Strasbourg Observers

Basu v. Germany and Muhammad v. Spain: Room for improvement in the Court’s first judgments on racial profiling

November 08, 2022

by Mathias Möschel

On 18 October 2022, the Third Section of the Strasbourg Court, decided two cases dealing for the first time with the question of whether and how far racial profiling by public authorities constitutes a violation of the Convention. The outcome is a mixed one. Whereas in Basu v. Germany the judges held that the facts constituted a procedural violation of Article 14 (prohibition of discrimination) in conjunction with Article 8 (right to private life), no violation was found in Muhammad v. Spain. The comparative and critical analysis of these two judgments provides interesting indications on what to keep and what to improve for similar future cases.


The fact patterns in both cases deal with individuals who claimed being singled out by police forces for ‘random’ identity checks only, or predominantly, due to their skin colour/race. In the German case, a father and his daughter, both German nationals of Indian origin, were checked by police shortly after their train had crossed the border from the Czech Republic into Germany. They were the only passengers whose documents had been checked and were the only non-White passengers on that carriage (§§5-6, Basu). As a result of the check and the humiliation suffered by it, the applicant stopped travelling by train for several months (§20, Basu). In the Spanish case, a Pakistani national with long-term residency in Spain was asked for his identity while walking on a busy street of Barcelona by national police forces. The versions as to why he was subjected to that check differ. The applicant claims that it was due to his identity whereas the police states that he had behaved disrespectfully toward them prior to the check. The applicant refused to follow the order, was arrested, and only released once he had decided to show his identity documents (§§5-10, Muhammad). As a result, he received an administrative notice for refusal to identify himself and insolent attitude towards the authorities.

In both cases, the applicants unsuccessfully challenged the racially discriminatory aspect of the identity checks they had been subjected to at the national level: first in administrative courts and then, via constitutional complaints, to the respective constitutional courts. In Basu, an internal police investigation was inconclusive, and the courts had dropped the case on formal grounds, namely that the applicant had no legitimate interest in challenging the lawfulness of the identity check. In Muhammad, the applicant had also brought separate criminal charges that did not raise the racial discrimination claims and that were eventually dropped.     


The Strasbourg Court (re-)characterized both cases as falling under the same legal basis: Article 14 in conjunction with Article 8 (§18 Basu and §43 Muhammad).

In Basu, the Court first addressed the question of admissibility raised by Germany on whether the case even fell into the ambit of Article 8. For that purpose, the Court explained that public searches and checks do not automatically violate a person’s right to private life and that for this provision to apply, in some situations a certain threshold of severity needs to be attained (§§22-23). It also mentioned that racial discrimination is deemed a particularly egregious kind of discrimination that requires special vigilance and vigorous reaction (§24). Considering both elements together, the Court held that the minimum threshold had been reached, because the applicant and his daughter had been targeted on account of specific physical or ethnic characteristics and that no other objective ground had been provided by the police agent for targeting him. Therefore, the case fell into the ambit of Article 8 (§§25-28). An almost identical reasoning on admissibility was applied in Muhammad (§§49-52)

Moving on to the merits, in Basu the Court only analyzed whether the State had investigated the racist attitudes behind the events. It essentially extended such duties from case law involving racist violence dealing with Article 14 violations taken in conjunction with Article 3 (prohibition of inhuman and degrading treatment) to this situation of Article 14 violations with Article 8 here. In that context, it had especially highlighted that authorities must do whatever is reasonable to secure evidence for a fully impartial and objective decision as well as to make sure that the authorities responsible for carrying out the investigation are truly independent (§§31-33). Applying these principles to the case, the Court held that the internal investigations made by the police on the potential racist motives could not be seen as independent and that the courts’ dismissal of the case on formal grounds could not be seen as an effective investigation at all. For that reason, it found a violation of Article 14 in conjunction with Article 8 (§§36-39) but was unable to find a substantive violation concerning the racial motive because the investigation was simply insufficient and did not offer enough material to rule on this specific aspect (§38). It also held that there was no need to examine the complaint under Article 13 (right to an effective remedy) (§§40-42) and that the claim for violation of Article 2 Protocol No. 4 (freedom of movement) under was manifestly ill-founded (§§43-44).   

In Muhammad, the reasoning was more detailed but led to a finding of no violation at all. As opposed to Basu, here the Court assessed both the procedural issue of whether an effective investigation had been made into the racist motives behind the identity check (§§53-76) and the substantive question whether that control had been racially motivated (§§77-103). As to the first aspect, the Court first established that there was an obligation to investigate the issue of racism (§§63-68) and then essentially concluded that the Spanish authorities had complied with its obligation to investigate possible racist motives in the sense that the legal framework had allowed the applicant to challenge the Spanish courts’ reasoned and motivated decisions (§§69-76).

As to the second aspect, the Court in citing its own precedent first repeated that it has recognized both direct and indirect discrimination under its case law on racist violence (§§92-93) and that the burden of proof shifts from the applicant to the government only once the applicant shows that a difference in treatment took place (§94). In any case the government in instances of racial prejudice “cannot be required to prove the absence of a particular subjective attitude on the part of the person/persons concerned” (§95). Applying these principles, the Court held that neither the applicant’s arguments of having been singled out based on race for identification nor the studies brought by third parties demonstrating the practice of racially motivated identity checks in Spain sufficed to shift the burden of proof to the government here. Indeed, it agreed with the national courts that the applicant’s recalcitrant behaviour and not his identity were the cause for the treatment (§§99-103) and thus found no substantive violation either.  

Both decisions had separate opinions. In Basu, there was one partial dissent by Judge Pavli. Muhammad instead was taken by a narrow 4-3 majority. Two judges concurred (Judges Elósegui and Serghides) and provided additional elements on why the majority opinion was correct, whereas two others authored single dissents, essentially disagreeing with the Court’s assessment of the investigation (§§1-6, Judge Zünd and §§4-8 Judge Krenc). It is unclear who the third dissenter is. The judges in both cases are almost the same, except that Judge Roosma substituted Judge Pavli in Muhammad. At the same time, Judge Pavli was aware of the Spanish judgment given his references to it in his dissent (§§15 and 20 dissent).


The Strasbourg Court has already found racial profiling in violation of Article 14 in conjunction with Article 2 of Protocol No. 4 at border entry checks (Timishev v. Russia) and also of Article 14 in conjunction with Article 3 when happening to an African prostitute working on the streets (B.S. v. Spain). Moreover, certain practices of stops and searches had already been found in violation of Article 8 of the Convention (Gillan and Quinton v. the United Kingdom and Vig v. Hungary). However, the Court had not yet decided whether stops and searches in an everyday context of ordinary people in public transportation or simply walking on the streets could also amount to race discrimination under Article 14 in conjunction with Article 8. In this sense, Judge Pavli rightly states that these two judgments are ground-breaking, while simultaneously indicating their shortcomings in his dissent (Basu, §§1-2 dissenting opinion). His critiques also serve as a basis for the first comments here before moving to the broader picture.  

Burden of proof, procedural and substantive obligations

First, how the burden of proof is (not) operationalized in these cases is questionable. As noted by Judge Pavli, the Court here is not merely asking the applicants to provide facts from which a prima facie case of discrimination can be presumed to shift the burden to the government. Rather, it seems to require full proof of racism and racial prejudice. In this sense, Judge Pavli highlights that also in the Court’s case law on racial violence, this shift is not applied because the underlying cases are criminal law ones where at national level the shift of burden of proof cannot be applied. However, it is applied in cases of indirect race discrimination (e.g. D.H. and Others v. Czech Republic, §§175-195). Such a difference leads to the ironic outcome that applicants are better protected in indirect than in direct discrimination cases with an underlying national criminal procedure before the Court (§§3-8 Judge Pavli dissenting).

As a related point, it is unclear why in a human rights court there should be any difference between burden of proof in criminal and in administrative cases as suggested by the concurring opinion in Muhammad (§§5-9, Judges Elósegui and Serghides in Muhammad). That would mean that different standards would apply depending on whether a case arose out of a civil, administrative or criminal law context. Especially problematic is the extension of the criminal law standard of proof beyond reasonable doubt to individual applicants in a human rights court. At the national level, rightly so, it is supposed to protect individual criminal defendants from state prosecutors who have all the information and power, but applicants before the Court are in a vastly different position. Overall, a careful and equal application of shifting the burden of proof in all cases involving (race) discrimination once the applicant has made a prima facie case of racism, whether direct or indirect discrimination or arising from criminal, civil or administrative law cases, should be the way forward.

Second, both judgments fail to explicitly and separately engage withanother key positive obligation of the State: putting in place a sufficiently protective legislative or regulatory framework. Even though both applicants raised the issue, the Court avoids it and focuses almost exclusively on the insufficient investigation against the kind of racial abuse witnessed by the applicants (§§9-13, Krenc in Muhammad and §14 Pavli in Basu). It is unclear why the analysis for the applicable legal framework and its sufficiency is described and clarified in more detail in the concurring opinion (§§26-30, Elósegui and Serghides in Muhammad). In any case, such an analysis should not be missing in future cases in the main arguments.

As to the focus on the second procedural obligation of running an efficient and independent investigation, it is commendable that the Court requires and, at least in Basu, finds that indeed that had not been the case. However, the consequence the Strasbourg judges then draw from it, namely that the absence of such an investigation and the connected material preclude them from assessing whether there is a substantive violation, creates what judge Pavli calls a “perverse incentive” (§4 Judge Pavli dissenting) for states to sweep such incidents under the rug and thus avoid any substantive liability for human rights violations by police forces or public authorities if and once a case reaches the Court. The Court has been using this shortcut in other cases, too (e.g. in a case dealing with harassment in a state ministry: Hovhannisyan v. Armenia, § 60). So, more broadly, a lack of effective investigation should not serve as a reason to dismiss any analysis into the substantive violation. On the contrary, it should serve as an indication the probably overall the system is ill-prepared and therefore allows for such incidents to take place.

Last, as to the assessment of the substantive question whether the identity checks – and not ‘simply’ the insufficient legislative framework and/or the ineffective investigation – were racially discriminatory, the Court’s reasoning remains rather thin and is mostly linked to shifting the burden of proof and the question of intent, prejudice and causal link between the identity check and racism. Ultimately, both judgments are an example of “procedural minimalism” (§§19-21, Judge Pavli dissenting in Basu).   

Locating the cases in a broader picture

While agreeing with Judge Pavli that the two judgments are ground-breaking, it might have been more correct to state that they could have been ground-breaking. Indeed, a recent European Parliament briefing as well as various country specific studies and/or country reports referred to the Court in both cases demonstrate how racial profiling is a broad and arguably structural problem in Europe. Moreover, already more than a decade ago, in 2009, the Human Rights Committee (HRC) had held that the plaintiff, an African American woman stopped by police forces for an identity check while in a Spanish train station, had been the victim of unlawful racial profiling in substantive violation of Articles 2 and 26 of the UN’s International Covenant on Civil and Political Rights (ICCPR) (Rosalind Lecraft Williams v. Spain). In this sense these two judgments only scratch the surface of the issue. But there are various ways for improvement.

One important way forward could be for the Court to consider its case law on violence against women (e.g. domestic violence) and apply the standards and practices developed in that context to cases involving racial violence and now also racial profiling. It would bear several consequences. First, in that case law, the Court regularly assesses whether the procedural obligation of putting in place a sufficient protective legal framework has been met (see e.g.: Volodina v. Russia §§125-133). Second, with gender-based violence the Court has been able to link individual situations to systemic and structural forms of gender bias by state institutions (see e.g.: Volodina v. Russia, and Tkhelidze v. Georgia). The Court has by and large so far not done that in the context of racial violence (see here and here)and now also of racial profiling. In fact, here it explicitly refused to give material provided by various international actors on structural and systemic racism in Germany and Spain any relevance and rather focused on a narrow, causal approach in the single cases before them (§100 Muhammad and the critiques in Pavli’s dissent in Basu at §§9-12). Ultimately, much in Basu and Muhammad seems to hinge on whose version of the facts seems more credible to the judges (both national and supranational) and whether the benefit of the doubt is given to national police forces or to the applicants. Fourth, the parallel with the Court’s gender-based violence case law might also help moving away from the distinctions on burden of proof in criminal, administrative or civil cases, something that is absent in that context. Indeed, the applicants in those criminal law cases are not asked by the Court to prove beyond reasonable doubt that gender bias motivated the state authorities.

The parallel with gender-based violence also allows to raise another point, namely whether these cases would not be conceptually better characterized under Article 14 in conjunction with Article 3 instead of under Article 14 in conjunction with Article 8? First, the Court has already applied this logic in a case of intersectional profiling and violence in B.S. v. Spain, with a rather more positive outcome for the applicant. Second, there has been a similar overall conceptual shift from Article 8 to Article 3 – with or without Article 14 – in gender based violence cases due to the Court’s better understanding of the harm in such cases (see e.g. Valiuliene v. Lithuania). Such a better understanding could also take place here. Third, while the shift from Article 8 to Article 3 might introduce a higher severity threshold, the Court has already stated that the severity threshold under Article 3 can be reached also without physical violence (Volodina v. Russia, §§73-75). So that could be extended to racial profiling as well, especially when considering that the Court itself recognizes the seriousness of race discrimination which asks for special attention from the institutions. Moreover, would the dignitary harms of being singled out in public by state agents because of one’s skin colour, not be sufficient evidence and arguments for lawyers and the Court as the master of characterization of facts to the case, to consider (re-)characterizing at least some of these cases as Article 3 in the future, even in the absence of physical violence?

Concluding on a positive note, the Court has the chance to rectify some of these elements soon in other cases on racial profiling arising in different legal systems: Wa Baile v. Switzerland, pending before the same section and Seydi and Others v. France before the Fifth Section. It will be able to take its own words that “racial discrimination is a particularly egregious kind of discrimination and, in view of its perilous consequences, requires from the authorities special vigilance and a vigorous reaction” (Timishev v. Russia, §56) seriously and apply it to its jurisprudence.

Print Friendly, PDF & Email

Related posts

Leave a Reply

Your email address will not be published. Required fields are marked *

1 Trackback