CANDIDATE SELECTION METHODS

Transcription

1 PA R T Y P O L I T I C S V O L 7. N o. 3 pp Copyright 2001 SAGE Publications London Thousand Oaks New Delhi CANDIDATE SELECTION METHODS An Analytical Framework Gideon Rahat and Reuven Y. Hazan A B S T R A C T The framework presented in this article supplies tools for delineating candidate selection methods, defines what is meant by their democratization and offers an analytical framework for cross-national comparison. The first section of this article raises the problems of classifying candidate selection methods and suggests solutions for them. Each of the next four sections offers a dimension for the classification of candidate selection methods: candidacy; party selectorates; decentralization; and voting/appointment systems. The sixth section defines the process of democratizing candidate selection, and demonstrates its implementation in the three largest political parties in Israel prior to the 1996 elections, via the dimensions of the analytical framework. The subsequent section assesses the repercussions of this democratizing phenomenon in general and provides empirical evidence drawn from the Israeli experience in the 1990s. The article concludes by examining the ability of political parties to comprehend and to overcome the consequences of democratizing candidate selection. KEY WORDS candidate selection democratization Israel party organization primaries Developing tools for the study of candidate selection methods is important in two respects. First, when we study party politics, appropriate tools enable us to draw a map of a major element in the party s internal power structure. Second, if we claim that the behavior of parties is affected by the nature of the electoral system, then the behavior of individual politicians must be affected by the nature of the selection method. This means that without analytical tools such as those supplied by electoral systems researchers (for example, Rae, 1967; Taagepera and Shugart, 1989), we lack an important factor for analyzing party politics. In light of the personification of politics and changes in party organization that lead to increased autonomy for the (200105)7:3; ;016987

2 PA R T Y P O L I T I C S 7 ( 3 ) individual politicians at the top (Katz and Mair, 1995; Katz, in this issue), this gap needs to be filled through further research. On the one hand, candidate selection methods are less stable mechanisms than state institutional mechanisms (such as electoral systems), and are relatively more prone to change. As such, they should be seen as reflecting party politics. On the other hand, these changes are not frequent enough to justify an inclusive treatment of them as only a mirror. Thus, candidate selection methods should be treated as institutional mechanisms that both reflect the nature of the parties and affect party politics. Until now, only a few studies of candidate selection methods have offered tools for comparative analysis (Gallagher, 1988a, 1988c; Ranney, 1981). The framework presented in this article sets out to achieve two goals: first, to supply tools for delineating candidate selection methods and for defining what is meant by their democratization; second, to offer an analytical framework that will enable a cross-national comparison of candidate selection methods. Candidate selection methods are only one aspect in the more comprehensive analysis of candidate recruitment. As Norris (1997) points out in her study of legislative recruitment, a broad range of selection actors and processes outside the party arena should be addressed when one answers questions regarding who becomes a candidate, and how and why this happens. This study is not about candidate recruitment in general, but rather about a particular and important aspect of it inside the party arena. While the broader context of selection is certainly important, parties are critical to the process. Thus, this article offers an institutional anchor for the study of candidate selection methods. As an analogy, electoral systems are also but one aspect in the more comprehensive study of electoral politics, yet they still supply us with useful tools for their analysis. 1 The unit of analysis is the single party, in a particular country, at a specific time. Only in cases where several parties in a particular country use similar candidate selection methods (usually due to legal requirements), where a single party uses a similar candidate selection method more than once, or when both similarities occur, will we make generalizations about the selection system at the first stage of the analysis. The first section of this article raises the problems of classifying candidate selection methods problems that stem from having to deal with a number of complex features and offers solutions for them. The next four sections offer four dimensions, or continua, for the classification of candidate selection methods, and elaborate how these dimensions correspond to the phenomenon of democratizing candidate selection. The analytical framework is, therefore, built around four major questions: 1 Who can be selected? Are there any restrictions on presenting candidacy in a given party? If so, how strict are these limitations? How much do they affect the size and nature of the potential candidate pool? 298

3 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S 2 Who selects candidates? Are there any restrictions on participating in a given party s candidate selection process? If so, what is their impact on the size and nature of the selectorate? 3 Where are the candidates selected? Are candidates selected by a national or a sub-national selectorate? If candidates are selected by a sub-national selectorate, is it a regional or a local one? Does the party allocate positions for functional representation, i.e., are candidates selected formally as representatives of social groups or sectors? 4 How are candidates nominated? Is candidacy determined by a voting procedure or are candidates simply appointed? The sixth part of this article defines the process of democratizing candidate selection, and demonstrates its implementation in the three largest political parties in Israel prior to the 1996 elections, via the dimensions of the analytical framework. The subsequent section assesses the repercussions of this democratizing phenomenon in general, and provides empirical evidence drawn from the Israeli experience in the 1990s. The article concludes by examining the ability of political parties to comprehend and overcome the consequences of democratizing candidate selection. Problems of Classification and their Solutions The concept of safe positions on the candidate list, or safe seats when dealing with majoritarian systems, is used quite freely in the research literature. Since we also use this concept, we define it according to the following parameters. Although the size of the party s legislative representation is not known in advance intra-party selection is made before the general elections parties and politicians tend to relate to their party s actual representation as the one that distinguishes safe list positions from unsafe positions. As for new parties that did not compete previously, and thus cannot relate to any existing size, we are forced to estimate according to their projected size, through the use of opinion polls. 2 The tools offered in the following sections can be easily used when analyzing a simple, one stage, uniform candidate selection method. Such a simple method is one in which all potential candidates face similar restrictions, and the selection of all candidates is made by the same selectorate, at the same level, using the same nomination method. Empirically, however, we face complex candidate selection methods methods in which, simultaneously, different requirements are set for candidacy for different positions on the list; selection is made by different selectorates; these selectorates use different nomination systems; and some candidates are selected at different types of locations, defined territorially and/or functionally. We distinguish between two kinds of such complexities. First is the mixed 299

4 PA R T Y P O L I T I C S 7 ( 3 ) candidate selection system. This is a method in which different potential candidates face different restrictions; or different candidates are selected by different selectorates, in different locations, or according to different nomination systems. Second is the multi-stage candidate selection method. This is a method in which the same candidates have to face more than one selectorate during the selection process. These selectorates may be defined differently in territorial and functional terms, and may also use different nomination systems. In this kind of process, screening happens through the candidate selection process, and not only by candidacy requirements. 3 In cases of both mixed and multi-stage candidate selection methods, classification becomes more complex. As our goal here is to offer a crossparty and cross-national analytical framework, we must try to offer a way to integrate one-stage, uniform candidate selection methods and different complex cases, such as mixed and multi-stage methods, into the same framework. The difficulty with mixed candidate selection methods can be addressed in two steps. First, there should be a separate analysis of each selectorate its location and its nomination system. Second, toward the goal of summing up the system, the relative impact of each procedure should be weighted by calculating the ratio of safe positions that are filled by the particular selectorates. The complexity of a multi-stage candidate selection method should be approached in a slightly different two-step method. First, there should be a separate analysis of each stage: defining the selectorate; its location; and its nomination system. Second, the relative importance of each stage should be estimated. If certain stages are found to be merely formalities, then they should be removed from consideration. When more than one stage has a real impact on the composition of the candidate list in terms of the safe positions then these stages should be weighted to produce results that will enable us to locate the system along the continuum suggested. The following sections will offer examples of the operationalization of such solutions. Candidacy The first dimension addresses the question of candidacy: Who can present his or her candidacy in the candidate selection process of a single party at a particular point in time? The restrictions applied to potential candidates will be classified in a continuum according to the level of inclusiveness or exclusiveness (Figure 1). 4 At one end, the inclusive pole, every voter can stand as party candidate. Some states in the USA are close to this pole. This phenomenon can be attributed to the fact that state laws, rather than party rules, regulate the 300

5 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S Figure 1. Candidacy candidate selection process. At the exclusive pole we encounter a series of restrictive conditions. An example is Obler s (1974: 180) account of the requirements that applied to potential candidates in the Belgian Socialist Party. According to these restrictions, a potential candidate must: (1) have been a member at least five years prior to the primary; (2) have made annual minimum purchases from the Socialist co-op; (3) have been a regular subscriber to the party s newspaper; (4) have sent his children to state rather than Catholic schools; and (5) have his wife and children enrolled in the appropriate women s and youth organizations. The Selectorate The selectorate is the body that selects the candidates. It can be composed of one person, or several or many people, up to the entire electorate of a given nation. The selectorates are classified in a continuum (Figure 2) also according to their inclusiveness or exclusiveness levels. At one extreme, the selectorate is the most inclusive the electorate that has the right to vote in the general elections. On the other extreme, the selectorate or rather the selector is the most exclusive, a nomination entity of one leader. Between these two extremes, the selectorates are classified according to their amount of inclusiveness. Each of the categories, or zones, that appears in Figure 2 contains sub-categories that are located along each zone, which together produce the overall continuum. For example, American primaries are located in the electorate zone. Methods such as the nonpartisan primary and the blanket primary, in which every registered voter can vote for candidates from both parties, would be located near the inclusive end of the electorate zone (Ranney, 1981). American closed primaries, which demand voters registration according to their party affiliation before the day of the primaries, are located toward the exclusive end of the electorate zone. The exact location of American primaries will therefore be dependent Figure 2. Party selectorates 301

6 PA R T Y P O L I T I C S 7 ( 3 ) on the restrictions that are defined by the different state laws (see Kolodny and Katz, 1992: 909; Ranney, 1981). Still, the electorate zone will not be occupied only by American cases. According to Kristjánsson (1998), from 1971 on several parties in Iceland adopted open primaries. These were usually conducted in some of the electoral districts, and every citizen in these districts could participate. An additional case, located toward the exclusive end of the electorate zone, is that of the Spanish Catalan party, which opened its candidate selection to registered sympathizers non-members who can register as party supporters without paying any membership fee (Hopkin, in this issue). European closed primaries (Newman and Cranshaw, 1973), as opposed to American closed primaries, usually mean party primaries (Gallagher, 1988c: ), in which the selectors are party members the second zone of the selectorate continuum. In this party members zone, we find different methods of party primaries. Such methods were adopted to select both the party leader and the party candidates in the major Israeli parties in the 1990s (Hazan, 1997), and were used in Iceland in some parties (Kristjánsson, 1998). The selectorates in the party members zone can be distinguished according to the restrictions on party membership, the additional requirements that are placed on members with a conditional right to take part in the party selectorate, and the level of accessibility of the selector to the selection procedure. For example, one rule that could restrict membership, or just the right to participate in candidate selection, is the payment of membership dues. Members participation may also be restricted by the request for a minimal party membership period prior to candidate selection, proof of party activity, etc. Accessibility may also be an important factor in distinguishing between such methods. Levels of accessibility and inclusiveness are higher if a party adopts such methods as postal ballots, tele-voting or spreading polling stations all over the country. A less accessible and inclusive method is an open party convention. While all members can attend such a meeting, it requires more effort on their part. In the selected party agency zone we find various party agencies that may be distinguished by different parameters. Inside each party, the relative size of each agency is a sign of its inclusiveness: conventions are usually larger than central committees, which in turn are usually larger than executive bodies such as bureaus. The terminology used in each country is rarely equivalent, and hence one must be cautious when inferring the extent of inclusiveness based solely on what a particular party calls a specific agency. In addition, the more inclusive party agencies contain delegates selected by party members, while the more exclusive ones include representatives who were selected by such delegates. The more inclusive selectorates in the non-selected party agencies zone are, for example, special selection committees whose composition is ratified en bloc by a selected party agency. The more exclusive selectorates in this 302

7 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S zone are represented by a gathering of party founders, in new parties, or an informal gathering of faction leaders in older ones. Israel s ultra-orthodox religious parties serve as an example of a very exclusive selectorate. In one party, Degel HaTorah, one Rabbi was authorized to decide the composition of the party list. In another party, Agudat Yisrael, the list was formed by the Council of Sages a body of Rabbis who are the spiritual leaders of the party factions. The Belgian parties, from the 1960s until the 1980s, serve as an example of a mixed system, one that uses different selectorates for selecting candidates of the same party. Inside the large parties, some candidates were selected by party members while others were appointed by local and central party agencies (De Winter, 1988; Obler, 1974). Locating such cases on the continuum requires one to weigh the impact of each selectorate. If half of the candidates were selected by party agencies and half by party members, then when summing up for comparative needs one can locate the selectorate in-between these zones. Multi-stage candidate selection processes include, by definition, different selectorates. In the British parties, small executive party agencies filter candidates and/or have the ability to veto their nomination. Most of the influence is held by selected party agencies or party members who decide who will be the party candidate (Denver, 1988). While it may be the case that those who are screened have no chance of being selected in the first place, and while the veto is activated only on rare occasions, one must still consider the impact of this agency. Populist candidates that might have had a chance among party members may be screened or checked, and selectorates may be sensitive enough to refrain from selecting someone who may be vetoed by the executive agency. Thus, a party such as the British Liberal Democrats in which members select candidates, but party agencies screen them would be located on the exclusive end of the party members zone. Figure 3 integrates our first two dimensions, presenting the party candidate selection method according to the level of inclusiveness or exclusiveness of both candidacy and the selectorate. This combination becomes relevant when analyzing the democratization of candidate selection processes. A high level of inclusiveness in one dimension combined with a high level of exclusiveness in the other would mean that the party leadership and/or the party apparatus retain control over the process. For example, the Belgian Socialist Party used inclusive selectorates in many districts party members but retained control over the composition of its candidate list through very restricted candidacy (De Winter, 1988). The Italian Communist Party included non-members as candidates, but this was done under the supervision of an exclusive and centralized selectorate (Wertman, 1988). 303

8 PA R T Y P O L I T I C S 7 ( 3 ) Figure 3. Candidacy and party selectorates in candidate selection * Each state s position may vary based on different state regulations. BSP Belgian Socialist Party 1960s Israel 1 two main parties 1996 Israel 2 ultra-religious parties 1990s Decentralized Candidate Selection Methods Party selection methods may be seen as decentralized in two senses, parallel to the concepts Lijphart (1984) proposed when he dealt with the division of power in federal and unitary democratic regimes. Decentralization may be territorial, i.e., when party local selectorates nominate party candidates such as a local leader, party branch committee, all party members or voters in an electoral district. Decentralization of the selection method may also be corporate, i.e., one that ensures functional representation for representatives of such groups as trade unions, women, minorities, etc. While decentralization based on territorial mechanisms, in order to ensure regional and local representation, is fairly straightforward, more complex mechanisms are required for ensuring functional representation via decentralization. There are two mechanisms to ensure functional representation. The first is the sectarian or social group district, where the candidates and the selectors are members of the same sector or social group. The second is the reserved place mechanism, which guarantees a minimal position on the list (or minimal number of safe seats in the case of single-member districts) 304

9 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S for a candidate or candidates belonging to a distinct sector or social group. As opposed to the district competitors, those candidates who are eligible for reserved places compete for their place on the list against all of the candidates, and are selected by the same selectorate. The reserved representation mechanism is implemented only if the candidates do not attain the reserved position, or a higher one. The first mechanism decentralizes both candidacy and the selectorate, while the second one implies the decentralization of candidacy alone. When candidates are selected exclusively by a national party selectorate, with no procedure that allows for territorial and/or functional representation be it a non-selected leader, a national party-agency or an electorate that selects all candidates from the whole nation then we have a method that is located in the centralized pole (Figure 4). At the decentralized pole, candidates are selected exclusively by party local selectorates and/or intra-party social groups and/or sectarian groups. In many European cases, the selectorate at the district level plays the crucial role in candidate selection. The Norwegian case seems to fall close to the territorial decentralization pole. Not only can national party agencies not veto candidacy that is determined at the district level, it is also the case that territorial representation is taken into account inside each district (Valen, 1988). Once again, we have to determine and weigh the impact of different selectorates at different levels in the case of a multi-stage selection method. A case in point is that of the Italian parties in the 1980s, in which central, provincial and local level selectorates took part in candidate selection. According to Wertman (1988), the provincial-level party agencies played the main role in candidate selection, vis-à-vis the center and the district level selectorates. Thus the Italian parties of the 1980s are in the middle area of Figure 4. Centralization and decentralization of candidate selection 305

10 PA R T Y P O L I T I C S 7 ( 3 ) the territorial continuum. Still, there were differences between the parties, which placed the Italian Communists, for example, closer to the centralized pole than their Socialist counterparts. Belgium supplies us with examples of both of the functional representation mechanisms which were used at the district level. In the Belgian Christian Social Party in 1961, the reserved place mechanism was used when it was decided that in some of the Brussels districts, Flemish and Francophone candidates would get every other seat on the party list. In 1965, separate intra-party sub-districts were actually established when Francophone and Flemish Party members in these districts selected, separately, Francophone and Flemish candidates for parliament (Obler, 1974). Voting Systems and Appointment Systems When the selection process includes a procedure in which votes determine whether someone will be placed as the party s candidate in the general elections, and his or her position on the list, we are dealing with a voting system. It should be noted that a voting procedure may be used by an appointment body of two people or more. However, this would not be considered a voting system unless two conditions are met: First, each candidate must be determined exclusively by votes, and not, for example, by an agreed-upon list or an allocation that is ratified by unanimous or majority vote; second, voting results must be presented officially to justify and legitimize the candidacy. When candidacy is determined without using such a voting procedure, we refer to this as an appointment system. In a pure appointment system, candidates are appointed with no need for approval by any party agency, or other organ, except the nominating organ itself. In a pure voting system, all candidates are selected through a voting procedure, and no other selectorate can change the composition of the list. This distinction is crucial when dealing with the level of control a party holds over the composition of its candidate list (Figure 5). When the list is appointed, its composition can be controlled. The list can express the balance that the party organs think is appropriate in response to both inter-party (electoral image, personal popularity) and intra-party (loyalty, factional, social and sectarian representativeness) pressures and demands. On the other hand, when the list is selected, the party organs do not control its composition, which is instead determined according to the aggregation of individual votes. Parties that use voting systems also tend to use representation correction mechanisms, such as functional and territorial districts or reserved seats, and multi-round and PR voting systems, to ensure a balanced list. There are also cases that can be located in-between these extremes. These appear in the continuum as appointment-voting systems. Such is the en bloc ratification vote that was used in Belgium. In many constituencies, party 306

11 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S Figure 5. Candidate nomination and party representational control members were asked either to vote for a model list a list of candidates determined by a local party agency or to express their preferences regarding the candidates. Only if more than 50 percent of party members did not ratify the model list were the other votes counted, and thus they did not have much weight (Obler, 1974). In Norway, the mix was more open to change in the ratification process. Lists that were recommended by a nominating committee were then ratified by a majority of a selected party agency, position by position (Valen, 1988). In multi-stage selection methods, it is possible that both appointment systems and voting systems will be used at different stages. In the British Liberal Party, a short list was appointed during the initial stage of the candidate selection process, but the final decision concerning those on the shortlist was determined by party members votes. There are also systems that mix both nomination systems: some candidates are appointed, while others are selected. In Belgium, some of the candidate lists of the three large parties were determined by member vote, others by member ratification of a regional agency list, and a few others were appointed by a national agency (De Winter, 1988; Obler, 1974). It is usually the case in small and more exclusive selectorates that candidates are appointed, while larger selectorates vote in order to choose their candidates. Still, these are two (almost) mutually exclusive categories. A voting system can, theoretically, be used in a selectorate of two or more people, and appointments can take place in bodies that include several dozens of people. Voting systems can be further distinguished on the basis of two elements (Table 1). The first is the position allocation formula, i.e., proportional representation (PR), semi-pr, semi-majoritarian and majoritarian systems. The distinction between these four kinds of voting systems is based on their potential level of proportionality. Proportional voting systems in this context will usually be personalized. For example, the three largest Irish parties in the 1980s, except for Fianna Fáil, used a personalized PR system to determine the composition of their candidate lists the single transferable vote (STV) system that was also used in the general elections. Semiproportional systems are those in which the number of votes each selector has is smaller than the number of safe seats being contested. This is the intraparty version of a limited vote electoral system. Semi-majoritarian systems are defined as systems where the number of votes that each selector receives 307

12 PA R T Y P O L I T I C S 7 ( 3 ) Table 1. Appointment systems and voting systems Category Sub-category Examples* Appointment Systems Pure Israeli ultra-orthodox parties Shas, Agudat Yisrael, Degel HaTorah ( ) With en bloc ratification Belgian BSP (partial usage of the Model List 1960s 1980s) Appointment-Voting With ratification and Norway (1960s 1980s) Systems correction possibilities Voting Systems One-round Majoritarian British Labour and Conservatives, LDP (1980s); Israeli Likud (1996) Multi-round Majoritarian Irish Fianna Fáil (1980s); Israeli Likud (1992), Shinui** (1992), and Mapam** (1992) One-round Semi-majoritarian Belgian Socialist Party (partial usage 1980s); Israeli Meretz (1996) Multi-round Semi-majoritarian One-round Semi-PR Israeli Labour (1992, 1996, 1999), Likud (1999), and Meretz (1999) Multi-round Semi-PR Israeli Ratz** (1992) One-round PR Irish Fine Gael, Labour, and Progressive Democrats (1980s) Multi-round PR * Data from Belgium, Obler (1974) and De Winter (1988); Norway, Valen (1988); Britain, Denver (1988); Ireland, Gallagher (1988b). ** Three parties that together composed the Meretz list established in is higher than the number of safe seats contested. While such a system is majoritarian as a majority block can be organized and can take over all the safe positions it is semi in the sense that incentives for organizing a plurality or majority block vote are weaker. 5 In a majoritarian system, the number of votes and safe seats is equal. In many cases, every position is contested separately, making the system almost parallel to single-member district elections. The second parameter distinguishes between one-round or multi-round selection methods. In the former, all safe positions are selected at one time, whereas in the latter the safe positions are filled gradually. The importance of this difference lies in the opportunities to control and/or balance the composition of the lists that a gradual selection process gives the parties. There is a connection between the voting system used in the final stage of the candidate selection process and the national electoral system. Where national elections are conducted in single-member districts, the voting system used in the candidate selection process must be majoritarian in order to produce a single candidate. For example, the exhaustive ballot was used by the British Conservative and Labour parties in the final stage of their 308

13 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S selection process, 6 while the Liberals used the majoritarian method of the alternative vote (Denver, 1988). On the other hand, when general elections take place in multi-member districts, the voting system need not be majoritarian. For example, in Ireland, the exhaustive ballot was used by Fianna Fáil in order to determine its candidate list, position by position, while the next three largest parties used a one-round STV method. There is also a connection between the selectorate size and the use of either one-round or gradual selection. In smaller selectorates, it is possible to adopt either method. However, when the selectorate is larger especially in those cases where it includes all party members or the entire electorate logistics make the use of one round almost a must. Democratizing Candidate Selection Methods Many parties now afford their ordinary members a greater voice in candidate selection than was once the case. (Mair, 1994: 15) Democratization of the candidate selection process is expressed by widening participation in the process, i.e. when the selectorate that is adopted following a reform of the candidate selection method is more inclusive than the previous one. Such a reform might immediately affect the nature of the nomination system, since the use of a selectorate that is larger than a few dozen usually requires adopting a voting procedure. Adopting more inclusive candidacy requirements, and either territorial or functional decentralization, may be labeled democratization. However, as long as the selectorate remains as exclusive as it was, these reforms cannot be seen as a true democratization of the candidate selection process, for two reasons. First, despite more inclusive candidacy requirements, the same limited selectorate still has full control over the final results. Second, decentralization might mean only that control of candidate selection has passed from the national oligarchy to a local oligarchy. 7 Only if decentralization encompasses a more inclusive selectorate can it be considered a democratizing process. In other words, decentralization can limit, maintain or expand the extent of intra-party democracy. The candidate selection methods adopted by certain parties in Israel, prior to the 1996 parliamentary elections, present an excellent case for the classification of candidate selection methods based on the dimensions elaborated in the analytical framework above. Moreover, the Israeli case can also be used in order to assess the causes of democratizing candidate selection, the extent of this democratization and its consequences. The Israeli party system is extremely multi-party in nature, and exhibits a wide variety of selection methods (Bar, 1996; Hazan, 1997; Rahat and Sher-Hadar, 1999b). This high variance stems from three conditions. First, although a Parties Law was adopted in Israel in 1992, it does not refer to 309

14 PA R T Y P O L I T I C S 7 ( 3 ) the way candidates should be selected. Therefore, in preparing their lists of candidates, parties are not legally bound to adopt any specific kind of candidate selection method. Second, the electoral system is a fixed list system with a single national constituency. Thus, when designing their selection methods, parties are not as pressured to respond to the electorate s interests as are parties in open list and personal electoral systems (Bogdanor, 1985; Carey and Shugart, 1997). Third, Israeli society is very heterogeneous, and many socio-political cleavages are reflected in the party system, resulting in many parties that represent different political sub-cultures. This section focuses on the democratization of candidate selection in three Israeli parties (or lists) prior to the 1996 elections: Labour, Likud and Meretz. These were the three largest outgoing parties of the 12th Knesset (1992 6), Israel s parliament. Three of the dimensions, or continua, presented thus far will be elaborated for each party. 8 Although candidate selection processes have undergone democratization in Israel since the 1970s a time when nomination by a closed inner circle of party leaders was the rule the pace of this development has been different from one party to the next (Brichta, 1977; Doron and Goldberg, 1990; Goldberg, 1980, 1994; Goldberg and Hoffman, 1983; Hazan, 1997). It is in the arena of the selectorate that the existence, or lack thereof, of a democratizing trend is of acute significance. The Israeli parties cover almost the entire spectrum of possible selectorates. At one end, near the exclusive pole of the continuum, we find the ultra-orthodox religious parties. The power to nominate these parties candidates lies in the hands of a single Rabbi, or a council of Rabbis, whose authority is based on their role as religious leaders. Close to the other end we find the three largest parties prior to the 1996 elections, all of which significantly expanded their selectorates and adopted party primaries. In 1992, the selection process of the Meretz alliance (three dovish parties: Ratz, Mapam and Shinui) was conducted solely by the central committees of the three constituent parties. In 1996, however, Meretz expanded its candidate selection process, and began to conduct it in two rounds. In the first round, the constituent parties three central committees (which numbered several hundred people each) filtered the candidates, i.e. each produced a short list of approved candidates from among those who presented their candidacy. In the second round, all Meretz members ranked the candidates from the three approved party lists and produced the final list of candidates it was this round that determined whether a candidate would be placed in a safe or unsafe position. In this second round, each member was given more votes than the number of safe seats. Thus, the more important round of the multi-stage method used by Meretz employed a semimajoritarian voting system. Meretz was, therefore, placed between the party members and the selected party agency areas of the selectorate continuum (see Figure 2), but was located closer to the former because the more inclusive selectorate had the more decisive role. 310

15 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S Israel s two major parties, Labour and Likud, are found even closer to the inclusive end of the selectorate continuum. Labour had already adopted a more inclusive candidate selection method in 1992 in 1988 it had still allowed its central committee to choose the list of candidates which it maintained in Partially in response to Labour s victory in the 1992 elections, Likud also democratized its candidate selection method and no longer allowed its central committee to produce the party list. In 1996, both parties held single-round party primaries in which all duespaying members were given the sole right to determine both the composition and the rank of the parties candidate list. These two parties were thus located in the party members area of the continuum, and used a pure voting system. However, the voting system used by Labour to select its national list was based on a semi-proportional allocation formula, while that of Likud was majoritarian. 9 Labour gave each of its members 11 to 15 votes, for what were expected to be approximately twenty safe seats. Since the party was expected to win more seats than the maximum number of candidates any member could vote for, this was a limited vote and was thus semiproportional. Likud, on the other hand, gave each member 19 to 20 votes, for what were also expected to be about twenty safe seats. This was indeed a majoritarian winner-take-all system. Since Israel s electoral system has only one national constituency, it is difficult to speak of territorial decentralization, but it is not impossible. Both Labour and Likud circumvented the national electoral system by decentralizing their candidate selection processes through the establishment of geographical districts (Hazan, 1999). For example, the major urban centers of Jerusalem, Tel Aviv and Haifa were each a territorial district in both parties, with secured positions on the party lists for whoever was selected in them. Only party candidates and members who were residents in the geographical district were allowed to run and vote. In addition, both Labour and Likud adopted functional districts as part of their candidate selection methods, characterized by geographically dispersed social group or organizational affiliation. Labour had two sectarian districts for its non-jewish members (Arabs and Druze), and two organizational districts for those living in cooperative or collective settlements (moshavim and kibbutzim); Likud had a single organizational district for those residing in regional councils or agricultural settlements. In both the geographical and the functional districts, it was the selectorate at the district level that played the only part in the candidate selection process. All three party lists also reserved positions on their lists for representatives from a particular social sector or a sub-group. Meretz, for example, used such functional decentralization to guarantee the representation of women and minorities (i.e. Arabs). At least three of the first thirteen candidates on the list had to be women, and at least one of the first ten had to be an Arab. Labour reserved positions on its final lists of candidates for women 311

16 PA R T Y P O L I T I C S 7 ( 3 ) and immigrants. Likud reserved positions for women, for an immigrant, a non-jew and a young candidate. These reserved positions were implemented only if the candidates could not reach them on their own. For example, if no female candidate won a high enough position in the party primary, the correction mechanism would then be implemented and the highest placed female candidate would be promoted to the higher reserved position on the list. Together, the geographical and functional districts, along with the reserved positions, made up over one-half of the safe seats in the Labour list for the 1996 elections, and over one-third of those in the Likud list. The parties thought that by democratizing their candidate selection methods, they would not be able to maintain representative lists, so they designed corrective mechanisms in order to circumscribe the decision made by their now more inclusive selectorates. In the decentralization of the candidate selection process based on districts, both candidacy and the selectorate were decentralized. That is, candidates and selectors were members of the same decentralized geographical, social or organizational group, and thus formed particular constituencies within the party. In the decentralization based on reserved positions, on the other hand, the candidates were the representatives of the social group, while the selectorate was the entire party membership. In other words, the territorial and functional district candidates had to compete within their own constituencies, while the reserved position candidates competed with the national candidates for the votes of the entire party selectorate. However, despite the more or less extensive elements of decentralization in these aspects of the candidate selection process, neither necessarily exhibits a true process of democratization, as defined here. The Consequences of Democratizing Candidate Selection In a number of parliamentary democracies, parties are now in the process of democratizing their candidate selection methods. The important question that arises from this new phenomenon is: Are parties still able to exert an important impact on the candidate selection process? If party lists are assembled not by the party organs, but, for example, by a more inclusive selectorate, such as the party members, the result could be the expunging of one of the more important functions of parties in parliamentary democracies. When the party leaders and organization can be circumvented by the politicians, it weakens the party and hampers its ability to aggregate policies and to present a cohesive ideological image. The ability of prospective politicians to appeal directly to the party membership thus changes the bases of both legitimacy and responsibility of the (s)elected party representatives. The result could be a drastic weakening of partisan discipline and cohesiveness, leading to a decline in the ability of the parties to function as a stable 312

17 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S basis for the political process and to operate effectively in the parliamentary arena. Former Minister Ze ev Begin (1996: 208), who won one of the top positions in the Likud Party primary in 1996, argued that The party organs... are essential for the creation of a political platform, social or economic, and for its promotion and implementation by the party s MKs [Members of Knesset]. Party discipline, which is so easily ridiculed, is a necessary condition for the proper functioning of a democratic regime. A random collection of MKs, who pander to their voters, each using more lowly tactics, is a dangerous obstacle. However, if the parties maintain, or reassert, control over certain phases in the candidate selection process, the phenomenon of democratization need not lead to a loss of control for the party organization, nor to a decline in its functional capacities. For example, if the party can filter the prospective candidates prior to allowing the more inclusive selectorate to rank them, or if its agencies produce the final list after the more inclusive selectorate has voted for the prospective candidates, the party can still remain the master of its internal fate. The democratization of candidate selection, if unchecked, can present a danger to the stability of both political parties and governing coalitions and, as a consequence, to the stability of parliamentary democracy in general. This phenomenon presents a paradox: Political parties, which are the functional bodies that operate in the democratic arena, should not become too internally democratic themselves. That is, while parties must try to avoid any constraints on democracy in the inter-party electoral arena barring those presented by thresholds, allocation formulas and restrictions on the inclusion of particular extremist parties they must at the same time implement constraints on democracy in the intra-party electoral arena in order to maintain control of their candidates and legislators. Candidates who are chosen by an inclusive selectorate owe their loyalty to their voters in the candidate selection process, and not only to their party. Such candidates are no longer assured of a future in politics by being loyal team players; instead they must stand out and be recognized not by the party leaders but by their inclusive selectorate. Democratizing candidate selection produces dual sources of legitimacy for candidates party legitimacy and popular legitimacy. The immediate results are: (1) a shortening of the political time-frame based on a constant fixation with elections; (2) a behavioral dependence on an amorphous group known only as the voters (i.e. the selectorate); (3) a tendency to act in a manner that largely disregards the group(s) with which the candidate is associated (i.e. the party, coalition, opposition, etc.); (4) a drastic increase of the rather basic political trend toward individualist and populist politics; (5) a significant growth in the influence of the mass media on politics in general, and on the candidate selection process in particular; and (6) a need to enhance financial resources in order to reach a wide voter base. 313

18 PA R T Y P O L I T I C S 7 ( 3 ) Former Labour Minister Uzi Baram (1996: 215), who won the top position in his party s primaries in 1996, aptly described the new situation in the Israeli parties: With the change of the intra-party selection method, the political parents were also instantly replaced. Instead of the 1,200 Party Center members, and a number of leaders managing the apparatus, the MK suddenly adopted 150,000 new parents. They chose him for the position of MK, and until his re-selection four years later, he is responsible only to them. Transferring the functional responsibility of choosing party candidates to a more inclusive selectorate could expand the influence not only of the individual support base of the candidates but also of well-organized groups, thus producing lists that include both independent and special interest candidates. In short, the narrow personal and special interests of a candidate selected by a more inclusive selectorate could overcome the more general party interests and the even wider voter interests that politicians must take into account. A legislature with parties that have openly democratized their candidate selection methods could splinter the legislative process because each legislator will attempt to satisfy numerous, divergent and even conflicting interests without party discipline circumscribing such a development. The disaggregation of parties, both institutionally and at the parliamentary level, finds clear evidence in Israel. After the democratization of candidate selection, one could witness the breakdown of the disciplined and highly institutionalized parties, which in the past were capable of mandating their legislative representatives to behave as a cohesive group dedicated to the pursuit of party policies and goals. One direct consequence was the whittling away of executive dominance in the sense of effective governmental control of the parliamentary agenda. For example, in the first 12 Knessets ( ), prior to the dramatic democratization of candidate selection, private members bills constituted 15 percent of the bills passed per Knesset. But in the 13th Knesset (1992 6), after Labour had already instituted primaries and other parties had decided to adopt this method of candidate selection as well, there was a dramatic increase, both absolutely and relatively, in private members bills: 54 percent of the bills passed were now private members bills and only 46 percent were government bills. Figure 6. Candidate selectorates and party cohesion 314

19 R A H A T & H A Z A N : C A N D I D A T E S E L E C T I O N M E T H O D S Moreover, of the private members bills passed, 33 percent were sponsored solely by MKs from opposition (non-coalition) parties, 42 percent by MKs from government (coalition) parties, and 24 percent were sponsored jointly by both government and opposition MKs. In the 14th Knesset (1996 9), 51 percent of the bills passed were private members bills and 49 percent were government bills. However, there was a 60 percent increase in the proportion of opposition private members bills passed. In all, 31 percent of successful private members bills were sponsored by government MKs, 14 percent were jointly sponsored, and 55 percent were sponsored solely by opposition MKs, which indicates that the challenge to executive dominance increased perceptibly during that period. 10 To the party organization and the individual legislator levels must be added the mass level. Empirical examples in the context of the 1996 primaries show how this method of candidate selection had a negative impact on the quality of party membership, and helped advance the internal demise of the parties. For example, candidates running in the primaries of one party attempted to recruit as many party members (and hence potential voters) as possible even from other parties thereby pushing the pursuit of quantity over quality to an extreme. Moreover, some of those who joined a party in order to support a particular candidate in the primaries and thereby influenced the party s list for the Knesset never intended to vote for that party in the subsequent Knesset elections. Partisan identification, beyond the momentary and the instrumental, seemed to matter little in the primaries. Indeed, both major parties had constituencies where the number of their dues-paying members was higher than the number of votes they received in the national elections only two months later (Rahat and Sher-Hadar, 1999a). The rather apparent instrumental relationship between party members and the party organization is further illustrated by the fact that the number of party members increased dramatically during the primary year and then sharply dropped off afterwards, thereby showing that membership had become merely the way by which one could vote in the primaries, and no longer a permanent link between the voter and the party. This increase and immediate decrease in party membership is similar to the more general phenomenon of party mobilization in the run-up to elections. However, in other countries, the rapid growth in party membership prior to elections has little to no effect on the makeup of the party list. It is rather a mobilization of forces to support the election campaign. In Israel, the party members were those who were responsible for producing the major party lists prior to the election of the 14th Knesset, but they were not necessarily those who actively participated in the parties campaigns. The ramifications of combining party mobilization and party primaries are, therefore, both different and more significant than the more general phenomenon of party mobilization alone. 315