#MeTwo: National Identity and Discrimination in Germany in Times of Migration

Analyzing the #MeTwo movement using Twitter data and R

2019-08-02 | Konstantin Gavras, Lisa Hehnke, Paul Meiners, Sandra Meneses & Juan Orduz

#MeTwo: National Identity and Discrimination in Germany in Times of Migration

Introduction

Shortly before the start of the Men’s Football World Championship in 2018, a debate on migration and the question of who a true German is erupted on the issue that Mesut Özil took a photo with the Turkish President Recep Tayyip Erdogan. The presumed reciprocal support of one of the two most famous Turkish men sparked a vivid debate on whether Mesut Özil, playing in the dress of the German Football national team, should distance himself from the policies and authoritarian tendencies displayed by the President of his father’s land. This debate quickly developed into a wider debate about who actually is German and whether it is possible that migrants and their children might even feel attached to two or even more heritages.

Based on this debate, social activist Ali Can developed a hashtag, which should be a crystallization and manifestation of the struggles, accusations and latent racism people not stemming from an autochton German heritage face in their everyday life. Inspired by the MeToo movement by women experiencing sexual harassment and violence from partners, superiors and even strangers, #MeTwo was born. Using this hashtag, everyone having experienced harassment and discrimination due to her migration background could unfold their story to make the general public aware of the problems people with migration background face in Germany. Based on this insight, reciprocal understanding of people with and without migration background should emerge, building a more sustainable and peaceful society.

To empirically test the implications and outcomes of this online movement, we at CorrelAid decided to scrape all N = 159114 tweets on #MeTwo from July 24 to August 03, 2018, and examine the content, scope, and temporal dynamics of this social media event, which now celebrates its first birthday. As such, we aim to answer the following questions:

  1. How did the #MeTwo movement emerge and how did it develop over the course of its lifetime?
  2. Which users have been involved in the online movement? Who were the most retweeted and favorited users? Do they tweet from personal accounts or from verified ones that are of public interest? How inclusive is the movement overall?
  3. How does the retweet network look like? Are there any key players that could potentially influence and shape the debate as the online movement continues? Did the network develop or change during the movement?
  4. Which content was shared and discussed? Which opinions and emotions are expressed in the tweets? Are the stories presented in the #MeTwo highlighting the staggering experiences people faced in the past, or do positive experiences and hopes for a better future dominate the discussion? Is there a connection to other prominent hashtags?

Development of an Online movement in Digital Environments

In an initial step, we simply plot the number of tweets and retweets associated with the #MeTwo Hashtag. As one can immediately see, this hashtag did trend extremely quickly and particularly strong. Already in the first hours of #MeTwo, there were more than 500 tweets per hour. On July 26 and 27, the online movement had its peak with more than 7000 tweets per hour only interrupted by the night hours in Germany. The retweets increase at the same pace, indicating that the online movement did spark wide interest within the Twitter community - the retweet count peaked at about 2300 retweets an hour on the first two days. On the following two days, there was still a reasonable amount of tweets and retweets, but with the online movement already losing its pace of the first two days. As for other online movements, MeTwo had a sharp peak in the beginning, which last for a few days and then - hopefully gets picked up by the offline society. After one week, we could not find any further substantial input for the MeTwo movement, bringing us to a preliminary conclusion that MeTwo seems to be comparable to other online movements, which spark quite fast, but decline from the internet after a short time period.

When seperating the tweets in original tweets and retweets, we see a ratio of 1:5, which is actually quite reasonable, given that many users do not get retweeted quite often. Thus, we are examining a hashtag, which spread considerably and which was thus able to spark a discussion about the reality of being German and about national identity in times, in which millions of people in Germany originate from foreign countries. The wide coverage of the hashtag allows us to empirically evaluate the content of the tweets as well as the networks emerging during the online debate on this topic. In a first step, however, we examine whether the debate did only occur on twitter or if it spread to other means of societal communication as well.

Timeline of MeTwo in- and outside of Twitter

While the first plots provided information about the total volume of tweets on #MeTwo, we now turn to the coverage of MeTwo in comparison with its appearance on Google Trends. This measure indicates how often certain search phrases are being used in comparison with its average search occurence. Thus, the peaks do not show absolute but relative values. In order to re-create the emergence and over-time coverage we do not only cover the search phrase of ‘MeTwo’ but also ‘Rassismus’ as well as the initial of the movement, ‘Mesut Özil’, we looked at these specific keywords.

In our first plot, we first compare the Google Trends index of ‘Mesut Özil’ with the Twitter activity on MeTwo. As we can see, Mesut Özil normally receives only small coverage in Google (perhaps mostly by hard-core football fans). However, after the outrage concerning his picture with the Turkish President Erdogan, Mesut Özil was googled exordinarily often. Yet, in Google this trend did only last for more of less two days after which Mesut Özil did not receive much of Google Trends attention any longer. As one can see, MeTwo sparked about one week after the incident together with a largeer political and societal debate about rassism, national identity and the criterions of being German. In order to check whether our asusmption about the essence of this debate holds true, we now turn to the coverage of the search phrase ‘Rassismus’ in Google Trends, since it is a common phrase associated with a public discourse on problems of national identity in times of migration.

As one can see, the search phrase ‘Rassimus’ trended at the exact time as ‘Mesut Özil’ and thus a few days before MeTwo started. Of course, this search phrase is much more common on average than ‘Mesut Özil’, thus we see a lot of variation in coverage over time. Interesting, however, is the fact that ‘Rassismus’ did not trend simultaneous with ‘MeTwo’ although one might have assumed that these two terms are highly associated.

Turning to the buzzword of MeTwo itself, we see that it experienced broader societal coverage (measured via Google Trends), but only during the Twitter debate on this topic and some weeks afterwards. As one can see, Google Trends and Twitter peaked on this search phrase on the exact same day, indicating that the online movement wasn’t restricted on Twitter only, but it also spread to a broader audience. As one can see, MeTwo was still relevant for the public until the end of August 2018, but faded afterwards. Summarizing, we can see that MeTwo was not only restricted to Twitter, but became publicly important. Furthermore, it is obvious that the debate on Mesut Özil sparked the online movement, which was heavily associated with the more general problem of rassism whilst discussing this issue.

Users and Participants in the MeTwo debate

Most active users (number of tweets, retweets, and favorites)

After these aggregated insights into the emergence and development of #MeTwo, we now turn to more fine-grained analyses, including both the accounts involved as well as the content articulated in their tweets. Here, we first examine the accounts which used the hashtag the most in their tweets and retweets. All of these accounts have impressive numbers of tweets and retweets on #MeTwo with more than 200, ranging up to 350 tweets, which is suprising given that the online movement did only last for about a week. As can be seen in the following plot, among the most active users are rather infamous accounts. Actually, the most active users with regard to the #MeTwo hashtag seem to be personal accounts with only very modest numbers of followers. However, all of them are extremely active twitter users in general.

Favorites are one of the most important currencies on Twitter, enabling us to examine which accounts were most prominent in a social media movement. The ten most favorited accounts in #MeTwo included members from civil society spreading the movement of Ali Can to a broader audience, which is key to successful online movement. As one can see Ali Can did not even receive the most favorites during the #MeTwo debate (alicanglobal: 6913), indicating that some even more famous social activists and members of civil society also took place in the debate, enabling it to make it more popular. In general, the 10 Twitter-accounts with the most favorites were the following:

  1. Hasnain Kazim (Journalist of SPIEGEL ONLINE)
  2. Miriam Davoudvandi (DJ and editor of splash! Mag)
  3. Shahak Shapira (artist, author and musician)
  4. missanphan (unknown)
  5. Mahret Ifeoma Kupka (curator, writer)
  6. problematash (unknown, but currently blocked)
  7. Ali Can (social activist, author)
  8. Oguz Yilmaz (Youtuber)
  9. JanaRennsteig (unknown)
  10. DieAgnosie (unknown)

As one can see, most of the highly favorited accounts are famous members of civil societies and have a migration background. This indicates at first that the debate was supported and developed through the active participation of these accounts. Furthermore, all seemed to have some direct or indirect personal relationship towards this topic, making it both societally and personally relevant for them. Yet, these famous Twitter-accounts seemed to have their share in the success of #MeTwo.

Turning to the number of retweets an account received, we find an extremely surprising picture. Among the top 10 of accounts having received most retweets there is actually no particular famous account. Although we are not quite sure, whether it might be an error in the display of the numbers, it seems that retweets did not depend on the popularity of the accounts during the #MeTwo movement. This, however, would indicate that #MeTwo has been a grassroot online movement with lots of people sharing their personal stories on national identity and acceptance within the German society.

Summing up the most active Twitter users on #MeTwo, we find that these accounts are primarily social activists and famous member of civil society (most have a migration background). Given that these accounts have lots of followers it becomes a bit clearer why MeTwo has been such a success.

Account status

Our previous results strongly indicate that social activists and popular members of civil society shaped the debate. To check whether the debate indeed was strongly influenced by specific groups of accounts, we classified all accounts into the following categories: * Verified account: Account is of public interest and thus officially verified by Twitter * Influencer: Account has more than 500 followers and its number of followers is at least three times higher than the number of followed accounts * Verified influencer: Account is both officially verified and an influencer (the most important accounts when trying to spread a social media movement) * Personal account: Account that is neither verified nor classified as an influencer

The following plot shows that the #MeTwo movement was actually quite widespread among the normal Twitter users. Far more than 40000 personal accounts took place in the movement, or at least retweet or favorized some tweets during the movement. Furthermore, we find support for the movement by a large share of verified and non-verified influencers amounting to more than 1000 individuals accounts. Lastly, also some verified accounts without a broader set of followers took part in the debate. These accounts, however, are mostly from political, social or academic institutions and should not contribute heavily to the movement. When taking a closer look at the non-personal accounts, it shows that many social activists simply are not verified by Twitter yet and elucidates that one should never fully rely on coding by third parties when analysing data.

As expected, the following plot shows us that personal accounts are by far more likely to retweet than tweet by themselves in the #MeTwo debate. For influencers, verified accounts and especially verified influencer this we find a much higher evenness in the frequency of tweets and retweets. Based on our knowledge on participation within social movements this can be easily explained using the idea of actual participants (gladiators) and spectators. It seems that actually all influencer and verified accounts who took part in the movement actually engaged in it, whereby lots of personal accounts did only retweet messages to spread the word. This is of course not at all reprehensible, but reflects our initial idea that #MeTwo follows the principle of classic social movements.

Tweeting about MeTwo in a Network

After analyzing how #We2 spread in our particular Twitter subpopulation, we now turn to the analysis of the retweet network during the debate. At first, we explain our directed retweet network. We define the nodes as follows: The source is the retweeting account and the target is the retweeted account. We define edges as a connection between two nodes if the source retweeted the target at least once. The coloring of the nodes follows the coloring of our categorization, with red nodes indicating influencers, dark blue personal accounts, light blue verified accounts, and purple verified influencers.

The layout of the network was calculated once for the entire time span. The only things that changes from one plot to the other is the activity, i.e. the connections between nodes. To make the plots more easily readable, we only show accounts that engaged at least 7 times with other accounts unsing the #MeTwo.

We plot the retweet network with the size of the nodes relative to their respective betweenness centrality. This measure represents the number of the shortest connections between different nodes pass through a node. Therefore, the higher the betweenness centrality, the higher the amount of information that has to pass through a node to reach another node in the network. The activities of nodes with high betweenness centrality are not only important for their own messages but for those of other as well, since they act as message distributors in the network.

In the first plot we look at the first two days of activity. We can see that the most central actors in the network are the personal accounts. Only one influencer with an already strong following appears central in the network. Accounts from verified users (for example media-related accounts) are almost completely absent.

The second plot shows the time between day 2 and 5 in our observation period. Similarly to the first two days, personal accounts are dominant. Essentially, only 6 accounts have high betweenness centrality.

During the last two days, we can see that activity between the actors in the network is starting to die down. But even here, no highly popular accounts appear to be able to profit from the activity of the hashtag.

Overall, the network seems to strenghten the results from previous analyses, when it comes to the most active category of users. Interestingly, the network appears highly interconnected, which means that most accounts are talking to each other. Even though we can see a small amount of clustered users, there are no clear “pro” or “contra” groups visible that do not interact. If there were separate conversations going on, the network layout algorithm would have arranged the nodes into clearly separable groups. However, this does not imply that all users agreed with each other. A retweet can also be used to express discontent with a message.

Tweet content

After examining the accounts who participated in the #MeTwo debate, we now turn to the actual content of the tweets. Ali Can’s original idea was to spark a discussion about people’s experiences with racism in Germany and the underlying question of national identity and belonging to Germany. As such, the focus of the hashtag should both be inward and outward looking as well as retro- and prospective, sparking a debate on national identity in a society having to deal with a new self-conception as an immigration society.

The next step of our analysis is structured by two main questions: First, what does the Twitter community actually tweet about their statements about their experiences with rassism and whether they feel accepted as Germans in Germany. Second, we examine whether the overall debate is more positive or negative. A positive sentiment would indicate that the Twitter community is willing to share their positive experiences and believes that Germany is able to find a good way to re-define its national identity, whereas a negative sentiment might indicate that either the debate has been captured by trolls or actual accounts mourning about the current stage of nationhood and their experiences as ‘new sort of Germans’.

Most common words

In a first analysis, we filtered the most common words that were used in the debate. In order to get a clean picture of these words, we removed stop words and user names prior to plotting. The remaining words show that ‘racism’ and ‘Germany’ were actually the dominating words structuring the debate in the #MeTwo movement. Turning to a word cloud with the 100 most used words, this picture becomes more nuanced, since words like society, foreigners, migration background, Turks, experiences, parents and teacher also show up. It seems that most of the debate is actually about the experiences Germans with migration background faced while growing up in Germany. In order to check how these stories untold, we now turn to analyses on the manner of writing about the individual experiences.

Co-occurring hashtags

A first indicator of the general notion of the content in the tweets comes with the hashtags also used when posting. These co-occuring hashtags provide a certain frame by which the content of the tweets can be interpreted. As one can see, rassism and metoo are the two most often co-occuring hashtags for the #MeTwo movement. These two hashtags highlight both the overall tone of the movement as well as its predecessor - the MeToo movement, sheding light at sexual harassment, violence and discrimination against women.

The third most often used hashtag shed lights to one fundamental problem of social movements - being hijacked by trolls and those feeling offended by emphasizing latent rassism in Germany. The so-called K.O. Challenge, initiated by Fatlind Albo, invoking young men with migration background to beat up autochton Germans. This, of course, attract the attention of right-wing activists associating with challenge with MeTwo and dis-credit all men with migration background as potential violent criminals. As this hashtag co-occured the third-most often in our data, we have to take care that a substantial share of tweets are posted by members of right-wing activist groups to discredit the MeTwo movement.

The remaining hashtags seem again to be related with the idea of MeTwo as such, although AfD might be highlighted as a political party denying the new realities in Germany and 29TemmuzDünyaGüzelLikGünü refer to the World Beauty Day, which we are not able to associate with the MeTwo movement at all. It seems a bit as if the movement did not only face right-wing trolls, but also interaction from users based in Turkey.

Tweet sentiments

The last layer of our content analysis concerns the sentiments associated with the respective tweets mentioning #MeTwo. Our sentiment analysis is based on a dictionary approach using the SentiWS dictionary by the University of Leipzig. The dictionary classifies which German words have a negative and positive meaning, respectively, and assigns numeric values to them. We then can simply map the words used in the tweets against the words included in the dictionary along with their values.

As we can see in our first analysis, the overall sentiment of the #MeTwo debate is rather balanced, however, with about 30 percent more positive than negative words. This rather high number of negative words indicates that many people had negative experiences when growing up in Germany. On the other side, however, we see lot of positive words, reiterating our assumption that many people also have made positive erperiences and have a positive stance towards this movement.

When examining the sentiment distribution over time, we see that we actually almost always have more positive words than negative ones. However, the margins between positive and negative words are steadily decreasing with only a few more positive than negative sentiments in the last days of the movement. Yet, in the first day one clearly sees the excitement which came along with this new social movement, transporting a positive reinterpretation of what Germany might look like in the future.

But which words are actually classified as positive and negative words in the #MeTwo debate? In the following overview we see that good, better, understanding, right, luck and love are the most frequently used positive words in the debate. This indicates that most users try to tell a positive story on their experiences but also their expectation for the future. On the negative side, we see problems, discrimination, hate and sorrow. This might refer to crucial negative experiences faced in the past, highlighting that people with migration background actually face problems of racism in the past, but perhaps up until now in their everyday life.

These findings can be visualized in a comparison cloud as well:

Tweets in foreign languages

Lastly, we examine whether the debate spread beyond a German-speaking subgroup on Twitter. Using automated language recognition functions, we classified the tweets in different languages. The following plot shows all languages with a minimum share of 0.001% that were present in the tweets. We see that the majority of the debate took place among German-speaking users. However, we also find some tweets in English, French, Japanese, Spanish and French. Although not being particular popular among foreign Twitter users, it seems that #MeTwo has somewhat spread to other European countries as well.

MeTwo and the Media

In a final step, we examined how the gate-keepers of societally relevant information - the media - took part in the MeTwo movement. In order to identify media accounts, we match all accounts refering to media institutions or journalism among the verified accounts to get a better picture of the total amount of journalists being willing to take up the story of Ali Can and #MeTwo.

As we can see, more than 200 media institutions and 150 journalists from all over the world reported on MeTwo. This is a considerably high number, showing that MeTwo was not only a Twitter movement, but spread to a broader audience, which can be targeted by the journalists. These actors are among the most important to lead social movements (especially in their online version) to success. They are able to spark public discourse about these topics and might even change public perception on topics like #MeTwo.

In order to check whether journalists and media institutions actually took part in the #MeTwo debate, we again examine the number of tweets, retweets, foavorites and followers for this sub-group of accounts. In the following plots, we present the results for these measures and compare them with the remaining verified accounts, not being affiliated to the media. At first, we can see that media institutions were actually tweeting far more than retweeting on #MeTwo. For journalists we find the exact opposite. Given the function of these accounts, this is very reasonable and confirms our assumption that #MeTwo has travelled to a broader public via the media. since institutions should not post personal stories, they should refer to shows in their program associated to #MeTwo using Twitter. This is a first indicator that #MeTwo has received a broad coverage also outside of the Twitter community. Journalists, however, actually took part in the debate - perhaps even because they wanted to share their stories as well. For non-media accounts, we find a comparable pattern to journalists.

Taking a closer look at actual media accounts being most active in the debate, Hasnain Kazim (journalist SPIEGEL), Düzen Tekkal (filmmaker and journalist), Tobias Huch (journalist and author), Dunja Hayali (journalist ARD) and Jens Clasen (editor Womens Health) stand out in particular. All of these perons are well-known to be active in fighting for a liberal and value-driven German society and have a large followership. Thus, they are both credible in speaking up in the #MeTwo movement and also able to mobilize a lot of people. Among the media institutions, Perspective Daily (online-magazine), SPIEGEL ONLINE and the Süddeutsche Zeitung are to be named as most active and influencial. Again, we see a clear pattern of left-liberal outlets being willing to spread the word on #MeTwo and helping the movement getting recognition and shaping public discourse.

Conclusion

Taken together, #MeTwo has been a rather successful online debate. With our Twitter analysis we tried to confirm existing assumptions about the debate and conveys some additional and interesting results. Our findings show that #MeTwo is both positive and negative in its tenor and, strongly influenced by prominent social media activists and journalists already focussing on societal problems concerning national identity and immigration. The hashtag seems to had a unifying effect on those persons, most often fighting for these issues by themselves. Personal accounts as well as the media and individual journalists, who are essential for social media events to become successful, were part of this movement. This explains why it was so successful in last year’s summer.

When compared to Ali Can’s other hashtag #We2 (see a short blog post here), we can see that #We2 is mainly driven and supported by political elites rather than the general population. Unlike #MeTwo with the debate on Mesut Özil’s retirement from the German national football team, there was no triggering event of broader public interest when #We2 emerged. As a result, the media response is largely absent and hence the potential of traditional media channels has not been fully exploited yet.

To conlude, #MeTwo went extremely viral and perhaps sparked a debate on national identity and Germanhood in a migration society. Using data science, we hope that we were able to provide insights about certain aspects of the movement and were able to show which aspects are crucial for social media movements to become successful.

Konstantin Gavras, Lisa Hehnke, Paul Meiners, Sandra Meneses & Juan Orduz

Konstantin Gavras, Lisa Hehnke, Paul Meiners, Sandra Meneses & Juan Orduz

Konstantin (@kongavras) is a Ph.D. candidate at the Graduate School of Economic and Social Sciences in Political Science and research associate at the Chair of Political Psychology at the University of Mannheim. Lisa (@DataPlanes) is a freelance social data scientist working to address current societal challenges in order to make a difference and drive change. Paul Meiners (@plmeiners) is a Ph.D. candidate at the Gradutate School of Politics in Munster and research associate at the department of Political Science at the University of Muenster. Sandra Meneses is data scientist at Mercedes Benz.io. Juan Camilo Orduz (@juanitorduz) is data scientist at TD reply.