Legal justifications for gender parity on the bench of the International Court of Justice: An argument for evolutive interpretation of Article 9 of the ICJ Statute

Abstract The UN General Assembly and the UN Security Council should amend their rules of procedure to create gender parity on the bench of the International Court of Justice. Only 3.7 per cent of all judges on the ICJ have been women. The UN Charter, ICJ Statute, and long-standing practice of the Court underscore the importance of representation, but the focus has been on geographical representation. Using the law of international organizations, combined with the law of treaty interpretation and international human rights law, this article argues that Article 9 of the ICJ Statute should be interpreted to include a requirement of gender parity. Established practice, subsequent practice, and the UN’s multi-decade gender parity in staffing policy establish an evolutive interpretation of what is required to fulfil equality at the UN and the ICJ. The nomination and election procedures for ICJ judges are sufficiently flexible to facilitate this interpretation.

gender in the 2020 ICJ election, 11 the UNSC and UNGA chose not to elect all three women candidates. Similarly, the 2017 and 2018 elections were missed opportunities, where men filled all five vacant spots, despite three of the nine nominees being women. 12 The UN Charter (Charter), 13 the ICJ Statute (Statute), 14 and the long-standing practice of the Court underscore the importance of representation, 15 but the focus has been on geographical representation. 16 This article assesses the possibility of expanding the well-established convention of equitable geographic distribution to include gender parity. It contends that the laws of established practice and subsequent practice have coalesced to require an updated interpretation of the Statute of the ICJ and the UN Charter, and that this evolutive interpretation mandates gender parity on the ICJ bench. Just as subsequent practice can modify treaty agreements for states, established practice of an international organization (IO)and in this case the gender parity practice of the UNcan modify the meaning of the terms of the IO's constituent instruments (Article 9 of the Statute and Article 8 of the Charter). Arguing that the obligation born of this evolutive interpretation is both an obligation of conduct and result 17 and falls on UN organs and likely UN member states, the article suggests amending the UNSC and UNGA rules of procedure to achieve this outcome.
Parity on judicial benches matters. Women's participation in the judiciary is key in part because '[t]he judiciary influences society at all levels'. 18 The ICJ impacts everything from environmental law, to human rights, to the rule of law, and remains uniquely important regarding its impact on general PIL as a field. 19 Scholars have published excellent work regarding gender equality at supranational courts. 20 Little has been written regarding gender parity on the ICJ bench, 21 11 'Gender must be considered in the upcoming election of judges to the International Court of Justice', joint statement of The Institute for African Women in Law and GQUAL, 30 October 2020, available at www.gqualcampaign.org/gender-mustbe-considered-in-the-upcoming-election-of-judges-to-the-international-court-of-justice/. and no parity arguments have been made relying on the combination of international human rights law (IHRL), the law of international organizations (IO law), and treaty interpretation. 2020 marked the UN's seventy-fifth anniversary, prompting reflection regarding whether the UN has fulfilled its promises. 22 The UN Secretary General (UNSG) Antonio Guterres utilized this anniversary to re-emphasize the UN's gender parity policy. 23 This article posits that via a combination of IO law and treaty interpretation, there are legal justifications for extending the UN's gender parity policy to the ICJ bench.
To build this argument, Section 2 of the article defines gender parity in the UN and ICJ context, explains the benefits of parity on judicial benches, and suggests that the best way to achieve parity is with a numerical, consequence backed rule. Section 3 analyses how the long-standing requirement of gender equality has evolved into gender parity in the UN context, and unpacks how the UN's decades-old gender parity policy in staffing can be understood as an evolutive interpretation of gender equality requirements in the Charter. On this basis, Section 4 builds a case that the previously examined UN parity practice amounts to binding IO law obligating UN organs and member states. Further, Section 4 links established practice to subsequent practice and examines the implications this has for interpretation of the ICJ Statute, leading to Section 5, which argues that gender parity can be read into Article 9 of the Statute via an evolutive interpretation. After these legal arguments have been made, Section 6 puts forward a proposal for how gender parity on the ICJ bench can be achieved in accordance with current ICJ judicial election procedures (Section 6.1), suggesting that the internal rules of both the UNSC and UNGA permit these organs to require gender parity in candidate lists and voting (Section 6.2). Section 6.3 considers how such changes to the rules of procedure would impact member states, and how they might be brought on board. Finally, Section 7 concludes that the UNSG, UNSC, UNGA, and member states all likely have obligations both of conduct and result to achieve gender parity on the ICJ bench. 24 The article ends with a call to translate its legal arguments into political persuasion.
2. The benefits of a rule mandating numerical gender parity Gender parity means 'an equal number of women and men'. 25 The UN Commission on the Status of Women interpreted it to mean no less than 50 percent women. 26 As the ICJ has 15 sitting judges, remedying its long history of inequality might begin with an eight women/seven men distribution, to be changed to an eight men/seven women distribution and then back again at a future time. Consideration should also be given to the representation of non-binary genders/genderdiverse persons and a protocol drawn up to address this.
Judges significantly impact international law's development and content. 27 Projects such as 'Feminist Judgments in International Law' imagine how law would differ if the people writing the judgments changed. 28 Judicial identity and diversity influence trial outcomes, 29 and the 22 T. Deen, 'After 75 Years, UN Claims 50:50 Gender Parity, But Falls Short of its Ultimate Goals', IPSNews, 18 September 2010, available at www.ipsnews.net/2020/09/75-years-un-claims-5050-gender-parity-falls-short-ultimate-goals/. 23 Administrative instruction, Temporary special measures for the achievement of gender parity, ST/AI/2020/5, 6 August 2020, available at www.undocs.org/ST/AI/2020/5. 24 Hungary v. Slovakia (Gabčikovo-Nagymaros), supra note 17, 7, at 77, para. 135; ILC Draft Articles, supra note 17. different contexts, perspectives, and approaches judges bring to the bench 30 impact the legal content of their decisions. 31 While a gender-balanced judiciary is not a panacea for all problems, 32 the benefits of parity on benches are numerous. Positives include increased legitimacy, representation, public confidence, access to justice, fairness, good governance, equality, rule of law, sustainable development, and much more. 33 It is no wonder, then, that the call for parity on international benches is decades old. 34 This article assesses the best options for achieving gender parity on the ICJ bench. After reviewing practice at other supranational courts, it concludes that a consequence backed rule to achieve numerical parity is the most certain path. The ACtHPR did not achieve parity until it implemented a penalty for failing to do so. 35 The Protocol to the African Charter called for 'adequate gender representation in [the] nomination process' 36 and that 'in the election of the judges, the Assembly shall ensure there is adequate gender representation'. 37 Yet, law alone did not result in this outcome, and neither did repeated requests from the court's Office of Legal Counsel asking that states nominate women judicial candidates. 38 The Office had to introduce a punitive measure disqualifying states that did not submit at least one woman candidate for parity to be achieved. 39 Requiring precise numerical parity and backing failure to achieve this with a penalty provides a guarantee that 'softer' rules do not. Grossman noted that '[f]or [international] courts where states were required by statute to take sex into account when nominating or voting for judges, a higher percentage of women sat on the bench' as compared to courts that did not have such a rule. 40 These results, while improved, were far from parity, totalling 32 per cent women judges at courts with 'gender balance' rules compared to 15 per cent women judges for tribunals without such 30 See, e.g., J. J. Dawuni 42 The absence of a precise numerical requirement 43 in this 'fair representation' rule has meant that at times that the percentage of women has been as high as 60 per cent and once there was even an all-women bench. 44 However, consecutively between 2015 and 2018, women's representation dropped to almost a third of the overall number of judges, 45 reminding one that without firm floors, '[t]here is no "acquis", i.e. the advancement of women may at any time fall behind existing achievements'. 46 For the ECtHR, a 2004 resolution stating that the Parliamentary Assembly that elects the judges would no longer 'consider lists of candidates where : : : the list does not include at least one candidate of each sex' 47 has resulted in at most 40 per cent empanelment of women. 48 The resolution was weakened via an ECtHR advisory opinion, 49 and presently women's representation is 34 per cent. 50 These examples suggest that a 50/50 penalty-backed parity rule would be beneficial for achieving and maintaining numerical gender parity on the ICJ bench. This article argues that achieving precise numerical parity is necessary to fulfil the right to gender equality. To support this claim, the next section builds a case that, in the UN context, gender equality has evolved to require gender parity.
3. How gender equality has evolved to require gender parity at the UN

Gender parity as an apex interpretation of gender equality
Gender equality is a long-established legal right, 51 and gender parity is a more recent interpretation of this right. 52 Gender equality and the related prohibition on gender-based discrimination 41 Ibid. The data on benches with gender balance rules compared the 2015 benches of the ICC, ECtHR, ACHPR, and the ad litem benches of the  43 See ICC, 'Procedure for the nomination and election of judges, the Prosecutor and Deputy Prosecutors of the International Criminal Court', ICC-ASP/3/Res.6, available at asp.icc-cpi.int/iccdocs/asp_docs/Resolutions/ICC-ASP-3-Res. 6-CONSOLIDATED-ENG.pdf, at A.2, A.6(c), A.11, A.12, B.20(b) and (c). No precise numbers are required and it is impossible to predict numerical outcomes on the basis of these rules. are represented in every major human rights treaty 53 and are international rules with customary status. 54 Some treaties address gender equality as a non-derogable right 55 that states have a 'mandatory and immediate obligation' to fulfil. 56 Such obligations would apply equally to the UN and its organs 57 and prohibit member states and the UN from discrimination based on sex. 58 Gender parity arises from gender equality in part because over time the prohibition on sex discrimination under international law, a negative obligation, has evolved to require substantive equality under international law, 59 a positive obligation. In other words, an obligation of conduct (not to discriminate) has evolved into an obligation of result (ensure gender parity). Human rights, including those in the Charter, were designed to evolve. 60 There are several legal interpretations of equality: formal, substantive, and transformative. 'Equality before the law' has historically meant formal equality, but IHRL has long emphasized the need to achieve substantive equality. 61 The Committee on Economic, Social and Cultural Rights stated that securing equality entails achieving equality in fact and eliminating indirect discrimination in law, policies, and/or practices. 62 Duty bearers, whether states or IOs, are obligated to positively fulfil these rights. 63 This interpretation of equality is bolstered by transformative equality, 'which sees full and genuine equality as likely to be achieved only when the social structures of hierarchy and dominance based on sex and gender are transformed'. 64 Transformative equality is an appropriate doctrine through which to approach parity on the bench because it focuses on including historically excluded groups and requires structural change to achieve this inclusion. 65 Gender parity on the ICJ bench corresponds to transformative equality's aims, such as overcoming cycles of disadvantage and promoting social and political inclusion and participation. 66 Because transformative 53 See, e.g., Charter, supra note 13, Preamble para. 2; Art 5, UDHR, supra note 51, at Preamble para. 5, Arts. 7, 21, 23(2); ICESCR, supra note 51, Art. 3; Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW), Art. 1.  equality represents cutting edge IHRL, some might counter that it is lex ferenda. However, the latest legal developments in human rights, such as the Convention on the Rights of Persons with Disabilities, which promotes transformative equality in almost every article, communicates that what was once a novel understanding of equality is better understood as the standard interpretation. 67 State support for this treatythere are currently 182 states partiesunderscores that transformative equality is now accepted as integral to IHRL. 68 That equality can result from parity requirements is supported by what IHRL calls 'special' or 'specific measures,' and domestic law calls quotas. Because formal equality means 'the state should not give preference to any one group and that people should be treated exclusively on their individual merits and regardless of group membership', 69 several treaties create an exceptional basis for such preferences in order to achieve substantive equality. 70 The CEDAW Committee's General Recommendation No. 25 notes that special measures should 'accelerate the equal participation of women in the political, economic, social, cultural, civil or any other field'. 71 For decades, the UNSG has used special measures in furthering its gender parity staffing policy. 72 The UN's internal justice system has confirmed through multiple decisions that special measures creating gender parity in UN staffing are valid. 73 Gender equality rights extend to gender on the bench. 74 International law enshrines women's rights to participate equally in public life including public service. 75 Member states and the UN must prevent both direct and indirect discrimination as a way to ensure greater gender diversity on the bench, 76 and 'take all appropriate measures' to ensure gender-equal representation at the international level, 77 including at courts. 78  participation in the international judiciary. 79 This is because substantive equality requires both representation and participation, 80 in part because courts are both political and representative. 81 In addition to aspiring to impartiality, rule of law, and judicial independence, courts are also political agencies where judges might act as political agents. 82 This is at least partly true at the ICJ, where voting patterns can map to a judge's national origin and/or favour the state that nominated the judge, 83 and where judicial elections can be a political process, 84 albeit one moderated by institutions and procedures intended to depoliticize. 85 Structural bias in international law and at the ICJ is well documented. 86 The underrepresentation of women in the international judiciary contributes to this and 'legitimate[s] the unequal position of women around the world rather than challeng[ing] it'. 87 Thus, gender parity on judicial benches is both a fulfilment of representation and participation in and of itself, and a means for securing more of it based on the decisionmaking patterns of women judges. 88

The UN's interpretation of gender parity as fulfilling gender equality in the Charter
The UN's commitment to gender parity in staffing is decades old, with early indications dating to 1970, 89 and the initial target year for achieving gender parity in UN staffing set at 2000. 90 The UNSG called parity 'fundamentally a right' 91 and stated that '[g]ender parity is : : : a crucial first step to orienting the system more strongly to deliver on gender equality'. 92 The UN Working Group has described parity 'as the ultimate measure of equality'. 93 In 2020 the UNSG 79 CEDAW Committee, supra note 54, para 5. Women are not a monolithic group and decide diversely. Additionally, not all women are feminists or concerned with gender equality, and like other genders they face structural constraints and bias in their decision making. However, the literature cited in this article suggests that women are more likely to decide judgments in favour of gender equality claims as compared to male judges (e.g., Boyd, supra note 29, at 78).  System-Wide Strategy on Gender Parity, United Nations (Strategy), 13 September 2017, available at www.un.int/sites/ www.un.int/files/Permanent%20Missions/delegate/17-00102b_gender_strategy_report_13_sept_2017.pdf. 91 Ibid. 92 Ibid. 93 HRC, supra note 52. reemphasized the urgent need to achieve gender parity in UN staffing. 94 The additional examples presented infra in this section suggest that, in the UN context, gender equality cannot be fulfilled without gender parity.
The UN's gender parity commitment is articulated in numerous ways, including the Nairobi Forward-looking Strategies of 1985, 95 the Beijing Platform, 96 various UNSC Resolutions calling for women's equal participation in decision-making, 97 and UNGA Resolutions such as 33/143 calling upon member states to 'assist the United Nations' in increasing the proportion of women employees 'by nominating more women candidates' to positions. 98 Parity practice is clear in the Human Rights Council's Consultative Group's guidelines on gender parity in international appointments, 99 and the creation of the Working Group on the issue of Discrimination against Women in Law and Practice. 100 The UN's parity strategy document states that parity is 'an imperative requested by Member States', 101 reflecting both state practice in favour of gender parity and the links between customary international law and the customary law of IOs. 102 Further, there is evidence that parity should apply to UN courts and tribunals and is compatible with geographical distribution. Ebrahim-Carstens notes that the UN's internal justice systemthe United Nations Dispute Tribunal (UNDT) and the United Nations Appeals Tribunal (UNAT)has always had a significant women's presence on the bench, and even at times a majority of women. 103 Both Tribunals' Statutes state that the appointment of judges shall take both geographical representation and gender balance into account. 104 Since 2015 both the UNSG and the President of the UNGA instruct member states to consider 'geographical distribution and gender balance' when appointing UNDT/UNAT judges. 105 The UN's internal administrative tribunals are significantly distinct from the ICJ, including that the administrative judges are appointed by the UNGA and ICJ judges are elected. The ICJ is of much greater political significance to states than the UNDT/UNAT. Further, gender 'balance' is not the same as parity and often results in less than 50 per cent empanelment of women. Despite this, there are several takeaways from the UNDT/UNAT example relevant to parity at the ICJ. The administrative tribunals demonstrate that there is precedent for both the UNSG and UNGA, key players in the ICJ judicial nomination and election process, to call for a gender balanceand to interpret this as requiring numerical parityin judicial appointments within the UN system. Similarly, it would not be unusual for the UNSG and the UNGA to instruct member states to take gender balance into account when nominating judges, and for member states to comply. And, the UNDT/UNAT example demonstrates that geographical distribution can coexist 94 Administrative instruction, supra note 23. with gender balance. In a nutshell, internal UN practice supports not only gender parity in staffing but gender parity in judicial appointments.
The UN's System-Wide Strategy on Gender Parity does not expressly mention ICJ judges. This is important, because a staffing policy is distinct from the political elections that empanel ICJ judges. How an internal administrative policy could create binding rules that apply to elections is the subject of Sections 5-7. For the purposes of this section, it is interesting to consider that, based solely on a fair reading of the plain text of the staffing policy, the position of ICJ Judges might be classed as a senior level of leadership within the UN. This interpretation can be made on the basis that 'the Strategy is intended to have system-wide application'. 106 The strategy stresses the need to look at the so-called 'top' positions because 'there is an inverse relationship across the [UN] system between seniority and women's representation-the higher the grade, the larger the gap in gender parity'. 107 It states that ICJ staffing as a whole has achieved more than 50 per cent employment of women, 108 and yet its bench lags behind.
The UN's System-Wide Strategy on Gender Parity also sheds light on the relationship between equitable geographic representation and gender parity. The UN's staffing policy holds geographic diversity and gender equality as 'parallel goals' that 'should be mutually reinforcing' and enacted 'simultaneously'. 109 Further, 'geographic representation cannot be used as an excuse not to achieve gender parity'. 110 Section 7 examines the ICJ's long-standing practice of equitable geographical distribution. The UN's parity policy's coupling of these targets indicates their equal importance. It also implies the ICJ's geographical representation achievements are not enough to fulfil equality under the Charter. A UNSG temporary special measure from 1996 allowing exceptions to geographical parity requirements in favour of gender parity in hiring demonstrates that, where there is tension between the two objectives, at least in some cases the UN's parity policy prioritizes increasing women's employment. 111 Viewed holistically, the UN's gender parity policy demonstrates its intention to evolve the meaning of gender equality under the Charter. 112 While IOs are not parties to their constituting documents, 'it is generally accepted that constituent instruments may be a source of international legal obligations for organisations.' 113 If the Charter is interpreted pursuant to the familiar tenets of treaty interpretation codified in Article 31 of the Vienna Convention on the Law of Treaties (VCLT), there is a case that the Charter contains obligations to uphold gender equality and therefore parity. 114 The strictest reading of the Charter regarding parity would result in obligations for the UNSG, and consequently all organs. 115 This would affect the way the UNGA/UNSC handle ICJ nominations and elections. These impacts are the focus of Section 8.
An evolutive interpretation of Article 8 of the Charter would entail that 'all United Nations organs should be obliged to respect Art. 8'. 116 The Charter has rarely been amended, but this 106 Strategy, supra note 90, at 3 (original emphasis). 107 Ibid., at 2. 108 Ibid., at 5. 109 Ibid., at 3. 110 Ibid. has not stopped significant modifications, 117 aided by the Charter's 'open texture'. 118 Article 8 implicates staffing at the ICJ in particular, as equal participation should occur 'in principal and subsidiary organs'. 119 Building on the fact that decades of UN practice demonstrate the existence of a belief in gender parity in staffing as a legal requirement, the next section analyses how the law of established practice transforms these repeated actions into binding rules of both conduct and result for both UN organs and UN member states.

So-called established practice
This section examines the so-called established practice, or institutional practice, 120 of IOs in light of gender parity on the ICJ bench. Established practice should be distinguished from subsequent practice, with the former pertaining to the organization and the latter to states parties and/or member states. 121 While established practice is less frequently discussed than subsequent practice, 122 it is wellsettled that established practice falls within the rules of an IO. In Personal Work, the Permanent Court of International Justice recognized that reference to international practice can help elucidate rules, characterizing resort to practice for ascertaining rules as 'not unusual'. 123 This early reliance on IO practice for rule ascertainment has been expanded and formalized. Multiple treaties reference IO's internal rules, 124 indicating that established practice of the organization is envisaged as falling within the rules of the organization. 125 The 2011 Draft Articles on the Responsibility of International Organizations 126 reflect an understanding of IO established practice as a source of binding rules for the IO and potentially its member states.
Established practice is a binding, quasi-customary law of an IO, based largely on the organization's secondary law. 127 It can be understood as a third source of IO law, potentially modifying the IO's constituent instruments. 128 The relevance of established practice as a source of binding law is particularly important for the argument infra regarding an evolutive interpretation of Article 9 of the Statute. 117 Established practice likely comprises 'the combined effect of numerous consistent acts'. 129 In The Wall Advisory Opinion, the ICJ considered a wide range of organ practice to find established practice. 130 An IO's established practice must not be 'uncertain or disputed' if it is considered binding. 131 At the same time, established practice should be viewed as evolving rather than static. 132 Section 3.2 of this article outlines the UN's numerous, consistent parity acts and a wide range of organ practice. These acts have continuously increased over time. As noted previously, the UN describes its gender parity policy as fulfilling the Charter's purpose. This is important for understanding established practice regarding parity as a binding IO rule: if the UN asserts its action is taken to fulfil its Charter, this is presumed to be correct, 133 particularly when the UN is acting to fulfil the competence of its staff. 134 On this basis, it is likely that, via the extensive pattern of practice coupled with prolonged acquiescence from member states, the parity in staffing policy has coalesced into 'Charter law'. 135 The duration of the parity policy aids this conclusion: the original target for gender parity staffing was 2000, member states supported this, and the policy has expanded in the last 20 years. Enduring state acceptance of IO practice can create binding obligations for states arising from this practice. 136 Thus, there is a case to be made that the UN gender parity policy is a customary rule for the organization based on established practice. This means the UN parity policy, even if not originally intended to do so, could bind member states and UN organs. While the Charter envisaged gender equality and did not discuss parity, the UN's established practice demonstrates that, in the context of the UN, gender equality now requires gender parity. This could impact the ICJ bench in part because established practice's scope is broad: 'it has the potential to at least add substantive rules to the law of the organization, which have not been included into the constituent instruments.' 137 Established practice is lex specialis, making internal IO rules preeminent when interpreting the IO's constituent instruments. 138 This is important for those who may wish to rebut such a rule, as it will create an additional challenge. Quayle notes that the ICJ has imbued established practice with strength and staying power, requiring new institutional practice to displace an existing practice-based rule. 139 IO established practice may generate CIL, 140 and may also modify how subsequent practice is understood. 141 VCLT Article 5 does not contain or define the words 'established practice' nor explicitly state that such practice could displace subsequent practice or modify a treaty's terms that pre-date the 129 Peters, supra note 121, at 630. practice. However, The Wall suggests such an application. 142 The next section examines how established practice and subsequent practice interact.

Subsequent practice through the lens of established practice
Interpreting the Charter requires considering member states' subsequent practice and evidence of UN established practice. Considering such evidence means that how Charter provisions are interpreted may evolve over time. Subsequent practice is a time-honored tool for treaty modification. 143 It is vital in the context of the Charter, which is unable to change dynamically owing to its elaborate internal amendment requirements. 144 Such practice can inform the Charter's application by circumventing international politics and power struggles between member states at the voting stages in the UNGA and UNSC. 145 In the Wall Opinion, the ICJ confirmed that the interpretation of a treaty could evolve over time by the practice of the IO formed under the treaty's provisions. In this case, the ICJ noted that Article 12 of the Charter was initially interpreted as precluding the UNGA from considering matters while the UNSC was still dealing with them. 146 However, Article 12's meaning evolved based on the subsequent practice of the UNGA, which repeatedly adopted recommendations for matters while these appeared on the Council's agenda. 147 The ICJ considered that the 'accepted practice' of the UNGA had 'evolved' and was consistent with Article 12(1), 148 allowing the UNGA to refer the Palestinian matter to the Court for its advisory opinion. 149 Organ practice stood in for the practice of states parties regarding evolutive interpretation.
A significant proportion of states parties must agree to the change for subsequent practice to modify the meaning of a treaty provision, 150 creating a higher threshold of state practice than is needed to find established practice. 151 It would be challenging for this article to quantify all UN member state positions on the UN's gender parity staffing policy, gender parity on judicial benches, or gender parity on the bench of the ICJ. Less challenging is noting that member states have endorsed or at least allowed the UN parity policy over several decades. That the UN's gender parity policy originated in the 1980s, and that the original target was 2000, are relevant in assessing the strength of the rules of the organization and member state reactions. Absence of member state opposition to 'consistent institutional practice' implies acquiescence to it and can substitute the need for agreement of all parties under Article 31 VCLT. 152 VCLT Article 5 makes clear that an IO's rules can modify and precede general rules, including Article 31(3)(b). 153 This means that established practice could change the way one weighs subsequent practice when considering the interpretation of an IO's constituent instruments. 154 142 Academics have debated whether subsequent practice can modify a constituent treaty rather than simply provide evidence that is relevant to interpreting or applying its provisions. 155 ICJ jurisprudence implies that organ practice can have law-creating effect even without a high threshold of member state support/with strong member state opposition. 156 Additionally, when interpreting subsequent practice vis-à-vis established practice, the balance changes: 'the established practice of the UN could have created a rule of the organization with the content that its subsequent practice does not strictly require the agreement of all the Member States'. 157 Some argue that subsequent practice is unsuitable for changing the composition of UN organs or the Charter's fundamental underpinnings. 158 It would be difficult to claim that gender parity changes the Charter's nature when the policy is 'reflective of core values that are as old as the Organization itself'. 159 This is the case even if one asserts that the parity policy is limited to the UNSG and thus only one UN organ, because the ICJ has confirmed that one organ's practice can establish agreement of parties. 160 Beyond the UNSG, the jurisprudence of the UNDT/UNAT links the parity policy to Article 8 in the Charter. 161 Ultimately, the combined established and subsequent practice impacts how gender parity can be read into the ICJ Statute. The next section applies the UN's established practice on gender parity to an evolutive interpretation of Article 9 of the ICJ Statute.

An evolutive interpretation of Article 9 of the ICJ Statute
If a term is deemed generic in nature and particularly if operation of the treaty in question is not time-bounded, the parties may be presumed to have intended the meaning of the relevant term to evolve over time. 162 The text of Article 9 is broad, flexible, and suggestive of the possibility for evolutive interpretation, as does its drafting history. 163 Article 9 reads: At every election, the electors shall bear in mind not only that the persons to be elected should individually possess the qualifications required, but also that in the body as a whole the representation of the main forms of civilization and of the principal legal systems of the world should be assured.
Charter is significant, expressed in both the instrument's preamble and substantive terms. These clauses indicate that the UN as a whole and the institutions that comprise it ought not only to espouse certain values but also to embody them. This is reinforced by Article 8 of the Charter: 'The United Nations shall place no restrictions on the eligibility of men and women to participate in any capacity and under conditions of equality in its principal and subsidiary organs.' 167 As Section 4 articulates, established practice of the UN indicates that gender parity is necessary to fulfil Article 8 of the Charter and that this is an obligation of both conduct and result. Thus, Article 9 is subject to the 'Charter law' developed by UN gender parity policies and must be interpreted in accordance with it.
Article 9 of the Statute must be read 'in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose'. 168 The key phrase is 'the representation of the main forms of civilization : : : of the world'. Regarding gender parity, one should question what was intended by the words 'representation' and 'main forms of civilisation', andmore importantlytheir meaning today. The Oxford English Dictionary states that representation refers to '[t]he fact or process of standing for, or in the place of, a person, group, institution, etc., esp. with the right or authority to speak or act on behalf of these'. This is the same meaning often given in political science and political sociology. 169 The question is, who may legitimately provide such representation? Arguably, this is someone who should be capable of 'standing in the place of' the people comprised from the 'main forms of civilisation'.
The term 'civilization' is fraught, as is its colonial history in international law. 170 Tzouvala rightly observes that 'civilization' is not 'a unitary legal concept lending itself to conclusive definition but : : : a mode of international legal argumentation'. 171 Yet, while many things are open to debate, this demographic fact is not. Rather, given women's historical exclusion from international law, focusing on women's global prevalence as a basis for representation on the ICJ bench is aligned with transformative equality's call to reconfigure institutions in order to re-centre the marginalized. A fair reading is that Article 9 cannot be fulfilled without gender parity, because there cannot be a representation of the main forms of civilization if 50 per cent of the members of each civilization are absent from the bench. Thus, the existence of a substantial gender imbalance on the ICJ's bench suggests that the pure requirement of representation is not being met. 172 This article proposes expanding the well-established convention of geographical distribution used to ensure diverse representation on the ICJ bench to include gender parity. 173 The Charter's text is suggestive that both gender equality and geographical representation should be considered together, 174 and the UN's parity policy promotes this interpretation. Originally, the reference to the 'main forms of civilization' was a means of allaying Great Power fears of their possible future exclusion from the bench. 175 The wording was left sufficiently open to permit smaller states to envisage their equal involvement, while the Great Powers saw Article 9 as an exception to the principle of state equality. 167 See Grossman, supra note 3, at 87. Today, equitable geographical representation is tied to the UN's system of regional groups. 176 The regional groups are another example of an extra-textual system that has been grafted onto vague Charter provisions. The history of regional groups, and how they evolved from an informal system to a formalized, powerful organizing principle that affects the composition of all of the most important UN offices, is relevant to gender parity on the ICJ bench. It evidences institutional flexibility at the UN, informed at least in part by legal norms, and in part by balance of power considerations. Article 9's wording is similarly vague, contrasting with the Rome Statute's Article 36(8)(a)(ii), which expressly requires 'equitable geographical representation' for ICC judges. 177 The convention of equitable geographical representation at the ICJ and the UN demonstrates established practice of reading in extra-textual representation requirements, providing a legal precedent for adding gender parity to the notion of representation in ICJ judicial nominations and elections.
Analysing the practice of equitable geographical distribution within the European Court of Justice (ECJ), Kenney examined how geographical parity in practice underscores why states historically have insisted on such representation. She asks, If one need not "stand for" in order to "act for" in Pitkin's definition-that is, if men can represent women and Italians can decide cases for Germans-why are Member States so determined to have their own representatives on the ECJ? 178 In Pitkin's framework, 179 in representation through gender parity women judges will be 'standing for' women rather than 'acting for' women, the latter referring to how elected legislators might 'act for' their constituents by directly representing their interest. Kenney concludes that states feel that equitable geographical representation is necessary for protecting their interests, and that this representation ensures compliance with the court's decisions by imbuing them with legitimacy. 180 This logic can be extended to gender parity on the bench. Even setting aside the substantive decision-making of women versus men judges discussed in Sections 2-3, women have no less a claim than member states of various IOs to equal representation on the basis of protecting their interests and ensuring the legitimacy of judicial decisions. Given this, the next section outlines a proposal for how to secure gender parity on the ICJ bench.

A proposal for achieving gender parity on the ICJ Bench
There are multiple ways that gender parity on the ICJ bench could be achieved. This article proposes practical pathways to secure this outcome. The subsequent sections consider how existing nomination and election procedures at the ICJ could incorporate gender parity; suggests changing the UNGA and UNSC rules of procedure; and addresses how member states might be brought on board with the proposal.

Incorporating gender parity into the ICJ's existing nomination and election procedures
The nomination and election procedures for ICJ judges are general and flexible. As with geographical parity, the Statute's broad stipulations allow for adopted conventions. This is a good thing for gender parity: the process for modifying the Statute is arduous, requiring an amendment to be approved by a two-thirds majority of the UNGA and ratified by two-thirds of the members of 176 Where can I get information about Regional Groups?, UN Library, 26 April 2018, available at ask.un.org/faq/14521; I. Winkelmann, 'Regional Groups In The UN', in H. Volger (ed.), A Concise Encyclopedia of the United Nations (2009), at 592-3. 177 Rome Statute on the International Criminal Court, 2187 UNTS 90, Art. 36(8)(a)(iii) (1998). 178 Kenney, supra note 3, at 131-2. 179 Pitkin, supra note 169. 180 Ibid.
the UN, including all five permanent members of the UNSC. 181 To date, the ICJ Statute has never been amended. 182 Therefore, an alternative model might be sought.
National groups at the Permanent Court of Arbitration (PCA) are key players in the nomination process. 183 Articles 4 and 5 of the ICJ Statute state that the UNSG initiates nominations by requesting national groups to nominate between two and four candidates to the ICJ bench. At this stage, the UNSG could encourage gender-equal submissions in the same language as is done for the internal UN justice system. The Statute provides that in preparing their shortlists a state should 'consult its highest court of justice, its legal faculties and schools of law, and its national academies and national sections of international academies devoted to the study of law'. 184 The Statute offers no guidelines or best practices and is silent on procedure at the municipal level, although academic writing sheds light into some national group consultations. 185 This absence of a formally codified procedure may facilitate new rules enshrining gender parity. It is worth noting that the research of Mackenzie et al. 186 and in person conversations between national group members and this article's author suggest that in some cases, candidates are picked by governments. 187 Still, national groups remain an important focal point for securing gender parity on the ICJ bench.
After national groups nominate candidates, the UNGA and the UNSC vote separately on the candidates. 188 Candidates receiving an absolute majority of votes from both organs are elected. 189 Under Article 9, electors should keep representation in mind as they vote. Article 9 thus presents an opening for creating a gender-parity rule that could guide and constrain UNGA and UNSC voting. Such a rule might be achieved through a UNGA resolution, similar to the ECtHR rule established by the Parliamentary Assembly. Civil society organizations have lobbied for this option. 190 However, UNGA resolutions are non-binding on states, and UNSC resolutions are only binding on states via Chapter VII, which is not relevant to gender parity on the ICJ bench. An alternative is that each organ could modify its internal rules to require both gender parity in candidate lists and gender parity in voting. The next subsection discusses why the UNSG/UNGA might be obligated to do so and how this can be done.
6.2 Changing UNGA and UNSC rules of procedure to create gender parity on the ICJ bench Because the UNGA and UNSC cast the final votes for ICJ judges, modifying the voting obligations of the UNGA and UNSC can create judicial gender parity. Amending the UNGA and UNSC rules of procedure is one avenue through which this could be achieved. Both the UNGA and UNSC have the power to make such changes, and there is precedent for them to do so.
Article 21 of the Charter gives the UNGA authority to adopt its own rules of procedure and the UNGA has amended these numerous times throughout the years. 191 Article 30 of the Charter empowers the UNSC to adopt its own rules of procedure, which it has done. The UNSC may 181 have even more flexibility regarding amending its rules of procedure given that it has chosen to keep these rules 'provisional '. 192 Unlike the UNGA's procedural rule 163, there is no written rule regarding how to amend the UNSC rules of procedure, although in practice adoption of such rules are made via affirmative vote of at least nine UNSC members, per Article 27 of the Charter. 193 The text of Article 30 across multiple languages affords significant flexibility to the UNSC in how it decides its own rules of procedure, 194 and it may do so on an ad hoc basis; 'in the case of formally established rules of procedure it would require the formal adoption of amendments to the existing wording'. 195 Thus, UNSC amendments regarding gender parity on the ICJ bench could be submitted and voted on per the normal rules. Such changes to the UNGA/UNSC rules of procedure would be legally binding upon each organ, 196 offering a strong foundation for both securing and maintaining parity.
Additional aspects of the Charter are suggestive that such amendments regarding ICJ elections would be both permissible and required. Article 103 of the Charter provides that the Charter shall take precedent over any internal procedural regulations, giving it primacy over both the UNGA and the UNSC's rules of procedures. Given that the UNSG has interpreted the Charter to require gender parity in all UN staffing, and that both organs have endorsed this view, modification of their rules of procedure to achieve gender parity on the ICJ bench is indeed appropriate.
Thus, both the UNGA and UNSC could modify their internal rules of procedure regarding the election of ICJ judges and create a candidacy system similar to that of the ICC's in which candidate lists that fail to contain sufficient numbers of women will lead to a request for a new list and another round of voting, 197 and with disqualification penalties for non-conforming lists like at the ACtHPR. The creation of the regional groups via a series of UNGA resolutions between 1957 and 1978 demonstrates the power of the UNGA to modify voting procedures at the UN for the purposes of equitable representation. 198 The Statute is flexible enough to allow the UNGA and UNSC to create rules that both reject candidate lists that do not meet gender parity thresholds and require gender parity thresholds to be met when casting election votes. For example, if seats remain unfilled after the first meeting held for election, the Statute contemplates that second and third meetings may take place, and if necessary, the UNGA and UNSC may form a joint conference for choosing names for remaining vacant seats. 199 This inherent flexibility renders it more likely that a gender parity rule would be compatible with the Statute.

Gaining member state support for gender parity on the ICJ bench
It is possible that, as was initially the case at the ECtHR, 200 member states might oppose gender parity requirements and the idea of updating the UNGA/UNSC rules of procedure. Amending the rules of procedure would require member state support, as the UNGA/UNSC are comprised of member state representatives. This is compounded by the fact that UN member states are inclined to work via consensus in both the UNGA and UNSC, 201 creating a higher bar to achieve parity. However, arguably there are legal grounds for requiring member states to support gender parity on the ICJ bench. Additionally, trends towards gender parity rules domestically and 192 supranationally indicate growing state practice in favour of parity on judicial benches. Further, lobbying at the municipal, regional, and international levels can create momentum that fosters member state support for gender parity on the ICJ bench.
Member states are bound by the fundamental human right of gender equality and are specifically instructed to create gender-equal access to the international plane. 202 Numerous states have employed these notions of substantive gender equality and gender parity to their domestic governments 203 and via their roles in drafting gender balance rules at supranational courts, indicating their acceptance of this formulation of gender equality. While some states initially rejected these obligations on the international plane, 204 in recent years more have championed gender parity in international representation. 205 Since states must nominate individuals to the selection process for the ICJ, logically, obligations stipulated in human rights treaties to ensure equal representation apply to states in this context. States' human rights duties necessitate ending discrimination through initiating and implementing affirmative measures for equal representation in international courts. 206 There is evidence that IHRL applies to aspects of the appointment process that are under the exclusive jurisdiction of the state. 207 When member states vote for appointments to international bodies, they cannot vote in a manner that contradicts or violates their international treaty obligations. 208 Additionally, IO law such as the UN's parity staffing policy may transfer UN obligations to member states. Organ practice can both create rules that bind member states and be indicative of state practice in favour of such rules. 209 Under the theory of transfer of powers, member states have imbued the UN with the power to bind them. 210 Member states remain partially or fully 202 responsible for an IO's breach of international law and therefore obligated to remedy this breach. 211 Consequently, if the UNGA and UNSC, given the roles they play in electing ICJ judges, were in breach of their gender equality obligations, states would be responsible for this and therefore required to remedy it. Further, if a state supports the organization's breach through its actions for example by failing to nominate women or enshrine the obligation to do so in policies or rules this support independently gives rise to state responsibility and therefore the need for remedy. 212 There is also the argument that states must prevent such breaches by IOs in order to uphold the integrity of international law as a system. 213 There are several reasons why the UNGA and UNSC may be obligated to ensure gender parity on the ICJ bench and that therefore member states would need to act to ensure this. 'IO organs are themselves subject to international human rights law', 214 and thus the obligation falls on the UNGA and UNSC. Moreover, the UNGA has explicitly confirmed its commitment to the UNSG's gender parity policy on numerous occasions. 215 One could argue that the UNSC has shown its commitment to gender representation in decision making through nine different resolutions on women, peace, and security. 216 A key point is that each organ's established practice indicates its understanding of the Charter, 217 and that UNSC/UNGA practice supports gender parity as a Charter requirement. As explained in Section 5, these policies have coalesced into established practice and therefore binding rules for the organization as a whole.
As these organs have not yet taken action to create gender parity on the ICJ bench, if one accepts their obligation to do so, they are in breach and must supply a remedy. 218 Responsibility for a breach presupposes such things as legal personality, that those deemed responsible had some say in the matter, and that they can meaningfully regulate their behaviour. 219 The UNGA and UNSC have a say in the matter and can regulate their behaviour because they cast the final votes for ICJ judges and can amend their internal rules that govern this voting; they have a direct impact on the ICJ's bench and modifying their voting obligations can create gender parity on it. Reparation for Injuries has made clear that IOs are required to provide remedies when in breach of an obligation. 220 While not expressly listed amongst possible remedies, 221 a direct method for rectifying the breach could be achieved via modifications by the UNGA and UNSC to their internal procedural rules. New rules requiring that candidate lists present gender parity, and that voting must take gender parity into account, can rectify the imbalance on the ICJ bench.
Even setting aside the notion that member states have gender parity obligations regarding nominations to the ICJ, the UNGA and UNSC have the power to create a new requirement to this effect. Member states might protest that such a change in the nomination and voting procedures creates new international law by transferring UN parity requirements to states, but organ power to this effect is well accepted. 222 Certain Expenses dictates that 'each organ must, in the first place at least determine its own jurisdiction'. 223 If the UNGA/UNSC were to modify their internal rules to create gender parity on the ICJ bench, they would be assumed in the first instance to have the competency to do this. State endorsement of the UNGA's role in the UNDT/UNAT judicial elections per Section 3 shows state practice that could favour this conclusion.
States might object that this interferes with their domestic affairs. However, while states have since the advent of the UNGA's first session asserted that Article 2(7) of the Charter prevents the UNGA from intervening in the domestic jurisdiction of a state including regarding human rights issues, the UNGA has repeatedly rejected this argument and declined to address so-called domestic matters on this basis except extremely exceptionally. 224 Rather, expansion of UNGA powers is common and settled law and the UNGA and UNSC have confirmed their key role in protecting and fulfilling human rights. 225 It is not unusual for the UNGA's rules of procedure to impact member states beyond the UNGA itself, for example regarding accreditation, admissions of new members, and elections. 226 In general, IOs and their organs have 'considerable autonomy' regarding their internal rules, 227 and this is so even when changes influence member states and even when they object. 228 When it comes to interpreting the constituent documents, and in this case, interpreting if the Charter requires gender parity on the ICJ bench, 'the balance of power shifts away from Member states to organs'. 229 This further empowers them to modify their internal rules even if member states were opposed.
Member states might argue that it would be impossible for them to comply with gender parity candidate requirements because they were incapable of locating and nominating sufficient numbers of qualified women. States could not be held responsible for an unworkable task. 230 However, this socalled 'pipeline' or 'pool' argument has been robustly debunked. 231 Further, state practice belies such reasoning: if member states can repeatedly make judicial nominations for the ICC, the ACtHPR, and UNAT/UNDT that satisfy gender parity, correspondingly they can for the ICJ. Against the specter that there are simply not enough women candidates for the job, it is important to bear in mind that currently only four additional qualified women need appointing to create gender parity.
Furthermore, while not always accessible for every region, data on women in the legal profession suggests that women comprise a sizeable proportion of high-level legal posts from which to draw ICJ judicial candidates. Women in OECD countries constitute more than 54 per cent of the judiciary, 232 and more than 30 per cent of all Supreme Court judges in Latin America are women. 233 Hennette Vauchez's analysis of the CVs of women candidates put forth by states for appointment to the ECtHR reveals both qualified women and the possibility that states over-but rather one well-connected individual, 236 is much more at fault. Thus, claims that the organs cannot affect the candidate list because it impinges on member state rights or creates an impossible task would likely fail on multiple grounds, paving the way for procedural change if the UNSC/UNGAand the member states that comprise themchose this route.
Despite these legal arguments, member states or other actors at the UN might oppose the suggested changes; identifying legal obligations does not guarantee compliance with them. It will be important to couple the doctrinal reasons articulated in this article with persuasion. 237 How much and what must be done to persuade states is an open question, as is where in the 'life-cycle' of a normnorm emergence, norm cascade, or norm internalization 238gender parity on the ICJ bench sits. To secure parity on the ICJ bench, a critical mass of supportive states is needed. 239 Lobbying efforts, such as those conducted by GQUAL, the campaign for gender parity in international representation, 240 might play a part in encouraging key national governments to support gender parity on the ICJ. Research suggests that the support of less than one third of states is needed to create such change; the support of 'critical' states and the 'socialization' of other states can turn a minority position into consensus. 241 Gender parity in international appointments remains salient and urgent. It is highly unlikely that the call for gender parity on international benches will disappear, increasing the possibility that member states will support parity on the ICJ bench.
7. Conclusion: Compelling legal reasons to create gender parity on the ICJ bench There are compelling reasons to consider gender parity on the ICJ bench and multiple legal, not just policy or moral, arguments in favour of parity. This article has posited that, over time, the customary law of the UN has evolved to support gender parity on the ICJ bench. This in turn affects the interpretation of Article 9 of the Statute regarding judicial selection, so that UN organs such as the UNGA and the UNSC are either legally permitted or legally obligated to ensure gender parity on the ICJ bench. Not only are there grounds for the UNGA and the UNSG to take parity measures 'establishing real accountability, backed by consequences' 242 such as those seen at the ACtHPR, the UNGA and UNSC can achieve this by modifying their internal rules of procedure. There is ample precedent for them to mandate such obligations via their procedural rules, even bearing in mind that this would impact member states. Such an outcome is supported by conventions on equitable geographical distribution. Gender should be given equal weight as geography: there is evidence that the two goals can be achieved symbiotically, and that it is at times appropriate to favour gender parity over geographical distribution. Importantly, only through clear and enforceable measures has parity thus far been achieved and protected. For these reasons, the UNGA and UNSC should act now to modify their rules and create a gender equal ICJ bench.