¹ Thirlway, H., ‘The International Court of Justice: Cruising Ahead at 70’, (2016) 29 Leiden Journal of International Law 1103 CrossRefGoogle Scholar.

² D. Terris et al., The International Judge: An Introduction to the Men and Women Who Decide the World's Cases (2007), 15.

³ Elsig, M. and Pollack, M.A., ‘Agents, trustees, and international courts: The politics of judicial appointment at the World Trade Organization’, (2014) 20 European Journal of International Law 1 Google Scholar; Danner, A. and Voeten, E., ‘Who is running the international criminal justice system?’, in Avant, D.D. et al. (eds.) Who Governs the Globe? (2010), 35 CrossRefGoogle Scholar; R. Mackenzie et al., Selecting International Judges: Principle, Process, and Politics (2010); Steinberg, R., ‘Judicial Lawmaking at the WTO: Discursive, constitutional and political constraints’, (2004) 98 American Journal of International Law 247 Google Scholar; Terris et al., supra note 2; Voeten, E., ‘The Politics of International Judicial Appointments: Evidence from the European Court of Human Rights’, (2007) 61 International Organization 669 Google Scholar; Voeten, E., ‘The Politics of International Judicial Appointments’, (2009) 9 Chicago Journal of International Law 387 Google Scholar; Wood, M., ‘The Selection of Candidates for International Judicial Office: Recent Practice’, in Ndiaye, T.M. et al. (eds.), Law of the Sea, Environmental Law and Settlement of Disputes: Liber Amicorum Thomas A. Mensah (2007), 357Google Scholar.

⁴ But see Keith, K.J., ‘Challenges to the Independence of the International Judiciary: Reflections on the International Court of Justice’, (2017) 30 Leiden Journal of International Law 137 Google Scholar.

⁵ T. Franck, Fairness in International Law and Institutions (1995), 346.

⁶ Robinson, D.R., ‘The Role of Politics in the Election and the Work of Judges of the International Court of Justice’, (2003) 97 Proceedings of the Annual Meeting (American Society of International Law) 277 Google Scholar.

⁷ Posner, E.A. and de Figueiredo, M.F.P., ‘Is the International Court of Justice Biased?’, (2005) 34 The Journal of Legal Studies 599 Google Scholar; Smith, A.M., ‘“Judicial Nationalism” in International Law: National Identity and Judicial Autonomy at the ICJ’, (2005) 40 Texas International Law Journal 197 Google Scholar.

⁸ By ‘vote share’ we mean the number of votes received by each candidate within each election relative to the total number of states that actually voted within either the General Assembly or the Security Council.

⁹ By ‘signaling device’ we refer to the game theoretic concept of using an action to convey information about intentions or preferences. See generally, J.D. Morrow, Game Theory for Political Scientists (1994). In the context of ICJ elections, governments that cast nomination votes for a candidate who has already been nominated are signaling their preference that other states co-ordinate vote choice on this particular candidate.

¹⁰ ‘Socialization mechanisms’ refers to the processes through which actors come to adopt norms of appropriate behaviour in social or institutional environments. In the context of ICJ elections, a large number of nominations for a given candidate could activate mechanisms of imitation or emulation of group preferences or peer pressure. See generally R. Goodman and D. Jinks, Socializing States: Promoting Human Rights Through International Law (2013); Johnston, A.I., ‘Treating International Institutions as Social Environments’, (2001) 45 International Studies Quarterly 487 Google Scholar.

¹¹ See also Keith, supra note 4.

¹² ICJ Statute, Art. 13(1).

¹³ The first elections, held in 1946, have been excluded from this analysis due to much more limited data on these candidates and the fact that the first election may be unique in terms of state motivations.

¹⁴ Judges elected by special election sit on the ICJ for the unexpired part of the term only.

¹⁵ ICJ Statute, Art. 2.

¹⁶ The UN regional groups consist of the five UN General Assembly geographical groups: Western European and Other Group (WEOG); Eastern Europe; Latin America and the Caribbean (GRULAC); Africa; and Asia. For the seats available to each regional group, there are also sub-regional conventions ensuring, e.g., the distribution of WEOG seats between Southern and Northern Europe states in WEOG, and African seats between Arab, Francophone and Anglophone candidates. Mackenzie et al., supra note 3, at 360.

¹⁷ Mackenzie et al., supra note 3, at 37–40.

¹⁸ ICJ Statute, Arts. 4–6.

¹⁹ Mackenzie et al., supra note 3, at 73–84.

²⁰ Robinson, supra note 6, at 279.

²¹ Mackenzie et al., supra note 3, at 84–95.

²² Rules of Procedure of the Assembly, Rule 92. There is no equivalent to Rule 92 in the Security Council's Provisional Rules of Procedure, but ‘in practice all ballots in the Security Council for elections to the Court have been secret’, see S. Rosenne, Law and Practice of the International Court: 1920-2005 (2006), 375.

²³ ICJ Statute, Art. 8.

²⁴ Cogan, J.K., ‘Competition and Control in International Adjudication’, (2008) 48 Virginia Journal of International Law 411 Google Scholar; Mackenzie et al., supra note 3, at 101, 106–7, 111.

²⁵ Mackenzie et al., supra note 3, at 75–7.

²⁶ See Hawkins, D. et al. (eds.), Delegation and Agency in International Organizations (2006)Google Scholar; Nielson, D.L. and Tierney, M.J., ‘Delegation to International Organizations: Agency Theory and World Bank Environmental Reform’, (2003) 57 International Organization 241 CrossRefGoogle Scholar; Pollack, M.A., ‘Delegation, Agency, and Agenda Setting in the European Community’, (1997) 51 International Organization 99 Google Scholar.

²⁷ Majone, G., ‘Two Logics of Delegation: Agency and Fiduciary Relations in EU Governance’, (2001) 2 European Union Politics 103 CrossRefGoogle Scholar; Alter, K.J., ‘Agents or Trustees? International Courts in their Political Context’, (2008) 14 European Journal of International Relations 33 Google Scholar.

²⁸ Alter, supra note 27; Grant, R.W. and Keohane, R.O., ‘Accountability and Abuses of Power in World Politics’, (2005) 99 American Political Science Review 29 Google Scholar.

²⁹ Voeten (2009), supra note 3, at 389.

³⁰ To be sure, we have not exhausted all the possible reasons for vote choice and indeed the reasons for nominating a particular candidate are many and may be idiosyncratic. For instance, nominations ‘can also be used as political tools: a state may wish to reward an individual or to banish an individual’. Mackenzie et al., supra note 3, at 70. However, because we are focusing on aggregate levels of state support for a given candidate (and not individual nomination or vote choices), we can say something about the candidate characteristics that appear to shape most states’ vote choices.

³¹ With some exceptions within the early elections, for which the Secretariat prepared a CV for the nominee, usually based on ‘Who's Who’ guides.

³² Danner and Voeten, supra note 3; Voeten (2007), supra note 3.

³³ Amerasinghe, C.F., ‘Judges of the International Court of Justice – Election and Qualifications’, (2001) 14 Leiden Journal of International Law 335 Google Scholar.

³⁴ Posner, E.A. and Yoo, J.C., ‘Judicial Independence in International Tribunals’, (2005) 93 California Law Review 1 Google Scholar.

³⁵ Helfer, L. and Slaughter, A.M., ‘Why States Create International Tribunals: A Response to Professors Posner and Yoo’, (2005) 93 California Law Review 899 Google Scholar.

³⁶ Elsig and Pollack, supra note 3, at 20; Steinberg, supra note 3, at 247.

³⁷ Enlightened positivism refers to a ‘school of thought which regards custom and treaties as the primary sources of international law, whose authority, so far as existing law is concerned, is higher than any conflicting rule of natural law and morality’. Kooijmans, P., ‘Two Remarkable Men Have Left the International Court of Justice’, (2000) 13 Leiden Journal of International Law 345 Google Scholar, citing Judge Sir Hersch Lauterpacht.

³⁸ An example of an expansive interpretation by the ICJ may be found in Reservations to the Convention on the Prevention and Punishment of Genocide, Advisory Opinion, 28 May 1951, [1951] ICJ Rep. 15. The Court considered that the newly adopted Genocide Convention contained principles binding on all states ‘even without any conventional obligation’. Akhavan suggests that the Court considered that the Convention's transcendent object and purpose led the Court to interpret it in a broad manner, and that this interpretation stands in stark contrast to the narrow interpretation of the Genocide Convention's Article IX adopted by the Court in Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Croatia v. Serbia), Merits, Judgment of 3 February 2015, [2015] ICJ Rep. 3. See more in Akhavan, P., ‘Balkanizing Jurisdiction: Reflections on Article IX of the Genocide Convention in Croatia v. Serbia ’, (2015) 28 Leiden Journal of International Law 893 Google Scholar.

³⁹ T.J. Bodie, Politics and the Emergence of an Activist International Court of Justice (1995). For concrete examples in the ICJ context, see Kooijmans, supra note 37, who places Judge Weeramantry and Judge Schwebel at the opposite sides of this spectrum, with Judge Weeramantry adhering to the natural school of law, and Judge Schwebel as a sound proponent of ‘enlightened’ positivism. See also, Anghie, A., ‘C.G. Weeramantry at the International Court of Justice’, (2001) 14 Leiden Journal of International Law 829 Google Scholar.

⁴⁰ While some commentators and ICJ judges have expressed little faith in the Court's ability to contribute to the resolution of politically sensitive disputes, others argue that the ICJ not only has the ability to, but often does contribute to the peaceful resolution of highly political disputes. See Coleman, A., ‘The International Court of Justice and Highly Political Matters’, (2003) Melbourne Journal of International Law 29 Google Scholar.

⁴¹ Voeten (2007), supra note 3.

⁴² This includes holding political office, working within government bureaucracy, diplomatic positions (including state delegates to international organizations), and solicitor generals.

⁴³ Voeten (2007), supra note 3, at 681.

⁴⁴ For international judicial experience, this probably depends on whether these prior positions were in an independent tribunal as opposed to an arbitration panel. As a first cut we have grouped all prior international judicial experience together, including experience on arbitration or investment arbitration panels. It may be more appropriate, however, to divide prior international judicial experience into more and less dependent positions.

⁴⁵ M. Barnett and M. Finnemore, Rules for the World: International Organizations in Global Politics (2004), 5–6.

⁴⁶ This does not include candidates who only acted as a government delegate to an IO or to a treaty convention. It does include candidates who were members of the International Law Commission (ILC), since ILC members do not sit as representatives of their government. Further research will separate out ILC membership from other positions within IOs.

⁴⁷ This does not include candidates who only served as a Judge ad hoc on the ICJ.

⁴⁸ These figures do not include candidates who only held lecturer or part-time lecturer positions.

⁴⁹ To some extent, this measure also captures a candidate's range of experiences and relies on the assumption that professional backgrounds have a cumulative effect on probability of independence.

⁵⁰ Alter, supra note 27.

⁵¹ This measure was constructed by coding whether a candidate holds a degree from the US/UK, a degree from a Western European institution, or degrees from both, on a scale from 0 (no prestigious degrees) to 2 (degrees from both US/UK and Western European institutions). Information was drawn from a candidate's CV and supplemented with outside biographical research when necessary. For regular elections, elected candidate mean = 0.9; unelected candidate mean = 0.72. For special elections, elected candidate mean = 0.71; unelected candidate mean = 0.62.

⁵² Elected candidates listed an average of 21.8 publications (regular elections) and 26.3 publications (special elections). Unelected candidates listed an average of 12.4 publications (regular elections) and 10.1 publications (special elections).

⁵³ Cogan, supra note 24.

⁵⁴ Candidates within the same election are not always on the ballot for each voting round; once they receive the required majority within a body, they are removed from the ballot. Voting continues on the remaining candidates until five individuals have received an absolute majority. If the General Assembly and Security Council did not select the same five individuals, voting resumes on all candidates except those upon whom the two bodies agreed.

⁵⁵ Wotipka, C.M. and Tsutsui, K., ‘Global Human Rights and State Sovereignty: State Ratification of International Human Rights Treaties, 1965-2001’, (2008) 23 Sociological Forum 724.Google Scholar

⁵⁶ Elsig and Pollack, supra note 3.

⁵⁷ For regular elections the number of P5 nominations has more or less remained the same. For special elections, P5 nominations typically depend on whether the slot to be filled was vacated unexpectedly by a P5 member.

⁵⁸ Only three P5 candidates appearing on the ballot were not elected during regular elections held in 1951, 1969 and 1978. They were all of American nationality (Abraham H. Feller; Herbert W. Briggs; and Myres S. McDougal), though none were nominated by the US national group.

⁵⁹ Arguably, whether a candidate's home country is a member of the Security Council at the time of election may also matter as their government can informally campaign other SC members prior to and during the voting rounds. However, aside from P5 members, very few candidates’ home countries were Security Council members at the time of their election.

⁶⁰ Danner and Voeten, supra note 3, at 61–2.

⁶¹ As an alternative test of political sensitivity, we include in three separate models a dummy variable for prior government experience (GOVPOL), a dummy for experience within academia (ACADEM), and both GOVPOL and ACADEM.

⁶² Elsig and Pollack, supra note 3.

⁶³ Since the Court's inception, 57 candidates have been up for re-election, 40 of whom were re-elected (70.2%), suggesting that states employ sanctioning through re-election in extreme cases only.

⁶⁴ Within models not reported we include in addition to EDUC the “prestige” of a candidate’s education (degrees obtained from US/UK and Western European institutions) and the number of academic publications listed as a measure of a candidate’s personal reputation, or in other words, his or her reputation amongst the legal community. Neither of these variables substantially changes results, and their substantive effects are minimal and insignificant.

⁶⁵ Average number of voting members during the initial voting round is 136.67 for the General Assembly and 14.05 for the Security Council.