Ten Theses on Constitutional Change in Turkey
This, the current Constitution of the Republic of Turkey is no longer the same constitution, the Constitution of 1982.  Yes, even partial, but sequential or re-iterated rounds of amendment can produce a new constitutional regime or material constitution. This is what happened in Turkey through amendment rounds in 1987, 1995, 2001, and 2004.  It was under European pressure in the first decade of the 21st century that the demand for formal constitutional replacement was adopted by Turkish political actors, supposedly in the place of the method of sequential amendments. Was the idea of an entirely new, civilian constitution wrong? No, for two reasons. The first is the problem of legitimacy, caused by tainted origins. This problem undermines the necessarily preservationist review function of the Constitutional Court. The second is the problem of freezing. Some features of constitutions, though formally changeable, are never sufficiently altered in reform, because incumbents significantly benefit from them. Such a feature e.g. is obviously the materially (if not formally) constitutional 10% electoral threshold in Turkey, but constitutional provisions for party closings belong here too, despite the fact that they have been partially reformed.  These reforms did not protect the Kurdish DTP (Democratic Society Party), www dissolved in late 2009. Yet with this said: an entirely new constitution could be worse than a thoroughly amended “old” one. So warned Chief Justice Solyom in Hungary in the 90s, and his fears were unfortunately realized by the FIDESZ Basic Law of 2011.  In Turkey a hyper-presidentialist constitution would be certainly worse than the current one, that has traveled a long way from its original presidentialist features.
Contrary to some Europeans beliefs, The AKP (the now ruling Justice and Development Party) has not been the only or main agent in Turkey’s process of constitutional reform, but rather only a late but important participant. Unlike its predecessor, the Welfare Party, it admittedly no longer disrupted the consensual amendment process in 2001 for special ideological reasons. In 2004, it even led one important consensual round with the full participation of the CHP (Republican People’s Party), the only opposition at the time. But it is equally important that in 2007 this governing party broke with the consensual method dominant till 2004. The long constitutional crisis was initiated by this break in consensual process, not merely by the insistence of the AKP on its choice for president, that had a positive side, but by its pushing through constitutional reforms on the institution of the presidency through the majoritarian-plebescitary path that until then was eschewed by the parties quite consciously. This was the source of a long term constitutional crisis punctuated by majoritarian amendments, that failed in 2008 and succeeded in 2009-2010. In both consensus was replaced by plebiscitary legitimacy.
The Constitutional Court has been more than merely an instrument of the deep state as many on the Left (generally opposed to constitutional review) tend to claim, re-enforcing charges by the milieu of the AKP. Indeed the Turkish Constitutional Court has not even consistently ruled against the AKP government. Yes on the issue of the parliamentary quorum in 2007 it sided with the CHP’s position, but in the immediately following decision on the referendum on the presidency it upheld the government’s right to go ahead, in the face of Pres. Sezer’s veto and the CHP’s challenge. Yes, in 2008, on the headscarf issue it invalidated the constitutional amendments of the AKP and MHP, but when it came to the AKP party closing project, it saved the governing party from dissolution. And finally, against the CHP project of blocking a referendum where the high judicial body’s own integrity and power were challenged through a packing scheme, the Court in 2010 once again upheld the right of the then 3/5 majority (and the president) to go ahead with its referendum, and was, as a result, indeed packed.  Unlike in Hungary, where there was also significant packing, there is no sign yet of any independence on the part of this new Court. 
The Court’s amendment review has been widely regarded as legally indefensible, and politically the equivalent of juristocracy. Those who see things this way seem to have no idea of the possibility of amendment review in terms of the either the basic structure doctrine of the Indian Court that does not have eternity clauses to rely on, and the doctrine of the German Court that does.  Amendment review in fact is even more clearly justified in Turkey than in India and Germany, given the much less representative nature of the TGNA (Turkish Grand National Assembly – the parliament) linked to an exclusionary electoral rule, one that is a key heritage of Evren’s junta. Those who oppose amendment review formalistically fasten on an untenably sharp distinction between procedure and substance according to the language of the Constitution. They even forget, that this very distinction was twice imposed by military controlled governments of Turkey, in the 1971 and in the Constitution of 1982. It was at those times defended by the most conservative jurists, but the Court never did or should have accepted the distinction (e.g. in decisions of 1976 and 1977).  Substance can and must be examined to know if the proper procedure was followed, for epistemological reasons rather than as the bases of the decision itself. From a political point of view finally neither the charge of juristocracy, or even that of the freezing the constitution could be upheld. The Court hardly rules Turkey, and has permitted important constitutional amendments not in its institutional interests. Since it takes either the president of the republic, or at least 110 deputies to bring an amendment to the court for possible invalidation, instead of juristocracy or blind conservatism, it is better to regard the Court before its packing as a consensus enforcing institution whatever the private intentions of judges. This raised the question after Court packing: is there any means of enforcing consensual process in constitutional change?
The packing of the Court, disguised by the large 26 item amendment package of 2010, was very likely connected to the problem of constitutional change, even if seen only in terms of the desire to counter-act a frozen process. That process was frozen only because the old Court was a possible roadblock in the face of majoritarian attempts to reform the Constitution. But why was the AKP government so intent at forcing through majoritarian change? Was it the impossibility of satisfying the European’s demands for a new constitution since, supposedly, no consensus whatsoever with the CHP was possible? First of all, the CHP has in the past showed very high readiness for consensus, helping to override Pres. Sezer’s vetoes in 2002 and 2004, one of these involved the all important issue of a constitutional amendment needed to allow Mr. Erdogan to return to parliament and assume the Prime Ministership. That was a very big issue and indicated the opposite than unwillingness to negotiate and compromise. So I don’t readily buy “the no consensus was possible” argument. If the CHP was no longer capable of consensus, perhaps the reason should have been sought and removed. Moreover, I find it hard to believe that in an epoch in which Chancellor Merkel and Pres. Sarkozy both moved to slow down Turkish accession, European desires could have been the main factor. The religious freedom issue (itself ambivalent: see the U.S.) on which certainly no consensus with the CHP was possible is a more plausible candidate for an explanation. Yet now that we see that even with a packed Court the AKP has been in no hurry to produce the required legislation permitting complete freedom of dress in all public institutions including universities , it is tempting to see the religious issue as a pretext. It is thus more instructive to ask what entirely new constitutional project was launched by the AKP government, once the Court was packed.
The candidate for the best explanation is the desire to replace parliamentary government, working quite effectively in Turkey, by a presidentialist form. I need not go into the well known motives of the PM in wishing to establish such a form, or decide when the project, not without important precedents in Turkey, was first conceived. I want to note only that at issue is at the very least strong presidentialism or hyper-presidentialism as S. Nino once called. Observing the struggles of American presidents in internal politics, it is very much doubtful that to a politician used to the strong power position of the Turkish PM that kind of presidency based on the strong separation of powers and the likelihood of divided government would be attractive. Others before Mr. Erdogan, like Nyerere and Nkrumah have openly rejected the U.S. model for similar reasons, developing the African concept of the “executive presidency.” So here too it is said that it will be presidentialism in a Turkish style, “a la Turque.” Actually, the parts of the scheme are much more important than the label. The change in question will be valuable to a future incumbent only if some combination of foreign policy leadership, significant decree and veto powers, strong appointment powers and the power of parliamentary dissolution could be achieved. In such a scheme, a prime minister would depend on the confidence of the president first and foremost not that of parliament. Indeed, it is important not to regard the provision for a prime ministership automatically as the institutionalization of a semi presidential system, especially in the form that was inaugurated with Mitterand’s first co-habitation. In most of the world, prime ministers are dependent on the president as are all other ministers. Even in France, the system remains highly presidential when the party of the president controls the national assembly.
I return to the problem of consensus, as Turkey too returned to consensual process. This may be surprising given two of my theses: the packing of the Court, that removes some of the need for consensus, and the project of presidentialism. On the face of it a consensual process involving the agreement of four parties is not likely to produce presidentialism, and thus the mere creation of parliamentary Constitutional Conciliation Commission of four parties with complete parity and a decision rule of unanimity seems to speak against the thesis of presidentialism as the motor behind the scenes. I do not concede this point. The reasons for the creation of the Conciliation Commission were normative as well as electoral. Given the recent international influence of negotiated forms of constitution making, even supporters of the AKP’s constitutional projects often concede that legitimate and successful constitution making should involve consensus. This matters, but ideology is rarely enough to cause things all alone. In Turkey the demand of consensus has been supported by obvious electoral results. Though as often emphasized by AKP supporters, the party has increased its share of its total vote twice since 2002 (the numbers: 34%, 47% and 49.8%), each time, presumably because voter learning, it has lost in the number of allotted seats (the seats: 363, 341 and 327). In 2011, the latter figure has fallen below the constitution amending 3/5 threshold. Even the imprisonment of deputies does not help, because the required proportion is of the absolute number. Thus the support of deputies from other parties for any constitutional project has become mandatory. 
From the outset two tracks were chosen for constitutional change, each involving more than one party. The first of these was the very important return to consensual methods, namely the creation of the Constitutional Conciliation Commission. But in contrast, e.g to Hungary in the 1990s, even as a consensual committee was constructed, there was no moratorium on ordinary amendments. This left open a second track, that has been used before in 2008 and 2010. Undoubtedly it was thought, that if it was possible in 2008 to get one party to support a specific amendment, it may be possible again. Moreover (this was repeatedly expressed in public statements) the very possibility of following the second more majoritarian route was a club over the Conciliation Commission, forcing agreement to avoid a worse constitutional outcome achieved through the route of a more partial amendment. Unfortunately this particular technique has another even more negative meaning. Indeed, one can speak of a subversion of possible consensus based on compromise, because a party that can force its will through without compromise is likely to wish to do so rather than give up anything it deems important. Thus, it is also possible, that the Conciliation Commission was meant by some AKP politicians to be mere window dressing, whose failure would legitimate the use of the other route. Certainly the overly high consensus requirement, again in contrast with Hungary,  speaks for the expectation that the Commission would be deadlocked. 
But which of the three parties could and would join the AKP in a presidentialist project? The MHP (Nationalist Movement Party) and the CHP have repeatedly announced their strong opposition, that would be difficult to retreat from. The Kurdish BDP (Peace and Democracy Party) however, though opposed in principle to the abandonment of parliamentary government, does not consider this the main issue, and thus for a price is capable of supporting the AKP on this. I am certainly not the first to argue that most likely the settlement process with the Kurds has a linkage, on the level of motivation or only function, who knows, with the new constitutional ambition of the AKP and its leader.  It is certainly not the only motivation. Islamic ideology that allows no distinction between Kurd and Turk is supposed to be a factor, real or not I cannot tell. In any case, ideology is never the only or the main motive force. There is also the electoral fact that the AKP gets a lot of Kurdish votes, and only by doing something for the Kurds can these votes be kept and even increased. This factor however is balanced in part by the fear of losing some Turkish ethnic votes in turn. Thus, the constitutional motivation cannot be discounted. Even if there is an electoral cost, the much strengthened presidency could govern with a much smaller majority in parliament.
Could such a Kurdish–AKP deal be sold to the electorate at all? The short answer is yes. I have had Turkish liberal interlocutors tell me that if an equitable deal would be given to the Kurds, “presidentialism”, not further defined, would be a small price to pay. I disagreed, but it is not my opinion that counts. In 1958, when De Gaulle offered peace with Algeria, I probably would have voted for Mitterand who spoke of the Vth Republic as a permanent coup d’état. Yet, today the long answer to the same question is probably no, presidentialism would not be supported. Not after Taksim square and the development of the Syria crisis. Not only the movement’s powerful discourse, but Erdogan’s reaction to it, did serious damage to his reputation, probably even within AKP elites. They after all do have other choices than creating a super –presidency for a leader who appears to be a loose cannon. Coupled with Erdogan’s analysis of the Egypt military take-over as an Israeli plot, his incredibly misinformed Syria posture, that counted on the Assad domino falling like others elsewhere, has completely undermined a previously agile and successful foreign policy. Unlike Pres. Obama, Erdogan was not saved by Lavrov. His errors on Syria, which were at first shared by the Americans and others, were entirely inexcusable for the leadership of a major country next door with a supposed understanding of the ethnic and religious situation of the neighbor. Interestingly, the PM responded to these setbacks not by accelerating, but slowing down the peace process with the Kurds. Does he realize something that commentators have not yet noted, concerning the end of his dreams for a presidential regime, led by himself? While the AKP can easily recover for electoral purposes given the quality of its opposition, this should not mean, this time around, support for Erdogan’s constitutional plans. The Hizmet movement led by F. Gülen and the newspaper Zaman have already turned against him. At Taksim Square the Turkish de Gaulle may have met his 1968 before he had his 1958.
 Only blind formalism or ideological thinking could claim that the 1989-1990 Hungarian Constitution, comprehensively reformed through amendment rounds, was still the Stalinist Constitution of 1949. After several rounds of important amendments the situation is similar in Turkey even if the changes were never completed.
 See the excellent summary by Özbudun and Genckaya Democratization and the Politics of Constitution making in Turkey (Budapest, 2009); See also my “From Democratization to Crisis: Three Phases of Constitution making in Turkey 1982-2008 in Quarterly Bulletin of third World Studies (2009) 49 (2)
 Of course both of these systems can be reformed if the party that has the relevant parliamentary majority can reform them in such a way as to draw the greatest advantage. This is what happened with party closings in the reform of 2011, with a result that it has become impossible to ban the governing party (not a bad thing of course) while it remains possible to ban a Kurdish party once again. As to the electoral threshold, the current announcement of a possible reform, either a new 5% threshold with small districts or a FPTP system would be the greatest benefit of the AKP itself, and surprisingly all other parties and even the BDP would lose a few seats. See Seyfettin Gursel “Electoral System Debate is opened” in Today’s Zaman October 1. And the following simulation table provided by the author:
|Votes shares (%)||45||6||26||13|
|Actual system (10% threshold)||308||35||147||60|
|Reformed system (5%)||337||32||141||40|
 See G. A. Toth ed. Constitution for an (un)Divided nation (Budapest, 2012)
 Packing should be defines as significant increase of size, where the responsible agency for this will largely control the appointment of new members. No only was TCC’s membership increased from 11 to 17, but 4 obviously known substitute members ere immediately added to the Court by temporary article 19 also approved in the same referendum. This point is systematically neglected by those who claim that there was no packing, and the two (eventually 3) new additions were named by parliament. And who controls that parliament? See A. Bali’s partially misleading “Unpacking Turkey’s ‘Court-Packing’ Referendum” Nov. , 2010. Can Yeginsu in “Turkey Packs the Court” in NYRBlog does get the real point, but also does not mention the relevant temporary article 19, that would have made his case even much stronger. Note that number of new judges that were to be added in Pres. Roosevelt 1937 scheme was a maximum of five. In Hungary in 2011 the number was increased from 11 to 15, though because of a retirement PM Orban was able to name 5 the, with two more to come. In Turkey 6 were added right away. The mechanism in Hungary is purely parliamentarian, but requires 2/3 each time, unlike the Turkish scheme that is 2/3 first, then absolute majority and finally a run-off of two.
 The relevant HCC decision was dramatic, because it invalidated part of a constitutional amendment, if only its temporary articles. Constitutional Court of Hungary: Decision 45/2012. (XII. 29.) “on the unconstitutionality and annulment of certain provisions of the Transitional Provisions of the Fundamental Law of Hungary.” See Renata Uitz The Return of the Hungarian Constitutional Court , Verfassungsblog Januarry 15, 2013. The FIDESZ government overcame this decision by lifting the articles directly into the constitution through the Fourth Amendment, an omnibus act. With respect to amendments the Hungarian Court recognizes only purely procedural review, that cannot rely on eternity clauses with respect to substance. Venice Commission (European Commission for Democracy Through Law) “Opinion on the Fourth Amendment to the Fundamental Law of Hungary” Venice, 14-15 June 2013.
 After reviewing some of the relevant country provisions, and the importance of eternity clauses, the Venice Commission goes on to say: “In Turkey too, the Constitution contains unamendable provisions. Article 148 of the Turkish Constitution provides that the Constitutional Court is limited to control the procedure of adoption of constitutional amendments, but it seems that the Court has a wider interpretation of its power to review constitutional amendment. In all these cases, the constitution has an inner hierarchy (unamendable provisions or basic principles) and ‘ordinary constitutional law’ is reviewed against these higher provisions or principles.” Venice Commission “Opinion on the Fourth Amendment to the Fundamental Law of Hungary.”
 C. Belge “Friends of the Court: The Republican Alliance and the Selective Activism of the Constitutional Court of Turkey” in Law and Society review (2006) 40 (3) who proves something quite other than she thinks. See my ”Democratic Constitution Making and Unfreezing the Turkish Process” in Philosophy and Social Criticism” (2010) 36 (3-4).
 Even in the current reform package, 3 years later, the removal of the ban is not complete.
 Unless of course one of the three new electoral proposals is adopted, each with a greater D (disproportionality index) than the current one to the benefit of the largest party. Any of them would put the AKP over 3/5, and two might over 2/3 with even 45% of the vote. See f.n. above.
 The numbers in Hungary were 5 out of 6 parties, and 2/3 of individual members. In Turkey all four parties must agree.
 In Hungary a positive result was achieved in committee, but was brought down I parliament where the old 2/3 amendment rule continued to apply. That too would be a further hurdle in Turkey.
 See Arato and Tombus “Learning from Success, Learning from Failure: South Africa, Hungary, turkey and Egypt” Philosophy and social Criticism (2013).