskursive Knotenpunkte in der britischen Europadebatte (Opladen: Leske + Budrich). Dyson, K. and Featherstone, K. (1999) The Road to Maastricht. Negotiating Economic and Monetary Union (Oxford: Oxford UP). Forster, A. (1999) Britain and the Maastricht Negotiations (Houndmills: Macmillan). Forster, A. (2002) ‘Anti-Europeans, Anti-Marketeers and Eurosceptics: The Evolution and Influence of Labour and Conservative Opposition to Europe’. The Political Quarterly, Vol. 73, No. 3, pp. 299–308. Gaddum, E. (1994) Die deutsche Europapolitik in den 80er Jahren (Paderborn: Schöningh). George, S. (1998) An Awkward Partner. Britain in the European Community (Oxford: Oxford UP). Giering, C. (1997) Flexibilisierungskonzepte für Europa (Munich: CAP). Gowland, D. and Turner, A. (2000) Reluctant Europeans. Britain and European Integration 1945–98 (Harlow: Longman). Hall, P.A. (1993) ‘Policy Paradigms, Social Learning, and the State’. Comparative Politics, Vol. 25, No. 3, pp. 275–96. Heclo, H. (1974) Modern Social Politics in Britain and Sweden (New Haven, CT: Yale UP). Helms, L. (1996) ‘Executive Leadership in Parliamentary Democracies: The British Prime Minister and the German Chancellor Compared’. German Politics, Vol. 5, No. 1, pp. 101–20. Helms, L. (2005) Presidents, Prime Ministers, and Chancellors. Executive Leadership in Western Democracies (Basingstoke: Palgrave Macmillan). Heseltine, M. (2000) Life in the Jungle. My Autobiography (London: Coronet). Howe, G. (1995) Conflict of Loyalty (London: Pan Books). Jordan, A. (2003) ‘The Europeanization of National Government and Policy: A Departmental Perspective’. British Journal of Political Science, Vol. 33, No. 2, pp. 261–82. Kaltenthaler, K. (1998) Germany and the Politics of Europe’s Money (Durham: Duke UP). Karama, M. (2001) Struktur und Wandel der Legitimationsideen deutscher Europapolitik (Bonn: Europa Union). Kassim, H. (2000) ‘The United Kingdom’. In Kassim, H., Peters, B.G. and Wright, V. (eds) The National Co-ordination of EU Policy. The Domestic Level (Oxford: Oxford UP), pp. 22–53. Katzenstein, P. (1987) Policy and Politics in Western Germany: The Growth of a Semisovereign State (Philadelphia: Temple UP). Katzenstein, P. (1997) ‘United Germany in an Integrating Europe’. In Katzenstein, P. (ed.) Tamed Power. Germany in Europe (Ithaca: Cornell UP). Kohler-Koch, B. (1998) ‘Bundeskanzler Kohl – Baumeister Europas?’ In Wewer, G. (ed.) Bilanz der Ära Kohl (Opladen: Leske + Budrich), pp. 283–311. © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
GOVERNMENT LEARNING IN GERMAN AND BRITISH EUROPEAN POLICIES
763
Lamont, N. (2000) In Office (London: Warner Books). Lawson, N. (1992) The View from No. 11. Memoirs of a Tory Radical (London: Doubleday). Lehmbruch, G. (2000) Parteienwettbewerb im Bundesstaat (Wiesbaden: Westdeutscher Verlag). Lijphart, A. (1999) Patterns of Democracy. Government Forms and Performance in Thirty-Six Countries (New Haven: Yale UP). Maier, M.L., Nullmeier, F. and Pritzlaff, T. (2003) Politik als Lernprozess? (Opladen: Leske + Budrich). Major, J. (2000) John Major. The Autobiography (London: HarperCollins). Mazzucelli, C. (1997) France and Germany at Maastricht. Politics and Negotiations to Create the European Union (New York: Garland). Mertes, M. (2001) ‘The Role of the Federal Chancellor during the Process of German Unification 1989/90’. Lecture delivered at the German Historical Institute, Washington, DC, 3 October 2001. Available at: «http://www.ghi-dc.org/mertes.pdf». Moravcsik, A. (1998) The Choice for Europe. Social Purpose and State Power from Messina to Maastricht (London: UCL). Paterson, W.E. (2005) ‘European Policy-making: Between Associated Sovereignty and Semisovereignty’. In Green, S. and Paterson, W.E. (eds) Governance in Contemporary Germany. The Semisovereign State Revisited (Cambridge: Cambridge UP), pp. 261–82. Peele, H. (1997) ‘Political Parties’. In Dunleavy, P., Heffernan, R., Cowley, P., Hay, C. (eds) Developments in British Politics 5 (Basingstoke: Macmillan), pp. 89– 109. Rhodes, R.A.W. (1995) ‘From Prime Ministerial Power to Core Executive’. In Rhodes, R.A.W. and Dunleavy, P. (eds) Prime Minister, Cabinet and Core Executive (New York: St. Martin’s Press), pp. 11–37. Sabatier, P.A. and Weible, C.M. (2007) ‘The Advocacy Coalition Framework. Innovations and Clarifiactions’. In Sabatier, P.A. (ed.) Theories of the Policy Process (Boulder, CO: Westview), pp. 189–220. Scharpf, F.W. (1997) Games Real Actors Play. Actor-Centered Institutionalism in Policy Research (Boulder, CO: Westview). Schmid, J. (1990) Die CDU (Opladen: Leske + Budrich). Schmidt, M.G. (2007) Das politische System Deutschlands (München: Beck). Seldon, A. (1997) Major: A Political Life (London: Weidenfeld & Nicolson). Smith, M.J. (1999) The Core Executive in Britain (Basingstoke: Macmillan). Stuart, M. (1998) Douglas Hurd. The Public Servant. An Authorized Biography (Edinburgh: Mainstream). Thatcher, M. (1984) ‘Europe: The Future’. JCMS, Vol. 23, No. 1, pp. 74–81. Thatcher, M. (1993) The Downing Street Years 1979–90 (New York: Harper Perennial). Tsebelis, G. (2002) Veto Players. How Political Institutions Work (Princeton, NJ: Princeton UP). Turner, J. (2000) The Tories and Europe (Manchester: Manchester UP). © 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd
764
NILS C. BANDELOW
Wiesenthal, H. (1995) ‘Konventionelles und unkonventionelles Organisationslernen’. Zeitschrift für Soziologie, Vol. 24, No. 2, pp. 137–55. Wilkesmann, U. (1999) Lernen in Organisationen (Frankfurt: Campus). Young, J.W. (2000) Britain and European Unity 1945–99 (Houndmills: Macmillan). Zohlnhöfer, R. (2003) ‘Partisan Politics, Party Competition and Veto Players: German Economic Policy in the Kohl Era’. Journal of Public Policy, Vol. 23, No. 2, pp. 123–56.
© 2008 The Author(s) Journal compilation © 2008 Blackwell Publishing Ltd