The Constituency Roles of MPs at the Federal and Länder Levels in Germany

Abstract

Germany has both a federal system of government and mixed member proportional (MMP) electoral systems at both federal and Länder levels. The paper explains how most patterns in the constituency work of German MPs at both levels are functions of the electoral systems employed, party competition and tactical considerations. District MPs have significant commitments to constituency work; overall list MPs do as well, though to a lesser extent. The article defines three types of list MP, between which there are variations in both the volume of constituency work undertaken and engagement with individual constituents or group interests. Relationships between members are strongly related to partisan contests, although the overall legitimacy of both federal and electoral systems means there is no dispute over the rules governing the rights of different types of member in constituency work.

Keywords:

• Constituency representation
• federal system
• mixed member proportional electoral systems

Notes

¹The core elements of representative democracy include a desire for the mutually independent action of representatives and those who are represented; the possibility but usual absence of frequent or severe conflicts between both; and an effective combination of responsiveness and leadership on the part of the representatives. Responsiveness, in turn, is a mixture of policy, allocation, service and symbolic responsiveness, as conceptualized by Eulau and Karps (1978: 55–71). ‘Constituency’ is understood as a complex concept; see Fenno (1978). Within the ‘geographical constituency’ (fixed by law) there are an MP's ‘re-election constituency’ (the people who vote for an MP) and the ‘primary constituency’ (the MP's strongest supporters among the selectorate). The theory of constituency representation underpinning this text is more extensively formulated in Patzelt (1993: 15–99).

²The best comprehensive information on German Länder parliaments can be found in Mielke and Reutter (2004).

³The Basic Law established only general principles of free elections. The electoral systems were established by special laws at federal and Länder levels. Legal provisions regarding parties are given by federal law. In both domains, rulings of the Federal Constitutional Court have supplemented the constitutional and legal arrangements for parties and party competition in Germany.

⁴Candidates enter parliament according to their position on the party list, beginning with those placed in the top positions. At the federal level and in most Länder, the order of election of candidates from the list cannot be changed by the voters.

⁵Between 1949 and 1990, there were on average 2.3 excess seats in the Bundestag per legislative term. This figure initially increased to more than ten after reunification because of the different structure of the East Germany's party system. For a discussion of this issue see Nicolaus (1995: 353–371).

⁶However, this rule is not applied if a party has won at least three direct seats. In this case it will have a share of parliament seats proportional to its share of the vote.

⁷Baden-Württemberg requires all candidates to run in a voting district, but allocates seats in parliament according to a party's share of votes. As a consequence, defeated candidates for a direct seat enter parliament like list MPs elsewhere, although voters have only one ballot. In Bremen, Hamburg and the Saarland there are no direct seats; all MPs are elected on party lists. However, Hamburg is about to introduce voting districts with direct seats.

⁸Among all Germans, 22% thought a party's share of Bundestag seats depended on its share of ‘second ballot’, 29% ‘first ballot’, 28% ‘both ballots equally’ and 19% ‘did not know’. See Noelle-Neumann and Köcher (1997: 834).

⁹Most of the empirical research on German MPs' constituency work has been carried out by the author, based on the 1994 survey of all German MPs at state, federal and European levels, and a network and observational study in 1996/97 of the constituency work of state and federal MPs in eight voting districts of Saxony and Hesse. For the main results see Patzelt (1996a) and Patzelt and Algasinger (2001: 503–527). Detailed narrative interview data on the constituency work of state, federal and European MPs in Bavaria, stemming from research carried out in 1989, can be found in Patzelt (1995b). Some of the results are also available in English; see Patzelt (1997a: 55–78; 1999: 23–52; 2003: 110–134). The research was originally informed by scepticism towards data gained from self-reported interviews and questionnaires. Subsequently, however, both observational studies and multi-level network analyses have validated most self-reported data.

¹⁰This slightly biased sample was weighted to make it as representative as possible.

¹¹In the Bundestag's first legislative term (1949–53) there were 31 direct seats won by independent and minor party candidates, in the second term (1953–57) 25 direct seats, and in the third term (1957–61) 7 direct seats. Subsequently, the 12th (1990–94) and 15th legislative terms (2002–05) saw one direct seat each won by FDP and Green candidates, who had large personal votes. Since 1990 the PDS has won up to 23% of the vote in East Germany and consequently has attained a few direct seats. See Schindler (1999: 3680).

¹² φ (phi) measures the strength of the relationship between two dichotomous variables. It ranges from 0 to 1.

¹³ η ² (eta squared) measures the strength of the relationship between an independent variable measured at the nominal level and a dependent continuous variable. It ranges from 0 to 1; its value can be interpreted as the percentage of variance in the dependent variable that is explained by the independent variable.

¹⁴Using this left–right-attitude scale as a covariate is a sound proxy for controlling party membership directly; both variables have a correlation of η ² = 0.46. It was impossible to include ‘party’ as a separate factor in the model, because in smaller parties there are either no MPs with a direct seat or not enough to compare. For further details see Patzelt (1997c: 1–2).

¹⁵In addition, case numbers would have been simply too small for separate multivariate analyses of Bundestag members.

¹⁶Just as for comparison between direct and list members it would be misleading to infer the effects of being a member of the Bundestag or of a Landtag simply from cross-tabulations. Again, hierarchical ANOVA models were used. The following effects were controlled for: representing rural vs. metropolitan constituencies; left vs. right political orientation; and the number of years spent as an MP. Controlling for the type of seat was impossible, because this would have meant controlling for differences between SPD and CDU/CSU on the on hand, and the Greens and the liberal FDP on the other. Again, case numbers dropped to a quarter when a question was included in only one of the four questionnaire versions, although this did not happen very often. For all details see Patzelt (1997b: 1).

¹⁷A comprehensive analysis of the ‘founding phase’ of East German parliamentarianism can be found in Patzelt (1995a) and in Patzelt and Schirmer (1996: 3–11).

¹⁸The research used a five-point rating scale where 1 = ‘I still work fully in my original profession’ and 5 = ‘I do not work any longer in my original profession’. Bundestag members measured 4.6 on average, while Landtag members measured 4.0.

¹⁹The research used a five-point rating scale where 1 = ‘I am politician by profession’ and 5 = ‘I am not politician by profession’. Bundestag members measured 2.0 on average, while Landtag members measured 2.6.

²⁰Fenno defines this as the ‘primary constituency’.

²¹More than a third of German MPs at Länder and federal levels have parliamentary functions at lower or middle levels of influence. The phenomenon of the backbencher who has no influential role in parliament at all certainly occurs, but this is not typical of most German MPs.

²²These include the FDP (liberals), the Greens (ecologists), and—at Länder level—right-wing radicals. Although the PDS is a major party in East Germany, it should also be included here as a tiny party in West Germany.

²³In all German parties nomination assemblies for party lists for elections both at Länder and federal level are convened at the Länder level. They are usually composed of the members of the parties' Land executive committees, delegates from the district party organizations, and the heads of the Land parties' sub-organizations concerned with, for example, women, youth, seniors, economy, labour and agriculture. Party leaders at the Land level and the most important second-rank party leaders lead negotiations before the nomination convention and come up with suggestions for a party list that is balanced in terms of representing different strands of political opinion, geographical interests, gender and age. Whereas the basic structure of the list is usually accepted by the nomination assembly, several sometimes dramatic changes to the actual candidates on the list as well as their list positions are quite normal. The main reason is that secret ballots are held to determine each list position. Consequently, nomination conventions usually are quite time-consuming events.

²⁴A single party won more than 55% of the first ballots in no more than 20% of the Bundestag voting districts in 1990 and 18% in 1994 (see Schindler, 1999: 240). Because direct MPs are elected not by a majority, but by a plurality of votes, the proportion of voting districts considered ‘safe’ is higher, but it still leaves many seats open to a credible contest.

²⁵Roughly 75% of all German MPs are members of leadership bodies at different levels of the party. Twenty per cent are members and 16% chairpersons of local executive committees; 38% are members and 22% chairpersons of district executive committees; and 22% are members and 4% (deputy) chairpersons of their party's executive committees at the Land level. MPs from the Green party were not allowed to formally combine a seat in parliament with a party leadership position until recently, but they have always exerted informal party leadership.

²⁶Conditions in East Germany after reunification were exceptional. So inclusion of East German MPs in this analysis would bias the ‘normal’ figures.

²⁷See Patzelt (1996a) for the full results of the 1994 study. Selected data are presented there in a highly condensed manner and without differentiation.

²⁸On balance, LMs are not doing less work than DMs. As can be seen in Table 3, they simply spend a comparatively smaller share of their working time on constituency work and a larger share on activities in the parliamentary arena. The key factor appears to be that the relatively few MPs of smaller parties, who usually are list MPs, simply have to concentrate on the parliamentary arena and cannot compete with the constituency work of the many direct MPs of major parties.

²⁹For discussion of the concepts of service and allocation responsiveness, see Eulau and Karps (1978). For a discussion of empirical evidence see (Patzelt, 1997c: 16–18).

³⁰The numbers of citizens represented on average by direct members also varies between the different Landtage: from approximately 43 500 citizens in Berlin to 135 000 in Bavaria.

³¹Cf. the discussion of Table 1. In addition, East German MPs have been excluded from this analysis. Because it is based on data from 1994 one would misinterpret ‘level-of-parliament’ effects for what have been really special characteristics of the newly established East German Länder parliaments.

³²The fact that local government is under the sole responsibility of the Länder is a key factor explaining MLs' special eagerness to serve as district and city politicians.

³³Nevertheless, MLs complain more than MBs about being unable to devote themselves sufficiently to their different duties. Perhaps the fact that their constituencies are smaller leads them to set themselves higher standards of availability and conduct of casework.

³⁴The Länder and their parliaments are basically responsible for public administration, local government, police, public infrastructure, cultural institutions, schools and universities, and for policies on water supply and forests. The Federal Government and the Bundestag are exclusively responsible for policy areas such as foreign policy and European affairs. In most other policy areas there are overlapping responsibilities for the cabinets and parliaments of the Bund and the Länder.

³⁵See, for example, the extended interview statements in Patzelt (1995b: 68–133).

³⁶When invitations to overlapping constituency events such as ceremonial openings, festivals and jubilees are received, MPs even from competing parties sometimes will co-ordinate who should accept which invitation and convey the greetings of absent colleagues. Nevertheless, this is the exception, not the rule, because systematic co-ordination is practically impossible. Most demands on an MP originate from quite different primary constituencies and simply lead to different, even though overlapping, patterns of constituency work.

³⁷In contrast, ‘downward hopping’ is rare. In 1994 only 6 of the 457 West German MPs in our ML sample had been members of the Bundestag previously.

³⁸The level of parliament does not interact with the three relationships discussed here. Therefore it is not necessary to cross-tabulate all varieties of level-of-parliament and type-of-seat effects and discuss them separately.