A Reflection on Turkey’s Contested Internet and Public Demand for Internet Freedom

Turkish media expert Dr. Bilge Yesil reflects on findings from the newly released report “Benchmarking Demand: Turkey’s Contested Internet.” In her post, Dr. Yesil examines what the report’s results mean for researchers, policymakers, and internet freedom activists in Turkey. Click here to read the full report.

Do a quick online search on Internet policy in Turkey, and it’s very likely that you will see reports and news stories with “censorship fears,” “decline in online freedoms,” and “Internet crackdown” in the title. It is also likely that you will read about how Internet and free speech activists in Turkey are decrying the AKP (Justice and Development Party) government’s heavy-handed approach to online communications, and maybe you will even get the sense that the level of support for Internet freedoms is quite high among ordinary users. You will also come across news stories that celebrate Turkish users’ expert circumvention of social media bans: “Battle-trained… users…quickly turned on VPN services that reroute access through other countries to conceal the point of access to a platform, effectively nullifying the [Twitter] blackout.”

 I bring up these talking points because they shed light on our assumptions about Internet users in Turkey—that they are concerned about online restrictions, do support Internet freedoms, and easily bypass social media bans. But I have often wondered who these users are. What is their socio-economic background? Are they the tech-savvy youth? Are they in fact the “upper classes…the ones who can afford the technology”? What do users think about Internet restrictions? How much of a difference do socio-economic background, education level and political party affiliation make in their approval/disapproval of the government’s Internet policies? Are there users who perhaps are not concerned about Internet censorship at all (gasp!) and do think that social media threatens traditional values? If so, who are they?

The OSU/Koc/Annenberg survey offers valuable insights into these questions and many more. To the best of my knowledge, it is the first of its kind to provide data that illuminates the demographic composition of users and non-users, their political party affiliations, and perceptions of and responses to online restrictions. According to survey findings, non-users in Turkey tend to be older, female, with primary school education. They come from larger households that have lower monthly incomes; self-identify as religious; and support the ruling AKP. In contrast, heavy users (those who use the Internet every day) are young, male, with secondary or higher education. They tend to come from smaller households with higher monthly incomes; self-identify as secular; and support the main opposition party CHP (People’s Republican Party) or no party at all. Light users (those who use the internet 2-3 times a week or less) are similar to heavy users except that they tend to be more religious and support the AKP (For more details, see Table 1). These findings resonate with other research data concerning the demographics of Internet users in Turkey. According to 2015 data, 40% percent of users are between the ages of 15 and 24, and 29% between 25 and 34. A 2014 survey has found that 34% of Internet users self-identify as middle class; 25% are high-school graduates, and 52% have a university degree.

The OSU/Koc/Annenberg survey provides granular data on demographic composition, frequency of use, popular online activities, level of familiarity with the country’s media policies (Figure 3), level of satisfaction with the quality of online content (Figures 4 and 5), and perceptions of social media as a source of threat (Figure 6). Due to space limitations, I wish to highlight this last point since it has been heavily instrumentalized by the AKP government to gather support for its Internet restrictions. In 2013, during the anti-government Gezi protests, then-Prime Minister Erdogan labeled Twitter a “menace” and “a curse on societies” that harbors “all sorts of lies.” And when the government banned Twitter in 2014, Erdogan rationalized it by portraying Twitter as a malicious, foreign company that’s part of a wider conspiracy to weaken Turkey, its national unity and social cohesion. However, it must be remembered that vilifying online content, Internet companies and users is not necessarily a new phenomenon. The AKP government has always relied on such populist tactics to rationalize the passing of strict Internet legislation, extensive filtering and blocking of websites, and the high number of content removal requests it has made to social media companies.

In light of this widely-distributed rhetoric, especially in the pro-government mass media, one cannot help but wonder if users and non-users agree that social media is a menace to Turkish society. According to the survey, 41% of non-users, 46% of light users and 39% of heavy users do think that social media promotes Western values as opposed to Turkish ones. Forty-three percent of non-users, 40% of light users and 36% of heavy users agree that foreign countries use social media to weaken or destabilize Turkey. Forty-five percent of non-users, 30% of light users and 34% of heavy users also think that social media threatens Islamic teachings and beliefs (For more details, see Table 4). It would be interesting to see the political party affiliations of users (both heavy and light) who agree with this view. Do politically conservative voters tend to have a more negative perception of social media than their more liberal, secular counterparts? Also, is there any correlation between users’ perceptions of social media and their consumption of broadcast and print news media sources? In other words, are users who consume pro-government news media more likely to have a negative view of social media?  I think citizens’ political leanings and broader issues relating to their news media consumption are important questions that future surveys can and should address.

Having made this recommendation, I must note that the OSU/Koc/Annenberg survey makes significant contributions to our understanding of Internet users in Turkey, specifically in regards to the role of political party affiliation in shaping citizens’ perception of and response to Internet censorship. (Once again, due to space limitations, I am not able to go into the methodology, but I invite others to share their feedback). In my opinion, and as the authors of the forward, Aysenur Dal and Golnoosh Behrouzian, note, the survey lays bare the fact that Internet freedom is a highly politicized issue in Turkey. For example, citizens who think that the Internet is “very censored or censored” tend to be opposition party voters and the politically un-affiliated. Sixty-nine percent of CHP, 62% of HDP, 54% of MHP voters and 47% of those who do not support a political party think that the “supply of Internet freedoms” is low (Figure 13). In contrast, 47% of AKP supporters think the Internet is “very free or free” while 29% think it is “very censored or censored.”

Another related finding concerns users’ feelings of fear and risk especially when it comes to online political expression. According to survey findings, 33% of all users admit that they “avoid certain websites, blogs, online conversations, etc. due to online monitoring by the government.” Thirty percent of users say they are “afraid to openly share with others online what they think about certain political topics.” Only 25% of AKP-affiliated users say they avoid certain sites, whereas this number is much higher among HDP-affiliated users (45%) who also happen to be the ones most concerned with sharing their political opinions online (39%) (For more details, see Table 7).

Despite the divergence of opinions on Internet freedom and censorship, there is one issue upon which the majority of citizens agree, and that is the censorship of pornographic or sexually explicit content. Approximately 70% of survey respondents (users and non-users) agree that such content should be censored (Table 5). We also see high levels of support for such censorship across party lines that range from 55% (HDP voters) to 78% (AKP voters) (Table 6).

Perhaps the most surprising finding of this survey for me is the low level of circumvention practices. As I noted above, we assume that users are savvy in bypassing social media bans, content filtering and blocking, etc, and we have this romantic notion of empowered users who readily change DNS settings or switch to VPN, rendering government restrictions ineffective. The survey findings tell us another story. The vast majority (84%) of Internet users in Turkey said they did not circumvent the Twitter or YouTube bans of 2014. Only 8% reported that they bypassed the bans “occasionally,” 4% “a fair amount,” and 5% “all the time” (Figure 19).

In terms of approval or disapproval of social media bans, 73% percent of Twitter users opposed the Twitter ban, and the disapproval rate for the YouTube ban among the users of the platform was 63%. In terms of political party affiliation, highest levels of disapproval came from CHP supporters (72% for Twitter and YouTube bans) followed by HDP supporters (61% for Twitter ban and 57% for YouTube ban). An interesting finding is the split among AKP supporters: While 29% of them approved the bans, 24% of them were in opposition (Table 8). Perhaps, in future studies, researchers can conduct interviews with AKP supporters to understand the factors behind their approval and disapproval of governmental restrictions.

There are certain methodological issues in the survey that can be improved or expanded upon, and researchers can probe some issues in more detail in the future. For example, surveillance and privacy in relation to online freedoms is one area where more data is warranted. In light of the new Internet Law and the Intelligence Law (both passed in 2014), I suggest future research focus on the government’s increasing surveillance powers, and implications on users’ online behaviors, specifically political expression. The survey has some preliminary findings about online privacy concerns and perceived risks, but future studies can probe users’ online experiences by asking where they access the internet (at work, school, home, cyber cafe?), if and what filters they use or are required to use, and what they think about the role of ISPs and social media companies in enhancing or diminishing online freedoms. For example, are users concerned about their institutions (high school, university) or employers collecting and sharing information with the government? What do they think about the role of ISPs and social media companies in data collection and retention given the fact the new Internet Law obliges ISPs to collect data on users’ activities for up to two years, and provide authorities with this data on demand? What do they think about the fact that the Turkish intelligence agency can request and obtain citizens’ personal data from any public or private institution (banks, schools, hospitals, ISPs) without a court order? What are their opinions on the ever-increasing number of users arrested for their social media posts? And of course, it is vital to learn how respondents’ opinions on these issues differ across demographic background and political party affiliation.

In conclusion, I want to note that the OSU/Koc/Annenberg survey is a very important step towards gaining a more refined understanding of Internet users in Turkey and their perceptions of online restrictions. Among the most significant findings of the survey are the lack of a universal demand for Internet freedoms among citizens of Turkey and the high levels of support for censorship of certain types of content.  Interestingly, these findings resonate with those of Annenberg’s earlier survey on Russian public demand for Internet freedoms. For example, in both Turkey and Russia, a sizable number of citizens think that foreign countries are using the Internet against them (40-43% and 42%, respectively). The percentage of those who think the Internet threatens political stability in their respective countries is also comparable (30-35% in Turkey, 24% in Russia). There are similarities regarding attitudes towards online censorship as well. In both countries, for example, the majority of respondents approve of censorship of online pornographic content (59% in Russia, 70% in Turkey). I think a significant takeaway from these surveys is the need to revisit our assumptions regarding public demands for Internet freedoms and to understand the underlying socio-cultural, political, and historic dynamics beneath (high) approval rates for online censorship.


About the Author

Bilge Yesil is an Associate Professor of Media Culture at College of Staten Island, City University of New York. She is the author of Video Surveillance: Power and Privacy in Everyday Life (2009) and the upcoming The Turkish Model? Media, Democracy and the Neoliberal Islamist State. She writes about internet regulation, surveillance, censorship and mediated activism in Turkey.

  1. Türkei: Studie untersucht Wahrnehmung über Freiheiten im Internet | netzpolitik.org

    […] Yesil, Dozentin an der City University of New York, hat die wichtigsten Erkenntnisse der Studie erläutert. Die Ergebnisse machten vor allem deutlich, dass das freie Internet ein stark […]

  2. An examination of Turkey’s Internet users by media expert Dr. Bilge Yesil | Delta Phi Epsilon – Epsilon Chapter

    […] media expert Dr. Bilge Yesil examines in a blog post what a new report, “Benchmarking Demand: Turkey’s Contested Internet” means  for […]

Leave a Reply