Explaining the dog that does not bark: Why do some localities in Turkey remain resistant to Islamist political mobilization?

Evren Celik Wiltse, South Dakota State University


Most studies on political Islam in Turkey have embraced a comprehensive, national level of analysis. From the classic “center-periphery” paradigm (Mardin, 1973) to recent works that explain the rise of political Islam (Gulalp, 2001; Arat, 2005; Turam, 2007; Gumuscu, 2010) or its downfall (Tugal, 2015), scholars focus mostly on macro transformations. These include socio-economic changes since the 1980s, new state-society dynamics, the rise of a “strong devout bourgeoisie” and successful bottom up mobilization efforts of political Islam in large urban centers. Few researchers (Turam, 2007; Tugal 2015) expand their analysis to include the transnational outreach of the Turkish Islamist experience.

In terms of their methods, many of the studies that explain political Islam in Turkey include key explanatory variables on a national scale. The Turkish state’s rigidly secularist disposition since the early Republic, effective mobilization of Islamist movements/parties as a reaction to top-down secularism, rise of conservative business networks since the neoliberal tide of the 1980s, expansion of civil society, urban marginalization, and women’s organizations all offer excellent explanations for the strengthening of Islamist movements and parties in Turkey.

While most of these studies successfully explain the rise of Islamist movements, many of them are not particularly sensitive to local variation and agency.[i] Even ethnographic studies that try to reach out to the marginalized sectors (White, 2013; Tugal, 2015) focus on the periphery of the largest metropolis, Istanbul – Umraniye and Sultanbeyli respectively.

Despite these compelling studies, an intriguing puzzle remains: why do certain pockets of Turkey seem relatively immune to the rising religious political wave?

This paper tries to explain the relative failure of Islamist politics in certain localities of Turkey, with reference to their distinct regional socio-cultural and economic characteristics. As Marschall Shalaby and Konak Unal demonstrate, local politics is not merely a venue for redistributive policy. Issues of identity and morality also play an important part, for they constrain “mayors’ ability to govern, manage conflict, and win elections.” (Marschall, Shalaby and Konak Unal, 2017:2) Briefly, this paper argues that having a distinct local socio-cultural identity and certain economic means to sustain it can make communities relatively immune to the inroads of Islamist politics. Due to methodological and practical complexities of accurately gauging the power of relatively non-transparent Islamic movements, electoral support for religious parties is used as a proxy in this paper. In addition to election data, the paper utilizes qualitative data from fieldwork conducted in Edirne and Kirklareli in the northwest, and Hatay in south-central Turkey.

Can a secular versus Muslim lens capture the whole picture?

At one level, Turkish society appears to be divided into secular and Muslim positions whose proponents are circling the wagons and demanding ideological and behavioral purity from their members. But neither term –secular or Muslim- does justice to the variety of the possible positions and their sometimes surprising combinations.” (White, 2013: 10)

After each election in Turkey, particularly since 2011, the predominant interpretation seems to be the “polarization” between modern, secular coastal cities against the conservative, Muslim core. Consolidation of multiple center-right and religious-right parties (ANAP, DYP, Genc, and Refah) under the Justice & Development Party (AKP), and left parties (DSP and CHP) under the Republican People’s Party (CHP) helps breed this perception. Since ultra-nationalists and Kurds are under constant existential crisis due to the 10 percent electoral threshold, scholars argue that Turkey has gradually replaced its chronic electoral volatility with a dominant party system, wherein the Islamist party takes the center stage and seculars take the main opposition seat. (Carkoglu, 2011) While Izmir symbolizes the stereotypical voter profile of the secularists with its modern, educated and elite outlook, places like Konya, Kayseri or Ümraniye symbolize the core constituencies of the religious political bloc. This paper tries to peel away some of these dichotomous assumptions about the secular and religious voters, elites and counter-elites of Turkey, by zooming in on three specific localities. (Map 1)

Map 1: Administrative Units in Turkey

Source: www.d-maps.com

Both Edirne and Kirklareli in the northwest are known to be historical strongholds of CHP. However, unlike other CHP strongholds, such as cosmopolitan Izmir, Eskisehir, or Ankara-Cankaya, boasting multiple universities, educated and affluent upper-middle classes, Edirne and Kirklareli are relatively small, economically less developed, and socially more parochial cities. In their region 45 percent of women active in the labor force (ages 15 to 49) work in the agricultural sector, which is 5 percent higher than the national average. (TNSA 2008: 51) Given their large rural populations that make a living from subsistence agriculture, it is difficult to paint these localities as the hubs of Turkey’s secular elites.

A third province included in this study is Hatay located in south-central Turkey with Syria on one side and the Mediterranean on the other. Hatay contains a medley of cultures, religions, and ethnicities. Housing the historic cities of Antakya (Antioch) and Iskenderun, Hatay prides itself for being the “cradle of the three Abrahamic religions.” It stands out for having Turkey’s only Armenian village, Vakifli, located on the legendary Musa Dagh.(Eski, 2015) Numerous holy sites of Islam and Christianity are scattered across the province. In the old city center of Antakya, one can see insignias on the facades of old stone houses that display the profession and faith of the residents. This is probably the only city in Turkey where one can hear church bells ringing regularly during the daytime. Yet it is not a typical secular/elite province such as Ankara, Izmir, or Eskisehir either. Compared to these cities, Hatay is much more religious and much less affluent. (see Table 1 below)

Table 1: Ranking of Turkey’s Provinces According to GDP per capita (2014)

1 İstanbul 21 EDIRNE* 41 Aydın 61 Malatya
2 Kocaeli 22 Kayseri 42 Amasya 62 Giresun
3 Ankara 23 Erzincan 43 Gaziantep 63 Ordu
4 Tekirdağ 24 Rize 44 Zonguldak 64 Tokat
5 Bilecik 25 Tunceli 45 Sivas 65 Ardahan
6 İzmir 26 Artvin 46 Afyon 66 Iğdır
7 Bolu 27 Balıkesir 47 Aksaray 67 Kilis
8 Bursa 28 Trabzon 48 Gümüşhane 68 Adıyaman
9 Antalya 29 Uşak 49 Kırşehir 69 Mardin
10 Eskişehir 30 İçel 50 Niğde 70 Bingöl
11 Yalova 31 Karabük 51 Bartın 71 Diyarbakır
12 KIRKLARELI* 32 Isparta 52 Çorum 72 Kars
13 Muğla 33 Konya 53 Sinop 73 Siirt
14 Çanakkale 34 Kastamonu 54 Elazığ 74 Bitlis
15 Denizli 35 Kütahya 55 HATAY* 75 Şırnak
16 Karaman 36 Kırıkkale 56 Yozgat 76 Muş
17 Manisa 37 Çankırı 57 Osmaniye 77 Batman
18 Sakarya 38 Adana 58 Kahramanmaraş 78 Hakkari
19 Düzce 39 Samsun 59 Erzurum 79 Van
20 Burdur 40 Nevşehir 60 Bayburt 80 Şanlıurfa
81 Ağrı

Source: TUIK. Calculated by the author based on Turkish Statistical Institute data on GDP of each province and the population of each province for 2014.

When we look at the voting patterns of these three provinces since 1991, we can observe certain variation that sets them apart from national trends. Chart 1 displays the support for religious parties, including the ruling AKP and its predecessors. Despite the progressively increasing vote share of the Islamist parties across Turkey, it shows that these three provinces consistently rank below the national average. Average support for the Islamist parties is about 4 percent lower in Hatay, and nearly 20 percent lower in both Edirne and Kirklareli.


Chart 1: Percentage of AKP Votes in General Elections in Edirne, Kirklareli and Hatay, 1991-2015:

Source: Compiled by the author from the Election Archives of the Turkish Supreme Election Council (YSK) www.ysk.gov.tr

When we look at the support for the CHP, variation of these three provinces becomes even more striking. All three provinces score much higher than the national average since 1991. On average, CHP receives 10 percent more votes in Hatay, and around 14-15 percent more votes in Edirne and Kirklareli.

Chart 2: Percentage of CHP Votes in General Elections in Edirne, Kirklareli and Hatay, 1991-2015:

Source: Compiled by the author from the Election Archives of the Turkish Supreme Election Council (YSK) www.ysk.gov.tr

Why swim against the tide?

The elite/secular center versus Islamist periphery template does not explain the strong support for the opposition CHP in small towns and villages of Edirne, Kirklareli, and Hatay, which are full of small-scale merchants and subsistence farmers. The majority of the population in these provinces is far from being highly educated, secular elites. In fact, both Trakya and Hatay are famous for their fertile agricultural lands, and outside of a few urban centers, there is also a significant portion of rural population. Nearly half of economically active women work in agriculture. The puzzle is: why would majority of these less privileged, rural residents turn their backs to the tangible benefits that come with a vote for the ruling Islamist party? Why do they insist on voting for the secular opposition?

Teune (1995: 14) argues that, “local structures and cultures are less amenable to change than are higher levels of political organization.” Constitutional amendments, bureaucratic or economic reforms are much easier to execute form the center, in comparison to breaking away the “familiar dominance of political influence” in the periphery. Consequently, we should not be too surprised to see these provinces in the periphery resisting the Islamist political wave that is impacting the metropolitan centers of Turkey. The type of morality and identity politics in which Islamist parties engage elsewhere, particularly in large urban centers, might undermine their success in these localities in the periphery of Turkey.

Religion is the crucial mortar that holds together the social fabric in all three of these provinces. However, the ways in which these communities incorporate religion are not necessarily the same way as many political Islamists. Compared to other parts of Turkey, religiosity in Trakya stands out in several aspects. First and foremost, the gender gap in Trakya is not as wide as the rest of the nation. Women and men work in the fields together and celebrate together in weddings and holidays. On average, more than 35 percent of women in the region between the ages of 15 and 49 are active members of the labor force as of 2008, this is 5 percent higher than the Turkish national average. (TNSA 2008) Marriage among relatives, including distant cousins and even close neighbors, is strictly forbidden. (Yedi kuşak akrabayla evlilik olmaz.) (Celik Wiltse 2014-16) This social norm eases the tension among sexes, making socialization much easier since they do not consider each other as potential spouses, or haram in all contexts.

Unlike other parts of rural Turkey, strict public segregation of sexes is neither practical, nor enforced in Trakya. Given the low population density and record low birth rates,[ii] men and women need to work side-by-side in the region. Women can work alone as day laborers, without being shunned by their communities. This relative gender parity also extends into property rights. Turkish civil code grants equal share of inheritance to all siblings, yet in most parts of Anatolia brothers are favored over sisters. However in Trakya, my respondents emphatically stated that all siblings get equal share, regardless of gender. (Celik Wiltse 2014-16)

Secondly, there is the unique way in which this devout Sunni Muslim population practices its religion and tradition. Complying with duties and transmitting them to the next generations are important goals that shape day-to-day conduct. However, they also insist on practicing these traditions the way their ancestors did, and react negatively to external influences. (Atalarımızdan, dedelerimizden gördüğümüz gibi ibadet ederiz.) (Celik Wiltse 2014-16)

In her impressive ethnographic work, Kimberly Hart captures the clash between different ways in which Islam can be experienced in any given community. The unorthodox Islam takes the glorified Ottoman past as its source of legitimacy, whereas the more modern and politicized form of Islam looks at the future to establish its ideal Islamic society. Hart compares and contrasts two villages in western Turkey. The first adheres to more traditional, heterodox practices, while the other one, influenced by the Islamic movements and brotherhoods (particularly by Süleymancılar), tries to “purify” religion from these backwards influences and seeks legitimacy in a future, Islamist form of polity. (Hart, 2013: 153-4)

An overwhelming majority of rural Trakya probably resembles the first village in Hart’s classification. They are proud of their Ottoman heritage, and religious ceremonies harken back to that golden era. (Hurriyet, 2016) Similar to Hart’s villagers, “they share a patriotic dedication to the nation-state, and an uncritical merging of Turkish ethno-national and Sunni Islamic identities.” (Hart, 2013: 23) Patriotism and discretion towards national institutions are noticeable in people’s daily habits as well. Across Trakya, there seems to be greater reverence for the public TV channels (TRT). The Directorate of Religious Affairs (Diyanet) serves as the sole arbitrator on religious matters, particularly among the middle aged and older generations. In almost every house, there is a Diyanet calendar used frequently as a reference, and most people tune in to TRT for news and Ramadan programs. In personal conversations, residents express reverence towards their imam, a civil servant appointed by the Diyanet and are decidedly negative about “outsiders” who come to preach them about “their own religion.” (Celik Wiltse, 2014-16) In sum, the pious rural population of Trakya adheres to a more traditional form of religion, and the mainstream Sunni interpretations of Diyanet, which repels the modern, puritan interpretations that are frequently advocated by Islamist movements. However, given “overtly political” turn that Diyanet is taking, (Fabbe, 2016: 12-16) there might be backlash or some lasting changes in the region.

The components of Hatay’s social fabric, and the ways in which religion is incorporated into the social life also serve as a salient barrier that, at a minimum, slows down the penetration of Islamist political movements in this province. Similar to Trakya, Hatay residents do not automatically fit the elite/secular profile either. However, while Trakya enjoys relative ethno-national and religious homogeneity, the opposite quality in Hatay shapes the social fabric of the province. Residents of Hatay pride themselves for having harmonious relations for centuries, despite their incredible ethnic, national, and religious diversity. Many of its residents are fluent in Arabic. Intermarriage among Muslims and non-Muslims, as well as among Sunnis and Alevis is not uncommon. People talk about a “delicate balance” that has been historically sustained in the region that enables each group to respect each other and mutually thrive. (Celik Wiltse, 2016)

In terms of their socio-economic outlook, Antakya, Iskenderun, Arsuz and Samandag all have vibrant communities with strong tourism and service sectors, renowned local cuisines, and a strong tradition of entertainment. Even the ongoing civil war in neighboring Syria does not seem to have dampened the nightlife in many of these cities. Compared to its conservative neighbors to the north, Hatay is more open, socially less restrictive, and overall, more accommodating of diversity and heterodoxy.

Having established strong footholds in the region, particularly in Gaziantep, Adiyaman and Urfa, the ruling AKP has been consistently flexing its muscles in Hatay, to gain electoral support. Many locals lament that it tries to do so by rocking the “delicate balances” in Hatay, and pitting the ethno-religious groups against each other. For example, prominent CHP politician and party spokesperson[iii] Selin Sayek Böke of Antakya was systematically harassed for her Christian heritage. (Haberturk, 2016) Locals state that in Hatay (population 1.5 million), most people knew the faith of their neighbors, but it was considered vulgar to express it publicly, let alone use it for political gain. However, since the AKP came to power, they claim each election tested ethnic and religious loyalties. (Celik Wiltse, 2016)

After several electoral defeats, the AKP resorted to extreme gerrymandering in Hatay. Three new municipalities were created in 2012, separating neighborhoods and villages that were consistently voting against the ruling party. According to most locals, the lines were drawn almost exclusively along sectarian lines, segregating the Alevis into separate enclaves. Defne was separated from Antakya city center. Neighborhoods that were long administratively and socio-economically part of Iskenderun were cut off and linked to the brand new administrative center of Arsuz, compelling people to travel 30 kilometers for basic services. Finally, Payas emerged arbitrarily, by dividing off the southern one-third of Dörtyol.

Map 2: Gerrymandering & its Impacts on Constitutional Referendum in Hatay

Map Authors: Dinesh Shrestha & Murat Keçeci, using ArcGIS 10.3. software, South Dakota State University, Department of Geography, September 2017.

As the 2017 referendum results illustrate, elections in Hatay remain a thorn in the side of the ruling AKP. Gerrymandering transformed Hatay into an extremely polarized province. In the 2017 referendum, Defne and Samandag scored record high pro-CHP votes (with more than 90 percent voting “no”), while Yayladagi and Altinozu next door were staunch pro-government constituencies (with over 70 percent voting yes). Despite these deliberate attempts to segregate the population between Sunni Turkmens and the rest, the AKP could not succeed in winning the metropolitan municipality of Antakya. During my interviews, local CHP leaders stated that collectively, they decided to not play the “sectarian game” that the AKP was trying to set up. (Celik Wiltse, 2016) The CHP nominated an openly pious, Sunni candidate for Antakya mayor, despite having its strongest electoral support from majority Alevi districts. Yet they were able to beat the heavyweight AKP candidate, Justice Minister Sadullah Ergin in the 2014 local elections.

The ruling party has continued to liberally apply sectarian pressure over Hatay. The recently established “village guard” (korucu) positions, which come with a rifle, steady paycheck and healthcare and retirement benefits, favor only Sunni villages. Arab Alevis of Samandag express growing sense of discrimination, especially in terms of refugee settlements practices, since the outbreak of Syrian civil war. Despite all the gerrymandering and sectarian public policies, Islamist politics have made only partial gains in the province. Local social dynamics and identities matter greatly in smaller provinces of Turkey.


Each province offers a densely rich palette of diversity in Turkey. Even the most homogenous, nationalist areas harbor variation in terms of ethno-nationalist heritage. Migrants from the Balkans or the Caucasus, nomadic Turkmens along the Mediterranean and settled Turkmens of central Anatolia, the Azeris in the northeast, Kurds in the southeast, and the Roma in the northwest all contribute to complex socio-cultural dynamics at the local level. Islamist movements in Turkey have to negotiate their way within this medley of distinct communities and identities. While some of these collective identities might be welcoming towards Islamic movements and parties, others might form effective barriers against them. How a group collectively defines itself at the local level, and the predominant norms and values they endorse, are important indicators of their relative predisposition towards Islamic mobilization. The works of Marschall et al (2017) clearly illustrate how local governments in Turkey have increasingly become critical venues to project the predominant norms and identity aspects of provincial communities.

Broad-brush explanations at the national level may not capture the significant variations and identity dynamics at the local level. Variables that may have strong explanatory value at the national level may not be as pliable at a local scale. This paper tried to illustrate that the average CHP voter in rural Trakya, who is fasting during Ramadan and probably saving money for a coveted visit to Mecca, is very different from the average CHP voter in Izmir. In fact, these two CHP voters might be very different from the average CHP voter in Hatay, who probably speaks fluent Arabic. In short, attention to local level politics offers significant insights. It can unpack some of the dichotomous ways of seeing Turkish electorate as monolithic pro-government Islamists in the heartland versus secular elites along the coasts. Second, local dynamics can help explain why, despite all their material perks and vast organizational networks, Islamist movements cannot gain headway in certain parts of the country. If localities have distinct socio-cultural identities that are inhospitable to political Islam and some economic means to resist the perks of patronage, they may stay outside the Islamist sphere of influence, despite the victory of Islamist politics at the national level. Comparative studies of localities that have similar socio-economic and identity characteristics but different voting behavior can offer important puzzles for political analysts.

[i] For an example of research that is particularly sensitive to local characteristics, see “Islamic Calvinists: Change and Conservatism in Central Anatolia” by the European Stability Initiative http://www.esiweb.org/pdf/esi_document_id_69.pdf Ironically, the local networks and bonds of trust in Kayseri province that are much praised in this report later became a serious liability. When the ruling AK Party and its once closest ally, the Gulenist religious network, parted their ways, businesses in Kayseri became targets and were overtaken by publicly appointed trustees.

[ii] Almost all surveys highlight low fertility rates in Trakya. TUIK data on population density in 2016 shows that Edirne and Kirklareli have nearly half the population density, compared to the Turkish average. (TR average, 104; Edirne, 66; Kirklareli, 56). Edirne also scores negative population growth in four of the last nine years that are recorded. Hatay on the other hand shows robust expansion with double-digit growth rates, and is one of the more densely populated provinces with a density rating of 267 in 2016. (Turkey average: 104). TNSA -13 survey results also show Trakya region having the highest rate (30%) of nuclear families in Turkey without children. (Average for Turkey is 17.9%) p.20. http://www.hips.hacettepe.edu.tr/tnsa2013/rapor/TNSA2013_ilerianaliz.pdf

[iii] She quit her position as the party spokesperson over post-referendum controversies within the party in Spring 2017.


Arat, Y. (2005). Rethinking Islam and Liberal Democracy: Islamist Women in Turkish Politics. Albany: State University of New York Press.

Carkoglu, A. (2011). Turkey’s 2011 General Elections: Towards a Dominant Party System? Insight Turkey, 13(3), pp.43-62.

Celik Wiltse, E. (2014-2016) Interviews with residents of villages in Kirklareli (İnece, Kocahidir, Akardere, Havsa), party members and election observers in Edirne, Hatay, Defne and Samandag.

Eski, B. (2015). Story of “Turkey’s Last Armenian Village” Vakifli. BBC Turkish, 22 April. http://www.bbc.com/turkce/haberler/2015/04/150422_son_ermeni_koyu [last accessed 8 Sept. 2017]

European Stability Initiative. (2005). Islamic Calvinists: Change and Conservatism in Central Anatolia. http://www.esiweb.org/pdf/esi_document_id_69.pdf [Accessed 6 Sept. 2017]

Fabbe, K. (2016). Forward Facing and Behind the Scenes: Shifts in Political Islam in post-July 15 Turkey. POMEPS Studies 22, Contemporary Turkish Politics, pp. 12-16. https://pomeps.org/wp-content/uploads/2016/12/POMEPS_Studies_22_Turkish_Politics_Web.pdf [Accessed 8 Sept. 2017]

Gulalp, H. (2001).Globalization and Political Islam: The Social Bases of Turkey’s Welfare Party. International Journal of Middle East Studies, 33(3), pp.433-448.

Gumuscu, S. (2010). Class, Status and Party: The Changing Face of Political Islam in Turkey and Egypt. Comparative Political Studies, 43(7), pp. 835-861.

Haberturk. (2016). Böke: Part of my family is Christian, the other part is Muslim. 10 Feb. http://www.haberturk.com/gundem/haber/1193672-boke-ailemin-bir-tarafi-hiristiyan-diger-tarafi-musluman [Accessed 8 Sept. 2017]

Hacettepe University. (2008) Turkey Population & Health Research Main Report. http://www.hips.hacettepe.edu.tr/tnsa2008/data/TNSA-2008_ana_Rapor-tr.pdf [Accessed 4 Jul. 2017]

Hart, K. (2013). And Then We Work For God: Rural Sunni Islam in Western Turkey. Stanford: Stanford University Press.

Hurriyet (2016). Territories Left Behind but Traditions Continue. 14 Sept. http://www.hurriyet.com.tr/topraklar-terkedildi-ama-gelenekler-yasatiliyor-40223564 [Accessed 1 Jul. 2017]

Mardin, S. (1973). Center-Periphery Relations: A Key to Turkish Politics? Daedalus, 102(1), pp. 169-190.

Marschall, M., Marwa S. and Konak Unal, S. (2017) Municipal Service Delivery, Identity Politics and Islamist Parties in Turkey. In POMEPS-GLD Workshop on Islamists and Local Politics. Gothenburg, Sweden.

Ministry of Family and Social Policy. (2013). Research on the Structure of Family in Turkey. http://ailetoplum.aile.gov.tr/data/54292ce0369dc32358ee2a46/taya2013trk.pdf [Accessed 5 Jul. 2017]

Teune, H. (1995). Local Government and Democratic Political Development. The ANNALS of the American Academy of Political and Social Science. 540(1), pp. 11-23.

Tugal, C. (2015). The Fall of the Turkish Model: How the Arab Uprisings Brought Down Islamic Liberalism. London: Verso.

Turam, B. (2007). Between Islam and the State: The Politics of Engagement. Stanford: Stanford University Press.

Turam, B. (2004). A bargain between secular state and Turkish Islam: politics of ethnicity in central Asia. Nations and Nationalism, 10(3), pp. 353-374.

White, J. (2013). Muslim Nationalism and the New Turks. Princeton, NJ: Princeton University Press.