Turkey’s domestic politics, public opinion and Middle East policy

This article looks at the domestic political conditions that help to shape the Turkey’s Middle East policy. The starting point is that the country’s foreign policy is not self-standing or determined solely by external circumstances, but is often affected by domestic divisions that shape party identities. We can classify these into the following categories: first, historically determined cultural and ethnic cleavages; second, public opinion on foreign policy issues; and, third, current domestic policy considerations that have important foreign policy implications. In tackling this agenda, certain restrictions have to be recognised. First, we have virtually no data as regards public attitudes on economic or social issues, so it has not been possible to say anything about these. Second, time series data are unavailable, as public opinion polls have only recently been made available in Turkey. This article is published as part of a collection on analysing security complexes in a changing Middle East.

Historical and cultural legacies and alignments T wo underlying features of Turkey's political culture have significant effects on its foreign policy, especially towards the Middle East. A defining alignment arises from the reforms implemented by Kemal Atatürk's government during the 1920s. These established the Turkish republic as a secular national state, in place of the multi-national Ottoman Empire, which had based its legitimacy on its attachment to Islam (at least, in the eyes of its Muslim subjects). In cutting the link between the state and Islam, the republic was, in effect, aiming to end the main cultural and historical bond between Turks and Arabs. This reorientation was not always straightforward, however. Since the state did not seek to deny the validity of Islam as a system of personal belief, but tried to accommodate it within a cult of ethnic nationalism, the official discourse sought to de-Arabise (even de-spiritualise) Muslim history and culture, rather than exclude the Turks from it (Aktürk, 2010).
Popular attitudes were also shaped by memories of the Ottoman defeat in the First World War, and the Arab revolt of 1915 in particular. For much of the succeeding generation, the attempt to keep the Ottoman Empire going in the Middle East was seen as a wasted effort, in which thousands of Turkish soldiers had been pointlessly killed. The clear implication was that Turkey should avoid involvement-especially military-in a treacherous and unrewarding region. The distant Yemen, in which countless wars had been fought, was chosen as an archetype. When the Caliphate was abolished in 1924, Kemal Atatürk dwelt on the price the Turks had paid for commitment to a pointless cause. "And what happened? Millions of them died, in every land they went to. Do you know how many Anatolian boys perished in the sweltering heat of the deserts of the Yemen (Lewis, 1955: 81)?" As a popular song laments: "before the barracks, the sad guitars are playing/ My feet are bare [a sign of mourning] and my heart is breaking/ For those going to the Yemen, the young girls cry/ … Those who go never return. I wonder why (Pope, 1990)".
Since the 1980s, some important revisions of these attitudes have taken place, both at the official and popular levels. The movement labelled as "Neo-Ottomanism" has helped to re-habilitate the Ottoman past, while side-stepping its defects. It began as a project to promote the multi-culturalism of an idealised vision of the Ottoman Empire in place of Kemalist monolithic ethnic nationalism. Following the end of the Cold War, it acquired a reference to foreign policy, as Turkey aimed to develop relations with those countries whose territories had once been part of the empire, both in south-east Europe and the Middle East. After the Justice and Development Party (Adalet ve Kalkınma Partisi, or AKP) came to power in the general elections of 2002, the main architect of these approaches was Professor Ahmet Davutoğlu, who served as foreign policy adviser to Prime Minister Tayyip Erdoğan until May 2009, when he became Foreign Minister. He took over the premiership in August 2014, when Erdoğan was elected President of the Republic. In a book published in 2001, Davutoğlu urged that Turkey possessed "strategic depth" and "multiple regional identities", which it had hitherto failed to exploit (Davutoğlu, 2001). He argued that it should abandon the old defensive and securitised attitudes of the cold war, imagining that Turkey was forever "surrounded by enemies". Instead, it should work for "zero problems" with all its neighbours.
This approach was applied, in particular, to Turkey's Middle Eastern neighbours, although Davutoğlu rejected the label of "neo-Ottomanism", emphasising that Turkey had no intention of trying to reconstruct its former empire (Davutoğlu, 2004). In the years preceding the "Arab spring", his policies resulted in what looked like a successful friendship offensive, developing strong and friendly links with the Arab world. These new approaches were not universally supported, however, since closer alignment with the Arab states and Iran was seen by strictly secularist opinion as backsliding towards Islamism. During the 1990s, attitudes towards Israel were an important touchstone, with militant secularists in the armed forces and parts of the intellectual elite supporting a more positive attitude towards the Jewish state than the cultural conservatives. Shifts in policy towards the Middle East thus reflected the dominant cultural cleavage in Turkish society. In defence of Davutoğlu's strategies, it is argued that they reflected pragmatic interests rather than religious sympathies: for instance, he did not de-emphasise Turkey's relations with the United States and EU, and the treatment of Muslim minorities in Russia and China was not allowed to become an obstacle in relations with Moscow and Beijing (Jung, 2011-12: 36). The conflict with Russia, which erupted in November 2015 was because of severe policy disagreements over Syria, not Islamism.
While it would be hard to argue that Muslim commitment has been the main inspiration of Turkish foreign policy under the AKP, the fact that the party is seen as Islamist by its domestic opponents has admittedly had a powerful effect on domestic power balances, and hence on the general direction of foreign policy. Traditionally, the military commanders, as the selfappointed guardians of Kemalism, exercised a powerful influence over Turkish foreign policy. This largely accounted for its highly securitised nature and suspicious if not out rightly hostile attitudes to neighbouring states. However, the AKP period has seen a dramatic reduction of the independent political power of the military, caused partly by constitutional and legal reforms designed to bring Turkey into harmony with EU norms, and partly by the revelation of alleged futile plots against the AKP government by members of the high command (Gürsoy, 2011(Gürsoy, , 2012 1 . After 20071 . After -2008, the decline in the political power of the military allowed Davutoğlu to adopt "soft power" approaches to regional states, emphasizing the benefits of cooperation rather than confrontation. Besides this perceived contest between the "Muslim democrats" of the AKP and its militantly secularist opponents, the challenge of Kurdish separatism has produced the second main cleavage in society-essentially between the Kurdish identity, and Turkish ethnic nationalism, which rigidly rejects any form of territorial division of the republic. This emerged during the 1980s with the launch of a campaign of violence by the PKK, or Kurdistan Workers' Party, with terrorist attacks on civilian targets and a ruthless counter-insurgency campaign by the security forces. The on-off struggle has continued since then, with periodic ceasefires and abortive attempts at a political settlement. In the meantime, the Kurdish minority has secured explicit representation in parliament, currently in the shape of the Peoples' Democratic Party (Halkların Demokratik Partisi, or HDP). Its most outspoken opponents are to be found in the Nationalist Action Party (Milliyetçi Hareket Partisi, or MHP) (Ünal, 2016). 2 This confrontation has important foreign policy implications, partly in relations with the western democracies, where human rights groups and liberal opinion generally have been harshly critical of Turkish government policies, but also in the Middle East, thanks to existence of large Kurdish minorities in Iraq, Syria and Iran. While Kurdish separatism appears to have been a less powerful movement in Iran than in the other countries affected, it has resulted in the establishment of the Kurdistan Regional Government (KRG) in Iraq, which now enjoys virtual independence, although it is constitutionally part of the Iraqi republic. This has crucial implications for Turkey, since the PKK has long used bases in Iraqi Kurdistan in support of its attacks in Turkish territory. After years of suppression by the Syrian Ba'thist regime, since the start of the Syrian civil war the Kurdish minority in Syria has also carved out a substantial zone of control in the north of the country, under the Democratic Unity Party, or PYD. This has strong links with the PKK, and is hence regarded with grave suspicion in Turkey. On the other hand, its militia wing, known as the People's Protection Units (YPG) enjoys full support from the United States, as an important element in the coalition fighting the ultra-Islamists of the Islamic State organisation (IS). This has provoked serious disputes between the Turkish government and the Obama administration. Since the Turkish Kurds generally enjoy good relations with Massoud Barzani's Kurdistan Democratic Party (KDP), the rulers of Iraqi Kurdistan, and some are supporters of the PKK, conflicting domestic pressures in foreign policy are created.
Public opinion and the two Gulf Wars, 1990Wars, -2003 Two striking examples of the effects of public opinion on Turkish policy occurred during the two Gulf wars of 1991 and 2003. On the first occasion, President Turgut Özal gave full diplomatic support to the American-led coalition which sought to end the Iraqi occupation of Kuwait. Further than this, he also sought to send a Turkish contingent to join the international force in the Gulf, and even considered opening a second front in the war against Saddam Hussein's regime, by sending Turkish forces across Iraq's northern frontier. However, his hands were tied by Article 92 of the Turkish constitution. This requires the passage of a special resolution by parliament to allow the government to declare a state of war, to send Turkish forces abroad, or receive foreign forces on Turkish soil, unless this is required by "international treaties to which Turkey is a party [in effect, the North Atlantic Treaty] or by the rules of international courtesy". Neither of the last two conditions applied in this case. Thanks to powerful parliamentary opposition, not only from the opposition parties, but also an important part of the ruling Motherland Party's parliamentary group, Özal was forced to back down In fact, the furthest the parliamentarians were prepared to go was to allow coalition air forces to use the joint Turkish-NATO air base at İncirlik, near Adana, for operations in Iraq (Hale, 2006: Chapter 2). This was of value to the coalition, which appreciated Özal's role, but demonstrated the persistence of public opposition to military intervention in the Middle East.
For Turkey, the dilemmas posed by the Gulf crisis of 1990-1991 were repeated, in a more acute form, in 2002-2003. When the AKP government came into office in November 2002 the Pentagon planners were preparing concrete proposals for the invasion of Iraq, although the claim that Saddam Hussein still possessed weapons of mass destruction was not proven. The US military planning was more ambitious than that of 1990-1991, as the Pentagon now proposed to place substantial land and air forces in southern Turkey, and to use Turkish supply routes, to support an invasion of Iraq from the north in addition to the main offensive in the Gulf. The AKP government, then under Abdullah Gül, was far from enthusiastic about this plan, but eventually decided that since the Bush administration was determined to go through with the invasion anyway, Turkey would be better off inside the American tent than outside it. It also insisted that Turkish forces should simultaneously be inserted into a "security belt" in northern Iraq (Ibid: Chapter 4). As in 1990-1991, the main obstacle it faced was fierce opposition from public opinion, including its own grass-roots supporters. Opposition was especially strong among the Kurdish minority, which opposed the proposed Turkish buffer zone in Iraq, as a mortal threat to the KRG. According to surveys carried out by the Turkish polling organisations Anar and Pollmark between December 2002 and September 2003, around three-quarters of their respondents opposed the US-led invasion of Iraq (Dalmış et al., 2005). Fundamentally, as in 1990-1991, the opposition was based on longstanding reluctance to get involved in a Middle Eastern war, especially one which was seen as unnecessary and unjustified.
The public opposition to the war evidently had powerful effects on the AKP's parliamentary group, including several leading members, such as the Speaker, Bülent Arınç and Deputy Prime Minister Ertuğrul Yalçınbayır, who spoke out against the American plan (Yetkin, 2004: 147, 151). The scale of this reaction became evident on 1 March 2003, when a motion to give the government powers under Article 92 (in effect, to allow the invasion plan to go ahead) was laid before parliament. Normally, the AKP would have had no difficulty in passing its draft, since it had an overwhelming majority of 365 of the 550 parliamentary seats. On this occasion, however, it was estimated that 68 AKP deputies voted with the opposition, with another 31 casting abstentions of absenting themselves. 3 With the motion lost, the government was effectively left without a policy on Iraq, but was at least saved from being sucked into the Iraqi quagmire.

Domestic politics, public opinion, and Middle East Crises, 2011-2015
More recently, a numerous opinion polls have appeared in Turkey, giving us far more information on how the voters feel about foreign policy. One needs to be cautious about this of course, since such polls are not always accurate, and foreign policy normally appears to have less influence on voters' decisions at election time than domestic issues such as the state of the economy, public security, and social services. Where public opinion backs a policy line which is quite impractical, or contrary to the country's international obligations, then governments are prepared to ignore it. For example, in a poll conducted by the Turkish Economic and Foreign Policy Research Centre (EDAM) in February-March 2012, 58 percent of respondents supported the idea that, in response to a possible nuclear threat from Iran, Turkey should develop its own nuclear weapons (Ekonomi ve Dış Politika Araştırma Merkezi, 2012). Thus is a project which has never been adopted by any democratically elected government in Turkey, and would be quite contrary to its obligations under the international Non-Proliferation Treaty. In response to the question asking how Turkey should react to the influx of refugees from Syria, over 55 percent of respondents surveyed in December 2013 suggested that the inflow should be stopped, with almost 30 percent even supporting the idea that the existing refugees should be sent back to Syria (Ekonomi ve Dış Politika Araştırma Merkezi, 2014). 4 The findings were confirmed by another poll conducted in April 2015 by the Turkey Research Centre of Istanbul's Kadır Has University (Kadır Has, 2015: slide 37). Again, these proposals were both impractical and contrary to recognised humanitarian principles.
On the other hand, public reluctance to support military intervention in Middle Eastern conflicts continued to have a powerful effect on government policies. Thus, in a survey conducted by the polling organisation Metropoll in October 2014, over 60 percent of the respondents opposed the idea that Turkey should intervene militarily in the Syrian civil war, even if this were part of a US-led or NATO operation, with only 22.5 percent in support. However, the government's eventual decision in July to allow coalition air forces to use the İncirlik base for attacks on the Islamic State (IS) forces, and to participate in joint attacks with coalition forces, would appear to have had some public support, with 20 percent of the Kadir Has poll respondents supporting the first proposal, and another 17 percent the second (Ibid: slide 10). Even before a horrific IS suicide bomb attack in Ankara on 10 October 2015, in which over 100 people were killed, 86 percent of Metropoll's respondents agreed that IS was a danger to Turkey, virtually equivalent to the PKK (MetroPOLL, 2014). However, a poll organised by the German Marshall Fund of the United States in July 2015 (well before the Ankara attack) also found that only 24 percent of its respondents thought that Turkey should join the international coalition against IS actively, with 38 percent believing it should stay out of the coalition entirely, and 23 percent that it should support the coalition in non-military ways (German Marshall Fund of the United States, 2015: 10). Similarly, in April 2015 the Kadır Has poll found that 65 percent of respondents recognised that IS was a threat to Turkey, but 47 percent thought Turkey should stay out of the struggle against it (Kadır Has, 2015: slides 35-36). Public opinion thus seemed ambiguous and inconsistent.
Broadly speaking, public support for the AKP's policies in the Middle East has been far from overwhelming, even among the party's own supporters. According to the German Marshall Fund survey, 70 percent of respondents thought that Turkey should first deal with its internal problems, with only 20 percent supporting the idea that it should play a more active role in the Middle East, the Balkans and Central Asia. The general public thus appeared to have a more realistic view of the country's priorities and capacities than Ahmet Davutoğlu, the supposed foreign policy guru. Overall, 51 percent of respondents disapproved of Turkey's current foreign policy, while 41 percent approved-although these figures are not very meaningful by themselves, since they do not tell us what aspects of foreign policy were supported or opposed, and by which sections of the public (German Marshall Fund of the United States, 2015: 9).
On specific issues, the AKP government took a clear stand on two aspects of policy towards the Syrian civil war. The first of these was that a "safety zone" should be established in Syrian territory along the frontier with Turkey, approximately 30(thin space)km deep and 100 km long, covering the section of the frontier between the Syrian towns of Jarablus in the east and Azaz in the west. This was territory currently occupied by IS, filling in the gap between a PYD-controlled zone to the east, and the short section of territory under the Free Syrian Army (FSA) which was aligned with Turkey and Saudi Arabia, to the west (Ergun and Kasapoğlu, 2015: 7). 5 The aims were to remove IS from the frontier region, to provide a secure and protected base for FSA operations, a zone in which refugees from the rest of Syria could be re-settled and supplied, and a territorial buffer to prevent the two sections of PYD-controlled territory from joining up (the other being at the western end of the international frontier). This plan was criticised as being quite impractical, thanks to the lack of international support, and serious military and legal obstacles . It was supplementary to consistent Turkish support for the FSA, plus some anti-IS Islamist resistance groups, by providing basing facilities and arms supplies. It appeared, however, that neither of these policies had more than marginal support from the Turkish public. In a survey conducted by EDAM in 2012, only 15 percent of those questioned supported the "safe zone" idea and 8 percent the continuation of arms supplies to the Syrian opposition (without saying which groups this referred to). Only 11 percent of the respondents favoured direct Turkish military intervention against the Assad regime, which admittedly had never been government policy. The majority (56 percent) opted for no intervention in Syria of any kind, or merely limited political and diplomatic activity (Ekonomi ve Dış Politika Araştırma Merkezi, 2012).
A prominent feature of the AKPs policies was its sharp opposition to the Ba'thist regime in Syria, which it regarded as primarily responsible for the outbreak of the civil war, and as barbaric as IS. It was far from clear that this unremitting hostility was shared by most of the public, however. In a poll conducted in March 2015, only a minority (44 percent) of Metropoll's respondents supported the idea that any solution of the Syrian crisis would have to include changing the regime, with 31 percent opposing it, and 25 percent maintaining that "it doesn't matter who rules Syria" (MetroPOLL, 2015a: 7). Seven months later, in October 2015, 49.9 percent of the respondents stated that they would agree to talks with the Assad regime if this were necessary for a solution, with 28 percent opposing and 22 percent not expressing any opinion on this idea. Significantly, the idea of talks with the regime enjoyed very nearly as much support among the AKP's own grass-roots 6 as the average of the respondent group as a whole (MetroPOLL, 2015b: 27).
This public reaction reflected criticisms of the government's Syrian policies by academics and the media. As the respected newspaper columnist and author Murat Yetkin pointed out, these had originally been based on the mistaken assumption that the overthrow of Assad would be swift and sharp, like that of Qaddafi, and that the Muslim Brotherhood would then take over, as it initially had in Egypt. This overestimated the cohesion on the opposition side, and the regime's survival power. As the war dragged on, Turkey found itself in a strategic cul-de-sac, continuing to identify the Assad regime as the primary enemy, while the western powers concentrated their attacks on IS, leaving Assad aside, and even cooperating with Russia, his main ally. Admittedly part of the resulting impasse was due to the fatal lack of consistent policy by Turkey's western allies. On 21 August 2013 the Assad regime used chemical weapons in a horrific attack on the Damascus suburb of Ghouta, reportedly killing over 1,000 people. A year earlier, on 20 August 2012 President Obama had threatened that if it used chemical weapons, the regime would cross a "red line" which would change his earlier decision not to intervene militarily in Syria. Accordingly, 5 days after the Ghouta attack, the US and British warships were deployed in the eastern Mediterranean in preparation for imminent missile strikes against determined Syrian targets. At the last minute, on 30 August, the President decided to back down, and later accepted a Russianbacked plan under which Assad's chemical weapons were removed under UN auspices (Hürriyet Daily News, 23 February 2016; Goldberg, 2016: 70-76). 7 The failure to act decisively was strongly criticised in Turkey, as elsewhere: besides undermining America's credibility as a world power, and allowing Russia to play a leading role. It prolonged the war in Syria, opening the door to the rise of IS as a much more serious threat to international security. Since the Obama administration refused to put "boots on the ground" in Syria, it outsourced its ground operations to the Kurdish militia of the PYD/YPG, refusing to have this classified as a terrorist organisation, in spite of its accepted links with the PKK. Naturally, Turkey complained bitterly about this, and had domestic public support on this issue, but could do little to prevent the US-PYD alliance.
Polls on other aspects of the government's Middle East policies suggested that they did not enjoy public support or that opinions were, at best, evenly divided. Asked what was the greatest threat to Turkey in a poll by EDAM in 2013, 36 percent of all respondents thought that this would be the establishment of a Kurdish state to the south of Turkey, with 12 percent identifying the worst threat as the establishment of an ultra-Islamist state in Syria. Predictably, in case of Kurdish respondents 8 these figures changed to four percent and 47 percent respectively. Although the government played down the danger that Iran might develop nuclear weapons, 10.6 percent identified this as the most dangerous threat to Turkey, with 16 percent citing the danger of an attack on Iran by either the United States or Israel (Ekonomi ve Dış Politika Araştırma Merkezi, 2013b). ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057/palcomms.2016.81 Elsewhere in the Middle East, Turkey's relations with Egypt were also contentious. As Prime Minister, Tayyip Erdoğan openly advocated the overthrow of Hosni Mubarak at an early stage of the "Arab spring". His outspoken support for the opposition won him widespread support in Egypt, especially among partisans of the Muslim Brotherhood, and he received a hero's welcome when he visited Cairo in September 2011 (Today's Zaman, 2011). Having hailed the election of the Brotherhood's candidate Muhammad Morsi as Egypt's President in June 2012, Erdoğan and Davutoğlu were sharply critical of his overthrow by coup d'état in July 2013, confirming that they both regarded Morsi as the legitimate President. The subsequent administration of General Abdul Fattah al-Sisi reacted sharply by expelling the Turkish ambassador from Cairo and accusing Turkey of interfering in Egypt's internal affairs. This prompted the Turkish Foreign Ministry to declare the Egyptian ambassador in Ankara persona non grata (in fact, he had earlier been withdrawn by his own government) ( BBC News, 2013). Turkish policy could be described as morally justified, but it had no practical effect, since Turkey had very little influence over the course of Egyptian politics. The results of a poll conducted by EDAM in 2013 suggested that public opinion was well aware of this, with 48 percent of those questioned saying that events in Egypt were the Egyptians' own affair, and Turkey should avoid criticism: another 12 percent thought that Turkey should be critical of Morsi's overthrow, but accept the result of the subsequent election. 9 Predictably, the AKP's own supporters, who could be expected to have a more favourable attitude to the Muslim Brotherhood than supporters of other parties, were more supportive of the government's line than the sample as a whole, but even among this group 37 percent thought that Turkey should leave Egyptian politics to the Egyptians. It thus appeared that on this issue, as on some others, the Turkish public was more realistic than the government (Ekonomi ve Dış Politika Araştırma Merkezi, 2014).
Relations with Israel were another critical issue. In May 2010 the Turkish cruise ship Mavi Marmara, which was carrying relief supplies to the besieged Palestinians in the Gaza strip, was attacked by the Israeli security forces in international waters, killing nine people and wounding some 50 others. In response, the Turkish government broke off ambassadorial relations with Tel Aviv and demanded an apology from the Israelis, with compensation for the victims' families and an end to the Gaza siege (International Crisis Group, 2010: 4-9). 10 On this issue, the AKP government appeared to have broad public support in Turkey. In a poll conducted by EDAM in 2013, only 7 percent of respondents suggested that relations with Israel should simply be restored without conditions. A measure of the strength of the opposition to the Israeli action was that supporters of the CHP, who in the past could have been expected to be more favourable to Israel, were in close agreement with AKP supporters on this issue. Attitudes towards Israel among Kurdish voters differed little from the national average (Ekonomi ve Dış Politika Araştırma Merkezi, 2013a). It thus appears that, thanks to its aggressive policies, the Netanyahu government had succeeded in alienating virtually all sections of opinion in Turkey.

Domestic political upsets and Middle East policy, June 2015-2016
Turkey's domestic politics passed through a turbulent period during the summer and autumn of 2015 . 11 In general elections held on 7 June, the AKP suffered the first electoral defeat in its history, winning only 258 seats in the new parliament, and thus falling 18 seats short of an overall majority. In subsequent negotiations with the Republican People's Party (CHP), as the next biggest party, the AKP failed to form a coalition government, allegedly due to interference by President Erdoğan (Kılıçdaroğlu, 2015). 12 Accordingly, the President exercised his powers under Article 116 of the constitution to declare new elections, with the date then fixed as 1 November 2015.
The following election campaign was overshadowed by a rising tide of violence by the IS and PKK militants, with the consequent resumption if military operations against both of them by the state. On 20 July a deadly bomb attack in the mainly Kurdish-inhabited town of Suruç, close to the Syrian border, killed 32 youth activists attending a meeting. The culprit was identified as a 20-year old Turkish Kurd, who had strong links with the IS terrorists in Syria. Opposition media argued that the government was partly to blame for the attack, since it had failed to secure the border, or to take a sufficiently tough line against IS. In response, PKK terrorists killed two policemen in their homes in the town of Çeylanpinar, ending the ceasefire which the PKK had declared two years earlier as part of an attempt to reach a political settlement with the Turkish government ( BBC News, 2015a, b). The resumption of terrorist attacks suggested that control over the PKK was now slipping out of the hands of its veteran leader Abdullah Öcalan, imprisoned in Turkey since 1999, and into those of hard-line leaders based in Iraqi Kurdistan, such as Murat Karayılan and Cemil Bayık. Worse was to come on 10 October, with the horrific suicide bomb attack in Ankara already referred to. In response, the Turkish air force joined in further attacks on the IS in Syria, but concentrated its main efforts against PKK bases in south-eastern Turkey and Iraqi Kurdistan. This was accompanied by large-scale land operations against the PKK, and arrests of suspected IS members or supporters. There thus seemed to have been a return to the dismal cycle of violence which had dragged on since the PKK's campaign had begun in 1984.
Naturally enough the worsening security situation, along with predictable economic issues, dominated the election campaign in September-October 2015, with little discussion of substantive foreign policy questions, such as Turkish policy towards the Syrian civil war or other middle eastern countries. The Republican People's Party urged that there should be important changes to Turkish policy in the Middle East, such as establishing contact with the Assad government in Damascus and the normalisation of relations with Israel, as well as cooperation with Egypt and non-interference in its internal affairs. 13 On the other hand, the section of the AKP's manifesto dealing with foreign policy was brief and sketchy-in sharp contrast to Ahmet Davutoğlu's previously ambitious rhetoric. In Iraq, the party undertook to continue its policy of supporting the establishment of a "comprehensive" administration (implicitly, embracing the Sunni and Kurdish communities as well as the dominant Shi'a). It had nothing to say about relations with the Kurdistan Regional Government in Iraq, although these were of crucial importance in the context of Turkey's internal Kurdish problem. Nor was there any mention of relations with Egypt. Policy towards Israel was touched on, but only to say that relations could not be normalised until Israel lifted its blockade of the Gaza strip (that is, besides paying compensation to the families of those killed on the Mavi Marmara). On Syria, the party undertook to continue aid to refugees from Syria and Iraq, but said nothing about the proposed "safety zone", aid to the opposition militias, policy towards the Assad regime, or the recent Russian intervention in the civil war (Ak Parti, 2015: 225). In sum, it was hard to escape the impression that most of the "Davutoğlu doctrine" had been quietly dropped, in the face of the multitude of problems it faced, but without any clear replacement. The idea that Turkey might act as a democratic mentor to the rest of the Middle East had also been abandoned, in the face of the evident failure of the "Arab spring" everywhere except in Tunisia, the land of its birth.
In the run-up to the November elections, opinion polls predicted that they would simply repeat the results of the previous June-or that, at most, the AKP might be returned to power with a wafer-thin majority. In the event the voters surprised the pundits, including senior members of the AKP (Yetkin, 2015), by giving the party 49.5 percent of the vote, and 317 of the 550 parliamentary seats. This represented a comfortable overall majority of 84 seats. Barring internal party splits, the AKP could be expected to stay in power for the next 4 years. 14 Hanging over the government, however, was Tayyip Erdoğan's ambition to convert Turkey from a parliamentary to a presidential republic in which, as President, he could enjoy wide political powers, reducing the Prime Minister to a subservient role. To achieve this would officially require amendment of the constitution, for which support of one or more of the opposition parties-or a part of them-would be required. 15 In practice, Erdoğan evaded the need for amendment by controlling the AKP's national organisation, and thus the parliamentary group, but he still apparently felt the need to formalise this situation. This move was opposed by some senior members of the party, with Prime Minister Davutoğlu apparently far from enthusiastic about the idea of a presidential republic (Hürriyet Daily News, 2016f, h). 16 When the final showdown between the President and Prime Minister erupted in early May 2016, Erdoğan clearly had the upper hand. Davutoğlu put up minimal resistance. On 5 May he announced that he would step aside on 22 May, when a special AKP party convention would assemble to choose a successor (Ibid, 2016g). On 19 May the party announced that Binali Yıldırım, formerly Minister of Transport and a close aide of Erdoğan, would be the sole candidate for the premiership. Three days later, his election by the convention was just a formality (Ibid, 2016j). As a result of what one of his critics called "Tayyip Erdoğan's power grab", the President achieved full executive power, with Yıldırım his faithful servant, in clear defiance of the constitutional convention that the President should be little more than a neutral figurehead (Yeşilada, 2016: 19).
While this domestic political drama was unfolding, the bitter conflicts in which Turkey was involved continued unabated. The destructive struggle against the PKK continued into May 2016, when Prime Minister Yıldırım announced the end of military operations (Hürriyet Daily News, 2016m). According to President Erdoğan, since the start of the campaign more than 7,600 militants had been "neutralized" (meaning either killed or captured)-a figure that apparently included IS terrorists as well as the PKK. Meanwhile, 600 police and soldiers had been killed, with an unreported number of civilians and 11,000 homes destroyed in the fighting (Yetkin, 2016b). In Syria, the bitter civil war dragged on, with the failure of successive ceasefires. The IS survived, to carry out repeated terrorist attacks as well as controlling a large chunk of Syrian as well as Iraqi territory. The conflict acquired a new and dangerous dimension in September 2015, when Russia carried out its first air strikes from its large base in Syria. In spite of Russian claims that this was part of the struggle against IS, it was clear that Moscow's main aim was to support its ally President Assad, by re-balancing the civil war in his favour. The clear sign of this was that the vast majority of the Russian air strikes were aimed not at IS but at the "moderate" opposition forces supported by Saudi Arabia and Turkey, who were fighting both IS and the Assad regime. The mounting toll of civilian casualties caused by these strikes was denounced by Davutoğlu as "vile, cruel and barbaric" (Hürriyet Daily News, 2016a). However, without risking a head-on military collision with Russia, in which it was doubtful that it would have the support of NATO (in effect, the United States), there was little Turkey could do in response. The nearest it came to doing this occurred on 24 November 2015 when the Turkish air force shot down a Russian SU-24 fighter jet which had strayed into Turkish territory. Both the two crew members were killed, one of whom was allegedly shot from the ground by a Syrian-Turcoman militiaman, after he had successfully ejected from the aircraft. In retaliation, the flow of Russian tourists to Turkey was cut to a fraction of its former three million per year. Russia also banned fruit and vegetable imports from Turkey, causing further damage to the Turkish economy (Ibid, 2016i, I).
What made Turkey's policy towards Syria even more problematic was that, while it was in direct confrontation with Russia, the PYD-PKK connection meant that it was also at odds with the United Sates. Like several European leaders, President Obama complained to Erdoğan about the lack of press freedom in Turkey. The war in Syria was the cause of more serious conflict, in which Turkey was accused of taking an insufficiently tough line against IS. After the downing of the Russian aircraft, Turkish ʼplanes had to be withdrawn from the coalition forces attacking IS targets from the İncirlik base, so as to avoid another direct confrontation between the Turkish and Russian air forces. Turkish land-based artillery was however engaged in repeated exchanges of fire with IS forces just across the border in Syria. More seriously, the "Kurdistan Freedom Falcons", identified as a proxy for the PKK, claimed responsibility for another deadly bomb attack in Ankara on 17 February 2016, causing 29 deaths, besides numerous injuries. Owing to its close connection with the PKK, Turkey blamed the PYD/YPG, and opened artillery fire on PYD positions in Syria, continuing the barrage for several says. As a result, Turkey was engaged in an open clash with America's main ally in Syria. Turkish spokesmen claimed that the PYD was not interested in destroying the IS, but carving out an autonomous Kurdish corridor along the Turkish border, with the help of Russia (Ibid, 2016b, c, d, e).
Conclusions and prospects: a return to "factory settings" in Turkish policy? After the change in the premiership, commentators naturally wondered whether this would mean a redirection of foreign policy, and the abandonment of Davutoğlu's goals in favour of a less ambitious or idealistic agenda (Bagdonas, 2015: 310-331). 17 Such a change had been foreshadowed over several years. In a penetrating paper, Ziya Öniş and Mustafa Kutlay see the period since 2011 as a "third phase" in the AKP's foreign policy in which the transition to an illiberal democracy, as Erdoğan expanded his personal power, drained it of the moral authority, or soft power, which it had exercised in its early years after 2002. Lower economic growth contributed to this effect (Kutlay and Öniş, 2016: 11-14). In historical perspective, this could be seen as a move back to Turkey's traditional strategies, relying more on hard than soft power, and more on realism rather than idealism. Significantly, the Kadır Has University's survey of public attitudes, conducted in 2015, suggested that ordinary citizens had reverted to the isolationist, "loan wolf" attitudes which Davutoğlu had rejected or maybe they had never abandoned them. Thus, asked what was Turkey's "closest friend", 39 percent of the respondents replied that Turkey had "no friends", with another 38 percent mentioning the fellow-Turkic republic of Azerbaijan, or the self-proclaimed Turkish Republic of Northern Cyprus (nine percent). In this ranking, middle eastern countries barely figured, with only Saudi Arabia and Iran having any significant score, with three and two percent respectively (Kadır Has, 2015: slide 13). As the journalist and academic Verda Özer suggested, Turkey seemed to be moving back to its "factory settings" in foreign policy, from which Davutoğlu had sought to break away (Özer, 2016).
Soon after Yıldırım's election as premier, it was suggested that Erdoğan would now opt "for a complete overhaul of Turkey's ties with countries in the immediate neighbourhood and beyond", including the restoration of relations with Israel, Egypt and Russia; (Demirtaş, 2016) (in fact, this was not a new decision, since a cabinet meeting under Davutoğlu in February 2016 had decided to try to mend the rift between the first two countries, and negotiations with Israel had been going on for some time) (Babacan, 2016). The first results came through on 28 June, when an agreement was signed normalising diplomatic relations with Israel. According to the agreement, Turkey would deliver humanitarian aid to Gaza and Israel would pay US$20 million in compensation to the victims of the attack on the Mavi Marmara. Meanwhile, a government spokesman claimed that there were "signs of softening" in Russia's attitude to Turkey, although a settlement was apparently some way off, (Hürriyet Daily News, 2016n, o). A reference by the Prime Minister to the Syrian struggle as a "meaningless war" (Ibid, 2016k) had also prompted speculation to the effect that Turkey might abandon its support of the "moderate" anti-Assad forces in Syria, but there was no clear sign that the government would go this far. Further afield, and looking to the not-so-distant future, there was a chance that US policy on Syria might change if, as widely expected, Hillary Clinton became the occupant of the White House in January 2017. Earlier in the war, she had urged that the US should put more effort into arming the non-Islamist opposition militias. Speaking in November 2015, she supported the Turkish plan for a "no-fly" or security zone in northern Syria, as well as the despatch of more US ground troops to the battle against IS. She was also critical of Turkey and the Gulf states for allegedly not doing enough to tackle Islamist fundamentalism, and insisted that Turkey must "lock down its border" with ISheld territory (Gambino and Roberts, 2015;Gearan, 2015;Goldberg, 2016: 73). However, this gave no hint as to whether the alliance between the Pentagon and the PYD would be continued, or how the "security zone" could be created in the face of fierce opposition by Russia. The restoration of the old spirit of cooperation between Turkey and its NATO allies also depended on a settlement of Turkey's internal Kurdish problem, and the restoration of democratic norms in Turkish politics. Almost certainly, Turkey's domestic political issues would continue to have a crucial effect on foreign relations, as they had in the past.