The results of the June 1996 election to the Parliamentary Assembly of the Czech Republic left Czech newspapers in the awkward position of not being able to identify a clear winner. One paper declared "two winners," and another, in its election supp lement, placed Czech premier Vaclav Klaus on the cover while inside declaring that the election belonged to his rival Milos Zeman. The close results make the confusion understandable, but amid the debate over the significance of the results for various p arties, it has gone largely overlooked that in at least one very specific sense the election confirmed a victory for the Czech Republic as a whole. The source of that victory is the maturation of the Czech Republic's political party system.

Although there has been considerable attention to the Czech Republic's parties from many Czech and some western political scientists, there has been little systematic attempt to understand the Czech Republic's party system as a whole. There has also been little comparison of Czech system to those of other countries, particularly countries in Latin America with arguably similar experiences of recent democratization. Comparison has been made even more difficult by the tendency of authors studying dif ferent sets of party systems to emphasize different features of those party systems. Where Giovanni Sartori [Sartori, 1976 #57] explores Western European party systems in terms of fragmentation and polarization, Scott Mainwaring and Timothy Scully [Mainw aring, 1995 #8] explore Latin American party systems using measures of institutionalization and Herbert Kitschelt [Kitschelt, 1995 #82] examines Eastern European party systems in terms of programmatic structuring. In this paper I will attempt to synthesize those characteristics of party systems that Sartori, Mainwaring and Scully, and Kitschelt believe to be important for democracy, and I will use those characteristics to examine the development of th e Czech Republic's party system. Rather than present new field research, this paper will use these more theoretical works together with the excellent existing research on Czech political parties and the results of the 1996 elections to provide an overvie w of the Czech political party system as a whole. From the data that is available, I argue that the Czech party system has already developed some of the characteristics of party systems which are believed to be important for the consolidation of a democr atic system of government. Even for the many characteristics which are not yet present in full, political developments in 1996 at least indicate a positive direction.

Characteristics of Party Systems

Many prominent researchers suggest that political party systems play an important role in shaping democracies [Lijphart, 1994 #19; Sartori, 1976 #57; Schumpeter, 1942 #58]. In Western democracies, parties serve as organized, two-way links between th e people and the government, both representing popular will and helping to shape it. Party systems bring these multi-purpose organizations into direct contact with one another, providing an arena in which they can compete with one another or cooperate to achieve policy objectives and ensure party survival. Many argue (although their precise prescriptions differ) that the quality of democracy may depend largely on the characteristics of the political party system.

The Questions

To understand party systems researchers have asked a number of different questions. Giovanni Sartori's 1976 Parties and Party Systems, represents an early and still influential attempts to look at party systems in a systematic and comparative way. Sartori argues that categorizing party systems by the number of component parties can offer capture important characteristics of the complexity and internal dynamics of a party system. as long as it is supplemented by a measure of ideological dist ance, especially in systems with more than two parties. He argues that "three-to-five parties, viz., limited pluralism, have very different interactions than six-to-eight parties, viz. extreme pluralism" and further that, "the fragmentation of the party system can reflect either a situation of segmentation, or a situation of polarization, i.e. of ideological distance"[Sartori, 1976 #57, p. 126]. The combination of large numbers of parties and large ideological distance between them has clear consequences for democracy: "the chances of survival of the polarized polities cannot be appropriately assessed. Their "external" fragility and exposure to exogenous crises--such as inflation for Weimar, and Algeria for F rance--cannot be doubted and remain a constant.... The thing we know for sure is that if a polity is centrifugal at all levels--electoral, parliamentary, and party leadership level--then it is doomed: It can only, and quickly, end in deflagration"[Sarto ri, 1976 #57, pp. 144-5].

Mainwaring and Scully in their 1995 Building Democratic Institutions examine the institutionalization of party systems in Latin America. They by no means reject Sartori's criteria but argue that fragmentation and polarization are not sufficie nt to understand the role of party systems in the development of democracy: "Parties in systems with similar structural features (for example, two distinct cases of polarized pluralism) can behave in very different ways. Conversely, parties in the same s ystem can change their behavior markedly without any change in the elements that form the system"[Mainwaring, 1995 #8, p. 28]. These authors find an often more appropriate measure in "institutionalization," the "process by which a practice or organizatio n becomes well established and widely known, if not universally accepted" [Mainwaring, 1995 #8, p. 4]. To assess the degree of institutionalization, they offer four specific characteristics: 1) "stability in the rules and nature of inter-party competition," 2) parties with "somewhat stable roots in society," 3) all major actors according "legitimacy to the electoral process and to parties," and 4) the "independent status and value of" party organizations [Mainwaring, 1 995 #8, pp. 4-5]. The first and fourth characteristics listed here in fact each include two related but distinct characteristics. The stability of "rules" of competition differs from the stability of "nature of competition" by which the authors appear t o mean a stable roster of players of the democratic game. The fourth characteristic likewise combines the extent of party organization and the autonomy of that organization from party leaders. Separating these to form six characteristics makes the analy sis more complicated but considerably clearer, particularly with regard to party organization. Whether the relevant characteristics number four or six, Mainwaring and Scully argue, their presence or absence has played a role in the success or the failur e of democracies of Latin America. "[I]nstitutionalizing a party system," they argue, "is important to the process of democratic consolidation. Whether or not an institutionalized party system exists makes a big difference in the functioning of democratic politics. It is difficult to sustain modern mass democ racy without an institutionalized party system.[Mainwaring, 1995 #8 pp.1-2].

Herbert Kitschelt [Kitschelt, 1994 #12] offers yet another set of characteristics of party systems in his attempt to identify the "programmatically structured party system." Kitschelt's three categories largely reinforce but do not simply repeat tho se of Mainwaring and Scully. He looks for 1) "the extent of programmatic structuring of party systems", 2) "the dimensionality and configuration of the programmatic structuring" and 3) "the proximity and symmetry of elite judgments and the representative ness of party leaders' issue stances for the political preferences of their voters"[Kitschelt, 1994 #12, p.1]. In explaining the first of these characteristics, Kitschelt focuses particularly on the degree to which "parties announce identifiable and diff ering commitments to realize binding political decisions and collective goods they intend to deliver to society, were their representatives elected to political office with sufficiently strong electoral support. As an alternative to programmatic structuring, party systems can be based on charismatic leader-follower relations and/or clientelistic networks of exchange"[Kitschelt, 1994 #12, p.1]. He explains these alternatives not only in terms of Max Weber's categories for legitimacy but also in terms of specific types of incentives to voters and politicians, and he argues that charismatic parties in their inherent instability will either disappear when the charisma fails or develop a new source of legitimacy, usually clientelist in f orm. The clientelist form, while itself stable, may undermine popular support for the democratic system by its constant use of material incentives which suggest that democracy is about payment for services rather than political choice. Kitschelt's secon d category--the dimensions and configuration of programmatic structuring--echoes Sartori's measure of the degree of polarization while recognizing that parties may compete along a different axis than the one defined by by Sartori or, indeed, that they may compete on more than one axis. Kitschelt's final category of the correspondence between party elites and their voters echoes Mainwaring and Scully's search for the stability of party roots in society, though Kitschelt looks at the same relationship in terms of the "responsiveness" of elites [Kitsc helt, 1994 #12, p.2]. Together, according to Kitschelt, these three characteristics shape democracy because "the structuring of party systems is a critical element of democratic 'consolidation,' a state of affairs in which all relevant collective choices are subordinated to democratically elected players and in which most relevant political players do not see opportunities to defect from the democratic system of rules"[Kitschelt, 1994 #12, p.1].

Synthesis

These characteristics identified by Sartori, Mainwaring and Scully and Kitschelt compete in explaining what it is about party systems that is important for democracy. They do not necessarily conflict, however, and many of the suggested characteristi cs are quite similar, even complementary. In organizing the characteristics identified by these authors, the most apparent distinction lies in the level of analysis (See Table 1.). Certain of the party system characteristics identified by these authors about qualities possessed by individual parties in themselves. These questions seek to know mainly about the dynamics of party organization and support and can be assessed without knowledge about other parties. These characteristics can provide informat ion about the party system when the information about the individual parties is added together and patterns or trends become apparent. Within this category, further distinctions can help to clarify and simplify the characteristics of party systems to be explored. All of the characteristics mentioned first group refer to the internal relationship between party leaders, party organizations and party supporters. In questions #4a and #4b, Mainwaring and Scully exp lore the extent and role of party organization. They ask on one hand whether a party is "territorial comprehensive", "well organized" and has its own resources, and on the other hand how party decision-making is organized and the extent to which "there i s a tendency toward routinization of intraparty procedures, including those for gaining control of the party" such that parties is not "subordinated to the interests of ambitious leaders"[Mainwaring, 1995 #8, p. 5]. The other three questions in this cate gory concern the relations between the party and its likely voters. Mainwaring and Scully's question #2 seeks to find out how deeply voters are committed to a party, and to find answers they look at the stability of party voting from particular groups a nd the penetration of party loyalty into everyday life. Conversely, Kitschelt's question #3 about the correspondence between the views of party elites and voters can indicate, over time, how committed party elites relate to their actual or potential voting base. Finally, and closely rel ated to both of these questions is Kitschelt's' question #1 in which he seeks to know why voters support parties: programmatic, clientelistic or charismatic incentives or some combination of these.

Two additional characteristics do not concern parties per se but the environment in which they operate. Mainwaring and Scully ask perhaps the most fundamental and difficult question for party systems in still-consolidating democracies. In their ques tion #3 they seek to find out the whether "the major political actors accord legitimacy to the electoral process and to parties"[Mainwaring, 1995 #8, p.5] They do not specify who "the major political actors" are, but in the course of their explanation th ey mention both party and non-party political elites as well as the voting public as major actors. A solid answer to this question should provide some indication to this legitimacy among all of these various actors. In addition to this question, Mainwar ing and Scully also ask about the formal and informal rules which shape a party system from outside. Although they do not ask about the nature of these rules, they again add a time dimension by asking in their question #1a whether those rules remain stable over time, without which the other aspects of the party system would themselves be unstable.

In addition to party-specific characteristics which can be assembled to yield a general picture, and broader system characteristics which provide the framework for party systems, each of these authors also discuss characteristics which refer to the i nter-relationship between parties in a system. These are relational characteristics because, for example, it would be impossible to know the number, variability, or ideological configuration of party competition without looking at all parties. Wi th these questions as with those mentioned above, further distinctions can be made which can clarify research efforts. Perhaps the most basic inter-party characteristic is Sartori's question #1 about the number of parties. According to Sartori's, the nu mber of parties must be considered along with the ideological distance, hence Sartori's question #2 about party polarization. But to deal effectively with polarization it is first necessary to know the axis whose poles are being examined. Here Kitschelt's exploration of the basis of party competition in his question #2--his willingness to consider multiple axes and schemes other than left-right polarization--complement and update Sartori's earlier work. Finally, with their concern for institutionalization, Mainwaring and Scully add a time dimension to these considerations by asking in #1b whether the parties competing and the basis of their competition remains stable over time.

In the Czech Republic in 1996 it is these inter-party questions which are particularly interesting, as the two parliamentary and senate elections will largely concern the specific inter-relationships between parties. The remainder of this paper will attempt to understand how these four inter-party questions can be answered when looking at the Czech political party system. Progressing one by one through a list of eleven questions certainly violates the principle of parsimony, but serves also to demo nstrate that regardless of which theoretical framework proves better at explaining the particular relationship between party systems and democratic consolidation, developments in the Czech political party system point in a democratic direction.

Inter-Party Characteristics

Several of the critical characteristics of a party system appear in the types of relationships between parties. To understand the Czech Republic's politics it is crucial to understand not only how parties differ from one another but also how the eng age with one another in the political realm. Questions about the number of parties, the intensity of their polarization, and the stability of their competition help to sketch out the competition of parties in the Czech Republic. The first and perhaps cri tical question, however, is what these parties are competing about in the first place.

Basis of Party Competition

The 1992 article by Kitschelt which led to a rethinking of the axes of competition in new Eastern and Central European democracies attempted to disengage the automatic association of economic, political and cultural stances contained in "left" and "r ight"and to allow for the existence of other combinations, especially since the collapse of Communism removed one of the longstanding guideposts for orienting political spectrums from left to right. Yet although party competition in Eastern Europe may in some places be defined across more than one dimension, it appears that the Czech political party system has begun to approximate precisely the "redistribution" v. "market" conflict that Kitschelt argued might not necessarily appear as a main dimension of competition. The best evidence for this argument is a research project organized by Kitschelt himself.

Extensive interviews with party elites at the regional level in 1993 revealed that unlike its neigbors in the region, Czech politicians tend to differ primarily on whether property should be distributed by the market or by the state. In analysis of those research results, Kitschelt argues that the Czech Republic's party system is "clearly of a uni-dimensional nature"[Kitschelt, 1994 #12, p. 36] and that this dimension is consistent with a left-right socioeconomic scale. Further analysis finds some specific differences between parties on a scale from "libertarianism" and "cosmopolitanism" to "authority" and "tradition." These differences, however, existed primarily within the governing coalition of the Civic Democratic Party (ODS), the Christian De mocratic Party (KDS), the Christian Democratic Union-Czech People's Party (KDU-CSL) and the Civic Democratic Alliance (ODA) [Brokl, 1994 #202]. While these differences caused occasional conflict between these parties, the questions of authority and tradition proved subordinate to the broader socio-economic questions in the formation and maintenance of the coalition within parliament.

Considerable changes occurred in Czech politics and the Czech political party spectrum between the 1993 interviews and the 1996 elections, but it appears that developments have, if anything, confirmed the initial trends noted by Kitschelt and Brokl. Most significant in this regard is the ability of a one-dimensional left-right axis in predicting a variety of related characteristics about political parties. [At a very basic level it appears that where parties place themselves from left to right is a lso where respondents in mass surveys place those parties. This left-right location also corresponds to where party supporters tend to place themselves from left to right [Vecerni'k, 1996 #207, p. 8; Vlachova, 1996 #209, p. 9]. Perhaps most interesting ly, this same rank ordering also corresponds closely to how parties interacted with one another during and after the 1996 campaign. The three parties widely characterized as on the right of the political spectrum have continued in coalition, while the two parties characterized as being on the left of the spectrum have remained in opposition. Of the coalition parties, the party most often characterized as center-right--KDU-CSL--has also been the one most willing to coo perate with the center-left Czech Social Democratic Party (CSSD). The only party which does not fit easily into this scheme is the Association for the Republic-Republican Party of Czechoslovakia (SPR-RSC) which emphasizes cultural and racial questions an d criticizes all other parties for improper leadership and communist-ties. The unique position of the marginal SPR-RSC, however, emphasizes the orientation of parties on a socio-economic issue dimension with only occasional attention to cultural and auth ority questions.

The neighboring Slovak Republic provides an illustrative counter-example which clarifies the importance of left and right in the Czech Republic. Where other issue dimensions have strong influence and do not coincide with left and right, they tend to create apparently ungainly combinations. Both the coalition currently governing Slovakia and its predecessor united self-proclaimed parties of the left with self-proclaimed parties of the right, but which have found it possible, even necessary, to coope rate because they agreed on other issues which appeared more critical. In Slovakia those issues have included national development and methods of using political power and correspond closely to the dimensions that Kitschelt in 1992 predicted might emerge in Central and Eastern Europe. According to Kitschelt's research project, similar issues have also had an effect in creating a multiple-dimension political scene in Poland, Hungary and Bulgaria. The same questions have not emerged with such deep influence in the Czech Republic, leaving parties to orient themselves and their coalitions around questions of redistribution.

Party System Size

Counting parties consistently and reliably is not always an easy task. Difficulties arise in how to count coalitions, splinter parties, parliamentary clubs and fringe parties. Sartori suggests counting all parties in parliament which have coalition or blackmail potential. He regards other parties as not-relevant for assessing the 'size' of the system. Mainwaring and Scully in their more recent work on party systems argue that Sartori's rules are not self-explanatory and instead propose using math ematical formula for party system size devised by Laakso and Taagepera [Mainwaring, 1995 #8, p. 29]. In many ways, however, the raw calculation is not the most interesting result of attention to party system size. More interesting is how that number can be used in conjunction with other aspects of the party system, and how it can be viewed in comparison to other party systems. For both of these purposes the methods used by Sartori and by Taagepera and Laakso yield similar results in the case of the CR.

Applying Sartori's method to the Czech Republic the raises the question of time. Sartori's rules for relevance of parties depend on the influence of parties "over time"[Sartori, 1976 #57] and not in a single election. Time is one commodity that par ty systems in post-communist countries have not had in great quantity, and it is therefore necessary to make a judgment of party size on the basis of three elections in a six year period. This may not be enough to determine the relevance of certain margi nal parties. With this caveat in mind, it is possible to count the number of parties in the CR's party system after each of its three elections. In the election of 1990 four electoral blocs won seats in the Czech National Council. By 1991 the largest o f these, Civic Forum (OF) had split into five smaller party groupings of various size, the largest of which were the Civic Movement (OH) and the Civic Democratic Party (ODS). The other blocs had split or lost members as well, and the total number of independent parliamentary clubs reached as high as 11. The 1992 election brought eight electoral blocs into parliament, one of which quickly separated into its two component parties for a total of nine parties in parliame nt. Splitting and recombination again took this total to eleven by 1994. The transience and marginal position of most of these smaller groupings in parliament, however, makes it difficult to actually judge party system size during the parliamentary sess ions according to Sartori's criteria. For the parliament which was elected in 1996 the task is at least for the moment quite a bit easier. Six parties were gained seats in parliament in 1996. All six of these had also gained seats in 1992. The tenuous balance of seats in parliament has also given all six of these parties either "coalition" or "blackmail" potential. By Sartori's reckoning system, the 1996 elections created a six-party system in the Czech Republic

Unlike Sartori's method, the method used by Taagepera and Laakso has the mixed blessing of neither requiring nor permitting discretionary decisions about whether particular parties are "relevant." The results of both the Sartori and Taagepera and La akso methods for each discernably distinct period in the Czech Republic's parliamentary party system are listed in Table 2. Although the two sets of calculations yield different raw numbers for party system size, they are show almost precisely the same pa ttern: a party system which began with a quite limited number of parties but rapidly acquired many more. Both the 1992 and 1996 elections brought a reduction in the number of parties. Shortly after the 1992 elections, repeated splitting had wiped out t he reduction and the number of parties returned to its high level. In the four months since the 1996 elections, by contrast, no parliamentary deputies have formed new parties or even moved from one to another, and the party size remains at the same level as after the election.

Assessing the meaning of these numbers is difficult in the abstract. Sartori argues that more than five parties reflects a condition of "extreme pluralism" while five parties or less may be considered to be limited pluralism [Sartori, 1976 #57, pp. 127-128]. In seeking to make use of Sartori's categories without using what they view as potentially ambiguous counting rules, Mainwaring and Scully translate raw numbers produced by Laakso and Taagepera's formula into Sartori's categories as follows: " Systems with an Ns between 3.0 and 3.9 usually correspond to Sartori's category of limited pluralism, while those with an Ns of 4.0 or higher usually correspond to the category of extreme pluralism"[Mainwaring, 1995 #8, p. 32]. Using these translations o f Sartori's categories, both counting methods again yield extremely similar results. Both place the Czech system within the category of extreme pluralism beginning in 1991 and continuing until the present day. For both counting rules, however, the 1996 results are the lowest since the election of 1990 and are just slightly above the respective lower limits for systems defined by extreme pluralism.

Sartori argues that party size yields meaningful results only if used in combination with measures of party system polarization but he also acknowledges that the number of parties alone helps to shape how parties interact with one another. Before go ing on to discuss polarization, therefore, it may yield certain insight to compare the size of the Czech Republic's party system with those of other democratizing countries in Latin America and East Central Europe. The complexity of Sartori's method and the comparability and availability of existing data using Taagepera and Laakso's formula gives the advantage the latter method. The sacrifices made by using this method are softened somewhat by the fact that in the case of the Czech Republic the results obtained through method this prove quite similar to those obtained through Sartori's. For the purposes of better comparison with the results found by Mainwaring and Scully, I will calculate party size for Slovakia and other countries in East Central Europe only on the basis of electoral blocs in parliamentary elections, ignoring for the moment the presence of formal or "crypto" coalitions and changes occurring between parliamentary sessions.

Compared to the party system size of twelve Latin American democracies measured by Mainwaring and Scully over the last 25 years, the Czech Republic's six year average of 3.7 falls just above the median. Looking only at the most recent elections (as of 1993 for the Latin American cases), the Czech Republic's 4.1 in 1994 falls precisely in the middle of the group of twelve, below Venezuela, Peru, Bolivia, Chile, Brazil and Ecuador, although its party system is still larger than that of Colombia, Parag uay, Mexico, Costa Rica, Argentina, and Uruguay [Mainwaring, 1995 #8, pp. 30-31]. Within East Central Europe, the Czech Republic's party size of 4.1 in 1996 is nearly a full point higher than that of the Hungary after the election of 1994 and slightly hi gher than that of Poland after the election of 1993. Its party size is slightly smaller than that of Slovakia since its most recent election in 1994. Unlike the Slovakia where the most recent trend has been slightly upward and Poland where the trend has been sharply downward, the Czech Republic shares with Hungary a trend which points gradually toward a smaller system. Only in the Czech Republic has the downward trend come without a change in the electoral system designed to achieve that result.

Polarization

Sartori acknowledges that "fragmentation of the party system can reflect either a situation of segmentation or a situation of polarization, i.e. of ideological distance"[Sartori, 1976 #57, p. 126]. Sartori defines his category of "pola rized pluralism" as a pure or polar type on a spectrum bounded at its other end by one-party totalitarianism. Sartori notes eight features of this pure type:

• the presence of relevant anti system parties. . . .
• the existence of bilateral oppositions. . . .
• the center placement of one party, . . . or of a group of parties. . . .
• polarization. . . as a fourth, synthetic characteristic
• prevalence of the centrifugal drives over the centripetal ones. . . .
• congenital ideological patterning. . . .
• the presence of irresponsible oppositions. . . .
• politics of outbidding or overpromising. . . "[Sartori, 1976 #57, pp. 132-139].

The Czech political party system contains two parties which in certain aspects can be identified as a anti-system party, a party which "would not change--if it could--the government but the very system of government"[Sartori, 1976 #57, p. 133]. KSCM has frequently noted its committment to democracy, but more radical opposition on the part of some members to market economy and the party's willingness to engage in exceedingly sharp criticism of the current governing coalition and to threaten it with v otes of no-confidence have given it a reputation as a potentially system-changing party. At another extreme, though not necessarily one diametrically opposite to the KSCM, the SPR-RSC has gained a reputation akin to anti-sytem behavior on the basis of in flammatory remarks against the Czech Republic's Rom minority and Sudeten Germans. In its public statements the party openly rejects cooperation with other parties, all of which it regards as tainted by open or hidden ties to the Communist regime. Neither of these parties has openly and consistently advocated the abolition of the existing system of government in the Czech Republic but both have raised enough doubts about their support for the system as to deserve close attention. They also meet other criteria of Sartori by creating between them a potentially bilateral opposition on the question of opposition or support for the former Communist regime. By remaining out of power they likewise threate n to avoid the moderating responsibilities of governing.

For a party to gain from being out of power, however, it must be able to capitalize on the mistakes of those in power. In the Czech political party system the KSCM and SPR-RSC have remained out of power not only through self-exclusion but through deliberate exclusion by other parties. Although in need of a partner on the left of the spectrum, CSSD has consistently rejected offers of cooperation by KSCM. Likewise, although it at first advocated a proportional system for assigning parl iamentary leadership positions, CSSD eventually accepted a system which offered such positions to CSSD and denied them to KSCM and SPR-RSC. These two parties have likewise remained on the margins in public opinion (See Figure 4.). The demographic profi le of KSCM places it in danger of losing support through the simple mortality of its older-than average voters. In 1996 the party fell to third place from its rank as the second largest party in both the 1990 and 1992 elections. Its vote totals also declined significantly, and even combined with the votes from other competing post-communist parties such as LB and SDL, the party still fared worse than it had in the two previous elections. For SPR-RSC, the election results proved rather more fortuitious and the party achieved its highest share of votes to date. But neither in terms of votes or percentages did the party repeat its dramatic increase between 1990 and 1992 and whether the party c an stop further leveling off depends on an increase in the number of Czechs who feel threatened by foreigners, minority groups and the course of post-communist transformation. Beyond a sustained economic downturn, it may be difficult for SPR-RSC to expan d its support without becoming somehow less anti-system.

At present, neither of the potentially anti-system parties seems capable of exerting a sufficiently strong pull to drag politics to the extremes. Correspondingly the remaining parliamentary parties--which received the support of 70% of voters and wh ich comprise 80% of seats in parliament--appear determined to pull politics toward the center. They have deliberately excluded potentially convenient agreements from parties which they see as anti-system even though doing so creates the problem of obtain ing a 50% majority in parliament without the support of the 20% of deputies who represent either KSCM or SPR-RSC. This difficulty created the need for some degree of mutual support between CSSD and the parties of the coalition led by ODS. Thus CSSD depu ties abstained to ensure a successful vote of confidence for the ODS-led government of Vaclav Klaus, and ODS along with its coalition partners asceded to the appointment of Milos Zeman, President of CSSD, as chairman of parliament. ODS and CSSD have also proven their willingness to moderate their political positions in search of voters at the center, and ODS materials prepared for the 1996 Senate Elections emphasize "Certainty for Citizens," a di stinct sign of moderation for a party which has in the past publicly emphasized the importance of market principles. Thus while political debate is in no danger of disappearing in the Czech Republic, the centripetal tendencies of the major parties have p revented the relatively large number of Czech parties from spiraling outward in a manner which would threaten democracy.

Stability of Competition

With their focus on institutionalization, Mainwaring and Scully invariably ask whether the characteristics studied by others have an enduring character. When they ask about the stability of competition they refer both to party system size and the to basis of party competition. In a system of stable competition neither the players change considerably nor do the sides they play for. Since the election of 1990, the CR has experienced significant change in the roster of parties, but at least since 199 4 political competition has been quite stable with change has occurring largely on the margins.

To measure the stability of a party system's players, Mainwaring and Scully look at the volatility of party representation in parliament using Pedersen's index, a measure of the share of seats that change party hands from one election to the next. B ecause a high number of seats changed party hands during first two parliamentary terms, I will also calculate Pedersen's Index for party delegations at the beginning and near the end of each parliamentary term. Pedersen's index becomes more difficult to calculate when more than one party makes a plausible claim to be considered as the successor of the original party. In the 1992 election both ODS and the OH claimed the mantle of the 1990 Civic Forum (OF). Although OH took a slightly larger share of OF parliamentary deputies, public opinion surveys show that ODS drew a majority of former OF supporters [Krejci, 1994 #99, p. 242]. Later surveys from the 1992 election show that 55% of those who claimed to have voted for OF in 1990 voted for ODS in 1992 while only 7% of OF voters chose OH [Krejci, 1994 #99, p.220]. For purposes of calculating electoral and intra-parliamentary volatility, I will consider ODS as the successor of OF. A similar suc cessor problem occurred after the 1992 election when the Communist Party-led Left Bloc (LB) split into separate Left Bloc (LB) and Communist (KSCM) parties. Although it was the KSCM deputies which left the LB coalition, it was also the KSCM which retaine d the organizational continuity of the original party and the support of nearly all original LB voters. I will consider KSCM as a splinter from LB for the calculation of intra-parliamentary volatility, but as the successor of LB for the purposes of calc ulating electoral volatility.

Table 3. shows the overall change in seats and calculations of Petersen's Index for the periods during parliamentary terms and from each election to the next. The numbers point to two important conclusions. First, even though both intra-parlamentar y and inter-election volatility remained quite prominent in the Czech political scene, both of these phenomena declined significantly from the first to the second parliamentary term. They also declined significantly from the first half of the parliamenta ry term to the second. As Table3. indicates, intra-parliamentary volatility dropped to 10% during the second half of the term. Public preferences also gained some stability during that period. As Table4. indicates, the results of an IVVM survey during October of 1994 (like virtually any of the monthly surveys taken during the same period) serves as a excellent predictor of which parties would gain seats in parliament in the elections which were held 21 months later. The poll even offers a fairly accurate prediction of the rank ordering of the major parties, though it did not predict the significant decline in preferences for ODA or the significant increase for CSSD. Approximate calculations from public opini on polling data suggest that the results of almost any hypothetical parliamentary elections in the Czech Republic between early 1993 and summer of 1996 would not have differed from the actual 1996 results by more than 15% of all seats in parliament [Rendl ova, 1996 #201].

The second notable result of calculating volatility, particularly intra-parliamentary volatility, is to confirm the observation made by Mansfeldova [Mansfeldova, forthcoming] and others that most changes within the parliamentary party system actually occurred only within a relatively small group of parties. During the period from 1992 to 1996 the parties of the governing coalition, though they together held a majority of seats in parliament, accounted for little more than 10% of all seats changing h ands during this period and the Petersen's index for these parties likewise hovers near 10%. CSSD shares with the coalition parties the tendency toward a net gain of deputies, though the significance of its gain makes it appear rather volatile in compari son. The remaining exchange of seats during this period was concentrated among deputies originally elected in one of the four remaining parties--LB, HSD-SMS, SPR-RSC, and LSU--even though these accounted for only 40% of the total number deputies in parliament. Deputies from this group of parties were nearly five times more likely than those of the others to leave their party to join a different parliamentary club, to form a new club, or to become indepen dent. It is a potentially positive sign for future low volatility that with the exception KSCM and SPR-RSC, the remainder of these more volatile parties did not succeed in gaining parliamentary seats in the 1996 election. This electoral result both excl uded those members of parliament most likely to move from one party to another and sent a clear message that splinter parties within parliament--unless they boast an extremely strong organization, like KSCM, or a prominent leader, like Miroslav Sladek of SPR-RSC--do not offer a strong basis for continued political success.

The trend toward increased stability of competition both in parliament and in public opinion offers an important counterweight to otherwise worrisome comparative statistics. The CR's mean electoral volatility of 29.5% over two electoral periods far exceeds the overall mean of 8.6% for Western Europe between 1885 to 1985 according to a study by Bartolini and Mair (cited in [Mainwaring, 1995 #8, p. 7]) and is almost twice that of France, which with an average of 15.2% was the most volatile of the coun tries studied. In comparison with Latin America, the Czech Republic falls near the high end of Mainwaring's and Scully's index of lower-chamber volatility. The Czech Republic's volatility stands above eight of the twelve countries studied including Mexi co (22.4%) and Paraguay (25.8%) and is only slightly lower than that of Ecuador (32.5%) or Bolivia (33.0%) [Mainwaring, 1995 #8, p. 8]. Within Central Europe comparisons are more immediate but are made difficult by intervening factors. Poland and Hungary both experienced higher volatility than the Czech Republic between their first and second free elections but in both cases the volatility may be explained in part by significant changes to the ele ctoral system between elections. The Slovak Republic has had less volatile party competition than any of its neighbors, but this may have much to do with the relatively short two-year spans between its elections. It also, however, reflects a certain coh erence in many of Slovakia's parties which appeared to be lacking in many of the parties in the center and left of the Czech spectrum especially in the first years after the elections of 1992. The rise of the CSSD as a large and visible alternative on th e center-left of the spectrum holds the promise of more stable competition among Czech parties in the future as long as it can maintain its unity. CSSD leaders maintain that this will not be a problem, but with only ten of 61 deputies having been parliam entary deputies of the party in the past, the task of maintaining discipline may well prove difficult.

Conclusion

The political party system of the Czech Republic has achieved a generally positive development in each of the inter-party characteristics which are associated with democratic consolidation. While the party system is large, it appears to be gradually shrinking in size. While it possess anti-system parties, these have not taken advantage of the still-large number of parties and polarized the system as a whole. Competition between parties has been oriented toward the center on one main, relatively si mple to understand axis. The identies of the competing parties and their political positions also appear to be increasingly stable and clear.

The positive development in this system, however, still does not allow the luxury of complacency. Many of the recent development confirmed by the 1996 elections merely reflects the escape from a disadvantageous starting point. Furthermore in other aspects the Czech political party system continues to face potential problems. Articles by Kopecky and Simicek note continued problems with weak party organization and weak linkages between members and leaders in most parties. Here too, positive develop ments have begun to appear but they represent only a beginning.

Only in the system-wide characteristics which shape the space for the Czech political party system are conditions clearly favorable. Parties do determine who governs the Czech Republic and the rules which affect parties have changed only slig htly and in ways that makes the party system (though not necessarily each individual party) stronger. The majority electoral system which is to be used for elections to the Czech Republic's new senate will offer certain advantages for larger parties at t he expense of smaller ones and induce incentives for further centripetal competition. Aside from the legislative and constitutional powers of the senate, the introduction of a new body dominated by the two or three largest parties will provide those part ies with new visibility and a new pool of elites who can also serve in other functions. The biennial election of senators will provide local and regional party organizations with additional regular functions and encourage the development of those structures.

While Sartori, Kitschelt, Mainwaring and Scully all emphasize the establishment of a strong and institutionalized party system, none of them see this as an end in itself. Rather, establishing such a system is just the beginning, a method of ensuring the consolidation of democracy based on effective and lasting institutions. The real questions to be asked of democracy do not concern the institutionalization of the party system but whether that system contributes to the effective and just rule of the people by the people. Those questions are just now beginning to be asked in the Czech Republic.

Brokl, L. (1994, November 11, 1994). Politicka kultura a politicky prostor Ceske republiky. Lidove noviny.

Kitschelt, H. (1994). Party Systems in East Central Europe. Consolidation or Fluidity. Paper presented at the Annual Meeting of the American Political Science Associaiton, New York.

Kitschelt, H. (1995). The Formation of Party Cleavages in Post-Communist Democracies. Theoretical Propositions. Paper presented at the Workshop on Public Opinion and Party Formation in Post-Communist and Post Authoritarian Democracies, Duke University.

Kopecky, P. (1995). Developing Party Organizations in East-Central Europe. Party Politics, 1(4), 515-534.

Krejci, O. (1994). Kniha o volbach (Book about Elections). Prague: Victoria Publishing.

Lijphart, A. (1994). Electoral Systems and Party Systems. A Study of Twenty-Seven Democracies, 1945-1990. New York: Oxford University Press.

Mainwaring, S., & Scully, T. R. (1995). Introduction: Party Systems in Latin America. In S. Mainwaring & T. R. Scully (Ed.), Building Democratic Institutions: Party Systems in Latin America (pp. 1-36). Stanford, California: Stanford University Press.

Rendlova, E. (1996). Volebni Preference v unor 1996 . Institut Pro Vyzkum Verejneho Mineni (Institute for Research on Public Opinion).

Sartori, G. (1976). Parties and Party Systems. New York: Cambridge University Press.

Schumpeter (1942). Capitalism, Socialism, and Democracy. New York: Harper and Row.

Simicek, V. (1996). Vnitrni "zivot" politickych stran. Politologicky casopis, 3(2), 105-121.

Vecerni'k, J. (1996, September 16, 1996). Levice a pravice jiste, stred nejisty. Lidove Noviny, pp. 8.

Vlachova, K. (1996, September 6, 1996). Nekolik poznamek k mateni pojmu. Lidove noviny, pp. 9.

Method	June 1990	Mid-1991	June 1992	Oct. 1994	June 1996
Sartori	4	(11)	8	(11)	6
Taagipera and Laakso	2.2	7.3	4.8	6.4	4.1

Table 2. Party system size according to various measurement techniques.

Party	1990-1992		1992-1996
	Intra-Parliamentary 1990-1991	Electoral	Intra-Parliamentary June 1992-Oct. 1994	Intra-Parliamentary Oct. 1994-Mar. 1996	Electoral
OF-ODS/ODS-KDS/ODS	-86	-61	-1	+1	-8
OF-OH/OH/SD/SD-LSNS	+44	-	+5	+1	-
OF-OFN	+20	-	-	-	-
ODA	+10	+14	+3	-1	-1
LDS	+5	-	-	-	-
ROI	+5	-	-	-	-
KDS*	+8	+10	0	-4	-
KDU-CSL	-5	-3	+1	+8	+3
CSSD	-	+16	+3	+4	+45
KSC/LB/KSCM*	-7	-3	+10	0	-13
DL	+7	-	-	-	-
LB*	-	-	-10	-2	-
LSU	-	+14	-8	-6	-14
HSD-SMS/CMSS/CMUS	-2	-9	-5	+6	-14
SPR-RSC	-	+16	-9	-2	+2
Independent	1		+11	-5	-
Parliamentary Seats Changing Hands	100	68	33	20	50
Pedersen's Index
(Change/# of Seats)	50%	34%	16.5%	10%	25%

Table 3. Intra-parliamentary and Electoral Volatility of Parties in the Czech Republic, 1990-1996

Party	Public Opinion, Oct. 1994 (share of support among voters expressing a preference).	Election Results, 1996
ODS	32	29.62
CSSD	17	26.44
KSCM	7	10.33
KDU-CSL	8	8.08
SPR-RSC	5	8.01
ODA	13	6.36
DZJ	0	3.09
DEU	1	2.8
SD/SD-LSNS	1	2.05
LB	1	1.4

Table 4. Party Preferences (Oct. 1994) and Election Results (June 1996) in the Czech Republic. ~