Let us imagine that, tomorrow, the Parliament were to reduce the legal time frame for abortion from 14 to 12 weeks. Is it certain that the Conseil constitutionnel would declare such a law unconstitutional? It is difficult to be positively affirmative. Such a law could well be interpreted as fulfilling the constitutional mandate of “determining the modalities” by which freedom to resort to abortion is exercised; and it is not certain that the Conseil constitutionnel would rule that a two-week reduction of the legal time frame for abortion infringed the “guarantee” requirement. For the guarantee to be called into question, the law would probably have to (more) drastically restrict the deadline – to 6 weeks of pregnancy, for example, as is now the case in several American states, such as Florida or Iowa. It is also conceivable that a legislative reform calling into question the Social Security coverage of abortion would be struck down – although here again, it is impossible to determine with certainty the delisting threshold at which the Conseil constitutionnel would place the cursor: 90%? 50%? Less?

This is because the term “guarantee”, while playing a key role in the wording of the constitutional amendment, is also largely indeterminate; and because the Conseil constitutionnel is the sole master of the interpretation of the Constitution thus revised.

The indeterminacy of the word “guarantee” can, however, play in all directions. While it is not clear at what point a retreat from the current equilibrium of the legal regime of abortion would trigger constitutional censure, it is easy to imagine that the term will be the subject of mobilisation in the opposite direction. As it has been repeatedly pointed out in the 18 months since the first proposals for the constitutional amendment, it is often not the legal status of abortion that poses a problem in France, but rather effective access to abortion. Inequalities in access across the country are well-documented and have multiple causes: the policy of closing health care facilities in line with the managerial turn in public health policies, the under-funding of sexual health care centers, the devaluation of the act of abortion in the budgetary nomenclatures for the provision of health services, and so on. Under these conditions, the semantic leverage of “guaranteed freedom” will undoubtedly provide the basis for future proposals and legislative amendments aimed to reinforce effective access to abortion.

From this point of view, it must be conceded that, however historic the vote of 4 March 2024, the rationale underlying the amendment is not that of the consecration of an individual right. The sentence that was added to the Constitution was inserted in Article 34, and this is far from insignificant. Unlike many foreign constitutional texts of the same period (the German or the Italian ones, for example), the French Constitution of 1958 does not contain a section specifically devoted to the enunciation of fundamental rights. These are set out, primarily, in the preamble to the Constitution (which refers to the 1789 Declaration of the Rights of Man and of the Citizen, to all the social rights enshrined in the 1946 Constitution, and to the 2004 Charter of the Environment), as well as, subsidiarily, in various articles of the Constitution. However, a different choice was made with regards to abortion, since, by supplementing Article 34, the constitutional amendment is designed, first and foremost, to define the powers of the legislator.

The Constitution thus commands that it is for legislation – and only for legislation – to “determine the modalities” by which the freedom to resort to abortion is exercised. Admittedly, legislation cannot do anything in the area, precisely because the freedom to have an abortion is constitutionally “guaranteed”. But the enshrinement of this constitutionally “guaranteed freedom” only comes into play here as a second intention, as a consequence of the first gesture, which is indeed to define and require legislative competence. In that respect, it is not an individual right in the same way as, for instance, the right to “the free communication of ideas and of opinions” elevated by the 1789 Declaration, or the freedom to unionise, to strike, or the right to asylum elevated by the 1946 Preamble, or even, the right to be free from arbitrary detention (article 66 of the Constitution) or from capital punishment (article 66-1 of the Constitution).

On the evening of the final vote, the president of the National Assembly, Ms Braun-Pivet, celebrated the adoption of the constitutional amendment on social media by claiming that, by “a historic vote”, France had turned abortion into “a right”. The irony of the message cannot be lost to observers, given the energy that was put into securing that the amendment precisely avoid referring to a “right”, favoring the – allegedly less robust and thus more conducive to political compromise – term “liberty” instead, throughout the entire amendment process. Braun-Pivet’s words only testify to a very real discrepancy between the political discourse around the amendment and its actual legal wording. Only the future will endow this 25^th amendment to the 1958 Constitution with its full and precise scope and meaning. For now, it can be read simultaneously as a mere conjuring of drastic setbacks, or a crystallisation of the existing legal abortion regime which is the result of 50 years of successive legislative interventions, or even – in the words of Braun-Pivet and others – as the elevation of a full-blown “right” to abortion. It now falls on the legislator to act on this newly explicit competence – under the power of judicial review.

The author thanks AOC Media, where a longer version of this blogpost was published in French on March 7, 2024, for allowing this edited translation.

The US Supreme Court’s Dobbs decision of June 2022 created shockwaves through large parts of the world, sending millions of American women back to a bygone era as several 19^th century abortion laws were automatically reinstated following the overturning of the 1973 Roe v Wade ruling, which had read constitutional protection for women’s right to abortion in the federal Constitution. This event resounded all the more loudly that it echoed other comparable setbacks – in Poland, in Hungary and elsewhere.

The constitutionalisation of abortion had already been considered in 2017 and 2019 in France, but without success. The setback created by the Dobbs ruling thus provided an opportunity to be seized. In the National Assembly, a cross-party agreement enabled the adoption of a bill that would have included the following wording in Article 66-2 of the Constitution, immediately following the constitutional elevation of individual liberty: “the law guarantees the effectiveness of and equal access to the right to voluntary interruption of pregnancy.”

When the matter was turned to the Senate, the second chamber accepted the principle of amending the Constitution but adopted a much weaker language. It proposed to include in Article 34 of the Constitution, a provision that lists the domains of competence reserved to the legislator, the statement that “the law determines the conditions under which a woman’s freedom to terminate her pregnancy is exercised”. This version appeared to be so watered down when compared to the initial vote that it even raised the question of its usefulness. Indeed, what would have been the point of including in Article 34 of the Constitution that the law “determines the modalities” of abortion? This was a truism, for the modalities of abortion have always been determined by the legislator.

Thus began a process of compromise between the two versions of the proposed amendment. The final bill, presented by the Government on 12 December 2023, would revise Article 34 of the Constitution by defining abortion as a field of legislative competence and essentially reproduced the language adopted by the Senate – speaking, notably, of a ‘freedom’ rather than a ‘right’ to abortion. However, the word “guarantee” was borrowed from the initial version adopted by the National Assembly. Hence the version that was voted by the joint session on 4 March 2024: “the law determines the conditions under which the freedom guaranteed to a woman to have recourse to a voluntary interruption of pregnancy is exercised.” Everything thus hinges, more or less, on the word “guaranteed.” It is the term that allowed compromise, which makes it possible to distance the reservations, if not the criticism, generated by the wording of the February 2023 Senate bill.

The occasion was historic, and understandably so: enshrining abortion in a constitutional text in 2024 is no mean feat – neither symbolically, nor technically. Symbolically, this makes the French constitution the first one to guarantee the freedom to have recourse to an abortion. The global silence of constitutions on this matter reveals the invisibilisation of reproductive issues by the modern constitutional paradigm. The only three constitutions to contain an explicit reference to abortion are those of Kenya (Art. 26§4), Somalia (Art. 15§5), and Eswatini (Art. 15§5); and they do so precisely to prohibit it almost entirely. Some of the more recent constitutions enshrine the notion of reproductive rights (Bolivia for instance), or the right to choose to have children (Slovenia for instance), but without explicitly guaranteeing abortion (see generally). It is against this backdrop that the pioneering nature of the French amendment needs to be read. The inclusion of reproductive issues in the Constitution affirms their profoundly political nature (political communities depend, radically and existentially, on reproductive work for their own perpetuation) based on a notion of equal citizenship (procreative freedom conditions access to all the dimensions –civil, political, social– of citizenship).

The amendment also has technical legal value, by affording increased protection to the freedom to have an abortion. In Dobbs, the US Supreme Court ruled that the reasoning of Roe v Wade fifty years earlier was “egregiously wrong” and that, in fact, the Constitution had nothing to say about abortion – and therefore, that there was no constitutional protection for abortion. It is safe to say that the constitutional revision that has just been adopted in France conjures up such a setback: abortion has clearly become a constitutional issue; the Constitution requires that the law respects women’s “guaranteed freedom” to have an abortion. The exact scope of that guarantee is, however, cause for caution, as Part II of this blog explains.

The author thanks AOC Media, where a longer version of this blogpost was published in French on March 7, 2024, for allowing this edited translation.

The Problematic Avoidance of the Indirect Discrimination Claims

In Dakir, Belcacemi, and SAS v France, the Court was faced with the claim that bans on the burqa/niqab are a violation of Article 14 ECHR, which prohibits discrimination. At first sight, the claim that banning face covering is indirectly discriminatory against Muslim women seems to be a strong one. Additionally, it is clear that, in both France and Belgium, the bans on the niqab place those Muslim women who wish to wear a full veil, and only them, at a particular disadvantage (as has been shown by empirical studies by the OSJI on France or researchers from the Ghent Human Rights Center).

The European Court of Human Rights, however, did not accept this line of argument. It held that the concept of indirect discrimination can apply only if the policy or measure that is thought to produce disproportionately prejudicial effects on particular group “has no “objective and reasonable” justification, that is, if it does not pursue a “legitimate aim” or if there is not a “reasonable relationship of proportionality” between the means employed and the aim sought to be realized” (SAS v. France, §161). Based on that principle, the Court upholds the bans, precisely because it has previously found them to pursue “objective and reasonable justifications” in encouraging “living together”. This mode of reasoning can, however, be criticized as somewhat tautological and threatening to the very relevance of the concept of indirect discrimination in human rights law adjudication. It precludes any possibility of finding that a measure or policy that is justifiable vis-à-vis human rights standards in general has discriminatory effects. In other words, the very moment any given measure or policy appears to pursue a legitimate aim, all arguments pertaining to its actual discriminatory impact become void and irrelevant.

Such a legal state of affairs confirms the highly uncomfortable position in which European law is progressively cornering Muslim believers –and Muslim female believers in particular. Recent developments at the Court of Justice of the European Union (CJEU, Grand Chamber, 14 March 2017, Asma Bougnaoui & ADDH v. Micropole SA, C-188/15; and CJEU, Grand Chamber, 14 March 2017, Samira Achbita & Centrum voor gelijkheid van kansen en voor racismebestrijding v. G4S Secure Solutions NV, C-157/15) seem to have rested on the notion that not all restrictions on religious freedom are discriminatory. Combined with the Strasbourg approach to burqa bans, the situation currently is one in which bans on face covering can only amount to indirect discrimination where they also fail to pass the proportionality test in Article 9(2). This leaves observant female Muslims largely unprotected by Article 14.

Proportionality and Polities

Last but not least, the Court was faced with the issue of proportionality. The Belgian ban that was being challenged provided for two possible levels of sanction in case of violations: first, a fine and then (in case of repeated offences), a jail sentence of up to 7 days. The French law challenged in SAS did not provide for any form of imprisonment. This further step walked by the Belgian legislator did not unsettle the Strasbourg court. Strikingly, all the developments pertaining to the proportionality of the ban are infused with references to the necessary restraint of the Court, the Convention’s subsidiary role (§54), the respect that needs to be paid to national authorities who are the only ones vested with direct democratic legitimacy (§54) and the very wide margin of appreciation of the State that ensues (§59).

That the Court would have so little to say about a ban that may lead Muslim women of strict observance to jail is not only striking but untimely. Contemporary times seem to be marked by a high-water mark of measures that result in excluding groups of people from the public space –a policy trend that seems at odds with the very purpose of human rights law. In a remarkable transversal and comparative book on this particular policy trend, Swiss legal scholar Daniel Moeckli demonstrates that shifts pertaining to both the privatization of public space and increased regulation thereof in relation with the emergence of a “security society” have led countries such as the UK, the USA or Switzerland to enact numerous laws and regulations that tend to expunge the public space from all those who stain its image and appearance as a sanitized, safe and discomfort-free space. Although the book does not (regrettably) include face veil bans in the scope of the inquiry, it is very tempting to view them as an additional ramification of the daunting exclusionary paradigm that the author astutely describes as taking over the policies in various Western countries. This, in turn, only reinforces the need to rethink the meaning of face veil bans and other measures that rest on the exclusion of new categories of “undesirables” with respect to what they say and reveal of our contemporary polities.

The “living together” requirement as a new cornerstone for the European “model of society”

The “living together” argument had already been instrumental to the 2014 SAS v. France ruling; and here again, it is endorsed by the Court. Remarkably though, there is little to be found in the Court’s reasoning on the meaning and scope of this new legal concept. Be it in the 2014 or the 2017 rulings, the Court merely: (i) grants the Governments’ desire to safeguard the possibility of “living together” as akin to the “protection of the rights and freedoms of the others”, which Article 9(2) of the Convention provides to be a legitimate aim for the restriction of religious freedom; and (ii) agrees that the “living together” standard can be interpreted as threatened by practices such as face covering, in that the concealment of the face may indeed hinder social interaction(s). But the Court itself does not say any more –especially on the merits of the concept. The only real discussion is to be found in the excerpts of rulings of national courts that the European Court of Human Rights cites but does not necessarily endorse. In that respect, because the Belgian Constitutional Court’s decision of 6 December 2012 upholding the burqa ban is much more extensively argued that its French counterpart of 2010 (which is not very difficult, as the latter shockingly “solved” the constitutional issues related to a legislative ban on the burqa in six short paragraphs…), the “living together” standard appears much more substantial today than it did after the 2014 ruling. Several important details stem from the Belgian Constitutional Court’s elaboration, including:

• the “living together” requirements are the token of a particular “model of society” in which the individual is said to prevail over his/her cultural, philosophical or religious background;
• they express and encapsulate a “common heritage” of fundamental values such as: the right to life, the freedom of conscience, democracy, equality between the sexes and the separation between churches and State;
• they forbid the individual from concealing his/her face, as the face is indispensable to social life.

None of these specifications weaken the words of warning of dissenting judges Nussberger and Jäderblom in SAS. They expressed their criticism of a very vague and abstract notion that did not, in their view, meet the standards that are usually required from concepts that allow for restrictions on those rights expressly guaranteed by the ECHR. In fact, one could argue that the 2017 rulings only strengthen the dissenting judges’ arguments. The European Court seems to agree with the Belgian Constitutional Court’s contention that the requirements of “living together” are part and parcel of a “model of society” that States may wish to pursue – and may obviously do so while enjoying a very wide margin of appreciation, for the Court heavily leans on the fact that it is not for it to meddle with national determinations of the “model of society” that the state wishes to promote (Dakir v. Belgium, §54 and following).