The news reports and commentary on Turkey in the middle months of 2010 have sounded alarmist themes. Analysts have warned that Turkish foreign policy is undergoing a reorientation away from the West, ominously foreshadowed by deteriorating relations with Israel. Commentators worry about creeping Islamization in domestic and foreign policy, a concern captured by pictures of headscarved women accompanying articles about Turkey’s eastward turn. Elsewhere, descriptions of Prime Minister Recep Tayyip Erdoğan’s increasingly assertive policies toward the Middle East are paired with allegations of a more authoritarian style of government by the ruling Justice and Development Party (AKP). Turkey’s September 12 referendum — resulting in the passage of a package of constitutional amendments with support from 58 percent of voters — offers the most recent occasion to revisit this increasingly critical portrait of Turkey in Washington and beyond.
Many analysts interpreted the referendum as the final showdown between the country’s secular establishment and Islamist forces, with the result definitively displacing the former and accelerating Turkey’s alleged turn from the West. Such accounts treat the referendum as a vote of confidence in the government with little attention to the substantive impact of the constitutional amendments that were the subject of the vote. In fact, that package of constitutional amendments chips away at the remaining illiberal provisions of a constitution originally drafted by the military junta that came to power by coup in 1980. Indeed, the referendum coincided with the thirtieth anniversary of that coup and the political polarization that preceded the vote reflected in some measure its still contested legacy. Accounting for the intense polarization around the referendum by recourse to the well-worn trope of Islamists arrayed against embattled secularists is both simplistic and misguided. A balanced analysis of the referendum results should reverse the emerging narrative on Turkey and restore a more realistic (and even optimistic) assessment of the country’s trajectory.
The 26 constitutional amendments at issue in the referendum would, in most other places, have been warmly welcomed as liberalizing improvements to a flawed constitution. They include provisions that: empower civilian courts while reducing the jurisdiction of military courts; strengthen gender equality and protections for children, the elderly, veterans and the disabled; improve privacy rights and access to government records; expand collective bargaining rights; afford individuals standing to bring constitutional challenges; and remove immunities long afforded to those responsible for the 1980 military coup.  The overwhelming effect of these provisions amounts to civilianizing the military coup-era constitution, strengthening individual freedoms and political rights, and undertaking much-needed judicial reform. The European Union, which Turkey seeks to join, consistently expressed support for the amendment package and welcomed its passage as another step toward convergence with European standards.  Once the results were released, President Barack Obama called to congratulate Prime Minister Erdoğan on the outcome. 
This portrait of international reception of the constitutional amendments hardly comports with claims that the referendum will enable consolidation of Islamist power in Turkey. Indeed, many commentators suspicious of the AKP’s motives have expressed frustration, exasperation and, at times, contempt for the support garnered by the constitutional amendments in the West. Why, they wonder, are Western “liberals” so easily enlisted in support of what their Turkish counterparts deem to be an Islamist threat? The Westernized elites of Turkey’s coastal cities — once the strongest bastion of support for Turkey’s EU bid — have lost their enthusiasm for Europe precisely as a result of their pessimistic assessment that European civil rights and political liberties inure to the benefit of the political forces in Turkey that they most despise.  That this referendum enjoyed European support only confirms, in their view, that the liberalization of the Turkish political order required to meet EU accession standards has become inimical to their own political interests and preferences.
Why was the amendment package considered so threatening by Turkey’s traditional urban elites (as well as some of their supporters in Washington and beyond)? The main substantive objections centered on two elements: Procedurally, the amendments were offered as a single package rather than allowing the electorate to vote on each provision individually. More importantly, opponents saw court packing in changes to the composition and selection process of the constitutional court and a board overseeing judicial appointments. Beyond these substantive objections lies the anxiety experienced by traditional urban elites in Turkey as they watch the demographic changes in Turkish society slowly reflected — through this constitutional referendum as well as prior reforms to the post-1980 political order — in new distributions of political power through state institutions. But before turning to these underlying anxieties, the specific objections to the amendments themselves ought to be considered.
The procedural objection has some merit. The AKP cobbled together a package with elements to attract a wide spectrum of the electorate, but had each amendment been put up for an individual vote, some might have been more vulnerable to rejection. That being said, the amendments were voted on singly by Parliament before being put to referendum. Indeed, a key liberalizing amendment that would have made it more difficult to ban political parties was dropped from the final package because it did not garner the requisite votes in Parliament.
The fate of the provision that fell out of the package during parliamentary voting is quite telling. In fact, the provisions of greatest concern to the opposition — those impacting the composition of the judiciary — may well have met the electoral test if voted on singly. Contrary to critics’ claims that much of the Turkish electorate did not understand the substance of the amendments, the one issue that was most hotly debated in the Turkish press in the leadup to the referendum was the allegation of court packing. But given the judiciary’s history of supporting the military and activism in closing or nearly closing political parties that garner significant electoral support from religiously conservative and Kurdish quarters, it is reasonable to imagine that a significant proportion of the voters may view the judiciary as anti-democratic. Accordingly, there is reason to think that those who voted in favor of the amendments package as a whole might also have supported judicial reform had it been separated from the package. By contrast, the amendments that were least controversial at the level of elite contestation — those improving individual freedoms and political rights — might have been compromised if voted on individually. For instance, provisions limiting the jurisdiction of military courts over civilians might well have foundered because of opposition among ultra-nationalists keen to maintain broad military authority in the Kurdish provinces. Ironically, given the liberal provenance of the procedural critique, disaggregating the vote on the amendments might have had the counterintuitive effect of retaining controversial provisions on judicial reform while undermining some of the improvements to individual liberties.
A second line of argumentation held that the single package of amendments was merely “democratic window dressing” designed to disguise the AKP’s underlying agenda to pack the court in its favor.  Analysts pursuing this line of critique also argued that piecemeal amendments were, in any event, inadequate, given the need to introduce an entirely new, civilian constitution to replace the coup-era document. Liberals who supported the constitutional amendment package in September acknowledged the critique that the reforms did not go far enough, while rejecting an all-or-nothing approach that would have postponed, once again, reforms in the areas the package did cover.
A comparison of the package put to referendum on September 12 with an earlier draft civilian constitution circulated in 2007 — and widely lauded as a liberal draft in line with European standards — belies the description of the package as an opportunistic hodgepodge. To the contrary, the four key areas of reform advanced by the earlier effort to replace the constitution — individual freedoms, political rights, judicial reform and civilianization — were all reflected in the 2010 amendment package. Indeed, the most contested provisions — related to the judiciary — were both part of a broader judicial reform strategy that was repeatedly applauded by the EU. Thus, the amendment package represented an array of reforms in four domains that remained aligned with previous reform initiatives and with EU accession requirements. On the other hand, while the provisions included in the package were not arbitrarily assembled, the limitations of what was proposed doubtless represented a degree of opportunism on the part of the government. In particular, the failure to address the 10 percent electoral threshold, by which parties must win 10 percent of the vote nationwide to win parliamentary seats, demonstrated that even as the AKP advocated these reforms, it also maintained its focus on the 2011 legislative elections.
The exceptionally high electoral threshold put in place following the 1980 coup benefits large national parties, like the AKP, while adversely affecting smaller and regional parties, such as the pro-Kurdish Peace and Democracy Party (BDP). Lifting the requirement or lowering it to be more in line with liberal standards (for instance, a 3-5 percent level), would certainly have been desirable. Such a change in the threshold would almost certainly have enabled more than a handful of parties to gain representation in Parliament in the 2011 elections, diluting the AKP’s majority.
The unwillingness of the AKP to address the electoral threshold, coupled with its failure to consult with the BDP during the parliamentary debate about the proposals in the spring, led the Kurdish party to call for a boycott of the referendum. The boycott was widely observed in several key Kurdish provinces, where higher turnout would likely have substantially increased the proportion of “yes” votes in the referendum.
In all but one province, those who did vote in the southeast overwhelmingly supported the referendum. This outcome is largely attributable to the “civilianizing” impact of the amendments on the military’s role in government. In particular, amendments that limited the jurisdiction of military courts and barred such courts from trying civilians were both important gains for Kurdish civil society. Thus, the boycott not only reduced nationwide turnout but did so from a sector of the electorate that would otherwise likely have supported the referendum at a rate far above the national average.
Yet this failing of the constitutional amendment package had a silver lining. First, the success of the boycott in much of the southeast demonstrated the extent of the BDP’s influence in those provinces. Following a period in which the AKP excluded the BDP from discussions of its Kurdish initiatives in the spring and summer, the referendum demonstrated that such a strategy is not viable. Further emboldened by its success in the referendum, the BDP has organized several civil disobedience campaigns, reflecting its status as the premier non-violent political representative of Kurdish interests. The BDP’s successful show of electoral force has yielded new instances of outreach by both the AKP and the Republican People’s Party (CHP) to the Kurdish party, themselves a promising sign of renewed engagement.  Second, the extensive debate about the electoral threshold in the runup to the referendum has generated momentum for reform on this key point, increasing the likelihood that the threshold will be lowered in the next round of reforms.
There is no question that the constitutional amendment package would have been improved had it been accompanied by measures lowering the electoral threshold. (Lowering the electoral threshold would require amendment of the Political Parties Law, rather than the constitution, as the current threshold is statutory, not constitutional.) Legitimate criticism of the shortcomings of the package, however, should be distinguished from substantive attacks on those provisions that were included. By reducing the role of the military in ordinary governance, undertaking judicial reform in the area of both military and civilian courts, and enhancing individual and political associational rights, the amendments that received substantial support in the September 12 referendum represent an important step toward liberalizing the Turkish political order. In the end, it might have been preferable to offer the amendments for referendum individually, but the procedural objection was not the real issue. The substantive concerns about the contents of the package — and particularly about its impact on the judiciary — are the core of the controversy. And they are largely baseless.
The Court Packing Charge
The allegations of threats to the independence of the judiciary and control over the appointments procedure would be incoherent but for the deep-seated skepticism about a democratic and participatory political order among the opponents of the reforms. Much of the concern stems from fear of what might happen if the judiciary and the state prosecutors became more representative of Turkish society as a whole, and not of the particular elites from which these groups have heretofore been selected. There are two sets of amendments at the center of the controversy — one concerns the composition of the Turkish Constitutional Court (TCC), the other the composition of the Supreme Board of Judges and Prosecutors (in Turkish, Hakimler ve Savcılar Yüksek Kurulu, or HSYK).
With respect to the TCC, the amendments in question increase the size of the Court from 11 permanent and four alternate justices to 17 permanent justices. Two related objections have been raised to this expansion. The first regards the institutions from which the pool of candidates may be drawn. The second regards the role accorded to the political branches in appointing the candidates. Both objections are driven by an additional layer of concerns. In Turkey, the elected branches of government (or hükümet) are understood to be separate from the unelected branches, which are more commonly referred to as the “state” (or devlet). The unelected branches are comprised of the senior judiciary and prosecutors, the military, and certain parts of the intelligence, national security and policing apparatus. The idea of separation of powers is understood not as the separation of the three branches of government, as would be the common American understanding, but as the separation of the elected and unelected branches of government. In particular, the subjection of the unelected branches of government to forms of democratic accountability is deemed to represent a direct threat to the autonomy of the state. Thus, the civilianization of the constitution — the subjection of the military to civilian authority — and the reform of the judiciary — greater democratic accountability in the appointments process — are both understood to undermine separation of powers by subjecting the state to majoritarian pressures. One can only assess concerns about the expansion of the TCC in light of the norm that the unelected branches of government are expected to serve as the guardians of official state ideology and elite interests and remain insulated from democratic pressures.
Concerns over the pool of candidates from which TCC members may now be drawn reflect the view that this highest court should be composed exclusively of candidates drawn from the self-appointing ranks of the elite judiciary. Prior to the amendment, the 11-member TCC was drawn primarily from among the five next highest courts in the country: the Court of Cassation, the Council of State, the Military Court of Cassation, the Military High Court of Administration and the Court of Accounts. Appointees from these courts accounted for seven of the 11 judges on the court, with the remaining four drawn from among senior administrative officers and lawyers and the Council of Higher Education (Yükseköğretim Kurulu, or YÖK). Promotions to appellate courts and the higher judiciary were controlled by the seven-member Supreme Board, itself comprised entirely of judges from the same courts and personnel from the Ministry of Justice. The narrow composition of the HSYK facilitated the imposition of an ideological litmus test on judicial promotions, ensuring that the high judiciary was a relatively politically homogeneous group. The selection procedure for the TCC was then much less important. So long as the pool of eligible candidates could be restricted primarily to the high courts, the promotion system producing the high judiciary would ensure relative ideological conformity.
Following the amendment, the Court will be larger and more broadly representative. In the 17-member TCC, nine of the judges will be continue drawn from the five next highest courts. The remaining eight members of the Court will be drawn from among: senior administrative officers and lawyers; judges and prosecutors from lower courts; YÖK; and one candidate chosen from a pool nominated by Turkish bar association presidents. Thus, whereas the TCC’s composition previously reflected a ratio of seven appellate court appointees to four appointees from other parts of the legal and judicial profession, after the amendment that ratio is now 9:8. This institutional change brings greater representation of the judicial and legal profession onto the Court in line with democratic judicial appointments procedures in Europe and beyond.
With respect to the appointment of particular candidates, claims that the amendment package expanded executive power are disingenuous. Prior to the amendment, all appointments were made by the president from among nominees selected by the high judiciary and other unelected state organs. Under the amendment, the president continues to make the majority of appointments in the same manner, from among a small set of nominees chosen primarily by the judiciary. A larger proportion of eventual seats on the Court will be drawn from a wider spectrum of the judiciary than was previously the case, but the pool of candidates remain restricted primarily to those chosen by the judicial branch. Further, while the role of the executive may be worryingly large in making appointments to the TCC, that role is not a consequence of these amendments. In fact, the democratically elected parliament is now accorded a role in the appointments procedure for the first time, enabling them to fill three of the 17 seats on the expanded Court. In other words, the amendment amounts to reducing, if marginally, the president’s sole authority to select justices from among the pool of judicial nominees.
These changes hardly amount to court packing, even if the AKP were to be guaranteed a durable parliamentary majority. The Venice Commission — the European Union’s Commission for Democracy Through Law, which advises the Council of Europe on constitutional matters — has noted that a parliamentary role in the selection of members of constitutional courts is the prevalent practice in Europe. In the words of the secretary of the Venice Commission, Thomas Markert, constitutional courts “can annul legislation adopted by parliament and therefore need some democratic legitimacy as well as sensitivity to political issues.”  Under the amended constitution in Turkey, parts of the senior judiciary and bar associations, not the parliament, nominate eligible candidates for parliamentary appointment. Moreover, the transition to an expanded TCC occurred by awarding the four current alternate justices — chosen under the pre-amendment procedures favored by the opposition — permanent seats. That left only two new seats to be filled on the expanded Court in the immediate aftermath of the September referendum. In the event, the Turkish parliament filled the two vacancies with one candidate from among three chosen by the Court of Accounts and one from three candidates nominated by the presidents of Turkish bar associations. Thus, although the expansion from 11 to 17 judges seems like a significant increase, only two new members were initially added to the Court under the new appointments procedure. If there is to be court packing by the government, evidently it will not be in the immediate aftermath of the referendum.
The second set of controversial amendments concerns the composition of the Supreme Board of Judges and Prosecutors. Here the amendments significantly alter the body’s makeup, but not in a way that enhances the role of either the parliament or the president in the appointments procedure. Prior to amendment, the constitution provided for the HSYK to be comprised of seven regular and five substitute members, with the minister of justice and the undersecretary to the minister of justice both serving as regular members (the latter ex officio). The remaining five regular members were all drawn from the senior judiciary. Under the amendments, the HSYK is dramatically expanded to include 22 regular members and 12 substitutes. While the minister of justice and the undersecretary continue to be members of the Board, their authorities are reduced, addressing concerns of undue executive influence. In particular, whereas the undersecretary’s presence was previously required to convene a meeting — giving the Ministry an effective veto over HSYK activities if it chose to boycott — HSYK decisions may now be taken in the absence of the members from the Ministry. More importantly, while the president, the Court of Cassation and the Council of State continue to select a proportion of members of the Board, the majority of the expansion draws on a completely new pool of candidates to be selected by judges and prosecutors across the country at lower-level administrative and judicial institutions. In other words, the HSYK will become more representative of the profession at all levels.
This reform was long sought by European officials who found the HSYK too narrow and insulated to offer a democratic standard of accountability in judicial affairs. For instance, the most recent European Commission report on Turkey’s progress on reforms toward accession noted that the HSYK “is not representative of the judiciary as a whole; only senior members of the Court of Cassation and of the Council of State are members of this Council,” and raised concerns about the “independence, impartiality and efficiency of the judiciary.” The Commission specifically noted that the HSYK was involved in what appeared to be politicized dismissals of prosecutors involved in high profile cases.  Given that the HSYK decides on promotions to the courts of appeal in the country, the prior arrangement had produced a self-perpetuating oligarchy of judges that applied ideological and other criteria to ensure that only like-minded members of the judiciary were promoted to senior positions. The expansion of the HSYK to enable junior judges and prosecutors to participate in the election process will enable broader judicial self-regulation, reducing the cliquishness that has characterized the promotions process for three decades.
Far from facilitating AKP control, this amounts to restoring control over judicial appointments and promotions to the judicial branch as a whole. Indeed, the claims that the AKP will necessarily benefit from such an expansion rest on the premise that a Board that better represents the demographic makeup of the judiciary will be more closely aligned with the AKP. On this account, HSYK expansion will inure to the benefit of the AKP because the party enjoys broader popularity in the judicial branch as a whole than it did among the senior judges that previously dominated the Board. This clout is either a result of its general electoral popularity — which is not an automatic advantage, but one the party can only maintain by remaining accountable to the electorate — or a result of a prior bias among the senior judges, which a more representative Board will correct. Either way, there is nothing about the expansion itself that introduces an illicit role for the political branches generally, or the AKP in particular, in the judicial appointments and promotions process. The elections that were held following the referendum to fill the ten directly elected slots on the HSYK were predictably decried by the opposition for creating a Board stacked with AKP-favored candidates.  Critics of such accusations countered that it was difficult to see how the AKP could have unduly influenced a secret ballot involving 12,000 judges and prosecutors, while EU officials deemed the elections for the Board legitimate. 
There is, however, a political backdrop against which concerns about a more representative HSYK can be better understood. Turkish courts are hearing a set of politically charged cases concerning a series of military plots against the civilian government, known collectively as the Ergenekon trials. These cases represent a rare instance in which civilian prosecutors have attempted to investigate retired and serving military generals. Traditionally, such investigations would have been inconceivable and the younger judges and prosecutors that have dared bring cases against the military have done so in the teeth of very serious opposition, often risking their careers. For instance, one young prosecutor chose to investigate a case involving the 2005 bombing of a bookshop in Şemdinli, where the perpetrators were caught by bystanders and turned out to be members of the security services. The three perpetrators were tried and convicted by prosecutor Ferhat Sarıkaya, who also suggested that more senior officers may have been involved and should be investigated. Once Sarıkaya began investigating connections to the higher ranks of the military, and notably Gen. Yaşar Büyükanıt, then chief of general staff, the HSYK intervened first by removing the prosecutor from the case and eventually by disbarring him. In another case, the HSYK stripped Osman Şanal of his prosecutorial powers when he undertook an investigation of another prosecutor, Ilhan Cihaner, on suspicion of involvement in the Ergenekon network. 
Most recently, the HSYK and the government were embroiled in high-profile disputes in 2009 and 2010 as the Board sought to use its power over appointments and promotions to remove or transfer judges and prosecutors involved in the Ergenekon prosecutions. The dispute brought the work of the HSYK to a standstill as members from the Ministry of Justice refused to convene meetings to enable the Board to approve its list of promotions and transfers. By the terms of the referendum, the presence of ministry officials is no longer required for HSYK decisions, so such paralysis will no longer occur. By the same token, the expansion of the HSYK has likely removed the threat of demotion, transfer and dismissal for prosecutors and judges involved in the Ergenekon cases. First, many of the members of the new Board will now be drawn from among the lower echelons of the judiciary, which have been more willing to exercise civilian jurisdiction over military personnel in criminal cases. Second, while the expansion of the Board will not necessarily produce a body that is broadly supportive of the AKP, it will certainly reduce the ideological conformity that was previously its hallmark.
The expansion of the HSYK must also be read in conjunction with other amendments, such as those giving judges and prosecutors the right to judicial review of dismissal decisions. Indeed, the expansions of judicial review under the amendments have the effect of limiting the capacity of both the military and the judiciary to engage in politically motivated purges. While this step generally works to strengthen individual rights and bureaucratic competence, it is also clear that these protections will (at least initially) disproportionately benefit a segment of the AKP’s constituency. There is no question that the primary targets of judicial and military purges in the last decade have been those deemed by the state to be unacceptably Islamist, a category that often overlaps with the AKP’s electoral base. Yet the fact that the provision of judicial review may at first benefit AKP supporters should not detract from an appreciation of the broader implications of the reform. Indeed, if it were not for the history of unfair targeting, there would be no reason to expect these reforms to aid the AKP constituency in particular. The promulgation of judicially monitored professional standards in appointments and promotions in the state bureaucracy, provision of judicial review to civil servants, and immunization of judicial processes from political intervention are all important — and liberalizing — accomplishments regardless of their initial, contingent beneficiaries.
The End of Guardianship
The coincidence of the referendum with the thirtieth anniversary of the 1980 military coup should have served further to reinforce the popularity of the civilianizing provisions of the amendments. Even as the media was saturated with referendum coverage, significant public attention was devoted to the anniversary as a time of reckoning. An exhibition in Ankara entitled “September 12 Museum of Shame,” which opened a week before the referendum, featured displays of torture instruments, reproductions of gallows and artifacts from the hundreds of thousands detained as well as those killed or disappeared during the coup.  In many ways, the demand across much of Turkish civil society for a definitive break with the legacy of the 1980 coup was a unifying current in an otherwise polarized electorate. Yet despite this consensus, deep fears pervaded the debate about the amendment package notwithstanding its liberalization of the military-era constitution. The many significant accomplishments of the amendments toward civilianization — eliminating military court jurisdiction over civilians, empowering civilian courts to try military officers, removing the immunity from prosecution of the 1980 military coup leaders — were all deemed less important than the alleged dangers of judicial reform.
What, then, accounts for the poisonous atmosphere around the judicial reform provisions of the amendment package? The concern, at base, is that the elected branches of government must be kept in check by unelected guardians of the Turkish regime. One lesson of the 1980 military coup, despite its brutal excesses, was that when civilian governments err, the Turkish military is the backstop for the political order. The authority of the military to intervene in the country’s political life has already been substantially curbed in the last decade by a combination of the electoral support enjoyed by the government, the strengthening of Turkish civil society and the willingness of lower courts to entertain prosecutions of military coup planning. But the guardianship role since 1980 has been shared between the army and other parts of the state bureaucracy, including the high judiciary. Each of the unelected branches acts to ensure that the core ideological tenets of the state remain stable, and that the elites most committed to those tenets retain their privileged access to state resources, ranging from subsidies to high-ranking office. Any attempt to subject such guardians to democratic accountability is met with genuine alarm by those elites in Turkey who fear the democratic reversal of their preferences. The fear is not that democracy itself is under threat but rather that democracy represents a threat to a particular configuration of political power and prestige in Turkey. Against this backdrop, the identity of the actors who proposed the amendments matters at least as much as the amendments’ content.
The judiciary has become the locus of political contestation precisely because of the increasingly politicized role the TCC has played in recent years in support of the status quo. The Court’s record has displayed a willingness to intervene against democratic preferences, whether by checking constitutional reform efforts by parliamentary majorities or through the device of political party closures. The starkest examples of the judicial check on democratic initiatives were the Court’s annulment in 2008 of constitutional amendments enabling headscarved women to attend universities and its closure in 2009 of the principal political party representing the country’s Kurdish citizens. Such decisions led some Turks, particularly among the urban elites of the Western cities, to view the Court as a guardian — together with the military and parts of the state’s civilian bureaucracy — of their preferred understanding of secularism and nationalism. What has been at issue in the political contestation between the AKP and these elites is not whether secularism and nationalism are to remain constitutive elements of Turkey’s political order, but rather whether the system can tolerate a reinterpretation of these values that would be more accommodating of private religious expression and ethnic and cultural pluralism.
This question has become increasingly urgent as the rising influence of the AKP’s constituency has led to accelerated elite contestation over the last three years. The AKP’s political fortunes in Turkey ride on the rise of a provincial middle class that is culturally more traditional and religiously more conservative than the elites that are accustomed to governing the country. The neo-liberal economic policies that have fueled Turkish growth, particularly in the last decade, have disproportionately benefited this relatively new bourgeoisie and increased its profile, even in the bastions of the traditional socio-economic and cultural elites of the country’s Western cities. For instance, many commentators within and outside of Turkey argue that the prevalence of women in headscarves has increased over the last decade. In fact, surveys show a decline in the proportion of headscarved women across the country. This disconnect between perception and reality is driven by the increased visibility of women who are religiously conservative and now affluent enough to shop, dine and live in the urban centers that were once the exclusive province of the country’s secular, Westernized elites.
The cultural shift that has accompanied Turkey’s economic growth does not bespeak creeping Islamization as much as it does a recalibration of the political and economic balance of power in the country between the traditional elites of the Western coastal cities and the new Anatolian business communities. The discomfort this shift has caused burst into open political confrontation in 2007 when the nomination of Abdullah Gül for the presidency brought with it the possibility that his headscarved wife would become Turkey’s first lady. The ensuing standoff was resolved then, as now, through voting. In nationwide parliamentary elections in 2007, the confrontation over the presidential nomination resulted in a significant increase in the AKP’s share of electoral support and enabled their choice to become president. Three years later, those opposed to the constitutional amendments were much the same forces that were aligned against Gül’s nomination. Again, results at the polling booth have demonstrated that a solid majority of the country’s electorate favors the amendments, even if it means giving greater voice to a civil society increasingly inflected by the social and cultural conservatism of the Anatolian middle classes.
While some find the cultural markers of this shift alienating, their unease should not cloud analysis of its political implications. The AKP has won significant pluralities in four national elections in the last decade — two municipal (2004, 2009) and two parliamentary (2002, 2007) — bringing a measure of political and economic stability to Turkey following the turbulence and polarization of the 1990s (marked by nine coalition governments from 1993-2002). During its time in office, the AKP has ushered in several significant rounds of constitutional and legislative reforms, each chipping away at the authoritarian legacy of the constitution put in place by Turkey’s military following the 1980 coup.
The results of September’s referendum are a testimony to the commitment of Turkish civil society to continuing the recent trajectory of democratization and liberalization, even in the face of increasing elite polarization. In characterizing the referendum results immediately following the vote, Prime Minister Erdoğan said that the message from the electorate is: “Yes to freedom. Yes to rule of law. No to the law of the rulers. The tutelage of the coup regime is over.”  The reference to the end of the guardianship system was one that the Turkish audience understood readily. Further, the referendum results show that commitment to ending the “tutelage of the coup regime” was far broader than the AKP’s electoral base. As the 2011 legislative elections will doubtless show, while the AKP commands a plurality of the Turkish electorate it cannot muster the near-60 percent majority enjoyed by the referendum. The depth of support for the referendum reflects a generational shift and the anti-military orientation of the electorate. Much of the Turkish commentary following the passage of the referendum confirms that it was more a vote to end the military’s role in the ordinary politics of the country — and to relax the suffocating grip of the state bureaucracy — than an endorsement of any particular party or politician.
The coverage of the referendum results in Turkey and in the West initially told a story of a regionally divided country. One Turkish commentator memorably described the results for an American audience as Turkey’s “red-blue divide.”  This misleading characterization soon gave way to more nuanced maps, showing remarkable consistency across the country in favor of the amendments. Despite highly distorted media coverage leading up to the referendum and the call for a boycott by the leading Kurdish party, nationwide results showed remarkable voter turnout (78 percent outside of the Kurdish provinces) and consistent majorities in favor of the amendment package in most regions. 
Rather than a new fault line in Turkey between the secular western coastal regions and the more religious interior, careful analysis of the referendum results showed that there was a clear majority in favor of the referendum in the great majority of provinces. For instance, Turkey’s largest city, Istanbul, voted in favor of the amendments by a margin of roughly 55 percent, despite being home to a large proportion of Turkey’s wealthy and Westernized elite. A neighborhood-by-neighborhood breakdown of Istanbul tells an interesting story. As an example, the relatively wealthy and staunchly CHP-supporting neighborhood of Beşiktaş voted 78 percent against the referendum, while the nearby neighborhood of Sultanbeyli, relatively poor and AKP-supporting, voted 86 percent in favor of the referendum. By contrast, in the southeast, two neighboring provinces that both resisted the BDP boycott call showed equally striking differences. Tunceli, the Alevi and Kurdish province where current CHP leader Kemal Kılıçdaroğlu was born, voted 81 percent against the referendum, while Bingöl, an adjacent predominantly Kurdish province, voted over 95 percent in favor of the referendum. These results belie efforts to locate an easy regional divide in Turkey and also point to another important fact: The rapid urbanization and accompanying demographic shifts that Turkey has undergone in the last three decades have unsettled the regionalism of “the west vs. the rest” that had previously characterized the country. Communities that migrated from rural parts of central Anatolia to the major and mid-sized cities are altering the composition of those cities in ways that are unpredictable but significant. None of the political parties in Turkey can take for granted the geographic base of their support and all of them need to reach out beyond their ordinary constituencies to become more truly national parties.
“Not Enough, But Yes!”
There were important lessons from the referendum for the principal political parties of the country. While the AKP enjoyed a significant electoral victory in the referendum, it nonetheless must take account of the portion of the electorate that voted against the package. In particular, given the liberalizing content of the amendments the government should have been able to build a consensus in support of the referendum. The AKP’s hand has surely been strengthened following the referendum — for instance, it is unlikely to face party closure under the new TCC, which now requires the votes of two thirds of the Court in order to shut down a party — but such an advantage is only durable as long as the party is able to command a substantial electoral mandate for its agenda. To ensure broad-based support for reform, one lesson of the referendum for the AKP is that future constitutional changes, and particularly the introduction of a new constitution, should be consensus-based rather than simply AKP-driven. Consensus would require genuine outreach to opposition parties to build elite convergence and avoid the polarization that preceded the referendum. That being said, the CHP and BDP would also have to decide that reform is a priority and abandon their roles as veto players.
In some ways, the CHP and the BDP emerged from the referendum strengthened. The CHP’s new leader, Kılıçdaroğlu, showed that he could run an effective campaign and rebuild the base of the party that had withered under the leadership of his predecessor, Deniz Baykal. But following the constitutional amendments, the CHP has no choice but to fashion a reform agenda of its own and indeed it has come out in favor of a new civilian constitution. The important question will be whether it is able to carve out a role for itself as a consensus builder despite its erstwhile obstructionism. The BDP, similarly, showed that its constituency could represent an important swing vote in the southeast and that the AKP cannot afford to exclude them from discussions. In many ways, the strengthening of these opposition voices during the referendum process is beneficial to the Turkish political system in providing clear alternatives for the electorate while also channeling elite contestation into political channels. The one party that suffered clear losses in the referendum was the ultra-nationalist, right-wing National Movement Party (MHP). Basing its opposition to the amendment package on its anti-Kurdish and pro-military stance, the MHP was unable to muster its own base against the referendum. Even Osmaniye, the hometown of MHP leader Devlet Bahçeli, voted 53 percent in favor of the referendum, despite the party’s aggressive “no” campaign.  In the aftermath of the referendum, each of the opposition parties would do well to transform its platform in a direction that appeals to a wider swathe of the nation.
The referendum showed a clear preference among the Turkish electorate in favor of political liberalization. In this context, a realistic assessment of the meaning of the constitutional referendum would endorse the slogan embraced by many who supported the amendments package: “Yetmez! Ama evet!” (Not Enough, But Yes!). While the particular amendments represent a clear improvement over the status quo and accomplish reforms required for EU accession, they do not go far enough in addressing the shortcomings of the 1982 constitution. For the parliamentary elections of 2011, the AKP has suggested that a new civilian draft constitution will be part of its campaign platform.  That initiative would be welcomed by those who criticized as opportunistic the piecemeal approach taken in the 2010 amendments package. In the meantime, the amendments endorsed in the September referendum should be understood as the most recent iteration of the incremental process of democratization that has been set in motion, with the support of the AKP, over the last decade. They do not eliminate the authoritarian and statist tenor of the constitution, but they ameliorate many of its excesses, as have the five prior constitutional amendment packages since 1999.
In light of the poisonous political climate that surrounded it, however, the passage of this particular package of amendments represents something more than prior rounds. The referendum was a much-needed rebuke to those who dismiss the Turkish electorate as easily manipulated and continue to treat the AKP as a stalking horse for creeping Islamization. Whatever the intentions harbored by members of the AKP or the opposition, the reforms represent another step in the direction of improved fundamental rights, judicial accountability and civilian control over government. While the test of some of these reforms will be in the implementation, there is much to celebrate for Turks as the thirtieth anniversary of the 1980 military coup ushers in a welcome set of amendments and confirms the ongoing commitment of the Turkish electorate to the path of political liberalization.
 The official English translation of the amendments package (which reflects the minor changes mandated by the TCC ruling) is available here: http://www.hurriyetdailynews.com/videos/justiceministryreferendum.pdf.
 Agence France Presse, September 13, 2010.
 Agence France Presse, September 12, 2010.
 The 2010 Report on Transatlantic Trends found a precipitous decline in Turkish popular support for EU membership. Transatlantic Trends, “Key Findings 2010,” available online at: http://www.compagniadisanpaolo.it/eng/file/pdf/TT_ING_2010_567.pdf.
 Haldun Gülalp, “The Battle for Turkey’s Constitution,” Guardian, September 4, 2010.
 Hürriyet Daily News and Economic Review, August 26, 2010; Today’s Zaman, September 26, 2010.
 Today’s Zaman, March 25, 2010.
 European Union Commission, Turkey 2009 Progress Report (October 14, 2009), http://ec.europa.eu/enlargement/pdf/key_documents/2009/tr_rapport_2009_en.pdf (at pages 11 and 69-70).
 Cumhuriyet, October 22, 2010.
 Hürriyet Daily News and Economic Review, October 22, 2010; Hürriyet Daily News and Economic Review, October 27, 2010.
 New York Times, February 20, 2010.
 Today’s Zaman, September 9, 2010.
 Washington Times, September 12, 2010.
 Aslı Aydıntaşbaş, “In Turkey, a Red-Blue Divide,” Wall Street Journal, September 17, 2010.
 The Turkish Supreme Board of Elections has published the official results, including provincial and municipal breakdowns, at: http://www.ysk.gov.tr/ysk/ReferandumSecimSonucServ?bilmece1=j5m4ls.
 Hürriyet Daily News and Economic Review, September 13, 2010.
 Hürriyet Daily News and Economic Review, September 12, 2010.