I am a professor of constitutional and international law at George Mason University Scalia Law School, and also deeply involved in legal matters in Israel, where I head the international law department at Israel’s largest think tank, the Kohelet Policy Forum. I had occasion to study the Istanbul Convention closely when Israel signed the convention, and found it went far beyond other similar treaties in its scope and revolutionary ambitions. I helped author a legal analysis that helped lead to the government abandoning plans to join the Convention.
The Convention imposes a variety of obligations that could be constitutionally problematic for many democratic countries. Understanding the Polish Constitutional Court is currently considering the legality of the Convention, I have undertaken a detailed analysis of the provisions that might be most problematic. Two notes of caution: while I am a constitutional lawyer, I am certainly not an expert on the Polish Constitution or its interpretation. In what follows, no firm conclusions are drawn about the meaning of that instrument. Instead, I focus on the Convention, and the particular provisions that are most likely to raise constitutional concerns in many democratic systems. In certain cases I refer to particular provisions of the Polish Constitution which seem relevant, by way of illustration. Second, this Brief focusses on eight substantive articles of the Convention (3(c); 4(3), 8; 9; 12(1); 13; 14; and 40, along with several procedural articles relating to the monitoring mechanism and the impermissibility of reservations. These are the provisions that seem most likely to be problematic, but it does not represent a judgement that other provisions, not discussed here, raise no questions under particular constitutions, including Poland’s.
The Convention is not just a treaty, it is a crusade. As the Council of Europe proudly states, the treaty is “a manifesto, laying down a vision of society in which … gender equality is fully achieved.” Unlike perhaps any other such international agreement, the Istanbul Convention has the express goal of remaking society. The Convention proceeds from a thesis that physical violence against women is not caused by the individual depravity of the perpetrators, but ultimately by distant “root causes.” These root causes include all social attitudes and traditions about the roles of men and women that cast men and women in differential, or “stereotyped” roles.
Starting with these unproven assumptions, the drafters of the Convention came to a radical conclusion: to prevent violence against women, governments must not only punish such crimes, but fundamentally remake society, and in particular people’s views and beliefs about the roles of women and men broadly speaking, outside the context of violence[1]. The people whose attitudes and beliefs will have to be eradicated have no demonstrable causal connection with violence against women. Rather, the treaty uses violence as a Trojan horse for a much broader agenda of mandating the governmental reshaping of society.
The ideological nature of the Convention can be seen in the reactions to legal criticism of it. Official reports by the Council of Europe and its experts describe concerns or criticism about the scope or effects of the Convention as “attacks against the Convention,” “backlash,” and part of an “ideological war” intended to “relegate [] women to a subordinate status in all aspects of life.”
There is nothing like this in any international human rights treaty.
A. Articles 3 & 4: non-biological definitions of “men” and “women”, and the issues of gay marriage and “trans” status
Art. 3(c) establishes the notion of non-biological gender by defining “gender” as being not something physical, but rather a cluster of “socially constructed roles, behaviors, activities and attributes.” As the Council of Europe boasts, this is “the first international treaty to contain a definition of gender as a socially constructed category.”[2] This particular provision has generated some of the most intense controversy around the Convention. The gender issue has deterred many countries, such as Croatia, the Czech Republic, Hungary, Latvia, Lithuania, the Slovak Republic, and until very recently, Ukraine, from signing the Convention. It has, as the Court well knows, been the focus of opposition to the Convention in Poland[3].
The Convention’s supporters, and the CoE itself, have responded to concerns over the gender provisions by denouncing those raising the issue as waging an “ideological war” against the Convention. A CoE parliamentary report complains that “There are credible reports that priests and other Church representatives play an active role in this campaign, sometimes literally preaching against the Istanbul Convention.”[4] Those concerned about inadvertently importing “gender theory” into their domestic legal systems are, according to the Convention’s supporters, not simply wrong, but evil, engaged in a “deliberate attempt to sabotage …equality between women and men and the end of a patriarchal mentality which relegates women to a subordinate status in all aspects of life.”
All this rhetoric obscures what the Convention actually means and how it will likely be interpreted. The CoE’s own promotional leaflet states that as a result of the Convention, “it is now recognised that women and men are not only biologically female or male, but that there is also a socially constructed category of gender.”[5] (emphasis added).
The claim of the Convention’s apologists that Art. 3(c) does not require the recognition of any particular genders is narrowly true. But it does prevent the state from insisting on a biological definition of men and women. The Constitution of Poland recognizes only two genders, men and women. Moreover, the Polish Constitution apparently assumes a biological definition of men and women (see, for example Art. 68(3)’s reference to “pregnant women”). Moreover, the Polish Constitution in both Art. 18 (marriage) and Art. 33 (equal rights), treats men and women as distinct and separate biologically-based categories – not social constructs. Similarly, the Constitution refers to people who bear children as “mothers” (Art. 71).[6] Those biological definitions create a manifest conflict with Art. 3(c) of the Convention.
CoE apologists for the Convention state that it makes “no reference …to the legal recognition of same-sex marriage.” Again, this is narrowly true, but fundamentally misleading. At least some scope for same-sex marriage follows directly form Arts. 3(c) and 4(3) of the Convention, when read together. The latter provision bars discrimination on the grounds of “sex, gender... sexual orientation” as well as “gender identity.”[7] The previous section, Art. 4(2) demands equality and non-discrimination between men and women, which means the categories in 4(3) go beyond simply “men” and “women.”. When read together with Art. 4(3), which includes “gender identity” as an impermissible basis for discrimination, and Art. 3(c)’s definition of gender as a social construct, one could conclude that it would be prohibited to ban a marriage between a man, and another man who identifies as a woman. The reason Convention defenders can claim it does not affect same-sex marriage is that the union just described would not be regarded by them as a same-sex marriage, but rather one between a man and a trans-woman.
The Convention also raises issues regarding “trans” and other non-binary genders. It is true that Art. 3 recognizes does not explicitly acknowledge a third “trans” category.” Nonetheless, Art. 3’s language can provide a legal basis for the recognition of “trans” and other gender identities, because it defines gender as non-biological. Art. 3 could, for example, provide a basis for men who identify as women in their social roles and behaviors to claim protection and legal status as women. Such a “right” is absent from any prior human rights instrument.
In the Convention’s reports and publications designed for an internal audience, the pretense that it does not encompass “trans” rights is dropped entirely. For example, GREIVO monitoring body routinely treats as women groups such as “transgender and intersex women (LGBTI)”.sup>[8] GRIEVO’s repeated references to “LGBTI Women” makes clear that it interprets that convention as including “trans women” in the Convention definition of “women.” Indeed, a significant feminist scholar of the Convention admits that GRIEVO “tellingly” includes “trans women… in its evaluation of the Convention’s implementation” as evidence that the Convention, properly interpreted, applies to such persons.”[9] This seems to be the superior reading of the Convention. Whether Poland giving legal status to “trans” identities violates the Polish Constitution is, of course, a question for this tribunal.
B. Articles 8 & 9: special status for private organizations
Art. 8 and 9 requires states to provide money to and collaborate with NGOs working on the subject matter of the Convention. Art. 8 mandates government funding (“support”) for such organizations.[10] While it does not specify amounts, choosing to not fund such organizations would violate the Convention. Art. 9 also requires the state to make such groups “partners in multi-agency cooperation or in the implementation of comprehensive government policies.”[11] According to the CoE, this requires formal intermeshing of state bureaucracy with the relevant NGOs (“agencies and NGOs should not act alone, but work out protocols for co-operation.”).[12]
These provisions raise several potential constitutional issues. First, they require the government to directly cooperate with certain private organizations, raising potential problems about the separation of public and private functions. One could read the Art. 153(1) of the Polish Constitution as requiring government functions be performed exclusively by “professional” and “impartial” civil servants as precluding including ideological NGOs in the execution of government functions.”[13] Furthermore, Art. 9 creates a permanent budgetary commitment for the benefit of certain favored, and often ideological private organizations, raising questions about equity and neutrality. These groups will not have to work through the normal parliamentary system to ensure that the receive funds; such funding will be permanently mandated, with only the levels up for discussion.
C. Requiring the government to reshape social values
Art 12(1) – “Parties shall take the necessary measures to promote changes in the social and cultural patterns of behavior of women and men with a view to eradicating …customs, traditions and all other practices which are based on the idea of the inferiority of women or on stereotyped roles for women and men.”
This mandate for government-run social change is extraordinary for a international human rights treaty. This would be a basis for challenging a wide variety of governmental practices in entirely unrelated to VAW, such conscription or combat service, or even differential laws regulating indecent dress for men and women.[14] Even the non-admission of women to priesthood can be viewed as “stereotyped” custom. The vast reach of this provision poses so many risks they are hard to list or enumerate.
But the problematic nature of Art. 12(1) goes further. It affirmatively requires the government to “eradicate” national “customs and traditions.” This means a party to the Convention mast actively be involved in the beliefs and private and religious practices of its citizens. The far- reaching scope this provision can be inferred from a pamphlet published by the Council of Europe to promote the Convention in countries hesitant to adopt it. Addressing the concerns raised by Art. 12(1), the document observes that “Eliminating gender stereotypes does not mean overturning all tradition and customs… The aim must therefore be to deconstruct the stereotypes affecting women and men that are invoked to justify such harmful traditions.”[15] Clearly to comply with the Convention great deal of innocent, non-criminal thinking on the part of citizens must be re-programmed. While the convention does not specify the measures of “eradication,” Art. 12 is mandatory – governments must take measures calculated to be effective, even if coercive. As the Explantory Memorandum honestly states, governments must take “measures that are necessary to promote changes in mentality and attitudes” of the population.[16] Stopping short of such measures because of constitutional concerns would violate the plain language of the Convention.
It is hard to overstate the sweeping nature of what this prohibition includes, and broad nature of the norms and viewpoints it requires governments to abolish. Some observations by the Scientific Advisor to the Convention’s drafting committee (and likely candidate for GREVIO membership) illustrate Art. 12’s broad sweep. Prof. Christein Chinkin states that Art. 12 seeks nothing less than to “transform gender relations.”[17] Whether this is a worthy goal or not is a political question; but the convention’s mandate for the government to achieve not mere equality but “transformative equality” seems incompatible with the distinction between private and public spheres that underpins the Polish (and other Western democratic) constitutions.
Prof. Chinkin also explains that Art. 12 “should also encompass mandatory anti-racism training [for judicial officials] and an understanding of the white male privilege in the culture of law enforcement and the administration of justice.” It is hard to understand what this has to do with the already broad goals of Art. 12, but is indicative of the expansive understanding of the provision taken by its most distinguished and authoritative exponents.[18]
The vast mandate for remaking private citizens beliefs and attitudes seems to conflict with many provisions of Poland’s Constitution. Governmental measures to destroy attitudes and traditions seem to conflict with the principle of governmental “impartiality in matters of personal conviction … in relation to the outlooks of life” (Art. 25); “private and family life” (Art 47); freedom of religion (Art. 53); and expression (Art. 54). Indeed, the Convention declares certain “outlooks on life” and certain “convictions” held by many religious believers to be anathema and requires the state to undermine them. There may be some way to square the contradiction between these provisions in practice, but on the face of it, there is a manifest contradiction.
D. Art. 13-14: Requiring the government to provide ideological education
These provisions deal with “education” by the government regarding the values of the Convention for the purpose of molding “attitudes” and “convictions” of citizens “very early on in life.”[19] The central provision dealing with education, Art. 14(1), which relates to curricular materials in schools, is broad in its sweep. Indeed, the Scientific Advisor for the drafting of the Convention, Prof. Christine Chinkin, argues that Art. 14(1) mandates sex education in public schools.[20]
However, in this case, a careful reading of the Convention suggests the provisions regarding school education are optional, due the words “where appropriate.”[21] However, the remainder of the education/public communications provisions notably lack that “where appropriate” opt-out. Thus Art. 13 creates an “obligation” for the government to conduct public information/propaganda campaigns about issues such as “non-stereotyped gender roles,”[22] without any “appropriateness” limitation. Similarly, Art. 14(2) requires the government to conduct “non-stereotyped gender role” indoctrination in informal educational contexts, such as Scouting movements, camps, etc. Here again, the CoE’s explanatory memorandum refers to this as an “obligation,” with state opt-out.[23]
In short, the Convention requires the government to involve itself actively on one side of a highly contested moral, political, and religious debate regarding the nature and inherentness of gender differences. It cannot be ignored that these are issues that divide citizens of good will in many Western democracy, are issues that form part of the platforms of political parties, and the creeds of many faiths. The Convention would appear to require the government to intervene on one side of these debates, especially when it comes to influencing minors. Thus these provisions seem to raise questions about their compatibility with Art. 25(2), Art. 48(1), 53(1)-(3). Again, it is beyond our expertise to opine on the interpretation of Polish Constitutional provisions. But the manifest intent of the Istanbul Convention is to involve the State in shaping the views and convictions of the citizens.
E. Art. 40: Free speech and the banning of jokes
This provision requires parties to penalize “sexual harassment,” which would seem benign, as Poland, like other states, already does so. Yet as with many measures of the Convention, the definition given to “sexual harassment” is expansive and unusual, going beyond what current Polish law, and the law of many other Western states, would currently regard as “harassment.” The Convention defines the harassment that must be prohibited as including “unwanted verbal, non-verbal or physical conduct of a sexual nature with the purpose or effect of violating the dignity of a person.” This includes things such as “jokes, questions, remarks” as well as “facial expressions, hand movements or symbols.”[24] Crucially, the Convention’s conception of harassment is not limited to workplace harassment, nor is it limited to situations of vulnerable persons, or where the offender holds a position of trust or objective power over the victim.[25] It also includes “online sexual harassment.”[26]
Penalizing pure expression, such as “jokes” and “questions” may infringe on principles of free speech, depending on how broadly that concept is defined.[27] For example, in the United States, there is a recognized tension between constitutional guarantees of free speech and sexual harassment rules even in the employment context.[28] Sexual harassment laws in the U.S. have not been extended outside the employment context – as the Convention would require – in recognition of the constitutional difficulties this would raise.[29]
The Convention’s proponents admit it would require Poland to go far beyond its current law, which regulates harassment solely in the workplace context, to penalizing “jokes” and “facial expressions” in the privacy of the home, or even online.[30] It is of course the role of the Tribunal to determine whether such laws violate the Polish notions of free speech, and the “freedom of expression” protected by Art. 54(1) and the protection of “private and family life” of Art. 47.
One of the features that distinguishes the Convention from most other human rights treaties is its broad prohibition on reservations, which eliminates the standard mechanism for a country to join such treaties while ensuring fidelity to its constitutional values. Combined with the Conventions extraordinarily broad scope and its demand for governmental social engineering, this is a recipe for constitutional difficulties.
More generally, the particular likelihood of the Convention raising difficulties under domestic constitutions arises not solely from the particular substantive provisions discussed above, but from the interaction of several structural elements – the allocation of interpretive authority to an independent commission; the impermissibly of reservations; and the ideological, revolutionary nature of the Convention’s worldview. As with many treaties, the precise scope of its substantive requirements is in many cases unclear. Human rights treaties often use broad language because they do not represent a tit-for-tat negotiation between parties, but rather aspirational Some provisions of the Convention may be susceptible to both constitutional and unconstitutional interpretations vis a vis Polish law (or any other country’s).
A. GREIVO and Interpretive Authority Over the Convention
The Convention contains many ambiguous provisions, including the ones discussed above, and thus its constitutionality may depend on how it is interpreted. Thus as important as what the language of the Convention says is who has the authoritative power to interpret what it means. It is almost certain that the Convention’s answer to that is not “the Polish Constitutional Tribunal.” In traditional treaty practice, parties to a treaty each interpret its requirements for themselves, absent any specific provisions for binding arbitration or adjudication. However, in the past several decades, a new practice has emerged under human rights treaties. Such treaties routinely create an official monitoring or oversight body to which member states must report on their domestic practices. Those bodies, such as the Human Rights Committee created be the International Covenant on Civil and Political Rights (ICCPR), have asserted for themselves final interpretative authority over those treaties. These bodies’ claims to special interpretive authority, while not clearly rooted in the texts of the respective conventions, has been widely accepted by international courts and scholars.[31]
This context is crucial for understanding the role of the Istanbul Convention’s monitoring body, known as GRIEVO (“Group of experts on action against violence against women and domestic violence”). The Convention empowers GRIEVO to “monitor the implementation” of the treaty by states (Art. 66(1)). While GRIEVO is not explicitly given any authoritative interpretive role, one cannot pretend that international lawyers do not generally regard similarly-constituted monitoring bodies created by other human rights instruments as having authoritative say on the meaning of those respective treaties.[32]
Like other such monitoring mechanisms, GRIEVO is empowered to issue “general recommendations” about the meaning of the treaty, separate from its review of particular countries. Such “general recommendations” are also generally considered to be legally binding. Take, for example, the treaty closest in subject matter to the Convention, CEDAW, which creates a monitoring committee similar to GRIEVO.[33] According to international law scholars, the “general comments” of the CEDAW Committee are “authoritative,” and that states are obligated to interpret CEDAW in line with the general comments.[34] As one scholar has commented, “the organs can adopt decisions beyond the control of individual member states’ constitutional organs, with important legal and practical effects for these states” and with “undesirable incursions into state sovereignty.”
The history of interpretation of human rights treaties has been to “hook” countries with general treaty language, which is subsequently given ever-broader and more aggressive interpretation by the monitoring body. Reassurances that GREIVO has no final say will likely turn into an insistence that “by ratifying or acceding …, a State party also acknowledges the competence of the monitoring body of such treaty … to examine its periodic reports, play the role of super partes interpreter of the treaty.”[35]
It would be unreasonable to dismiss constitutional concerns in such a situation because a given provision of the Convention is arguably susceptible of a narrow interpretation. Here, the ideological nature of the Convention must be considered. As has been discussed above, the Convention is the product of a very particular social philosophy, that physical violence against women is necessarily the result of deep social “root causes” including basic social “stereotypes” about men and women. Constitutional concerns such as those raised in Poland are described by the Convention’s experts as an “attack on the progressive normative framework established by European nations,”[36] and as mere sexist “backlash.”[37] Given this, it is unlikely that GRIEVO would be concerned about interpreting the Convention to make it compatible with idiosyncrasies of Poland’s constitutional or other national values. Instead, they will likely interpret it in a “progressive,” i.e., increasingly broader fashion, that would likely raise the constitutional problems discussed above.
In the event that GRIEVO would adopt an interpretation of the Convention that would infringe on Poland’s Constitution, the country would face a conflict between Art. 9 of the Constitution and other provisions. It would also again find itself in the awkward diplomatic position of seeming to “reject” binding international in favor of domestic law. However, it would be in a much weaker position rejecting the Convention’s application in a particular case if the Constitutional Tribunal had previously held the Convention to be constitutional.
B. The Prohibition of Reservations
The Istanbul Convention prohibits almost all reservations, except to certain specifically enumerated provisions.[38] The few reservations that are permitted do not concern any of the constitutionally suspect provisions discussed in this brief, and thus cannot remedy constitutional problems arising from them.[39] The convention’s significant limitation on reservations eliminates a safeguard used in most multilateral treaties to allow countries with diverse constitutional features to join the same instrument. As is well know, Poland issues made a series of reservations, taking full advantage of those permitted by the Convention. Yet it also made an additional “declaration” that it “will apply the Convention in accordance with the principles and the provisions of the Constitution of the Republic of Poland.” To the extent that this means that the Convention would not be applied when inconsistent with Polish constitutional law, it amounts to a reservation – a modification of the legal effect of the agreement. As such, this reservation, while completely reasonable in motive, and of the kind often made by countries in joining human rights agreements, is nonetheless clearly prohibited by the Convention.
The reservations restriction is what distinguishes the Convention most distinguishes it from other subject-related treaties that Poland has joined, such as CEDAW This was a conscious design choice on the part of the drafters. With the exception of those provisions for which reservations are allowed, the drafters believed the remaining provisions must be accepted in all their details by member states.[40] They were well aware that limiting reservations can affect breadth of participation.. Even within Europe, there is considerable diversity among countries – Protestant and Catholic; monarchy and republic; post-Communist or not; north and south – and their constitutions vary as well. The Council of Europe, in drafting the Convention, took the risk that its restrictions on reservations would preclude broad participation. If Convention proves incompatible with the Polish Constitution, it is not through the inflexibility or extremism of the latter, but of the former.
Human rights treaties are typically broadly worded, yet designed to be joined by countries with a wide range of domestic constitutional, cultural, and political systems and constraints. A major part of the rationale of such treaties is to get broad, even universal adherence, to anchor the relative norms in international discourse, rather than to be specifically legally enforceable. It is unlikely that any treaty language would frictionlessly fit all these systems. This would be an intractable obstacle to human rights treaties seeking broad membership – were it not for the use of reservations.
Reservations are statements made by a country upon signing or ratifying a treaty that modify the legal effect of the relevant treaty.[41] In short, countries can use reservations to opt out of what they may regard as problematic provisions, while still becoming a state party to the treaty. The allowance of reservations is seen as crucial to human rights treaties achieve one of their major goals – membership by many countries. Since such treaties fundamentally depend on self- enforcement, their principal purpose is declarative. This can be achieved by winning a broader, if thinner, adherence to their overall purposes.
Scholars understand multilateral treaties to be a package of interrelated design choices. Primary among these is the “broader vs. deeper” tradeoff between universal ratification and binding commitment. Similarly, reservations are more often used when treaties have obligations that are substantively broad, and definitive (i.e., “shall” rather than “shall endeavor” language).[42]
Human rights treaties do not typically hope to achieve global compliance with all of their often- vague provisions; rather, they seek to achieve a global consensus on top-level norms. Moreover, unlike multilateral arms control or trade treaties, which typically do ban reservations, human rights treaties are not based on a logic of reciprocity among states. Rather, the govern states’ relations with their own populations. For example, one signatory would not observe the relevant rights among its population simply because anther one does, nor would a signatory violate its own citizens rights in “retaliation” for such violations by other member states.
As a result, every major human rights treaty allows reservations, and has reservations made to it.[43] The United States attaches a prepackaged set of “reservations, declarations and understandings” to treaties it joins, stating “That nothing in the Convention requires or authorizes legislation or other action by the United States of America prohibited by the Constitution of the United States as interpreted by the United States.”[44] Neither the importance nor subject matter of such treaties has prevented reservations.[45] The U.S. and others made reservations even to the Genocide Convention. And CEDAW, the closest subject-matter treaty to the Istanbul Convention, not only does not restrict reservations,[46] has the largest number of reservations of any treaty,[47] covering idiosyncratic national issues from religious practices (Israel) to monarchic succession (Monaco and numerous other constitutional monarchies).
The Istanbul Convention is highly unusual in that it almost entirely disallows reservations. This has in recent years become a feature of a growing proportion some Council of Europe human rights conventions.[48] This is apparently due to a conscious policy shift within the Council’s treaty office, which seeks to make more binding the increasingly reticulated obligations of its conventions.[49] Yet because of its explicit goal of engineering gender relations, Istanbul Convention is far broader in its subject matter and potential sweep than the limited number of Council conventions substantially restricting reservations. The conventions against child trafficking, and sexual abuse, or trafficking in humans or their organs were not written as “manifestos” with a particular “vision of society.” Thus the Istanbul Convention contains a unique combination of a broad, detailed, intrusive provisions combined with near-total inadmissibility of reservations.