Inter-party mobility in the Turkish Grand National Assembly: curse or blessing?
?lter Turan, ?eref ?baand Ay?e Zarakol
Thematic Issue N°3 , Being a MP in contemporary Turkey.
Historically, there have been high levels of inter-party mobility in the Turkish parliament. The military makers of the 1982 constitution felt that party changing not only promoted political instability and was unethical, but they also feared that it might undermine the two party system they were trying to build. They introduced measures to prevent it. These failed to stop the mobility of deputies who found many ways to circumvent restrictions to move between parties. The restrictions have since been repealed. Such mobility has contributed to a peaceful transition from military to civilian rule. New parties have been formed, others have merged through party changing, while parties closed down by the constitutional court have found an opportunity to get reorganized as a parliamentary party. Inter-party mobility done for political and personal benefit, on the other hand, raises questions of political and personal ethics. Party changers and non-changers do not appear to have different attributes. Political contexts influence party changing decisions by individual deputies and factions.
An examination of the records of the TBMM (Turkish Grand National Assembly) reveals that from 1983, the year marking the return to civilian politics after a military intervention that lasted three years, until the summer of 2005, no less than 812 members of parliament have left the party on whose ticket they had been elected1. This figure includes three different contingencies: an individual deputy may leave his party to become an independent or to join another; a party may be closed down by the Constitutional Court for having violated the Political Parties Law whereupon some of the now ?partyless? deputies may join another or proceed to establish a new party2, and finally two or more existing parliamentary parties may unite to form a new one. The incidence of party changing, that is the number of times deputies may have changed parties during the same period, is much higher than the above figure deriving from the fact that an individual may change his party several times during one legislative term3. Inter-party mobility in the Turkish legislature is not unique to the post-1983 period. It has been a frequent occurrence ever since the transition to competitive politics in 1946 (Turan 1985) although it appears to have become more commonplace in recent years.
Table 1: Number of deputies changing parties*
*The data have been compiled from the records of the Turkish Grand National Assembly, Division of Laws and Resolutions.**The size of the parliament was increased from 400 to 450 in 1987 and to 550 before the 1995 elections.***The numbers are as of August 2005. A few more deputies have since left both government and the main opposition parties.
 Inter-party mobility in a legislature is not confined to Turkey. The realignment of party preferences in the American South, for example, has been realized in part through congressmen and senators who have changed their parties. Usually in their case, however, the Democratic incumbents have simply preferred to run on the Republican ticket at the next election. They have not changed their party affiliation during the legislative term for which they have been elected. Similarly, in Britain, the Liberal Democratic Party was formed in 1988 by a union between the Liberal Party and the Social Democratic Party, the latter already having some parliamentary representation4. Yet, these appear to be occasional developments that usually occur as part of an overall realignment process in the party system. Their occurrence would not, for instance, allow us to characterize either the United States or Britain as systems with high inter-party mobility in their legislatures. The Turkish case is untypical both in magnitude and the richness of reasons, contexts, styles and methods that are associated with it. Therefore, it merits closer examination.
 This article will examine inter-party mobility in the Turkish Grand National Assembly from 1983 to the current period. During the period under study, in addition to what may be called routine party changing that was observed extensively during the 1946-1980 interim (Turan 1985), three types of non-routine movement between parliamentary parties have occurred. First, the military junta that assumed political power in 1980 proceeded in 1981 to abolish all existing political parties and try to build an entirely new party system. After the 1983 elections, Turkey went through a process of normalization of political life that included the partial restoration of the pre-1980 parties. This included considerable inter-party mobility as deputies elected on the tickets of the military approved parties moved to the re-established pre-1980 parties that had deeper roots in society. Second, on several occasions, the Turkish Constitutional Court closed down political parties for having violated the provisions of the constitution and the Political Parties Law which ban parties, among others, from using religion for political ends and inciting ethnic divisions. The deputies of the defunct parties went on either to join others or to form new parties. Finally, in the process of the post-1983 reconstruction, some parties united with others.
 Routine party changing, by contrast, occurs under conditions that may be characterized as ?politics as usual?. At any given moment, an individual deputy may want to leave his parliamentary party and join another one for a variety of reasons (Turan 2003: 164) having to do with ideology, constituency service, conflicts in his electoral district or self interest such as reelection. Similarly, a group of deputies such as members of a faction may choose to depart their party and move to establish a new one if they lose hope in their current party or if they sense that they can do better by starting a new party (Turan 1985: 24).
 Routine and non-routine inter-party mobility may take place within a parliament simultaneously. Our data does not allow us at this time to judge whether an act of inter-party mobility committed by a deputy is routine or non-routine. It may sometimes be difficult to classify a particular event except in ex-post fashion, but behavior usually occurs within a political context the parameters of which are known. With further refining of our data, we hope that in the future we shall be able to differentiate the routine and non-routine changers. In this article, both types of mobility will be examined. The first section of the article will take up the specific context of non routine mobility. Next, it will proceed to examine the general context of routine mobility with emphasis on the characteristics of political parties and the party system that seem to favor party changing. It will conclude by looking at the characteristics of the party changers themselves.
 The MGK (National Security Council) comprising the Turkish chief of staff and commanders of the four branches of the military ? which in Turkey includes the national gendarmerie ?, took over the government of Turkey on September 12, 1980. The military takeover had been expected. The party system had experienced increasing polarization since the late 1960s. After 1973, a series of coalition governments had ruled the country in an haphazard way under very high inflation rates, worsening economic conditions and intensifying public disorder that resulted in the loss of more and more lives every day. Parties had failed to unite even in the face of most urgent national problems. When the MGK took over, the TBMM (Grand National Assembly) had gone past one hundred rounds in trying to elect a new president for the country.
 After taking over, the military leadership observed that while many of the problems they had to deal with could be traced, if in part, to the totally selfish and less than responsible behavior of political parties, these very parties were now trying to resist the military policies to restore law and order. After some hesitation, they decided to terminate this struggle to dominate the political space. They judged that the existing parties and the party system was divided, divisive, corrupt and beyond reform. In 1981, they decided that the existing political parties should be abolished and a new party system that would, according to their judgment, serve the needs and interests of the country, should be constructed. The ideal system, the commanders felt, was a two party system with a moderate party on the left and the right each. In preparing for the transition to competitive politics, they tried to construct a system that would promote the evolution of the party system they envisioned. The particular instruments they devised included the introduction of a 10% national as well as district level thresholds Election of Deputies Law, no. 2839, dated 10 June 1983, art. 33) to keep small parties from achieving electoral success and a stringent geographical organizational requirements to insure that only parties organized throughout the country could take part in the elections. While the constitution said (art. 68, elaborated in art. 8 of the Political Parties Law no. 2820, dated 22 April 1983) that citizens were free to form political parties, the military leadership initiated a temporary clearing system for the national officers of political parties to take office. In this way, parties could be prevented from being operational before the 1983 elections, giving a head start to the two parties for which the military leadership had handpicked leaders.
 The military leadership appears to have entertained the hope that the two party system they were proposing would take hold and achieve success, thereby eliminating any popular inclination to go back to the old parties and the reconstitution of the pre-1980 system. Yet, the commanders displayed some flexibility for reasons that have never been fully explained, and allowed a third party to participate in the elections. By the time the elections came, public peace had been restored, the rate of inflation had begun to come down while the heavy handed manner with which the military was trying to shape the politics of the country had begun to generate frustrations and dissatisfaction. In the elections, ANAP (Motherland Party), the third party that had been reluctantly allowed to become operational, scored astounding success, gaining enough seats to be able to form a government without having the rely on the support of other parties.
 The results of the 1983 elections placed the military favored parties in an awkward position. The MDP (Nationalist Democratic Party), favored most by the military had come in a poor third, the HP (Populist Party) an unimpressive second. Their sense of weakness was further exacerbated by the fact that political parties that were established to continue the tradition of the pre-1980 parties scored major victories in the local elections held in the Spring of 1984. The outcome was resounding defeat for the two officially sanctioned parties. The local elections showed that these parties had no backing among the voters and would be eliminated from the parliament after the next national elections. Yet, some reciprocity of interests existed between the parties that appeared to enjoy popular support but had no representation in the parliament and those that had deputies in the parliament but had little popular support. Linking the deputies to parties that had broader support among the electorate took different forms on the left and the right.
 On the left, the HP had a leadership struggle between the founder of the party, Necdet Calp, who wanted to preserve the autonomy of his party and Ayd?n Güven Gürkan who advocated that parties of the social democratic tradition should unite. At the party congress on 27 June 1985, Professor Gürkan won the presidency of his party. He began shortly to bring about a unification of his party with the Social Democratic Party and succeeded in reaching an agreement. By early November, the SHP (Social Democratic Populist Party) had come into being with Ayd?n Güven Gürkan as its head. The deputies, by virtue of the fact that their party had decided to come together with another to form a new party, now became representatives of a party different than the one on whose ticket they had been elected.
 On the right, the MDP experienced greater difficulties. The government party, the ANAP, rather than the MDP was seen as the party of the moderate right. The ANAP already enjoyed a majority in the parliament and was not interested in negotiating unification with another party that did not have much voter support. The DYP (True Path Party), on the other hand, that had been established to represent the tradition of the pre-1980 AP (Justice Party) felt itself to be a victim of the military?s policies and did not want to work with the MDP leadership that identified closely with the military junta. Many deputies were in fact interested in going to either the ANAP or the DYP, but were deterred by a provision that the military leadership had interjected into the new constitution in order to prevent party changing in the parliament. The military leadership had identified party changing in the parliament not only as contributing to the political instability of the period that preceded their intervention, but also as a manifestation of political corruption. Article 84 of the constitution, therefore, called for the expulsion of the party changers from the parliament. Furthermore, such persons could not be designated as candidate by the national organs of another party during the elections immediately following the term during which they had changed their party. One formula that parties developed to circumvent the constitutional provision was for the party to dissolve itself. If the MDP deputies were left without a party, they were free to join another. On May 4, 1986, MDP held its national congress and voted to dissolve itself. Within days, the newly independent deputies began to join either the DYP or the ANAP. Another formula used later was called the hülle or ?pretend marriage? party. The pretend marriage party was a party that a deputy or a group of deputies who wanted to change their party would establish. It would then hold a convention and decide to merge with a political party to which the deputies wanted to move in the first instance6. The constitutional provision banning party changing (and forcing both parties and individual deputies to find ways to circumvent the laws) was finally changed in 1995 as part of a package of constitutional amendments on which broad consensus was obtained in the legislature (law no. 4121, dated 23 July 1995, art. 9). The privilege of deputies to change parties without suffering legal consequences was restored.
 As can be seen from the preceding discussion, inter-party mobility contributed to the normalization of parliamentary politics after the restoration of civilian rule in 1983 by strengthening the linkages between the voters and parliamentary parties. In doing this, such mobility also contributed to the reconstruction of some continuity in Turkish party life. In fact, this latter function was performed once again later in the case of the SHP, the successor the CHP (Republican People?s Part), the founding party of the Republic. In 1992, the decision to close the pre-1980 parties permanently was repealed. Many of the pre-1980 parties were reestablished. They enjoyed a name recognition and some voter attachment deriving from their history. Furthermore, they were entitled to receive back the money and buildings they possessed at the time of their closing. Almost all pre-1980 parties were reopened, then proceeding to unite with the new party that had assumed the legacy of the old. The SHP followed a different path and united with the CHP under the banner of the latter, consolidating the fragmented party system while restoring party continuity (Turan forthcoming).
 Both the Turkish constitution and the Political Parties Law contain provisions that ban parties, among others, from conducting activities against the fundamental values of the republic such as laicism and extending support to terrorism. In terms of its effect on the parliament, none of the party closures have been more significant than that of the RP (Welfare) and later FP (Virtue parties) for having acted against the laicism principle of the republic. The RP was established after 1983 to continue the tradition of religious conservatism which had been represented by the MSP (National Salvation Party) of the pre-1980 period. Its religious approach had always constituted a cause of concern among the staunchly laicist republican establishment. The constitution of 1982, based on the pre-1980 experience, contained strengthened provisions and institutional arrangements against those parties that appeared to undermine the fundamental values of the republic as a result of their activities. The RP was taken to court by the Chief Public Prosecutor of the Republic in May 1997, arguing that the activities of the party, taken together, violated the principle of laicist republic deeply embedded in the constitution. In January 1998, in a majority decision, the court agreed with the Chief Public Prosecutor and decided that the party should be closed permanently7. Part of the same decision was that those deputies whose actions have constituted the grounds on which the party was closed, would lose their seats in the parliament. When a party is closed, its deputies become independents. They can join another party or form a new party.
 The RP lost many of its leaders as a result of the court decision but those remaining quickly established the FP to continue the tradition of the MSP/RP. Their insistence on the continuation of tradition produced the expected outcome. The FP was taken to court by the same prosecutor and was closed once again in June 2001. This time, however, the decision produced a somewhat different outcome for the party. A group of deputies blamed the party leadership and its philosophy for the unfortunate fate that seemed to be constantly befalling them. Therefore, two parliamentary parties became established. Those who wanted to continue with the old tradition proceeded to establish the SP (Felicity Party) whereas those who wanted to part with tradition opted for the new AKP (Justice and Development Party). The new party insisted that it was a conservative party, respectful of republican principles and dedicated to continuing building a modern country.
 The court decisions have stimulated mobility among parliamentary parties. The founding of the AKP clearly became a possibility mainly by the occasion of party closure. If the FP had not been closed by the Constitutional Court, the founders of the AKP would have had greater difficulty in finding an appropriate opportunity to start a new party although their presence as a faction was already known. The establishment of a new party by some of the deputies of the defunct FP, provided an opportunity for the latter to have immediate parliamentary presence and all the benefits that it brings, without having had to organize outside the parliament first, and then, to achieve electoral success, trying to convince all that you were a credible organization worth supporting.
 Non-routine inter-party mobility is a phenomenon that is the product of an unusual context. It is not often that all political parties in a country are closed down and there emerges a process through which old politics may assert itself and eventually blend in with the new mainly through mobility in parliament. Again, it is not frequent nor is it desirable that a political party that is represented in the legislature is closed through a court decision. Yet, these have been important in the recent history of Turkey and as such, have deserved examination. But, as has been indicated already, inter-party mobility in the parliament in Turkey is also highly routinized activity. The persistence of such activity is unique and warrants close examination. It is to this routinized phenomenon that we now turn.
 Party changing occurs in an environment in which political parties, factions within political parties and individual deputies may have different priorities. Each may have its own goals, purposes, interests and agendas. There are trends in the political environment that affect the behavior of all. It seems reasonable to begin by examining the characteristics of the political environment that influences the behavior of all actors before turning to the examination of the actors.
 Referring to the characteristics of the party system in the 1970s, Özbudun identifies volatility, fragmentation and polarization as problematical tendencies (2000: 74). These have persisted into contemporary times despite the efforts of the 1980 military interveners to design a constitutional and legal system that was specifically intended to avert them.
 Many observers have traced the growing fragmentation of the party system to the introduction of a proportional representation as an electoral principle in the 1961 constitution. Sayar? (2002: 27) for example, argues that ?the plurality formula that was used between 1950 and 1960 discouraged the fissiparous tendencies within parties.? While the roots of fragmentation may lie in socio-economic change, it appears to be the case that a PR (proportional representation) system offered more of an incentive for ?dissenting parliamentarians within large parties? to leave their parties and form their own. Beginning with the late 1960s, the major parties continued to give birth to a number of small parties, some of which survived while others were buried in the Turkish political graveyard.
Table 2: Number of parties contesting elections and represented in the TBMM*
*The number of parties taking part in elections and in parliament have been compiled from information contained in Tuncer (2003).** Douglas Rae?s fractionalization of the vote index comes from Özbudun (2000: 77), except for 1999 and 2002 that have been calculated by Ilter Turan.
 Fragmentation of the party system has also been in part a product of military interventions that hoped to create a full or partial tabula rasa in the party system. The Democratic Party, the government party during 1950-1960 was closed down by the National Unity Committee in 1960. During the return to competitive politics, several parties were established to claim its legacy. But more comprehensive fragmentation came after the 1980 intervention which had proceeded, as has already been explained, to abolish all political parties. ?As civilian politics were restored, parties other than those the military allowed came into being, many of them representing the pre-1980 parties. On occasion, more than one party claimed the same pre-existing heritage? (Turan 2003: 156). Both some of the new parties and those claiming connections with the pre-military intervention times have managed to survive long after the termination of the military intervention. This has sometimes been true even in the case of the re-establishment of the pre-1980 parties, the presence of which would presumably remove the need for the continuation of the surrogates.
 A well intentioned measure (when) to extend funding to political parties from the public treasury in order to free them from dependence on well endowed actors that pursue their self interests, has unintentionally rendered the life of small parties easy and more attractive, enticing their formation in the parliament while reducing the costs of surviving as a small party. Although the criteria that are used in extending financial support to political parties have varied over time, parties with a minimum number of deputies in the parliament have always been assured of an allocation. In fact, in June 2005, when inter-party mobility in the parliament appeared to be picking up with some deputies being attracted away from the government party, the latter passed legislation raising the number of deputies that would be needed for a party to receive public funding from three to 20. The act was vetoed by the President who said that the measure was unfairly undermining political competition.
 There is general consensus among the observers of the Turkish political scene that patronage or clientelistic politics characterizes the party system. (Güne?-Ayata 1992, 1994). Patronage, in the form it has taken in Turkey, has simply come to mean the distribution of public resources to party activists and supporters. ?Patronage is better distributed in small networks. In fact, the larger parties have had more difficulty in meeting the demands of their constituents and have experienced centrifugal pressures, since catering to the demands of some creates jealousy in others? (Turan 2000: 156). In large organizations where getting sufficient and reliable information is often more difficult, it is always possible to think that your neighbors are getting more and become dissatisfied. It is never possible to deliver to all demands of a large body of clients satisfactorily. Under such circumstances, however, clients can increase their value by threatening to leave when their expectations are not met. The rewards of being in government so as to partake in the public resources often prove sufficient to lure individual deputies as well as factions from other parties to join government parties.
 The fragmentation of the party system has raised the need for coalition governments. In coalition governments, small parties get to enjoy more power than their share of the vote and seats in the parliament would suggest. ?The possibility of becoming a coalition partner serves as a disincentive for small parties to unite? (Turan 2000). Yet most coalition governments in Turkey, before and after 1983, have not functioned and performed well. There are built in problematical tendencies that are dysfunctional for the successful operation of coalition governments. To begin with, in a fragmented system in which many of the parties are breakaways from older parties and therefore ideological relatives to them, voters can easily shift from supporting one party to another. This potential for electoral volatility renders coalition partners insecure, creating pressures on them to maximize gains during what is likely to be a short tenure while also trying to maximize the distance between themselves and the ideologically relative parties, in order to establish a distinct identity in the minds of the voters so that their support may be retained. These pressures do not lend themselves to a long term ?cooperate and everybody wins? mind set and ends up producing unstable coalitions that fail to address the problems of the electorate at large. Similarly, the opposition is often unreasonably anxious to criticize governments and bring them down, and fails to display that appearance of responsible opposition that would become the government when the time comes. Çarko?lu et al. (2000: 32) have noted that the failure of parties in the Turkish parliament to deliver to their supporters leads the voters to engage in a never-ending search for alternatives. New parties aiming to attract these floating votes are established both within and outside the parliament. In this context, the 10% threshold initially intended to discourage fragmentation, has evolved into a barrier that stands in the way of new entries into the system and therefore its renovation.
 In conclusion, the fragmentation in the party system is both a cause and an outcome of electoral volatility that have come to pervade the Turkish political system in recent years or even decades. Fragmentation, in addition to rendering parties ineffective and discredited in the eyes of the electorate, has also led to the emergence of many parties, whose attributes the voters find difficult to discern. The attachment people had toward a specific political parties has been eroding, major shifts of votes between parties have become typical. Since the elections of 1991, for example, the parties that have prevailed in one parliamentary term have been the losers at the next. Parties that no one imagined would achieve electoral success have done so for one term only to go back into political oblivion. Such circumstances constitute parts of a political environment that does not discourage inter-party mobility in the parliament.
 Several characteristics of the Turkish political parties strike even the casual observer. Turkish political parties are organizationally weak, they have poor links with the broader electorate and they are leader dominated. Organizational weakness and poor links with society are, in fact, related phenomenon with historical origins. The political organization that established the Republic and ruled it as a single party between 1923 and 1946 was the CHP. Representing the leadership of the modernizing elite, the RPP had achieved nationwide control by making alliance with the local notables. It was not a party that was capable of mobilizing the masses for social and economic change. Rather it wanted the citizens to be passive recipients of modernization programs. It relied on the bureaucrats and local notables to elicit compliance and to mobilize votes during the elections. The party organization was not expected to produce new ideas or programs. Rather, such matters were the responsibility of the party leader who had a key role in giving direction to the party and its activities. The position of the leader was fortified by the powers he possessed in designating candidates for electoral positions, by a strict understanding of discipline in which criticizing the leader in itself was often deemed to be a violation of party discipline, the power to dismiss the heads and the executive committee members of local branches, to replace them if necessary and even to close the branch.
 The transition to political competition was initiated by some of the CHP leaders that had a different vision as to where the party should be heading. Their understanding of politics had been shaped under the single party system and they emulated the CHP as the model party after which they should build their own party (Turan 1969). The result has been narrowly based organizations in which the leader prevails. One evidence that none of the parties managed to evolve into a mass organization is that membership rosters of Turkish parties are poorly and irregularly kept8. Another is the fact that, although the option of holding primary elections is open, candidate designation for national elections has been taken over by the national organs of political parties without producing significant resistance on the part of local organizations. Similarly, one of the reasons why fragmentation of parties has proven easy is that it concerns not a large group of members but a much smaller number of activists and members of cliques who can move to a new party with relative ease.
 To restate, Turkish political parties are leader-dominated organizations. Leadership change is not usual or patterned. Attempts to change the leader even after major electoral defeats are exceptions and usually result in defeat of the challengers. In general, leaders rule with an authoritarian style within which challenges or modest criticisms are construed as acts of disloyalty9.
 The particular party structure described here leads to particular forms of behavior on the part of the deputies that are closely linked with inter-party mobility. Deputies try to get into the narrow circle around the leader in order to be noticed, to offer their services, to secure services for constituents if the party is in power, to acquire visibility by being given jobs such as talking on behalf of the party on the floor of the parliament or during a national party convention, to be considered for a ministerial position should the party involved in forming the government and above all to be designated as a winnable candidate10 in the next election.
 What is expected from the deputy is submissive behavior. Problems arise between the leadership and the deputies under a set of different circumstances. A deputy may feel that he is not accorded proper rewards for his loyalty to the leader. He may not be the beneficiary of sufficient patronage, for example. His demands for services to his constituency may be ignored if his party is in government. Or he may feel that he has been passed over for a position, be it in party or ministerial post, and decide that he has no future in the party. Finally, he may get the feeling that his chances of being nominated are not high. These contingencies are but examples of situations under which deputies will consider leaving their party for one where they think their interests would be better served.
 Other contingencies also exist. It is sometimes the case that serious differences may emerge on ideology or policy positions between the party and a deputy or a group of them. The departure of those disagreeing with the party is then a natural outcome. Equally likely, however, is a situation in which the deputy already has ideological or policy differences with the leadership that have become public and he already risks expulsion. To avoid expulsion (sometimes disciplinary proceedings against him may have already started), departure is a preferable option.
 Finally, there are non-political considerations that may defy classification. For example, a deputy of the DSP (Democratic Left Party) of Bülent Ecevit was asked to by the leader to leave the party when bribing allegations about him reached the public11. Sometimes, it is said that individual deputies have been offered material or other rewards for leaving their party. While such speculation was heard of frequently during the time of coalition governments that could come down with a few votes, they have dwindled under the one party government of AKP after 2002. But such speculation was recently revived when the government party received a deputy from the opposition CHP who was ostensibly under investigation for tax irregularities (Zaman, 4 May, 2004). One deputy in 1999 left his party two days after the election for it appeared that he had hoped but failed to get on the ticket of the party which he eventually joined. One deputy achieved fame for having changed his party affiliation six times during one term. He seemed to have no sense of what he was doing. His name had been picked as a candidate from among many who had sent in applications to Bülent Ecevit. The lack of local organization meant that the candidates had not gone through a local screening or an evaluation process.
 Political parties have organizational interests. They want to be represented in the parliament. They want to become the government party or at least partners in coalition governments. If they are in government they want to stay there. If they are not in government, they may desire to bring the government down. They want to expand their bases of support among the voters. They are open to expanding their numbers in the parliament. Such goals may sometimes be served through measures that involve inter-party mobility within the parliament. Here are some examples.
 As has been mentioned, the Turkish electoral system has a 10 % national threshold provision. The additional district level thresholds were ruled unconstitutional by the Turkish Constitutional Court in 1995. The electoral system favors the higher vote getter at the expense of the lower. The Political Parties Law, on the other hand, does not allow for political parties to form electoral alliances. These constraints have led parties to look for ways of getting around them to place their representatives in the parliament. One way that has been found is for a party that has some support but is not expected to achieve the threshold, is to negotiate with a more promising ideological relative (even if distant) for a number of eligible positions in return for instructing their voters to support the latter during the elections. If an agreement is reached, then the persons who the small party wants to offer as candidates resign from their party and become members of the party with whom the agreement has been made. It is expected that the votes such a formula brings will not only generate some seats for the small party which otherwise would not have had anyone elected under its own name, but it will also bring in additional seats for the bigger party. After the elections, the members of the small party will return to their home base, but having achieved parliamentary representation. This formula was first tried in the 1991 elections by the SHP that cooperated with HEP (People?s Labor Party), representing Kurdish ethnic nationalism. It is not clear that the pact enhanced the electoral position of the SHP. Some of the HEP deputies created incidents during the oath taking ceremony. Within a month, 16 of the 22 HEP deputies had resigned from the SHP. Eventually the party was closed by the Constitutional Court and some of its members were convicted, losing their parliamentary seat12.
 During the same election, the RP had made an arrangement with the ultra-nationalist MÇP (Nationalist Work Party), the successor to the pre-1980 MHP (Nationalist Action Party). Within a month of the opening of the legislature, the 18 deputies with nationalist credentials resigned from the RP and returned ?home?. During the elections of 1995, the same strategy was tried between ANAP and the BBP (Great Unity Party) that had broken off from the MÇP, giving the latter seven seats in the parliament. These electoral partnerships have not been tried after 1995, reducing interparty movement in the legislature. It may be that as the ties between parties and their voters loosened up, the ability of partners to command the votes of their ?supporters? came ever under question. Furthermore, the ANAP alliance with the BBP was met with such disapproval by the rank and file in the former that it became difficult to judge whether the partnership brought more benefits or losses to the major partner.
 Turkish political parties always keep an open door to deputies from other parties and welcome them among their ranks most of the time. Some judge that this is a reminder to those who may be entertaining leaving their party that they are not indispensable since replacements can always be found. Sometimes, parties may initiate an open or discreet search to get new members. Parties that do not have any deputies in the parliament always welcome having one or more of their own in the house. Parties that are represented by a few deputies are interested in reaching the magic number 20, which the Standing Orders define as the number needed for a parliamentary party to qualify as a ?party group?. This designation bestows immense benefits on a parliamentary party since party group is accepted as the basic parliamentary actor in most of the routine activities of the parliament. Shortly before the completion of this writing, for example, ANAP having failed to achieve the national threshold during the 2002 elections, had managed to get together more than 20 deputies by transfers from other parties so as to constitute a party group.
 The desire to have more deputies in a party has, on occasion, led to resorting to unethical practices. These may be anything from a promise for reelection, appointment to a ministerial post to substantial material rewards such as cheap credits from a public bank and even worse. Impressionistically, along with the general decline of inter-party mobility, it appears that the incidence of such movements have decreased not only because public scrutiny may have intensified but also because the current parliamentary arithmetic of only two major parties, with the government party possessing a substantial majority has rendered a party?s enticing others to join it an unnecessary exercise. It may be remembered that the so called deputy markets opened at the time of the formation of coalition governments.
 A faction refers to a group within a party that has some characteristic that differentiates it from the rest and has a certain ability to act together. Membership in a political faction is not clearly defined. Some are identified as leaders, others as belonging to it or even as fellow travelers. While factions are often just a group within a party, there may be instances where they may constitute subjects of inter-party mobility. First, a faction may decide to leave a party if it feels that the electoral changes of its members have been considerably reduced. Second, a faction may in fact develop into a proto-party and depart the party in which it is organized to found it own party. Third, a faction may leave a party to go to another, if it judges that it will be expelled from its home party if it does not move. These conditions may all exist in the same party at the same time. The transition to competitive politics in 1945-1946 was an early example of factions as a source of inter-party mobility. A more recent example may be found in the establishment of YDP (New Democracy Party) of ?smail Cem who served as Foreign Minister in Ecevit governments in 2002. Three high ranking officials of the DSP, all of them bearing ministerial rank, planned to take over the party from Bülent Ecevit whose health appeared to be failing. They were exposed before they implemented their plans. They and some of their supporters left the party. Had they not done so, they would have been expelled. Earlier, in 1996, a group of deputies had again rebelled against the decision of Bülent Ecevit not to support a motion of censure against Prime Minister Erbakan who had given a dinner at the prime minister?s residence to which leaders of religious orders wearing religious garbs had been invited. The same group of deputies had also asked Turkish Federation of Labor at an earlier time to persuade Bülent Ecevit to be more open to the unification of the center left parties. They left the DSP and joined the CHP.
 In an earlier study, an attempt was made to see if there were differences on the political characteristics of those who were involved in inter-party mobility and those who stayed with their parties. It was found at that time, that the level of localism, meaning whether the deputy was representing the province in which he was born or some other province, had little relationship with his party changing. It was also observed that there was a higher incidence of party changing among those that represented economically less developed regions. Finally it was observed that those who changed parties were more likely to have had longer parliamentary tenure (Turan 1985).
 For the 1983-2002 period, data was available on localism of deputies, their tenure (the number of times they have been elected to the parliament after 1983), their level of education, whether they have held ministerial positions and the socio-economic development level of the districts they represent. It is interesting that although the search for relations included only those deputies that changed their parties (i.e. excluding those who were left without a party because the court closed their party down), none of these variables exhibit any statistically meaningful relationships. Such lack of relationships may point to the homogenization of the deputies. Most are educated, most are local, there are virtually no differences between those who become ministers and those who do not, and all deputies have an equal proclivity to change their parties rather than only those from the less developed provinces.
 We have to conclude, with reluctance, that other information which does not appear in the rosters of the Turkish Grand National Assembly but which may be collected through interviews may be needed to discern the politically relevant attributes and to understand better the logic of those who move to other parties. Such information might include whether the deputy has an independent electoral base or whether he has been carried into office mainly by his party and the nature of relationship of the deputy with the local party organization (e.g. does he belong to a local faction, are there major rivals in his district etc.). Besides, what kinds of resources are available to him to get the attention of the party leadership? What options are open to him and what losses he would incur should he fail to get reelected? Has the deputy been identified with an existing faction within the party for a long time? What other networks within and outside the party is he connected with, including civil society organizations? Has he been a member of another party earlier? How was he recruited to politics and political office? And a host of other questions.
 The high level of inter-party mobility in the TBMM was a phenomenon that the founders of the 1983 constitution viewed as being highly problematical and tried to prevent. The particular measures they devised failed to stem the movement of deputies between parties in the parliament. Such mobility, far from producing undesirable outcomes such a governmental and broader instability, has contributed to a peaceful transition of Turkish politics from military to civilian rule. Further, mobility between parties is an instrument of making adjustments to changes in the political environment, a source of flexibility that may reduce tensions in the system. Is inter-party mobility then a blessing?
 Such positive contributions to political stability on a variety of occasions should not lead us to overlook the problematical aspects of inter-party mobility in the parliament. Certainly, there exist a set of multi-dimensional questions to be addressed. On the political front, it may be fair to ask if it is ethical to leave the party framework from which a deputy or a group of them have been elected? Expressed differently, is inter-party mobility more a curse than a blessing?
 This is a question of political philosophy that voters, party leaders, deputies should all address. There is no single simple or pre-ordained answer that can be offered easily. In the more appropriate domain of personal ethics, the question as to whether personal benefits are a legitimate basis for inter-party mobility, the answers is a qualified no. It is qualified only in the sense that it is not easy to make judgments without knowing the context of behavior, the nature of the benefit received, etc.
 Finally, should something be done about inter-party mobility? Apart from prosecution under the existing criminal code, whatever should be done, should be done in the domain of politics, not law. If citizens do not like what politicians that they have elected have done, it is within their power to change them at the next elections. Our analysis shows that inter-party mobility cannot be branded as good or bad without the consideration of the specific conditions under which it occurs and the outcomes it produces.
Bekta?, Arsev (1993) Demokratikle?me Sürecinde Liderler Oligar?isi, Istanbul, Ba?lam.
Çarko?lu, Ali; Erdem, Tarhan; Kabasakal, Mehmet (eds.) (2000) Siyasi Partilerde Reform, ?stanbul, TÜSES.
Erdem, Tarhan (1994) CHP üye Kay?tlar?n?n Merkezde Toplanma Projesi, 1978, Istanbul.
Güne?-Ayata, Ay?e (1992) CHP: Örgüt ve ?deoloji, Ankara, Gündo?an, 1992.
Güne?-Ayata, Ay?e (1994) ?Roots and Trends of Clientelism in Turkey?, in Güne?-Ayata, Ay?e; Roniger, Luis (eds.), Democracy, Clientelism and Civil Society, Boulder, Lynne Rienner.
Heper, Metin; Evin, Ahmet (eds.) (1988) State, Democracy and the Military: Turkey in the 1980?s, Berlin, De Gruyter.
Heper, Metin; Evin, Ahmet (eds.) (1994) Politics in the Third Turkish Republic, Boulder, Westview Press.
?nönü, Erdal (1999) An?lar ve Dü?ünceler I, ?stanbul, Yorum.
Özbudun, Ergun (2000) Contemporary Turkish Politics: Challenges to Democratic Consolidation, Boulder, Lynne Rienner.
Sayar?, Sabri (2002) ?The Changing Party System?, in Sayar?, Sabri; Y?lmaz, Esmer (eds.) Politics, Parties and Elections in Turkey, Boulder, Lynne Rienner, pp. 9-23.
Tunçay, Suavi (n.d.) Parti ?çi Demokrasi ve Türkiye, Ankara, Gündo?an.
Tuncer, Erol (2003) Osmanl??dan Günümüze Seçimler: 1877-2002, Ankara, TESAV.
Turan, ?lter (1969) Cumhuriyet Tarihimiz, Istanbul, Ça?layan.
Turan, ?lter (1985) ??Changing horses in midstream?, party changers in the Turkish Grand National Assembly?, Legislative Studies Quarterly 10, pp. 21-34.
Turan, ?lter (2000) ?Volatility in politics, stability in parliament: an impossible dream? The Turkish Grand National Assembly during the last two decades?, Journal of Legislative Studies 9(2), pp. 151-176
Turan, ?lter (forthcoming) ?Old soldiers never die: the Republican People?s Party of Turkey during the last decades?, Southern European Politics.
1 This statistic has been compiled by Associate Professor Dr. ?eref ?ba, who is the Assistant Director of the Division of Laws and Resolutions of the Turkish Grand National Assembly. He compiled the list of the members of parliament who changed their parties and provided a note on the constitutional provision restricting party changing and its being amended in 1995. Ay?e Zarakol, an ABD student in the Department of Political Science of the University of Wisconsin, coded the social background characteristics of deputies, their changing of parties and executed the data processing. Finally, I would like to thank my former graduate assistant Oya Memi?o?lu for newspaper search on party changers.
2 In case a party is closed by the Constitutional Court, the national officers and the parliamentary officers of the party as well as members whose actions have led to the closing of the party lose their seats while the rank and file become independents, making it possible for them to join other parties or to form new parties. The law bans, however, new parties that can be depicted as the continuation of parties that have been closed down.
3 The record breaker was a deputy elected from the DSP during the 20th legislative term (1995-1999). After staying with his original party for a year or so, Kubilay Uygun moved to DYP, but four days later returned to his original party. He resigned three weeks later to go once again to DYP. After about a year there he moved to the MHP, only to leave it less than three weeks later to join the recently established Democratic Turkey Party (DTP). Five months later, shortly before the elections, he became an independent. Kubilay Uygun?s parliamentary journey has been compiled from the records of the Divison of Laws and Resolutions of the Turkish Grand National Assembly.
5 For a comprehensive examination of the period concerning party realignments and the consolidation of competitive politics, see Heper et al. (1988, 1994).
6 The expression hülle or ?pretend marriage? was named after the ancient Islamic practice devised to discourage people from arbitrarily divorcing their wives. If a man had divorced his wife three times, he could not remarry her until she had been married to someone else. To insure that a marriage to someone else would not be consummated, an elderly, impotent man would be found and paid a modest sum to marry temporarily the divorcee so that she could remarry the former husband again.
8 Tarhan Erdem (1994) has an excellent book describing the difficulties the project of renewing membership rosters, an effort which was strongly resisted by the provincial organizations.
9 There is considerable writing on oligarchic leadership and intra-party democracy. As examples, see Bekta? (1993) and Tunçay (n.d.).
10 Winnable candidate refers to the rank of the candidate on the party ticket. Turkey has multi-member electoral districts and a proportional representation system. This means that only a certain part of the ticket will win, depending on the percentage of the vote a party gets. Naturally, the higher the rank of the candidate on the ticket, the more likely that he/she will win the election.
11 Birlik Ajans News Agency, March 4, 2005.
12 Erdal Inönü, the president of the party at the time, in his memoirs (1999: 256) offers a different explanation. He says that this alliance was not intended to increase their votes. Rather, it was the extension of a helping hand to HEP that faced an unreasonable threshold.