How Greens govern. The experience of the Green party in Länder governments in the Federal German Republic.

How Greens govern. The experience of the Green party in Länder governments in the Federal German Republic. Geoffrey K. Roberts (Department of Government, University of Manchester, UK) `They thought they were in power, but really they were only in government' [Kurt Tucholsky, cited in Kuhnert, 1998:56] Introduction The German Green party developed from the conjunction of several influences in the late nineteen-seventies. Some of its early members had been participants in the `1968' movement; some, disillusioned by that party's static politics, came over from the Social Democratic Party (SPD); some had been associated with the citizen initiatives (Bürgerinitiativen) of the early nineteen-seventies; some were `conservative' environmentalists, who later formed their own right-wing environmental party: the ÖDP (Ökologisch-Demokratische Partei) round Grühl, the former Christian Democratic Member of the Bundestag; some had been in the forefront of the anti-nuclear power protests of the nineteen-seventies. Following unsuccessful electoral campaigns in the Lower Saxony and Hamburg Land elections in 1978, (1) a Green party won seats in the 1979 Bremen Land election and - even more impressively, given the size and semi-rural structure of that Land - in the 1980 Baden-Württemberg Land election. A decision was made to form a federal party organisation in 1980, to enable the Greens to compete in the Bundestag election that year. The publicity their campaign received was out of all proportion to their very disappointing result (1.5 percent). Further successes in Land elections were achieved, however: in 1981 in Berlin (7.2 percent); in 1982 in Hamburg (7.7 percent), in Lower Saxony (6.5 percent) and in Hesse (8.0 percent). In the 1983 Bundestag election the Greens won seats for the first time. Since then, they have remained in the Bundestag apart from the period 1990-94, (2) entered the European Parliament and obtained seats in every West German Land parliament (and, in various party alliances, briefly after reunification in some East German Land parliaments also). The West German Greens merged with the East German Green party immediately after the Bundestag election of December 1990, and in 1993 conducted a further merger, with Alliance `90, the party formed from the various citizen movement groupings in East Germany. (3) They have been partners in 13 coalition governments at Land level, and since 1998 have been junior coalition partner in the SPD-led federal governing coalition. (4) This paper examines the experiences of the German Green party in Land-level coalitions. A listing of these coalitions is provided as an Appendix. The paper will focus upon the contexts, processes, and constraints affecting the processes, of coalition formation, and on factors accounting for the fate of those coalitions (whether terminated before an election, terminated by an election, or continued after an election). A concluding section will attempt to identify some comparative generalisations which may apply to Green participation in coalitions at Land level. 1

The familiar theoretical statements concerning coalition formation, regarding minimal- and minimum-winning coalitions and ideological proximity, apply as much to Land-level coalition formation in Germany as to federal coalition formation. In both cases, the type of coalition formed (consisting usually of only two, or at most three, parties) and the type of party system in existence differentiate the German cases from, say, cases of coalition formation in Italy, Fourth Republic France or Belgium, where a much larger number of parties have seats in the legislature, where the number of politically-feasible coalition options is greater, and where coalitions frequently consist of four or more parties [Roberts, 2000:197-8] However, the distinction between minimum-winning and minimal-winning coalitions, or even between these and over-size coalitions, is not of great significance in analysing and understanding the reasons for the creation of Land-level coalitions involving the Greens. There may well be interactions between the federal and Land levels of government which have to be taken into account in either case: a federal coalition may form because of successful experimentation with such a coalition at Land level, or a Land coalition may be difficult to create because it would be incongruent with the federal coalition. In the case of coalitions involving the Green party, there is an additional factor involved: the difficulty which the Green party has in accepting the compromises and bargaining which are not only part and parcel of the coalition formation process itself, but also a necessary element of the day-to-day business of governing by coalition, and which make coalitions involving the Green party potentially more fragile than other coalition patterns. One can assert that the Green party is the only significant party in contemporary Germany where the idea of being partner in a coalition is a contested issue (as contrasted to the issue of which of several coalition options to pursue). Strom's typology of parties as vote-seeking, office-seeking or policy-seeking is very relevant here, since the emphasis which the Green party places upon policy, as a major part of its party identity, may hamper its efforts in relation to vote-seeking or holding office [Rüdig, 2000:48]. Certainly in the early years of the party's history, any discussion of whether or not to make the party available as a coalition partner could arouse considerable hostility within the party, especially (but not exclusively) from the `fundamentalist' faction. The party s federal congresses at Duisburg (1983), Hamburg (1984) and Hagen (1985) all demonstrated this, though by 1985 there was a greater readiness to consider all forms of parliamentary participation, ranging from opposition to government, though seeking power at any price was explicitly rejected [Raschke, 1993:902, 904-6]. As will be seen below, support within some Länder parties was more positive in relation to the coalition issue. By 1987, there was even a suggestion that a coalition with the CDU might one day be feasible, though not at that time [Raschke, 1993:911]. Co-operation with the CDU at local council level has occurred quite frequently [Hoffmann, 1997]. It is important also to draw attention at this stage to the time dimension in any discussion of coalitions, but especially those involving the Greens. This is important in at least three ways. First, coalitions are not discreet, isolated instances. They affect future potential coalitions, and are themselves affected by any previous coalition experience of the potential or actual coalition partners [Pridham, 1986:20]. Second, the parties themselves undergo a learning process over time. The pioneering experiences of the Greens in the first Hesse red-green coalition affected the whole party and its attitudes to coalition participation, as well as the Hesse Land party in 2

particular. Thirdly, the party and the party system undergo change and development over time additional to the changes produced by a `coalition learning' process. The SPD of the nineteeneighties was not the same as the SPD today under Schröder's leadership. The Greens today are a more pragmatic party and more liberated from the constraints of fundamentalist factions than they were up to 1990. All this does mean that any comparisons drawn between aspects of coalition experience of the Greens in Land governments at different time periods must be heavily qualified by such temporal factors. The coalition formation process The Greens have joined in governing coalitions at one time or another in nine of the sixteen Länder. In three Länder they have been partners in a coalition for more than one legislative period: Hesse three times (two of which were consecutive); Schleswig-Holstein and North Rhine-Westphalia each twice (in both these Länder these were consecutive terms). This gives a total of 13 cases of Green participation in Land coalitions. This section will focus upon the coalition formation process, drawing upon those 13 cases. The most significant of the cases of coalition formation involving the Green party was undoubtedly the creation of the first red-green coalition in Hesse, in December 1985, because it was a `first', and because it was formed at a time when the fundamentalist opponents of coalitions were still potent within the party. The Greens had agreed to support a minority SPD government in June 1984 on the basis of a formal written agreement, and the experiences of co-operation and discussion between the two parties over an eighteen-month period encouraged Green party politicians to progress towards the formation of a coalition. This was not an easy task, mainly because of internal opposition to a coalition within the federal and the Land Green party organisations. Negotiations commenced following a Land party congress of the Greens in October 1983 at which 80 percent of delegates voted to support negotiations with the SPD. In November 1983 a federal congress of the Greens heatedly debated the wisdom of conducting coalition negotiations in Hesse, and in January 1984 the federal Executive Committee of the Greens wrote to all party members in Hesse criticising such negotiations [Raschke, 1993:901-2]. Once negotiations had been completed between the Land party teams, and decisions had been made concerning allocation of ministerial portfolios, the Hesse Greens had to get their nominations accepted by the federal Executive Committee, and then to obtain approval from the Land party delegate congress. The fundamentalists within the Hesse Green party, opposed as they were to the coalition, made an error in sending Jutta Dittfurth (a prominent fundamentalist) as their spokesperson to convince the Land congress to reject the coalition. Her strident tone and uncompromising style of debate were counter-productive, and over 70 percent of delegates voted both for the coalition treaty and for the nominations for ministerial portfolios [Krause-Burger, 1997:166]. The fact that a survey had shown that 83 percent of prospective Green voters wanted a red-green coalition in 1983 may have influenced the delegates in their decision [Scharf, 1989:171]. Experiences of successful co-operation with the SPD in local councils in Hesse, such as Kassel, sometimes as formal coalitions, also benefited those Greens who wished to extend such co-operation to the level of Land politics in the form of a coalition [Scharf, 1989:168, 170]. 3

A red-green coalition was formed in Berlin in 1989. Though prior to the election the SPD candidate for lord mayor, Momper, rejected the notion of a red-green coalition, declaring the party to be `not competent to govern' (`nicht regierungsfähig') [Appel, 1996:343] and was initially cool towards it after the election, the election result which gave the SPD and Christian Democrats an equal number of seats (and saw, surprisingly, the Republicans winning 11 seats) meant that only a grand coalition or a red-green coalition could be formed. Momper opened talks with the CDU and the Greens (in Berlin, known as the Alternative List: AL), demanding from the Greens four basic conditions, including acceptance of the status of the Allies in West Berlin. (5) A delegate meeting of the AL unanimously agreed to commence negotiations with the SPD, and later approved the coalition agreement. Talks between the SPD and the CDU quickly came to nothing. The Greens received 3 of the 13 ministerial portfolios. The Hesse experiment encouraged the Greens and the SPD in Lower Saxony to form a coalition government after the Land election in 1990. Successful co-operation in opposition and, as in Hesse, experience of coalitions at local council level in Lower Saxony, smoothed the way to such a coalition [Jun, 1994:204] The FDP both before and after the election had rejected the option of a coalition with the SPD. The SPD had kept its coalition options open, though realising that the only likely coalition partner would be the Greens. Coalition negotiations took three weeks, spread over 8 meetings, involving talks first within each party's negotiating team, then between the two teams, and sometimes involving contributions from invited experts [Jun, 1994:196-7]. This coalition was the only one formed to date where the Greens were not able to obtain the Environment portfolio: the Social Democrats had promised to appoint an independent expert, Greifhahn (former Business Manager of German Greenpeace) to that post, but the SPD made concessions concerning environmental policy in order to prevent negotiations from failing [Jun, 1994:198-9]. The Greens' Land delegate conference approved the coalition treaty by 77 votes to 45. The Greens obtained 2 ministries and 3 positions of state secretary. Two of the next cases in chronological order were both particularly unusual: the `traffic light' coalitions formed in Brandenburg and Bremen from the SPD, FDP and Greens (in Brandenburg: Alliance '90). In Brandenburg the coalition following the 1990 Land election was created in a period when the party system of the new Länder was still in its formative phase, and the East German versions of the SPD and FDP were still different from their West German counterparts; the FDP, for instance, had members and political elites drawn from the old block parties (the LDPD and NDPD) as well as from liberal parties formed less than a year earlier [Roberts, 1997:85-7]. There appeared to be no special problems concerning the compatibility in a governing coalition of the SPD, the Liberals and a party based on the citizen movement of the former German Democratic Republic. In Bremen the following year that `traffic-light' coalition pattern was imitated. The SPD had lost its overall majority thanks to a decline in voteshare of over 11 percent. A two-party coalition with either the FDP or the Greens would have been possible for the SPD, but each of these coalitions would have possessed only a very small majority. A grand coalition with the Christian Democrats was not then feasible politically. 4

Some prominent SPD members resigned from the Bremen SPD in protest at this coalition with the Greens, and the Greens only reluctantly accepted the proposal, consenting by a majority of just 1 vote at a Land party delegate meeting to enter into coalition negotiations. However, the coalition treaty produced by those negotiations was accepted by 150 votes to 56 a few days later [Roth, 1992:288-9]. In both these cases the Greens received the Environment Ministry portfolio, along with one other Land ministry. The same year, 1991, a `second edition' of the Hesse red-green coalition was formed. This, too, represents an interesting case, because it was created having regard to the experience of the first coalition, which had come to a premature end amid mutual recriminations concerning policy and tactics (see below). Again, experience of co-operation in opposition and in local councils where red-green coalitions had functioned successfully assisted the coalition formation process; indeed, the new prime minister, Hans Eichel, had been lord mayor of Kassel and in that office had acquired personal experience of such co-operation across the parties [Jun, 1994:207]. The fact that, in contrast to Lower Saxony, the only alternative coalition option was a grand coalition which was politically unfeasible, also made negotiations easier. Fischer, as de facto leader of the Hesse Greens, had prepared in advance for coalition negotiations and for their positive reception within the Land party, so a coalition treaty, after six weeks of negotiation, was accepted without significant dissent by the Green party membership delegate conference. The Greens were given 2 ministerial posts, and Fischer, as Environment Minister, was officially listed as deputy prime minister. He was also given responsibility for European contacts and for Bundesrat affairs, which offered the Greens an opportunity to compensate in the federal legislature for their exclusion from the Bundestag following their poor result in the 1990 federal election [Johnsen, 1993:807]. The fact that the SPD now agreed with the Greens to work towards abolition of nuclear energy provision in Hesse made Fischer's task of `selling' the coalition to his members much easier [Krause-Burger, 1997:192-3]. Following the 1994 Land election in Saxony-Anhalt, yet another format for a red-green coalition was developed: a minority coalition dependent upon toleration by the Party of Democratic Socialism (PDS) - the post-communist version of the former ruling Communist party of the German Democratic Republic. This type of coalition was necessary because of the electoral arithmetic: the PDS had obtained a substantial vote; the FDP had failed to win any seats and the SPD wanted to avoid the only other politically-realistic option, a grand coalition with the CDU. In this coalition the Greens were given the Environment ministry: their only post in the 9- member cabinet. The fragile position of this coalition, required as it was to avoid any initiatives that would cause the PDS to withdraw its acceptance of the government's programme, was a major constraint which compelled a high degree of co-operation between the Greens and the SPD both within the cabinet and in the coalition committee [Fikentscher, 1999:238-9]. The Hesse coalition was able to continue after the 1995 Hesse Land election. The negotiations were relatively brief and without serious problems, either between the parties or within the parties. The Greens had been assisted in the campaign by the good economic situation in Hesse, which had allowed the environmental issue to acquire salience for the 5

electorate (20 percent regarded it as the most important issue: second only to the 44 percent who regarded unemployment to be most important [Forschungsgruppe Wahlen, 1995:43, 52]). Low turnout (66.3 percent) benefited the Greens, as did the fact that, under the Hesse two-vote electoral system, the party got 30 percent of its list votes from voters who gave SPD candidates their constituency vote [Forschungsgruppe Wahlen, 1995:24]. That same year, 1995, saw the creation of a red-green coalition in the most populous Land in Germany: North Rhine-Westphalia. Here the Greens had been relatively weak, failing to win seats in the Land parliament until 1990. The party doubled its vote in 1995 from 5 percent to 10 percent, forcing the SPD (which had previously been able to govern alone) to seek a coalition. Changes in the social structure in the Land were benefiting the Greens at the expense of the trade-union-supported SPD, and (though very much in second place as `most important issue' for voters) the environment was seen as a matter of some considerable concern. Voters had come to dislike single-party government: even 34 percent of SPD supporters did not want their party to govern alone, and 36 percent of all respondents supported the idea of a red-green coalition. This may account for the estimate by INFAS that 225,000 former SPD voters switched to the Greens in 1995, accounting for half their new votes [Feist and Hoffmann, 1996:265]. The Greens in North Rhine-Westphalia had succeeded over the preceding five years in transforming their rather radical image from to a more pragmatic one, and as a party which was a feasible governing partner [Forschungsgruppe Wahlen, 1995(a):51-2]. Negotiations were long and complicated. The demands of the Greens concerning energy production and transport in particular were difficult to make congruent with SPD requirements, dependent as that party was on trade union support. The coalition treaty contained a compromise about the main issue of inter-party controversy: the extension of brown-coal open-cast mining at Garzweiler, and on issues such as the appointment of additional teachers, gene technology and major road-building projects [Das Parlament, 7 July 1995]. Some compromises, such as a failure to prohibit night flights entirely at Düsseldorf airport, were difficult for the Greens to `sell' to their grass-roots membership [Kropp and Sturm, 1998:100]. As well as the Environment Minister, the Greens were given the Construction portfolio also. A similar situation then arose in Schleswig-Holstein in 1996. In the 1992 election the Greens had just failed to win seats in the Land parliament for the first time (4.97 percent), but secured 8.1 percent in 1996, enough - combined with a sharp decline in the SPD share of the vote - to compel the SPD to consider a red-green coalition, despite a rather reluctant attitude toward such a coalition of Simonis (the SPD prime minister in Schleswig-Holstein) during the campaign itself. The red-green coalition was the most popular option for the electorate: 30 percent wanted that coalition, compared to 21 percent who wanted a grand coalition. 58 percent of those who supported the SPD and 83 percent of those who supported the Greens wanted a red-green coalition [Forschungsgruppe Wahlen, 1996:43]. Certain issues led to considerable controversy in the discussions about the coalition treaty (the A20 motorway project, for example). The Greens seemed to have expected the Schleswig-Holstein SPD to still be rather a left-wing party, but 8 years in office had given it a pragmatic and `responsible 6

image, which revealed itself in coalition negotiations which were severely constrained by financial restrictions [Das Parlament, 24 May 1996]. Though initially 5 of the 13 Green area (Kreis) party organisations voted against accepting the coalition treaty with the SPD, the Land party delegate conference approved it by a clear majority [Mnich, 1996:640]. The Greens received 2 ministers (including the Environment portfolio) in a 9-member cabinet. A red-green coalition was formed in Hamburg in 1997. The failure of the STATT-party, which had previously governed with the SPD, to secure any seats in the Land election and the absence of the FDP from the Land parliament left the SPD with a choice between a grand coalition and a red-green coalition. A post-election survey found that 48 percent of respondents wanted a grand coalition, and only 39 percent a red-green coalition [Brunner and Walz, 1998:287]. The 13.9 percent vote-share of the Greens was an especially startling performance given the fact that the environment was not among the top five most important issues among respondents (only 7 percent named it as most important) [Forschungsgruppe Wahlen, 1997:42]. The relatively strong position of the Greens nationally at the time, and the constructive work of the Greens in opposition in Hamburg over the previous legislative period (regarded by respondents as slightly better than that of the CDU there) partly accounted for the support given to the Greens, as did the fact that Hamburg is one of the most favourable areas for the Greens in terms of its post-industrial social structure [Forschungsgruppe Wahlen, 1997:52-3]. The Greens secured 3 ministries (including that responsible for the environment, and one of its ministers named as deputy lord mayor) in an 11-seat cabinet. In the year 2000, the red-green coalitions in North Rhine-Westphalia and Schleswig-Holstein were both renewed following Land elections. In Schleswig-Holstein the Greens lost vote-share in the Land election, though the SPD increased its vote-share at a time of reaction against the Christian Democrats because of the party financing scandal. The presence in the Land legislature not only of the FDP (with more seats than the Greens) but also the party of the Danish-speaking minority, the South Schleswig Voter Association (SSW) complicated the coalition arithmetic. The Greens' contribution to the Land government in the previous four years was rated (on a +5/-5 scale) at -0.7, and the leading candidate of the Greens, Fröhlich, was not known to over half the electorate, despite this being the second election in which she had occupied that position, and was rated negatively by respondents (and only at plus 1.1 by Green voters: the lowest of the top candidates' ratings by their own supporters) [Forschungsgruppe Wahlen, 2000:37, 55-6]. Survey data showed that many Green voters were actually SPD supporters, who had used the two-vote electoral system (applied in Schleswig- Holstein for the first time) to cast a tactical vote for a red-green coalition. The fact that the Greens only received 4.4 percent of first votes, but 6.2 percent of second (list) votes, whereas the SPD received 47.6 percent of first votes but only 43.1 percent of list votes, demonstrates the significance of such `split voting' [Forschungsgruppe Wahlen, 2000:54]. The SPD, as in North Rhine-Westphalia, decided to continue with the existing red-green coalition, even though its majority of 3 seats was lower than the 7-seat majority which a coalition of the SPD and FDP would have possessed. This preference was supported by the electorate (27 percent preferring red-green; 12 percent a coalition with the FDP) and overwhelmingly by respondents 7

who supported the SPD (50 percent preferred this option) or the Greens (84 percent preferred it) [Forschungsgruppe Wahlen, 2000:56-7]. In North Rhine-Westphalia the two governing parties both lost vote-share in the 2000 Land election: the SPD 3.2 percent, the Greens 2.9 percent. The FDP returned to the Land parliament after a five-year absence, and with a vote share of just under 10 percent. The electorate rated the Greens' performance in government and the Land and federal Green parties all negatively, though Höhn, the candidate who headed the Green party list and who had served in the 1995 coalition, was rated positively (plus 0.2 on a +5/-5 scale), and was now known to 65 percent of respondents, compared to only 20 percent in the 1995 campaign [Forschungsgruppe Wahlen, 2000a:46-7, 63]. The SPD had the option of forming a coalition with the FDP, which would have enjoyed a larger majority (21 seats) than a red-green coalition (7 seats). Voters were fairly equally divided: 29 percent preferred red-green, 20 percent a coalition of the SPD and FDP, but SPD supporters preferred the red-green option by a 2:1 majority [Forschungsgruppe Wahlen, 2000a:65]. The fact that the SPD had at least one alternative option to the red-green coalition gave it a stronger hand in dealing with the Greens in coalition negotiations, and its continuance of the red-green coalition in North Rhine- Westphalia was a signal that the party was still loyal to the red-green coalition at federal level [Jesse, 2000:303, 305]. As well as these 13 cases of Land-level coalitions involving the Green party, it is important to note cases where, politically as well as arithmetically, the Greens might have been involved in a Land-level coalition. In Hamburg in June 1982 a red-green coalition would have been arithmetically possible; failure to construct such a coalition led to new elections in December (following which talks were held concerning Green `toleration' of an SPD minority government, but these talks came to nothing [Heinrich, 1999:130]). In 1987 also in Hamburg the SPD preferred the FDP to the Greens as coalition partner. In Hamburg in 1997 the SPD formed a government with the STATT-party rather than with the Greens. In 1991 the SPD became the largest party in Rhineland-Pfalz, displacing the CDU. The Greens and the FDP secured 7 seats each. The SPD decided to form a coalition with the FDP rather than the Greens, a coalition which was continued in 1996. What conclusions can be drawn about the coalition formation process from the 13 cases where the Greens participated in Land-level coalitions, and from the cases where it might have participated, but did not? It is clear that there are significant differences among the 13 `successful' cases, leaving aside the time dimension referred to in the introductory section of this paper. First, one case produced a minority government `tolerated' by the PDS (Saxony-Anhalt) and two (Brandenburg and Bremen) produced three-party `traffic-light' coalitions, including the FDP as well as the SPD and Greens. Second, the Berlin and Brandenburg cases involved parties which did not contest the Land election prior to the coalition as `the Greens'. Third, three cases are essentially continuations of other cases, in that the red-green coalition formed after the Land 8

election in Hesse (1995), Schleswig-Holstein (2000) and North Rhine-Westphalia (2000) were no more than slightly revised editions of incumbent coalitions. When one tries to draw out even the most tentative generalisations otherwise from these cases of coalition formation, they seem more remarkable for their differences than their similarities. Electoral success is a necessary, but not a sufficient, explanation: clearly the Greens have to obtain at least 5 percent of votes to be a `player'. In 6 cases the Greens had lost vote-share at the Land election, in 6 cases it had increased vote-share (the 1990 Brandenburg case has no previous Land-level voting data for comparison). Indeed, in some Land elections where the Greens improved their share of the vote, such as Hamburg in 1993 and Lower Saxony in 1994, the Greens were not able to participate in a coalition. In the Hamburg case, the Greens' hard line led the SPD to seek to form a government with the new STATT-party; in the Lower Saxony case the failure of the FDP to win enough votes to obtain representation in the Land parliament enabled the SPD to govern alone. Support among SPD and Green supporters for a red-green coalition (or its continuance) is a positive factor, of course, in cases where the SPD possesses other options: usually either a grand coalition or a coalition with the FDP. Experience, either within the same Land (as in Hesse, North Rhine-Westphalia or Schleswig-Holstein) or in other Länder, of red-green coalitions, co-operation or coalition at local government level between the two parties, and since 1998 experience of a red-green coalition at federal level, are all factors which may promote rather than discourage creation of coalitions involving the Greens. The existence (or even, for a year prior to the 1998 Bundestag election, the probability) of a redgreen coalition at federal level works in another way as well: it makes the creation, continuation or preservation of red-green coalitions important in relation to national politics. It was rumoured that the North Rhine-Westphalia red-green coalition would have collapsed in 1998 had the Bundestag election not been scheduled for that Autumn. Finally, it should be noted that, where it is feasible, continuation of a coalition reduces negotiation costs after a Land election, a consideration which applies perhaps especially to red-green coalitions [Kropp and Sturm, 1998:21]. The fate of coalitions involving the Green party Are red-green coalitions (or `traffic-light' coalitions) successful? How do they come to an end? Of the 13 cases, 3 (those in Hamburg, North Rhine-Westphalia and Schleswig-Holstein) are still in existence at the time of writing. Another 3 survived for a full legislative term and were then renewed after a Land election (in Hesse, North Rhine-Westphalia and Schleswig- Holstein). Another 3 terminated because a Land election made continuation impossible or politically superfluous: in Lower Saxony in 1994 (because the SPD could govern alone), in Saxony-Anhalt in 1998 (because the Greens won no seats in the Land parliament) and in Hesse in 1999 (where the coalition was replaced by a Christian Democrat-FDP coalition). The remaining four cases were all brought to a premature end. These are the most interesting to examine, not simply because they happened all to be among the first experiments by the Greens with coalition government responsibilities, but also because of the events surrounding their termination. 9

The 1985 coalition in Hesse ended after just 14 months when Fischer's threatened resignation over nuclear power issues was eagerly accepted by Börner, the SPD prime minister. The Greens in the Hesse parliament and government quickly became frustrated by the fact that they joined the government when the legislative term had run half its course and the budget for the coming year had been settled. Constraints on Fischer, ranging from hostility and obstruction from his own federal party (he had difficulty in being placed on the party's circulation lists for key documents, for example), to the powers exercised by the federal government over environmental matters and the competencies of other Land ministries overlapping with Fischer's functions as Environment Minister, (6) as well as the limitations on action placed on Fischer by the coalition treaty (he could not react quickly following the Chernobyl disaster in 1986 by accelerating closure of nuclear power plants in Hesse, for example) all prepared the way for a crisis. This occurred because of disagreements between Fischer and the SPD over a licensing decision for a nuclear power plant at Hanau. When Fischer made a speech in which he insisted that the Greens would not support this, Börner referred to this as `shameless' and affected to identify an offer of resignation in Fischer's speech. Fischer claimed that he had been dismissed, but, either way, the coalition came to a premature and unhappy end [Krause-Burger, 1997:177-9]. The Berlin coalition between the SPD and Alternative List, though marked in its early period by mutual trust and co-operation, became affected by disputes between the two partners over issues such as the sale of the Potsdamer Platz site to Daimler Benz, support by the Greens of a strike by Berlin Kindergarten teachers, and by differing attitudes to German reunification, disputes which were not easily resolved within the coalition committee (Koalitionsausschuss). The decision by the Berlin government to use force to end occupation by squatters of housing in Friedrichshain, (East Berlin) led the AL to withdraw from the coalition two months prior to the 1990 city election [Heinrich, 1993:813-4, 818]. This premature termination of the coalition was of tactical value to both parties, who wished to position themselves favourably especially with regard to their core supporters, ahead of the scheduled all-berlin Land election on 2 December 1990 [Jun, 1994:219]. The `traffic light' coalition in Brandenburg ended prematurely when a series of conflicts within the coalition culminated in Alliance '90 deciding that they could no longer serve under prime minister Stolpe. The evidence of his connections in the GDR with the secret police and his refusal to clarify his involvement before an investigative committee made Stolpe's position dubious [Das Parlament, 25 March/1 April 1994]. The `traffic-light' coalition in Bremen held together longer than most observers had expected. The conciliatory talents of the lord mayor, Wedemeier, restrained the two smaller parties in the coalition from quarrelling too much. However, a dispute arose concerning designation by the Green's Land Environment Minister of a site in Bremen as a protected area, when the FDP had wanted it classified for commercial investment purposes (the `Piepmatz affair'). This provoked the FDP to join with the CDU in a vote of no confidence in the Environment Minister. This motion was successful. The vote of 55-45 meant that at least 5 SPD members of the Land 10

parliament must have voted for the motion, perhaps as part of pre-election manoeuvring. A premature Land election had to be called [Das Parlament, 3 March 1996]. From these four cases, it should be noted that three were not `normal' examples of red-green coalitions: two were `traffic light' coalitions (one involving Alliance '90) and one involved the Alternative List. All were early in the brief history of Green experience in Land government, suggesting that perhaps early experience, together with the decline of the influence of the more radical factions within the Green Länder parties, later led the party to adopt a more pragmatic and long-term approach to coalition co-operation. It is also likely that the SPD also learned from these experiences, and tried to ensure that coalition treaties would anticipate as much as possible the possible issues that may otherwise lead to inter-party conflict (see below). It should be asked why coalitions involving the Greens could not continue after the Land elections in Lower Saxony, Saxony Anhalt and Hesse. The reasons in each case are different. In the 1994 Lower Saxony election the Greens could consider themselves especially harshly treated by electoral fortune. The party secured 7.4 percent of the vote, an increase of 1.9 percent. It was regarded more favourably by the electorate (on a scale ranging from +5/-5) than all other parties except the SPD, having a score of zero compared to its 1990 score of minus 0.6 [Forschungsgruppe Wahlen, 1994:50] and the coalition had worked reasonably well for four years. However, the failure of the FDP to win seats in the Land parliament, thus removing a component of the opposition, meant that the SPD could govern alone even though it had itself added a mere 0.1 percent to its vote-share. In Saxony-Anhalt the situation was much more straightforward. The Greens failed to win seats in the Land parliament in the 1998 Land election (obtaining a vote-share of 3.2 percent, compared to 5.1 percent in 1994), so the party was not available as a partner for the SPD. Federal and Land-level factors worked against the Greens in that election. The party had recently gained notoriety for its proposal at its federal congress to raise petrol prices by taxation to 5DM per litre; the economic situation in Saxony-Anhalt was dire and environmental issues were very low on voters' scales of priorities; the Land party itself and its leading candidate, Heidecke, were regarded with only mild favour (each scored plus 0.2 percent) and its work in government was regarded negatively (minus 0.7 percent (on a +5/-5 scale) [Forschungsgruppe Wahlen, 1998:34, 51-2]. The Hesse case in 1999 is the most interesting in this set. The Greens lost over a third of its vote-share compared to 1995 (7.2 percent, down 4 percent), which, together with a mediocre performance by the SPD (an increase of only 1.4 percent), gave the coalition parties just 54 seats, compared to 56 seats for the CDU and FDP (who managed to get 5.1 percent and thus return to the Landtag after four years' absence). Unemployment was the priority issue for voters, with environmental issues in fourth place. The Greens suffered in part because of the unpopularity then of the Schröder red-green coalition; the performance of the Greens in government federally was scored at minus 0.5 by respondents, and in Hesse at minus 0.6 on a +5/-5 scale). Hinz, the Greens' leading candidate, was rated the worst of all four top candidates both by voters generally and by Green voters; 59 percent of voters gave no response because she was not known to them [Forschungsgruppe Wahlen, 1999:47-8, 53, 67-8]. The CDU also benefited from a controversial petition campaign against the proposals of the Schröder government to 11

allow foreigners seeking German nationality to possess `double citizenship' [Schmitt-Beck, 2000:6-7]. It would seem, then, that factors concerning the Green party, its image and performance as well as external factors, such as changes in the economic situation and the performance of other parties, may account for `electoral' termination of coalitions involving the Greens. Of coalition treaties and other matters The development of formal coalition treaties at Land level, imitating those applied at federal level, has been an especially significant factor in the formation and behaviour of coalitions involving the Greens. They have negative and positive functions. Negatively, they can anticipate likely controversies which otherwise would arise as matters of dispute between the coalition partners, and either `sterilise' them (by postponing decision upon them during the life of the coalition) or `neutralise' them by making explicit what decision or what mode of decisionmaking would apply for the life of that government (for instance that the parties could promote differing solutions, or that one or other of the partners would have a right of veto). The issue of open-cast coal mining in the 1995 North Rhine-Westphalia government and the A20 motorway for the Schleswig-Holstein 1996 government are examples of issues where `agreements to differ' were recognised in the coalition treaties [Kropp and Sturm, 1998:180-2]. The issue of expansion of Hamburg s harbour and airport were settled in the 1997 coalition treaty broadly in line with SPD demands, rather than with the position of the Greens [Brunner and Walz, 1998:288]. Treaties may also include more general agreements that neither partner will vote with opposition parties in an attempt to override the other coalition partner, and rules concerning how the Land votes in the Bundesrat will be cast (or not cast) in various circumstances [Kropp and Sturm, 1998:183-5, 191, 122]. The creation of coalition committees which would, among other tasks, seek to find compromises on disputed matters, is another `negative' attribute of such treaties [Kropp and Sturm, 1998:188-91]. Positively, treaties can propose initiatives which the coalition will undertake in the Bundesrat, for example concerning energy policy, and do this in red-green coalitions more than in other types of coalition [Kropp and Sturm, 1998:138-9, 195-6]. This may include attempts to `rectify' or obstruct policies from the federal government where the ruling coalition consists of the Christian Democrats and the FDP [Heinrich, 1999:136]. Not everything controversial could (nor should) be covered by provisions in a coalition treaty. Unforeseen events and developments as well as questions of interpretation of the treaty itself will give rise to stresses and strains within any coalition, but especially one involving the Greens, where perhaps the grass-roots members are less gratified by government participation than the leading politicians. So, for example, conflicts concerning the expansion of Dortmund regional airport in 1996 threatened the continued existence of the North Rhine-Westphalia redgreen coalition. The Greens wanted to eliminate the Land financial contribution of DM20 million (the federal government and European Union were paying substantial amounts also). Only assurances from the SPD that the expanded facilities would not be used for tourism persuaded the Greens to give way on this. 12

There are other constraints that can be important when examining coalitions in which the Green party is a partner. In many cases, budgetary constraints are a dominant factor, and may be recognised as such in coalition treaties: for example, in the Hesse and North Rhine- Westphalia coalitions formed in 1995, the Schleswig-Holstein coalition formed in 1996, and the Hamburg coalition formed in 1997, as well as in both the `traffic light coalitions' [Kropp and Sturm, 1998:159, 161-4; Brunner and Walz, 1998:288]. This can be employed by the SPD as a barrier to Green party initiatives, since the larger party can claim that new initiatives can only be introduced by eliminating the costs of some other programme within the ministerial sphere of responsibility of the Green party. The fact that several areas of Land policy (including aspects of environmental and transport policy) are shared with, or dominated by, policymaking at the federal level, and, increasingly, at the level of the European Union, also limits and frustrates initiatives which the Greens in Land coalitions might wish to undertake. Coalition negotiations produce such coalition treaties. These negotiations also are concerned with allocation of ministerial portfolios. Junior coalition partners in particular may place great store on the acquisition of particular ministries: at federal level, the CSU has frequently received the Agriculture portfolio in CDU-led coalitions and the FDP has provided the Economics Minister in such coalitions for many years. For the Greens, of course, the Environmental Ministry is their priority, in federal and Land coalitions. As stated above, in all Land coalitions except the Lower Saxony red-green coalition the Greens have provided the Environment Minister. This is not such a big deal in terms of coalition bargaining. Ministerial portfolios have two values in coalition negotiation. They can be symbols of power within the coalition: the `classical' portfolios in federal coalitions (Foreign Affairs and the Interior Ministry, Finance, Justice and Defence), the Economics, Education, Interior and Justice Ministries in Land governments. They can also be of value to a particular party because of their intrinsic importance to that party's policy agenda: the Environment Ministry for the Greens is such a case [Kropp and Sturm, 1998:33-4, 150; Hogwood, 1999:29-30]. This may mean that allocation of such ministries of special value to only one of the coalition partners does not become a contested issue in coalition negotiations. Because Land-level ministerial portfolios often have to be portmanteau ministries, the addition to or subtraction from such ministries of particular areas of responsibility can also be of significance in coalition bargaining. Cases relevant to Land coalitions involving the Greens include the failures of the Greens to add urban planning to the Environment Ministry in North Rhine-Westphalia in 1995 [Feist and Hoffmann, 1996:270] and energy policy to the Environment Ministry's responsibilities in the Schleswig- Holstein coalition in 1996 [Mnich, 1996:640], while in 2000 in North Rhine-Westphalia other planning responsibilities were taken from the Green Environment Minister and transferred to the prime minister's office, as an indication of the weaker situation of the Greens in the renewed red-green coalition, since the FDP would have been a stronger partner for the SPD after the 2000 Land election. On the other hand, in the Hesse coalition formed in 1995, the Greens were given the Justice Ministry (von Plottnitz, a lawyer who had defended members of the Red Army Faction, took this post, which was therefore a controversial appointment indeed!), and Fischer's successor as Environment Minister, Blaul, had family and youth affairs added to her portfolio of responsibilities [Das Parlament, 8 March 1996]. 13

Two other comments must be made concerning allocation of Land ministries. First, in several cases (as in the Schröder coalition at federal level) a Green minister has also been given the position of deputy prime minister or lord mayor (in Hesse and in Hamburg, for example), which provides `added value' to the Green's position at the end of negotiations. Second, ministerial appointments are not the only form of governmental patronage with which coalition negotiations are concerned: state secretary posts and other strategic staff positions in ministries also come into the equation, as they did in Lower Saxony in 1990, for instance, in the case of the Environment Ministry [Jun, 1994:198-9]. Some tentative conclusions A number of - very tentative - conclusions can be drawn from this analysis of 13 cases of Green party participation in Land-level coalitions. 1. Obviously, for the Greens to participate in a coalition they need to secure at least 5 per cent of votes in a Land election, since otherwise the party would be absent from the Land parliament. (7) However, Raschke has recently suggested that the Greens benefit from such a 5 percent clause in terms of party integration: were there to be a 3 percent clause instead, there would be two different Green parties, one left-wing, one pragmatic, he claims [Raschke, 2001:28]. 2. This has implications for aspects of the Land electoral campaigning of the Greens, such as the recognition-level of the top candidate on the party-list, public perception of the record of the Greens in coalition either in that Land or elsewhere (including the federal government since 1998) or in opposition in that Land, the image of the party in terms especially of its factionalism, and the salience of environmental issues for the electorate. 3. However, there is no strong connection between Green party's electoral performance (except for the 5 percent requirement) and the likelihood that it will be required as a coalition partner, since (a) in 50 percent of the cases where the Green party entered a coalition its voteshare in the Land election had actually decreased compared to the previous election and (b) in some cases (even, as on Lower Saxony in 1994 where the party had been in coalition) when its vote-share increased compared to the previous Land election, it was not required as a coalition partner. 4. Loyalty to its coalition partner, effective co-operation in office and the influence of the `model' of the existing federal coalition have been factors in the decision of the SPD to renew an existing coalition with the Greens even when another option (i.e. coalition with the FDP) was available, and even when a red-green coalition would possess a smaller majority than such an alternative SPD-led coalition (e.g. North Rhine-Westphalia and Schleswig-Holstein in 2000). It should be noted that the SPD enjoys additional power in a red-green coalition where it has an immediate alternative option of forming a coalition with the FDP: the so-called `golden scenario enjoyed also by the FDP on occasion, and never available to the Greens. Hogwood's stress on the compatibility of coalition partners, and the record of credibility and trust built up between them, is relevant here [Hogwood, 1999:15, 34-6]. The SPD and Greens have proved reasonably compatible in most Land coalitions, though Flach's warning should also be noted: `A coalition is not a marriage. It is nothing more than a temporary functional alliance of two or 14