The new Middle East: Jewish-Israeli exclusion of Palestinians in Facebook advertising
First Monday

The new Middle East: Jewish-Israeli exclusion of Palestinians in Facebook advertising by Yifat Mor and Ifat Maoz



Abstract
Previous studies have pointed to practices of disconnectivity and disengagement on Facebook, both generally (Stroud, 2010; Sunstein, 2009) and in the specific context of the Israeli society as a divided society in conflict (John and Dvir-Gvirsman, 2015; John and Gal, 2017; Gal, 2019). Our study aims to expand on these previous findings by examining practices of disconnectivity and disengagement in a different setting: among a professional group of Jewish Israeli advertisers who discuss their attempts to exclude Arab-named profiles from responding to their Facebook campaigns. The analysis focuses on a major Facebook group of social media marketing experts in Israel. This group is an open group consisting of 15,789 members, most of them Jewish-Israelis. We examined posts that were published in this group between 2013–2017 and dealt with Israeli-Arab audiences. Our findings emphasize the moral and ethical aspects of practices of disengagement and disconnectivity that — in the case studied here — help normalize and justify exclusion.

Contents

Introduction
Theoretical background
Methods
Findings
Discussion and conclusions

 


 

Introduction

This study aims to extend the discussion on social media practices in divided societies embedded in a protracted, asymmetrical, intractable conflict.

Previous studies have pointed to practices of disconnectivity and disengagement on Facebook, both generally (Stroud, 2010; Sunstein, 2009; Rymarczuk and Derksen, 2014) and in the specific context of the Israeli society as a divided society in conflict (John and Dvir-Gvirsman, 2015; John and Gal, 2017; Gal, 2019), as well as among Palestinians as a marginalized ethno-national group embedded in asymmetrical conflict (de Vries, et al., 2015; de Vries, et al., 2017).

Our study aims to extend these previous findings by examining practices of disconnectivity and disengagement among a professional group of Jewish Israeli advertisers who attempt to exclude Arab-named profiles from responding to their Facebook campaigns. While previous research focused on affordances and tools of Facebook’s user interface which allow users to curate their own newsfeed by hiding, unfollowing or unfriending specific users, pages or content (Stroud, 2010; Sunstein, 2009; Rymarczuk and Derksen, 2014; John and Dvir-Gvirsman, 2015), here we focus on Facebook’s advertising system that allows advertisers to influence other users’ newsfeed in far greater numbers. Targeted marketing is based on the idea that advertisers can include or exclude audiences from their campaigns based on demographic traits, interests and behaviors in order to achieve high campaign efficiency. However, advertisers can employ the system’s affordances to achieve different goals and to exclude certain users from exposure to content or opportunities (Speicher, et al., 2018).

Our findings emphasize the moral and ethical aspects of such practices of disengagement and disconnectivity, that in the present case study help normalize and justify exclusion. In the following sections we present the theoretical background of our study.

 

++++++++++

Theoretical background

Disconnectivity, disengagement and boundary work on social media

The Internet is a fertile ground for varied opinions and voices. Yet while every person, community or party can potentially express their thoughts and beliefs online, being heard online is a different story (Magalhães, 2018); publishers and content creators have far less control when it comes to online distribution for varied reasons. One of the main characteristics of Internet use is ‘selective exposure’ — the ability to choose the sort of content a user is exposed to, combined with users’ tendency to prefer content that matches their political predispositions (Stroud, 2010). Several studies have shown that this phenomenon, also known as ‘echo chambers’ or the ‘filter bubble effect’, is corelated with polarization of political opinions and decrease in civic deliberation and civil discourse (Mutz, 2006; Sunstein, 2009; Stroud, 2010; Cirucci, 2017).

Early studies on social media sites suggested that these platforms can enable users to encounter more diverse views and opinions through their social networks, thus leading to increased moderation in political opinions (Bimber, 2004; Papacharissi, 2002). Yet recent studies pointed at various factors that take part in determining whether or not a user will be exposed to content published by his or her social network and revealed users’ methods and tactics to enforce selective exposure in their own newsfeeds. First, the user can adjust his or her social network through the act of unfriending or unfollowing certain users or content (John and Dvir-Gvirsman, 2015; John and Gal, 2017) or by liking pages that present positions and views similar to those he or she holds in the first place. According to John and Dvir-Gvirsman (2015), unfriending is a practice enabling Facebook users to shape their political surroundings by muting certain opinions and/or the people stating these opinions. Second, the user is never exposed to everything that is posted by people and pages in his or her network — the platform itself is involved in determining which posts the user will see in the newsfeed and which posts will be hidden. In order to construct the newsfeed of each user in each moment, the platform uses an algorithm that makes decisions based on factors such as explicit or implicit norms, cultural presumptions or economic aims (Gillespie, 2015). The algorithm serves the platform’s interest in designing the social network to be more appealing to the user and assuring that he or she will want to spend more time on the platform (Gillespie, 2015; Magalhães, 2018). It is therefore claimed frequently that the algorithm causes audience fragmentation, amplifies the tyranny of the majority and promotes a populist media experience that is profitable and normative (Harper, 2017). Another important factor in determining which content will appear on each user’s feed is Facebook’s advertising system. Advertisers and publishers target the audiences relevant to their ads and choose which users will be potentially exposed to their content. Previous research has pointed to the moral and societal dilemmas concerning this ability of digital advertising platforms to support exclusion (Light, 2014; Light and Cassidy, 2014; Speicher, et al., 2018; Pallitto, 2018).

The context of the Israeli-Palestinian conflict

The Israeli society is embedded in the continuous, protracted and intractable conflict between Israeli Jews and Palestinians. As an intractable conflict, the Israeli-Palestinian conflict plays a major role in shaping the ethos, viewpoints, collective traumas and socio-cultural representations in the societies involved in it (Bar-Tal, 2013, 2007; Bar-Tal, et al., 2009; Morag, 2018, 2012, 2008). According to Bar-Tal (2007; 2000), ethnopolitical conflicts are commonly accompanied by psychological phenomena such as mutual prejudice, delegitimization and dehumanization (Bar-Tal, et al., 2012). Maoz (2011, 2004) defines two main sociopolitical characteristics of Jewish-Palestinian interaction in the context of the conflict. First: aggression, confrontation and competition alongside coexistence and cooperation. Second: inequality in access to and control of cultural, social and financial resources to which Israeli-Jews have greater access to and control of. This asymmetry is also manifested in Jewish-Palestinian computer–mediated interactions (Ellis and Maoz, 2007; Mor, et al., 2016).

Facebook in Israel serves as a popular platform for public interactions and discussions, both within the group of Israeli-Jews or Israeli-Arabs and between the two groups. In a previous study which examined Facebook as an arena for dialogue between groups in conflict in Israel, we found that when Israeli-Jews were exposed to moderate and peace-seeking Palestinian voices on Facebook, it facilitated a positive dialogue between members of the two groups on the platform. Yet, when Israeli-Jews were exposed to the Palestinian pain and suffering, it led to a predominantly negative intergroup exchange (Mor, et al., 2016).

Interestingly, psychological patterns that characterize intergroup conflict are also reflected in polarized discussions on social media platforms within the group of Israeli Jews in Israel (Harel, 2018). In a study examining posts and comments in a highly popular Facebook page associated with a rightist agenda, Harel (2018) pointed out that a core issue in the discussions revolved around the drawing of boundaries of the Israeli-Jewish in-group and the placement of the “leftist” group of Israeli-Jews inside or outside these boundaries. Boundary work on Facebook within the Israeli-Jewish in-group works in both directions: Gal (2019) examined the use of humor and irony as participatory boundary work on Facebook in left-wing mockery of an extreme right-wing group. She suggests that the combination of the medium — Facebook, a decontextualized digital environment and the key — ironic humor that is characterized in polysemy — foster misinterpretation that translates humorous interactions into tools of segregation, empowering one group and marginalizing the other one.

Political segregation and separation take place not only on the discursive level, but also through actions enabled by Facebook’s platform like unfriending or unfollowing; John and Dvir-Gvirsman found that during the Israel-Gaza War of 2014, political expression and discussion on Facebook in Israel often lead to heated discussions that in some cases ended with political unfriending — termination of weak social ties over political disagreement within the group of Israeli-Jews (John and Dvir-Gvirsman, 2015). Furthermore, during the Israel-Gaza War, acts of disconnectivity that were initiated on Facebook resonated outside of Facebook to other aspects of life. According to Schwarz and Shani (2016), for some Israeli-Arabs who worked in Israeli-Jewish companies, the cost of expressing their support for Palestinian citizens during the War was losing their jobs. In other cases, some Israeli-Jewish groups called to boycott Israeli-Arab businesses.

Against this background, this study aims to expand the understanding regarding Facebook uses and behaviors among Israeli-Jewish users in the context of the Israeli-Palestinian conflict by examining commercial use of Facebook, via Facebook’s advertising system. Unlike using one’s own Facebook profile to post content and gain exposure (or restrict it) through Facebook’s advertising system, advertisers have far better control in determining who will be exposed to their content, and who will be excluded. Through the analysis of discussions in professional Israeli-Jewish advertisers’ groups we aim to explore the perceptions and opinions of Israeli-Jewish advertisers on Facebook’s advertising system as well as the targeting practices they have developed in the context of the Israeli-Palestinian conflict. Specifically, we aim to examine the discourse around the phenomenon of receiving unintended engagement from profiles with Arabic names on Jewish-Israeli Facebook advertising campaigns and the practices suggested and used to deal with this phenomenon.

Online advertising, retailers’ surveillance and users’ discrimination

Online advertising, and specifically social network advertising, allows marketers to target their audience meticulously. On Facebook for instance, advertisers can target potential consumers by combining information the users have shared directly on the platform (such as name, age, etc.) with information gathered from users’ activities on the platform (such as which pages they liked); and information purchased from third-party data providers (Book and Bronk, 2016; Xia, et al., 2016). The accumulation of this reach data may include the users’ educational records, current or past employment data or even details about significant life events like marriage (Xia, et al., 2016). The implications of advertisers’ use of big data and of their collecting and analyzing information on users’ behavior is receiving increased critical attention. Studies on the use of big data often point to its lack of accuracy: the quantitative data collected on users contains merely traces of digital actions and cannot explain how or why these actions were taken and thus fail to deal with the complexity and context of human behaviors (Gillespie and Seavers, 2016; Markham, et al., 2018). Moreover, there is the issue of privacy. While some of the data collected is provided by users, other data is collected without the knowledge or consent of users (for example, their current location that is retrieved from the GPS component on their mobile phone). Collecting this information allows advertisers to micro-target potential customers based on the accumulation of detailed personal profiles (Auerbach, 2013). Another problem revolves around information analysis; it is often biased due to pre-existing values and perceptions that are structured into the algorithm by individuals or companies designing programs or processes. Even the simple act of defining categories for an algorithm can cause systematic bias in analysis (Shorey and Howard, 2016; Cirucci, 2017; Markham, et al., 2018). And while the data that are being gathered may assist advertisers in approaching relevant audiences, it may also be used to prevent knowledge of and access to desirable goods as well as the obtainment of social status from certain individuals or groups (Pallitto, 2018; Wilson and De Paoli, 2019).

Turow, et al. (2015) discuss the new forms of social discrimination that can be caused by transformations in digital retailing: merchants may change pricing or make certain products unavailable to some clients based on the information they hold on them and presumptions they make regarding the information they have collected. The use of this information establishes the ‘quantified individual’, in which a person is judged solely by the measurable traces of his or her digital actions. As a result, communities or social segments may be excluded or discriminated mainly on the basis of their presumed economic status. For example — if the retailer believes that a potential client will be willing to pay more for a certain product based on his or her previous behavior or assumed income, he might offer them a certain product for a higher price. While digital advertising platforms are designed to serve the needs of advertisers, the question about their accountability regarding unethical use of their affordances by ignorant or malicious advertisers remains unanswered (Turow, et al., 2015; Hemsley, et al., 2018). As Papacharissi (2018) noted, technology is generated by humans and when designing our future, it is important that those who create new technologies remember to be humane.

While most research regarding the evolving landscape of digital advertising has focused on the U.S. context, this study examines practices of disconnectivity and disengagement in the context of the asymmetrical ethnonational conflict between Israelis and Palestinians — among a professional group of Jewish Israeli advertisers who discuss their attempts to exclude Arab-named profiles from responding to their Facebook campaigns.

Studying Israeli-Jewish advertisers’ opinions, perceptions and techniques of ethnicity-based exclusion on their Facebook campaigns and their discourse regarding the phenomenon of receiving unintended engagement from profiles with Arabic names on their campaigns can expand our understanding of digital marketing in the context of protracted ethno-national conflict and social divide. The Facebook advertising system is a global service, and the market of digital advertising and specifically social network advertising is rapidly growing worldwide. Therefore, the findings of this research regarding targeting practices on the Facebook advertising system in the context of ethnic-national conflict and social divide may be relevant to other regions and contexts in the world. Following the Oslo agreements in 1993, Shimon Peres coined the term “The New Middle East” to describe a strategic plan to enhance cooperation between Arab and Jews in the Middle East in the realms of diplomacy, economy, and dealing with complex challenges such as environment, refugees or regional security. Nowadays, it seems that the narrative of cooperation is being replaced with a new narrative of separation, in which digital affordances are being used in order to build higher, virtual walls.

 

++++++++++

Methods

Research corpus

In contrast to Facebook pages in which the page administrators create most of the content, Facebook groups operate as open arenas for discussions in which the group administrators act mostly as moderators. Our analysis focuses on an open Facebook group for advertisers in Israel, in which discussions are conducted in Hebrew and most of the participants are Israeli-Jews. We examined discussions that were conducted within the group between August 2013 and February 2017 which dealt with the phenomenon of receiving unintended engagement from profiles with Arab names on advertising campaigns that were designated for Israeli-Jewish audiences in Israel.

Sampling and data collection

The analysis focuses on an open Facebook group, “Internet Advertising Professionals,” that was founded by an Israeli marketer in order to provide an arena for the exchange of professional knowledge between group members on issues such as new marketing-related features in digital platforms; marketing methods; strategies and tips; instructions and explanations on best uses and practices of Facebook and Google marketing tools.

In order to allocate and retrieve the relevant group discussions, we used the search tool that enables searching and filtering content so that posts and comments that contain a specific key word are shown. We conducted searches for keywords such as: “Arab”, “Arabs”, “Arabic” and “Cousins,” which is a phrase Israeli-Jews often use to describe Arabs.

This search yielded a data set containing 52 posts and 797 comments following these posts that were published between August 2013 and February 2017 and dealt with Israeli-Arab audiences. All posts and comments were copied to a document for further analysis.

Research site

The group studied here was founded on 25 July 2011 by an Israeli-Jewish marketer and consists of 15,789 members. The group description in the group’s &llsquo;about’ tab portrays it as an open arena for advertisers’ discussions regarding advertising platforms such as Facebook, Google and LinkedIn.

The groups rules, as defined by the admin in the group’s ‘about’ tab, state that: 1) members are not allowed to post any job searches or opportunities; 2) Participants are requested to be kind, helpful and respectful; 3) Group members are not allowed to ask for recommendations regarding digital advertising professional study programs; 4) Participants should ask for the admin’s permission to post any content that is not directly related to professional questions regarding the advertising platforms; 5) No spam and no self-marketing; and 6) For questions regarding digital advertising courses and schools, participants are referred to another Facebook group that focuses on these topics. The group is highly active, with approximately 500 new posts per month.

Analysis

Our analysis is inspired by the grounded theory approach (Glaser and Strauss, 1967), which emphasizes the construction of theories and concepts based on data that was gathered in the research process (Strauss and Corbin, 1998). In line with this paradigm (Berg, 2004; Strauss and Corbin, 1998), we conducted a horizontal reading of the Facebook groups’ posts in order to identify relevant main themes. The units of analysis were the posts and comment threads that followed them. Several more readings led us to narrow down the number of themes identified in the first stage by merging similar categories and focusing on the ones that were found to be the most prevalent within our sample as well as relevant to the focus of our research. Finally, the posts — together with the comments and replies that followed them — were categorized according to the themes. Four major themes were derived from the analysis and are presented in the next section.

 

++++++++++

Findings

We present here four major themes that have emerged from our analysis of posts as well as from comment threads following these posts.

The first three themes portray a discursive process of justifying the attempts to exclude profiles with Arab names. This is done through focusing on technical aspects, explanations and practices and through avoiding addressing moral considerations or the socio-political context of Jewish-Arab relations in Israel. The fourth theme presents an alternative and more inclusive approach that attempts to bring back into consideration moral aspects and a socio-political context.

1. Discussing the phenomenon of unintended engagement as ‘a problem’ while attempting to avoid appearing as racist

Advertisers create campaigns on Facebook with the goal of receiving engagement (likes, comments and shares), clicks to their Web sites and purchases from potential customers. Yet, the posts and comments analyzed here did not tend to refer to profiles with Arab names as potential customers. Rather, receiving engagement from such profiles on advertising campaigns is markedly portrayed in the data analyzed as ‘a problem’ which elicits the seeking of professional advice in order to understand why this phenomenon occurs and how can it be prevented.

Interestingly, while bringing up the topic of receiving unintended engagement from profiles with Arab names and asking for advice on how to exclude these users from their advertising campaigns based on their ethnicity, Israeli-Jewish advertisers employed different tactics in order to avoid being seen as racists.

One such practice used by participants was declaring that personally, they don’t have anything against Arabs, while describing the motivation to exclude as coming from their client:

“A Facebook question (and forgive me for lack of Euphemism)
My client is very unhappy about receiving a lot of likes from Arab users in a paid campaign. Even though they live nearby and could serve as potential clients. I know it sounds bad. But this is the client ... He sat down and deleted all the unwanted likes he received ... The question is what do you suggest doing? And how can you avoid unintended likes in advance???”

(Posted 12 March 2015 | 7 comments)

By mentioning that they are not against Arabs, and that their clients asked them to find a way to avoid unintended engagement on paid campaigns, the participants attempt to distance themselves from the motivation to exclude. By doing so, they attempt to avoid being seen as racists.

Another practice used was employing a statistical perspective describing ‘the problem’ in terms of a lack of variation in the campaign’s audience:

“Did someone find out — how to advertise on Facebook, and not receive every Arab in Israel as a new like on your page? Not that I have anything against them, but a little statistical variation would be nice 😊”
(Posted 6 October 2013 | 14 ‘Likes’, 73 comments)

Another post attempts to justify the question by focusing on profit while again noting that the writer doesn’t have anything against Arabs, but ...:

“Hi friends, a little question, what else can I do about my Facebook advertising, aiming mostly to the central district of the country? I get a lot of Arabs with Arab names (I don’t have anything against Arabs) they just don’t buy anything ...”
(Posted 30 June 2015 | 1 ‘Like’, 8 comments)

The quotes above indicate that Israeli-Jewish advertisers who participated in the discussions analyzed and employed different conversational tactics in order to minimize the risk of being perceived as racists, while discussing the phenomenon of receiving unintended engagement from profiles with Arab names on their campaigns and presenting it as ‘a problem’ that they are attempting to address.

2. ‘System inadequacy’: Discussing the phenomena and its suggested explanations in technical terms

The discussions in the Israeli-Jewish advertising group we explored also contained varied explanations and speculations as to why this phenomenon of unexpected engagement from Arab-named profiles occurs. In line with the findings presented above, these explanations tended to focus on technical terms and on the inadequacy of the system in an attempt to justify presenting this phenomenon as a problem that needs to be addressed.

2.1. Explanations associating the phenomenon with Facebook’s advertising system’s inaccuracy or inadequacy

Facebook’s advertising system provides advertisers a broad, yet limited set of targeting tools and options. While different societies around the world hold different norms and expectations regarding targeted marketing and information sharing on social media (Stanger, et al., 2017; Jung, et al., 2016), the system is uniform. Moreover, some of the information that Facebook offers advertisers relies on users’ self-reported data such as age, gender or workplace and thus is not highly accurate and could lead to unexpected targeting results. Finally, Facebook’s advertising system is constantly changing and evolving, which sometimes causes temporary technical bugs or changes in the performances of different functions in the platform.

Interestingly, participants in the discussions analyzed suggested that the origin of the phenomenon of unintended engagement is related to a technical malfunction or to the system’s inadequacy. For example, in a discussion from October 2013, a participant stated that geographical information that Facebook is collecting is inaccurate:

“And you think Facebook can distinguish between Jenin [a Palestinian city, Y.M.] and Natanya? [an Israeli, mostly Jewish city, Y.M.] I mean that its distinguishing abilities are accurate enough? ... To say that Facebook’s location indicators are exact is still a little pretentious ...”
(Posted 6 October 2013)

More specifically, advertisers who took part in the examined discussions focused on the geographical targeting function on Facebook’s advertising system, in which the default setting for including a location is to add a radius of 25 miles around it. In the quotes below advertisers refer to this default feature of the system, given that in Israel, Jews and Arabs live in much closer proximity to each other.

The system’s geographical targeting function was very clearly addressed in a discussion from May 2014:

“As a default the system adds a radius of 25 miles to the location you specified, make sure to cancel it ...”
“I did the thing with the 25 miles — that was probably the problem 😊 thanks everybody!”
(Posted 31 December 2014)

In a more recent discussion, a commenter further explained:

“Most of the targeting doesn’t work because of diverse cities such as Tel Aviv, Ramla, Be’er Sheva, Lod, Nazareth, Jerusalem, Acre, Haifa and thirty percent of the population lives in these cities.”
(Posted 1 February 2017)

Generally, these quotes associate the phenomenon of receiving unexpected engagement from profiles with Arab names with errors in Facebook’s advertising system or with Facebook’s inability to address the unique characteristics of the Israeli market. These quotes frame the phenomenon of receiving unexpected engagement as a technical ‘problem’ that should be solved using technical tools and practices.

2.2. Explanations associating the phenomenon with frauds on Facebook

Another common claim regarding the phenomenon attempts to avoid viewing unexpected likes from profiles with Arab names as authentic engagement — by claiming that these profiles are fake. Some advertisers claimed that the phenomenon is part of a fraud in which companies use “like farms” to sell fake “likes” illegally to pages in order to increase their number of followers cheaply. “Like farms” are a network of fake Facebook profiles, operated by humans or by bots for the purpose of massively ‘liking’ pages that pay illegally for this service. In order to avoid the system’s suspicion and behave like ‘normal’ users, these bots are directed to ‘like’ other varied content — so as to imitate the choices and behaviors of regular users. In the framework of this explanation, the unexpected ‘likes’ by Arab-named users were presented as likes intended to give credibility to fake ‘like farm’ bots. For example:

“People (mainly Arabs) whose job is to like pages in exchange for money, usually like many pages for which they don’t get paid as well in order to retain an impression of regular activity [liking pages randomly rather than only liking pages that pay for it, y.m] so that Facebook won’t block them.”
(Posted 10 March 2014)

In another discussion, an individual claimed that there was a need to distinguish between true and fake profiles of Israeli-Arabs:

“Friends, it’s not a matter of racism ... the high-quality Arab audience is buying, and buying well, I know from personal experience ... but they also complain about these low-quality likes of the fake Arab profiles and there are tons of those on Facebook ... And they are the subject of this whole discussion ... And if they were British or French or Russian we would raise the problem just the same ...”
(Posted 6 October 2013 | 1 “Like”)

This participant states that due to the existence of fake profiles with Arab names, discussing unintended engagement as a problem should not be considered as racist. Generally, the quotes presented in this theme suggest that the engagement received is not just a technical incompatibility but rather the outcome of illegal activity on Facebook, conducted either by people or bots in order to retain an impression of regular activity.

3. Normalizing exclusion and discussing technical practices

Facebook’s advertising system is designed to enable marketers to define the relevant audience for their campaigns. While the system allows the filtering of target audiences by using various categories such as location, interests, demographics or behaviors, it does not allow the exclusion of users based on their ethnicity or language. And so, in order to bypass the system’s restrictions, participants in the discussions we examined came up with various exclusion techniques. For example, most Israeli-Arabs reside in specific cities and villages, and thus the exclusion of these locations can minimize the chances that the campaign will reach Israeli-Arabs. Thus, some of the participants agreed to share a list they made of all the Arab cities and villages in Israel so that others can easily copy and use it:

— “There’s an attached file of Israeli-Arab cities and villages.”

— “What she was trying to say to you is that you should copy the list of Israeli-Arab cities and villages and set them as excluded.”

— “Or, you can choose main cities with no radius around them, but not Haifa or Tel Aviv or Jerusalem — there are Arabs there too ... Or exclude all the Israeli-Arab cities and villages.”

— “You can generate an audience in Facebook’s Power-editor by excluding them one by one once and then keep using this audience in other campaigns.”

— “One by one? I copy the entire list and paste it with one click.”
(Posted 30 June 2015)

On another comment thread, one user suggested:

“Exclude their interests such as Al Jazeera, Arabic language ... choose Hebrew and English in Facebook’s interface language, and as mentioned above, if you target locations with radius, you must exclude their cities and villages. And also, target only women, Arab women ‘Like’ less than Arab men.”
(Posted 16 February 2017)

These remarks exemplify the various exclusion methods applied by Jewish-Israeli marketers in order to bypass the inability to exclude audiences based on their ethnicity directly via Facebook’s advertising system. These techniques can be interpreted as ‘platform-hacking’ — using the platform’s features and affordances discordantly with the uses intended for them by the platform’s designers. It seems that the framing of the phenomenon as a technical problem enables participants in these discussions to discuss and to share exclusion methods with other group members while avoiding the moral and social consequences of using such techniques and minimizing the risks of being seen as racists.

4. Is it unprofessional to be moral? Attempting to bring moral considerations and the socio-political context back to the discussion

The dominant approach conveyed by advertisers in the discussions explored here, treated the phenomenon of receiving unexpected engagement from profiles with Arab names as a technical problem. Accordingly, participants commonly addressed the phenomenon by using technical tools and tactics that enable excluding such profiles from their campaigns. Yet, other participants suggested an alternative point of view that does take into account the socio-political and moral aspects of Jewish-Arab relations in Israel.

For example, one commenter pointed to the problematic nature of exclusion and was portrayed by another commenter as talking in an unprofessional way:

“— Too bad marketers and advertisers think that 20 percent of the population is irrelevant to them.

— And too bad some people talk in an unprofessional way ...

— No reason to be offended, just like some campaigns are targeted only at men — that doesn’t make the advertiser chauvinist. And I have nothing against them except that they are not potential buyers.”
(Posted 6 October 2013)

Similarly, another advertiser wrote a post complaining about his client:

“My client: ‘let’s redefine what sort of customers I want you to get me: no Russians, no Arabs, no Ethiopians, only people from the center of Israel who earn such and such money a month’ — Me: ‘Are you looking for a customer or for someone to marry? What do you want me to write in the landing page? Russians/Arab/Ethiopians out? Even as a joke it sounds bad’... what’s up with people?”
(Posted 1 August 2013 | 8 “Likes”, 7 comments)

These remarks indicate that when the problematics of exclusion is brought into discussion, it is encountered by resistance and by justifications and rationalization of targeting only certain populations.

To summarize, these discussions that we examined in this section prevalently treated the phenomenon of receiving unintended engagement from profiles with Arab names as a technical problem that thus requires technical solutions. Yet some participants also treated the problematic aspects of exclusion.

The first theme exemplifies how advertisers frame the phenomenon as a problem, while attempting to avoid appearing as racists or referring to moral considerations and to the socio-political context of the relations between Jews and Arabs in Israel.

The second theme demonstrates how advertisers focus on the technical aspects of the phenomenon by explaining it in terms of ‘system inadequacy’. By doing so, they ignite the discursive process of justifying attempts to exclude profiles with Arab names and normalize the discussion around practices of exclusion on Facebook’s advertising campaigns as presented in the third theme.

Finally, the fourth theme presents an alternative point of view on the phenomenon that encourages inclusivity and attempts to bring the moral aspects and the socio-political context back into consideration.

 

++++++++++

Discussion and conclusions

This study extends the discussion on social media practices in divided societies embedded in a protracted, intractable conflict in several ways.

Disengagement and exclusion

Previous studies have pointed to practices of disconnectivity, disengagement and participatory boundary work on Facebook, both generally (Light and Cassidy, 2014), and in the specific context of the Israeli society as a divided society in conflict (John and Dvir-Gvirsman, 2015; John and Gal, 2017; John and Gal, 2018; Gal, 2019), as well as among Palestinians as a marginalized group embedded in asymmetrical conflict (de Vries, et al., 2015; de Vries, et al., 2017).

Our findings expand on these previous findings by showing that disconnectivity, disengagement and participatory boundary work are also practiced among professional groups of Jewish Israeli advertisers who attempt to exclude Arab-named profiles from responding to their Facebook campaigns. While previous research focused on discursive practices of boundary work on social media such as humor and irony (Gal, 2019), or technical practices on the private use level such as unfriending, unfollowing or blocking specific users (John and Gal, 2018), our research extends these practices to the macro level, using technical tools to enforce boundary work in far greater numbers. Our findings thus emphasize the moral and ethical aspects of disengagement and disconnectivity practices that in the case studied here clearly take the form of practices for normalizing and justifying exclusion.

Exclusion justified

Previous studies of social media in protracted, asymmetrical conflict have demonstrated the use of Facebook for the delegitimization of outgroups (Mor, et al., 2015; Mor, et al., 2016; de Vries, et al., 2015; de Vries, et al., 2017; John and Gal, 2018; Gal, 2019; Harel, 2018) and specifically of the delegitimization of Palestinians on Facebook by Israeli-Jews (Mor, et al., 2015; Schwarz and Shani, 2016; Harel, 2018). This study focuses on a seemingly silent but actually meaningful form of delegitimization that is achieved without using terms that appear as derogatory and while employing justifications to normalize these practices.

This normalization and justification of exclusion on technical and professional grounds may be an indication of the ways in which more liberally oriented groups in Israel that perceive themselves as non-racist also practice exclusion and delegitimization that is manifested in attempts to disconnect, disengage and distance themselves while emphasizing non-racist attitudes (Maoz, et al., 2002).

Using Facebook to disengage, disconnect and to exclude

On a broader level, our study extends previous discussions on the use of Facebook affordances to disconnect and to exclude, and on the moral and societal dilemmas involved in the ability of this platform to support exclusion (Light, 2014; Light and Cassidy, 2014; Speicher, et al., 2018). Interestingly, it is the perceived ineffectiveness of the Facebook advertising system at exclusion, and the explanations and solutions suggested to address these limitations in the Israeli context, that clarify the extent to which this platform is expected to help exclude and distance certain populations when functioning optimally.

Previous studies on the potential for discrimination in Facebook advertising focused mainly on the U.S. context and explored exclusion techniques that allow unintentionally ignorant or malicious advertisers to exclude users from advertising campaigns based on their affinity to attributes that are highly correlated with certain ethnic groups, religions or sexual orientation (Speicher, et al., 2018). Our study reveals new techniques that are relevant to the Israeli context and leverages Facebook’s targeting methods as well as the data provided by Facebook to advertisers to exclude Israeli-Arab users from Israeli-Jewish campaigns. Moreover, our study reveals the discursive process that allows advertisers to discuss and share such techniques openly within what they perceive as the normative framework of a professional discussion on targeted advertising.

Advertising platforms innovate new efficient methods for targeting users for the benefits of advertisers yet seem to overlook the potential outcomes of unethical usage of these methods (Speicher, et al., 2018). Our findings point to the crucial need for a professional ethical framework that would guide both advertisers and platforms in the rapidly evolving landscape of online advertising.

Limitations and directions for future research

This study focused on discussions that appear prominently on a Facebook group of professional Jewish Israeli advertisers. In order to receive a fuller and more nuanced picture, it is suggested that future studies also attempt to interview these advertisers and uncover the potentially complex relations between online and off-line attitudes and behaviors (de Vries and Maoz, 2013; de Vries, et al., 2015; de Vries, et al., 2017) as well as their broader attitudes towards Jewish-Arab relations and contact in Israel (Maoz, 2011; Mor, et al., 2016).

Future studies should also attempt to gauge the extent to which the exclusion practices discussed are actually employed, as well as further identify the conditions and circumstances in which a more inclusive approach is adopted (Mor, et al., 2016) and the practices of inclusion involved in such cases.

Future research should also examine more closely, using several research methods including interviews and social-psychological experiments, the ways in which the characteristics of social media and advertising platforms can affect connectivity vs. disconnectivity and inclusion vs. exclusion between groups embedded in internal and external ethnopolitical dispute and conflict. End of article

 

About the authors

Dr. Yifat Mor is a postdoctoral fellow at the Swiss Center for Conflict Research, Management and Resolution, the Hebrew University of Jerusalem, Israel. Her research focuses on political expression and discussion on Facebook in Israel, as a society embedded in a protracted ethno-political conflict.
E-mail: yifat [dot] mor1 [at] mail [dot] huji [dot] ac [dot] il

Ifat Maoz is a Full Professor, Head of the Department of Communication and Journalism at the Hebrew University; Director of the Swiss Center and MA Program for Conflict Research, Management and Resolution and holds the Danny Arnold Chair in Communication.

Her current research program focuses on the dynamics of social media in intractable, asymmetric conflict and on public opinion in conflicted, and divided societies.
E-mail: msifat [at] gmail [dot] com

 

References

D. Auerbach, 2013. “You are what you click: On microtargeting,” Nation (13 February), at http://www.thenation.com/article/you-are-what-you-click-microtargeting/, accessed 10 August 2019.

D. Bar-Tal, 2013. Intractable conflicts: Socio-psychological foundations and dynamics. Cambridge: Cambridge University Press.

D. Bar-Tal, 2007. “Sociopsychological foundations of intractable conflicts,” American Behavioral Scientist, volume 50, number 11, pp. 1,430–1,453.
doi: https://doi.org/10.1177/0002764207302462, accessed 10 August 2019.

D. BarTal, 2000. “From intractable conflict through conflict resolution to reconciliation: Psychological analysis,” Political Psychology, volume 21, number 2, pp. 351–365.
doi: https://doi.org/10.1177/0002764207302462, accessed 10 August 2019.

D. Bar-Tal, Y. Rosen and R. Nets-Zehngut, 2009. “Peace education in societies involved in intractable conflicts: Goals, conditions, and directions,” In: G. Salomon and E. Cairns (editors). Handbook on peace education. New York: Psychology Press, pp. 21–43.

D. Bar-Tal, K. Sharvit, E. Halperin and A. Zafran, 2012. “Ethos of conflict: The concept and its measurement,” Peace and Conflict, volume 18, number 1, pp. 40–61.
doi: http://dx.doi.org/10.1037/a0026860, accessed 10 August 2019.

B.L. Berg, 2004. Qualitative research methods for the social sciences. Fifth edition. Boston Mass.: Pearson Education.

B. Bimber, 2004. “The Internet and political fragmentation,” paper presented at the Democracy in the 21st Century Conference (University of Illinois Urbana-Champaign).

T. Book and C. Bronk, 2016. “I see you, you see me: Mobile advertisements and privacy,” First Monday, volume 21, number 3, at https://firstmonday.org/article/view/6154/5215, accessed 10 August 2019.
doi: https://doi.org/10.5210/fm.v0i0.6154, accessed 10 August 2019.

A.M. Cirucci, 2017. “Normative interfaces: Affordances, gender, and race in Facebook,” Social Media + Society (28 June).
doi: https://doi.org/10.1177/2056305117717905, accessed 10 August 2019.

M. de Vries and I. Maoz, 2013. “Tracking for peace: Assessing the effectiveness of track two diplomacy in the IsraeliPalestinian conflict,” Dynamics of Asymmetric Conflict, volume 6, numbers 1–3, pp. 62–74.
doi: https://doi.org/10.1080/17467586.2013.861074, accessed 10 August 2019.

M. de Vries, A. Simry and I. Maoz, 2015. “Like a bridge over troubled water: Using Facebook to mobilize solidarity among East Jerusalem Palestinians during the 2014 war in Gaza,” International Journal of Communication, volume 9, at https://ijoc.org/index.php/ijoc/article/view/3581, accessed 10 August 2019.

M. de Vries, N. Kligler-Vilenchik, E. Alyan, M. Maoz and I. Maoz, 2017. “Digital contestation in protracted conflict: The online struggle over al-Aqsa Mosque,” Communication Review, volume 20, number 3, pp. 189–211.
doi: https://doi.org/10.1080/10714421.2017.1362814, accessed 10 August 2019.

D.G. Ellis and I. Maoz, 2007. “Online argument between Israeli Jews and Palestinians,” Human Communication Research, volume 33, number 3, pp. 291–309.
doi: https://doi.org/10.1111/j.1468-2958.2007.00300.x, accessed 10 August 2019.

N. Gal, 2019. “Ironic humor on social media as participatory boundary work,” New Media & Society, volume 21, number 3, pp. 729–749.
doi: https://doi.org/10.1177/1461444818805719, accessed 10 August 2019.

T. Gillespie, 2015. “Platforms intervene,” Social Media + Society (11 May).
doi: https://doi.org/10.1177/2056305115580479, accessed 10 August 2019.

T. Gillespie and N. Seaver, 2016. “Critical algorithm studies: A reading list,” Social Media Collective, at https://socialmediacollective.org/reading-lists/critical-algorithm-studies/, accessed 10 August 2019.

B.G. Glaser and A.L. Strauss, 1967. The discovery of grounded theory: Strategies for qualitative research. Chicago: Aldine.

T. Harel, 2018. “Psychological characteristics of a polarized political discourse in social media platform in Israel,” unpublished Master’s thesis, Hebrew University of Jerusalem.

T. Harper, 2017. “The big data public and its problems: Big data and the structural transformation of the public sphere,” New Media & Society, volume 19, number 9, pp. 1,424–1,439.
doi: https://doi.org/10.1177/1461444816642167, accessed 10 August 2019.

J. Hemsley, J. Jacobson, A. Gruzd and P. Mai, 2018. “Social media for social good or evil: An introduction,” Social Media + Society (19 July).
doi: https://doi.org/10.1177/2056305118786719, accessed 10 August 2019.

N.A. John and N. Gal, 2018. “‘He’s got his own sea’: Political Facebook unfriending in the personal public sphere,” International Journal of Communication, volume 12, pp. 2,971–2,988, and at https://ijoc.org/index.php/ijoc/article/view/8673/2410, accessed 10 August 2019.

N.A. John and N. Gal, 2017. “‘I can’t see you any more’: A phenomenology of political Facebook unfriending,” AoIR Selected Papers of Internet Research, at https://journals.uic.edu/ojs/index.php/spir/article/view/8456, accessed 10 August 2019.

N.A. John and S. Dvir-Gvirsman, 2015. “‘I don’t like you any more’: Facebook unfriending by Israelisduring the IsraelGaza Conflict of 2014,&dquo; Journal of Communication, volume 65, number 6, pp. 953–974.
doi: https://doi.org/10.1111/jcom.12188, accessed 10 August 2019.

J. Jung, S.W. Shim, H.S. Jin and H. Khang, 2016. “Factors affecting attitudes and behavioural intention towards social networking advertising: A case of Facebook users in South Korea,” International journal of Advertising, volume 35, number 2, pp. 248–265.
doi: https://doi.org/10.1080/02650487.2015.1014777, accessed 10 August 2019.

B. Light, 2014. Disconnecting with social networking sites. Basingstoke: Palgrave Macmillan.

B. Light and E. Cassidy, 2014. “Strategies for the suspension and prevention of connection: Rendering disconnection as socioeconomic lubricant with Facebook,” New Media & Society, volume 16, number 7, pp. 1,169–1,184.
doi: https://doi.org/10.1177/1461444814544002, accessed 10 August 2019.

J.C. Magalhãe, 2018. “Do algorithms shape character? Considering algorithmic ethical subjectivation,” Social Media + Society, (2 May).
doi: https://doi.org/10.1177/2056305118768301, accessed 10 August 2019.

I. Maoz, 2011. “Does contact work in protracted asymmetrical conflict? Appraising 20 years of reconciliation-aimed encounters between Israeli Jews and Palestinians,” Journal of Peace Research, volume 48, number 1, pp. 115–125.
doi: https://doi.org/10.1177/0022343310389506, accessed 10 August 2019.

I. Maoz, 2004. “Coexistence is in the eye of the beholder: Evaluating intergroup encounter interventions between Jews and Arabs in Israel,” Journal of Social Issues, volume 60, number 2, pp. 437–452.
doi: https://doi.org/10.1111/j.0022-4537.2004.00119.x, accessed 10 August 2019.

A.N. Markham, K. Tiidenberg and A. Herman, 2018. “Ethics as methods: Doing ethics in the era of big data research — Introduction,” Social Media + Society (19 July).
doi: https://doi.org/10.1177/2056305118784502, accessed 10 August 2019.

Y. Mor, Y. Ron and I. Maoz, 2016. “‘Likes’ for peace: Can Facebook promote dialogue in the Israeli-Palestinian conflict?” Media and Communication, volume 4, number 1, pp. 15–26.
doi: http://dx.doi.org/10.17645/mac.v4i1.298, accessed 10 August 2019.

Y. Mor, N. Kligler-Vilenchik and I. Maoz, 2015. “Political expression on Facebook in a context of conflict: Dilemmas and coping strategies of Jewish-Israeli youth,” Social Media + Society (8 October).
doi: https://doi.org/10.1177/2056305115606750, accessed 10 August 2019.

R. Morag, 2018. “Blood relations and non-consensual ethics: Israeli intifada documentaries,” Post Script, volume 36, numbers 2–3, pp. 75–85.

R. Morag, 2012. “Perpetrator trauma and current Israeli documentary cinema,” Camera Obscura, volume 27, number 80, pp. 93–133.
doi: https://doi.org/10.1215/02705346-1597222, accessed 10 August 2019.

R. Morag, 2008. “Chronic trauma, the sound of terror, and current Israeli cinema,” Framework>, volume 49, number 1, pp. 121–133.
doi: https://doi.org/10.1353/frm.0.0003, accessed 10 August 2019.

D.C. Mutz, 2006. Hearing the other side: Deliberative versus participatory democracy. Cambridge: Cambridge University Press.
doi: https://doi.org/10.1017/CBO9780511617201, accessed 10 August 2019.

R.M. Pallitto, 2018. “Irresistible bargains: Navigating the surveillance society,” First Monday, volume 23, number 2, at https://firstmonday.org/article/view/7954/6643, accessed 10 August 2019.
doi: https://doi.org/10.5210/fm.v23i2.7954, accessed 10 August 2019.

Z. Papacharissi (editor), 2018. A networked self and human augmentics, artificial intelligence, sentience. New York: Routledge.

Z. Papacharissi, 2002. “The virtual sphere: The Internet as a public sphere,” New Media & Society, volume 4, number 1, pp. 9–27.
doi: https://doi.org/10.1177/14614440222226244, accessed 10 August 2019.

R. Rymarczuk and M. Derksen, 2014. “Different spaces: Exploring Facebook as heterotopia,” First Monday, volume 19, number 6, at https://firstmonday.org/article/view/5006/4091, accessed 10 August 2019.
doi: https://doi.org/10.5210/fm.v19i6.5006, accessed 10 August 2019.

O. Schwarz and G. Shani, 2016. “Culture in mediated interaction: Political defriending on Facebook and the limits of networked individualism,” American Journal of Cultural Sociology, volume 4, number 3, pp. 385–421.
doi: https://doi.org/10.1057/s41290-016-0006-6, accessed 10 August 2019.

S. Shorey and P.N. Howard, 2016. “Automation, algorithms, and politics | Automation, big data and politics: A research review,” International Journal of Communication, volume 10, at https://ijoc.org/index.php/ijoc/article/view/6233, accessed 10 August 2019.

T. Speicher, M. Ali, G. Venkatadri, F.N. Ribeiro, G. Arvanitakis, F. Benevenuto, P. Gummadi, P. Loiseau and A. Mislove, 2018. “Potential for discrimination in online targeted advertising,” Proceedings of the First Conference on Fairness, Accountability and Transparency, pp. 5–19, at http://proceedings.mlr.press/v81/speicher18a.html, accessed 10 August 2019.

N. Stanger, N. Alnaghaimshi and E. Pearson, 2017. “How do Saudi youth engage with social media?” First Monday, volume 22, number 5, at https://firstmonday.org/article/view/7102/6101, accessed 10 August 2019.
doi: http://dx.doi.org/10.5210/fm.v22i15.7102, accessed 10 August 2019.

A.L. Strauss and J.M. Corbin, 1998. Basics of qualitative research: Techniques and procedures for developing grounded theory. Second edition. Thousand Oaks, Calif.: Sage.

N.J. Stroud, 2010. “Polarization and partisan selective exposure,” Journal of Communication, volume 60, number 3, pp. 556–576.
doi: https://doi.org/10.1111/j.1460-2466.2010.01497.x, accessed 10 August 2019.

C.R. Sunstein, 2009. Going to extremes: How like minds unite and divide. Oxford: Oxford University Press.

J. Turow, M. Hennessy and N. Draper, 2015. “The tradeoff fallacy: How marketers are misrepresenting American consumers and opening them up to exploitation,” Annenberg School for Communication, University of Pennsylvania, at https://www.asc.upenn.edu/news-events/publications/tradeoff-fallacy-how-marketers-are-misrepresenting-american-consumers-and, accessed 10 August 2019.

A. Wilson and S. De Paoli, 2019. “On the ethical and political agency of online reputation systems,” First Monday, volume 24, number 2, at https://firstmonday.org/article/view/9393/7730, accessed 10 August 2019.
doi: http://dx.doi.org/10.5210/fm.v24i2, accessed 10 August 2019.

C. Xia, S. Guha and S. Muthukrishnan, 2016. “Targeting algorithms for online social advertising markets,” ASONAM ’16: Proceedings of the 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, pp. 485–492.

 


Editorial history

Received 24 February 2019; revised 29 July 2019; accepted 8 August 2019.


Creative Commons License
This paper is licensed under a Creative Commons Attribution 4.0 International License.

‘The new Middle East’: Jewish-Israeli exclusion of Palestinians in Facebook advertising
by Yifat Mor and Ifat Maoz.
First Monday, Volume 24, Number 9 - 2 September 2019
https://journals.uic.edu/ojs/index.php/fm/article/view/9734/8067
doi: http://dx.doi.org/10.5210/fm.v24i9.9734





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.