For years, critics of investor-state dispute settlement (ISDS) have described its arbitrators as “pale, male, and stale,” a line usually traced to Michael Goldhaber’s 2004 essay asking why women so rarely chair major tribunals. The complaint was about legitimacy: a small, look-alike group of repeat players decides cases that test how states regulate. Since then, institutions have started publishing gender statistics, diversity pledges have spread, and the numbers have moved. Aggregate numbers can hide a lot, though. As more women are appointed, it helps to ask who is reaching these seats, and through which doors.

To find out, I built a dataset of 148 treaty-based disputes filed against Middle East and North Africa (MENA) states between 2000 and 2025, drawn from the ICSID case database, the Permanent Court of Arbitration, and UNCTAD’s Investment Dispute Settlement Navigator. It records 387 named appointments held by 184 people, coded by tribunal role, nationality, and gender. MENA is a good place to ask the gender question. The caseload is large and recent: 86 of the 148 cases were filed in 2016 or later. Many disputes also turn partly on the host state’s own law.

At first glance, the trend looks like progress. Women hold 21 percent of presidencies and 18.3 percent of respondent-appointed seats. Female chairs went from 9.1 percent of cases filed in 2000–2010 to 25 percent in 2016–2020 and 43.8 percent in 2021–2025. The chair runs the procedure, frames the issues, and usually drafts the award, so this is real change in the seat that carries the most power.

The gains are uneven, though. Women hold only 4.7 percent of claimant-appointed seats. No woman held one in cases filed from 2000 to 2010, and in every period since, the share has stayed below 6 percent. The pattern follows who makes the choice. Presidents are usually picked by the co-arbitrators or by an appointing institution. The claimant seat is filled by one party acting alone. This fits Sarah Rudolph Cole’s observation that institutional appointments have done more for diversity than party appointments.

Even in the chair, the gains rest on a few people. One arbitrator accounts for 30.8 percent of all female presidencies in the dataset, and a single appointee accounts for 43.5 percent of female respondent appointments.

Now the number that I think should change how we read these statistics. Across every party-appointed seat in these cases, claimant and respondent combined, women with a MENA nationality account for zero percent. They appear only as chairs, and there in just 3.2 percent of chair appointments. The women being added are overwhelmingly not from the region whose public law is on trial.

An intersectional reading helps here. Counting “women on tribunals” treats gender as one separate axis and assumes the women counted are interchangeable. A woman who is also a MENA national has to get through two gates. The gender gate has started to open through institutional channels. The regional gate has hardly moved; MENA nationals of any gender hold just 6.3 percent of chairs. She is stopped by whichever gate is narrower when the appointment is made. The overall gender figure rises because the new seats go to women who only have to clear one of them.

Some of this has a long history. Egypt never passed a law barring women from the bench, yet Tahani al-Gebali became the country’s first female judge only in 2003, when she was appointed to the Supreme Constitutional Court. She was already an international commercial arbitrator, as Engy Abdelkader recounts. In Saudi Arabia, women were admitted to court practice only in 2013, and the Shura Council rejected proposals to appoint female judges in 2018 and again in 2020, according to Samah al-Agha’s study. Even so, a Saudi court accepted the Kingdom’s first female arbitrator, Shaima al-Jubran, in 2016. Arbitration opened to women there before the courts did. The investment appointment market has not caught up.

Other barriers are personal. The Lebanese arbitrator Nayla Comair-Obeid has recalled a co-arbitrator who protested her nomination as president because he “did not want to be chaired by a woman,” an episode Harshad Pathak describes. Gender also compounds with region across the field. Andrea Bjorklund and her co-authors counted only three appointments of female, non-white arbitrators from developing states among nearly a thousand ICSID appointments between 2012 and 2017. The zero in my data fits that larger pattern.

The gap also has a practical side. English is the language of proceedings in 92.2 percent of the cases where the language is reported, and Arabic is the sole language in 2.2 percent. Yet these tribunals regularly apply Egyptian, Libyan, or Jordanian law and Arabic treaty texts. In Agility v. Iraq, the tribunal heard a linguistic expert on the Arabic text of the OIC Investment Agreement because no member of the tribunal was a native Arabic speaker. Two tribunals have read Article 17 of that same treaty in opposite ways, in Al Warraq v. Indonesia (2012) and Itisaluna v. Iraq (2020). MENA women are among those who could read it firsthand.

None of this depends on the motives of any appointing party, and the figures alone do not prove bias. My claim is narrower: gender gains cannot stand in for representational legitimacy when progress on one axis leaves the other untouched.

Party autonomy is central to arbitration, and nothing here asks parties to give up their choice. The pool they choose from can still grow. Institutions could publish appointment data that cross gender with region, so progress for some is not read as progress for all. When institutions and co-arbitrators select a president, they could treat Arabic-language ability and working knowledge of the host state’s law as real qualifications, since language already counts as a criterion in arbitrator selection. MENA states could also use their seats on the ICSID Panel of Arbitrators to name women from the region, which puts new names in front of every party. The bench is becoming less male. For MENA women, it is not yet becoming open.