Post-mortem on the ICC’s Prosecutorial Election: Lessons Learned and Questions for the Future – Natalie Hodgson

On 12 February 2021, Karim Khan QC was elected as the third Prosecutor of the ICC. This piece revisits the long process to that election, focusing on two key issues: the role of the Committee for the Election of the Prosecutor and the need for consensus. The election brought to the fore multiple important issues for states, civil society, academics, lawyers and the Court to consider. Now that the election is over, it is important to take this opportunity to reflect on some of the lessons learned for the future. 

On 12 February 2021, Karim Khan QC, a barrister from the United Kingdom, was elected as the third Prosecutor of the ICC. The election process was long and fraught. This election was the first time that states parties did not elect the Prosecutor by consensus; four men were nominated to contest the final ballot. Only one of those nominees, Fergal Gaynor, was among the shortlisted candidates identified by the Committee for the Election of the Prosecutor (CEP). The other three nominees were drawn from the CEP’s longlist – Carlos Castresana Fernández, Francesco Lo Voi and Khan. 

The way the election ultimately unfolded was far from what was envisaged when the process to elect a new Prosecutor commenced in 2019. This piece focuses on two key issues that emerged during the election – the role of the CEP and the need for consensus. (Other issues that arose, such as the need for improved vetting of candidates and female underrepresentation among candidates, have been discussed elsewhere.) In reflecting on these issues, this piece hopes to identify some lessons and questions for future prosecutorial elections.

The Envisaged Process and its Deviations 

In April 2019, the Bureau of the Assembly of States Parties (ASP) released its Terms of Reference for the election of the Prosecutor. It identified a process whereby the CEP would be formed to receive applications from interested individuals, prepare a longlist of candidates to interview, and establish by consensus a shortlist of three to six of the most highly qualified candidates for the role. The creation of the CEP was a welcome step towards ensuring that the ICC’s next Prosecutor would be elected on the basis of qualifications and merit, not politics. The CEP consisted of five members, one representative for each regional group, and was assisted by a panel of five independent experts, also comprising one expert for each regional group. 

Once all the applications had been received, the CEP generated a longlist of 16 candidates (reduced to 14 when two candidates withdrew) to interview. From this longlist, four candidates advanced to the shortlist: Morris A Anyah (Nigeria), Fergal Gaynor (Ireland), Susan Okalany (Uganda) and Richard Roy (Canada). 

Historically, the Prosecutor has been elected on a consensus basis. That is, states agreed on who to elect as the next Prosecutor; this individual was the only person officially nominated for the role. However, in this election, some states immediately raised concerns about the shortlist generated by the CEP. Given that the current Prosecutor, Fatou Bensouda, is from The Gambia, the nominations of Anyah and Okalany conflicted with states’ tacit rule of regional rotation for the position. Additionally, as James Stewart, current Deputy Prosecutor, is also Canadian (and not due to leave office until 2022), concerns were raised that Roy’s nomination would conflict with Article 42(2) of the Rome Statute (requiring the Prosecutor and Deputies to be different nationalities), although others suggested that this was not a fatal issue. On 13 July 2020, Kenya wrote to the President of the ASP, describing the shortlist as a fait accompli weighted in favour of Gaynor. 

As the election continued, it became increasingly clear that consensus would not be reached on any of the four shortlisted candidates. On 13 November 2020, the decision was made to expand the list of potential candidates to include longlisted candidates. The longlisted candidates were asked whether they wished to continue being considered as a candidate and were provided with a copy of their appraisal by the CEP. Five candidates withdrew during these stages, leaving five remaining longlisted candidates who were publicly announced as additional candidates for the position – Carlos Castresana Fernández (Spain), Francesco Lo Voi (Italy), Robert Petit (Canada), Brigitte Raynaud (France) and Khan (United Kingdom).

Discussions to find a consensus candidate continued. The ASP annual meeting, where states had been due to elect the new Prosecutor, came and went. By February 2021, with consensus no longer appearing possible, the decision was made to take the matter to a contested election at the second resumption of the ASP. Four candidates were nominated by states – Castresana Fernández, Gaynor, Lo Voi and Khan. Khan won in the second round of voting, with 72 votes cast in his favour out of a total of 122 votes. Thus, a process which seemingly began with the intention that states would select a Prosecutor by consensus from a shortlist created by the CEP ended with states voting in a contested ballot where three of the four nominees had not made the CEP’s shortlist – including the Prosecutor who was ultimately elected.

The Role of the CEP

The twists and turns in the election raise a number of questions as to how future elections should proceed. The first of these involves the role performed by the CEP.

An issue first arose when the CEP announced its shortlist of candidates. The CEP was acting in accordance with its Terms of Reference (at [19]) when it provided a shortlist that had ‘due regard to… geographical balance’ (see also the CEP’s Report at [23] and [40]). However, the geographical balance provided by the CEP conflicted with states’ expectations of regional rotation, which meant that some states perceived certain candidates to be unelectable. All four of the candidates eventually nominated to contest the election were from countries within the Western European and Other States Group.

Regional rotation is not a requirement found in the Rome Statute – the Rome Statute only requires that the Prosecutor and Deputy Prosecutors be of different nationalities (Art 42(2)). Thus, the CEP was correct to perform its duties according to its Terms of Reference, which required ‘geographical balance’. Perhaps the issue of nationality would have been less consequential if the CEP had provided a shortlist of six candidates (as was permitted at [16] of the Terms of Reference), which may have presented states with additional options to consider. However, as indicated by Ambassador Sabine Nölke, Chair of the CEP, the decision to shortlist four candidates instead of six was based on the CEP’s assessments of which candidates were the most qualified and who the CEP could reach consensus on. The CEP also indicated (at [40] of their Report) that they were somewhat limited in their ability to provide a diverse shortlist (with respect to geographical balance, gender and the representation of the principal legal systems of the world) by the available pool of candidates. Accordingly, for future elections, thought must be given to: (1) how to facilitate the career development of individuals from underrepresented geographical groups, to expand the group people sufficiently qualified to apply for the position of Prosecutor; (2) how to encourage qualified candidates to apply for the position of Prosecutor; and (3) to what extent (if at all) the CEP should be expected to be responsive to states’ tacit geographical expectations in its selection and shortlisting processes – and, if not, how the discrepancy between the CEP’s Terms of Reference and states’ expectations should be accommodated and/or addressed. 

Once it was evident that some states had issues with some of the candidates, the question arose as to whether and how to diverge from the CEP’s shortlist to consider other candidates. Civil society argued that states should ‘stand by their commitment to a process for election of the next prosecutor that is transparent, merit-based, and free from political interference’, being particularly concerned about the prospect of new candidates being nominated who had not been subject to the same level of due diligence and scrutiny applied to other candidates. In contrast, others (see Professor Gregory S Gordon and Gunnar Ekeløve-Slydal) argued that process should not trump substance; that is, the most important consideration was ensuring that the best candidate for the job was elected – preferably with consensus from states. The President of the ASP’s decision to open the list of potential candidates to those who had been longlisted by the CEP – meaning that the additional candidates had been subjected to the same process and scrutiny as the shortlisted candidates – appears to have been the best way to balance civil society’s concerns with the desire to secure a consensus candidate. 

The CEP’s Terms of Reference acknowledged that ‘[t]he CEP and subsequent consultation process is supplementary to the formal procedure set out in the nomination resolution’ (emphasis added). That is, the CEP’s shortlist was never binding. States parties always retained the ability and right to submit a nomination (including of an individual not on the shortlist or longlist) during the formal nomination period. Nonetheless, as a general rule, one would think that the CEP’s shortlist should generally be adhered to and only diverged from in exceptional circumstances (as was seemingly the case in this election). The risk is that, despite its exceptional circumstances, this election normalises similar diversions in the future.

Professor Kevin Jon Heller has written about how the CEP was established – including that the Terms of Reference were drafted by the President of the ASP and adopted (by a ‘silence procedure’) by the Bureau, with limited input from the broader ASP. Given the issues that have arisen in this election, it is important for future elections that the CEP’s processes are not only rigorous but have the support of the ASP. Involving the ASP to a greater extent in the CEP’s creation may give the CEP greater legitimacy among states, which may help to ensure that any future shortlist generated by the CEP is broadly respected and only diverged from if exceptional circumstances arise. 

The Necessity of Consensus

The driving imperative of the length of the election process was the desire to achieve consensus among states as to which candidate should be elected. The election of the Prosecutor by consensus is practically desirable because it means that the Prosecutor ostensibly has the support of all the ICC states parties. Additionally, consensus is symbolically important, given that the Prosecutor acts on behalf of the international community in investigating and prosecuting international crimes. 

The events of this election raise the question of whether consensus is an objective that should be retained. As the ICC has developed as an institution, encountered new challenges and faced setbacks that have affected its functioning and the administration of justice, numerous debates have occurred as to the appropriate role of the Prosecutor and the best strategy for the OTP. It is natural that states might disagree as to the key priorities of the OTP and the traits a Prosecutor needs to successfully pursue those. As noted by the CEP (at [37] of its Report), ‘[t]here is no such thing as a “perfect” candidate’; all of the candidates had areas where their experience was not as extensive as considered desirable. Therefore, it is reasonable and legitimate for states to disagree as to which candidate is best suited for the job. Consensus may no longer be a feasible outcome in a prosecutorial election.

Nonetheless, there are benefits to retaining a pursuit of consensus. Such a requirement encourages debate, discussion and reflection among states about the most important attributes for a Prosecutor and the candidate that best demonstrates those attributes. The requirement for consensus may also operate to somewhat prevent the political vote trading that characterises the judicial elections. Therefore, the question of whether states should continue to aim for the election of a Prosecutor by consensus is an important issue that states should discuss prior to the next election.


The ICC’s prosecutorial election has raised a number of important issues. The election has generated discussion and debate around the functions that the Prosecutor should perform, and the qualities needed to be an effective Prosecutor. It has brought to the fore the problem of sexual harassment in international criminal law, as well as a lack of gender and regional diversity, particularly among individuals at higher levels in international institutions. It has also raised questions about process: how the Prosecutor should be elected and what is needed for this process to have legitimacy. This was by no means a straightforward or uneventful election. But it has given states, civil society, academics, lawyers and the Court plenty of food for thought. It is important that we take this opportunity to think about the issues raised throughout the election and how we can improve in the future.   

Natalie Hodgson is a Scientia PhD Candidate in the Faculty of Law & Justice, UNSW Sydney.

Suggested citation: Natalie Hodgson, ‘Post-mortem on the ICC’s Prosecutorial Election: Lessons Learned and Questions for the Future’ on ILA Reporter (19 March 2021) <>