By Tim Dekkers, Maartje van der Woude & Robert Koulish
Migration control has seen an influx of information technologies in order to control mobilities. The traditional means of securing borders with walls, gates and border guards are not considered to be sufficient anymore or, as is the case at the internal borders of the European Union, legislation prevents the implementation of such measures. This poses a problem. Migration is increasingly viewed as a security issue and migrants as security threats, but freedom of movement benefits trade and tourism resulting in economic growth. EU member states therefore face the dilemma of security versus freedom of movement. In an age of rapid technological development, policymakers often turn to new technologies to address this issue. Risk assessment technologies, for example, would be able to analyze immigration related information in order to distinguish high risk from low risk individuals allowing migration authorities to efficiently and effectively target only high risk individuals while letting low risk individuals pass. In other words, risk technologies would be able to distinguish wanted from unwanted individuals in cross-border mobilities. As such technologies are assumed to operate on objective and unbiased risk criteria, the implementation of risk tools would result in an objective decision-making process inhibiting potential subjective biases that decision-makers could have.
The application of risk technologies and their claimed benefits are contested, however. One of the main criticisms is that for risk assessment technologies to work, users would have to exactly follow the assessments made by the technologies. If users were given discretion to interpret the presumed objective risk assessments, the objectivity of the decision could be called into question. Critics of risk assessment technologies are also of the opinion that eliminating discretion from the decision-making process would be a very difficult task. Human agency is a force to be reckoned with.
Despite the prevalence of risk technologies in migration control and human agency potentially undermining the premise of these technologies, empirical research on how they are used in practice is scarce. By means of a case study of the Dutch Mobile Security Monitor (MSM) we aim to shed light on the issue of risk tools and human agency. While the Schengen Agreement has abolished permanent border control inside Europe’s Schengen area, it does allow for police activity in border areas as long as this does not equal systematic border control. The MSM is the Dutch interpretation of the Schengen Agreement’s border-policing provision and is carried out by the Royal Netherlands Marechaussee (RNM), a paramilitary organization responsible for border policing in the Netherlands.
MSM checks can be found at airports, trains and highways that cross the border, but the underlying research that my two coauthors and I conducted focuses only on the highways. In accordance with the Schengen Border Code, not every vehicle or individual that crosses the Dutch border can be stopped. Rather, Marechaussee officers must select some vehicles out of the cross-border traffic. In 2012 the RNM introduced a smart camera system called Amigo-boras to assist the selection process. Using risk profiles, the camera system would be able to identify high-risk vehicles which could then be stopped by RNM officers. Much like many risk technologies, Amigo-boras is presented as a means to make the decision-making process more objective by employing risk profiles based on statistical analysis. Considering the possible tension between human agency and risk tools, it would be interesting to see how the camera system is used in the selection process.
To understand how Dutch police are using MSM, we observed RNM officers in the field and spoke with policymakers and officers. The interviews with policy-makers indicate that the system is meant to create a balance between the risk technology and human discretionary decision-making. The system gives an indication of a high-risk vehicle and RNM officers can use this information to decide whether or not to stop a vehicle. In using the system, RNM officers are therefore not bound by the risk assessment of the camera system and can still exercise their discretion. During the fieldwork, officers also indicated that this was very much necessary. Officers did not like the system, as the risk profiles were very general. The cameras would only analyze the license plate of a vehicle and determine the country of origin. If this country was considered to be high risk the system would signal the officers. Officers explained that while the license plate is an important indicator for the selection process, many other factors were needed to make a proper selection. In addition, officers were of the opinion that the risk profiles were redundant in the first place because they were also able to see license plates from the side of the road. To officers, the intended balance between technology and officers in the decision-making process was not achieved. Yet, this may be just their perception. Risk technologies being information technologies, quantitative data about how the system is working could potentially show otherwise. Surprisingly, the RNM claims to have no data on the use and outcome of the Amigo-boras system. The system only tracks which vehicles are flagged as high risk, but not if the vehicle is actually stopped or what the outcome of the stop is.
The case of Amigo-boras allows us to learn several things about the use of risk technologies in migration control. First, risk technologies could be nothing more than a veneer of technology in a decision-making process that is otherwise highly dependent on discretionary decision-making, raising questions as to what extent they make the decision-making process more objective. Second, while risk technologies are designed to remove biases of human decision-makers, they also could amplify biases that become ingrained into a technology that is assumed objective. Amigos-boras illustrates this concern. The risk profiles used by Amigo-boras were based on the idea of high-risk countries, hinting towards stereotypical assumptions about specific nationalities. Third, if no quantitative data can be produced on the system’s role in the decision-making process, it is easy to justify unjust decisions by ascribing the decision to the presumed objective risk technology. Technologies such as Amigo-boras can therefore be used as a shield for holding individual officers accountable. With this in mind, it is important to gain more empirical insight in the use of risk technologies in migration control to see if they are as beneficial as they are presented to be.
If you would like to learn more about risk technologies in migration control, please read our recent publication in the European Journal of Criminology.