Datafying Islamophobia in Catalonia
Entire Muslim communities are targeted and surveilled by an expanding “preventive” sociotechnical system
By Aitor Jiménez (University of Melbourne/ ARC Centre of Excellence for Automated Decision-Making and Society) and Ainhoa Nadia Douhaibi (Universitat Oberta de Catalunya)
After more than 30 years of living in Catalonia, Mohamed Said Badaoui — who is married to a Spanish citizen and the father of three children, all born in the country — decided to apply for citizenship. Not only did the authorities deny his request, they ordered Badaoui’s expulsion from the country under the article 54 of the Spanish Immigration law. They alleged that he was “radicalized,” and that his case was a matter of national security. Badaoui had no criminal record, nor previous contact with law enforcement agencies; instead, the police report backing up the expulsion procedure was filled with data referring to his activist and religious life. A quick look at the police record shows what reliable data indicating radicalization means for police agencies: a handful of poorly interpreted Facebook posts, contact with anti-Islamophobic activists, strong ties with local Muslim association, and the respect and recognition of the Muslim community. What in any other case would have been indicators of the natural exercise of civil and political freedoms — things like attending meetings and engaging with community members — instead triggered a life-threatening deportation order. How did this come to happen?
Catalonia is home to the largest Muslim communities in the Iberian Peninsula; roughly 8 percent of its population follows the Islamic tradition. The majority of Muslims work in precarious jobs or don’t have jobs at all. Roughly 20 percent of the migrant population is unemployed, compared to just over eight percent in Catalonia (INE, 2022). They tend to live in impoverished and deprived zones with less access to public resources and green areas. Traditionally migrant neighbourhoods such as la Barceloneta, el Raval, and Poblenou in Barcelona are among the most affected by the touristification and gentrification unleashed by foreign investment firms. With scarce jobs, and skyrocketing rents and living costs, thousands of families are forced to live in slums and industrial areas with extremely poor living conditions. Muslim communities are also targeted for their beliefs and culture. Despite their demands, 90 percent of Muslim students do not enjoy the same right to religious class in the public education system as their Catholic Christian counterparts. Despite comprising just 3.1 percent of the population, people of Maghrebian background represent 16 percent of Catalonia’s incarcerated population. As countless academics and activists have pointed out, this is not a matter of rampant criminality among a very specific and identifiable segment of population, but a consequence of racial profiling. Episodes of discrimination are not an accident, but functional elements of a system of domination that is intended to reinforce structural gender, racial and class inequalities through a sociotechnical system that encompasses all sorts of surveillance, repressive, legal, political, economic, educational, and military instruments.
From the early days of inquisition to the latest algorithmic developments, the social construction of the Muslim as a social enemy has helped to shape both the Spanish identity and the Spanish state’s surveillance and repressive apparatuses. Muslims are labelled everything from job-stealers to terrorists. The second decade of the 21st century has seen the proliferation of heavily racialized surveillance and carceral geographies. As anti-immigrant raids in the United Kingdom, the United States, and Australia demonstrate, border reinforcing technologies now extend to every territory, every street, every working place. The “exceptional” and “temporary” powers to surveil and punish —conferred to public authorities in order to fight the “war on terror” — are now well established public practices.
Since 2016, Catalonia has been implementing the Catalan Protocol for Prevention, Detection and Intervention in Processes of Violent Extremism (PRODERAE) in schools, local police stations, prisons, and social services. PRODERAE is part of the wider Special Counter Terrorism Policing Operational Program. Most details of the PRODERAE remain unavailable to the public and hence hidden from democratic scrutiny, but thanks to a leak, we have had access to some documents and a non-official recording of the PRODERAE training. Our early findings show how mosques and entire Muslim communities are targeted and surveilled by an expanding “preventive” sociotechnical system. As a result, an army of educators, social workers, and police officers are entrusted with gathering information from endless data points and asked to report to their civil and police superiors the most subtle changes in individual and collective behaviour. For instance, teachers are taught by police agencies that the quotidian manifestations of religiosity — such as the adoption of “Islamic” dress codes or collective prayer — could be symptoms of radicalization. This information is then used to threaten vulnerable with criminalization, family separation, and even deportation.
The PRODERAE lays out four areas in which to evaluate the risk of a subject: personal development, school context, family context, and social context. To obtain information the system relies on a vast array of agents, technologies, and points of data extraction, members of the community, educators, social workers, police officers and intelligence services. The Catalan government has deployed considerable resources to provide training on the use of these tools to educators, lawyers, social workers, and police officers — blurring the boundaries between welfare and policing, street surveillance and cyberwarfare.
Under personal development, among the qualities the tools evaluate negatively are “the difficulty of managing emotions,” “the difficulty of building a multiple identity,” the “proximity to radicalized peer groups,” and “low expectations of success.” Elements such as the dress code (hijab, niqab), personal appearance (beard), as well as dietary and leisure habits (halal, alcohol consumption), are surveilled with special interest. In educational contexts, teachers are asked to pay special attention to “the lack of bonds between peers” and “the difficulty of (the teacher) establishing bonds with students,” which are considered to be risk indicators. In the family context, “low family participation and involvement in school activities” and “the sense of belonging” are among the elements to consider in measuring potential radicalization. Under social context, the instruments look at “the influence of social networks,” or if the individual belongs to “socioeconomically disadvantaged contexts.” Information collected is transferred to the Territorial Evaluation and Monitoring Board, where police officers and education inspectors will assess the indicated risk. This could lead to further investigation, wiretapping, raids, detentions, and deportations.
Given the opacity, secrecy, and lack of transparency guiding cases of alleged radicalization, researchers, activists and even politicians struggle to access critical information. What data gathering tools — both analogical and digital — are authorities using? How is the data gathered across services being stored, processed, and analyzed — and by who? Are these data sets feeding automated decision-making systems used in the public sector? Who is entrusted with overseeing these data-intensive tasks? Have these instruments passed any form of algorithmic auditing and impact assessment? We have posed these and other questions to the Spanish and Catalan authorities, and received no response.
The system to prevent terrorism operates as a self-fulfilling prophecy: what risk assessment tools may flag as threatening symptoms of radicalization are often mundane and even contradictory facts. For example, authorities might believe that a young Muslim is up to something because he exercises too much, or is too sedentary. Similarly, young Muslims who follow religious rituals may signal fundamentalist tendencies, but not following religious mandates may be, in the eyes of the police services, a worrying symptom of latent lone-wolf tendencies.
A vast sociotechnical web of analog and digital technologies controls the lives of thousands of Muslim people in Catalonia. The surveillance apparatus deployed for gathering the data of vulnerable populations, and the extensive use of actuarial and algorithmic methods, amounts to a datafied surveillance state. Every Facebook like, attendance at a meeting, public expression, social exchange, and online purchase is seen as data worth gathering — information that perhaps, one day, could pre-empt a “future terrorist.” The more common result is that the people being surveilled are the ones who are terrorized.
On October 19 the Catalan Parliament’s Committee of Spokespersons declared its support for Mohamed Said Badaoui, calling it a case of “state Islamophobia” and emphasizing how opaque procedure against activists violates essential judicial and procedural rights. Then at midnight on November 20th, after a month in the Barcelona immigration detention center, Badaoui was put on a plane — escorted by 13 police officers — and deported to Morocco. There is no doubt that data-centric technologies played a role in legally producing Badaoui as a terrorist. Yet it would be naive to think that “fixing” the high- and low-tech risk assessment tools that led to to this situation would change his plight or result in a significant change in future “counterterrorist” operations. Still, by acknowledging such political fundamentals we can begin to grasp how structural racism fuels technological developments, and how deeply technologies are connected to the political and economic environments in which they were built. Understanding how sociotechnical systems operate may open up new ways to imagine emancipatory technologies of the oppressed.