How widespread is racial profiling?

A difficult question to answer in migration matters is to what extent the police are allowed to act on substantiated presumptions and statistics regarding the (criminal) behaviour of certain ethnic or national groups. Thus, we hear stories of certain minorities being targeted in police or border controls without any apparent reason; but also about parallel societies in areas where crime is rampant, with Germany currently being confronted with organised crime perpetrated by large extended families or clans often with an Arab background.

A recent overview provided by the Council of Europe shows how ethnic profiling is still a persisting practice in Europe. It refers to ethnic profiling as “the use by the police, with no objective and reasonable justification, of grounds such as race, colour, languages, religion, nationality or national or ethnic origin in control, surveillance or investigation activities”.

According to Open Society Foundations, “law enforcement engage in ethnic profiling when they base their actions on ethnicity, race, religion, or national origin instead of an individual’s conduct or objective evidence”. These actions may be carried out by all ranks of law enforcement.

In some countries, the ethnic element is already reflected in the terminology used. Germany, Italy, Poland, and Sweden employ the terms Staatsangehörigkeit, cittadinanza, obywatelstwo, and medborgarskap to describe the legal link of citizenship. The words Nationalität, nazionalità, narodowosc and nationalitet are less frequently used to describe this legal link because of their ethnic connotation.

As a general rule racial profiling is prohibited, not just in criminal matters. In France, soccer club Paris Saint Germain was fined for illegally racially profiling young players as part of their recruitment process. The line between racism and the justified targeting of certain ethnic groups can be thin and difficult to draw. Some outcomes of the prohibition of ethic registration are also questionable, as targeting ethnic groups who are overrepresented in criminal statistics is officially prohibited, while the law often makes an exception for the same groups if registration of their ethnicity is considered vital to implement positive discrimination policies. The risk of racial profiling also appears greater if access to the territory cannot be denied at an earlier stage. Worried about Albanian criminals’ access to the Schengen area, the Dutch government recently unsuccessfully pleaded with the European Commission for a reintroduction of a visa duty.

Traditionally, ethnic profiling was driven by unspoken biases in exercising discretionary powers. With the arrival of machine-learning algorithms, however, some of these biases have become formalized. The Council of Europe notes that although still in its experimental stages, the use of such algorithms in the criminal justice systems is becoming increasingly common and the UK, the Netherlands, Germany and Switzerland, among others, have been testing them, notably in the field of “predictive” policing. Algorithms are now used to predict where crimes may occur and how best to allocate police resources; to assess the risk of reoffending for the purpose of deciding whether or not to prosecute; and to forecast reoffending in the context of decisions on remand in custody, sentencing and parole.

The danger of “social sorting”, based on prejudices and social stereotypes that will unwillingly always corrupt the surveillance tools, seems confirmed by research concluding that “systems underpinned by machine learning will inevitably reproduce the inherent biases present in the data they are provided with, producing predictions that reinforce or even amplify improper data recording practices”. The big challenge is thus to overcome the problem identified by the Council of Europe that “feeding an algorithm with data that reproduces existing biases or originates from questionable sources will lead to biased and unreliable outcomes”.

A paper by Alpa Parmar, a researcher investigating the nexus between race and criminalization, deals with the intersection between policing, immigration law and racialization. Although this is not a new phenomenon, Parmar feels “the political, conceptual and operational move to formally conjoin policing with immigration in contemporary times is troubling because police work has long been reputed for using prejudice as a tool through which the ends are thought to justify the means”. What she did not expect to witness so clearly in her interviews with police officials in the UK was “the role of public assistance provided to the police in their role regarding migration control”. Parmar feels the police “rationalised public reporting through tandem narratives of crime risk (of terrorism, fraud, abuse) and mounting pressures on public services”, and suggests that possible racist intent was not considered as a motive by the police.

On the positive side, the fight against ethnic profiling is firmly on the agenda and the Council of Europe lists to a growing body of national and international case law on the issue. Thus, in Lingurar v. Romania the European Court of Human Rights found violations of Article 14 (prohibition of discrimination) of the European Convention on Human Rights in conjunction with Article 3 (prohibition of inhuman or degrading treatment) in a case of a racially motivated raid by police on the Roma population. This is a key ruling as it is the first time the Court explicitly used the term “ethnic profiling” with regard to police action.

Author: Dr. Olivier Vonk