Veröffentlichungsreihe der Forschungsgruppe Internationale Beziehungen
P 91 - 302
The Endless Election:
1990 in the GDR
Thomas R. Cusack and W olf-Dieter Eberwein
Publication Series of the International Relations Research Group Reichpietschufer 50
D 1000 Berlin 30
The views expressed in this paper are those of the author and not necessarily those o f the International Relations Research Group.
The assistance o f Carsten Johnson, Katrin Gaertner, Jörg Stangl, and Susanne Fuchs in helping to prepare and organize the data sets used in the analysis reported in this paper is gratefully acknowledged.
The fall of the SED regime led to the first free election in the GDR. The M arch 18, 1990 Volkskammer election was quickly followed by another three, the communal, state, and then the first all-German Bundestag elections. This paper reviews some o f the arguments about the outcomes o f these elections, describes the outcomes at both the national and state level, and analyzes some aspects o f the election outcomes.
Using data at the Kreis level (227 electoral-administrative units), evidence is provided that suggests at least part o f the notable decline in electoral participation over this series o f elections appears to be related to an exit from the electoral system by SED-PDS supporters. In addition, the relative success and failure o f the CDU and the PDS appear to be related to, respectively, the ability and inability o f these two parties to appeal to the interests and sentiments o f the working class in the area of the former GDR.
ZUSAMMENFASSUNG
Der Niedergang des SED-Regimes hat zu den ersten freien W ahlen in der DDR geführt. A uf die W ahlen zur Volkskammer am 18. März 1990 folgten in schneller Abfolge gleich drei weitere, die Kommunalwahlen, die W ahlen zu den neu geschaffenen Länderparlamenten und schließlich die ersten gesamtdeutschen Wahlen. Zunächst erfolgt ein Überblick der Erklärungen zum Ausgang dieser Wahlen. Im Anschluß daran werden die Ergebnisse der Wahlen insgesamt als auch auf Länderebene dargestellt und ausgewählte Aspekte näher analysiert. A u f der Grundlage der Ergebnisse der 227 Kreise wird deutlich, daß der beachtliche Rückgang der Wahlbeteiligung zumindest teilweise mit dem Rückzug der SED-PDS W ähler aus dem Wahlsystem erklärt werden kann. Darüber dinaus scheinen der relative Erfolg der CDU bzw. der Mißerfolg der PDS mit der Fähigkeit der beiden Parteien verknüpft zu sein, die Interessen und Gefühle der Arbeiterschaft in der ehemaligen DDR in Unterstützung umzusetzen.
Introduction ... 1
Alternative Explanations ... 2
Description and A n a ly sis... 7
C onclusion... 21
References ... 24
Friday o f the social sciences by von Beyme (1990a:475). Few had theorized about and even fewer had predicted the end o f the former Stalinist regimes and with them the collapse o f the post World War II bipolar structure and Soviet hegemony in the region. All o f the East European states are now enmeshed in a transition that many hope will lead to stable democratic order. W hether this will occur or not is, as with the breakdown, something social science cannot answer today. But, in the case o f the German Democratic Republic (GDR), this may be less problematic in both theoretical and practical terms. The German problem, at least in the eyes o f some Germans, was solved with the unification. O f all the nations that are now in the process o f transition to democracy, the once GDR is in a privileged position. "The former GDR is about to slip into a new paternalistic relationship, in order to allow all its problems to be solved by a big brother" (von Beyme, 1990b: 180).
This last statement may be somewhat misleading in that it could be taken to mean that the citizens o f the former GDR have completely renounced any further action within the political sphere. There is no sign o f that and there will be many opportunities in the future where their actions in this sphere will not only influence their own lives but also those of others. The East German voters played a critical role in deciding which party or parties should be given the responsibility to solve the massive problems that are now on the agenda o f the Federal Republic. This opportunity arose in the first elections o f the united Germany in December 1990. This election was the last in what at times seemed an endless series beginning with the first free elections to the People’s Chamber (Volkskammer) in M arch 1990, followed by the local elections in May and continued by the first elections o f the newly reestablished Länder (states) in October 1990.
These elections in the former GDR offer an opportunity for systematic empirical analysis o f the special case o f the German transition process. This might serve as a starting point for understanding the evolution o f new patterns o f political behavior within a population that had not had the opportunity to express freely its own preferences. Such understanding might help as well to shed light on some o f the problems Germany confronts as the attempt is made to integrate it into a single political community.
In these four elections the Christian Democrats (CDU), the strongest ally o f the Socialist Unity Party (SED) in the former block party system, rose like a phoenix from the ashes and consistently attracted the largest num ber o f the votes. W hat struck Western observers was the fact that this party found its strongest base o f support among the blue collar workers, a group many, based upon W estern experience, had anticipated would lean heavily toward the Social Democratic Party (SPD). Indeed, given the socio-economic makeup o f this former w orkers’ state, the SPD was expected to emerge as the dominant political force. As things turned out, it only managed to achieve the status o f the largest minority parliamentary grouping. Furthermore, the revolutionaries, who had played a decisive role in forcing the Stalinist system to declare default, were marginalized politically by the electorate.
The fundamental political choice expressed in these elections needs to be explained.
The predominant explanations of electoral behavior in the W est (cf. Falter and Schumann, 1989) are o f limited use. Free elections had never taken place in the former GDR and the new party system was only in its formative stage. Despite this, voting behavior showed a consistent pattern o f a "conservative majority," as Feist (1990) has called it, across all o f the four elections. W hether this will contribute to overcoming the cleavages o f the old exclusionary system or lead to establishing a yet another one between this new conservative majority and a fragmented Left remains to be seen.
In the next section we review some of the explanations that have been offered to account for the voting behavior o f the electorate in the area o f the GDR. In the third section we turn to a description o f how the patterns o f voting behavior have developed in this quick succession o f elections. This is followed by a preliminary assessment o f one explanation for a major outcome of all of these elections, viz., the popular acceptance o f a party with a program o f radical change in the political-economic sphere. The description and analysis presented here draws upon data provided by various statistical offices. These data include information on electoral results and socio-economic characteristics o f the 227 Kreise or districts within the region o f the GDR. The paper concludes with a brief discussion o f the problems and prospects confronting Germany in the integration process and what the results from these elections might entail.
Alternative Explanations
The results o f the first free elections came as a surprise to many, if not most, observers.
Results from the first public opinion polls, taken at the beginning of January in 1990, depicted the SPD as the very likely winner by a large margin. The CDU and its Allianz appeared to be a weak and not very popular contestant. Within a few weeks, however, the picture changed dramatically. The Social Democratic Party, expected to win the first free elections, was overtaken by the Allianz für Deutschland, dominated by the East German Christian Democrats.^ As the three elections later in the year revealed, this was not another ephemeral spike on the graph measuring the public pulse. W hy did this happen?
There is one point o f agreement among the analysts o f electoral behavior in East Germany: it can not be explained by reference to West German voting behavior nor is the theoretical apparatus used in accounting for that behavior applicable to East Germany (Feist, 1990; Gibowski, 1990; Gibowski/Kaase, 1991; Roth, 1990). Explaining the election outcomes at this stage is still something o f a speculative enterprise. Three different clusters
1) In the poll taken by Infas in January 1990 the SPD achieved a 66 percent approval rating. In the polls of the Zentralinstitut für Jugend at the end of January and early February the SPD ranked highest with support levels of 53 and 65 percent, whereas the CDU only obtained 13 per cent in the first of these polls. In the last poll taken by Allensbach one week before the Volkskammer election support for the CDU had risen to 47 percent while the SPD’s had declined to just 27 percent (see Sieger, 1990: 333, Table 3).
o f explanations can be distinguished.^ The first is based on reasoning by historical precedent (Gibowski, 1990:7). The second deals with an argument about the "unprepared voter," the plaything o f the manipulative politician (Sieger: 1990:335). The third explanation is based on a model o f rational choice behavior where issue orientation prevails.
There are two aspects to arguments based on historical precedents. In the first an argument is made that Germans in general vote for the Right in times o f crisis (Feist, 1990:
233) and that the German working class in particular supports the Right in such periods (Hanschmidt, 1990). This sort of reasoning when applied to the DDR elections could be construed to mean that the CDU in the East was viewed by the electorate as a conservative force. This interpretation is questionable to the extent that the Eastern wing o f the party has taken a strong postion in support o f the social dimension o f the social m arket economy concept and is far less market-oriented than the CDU in the region o f the former FRG (Fitzmaurice, 1990).
Second, there is an appeal to the force o f historical voter alignment in the region o f the former GDR. Gibowski (1990:7) argues that the SPD was strong in the region o f what was to be the GDR during the W eimar Republic. That led him to conclude that the SPD should have fared better in the 1990 elections. But one must note that the assumption upon which this argument rests does not appear to be well founded. N either the size o f the electoral support for the Social Democrats during the Weimar Republic in this region nor the party’s relative success could have warranted any optimism. Table 1 reveals that in the five general elections between 1928 and 1933 the Social Democrats never received much more than 30 per cent o f the vote. Mecklenburg and Magdeburg in the 1928 elections were the only outstanding exceptions. In sum, the historical alignment o f East German voters would not lead to the conclusion that an historical stronghold for the SPD has suddenly attached itself to the Bundesrepublik.
The second cluster o f explanations relates to assumptions about the "unprepared voter" (Sieger, 1991: 335). This could be stated as follows: Having never been exposed to the low theater o f western election campaigns, and with little experience in evaluating alternative party programmes, the electorate is easily influenced by the campaign. This can account for the high volatility o f the vote intentions expressed in the polls and the rapidly shifting moods.2 3) The CDU (West) invested far greater effort and money in the electoral campaign than any other party. The interpretation o f the first general elections in the GDR with the overwhelming success o f the CDU as a plebiscite for Chancellor Kohl (Gibowski, 1990: 21) also falls under this rubric.
2) In addition to the studies referenced here there are a number of other reports on the GDR elections, including Berger et al. (1990), Feist/Hoffmann (1991), INFAS (1990), and Jung (1990).
3) Sieger (1990: 336-7) has documented this volatility or moods with respect to the issue of support of socialism and unification.
Table 1 Support for the SPD During the Weimar Republic*
Region June 1928 Sept. 1930 July 1932 Nov. 1932 March 1933
Berlin 34.0 28.0 27.9 23.8 22.5
Potsdam I 34.6 28.7 26.7 23.5 20.8
Potsdam II 30.6 25.9 26.3 22.8 20.6
Frankfurt 33.1 26.6 23.5 22.7 18.6
Magdeburg 43.0 37.2 32.5 31.0 27.6
Merseburg 23.8 19.5 19.8 19.3 16.4
Thueringen 33.3 28.9 22.1 21.6 19.2
Dresden 39.1 34.7 31.1 32.2 30.1
Leipzig 37.0 34.9 33.1 32.2 30.1
Chemnitz Zwickau
33.5 28.3 22.4 zz.3 21.3
Mecklenburg 41.7 35.2 31.3 30.5 26.5
Total 29.8 24.5 21.6 20.4 18.3
Source: Milatz, 1965
This type o f explanation is problematic for three related reasons. First, as Sieger (1991) warns, there may be problems in the polling results themselves upon which the expectations and interpretations are based. Second, even if the volatility o f the polling results during the period prior to the Volkskammer election can be interpreted as supporting the "unprepared voter" argument, one is still confronted with the need to explain the stability o f the voting behavior across the following four elections. And finally, the unprepared voter hypothesis detaches voting behavior and voting intentions expressed by the individual citizen from the historical context in which she or he has grown up.
The third type o f explanation, outlined by Roth (1990), has a degree o f plausibility.
He assumes rational choice on the part o f the individuals. According to this explanation voters made their decisions as a result o f their individual preferences in the context o f the given set of social and political conditions. If these conditions change, the voting behavior will change as well. Roth (1990) argues that the degree to which issues and not so much party preferences determined the election results is one o f the most striking findings, especially when compared to the electoral behavior in West Germany.
M ost observers would agree that a critical motive for the GDR voters during this election was the question o f unification. But this has to be viewed in terms o f a subtle but nonetheless important distinction, i.e., not whether one desired unification but rather the speed with which one wanted unification to occur. Thus, it is difficult to distinguish
amongst voters in terms o f which party they supported and where they stood on the question of whether unification should occur. Over 90 percent o f the population favored unification (Roth, 1990: 383, Table 7). However, there was a systematic diversity on the tempo question. A lm ost 58 percent o f the supporters o f the Allianz für Deutschland (the CDU, the conservative German Social Union, DSU, and the Demokratische Aufbruch, DA) favored a fast unification process. Only 26 percent o f the SPD partisans shared this opinion.
Two-thirds o f the supporters o f Bündnis90 and three-quarters o f the PDS supporters preferred a slow unification process.
This type o f explanation is plausible, but individual rationality is not in itself fully satisfactory because it downplays the socio-economic context determining the final choice.
Gibowski (1990: 12) points to the fact that in the industrial regions o f the South, where the failure and mismanagement o f the Communist system is the most obvious, as compared to the capital Berlin and the North, the unification model propagated by the CDU found great appeal. This amendment to the rational choice explanation is not only plausible but also capable o f accounting for the most distinctive feature o f the electoral behavior in East Germany, when compared to that in the West Germany, namely the overwhelming success o f the CDU among the blue collar workers. Socio-economic context would also account for the relative stability o f the election results that the individual rational choice hypothesis might not fully accommodate.
The percentages o f the votes received from the different socio-economic groupings by four parties, the CDU, the SPD, the PDS, and Bündnis90, are listed in table 2. W e have included the results for the first free election, the Volkskammer, and the elections to the Bundestag. Included as well, for comparative purposes, are the corresponding socio-economic distributions for the West German voters in the Bundestag election.
Turning first to the CDU one is struck by the relatively high level o f support it achieved among blue collar workers. By West German standards it displaced the SPD as the workers ’ party. By contrast, the CDU has done less well in attracting support from other socio-economic groupings. Its support in the regions o f the form er GDR has been significantly lower (relative to levels achieved in the former FRG) in three o f four other categories: managers/intelligentsia, white collar workers, and students. Among the self-employed its relative support has been lower, but only marginally so.
This attraction to the CDU felt by the blue collar workers helps to account for the relatively poor showing by the SPD. In contrast with the situation in the West where the Social Democrats obtained approximately 47 per cent of the blue collar vote, the workers of the ex-GDR provided less than a quarter o f their vote in the two major elections. Other groups that typically provide significant support to the SPD in the West, viz., white collar workers and students, were also clearly less enamored by this party in the East.
The PDS, as one can see, lost considerable ground among all socio-economic groups between the first and second general elections. The successor to the party claiming to
represent the interests o f the working class obtained just 11.9 per cent o f the vote from this group in the Volkskammer elections in March; this share shrank to little more than five per cent by the December general elections.
The Bündnis90, representing the former revolutionaries o f the GDR, did not enjoy much support in any o f the various social strata. This may be related to the marginal position that this primarily intellectually based aggregate o f various social movements occupies.
The only striking result is that this group attracted 25 percent o f the student vote in the Bundestag elections.
Table 2: Party Affiliation and Socio-economic status
Party Affiliation
Blue collar Workers
White collar Workers
Managers -Intelligentisa
Self- Employed
Students
CDU 1) 47.40 38.50 37.30-22.10 53.50 28.50
2) 49.80 37.00 34.40 50.30 18.10
3) (39.00) (43.00) (46.70) (56.90) (32.10)
SPD 1) 22.20 20.60 21.60-22.7 12.80 17.40
2) 24.60 25.10 24.10 16.20 23.10
3) (46.7) (35.90) (31.90) (17.20) (33.5)
PDS 1) 11.90 19.20 20.20-31.00 4.70 19.70
2) 5.30 9.90 13.10 4.30 16.40
3) n.a. n.a. n.a. n.a. n.a.
Böndnis90
1) 1.60 3.60 2.3-6.4 2.50 6.20
2) 5.00 10.30 6.00 4.70 25.00
3) n.a. n.a. n.a. n.a. n.a.
1) Volkskammer, May 1990 (Gibowski, 1990)
2) Bundestag, December 1990, former GDR (Forschungsgruppe Wahlen, 1990) 3) Bundestag, December 1990, former FRG (Forschungsgruppe Wahlen, 1990) for the Biindnis90 the results for the Bundestag include the Greens as well.
Class based voting in the broadest sense, when interpreted in terms o f rationally grounded choice on the part o f voters, may indeed help to account for the results thrown up by the rapid parade of elections held in the region o f the former GDR. W e will turn to a limited effort at trying to establish the plausibility o f this interpretation in the next section when we analyze some o f the results from the first three elections and w hat they meant for the starkest alternative offered to the voters, viz, the CDU as the self-proclaimed vehicle for radical and fast change for the better and the PDS, the embodiment o f the burdened past. But first we turn to a brief description o f various aspects o f the election results.
Description and Analysis4)
The first o f the series o f elections in the GDR was the Volkskammerwahl, the election to the People’s Chamber held on M arch 18. Here, based on a nation-wide proportional system, 400 representatives were elected. Twenty-four parties or groupings (so called lists) contested this election. Shortly thereafter, in early May, the Kommunalwahlen, or Communal election, occurred. Depending upon the Kreis, or electoral district, varying numbers o f candidates were elected to local offices. The third, the Länderwahlen or state elections, took place on the 14th o f October, after the GDR had merged with the Bundesrepublik. Finally, the Bundestagswahl or election took place on December the 2nd.
It cam e as no surprise that the first election mobilized a relatively large num ber o f the voters. As can be seen in Figure 1, participation in the first free election was very high, m ore than 93 percent. This is six percent more than the average o f the three recent elections to the Bundestag in the former Bundesrepublik. By comparison, in Czechoslovakia, the participation rate was as high as 96.8 percent in the first general elections, 90.6 percent in Bulgaria and 86 percent in Romania, whereas the corresponding figures for the first general elections in Hungary and Poland were about 70 and 62 percent respectively. The relatively high level o f electoral mobilization in the GDR may have been due to the fact that elections were clearly cast in terms o f unification now. But the differences across these elections in the East European countries clearly need to be explained.
The participation rate in the communal elections dropped to approximately 75 percent. Nonetheless, this figure was once more higher than that which had been observed in the form er Bundesrepublik during the 1980’s, where participation in a sample o f 50 local elections during the early 1980’s was 67 percent on average.5* In the communal elections in Poland the comparative figure was just above the forty percent level. It may well be that people did not have either a clear understanding o f the importance o f the local
4) The electoral data used in the section are drawn from the following sources: Wahlkommission der DDR (1990), Wahlkommission der DDR und Statistisches Amt der DDR (1990), Gemeinsames Statistisches Amt in Berlin (1990). The socio-economic data were generously supplied by the Staatliche Zentralverwaltung fuer Statistik.
5) A remarkable aspect of the Kommunal elections in May was the very large share of invalid votes. On average, approximately 6.5 percent of all votes were declared invalid in this election. This was far and away higher than the levels registered in the Volkskammer election (slightly more than half a percent), the Landtag elections (slightly less than 3 percent) and the Bundestag election in the area of the former GDR in December (1.6 percent). Voting procedures in this election were more complicated than in the others, as indeed they are inm ost Kommunal elections in the original Bundesrepublik. However, in the latter, invalid votes accounted for slightly less than two percent of all votes in a sample of elections from 50 Gemeinden in the period from 1981 through 1985. This two percent figure is about twice the level that one generally finds in other types of elections in the Bundesrepublik (.94 for Bundestag elections during the 1980’s and 1.15 for the Landtag elections during the period of 1985 through 1989).
It is clear that the complexity o f the voting system employed in local elections diminish the capacity o f the electorate to successfully translate their preferences into valid votes, and this effect was very pronounced in the GDR Kommunal election.
elections or that they considered them much less critical. Another reason for the decline may be that the high level o f political mobilization observed in the first election simply could not be maintained.
Election fatigue or weariness continued into the third set o f elections. Participation rates were below those that have recently prevailed in comparable elections in the area o f the original territory o f the Bundesrepublik. The average across the new Länder was 66 percent, a figure eleven points lower than comparable elections during the late 1980’s in the area o f the former Bundesrepublik. These Landtag elections, again, were not posing the same basic options as the first election. Thus, people may have been less concerned, or else they underestimated their importance. It is clear that these elections were to a much larger degree influenced by the personality factor. The two outstanding cases are Stolpe, running for the SPD in Brandenburg, and Biedenkopf, campaigning for the CDU in Saxony.
Both attracted a significantly larger share o f the vote for their respective parties as compared to the GDR average.
In the fourth election, which took place in December, the trend was reversed.
Participation rose to nearly 75 percent. Even though this figure is slightly less than the participation rate observed in the area of the former Bundesrepublik^, that election was obviously considered m ore important than the two preceding ones by the GDR population at large. This also makes sense given the alternatives offered by Kohl and Lafontaine. The former constantly stressed the advantages o f unification; the latter emphasized the serious follow-up problems that speedy unification entailed.
As noted above, the decline in participation may have been partially a response to the sheer frequency o f elections in such a limited period o f time. There may indeed have been other factors at work. One leading hypothesis here deals with the potential for exit from the electoral system brought about through resignation on the part o f voters who see that the electoral landscape militates against the possibility o f their interests and beliefs being translated into the sort government program they would support. Clearly one group likely to fit into this category would be members and supporters o f the former regime party, the PDS. Having been forced to surrender a significant amount o f power prior to the elections, confronted with a generally hostile electorate, and then suffering a major defeat in the first free election, that to the Volkskammer, it may indeed be the case that this party’s supporters would have seen the tide o f history running against it. W ere that indeed the case, then one might expect an impact to be registered on their willingness to continue to participate in electoral politics.
6) Notable here though is the fact that participation by voters in the area of the former Bundesrepublik was at a much lower level than in the three previous Bundestag elections.
Figure 1
Participation Rate (%)
Electoral Participation in the Four Elections
100
Kommunal
(Maj) Bundestag
(December)
Elections
C o m p a ra b le B R D E le c tio n s
Data limitations prevent us from directly testing these conjectures. W e have no information, for example, from surveys on the attitudes and behavior o f SED-PDS members over this series o f elections. The data available to us is at the aggregate and not individual level and so inference here is at best indirect. We would expect districts with large concentrations o f party members to show relatively large declines or lower levels o f participation in the elections succeeding that to the Volkskammerwahl. Even here we are restricted in that party membership data at this level is not available. So our test of this conjecture must be very indirect in that we have to rely upon information on how voting districts oriented themselves toward the PDS in the initial election and what the apparent impact o f this was on succeeding participation rates.
Figures 2a and 2b plot the Kreis-level participation rates in the Kommunal and Landtag elections against the level o f voter support received by the PDS in the March Volkskammer election. In both instances a strong negative association is apparent. Indeed, in regressing the two participation rates against PDS support in the earlier election, the fits o f the equations are quite good (the IP 's are both above 40 percent). In the case o f the Kommunalwahl the estimated parameter indicates that for each two-percent level o f support provided to the PDS earlier, participation was one percent lower. This effect seems to have increased by the time o f the Landtag election in October where for each two-percent level o f support provided to the PDS earlier, participation was 1.2 percent lower.
Figure 2a
SUPPORT FOR PDS IN VOLKSKAMMERWAHL
PA R TIC IPA TIO N R A TE IN KOM MUNALW AHL
Figure 2b
SUPPORT FOR PDS IN VOLKSKAMMERWAHL
PART1CPATION
RATE IN 80
LANDTAG SW AHL 78 7 6 7 4
72 70 6 8 6 6 6 4
62 60 5 8
56 54 52
PDS VOTE S UPPORT IN VOLKSKAMMERWAHL
As alluded to above, the first free election, that to the Volkskammer in March, led to a significant defeat for the party that had ruled the GDR during the last four decades.
Under its new name, the Party o f Democratic Socialism, it managed to achieve less than 17 percent o f the vote (see Table 3). Two o f the parties with significant contacts and support from their sister parties in the former Bundesrepublik achieved greater measures
o f support, although their respective performances were quite unanticipated. Thus, the CDU, a form er block party within the old SED regime, obtained nearly 41 percent o f the vote while the newly founded SPD, which had been anticipated by many to be the winner of this election, managed only to obtain about 22 percent o f the vote. In all, 24 parties and groupings competed in this election and most were capable o f obtaining very limited voter support. Outside o f the big three, the CDU, SPD and PDS, most of the remaining votes went to such parties as the FDP, a former block party, with slightly more than 5 percent o f the vote, the DSU, a party supported by the Bavarian CSU, with slightly more than 6 percent, the DBD with around 2 percent, and another approximately 5 percent going to various parties and various citizens movement groups (BUR) that had been very active in the events that led to the downfall o f the SED regime.
The basic pattern since the first election involved the general shrinkage o f support for the various minor parties, the solidification o f both the CD U ’s role as the dominant party and that o f the SPD as the second largest group, the continued decline o f the PDS, and a notable rise in support for the traditional swing party o f the West, the FDP. Results from the local elections in May might not appear to conform to this pattern. However, it needs to be noted that during this election many o f the parties formed a bewildering variety o f electoral alliances in different localities. As a result, it is not possible to assess the full level o f support across the country for the different parties. By the time o f the elections to the state parliaments in October, it is clear that the pattern described above had revealed itself. The Bundestag election results for the area o f the former GDR once more conformed to this pattern.7)
7) The CDU merged with the German Farmers’ Party (DBD). The FDP is the merger of the three liberal parties/groups.
Table 3
The Changing Patterns of Voter Support for the Parties
GDR Bundesrep.
East West
Volkskammer (March)
Kommunal*
(May)
Landtag (October)
Bundestag (December)
Bundestag (December)
CDU 40.85 30.59 43.95 41.80 44.30(CDU/CSU)
SPD 21.87 21.05 25.36 24.30 35.70
FDP 5.28 6.39 7.81 12.90 10.60
PDS 16.36 13.99 11.50 11.10 —
BUR 4.88 2.42 6.67 6.10 —
DSU 6.31 3.39 2.44 1.00 —
BAU — 1.97 — — —
DBD 2.18 3.45 — — —
OTHER 2.77 16.75 2.26 2.80 9.40*
CDU: Christian Democratic Union SPD: Social Democratic Party FDP: Free Democratic Party PDS: Party of Democratic Socialism
BUR: Consisting of the following parties/groups: Bundnis90, New Forum, Initiative for Peace and Human Rights, Independent Womens’ Association, and The Greens DSU: German Social Union
BAU: Farmers’ Union DBD: German Fanners’ Party ' - Includes the Greens/Alternative List
The general tendencies that describe the whole o f the former GDR seem to have been consistently played out within the six regions (the five states and the eastern part o f Berlin, that now together with the western part o f the city forms the state o f Berlin) This can be seen in Table 4. Nonetheless, it is clear that there are divergent profiles distinguishing these areas. In the main, there is an indication of a North-South split with the CDU generally dominating but more so in the South than in the North. The SPD has managed at least to be competitive in parts o f the North, and the PDS has retained a relatively stronger position outside o f the South, particularly in the Brandenburg and Berlin areas.
This North-South split is an indication that socio-economic conditions may have been a m ajor factor determining the election outcomes. In the industrialized South, the area most affected by the overall deterioration in regime performance and thus most susceptible to dissent, the CDU did in fact emerge the strongest. Neither the revolutionaries nor the SPD were capable o f presenting an attractive alternative. In contrast, in the most highly urbanized area, (East) Berlin, where one finds the highest density o f the former privileged beneficiaries o f the old system, the PDS obtained its highest levels o f support.
Table 4: Regional Variation in Party Support Across the Elections
Mecklenburg-Vorpommern
Election: V K L B
Party/Grouping:
CDU 36.34 26.28 38.58 41.20
SPD 23.40 20.11 27.15 26.60
FDP 3.59 5.40 5.44 9.10
PDS 22.81 19.04 15.45 14.20
BUR 4.37 3.97 9.17 5.90
DSU 2.39 1.14 0.76 -
BAU - 4.16 - -
DBD 4.89 6.34 - --
OTHER 2.21 13.57 3.46 2.80
Table 4, continued Brandenburg
Election: V K L B
Party/Grouping:
CDU 33.56 24.40 29.60 36.30
SPD 29.91 27.83 38.30 32.90
FDP 4.70 5.47 6.63 9.70
PDS 18.30 16.59 13.71 11.00
BUR 5.41 2.37 9.20 6.60
DSU 3.32 1.63 1.00 -
BAU - 2.84 - -
DBD 2.82 4.62 -- -
OTHER 1.99 14.25 2.10 3.50
Sachsen-Anhalt
Election: V K L B
Party/ Grouping:
CDU 44.54 30.85 39.26 38.60
SPD 23.70 22.94 26.00 24.70
FDP 7.71 9.29 13.54 19.70
PDS 14.01 12.73 11.77 9.40
BUR 3.97 2.48 5.18 5.30
DSU 2.40 0.88 1.71 -
BAU - 1.93 - -
DBD 1.79 3.46 - -
OTHER 1.85 15.45 2.53 2.30
Table 4, continued Thüringen
Election: V K L B
Party/Grouping:
CDU 52.55 35.10 45.65 45.20
SPD 17.52 19.51 22.77 21.90
FDP 4.64 6.78 9.25 14.60
PDS 11.38 9.49 9.57 8.30
BUR 4.11 2.24 6.32 6.10
DSU 5.76 3.27 3.32 -
BAU -- 1.77 -- -
DBD 1.41 3.36 -- -
OTHER 2.63 18.48 2.44 3.90
Sachsen
Election: V K L B
Party/Grouping:
CDU 43.41 35.94 53.97 49.50
SPD 15.05 14.39 19.14 18.20
FDP 5.69 6.62 5.24 12.40
PDS 13.60 10.01 10.08 9.00
BUR 4.71 2.55 5.49 5.90
DSU 13.09 7.24 3.64 -
BAU - 1.36 - -
DBD 1.88 2.64 --
OTHER 2.56 19.24 1.66 5.00
Table 4, continued Berlin (Ost)
Election: V K L B
Party/Grouping:
CDU 18.38 17.73 25.00 24.30
SPD 34.94 34.13 32.10 31.30
FDP 3.02 1.18 5.60 7.80
PDS 29.98 30.11 23.60 24.80
BUR 9.04 0.00 11.40 8.80
DSU 2.22 1.03 - -
BAU 0.45 0.00 - -
DBD - 0.03 - -
OTHER 1.97 15.78 1.90 3.00
Not only have the aggregate as well as state level outcomes within these elections been running along persistent lines, it is also clear, when one examines party support at the Kreis level, that a solidification process has been underway. Table 5 provides information on whether individual districts have tended to support the larger party groupings in similar or dissimilar ways. This table lists the regression coefficients (where constants have been suppressed in the estimation) and correlation coefficients (in parentheses) describing relationships amongst voter support for five different parties/groupings at Kreis level across the first three elections. The correlations between support levels for the three initially largest parties, the CDU, the SPD, and the PDS, are quite high. The regression coefficients suggest that while the CDU has generally been holding its own, the SPD has tended to systematically receive greater support over that which it had previously, and the PDS has continuously lost ground in those areas where it had been strong. Support for the FDP geographically was not very stable between consecutive elections, but was relatively more systematic in its geographical concentration between the first and third election with its support levels generally rising by a factor of 0.5. A similar pattern holds for the conglomeration o f the civic movement parties.
The voting behavior in the GDR, in sum, can be seen to be a relatively stable expression o f the preferences o f support o f the electorate and not subject to random variation. In the next pages we will try to translate the general argument developed above into a testable proposition. We are restrained in that we can only formulate it for the aggregate level o f analysis, the Kreis level. There are 227 such units, which gives us a sufficiently large number of cases for statistical manipulation.
Table 5: Stability in Party Support at the Kreis Level
Volkskammer -Kommunal
Elections
Volkskammer -Landtag Elections
Kommunal -Landtag Elections
CDU .80
(•82)
.98 (.71)
1.22 (•79)
SPD O\ 00 COG"
1.13 (•91)
1.27 (•75)
FPD 1.39
(•41)
1.47 (•72)
.87 (.46)
PDS .89
(•95)
.71 (•90)
.71 (•91)
BUR 1.2
(•26)
1.57 (-73)
.97 (.27)
Our basic expectation is that the CDU stood to gain voter support to the extent that a district’s population had been relatively disadvantaged as a result o f the 40 years o f SED rule. Equally the case, w e expect areas wherein greater benefits had been provided to the population would be less likely to support the CDU and more likely to support its polar opposite, the PDS.
Figures 3 and 4 lend credence to this basic expectation. Thus, in Figure 3 we have plotted the district level support for the PDS in the Volkskammerwahl against the relative share o f the labor force engaged in production (the latter measure being a surrogate for the district level strength o f the blue collar class). As the distribution o f points as well as the results reported from a regression analysis included within the Figure demonstrate, the basic expectation meets with empirical support. Thus, the greater the presence o f blue collar workers, the lower the level o f support for the PDS. Conversely, in areas where some physical manifestation of relative privilege (in this instance, the proportion o f housing units equipped with bath facilities) abound, support for the political movements most identified with a radical change from the old regime (here we use support for the Allianz parties and the liberals) would be generally lower. Again, the graphic and statistical evidence provided in Figure 4 are supportive of our expectations.
A more encompassing demonstration o f the utility o f our leading hypothesis can be seen in the analyses reported in Tables 6a and 6b. Here we estimate for each o f the first three elections, models that are intended to capture the effects o f relative advantage and disadvantage on the propensity of voters to support the party of radical change (the CDU) or the party o f the past (the PDS). The models are incorporated in equations that are
estimated with a logistic regression technique o f the form used by Brown (1982) in his ecological analysis o f Nazi Party support during the Weimar Republic. The basic equation employed in the analysis o f party support at the district level is as follows:8*
L = b, + b ß C W + b ^ D V + b^SIZE + e
where L = l n ( j ^ , or the log o f the odds ratio, and P is the proportion o f the electorate within each district voting for the party. The independent variables include BC W , the share o f the district’s labor force engaged in production, ADV, the degree o f advantage in physical amenities enjoyed within the district, and SIZE, a ordinal variable measuring the relative size o f the community to which the district belongs, with the highest score given to districts in large cities. Our expectations are as follows. Heavily blue collar areas would be more supportive o f the CDU and less supportive o f the PDS, i.e., b twould be positive in the CDU vote share equations and negative in the PDS equations. Both b2 and b3 would have negative signs in the CDU equations and positive signs in the PDS equations. Note that while we believe that it is more likely that urban settings would provide greater amenities, our expectation was that the ordinal measure available to us would not be sufficiently discriminating and so each o f the equations was reestimated with the SIZE variable excluded. Note also that as is appropriate with analysis o f ecological data based on units o f disproportionate size, the observations have been weighted by relative population share (cf., Brown, 1982).
The results o f the logistic regression analyses reported in Tables 6a and 6b generally are supportive o f our leading hypothesis. Thus, in the CDU equation, br is consistently positive and always statistically significant, suggesting that districts with heavy concentrations o f blue collar workers were more likely to support the CDU. The mirror image holds for the PDS. Thus, b y in the PDS equations is consistently negative and statistically significant across the election-specific equations. On the other side, it is clear that at least one measure o f relative advantage, A D V , captures the effect w e posit, namely, the tendency to vote against the party o f radical change and to vote in support o f the party o f the old regime. The SIZE variable is relatively erratic in terms o f its measured effects on voting behavior and indeed in the two instances where it is significant, in the equation for the CDU vote in the Kommunalwahl and in the equation for the PDS vote in Volkskammerwahl, the estimated sign is opposite that predicted. Nonetheless, it is clear that our basic expectation meets with a good measure o f support in these regression results.
8) While there were originally 227 distncts in the GDR region at the beginning of the series of 1990 elections, variation in the number of districts included within our analysis was dictated by such factors as (1) the need to exclude one notional district wherein citizens living overseas are aggregated, (2) the failure of a party to contest the election in a district or to have joined in an electoral coalition within a district, and (3) the merger of one of the districts in the Halle region after the Volkskammerwahl.
Figure 3
Volkskammerwahl: Industrial Workers and PDS VOTE
Right (Allianz
&BFD ) Electoral Support (RTVOTE)
PDS Electoral Support (PDSVOTE)
Figure 4
N um ber o f Housing Units (per thousand) Equipped w ith BatIW Showers
Table 6a
Socio-Economic Factors and Support for the CDU:
CDU Support Level in
Election: Volks.: Volks.: Komm.: Komm.: Land.: Land.:
Independent Variable:
Industrial Workers as a % o f Labor Force:
B C W
0.048 (10.09)
0.047 (12.72)
0.034 (8.196)
0.032 (7.918)
0.027 (5.067)
0.021 (5.018) Size of Community:
SIZE
0.009 (0.269)
0.094 (2.786)
0.062 (1.823)
% of Households with Bath Facilities: A D V
-0.002 (-6.23)
-0.002 (-6.24)
-0.002 (-6.76)
-0.002 (-6.52)
-0.003 (-8.46)
-0.003 (-8.45)
Constant -1.863
(-4.03)
-1.794 (-4.68)
-1.647 (-3.48)
-0.982 (-2.37)
0.322 (0.634)
0.855 (2.040)
53.21 53.4 42.61 40.72 35.04 34.33
F-statistic
Number of Kreise
86.7 226
130.5 226
2.2 208
72.1 208
39.5 215
56.9 215 (t-stats. in parentheses)
Table 6b
Socio-Economic Factors and Support for the PDS:
PDS Support Level in Election:
Independent Variable:
Volks.: Volks.: Komm.: Komm.: Land.: Land.:
Industrial Workers as -0.034 -0.029 -0.04 -0.036 -0.031 -0.033 a % of Labor Force: (-7.682) (-8.18) (-8.92) (-10.0) (-8.35) (-11-3) B C W
Size of Community: -0.059 -0.039 0.017
SIZE (-2.03) (-1.30) (0.712)
% of Households with 0.002 0.002 0.003 0.003 0.002 0.002 Bath Facilities: A D V (7.123) (7.108) (9.028) (8.961) (8.009) (8.007)
Constant -1.197 -1.706 -1.395 -1.68 -1.788 -1.646
(-2.73) (-4-73) (-3.23) (-4.50) (-4.98) (-5.59)
R 1 F-statistic
Number o f Kreise
40.22 49.0
215
39.34 70.4
215
53.42 85.5
221
53.28 127.0 221
51.42 76.5
215
51.53 114.8 215 (t-stats. in parentheses)
Conclusion
The elections in the former GDR and the first all-German elections have resolved the question about which parties the voters in both parts consider the most competent to solve the problems o f unification during the next legislative period. The victory of the conservative-liberal coalition, in power since 1982 in the former FRG, is uncontested. But this electoral success may yet prove to have been far easier than the responsibilities that accompany it. Managing the integration of two distinct societies into a single community, in contrast, is a formidable task, the dimensions o f which become more apparent with each passing day.
In the seventies Schweigler (1975) argued that the German problem had been resolved: "Two Germanies, not one." If the two societies are now completely distinct, as Schweigler and many others believe, the formation o f an integrated community will be extremely difficult. B est’s (1990) empirical analysis, based on the communication approach developed by Deutsch (1964), shows that various transaction indicators reveal that ties between both communities have weakened over time. This suggests that the distance between the two German peoples has increased. To reverse that process may be difficult.
There are problems with the Deutsch paradigm, as Best suggests, in that it is defined in terms o f macro-level behavior and ignores the fact o f the individual consciousness of unity. This consciousness is not changing as fast as transactions are. There is w hat Ogbum calls a "cultural lag." Although the chances for unification were never estimated to be high, the available data indicate that German unity has always been alive in the minds of the people in the West.9) W hile the evidence is anything but conclusive, W est German public adherence to the principle o f German unity may contribute to the creation o f a new sense o f community and facilitate the task o f integration.
The attitudinally favorable predisposition towards unification and integration is clearly overshadowed by the actual differences between the political cultures o f the two countries. These differences may prove to be difficult obstacles to overcome. Some findings reported by Feist (1991) attest to this.10) First, she shows, perhaps not unexpectedly, that the W est Germans have a relatively postmodern view o f society. In contrast, people in the form er GDR hold to a much more traditional or materialist outlook on life. Second, Feist (1991:26) points out that relative to citizens in the West, the East Germans have far less trust in existing political institutions. Third, she found that the willingness to restrict the franchise and other forms o f political participation was noticeably greater in the East. Feist also points out that there are some fundamental differences in the way in which citizens on both sides o f the former border conceive o f democracy. Three
9) Best points out that the many surveys concerning the attitudes of the West German people towards unification and commissioned mainly by the Federal Ministry of Inner German Affairs are still not available.
10) Feist’s report is based on a survey conducted by INF AS between September and November of 1990.
distinct views are common in the West. These include (1) a conception o f democracy based upon the principle of leadership provided to the uninformed masses by a competent elite, (2) a vision based upon the principle o f the representation o f various political interests, and (3) a participatory model o f direct democracy. In the East the elitist and representational concepts o f democracy are merged into one. The conflicting principles o f elitist leadership and representation are not perceived as mutually exclusive but interdependent.
Convergence between the two sides only occurs with respect to the third model, i.e., direct democracy through participation by the people.
In a survey conducted in both parts o f Berlin in the spring o f 1990 by Klingemann (1990), relevant behavioral predispositions were uncovered in terms o f anomy, authoritarianism, and ethnocentrism. In all three cases the proportions o f people with these predispositions are higher in the East than in the West: 58 to 24 per cent for anomy, 62 to 30 percent for authoritarianism, and 26 to 11 per cent for ethocentrism (cf., Feist, 1990:
24).
Differences in value orientation (modern vs. traditional), political trust, democratic norms, and politically relevant predispositions, are in line with w hat one can and should expect given the different histories o f the two parts o f Germany. But their importance cannot be gainsaid in that they signal, as Feist (1991:32) argues, a more limited willingness on the part o f the East Germans to support the democratic order during times o f crisis.
And while some eschew the use o f that term in the present circumstances, it is clear that rising unemployment, the climbing cost o f living, growing disillusionment with the results o f unification, a heightening sense amongst many East Germans o f being "second-class citizens," and the unabated exodus o f the younger and better trained to the West, would not normally be considered ingredients in a recipe for a stable democratic order.
Frustrations are rising because the expectations about the fruits o f unification have not been met. While surveys indicated that few expected the situation to improve immediately (see, among others, Roth, 1990), it is clear that not many people expected the wrenching changes that have occurred. This might reinforce antidemocratic attitudes and further reduce trust in existing institutions, thereby laying the groundwork for widespread protest and violence.
Superficially, at least, the integration process is on track. One indicator o f this is a convergence of the party system in both parts o f Germany. But, as the electoral studies reported and our own results show, the majority party, the CDU, has a constituency in the East that is quite different from w hat it has in the West. The same is true o f the SPD (Feist, 1991:32). The speed and efficiency with which the political parties are capable and willing to adjust to these new conditions and reconcile internally the diverging interests is therefore an important element in the integration process.1^ For the CDU, the rational choice
11) Schacht (1990) discusses this specific issue from the perspective of the Social Democratic Party.
explanation in conjunction with the systematic contextual effect o f socio-economic conditions points to the risk that this party will be penalized by the electorate in the next round if it fails to meet its voters’ expectations.
W ithin the East the voting patterns show the polarization between the CDU and the PDS while at the same time revealing a fragmentation on the Left. There are indications pointing to declining support for the PDS and increasing attraction to the SPD. It may be then that the fragmentation within the Left will disappear with time. The effect o f this m ight work in both directions, either contributing to the political polarization within the East and reinforcing the East-West cleavage, or, alternatively, strengthening the legitimacy o f the new democratic pluralist structures and enhancing political stability.
As Deutsch et al. (1957:40) remind us, the successful amalgamation o f communities rests on the balance between capabilities and loads. The loads o f the integration process can be identified with increasing precision. They are still growing. As far as the capabilities needed to cope with them, there is not much one can say at this point. One can only hope that the existing political system structure will enhance the mobilization of the capabilities required to bring the process o f integration to a successful conclusion.
References
Berger, Manfred, Wolfgang G. Gibowski und Dieter Roth (1990) Ein Votum für die Einheit. D ie ZEIT, 23.3.1990.
Best, Heinrich (1990) Nationale Verbundenheit und Entfremdung im Zweistaatlichen Deutschland. Theoretische Überlegungen und empirische Befunde. Kölner Zeitschrift fü r Soziologie und Sozialpsychologie 42(1), 1-19.
Beyme, Klaus von (1990a) "Die vergleichende Politikwissenschaft und der Paradigmenwechsel in der politischen Theorie." Politische Vierteljahresschrift, 31(3), 457-74.
Beyme, Klaus von (1990b) Transition to Democracy - or Anschluß! The Two Germanies and Europe. Government and Opposition 25(2), 170-90.
Brown, Courtney (1982) The Nazi Vote: A National Ecological Study. American Political Science Review 76 (2), 285-302.
Deutsch, Karl W. (1964) Transaction Flows as Indicators o f Political Cohesion. In: Ph. E.
Jacobson and J. V. Toscano (eds.) The Integration o f Political Communities.
Philadelphia, Pa.: Lippincott.
Deutsch, Karl W., Sidney A. Burrell, Robert A. Kann, M aurice Lee Jr., Martin Lichtermann, Raymond E. Lindgren, Francis Loewenheim, and Richard Van W agenen (1957) Political Community and the North Atlantic Area: International Organization in the Light o f Historical Experience. Princeton, NJ: Princeton University Press.
Falter, Jürgen, und Siegried Schumann (1989) Methodische Probleme von Wahlforschung und Wahlprognose. Ams Politik und Zeitgeschichte -B eilage zur Wochenzeitung D as Parlament, B 43/89, 3-14.
Feist, Ursula (1991) Zur politischen Akkulturation der vereinten Deutschen. A us Politik und Zeitgeschichte - Beilage zur Wochenzeitung D as Parlament, B 11/12, 21-32 Feist, Ursula (1990) Votum für einen konservativen Modemisierungskurs. Analyse der
Volkskammerwahl in der DDR. Gewerkschaftliche Monatshefte Nr. 4/90, S. 233-40.
Feist, Ursula, and Hans-Jürgen Hoffmann (1991) Die Landtagswahlen in der ehemaligen DDR am 14. Oktober 1990: Föderalismus im wiedervereinten Deutschland - Tradition und neue Konturen. Zeitschrift fü r Parlamentsfragen, 22(1), 5-33.
Fitzmaurice, John (1990) Eastern Germany. Electoral Studies 9(4), 327-336.
Forschungsgruppe W ahlen E.V. (1990) Bundestagswahl 1990: Eine Analyse der Ersten Gesamtdeutschen Bundestagswahl am 2. Dezember 1990. Mannheim: Berichte der Forschungsgruppe Wahlen, Nr 61.
Gemeinsames Statistisches Amt in Berlin (1990) Wahlen zu den Landtagen der Länder Mecklenburg-Vorpommern, Brandenburg, Sachsen-Anhalt, Thüringen, undSachsen am 14. Oktober 1990. Berlin: Gemeinsames Statistisches Am t in Berlin.
