Brief 47: Refugee Messaging and Host Society Attitudes in Turkey
EGAP researchers: Tolga Sinmazdemir, Thomas Zeitzoff
Other authors: Anna Getmansky
Geographical region: Europe
Research question: What factors influence citizen perceptions of refugees and do these factors also influence attitudes towards domestic conflict?
The arrival of refugees from Syria’s ongoing civil war has been linked to heightened tensions in the neighboring countries of Turkey, Lebanon, and Jordan. At the time this study was conducted (June 2014), there were already about 1 million Syrian refugees in Turkey (currently there are more than 3 million). Turkey continues to host more Syrian refugees than any other country. In Europe, the influx of Syrian refugees has been connected to increased xenophobia, a surge in support for far-right parties, and instances of violence between locals and refugees. Indeed some studies that use cross-national data find a positive correlation at the national level between influx of refugees and conflict in receiving countries.
Previous research presents several ways in which refugees’ arrival may trigger instability in host countries. These include increased economic competition between the refugees and the local population, disruption of the ethnic balance in the host society, and arrival of people with fighting experience and ties to rebel groups. Importantly, the negative externalities of refugees’ arrival—economic, ethnic, and security-related—do not have to be real to have a destabilizing effect. Fueled by elite discourse and social media, negative perceptions of refugees may be sufficient to increase tension and make violence more likely. In this study, the authors directly explore how messages about refugees can influence the perceptions of refugees and affect attitudes and preferences of local individuals.
The authors conducted a survey experiment on 1,257 Turkish citizens who resided in southeastern and central Turkey in June 2014. The survey experiment presented different messages about the possible effects of hosting refugees—increased economic burden, disruption of ethnic balance, and ties with rebels, as well as a positive message of saving innocent women and children. These messages were tailored to resemble elite cues as they appear in the media. The authors were interested in how these messages affect the locals’ perceptions towards Syrian refugees, and attitudes towards the Turkish-Kurdish peace process. In addition to these messages, the authors also examined how partisanship, ethnicity, religiosity, and actual exposure to refugees in the course of the respondents’ daily lives affected their attitudes towards the refugees. The sample covered areas in Turkey with varying presence of refugees, as well as varying degree of exposure to past political violence (the Turkish-Kurdish conflict). The sample included both Kurdish and non-Kurdish respondents (Kurds being the primary ethnic minority in the sample area). The survey was conducted through face-to-face interviews.
The authors of this study report several important findings. First, negative messages about Syrian refugees’ effect on Turkey lead to negative views about the refugees, but primarily among the ethnic majority (non-Kurdish respondents). Among the negative messages, the authors find that emphasizing a possible link between the refugees and militants can have a negative effect on the views of refugees. Interestingly, the authors find that positive messages about most Syrian refugees being innocent women and children also have a negative impact on the locals’ attitudes towards refugees. Second, the authors find some evidence of co-ethnic bias. Turkish-Kurdish respondents are more positive towards Syrian-Kurdish refugees than towards other refugees. Non-Kurdish respondents do not exhibit such attitude. Relatedly, Sunni refugees are not as disliked by Turkish respondents (the majority of whom are also Sunni) as other refugees. Third, the authors do not find that information about Syrian refugees affects respondents’ positions on the Turkish-Kurdish conflict (support for the peace process). Fourth, respondents who report higher exposure to refugees in their daily lives (seeing them on the street, in public transportation, engaging in commerce with refugees, etc.) exhibit more negative attitudes and have a higher threat perception towards refugees. Finally, the authors find that partisanship matters. Supporters of the ruling Justice and Development Party (AKP) are less likely to view the refugees as a threat.
Given the political upheaval from the Syrian refugee crisis in host countries, finding ways to reduce prejudice towards refugees is an extremely important public policy issue. The findings from this study have several implications for designing policies to promote refugees’ integration into their host societies. First, the authors show that negative messages about mostly-Muslim refugees can negatively impact attitudes towards them even in a middle-income, Muslim-majority country such as Turkey. Based on these findings, the authors can expect the effect of such messages to resonate even more strongly in non-Muslim host countries (i.e. Europe, the US, or Canada). This finding highlights the importance of correcting negative misperceptions about refugees, particularly those that portray them as security threats. This suggests that it is crucial to reassure the local population that refugees are not a security threat—one example would be by addressing the concern that Syrian refugees may be connected to militants. Second, the authors find that a positive message about women and children does not necessarily increase support for refugees and may backfire depending on how the locals interpret this information. This is relevant for government bodies and NGOs seeking to promote positive views of refugees among the host population. Finally, and perhaps most importantly, the authors show that while locals can be swayed to hold negative views of the refugees, this does not spillover to hardened attitudes on domestic conflicts such as the Turkish-Kurdish conflict. Messaging is more likely to affect attitudes than established policy positions.