In November 2016, the Office of the Prosecutor (OTP) at the International Criminal Court (ICC) produced an extensive report documenting the status of ten “preliminary examinations” it is conducting. These examinations cover situations in which crimes under the jurisdiction of the Court may have been committed but the prosecutor has not yet decided to open a full investigation.Footnote 1 The OTP's reporting practice has shed additional light on a process that has been opaque for much of the Court's existence and that has attracted relatively limited scholarly and specialist attention.Footnote 2
More regular and detailed OTP reporting, and information from other sources, makes closer consideration of the preliminary examination phase possible. Even if speculative in certain respects, this analysis is important given the attention and criticism that the ICC's process of selecting situations for full investigation has attracted. The Court's focus on African conflicts, in particular, has provoked tension between the Court and several African leaders and apparently contributed to several African states’ withdrawal from the Rome Statute (the Statute).Footnote 3 Understanding the Court's situation-selection process requires greater scrutiny of the preliminary examination phase.
This article focuses on a particular aspect of the ICC's preliminary examinations: the distinction between examinations conducted on the prosecutor's own initiative (proprio motu) and those conducted as a result of member state or United Nations Security Council referral. The OTP has insisted that it conducts the same type of preliminary examination regardless of the way in which Court involvement was triggered. I argue that the prosecutor has, in practice, developed two quite different processes. For member state and Security Council referrals, the prosecutor conducts a review tilted sharply toward opening a full investigation. In proprio motu situations, the presumption appears to be reversed. This bifurcation has important implications for the Court's docket and, ultimately, for its independence. The ICC statute is an elaborate compromise between the prerogatives of states and the interests of justice, and the apparent deference given to state and Security Council referrals provides important perspective on how the prosecutor is striking that balance. The divergence between the OTP's stated policy and its record also sheds light on how the prosecutor's office is navigating the myriad pressures it faces.
I. The Preliminary Examination Process
Most international courts and adjudicative bodies reserve for member states the power to initiate court action. The International Court of Justice, the World Trade Organization's dispute settlement system, and the Law of the Sea Tribunal, for example, all require that member states affirmatively trigger adjudication.Footnote 4 By contrast, the Rome Statute “vests the power to investigate and prosecute the politically sensitive crimes within its broad territorial sweep in a single individual, its independent prosecutor.”Footnote 5 The ICC prosecutor's responsibility to select investigations from a range of situations places her in a very different position from most international judicial officials—but also from the prosecutors of other international criminal tribunals. The instruments that created those tribunals, including for the former Yugoslavia and Rwanda, defined the territorial and temporal boundaries within which the prosecutors were to work; the ICC prosecutor has few such limitations.
The Rome Statute attempts to address this new reality in part through the preliminary examination, “an important and necessary innovation compared to the pre-trial procedure of former International Criminal Tribunals … .”Footnote 6 The Statute makes the prosecutor's office responsible for “receiving referrals and any substantiated information on crimes within the jurisdiction of the Court, for examining them and for conducting investigations and prosecutions before the Court.”Footnote 7 The phrase “preliminary examination” appears only once in the Statute and is not precisely defined.Footnote 8 It has been usefully described as “the investigative steps which the Prosecutor may take after he or she is seized of a situation but prior to his or her determination of whether there is a reasonable basis to proceed with an investigation … .”Footnote 9 There is little in the Statute regarding how the prosecutor should conduct this preliminary work, and one commentator has argued that the document “offers no real guidance on the criteria that the Prosecutor is to apply in making determinations about which situations to pursue and which ones to ignore.”Footnote 10 Indeed, several provisions create confusion about the standards that apply to these examinations and, in particular, how much deference the prosecutor should give to state preferences.
The Statute provides that three well-known mechanisms may “trigger” ICC inquiries.Footnote 11 First, a member state may refer a situation to the Court. Second, the UN Security Council, acting under Chapter VII of the UN Charter, may do so. Finally, the prosecutor may seek to initiate an investigation on her own.Footnote 12 This proprio motu power of the prosecutor was debated at length during the Rome Conference; it survived despite the strong opposition of several countries, including the United States.Footnote 13 The three principal trigger mechanisms—Council referral, member state referral, and proprio motu initiation—correspond to very different environments for the Court in terms of state support for its work. At one end of the spectrum, a Security Council referral means that the leading international authority on peace and security—and therefore several of the most powerful states—desire Court involvement in a given situation.Footnote 14 A member state referral means that at least one government, likely although not necessarily one directly affected by alleged crimes, seeks Court action.Footnote 15 By contrast, a proprio motu situation could mean that the prosecutor has received information about possible crimes primarily from nongovernmental sources and that no member state supports—or is willing to acknowledge support for—Court action. A key question is whether the prosecutor should conduct preliminary examinations with the same receptivity to opening a full investigation notwithstanding these potentially significant differences in state support.
Certain provisions of the Statute suggest uniformity in the preliminary examination process. Article 53, which guides the prosecutor in determining whether to open a full investigation, provides only that the prosecutor shall evaluate “the information made available to him or her” and determine if there is a “reasonable basis” to proceed with a full investigation.Footnote 16 The wording of Article 53 suggests that this test is “applied in the same manner to all triggers … .”Footnote 17 The provision also implies that a full investigation is the default course of action; the prosecutor shall initiate an investigation unless there is no reasonable basis to proceed.Footnote 18 Yet other elements of the Statute lead to the conclusion that the trigger mechanism should affect the preliminary examination process. Article 15, which describes the proprio motu power of the prosecutor, provides that the prosecutor shall petition to open a full investigation, “if [she] concludes that there is a reasonable basis to proceed.”Footnote 19 Juxtaposing the phrasing of Articles 15 and 53, the OTP itself has noted that the prosecutor begins referral and proprio motu preliminary examinations at different points:
Where the prosecutor receives a referral, Article 53 provides that the Prosecutor shall initiate an investigation unless he determines that there is no reasonable basis to proceed under the Statute … . When the Prosecutor receives a communication, the test is the same but the starting point is reversed: the Prosecutor shall not seek to initiate an investigation unless he first concludes that there is a reasonable basis to proceed.Footnote 20
The statutory case for differentiated preliminary examinations does not rest solely on the relationship between Articles 15 and 53. The provision on preliminary admissibility challenges, Article 18, also distinguishes between different triggers. It provides mechanisms through which states can demonstrate to the Court that they are investigating a given situation and thereby prevent the prosecutor from opening a full investigation due to complementarity. Yet Article 18 refers only to investigations begun through member state referral or proprio motu action.Footnote 21 The omission of the Security Council trigger implies that states cannot challenge admissibility, at least at an early stage, when the Council has referred a situation. The unstated corollary is that the prosecutor need not conduct a full admissibility assessment during the preliminary examination of a Council referral.Footnote 22 The apparent removal of such a significant obstacle has led some observers to argue that the Statute creates a “fast-track” process for Council referrals.Footnote 23
Viewed together, Articles 15, 18, and 53 suggest different levels of substantive review during the preliminary examination phase. One scholar has advocated several distinct standards, with Security Council referrals receiving the least scrutiny from the prosecutor and proprio motu situations the greatest.Footnote 24 In addition to the textual support described, this interpretation has a political logic, at least from a perspective that is respectful of state authority. The prosecutor might feel most confident that a proposed investigation has merit when the fifteen-member Security Council has recommended it. Referral by a member state also gives the prosecutor some reassurance but deserves greater independent analysis. The prosecutor should be most cautious when no state has referred. However plausible, the Statute certainly does not require this reading, and the uniformity suggested by Article 53 in isolation is an obstacle to this interpretation.
The relevance of trigger mechanisms to preliminary examinations is clearer in procedural terms. Most important, the prosecutor must seek the approval of a pretrial chamber of judges before opening a full investigation proprio motu.Footnote 25 This requirement was included to ease concerns about the danger of an unchecked prosecutor.Footnote 26 Unlike in referral situations, therefore, the OTP conducts its proprio motu preliminary examinations knowing that it must present evidence to a panel of judges in order to launch a full investigation. Procedural distinctions also appear when the prosecutor decides not to seek a full investigation. Absent a state or Security Council referral, the prosecutor has an obligation only to notify the sources of the information that led to the examination.Footnote 27 These sources have no recourse or means to appeal the decision.Footnote 28 By contrast, the Statute affords the Security Council and member states that have referred situations the opportunity to request—via the pretrial chamber—that the prosecutor reconsider her decision.Footnote 29 The prosecutor therefore faces enhanced scrutiny of the preliminary examination in two contexts: (1) when she affirmatively decides to pursue a proprio motu situation; and (2) when she decides not to pursue a referral situation.
Given clear procedural and plausible substantive distinctions between proprio motu and referral situations, the OTP might have articulated an explicit preference for opening investigations in referral situations. Instead, it has rejected the notion that a lesser degree of scrutiny should apply for state or Security Council referrals.
As required by the Statute, the Office's preliminary examination activities are conducted in the same manner irrespective of whether the Office receives a referral from a State Party or the Security Council, or acts on the basis of information on crimes obtained pursuant to article 15.Footnote 30
This interpretation amounts to an assertion of the OTP's independence in conducting preliminary examinations and determining which merit full investigation. I argue below that the prosecutor has struggled to maintain this interpretation in practice.
II. The Court's Divergent Record
To date, the prosecutor's office has made public twenty-three preliminary examinations. The table below lists all known preliminary examinations by their current status:
Below, data on the outcomes and duration of preliminary examinations are disaggregated into referral and proprio motu categories.
Table 1. Preliminary Investigations by Status
Table 2. Preliminary Examinations of State or UNSC Referrals
The picture changes considerably when prosecutor-initiated preliminary examinations are considered.
Table 3. Preliminary Examinations of Proprio Motu Situations
Several patterns emerge. First, preliminary examinations initiated by the prosecutor rather than by referral rarely lead to full investigations. Of the fourteen preliminary examinations initiated by the prosecutor, only three (21 percent) have led to full investigations. And in one of those three cases (Cote d'Ivoire), the sitting government had affirmatively sought an investigation through the use of an Article 12(3) declaration, making the situation functionally equivalent to a referral. By contrast, seven of the eight examinations initiated by state or Security Council referral have been followed by full investigations (87 percent). The pace at which a final determination is made also varies dramatically between the two sets of situations. The average duration of a preliminary examination initiated by the prosecutor is almost five years. The average duration of a preliminary examination initiated by state or UN Security Council referral is less than eight months. In cases of Security Council referral, the prosecutor has moved particularly fast. The Darfur preliminary examination lasted about three months, and the Libya inquiry became a full investigation in a matter of days.
The differences in the prosecutor's approach to these groups of situations are stark enough to suggest that, effectively if not formally, there are two distinct preliminary examination processes. In cases of state or Council referral, the office moves relatively quickly and with an eye to opening a full investigation. Without state support, however, the prosecutor proceeds slowly and is disinclined to open a full investigation. There are a range of possible explanations for the disparity in the outcomes of preliminary examinations, and this section considers several of the most likely.
Substantive: The Scale of the Alleged Crimes
The simplest—and, legally, the most satisfactory—explanation for the divergence would be differences in the nature or scale of the alleged crimes. The ICC is designed to address the world's most serious crimes but must prioritize certain situations over others. It is possible that situations involving the most severe and widespread crimes arrive at the Court either via state or Security Council referral while the prosecutor herself tends to initiate inquiries into somewhat less grave situations. There are several reasons this might be the case. First, some of the most serious situations can only reach the Court via Security Council referral, which is required when there is no territorial or nationality basis for Court involvement. Second, member states facing serious and sustained civil conflict may see Court involvement as a way of attracting broader international support and may therefore refer those situations to the Court.Footnote 33 Finally, the OTP's practice of soliciting state referrals may have encouraged referral in situations where the prosecutor had already determined that a situation was grave enough to warrant Court involvement. In the Democratic Republic of Congo and Uganda, the OTP has described “the use of proprio motu powers to identify a situation followed by a [solicited] referral by the territorial State … .”Footnote 34 This dynamic means that a full investigation effectively prompted by the OTP's independent action would be formally categorized as a referral.
Assessing whether relative severity accounts for the disparity in outcomes raises several conceptual and methodological challenges. A wide range of possible crimes come within the Court's jurisdiction, and the Statute provides little guidance on whether certain of these crimes should be considered the most serious. The concept of “gravity,” referenced at several points in the Statute, is the most relevant to this inquiry, yet neither the Statute nor the Court have made clear how to determine the gravity of given situations. In early discussions of the concept, the prosecutor outlined several relevant factors:
The most obvious of these is the number of persons killed as this tends to be the most reliably reported. However, we will not necessarily limit our investigations to situations where killing has been the predominant crime. We also look at number of victims of other crimes, especially crimes against physical integrity. The impact of the crimes is another important factor.Footnote 35
The OTP's description of its gravity analysis has changed relatively little since that time. According to a 2007 policy paper, “the Office considers the scale of the crimes, the nature of the crimes, the manner of their commission and their impact.”Footnote 36 That formulation is repeated almost verbatim in the more recent OTP analyses.Footnote 37 ICC judges have also entered the debate on gravity, but ultimately have insisted that that the term must be interpreted flexibly and is not susceptible to a precise formula.Footnote 38 Most analyses have concluded that gravity should incorporate both quantitative and qualitative factors.Footnote 39
Here I use estimated civilian deaths as a proxy for the severity of a situation. Civilian deaths are defined as those deaths produced as a direct result of violence and exclude broader measures incorporating, for example, excess mortality rates and deaths as a result of deteriorating health and sanitation conditions. This metric is imperfect in several respects. It does not take into account grave crimes, including torture and rape, which do not necessarily produce civilian deaths. It does not distinguish between accidental killing of civilians (which might not be criminal) and intentional killings (which likely are criminal). It also omits consideration of important war crimes such as execution of military personnel who have surrendered. Yet number of civilian deaths is a measure that the Court and other international actors use frequently, and it allows for a broad comparison of very different situations. For full investigations, the estimates provided are for civilian deaths between the time the Court acquired jurisdiction and the time the investigation was launched. For ongoing or closed preliminary examinations, the totals are for civilian deaths for the period during which the ICC likely has jurisdiction. Whenever possible, the OTP's own estimates of civilian deaths are employed.
Table 4. Comparative Gravity of Referrals and Proprio Motu Situations
The data suggests that the relative severity of conflicts does help to explain the divergence in preliminary examination outcomes. Several proprio moto situations have markedly lower civilian death totals than those triggered by referrals. The situations in Guinea, Honduras, Korea, Iraq (taking into account significant jurisdictional limitations), and Venezuela have produced relatively low civilian death tolls. In several of these situations, there were allegations of other serious crimes, including torture and rape, but these crimes were also alleged to have occurred on a relatively limited scale.
There remain significant anomalies, however. Afghanistan has been the most salient; the civilian death toll there has long exceeded that in most other situations the ICC has chosen to investigate and yet the prosecutor has deliberated for more than nine years. Georgia also remained a preliminary examination for a period far in excess of the average. Colombia and Nigeria, which remain in the preliminary examination phase, feature civilian death tolls that are comparable to situations that have been investigated. These situations make consideration of other possible explanations for the pattern important, even if they are not susceptible to a similar kind of analysis.
Procedural: Admissibility and the Pretrial Chamber
The Statute's differing procedural requirements, discussed above, represent another possible explanation for the divergence between proprio motu situations and referrals. One important procedural factor is the OTP's admissibility analysis, and particularly its assessment of whether a state is undertaking its own investigation of alleged crimes. The OTP and the judges have determined that a self-referral that occurs in the absence of state action to prosecute relevant crimes constitutes an effective waiver of complementarity.Footnote 63 “The absence of national proceedings, i.e. domestic inactivity, is sufficient to make the case admissible.”Footnote 64 That interpretation significantly simplifies the OTP's analysis during the preliminary examination and may therefore facilitate the move to a full investigation in referral contexts. In proprio motu situations, by contrast, the OTP has often spent considerable time and effort monitoring national proceedings and attempting to determine whether they constitute an adequate response. The prosecutor has also embraced the concept of “positive complementarity,” which encourages the prosecutor not simply to determine whether adequate national investigations are occurring but to encourage them, sometimes through extended dialogue.Footnote 65 Territorial states that oppose ICC involvement can draw out the complementarity analysis by altering their accountability policy or announcing new initiatives that require further OTP analysis.
The role of the pretrial chamber in proprio motu situations may be even more consequential. As discussed, the prosecutor must seek authorization from a pretrial chamber before opening a full investigation proprio motu, a requirement that does not exist for investigations triggered by referral. One of the drafters of that provision has argued that the requirement “was intended to provide judicial ‘internal’ safeguards for the Prosecutor's decisions and compensate for the absence of a referral from external actors.”Footnote 66
The judges have not treated this process as a mere formality. The situation in Kenya presented the judges the first opportunity to consider whether to authorize a proprio motu investigation. The chamber deliberated for four months before authorizing an investigation, and one of the three judges dissented. Footnote 67 German judge Hans-Peter Kaul acknowledged that the prosecutor faces a low evidentiary threshold but insisted that the Statute “requires a full, genuine and substantive determination” that is not “of a mere administrative or procedural nature.”Footnote 68 With respect to Kenya, Kaul was not convinced that a state or organizational policy of attacking civilians existed. Without this element, he did not believe the prosecutor had demonstrated a reasonable basis to believe that crimes against humanity had been committed.Footnote 69 More broadly, he worried that a court that did not rigorously police its reach might become a “hopelessly overstretched, inefficient” institution.Footnote 70
If the judges have indicated that they will scrutinize prosecutorial requests to open investigations proprio motu, they have also questioned prosecutorial decisions not to investigate in referral contexts. In 2015, the prosecutor announced her decision not to open an investigation of the 2010 Israeli raid on a flotilla of ships bound for the Gaza Strip. That incident was referred to the Court by the Union of the Comoros, an ICC member. Comoros sought reconsideration of the prosecutor's decision, and a panel of judges supported that request. In so doing, the judges criticized aspects of the prosecutor's reasoning for declining an investigation.Footnote 71 Together, the Kenya and Comoros experiences might give the prosecutor pause about seeking a full investigation in a proprio motu context or declining to investigate when there is a referral. In the context of limited resources and situations of comparable gravity, a prosecutor might be inclined to favor referral situations as the path of lesser resistance from the judges.
Practical: Investigations Without State Support
A third potential explanation for the pattern relates more to the feasibility of potential investigations than to procedural requirements. Proprio motu investigations may be rarer because the absence of a referral implies a lack of logistical and administrative support from key state actors. In an environment of state disinterest or hostility, the OTP may be simply unable to pursue an investigation even if it desires to do so. A full investigation generally requires access to the territory, the provision of security, government information, and official permission to conduct interviews. In some contexts, the OTP has suggested that the lack of state support might militate against a full investigation purely on these feasibility grounds:
Will the necessary assistance from the international community be available, including on matters such as the arrest of suspects? In short, will it be possible in all reality to initiate an investigation at all? These are not matters which need normally trouble a domestic prosecutor, but they are all relevant to an ICC prosecution and they all underline the necessity of State support for the Office of the Prosecutor in the bringing of any investigation.Footnote 72
The OTP has also noted the corollary: that when there has been a referral, and particularly when that referral comes from the state where the alleged abuses occurred, the office can be confident that some political will exists to support an investigation.Footnote 73 The office has also suggested that a referral from a state other than the territorial state has relevance in terms of the feasibility of an investigation. “Even if a referral comes from a third State not involved in the alleged crimes, the referral will indicate support for the involvement of the Court from that part of the international community.”Footnote 74 While they have not crystalized into an explicit policy of privileging referrals, these formulations strongly suggest that the OTP considers the likelihood that states will support investigations but also, more speculatively, their willingness to support eventual enforcement of arrest warrants.
The institutional logic of this approach is evident: a Court investigation in the absence of a referral may put the Court in an untenable position. The prosecutor may lack the state support required to pursue and secure indictments. Even if indictments do eventually issue, these may remain unenforced, weakening the Court's credibility. According to a former OTP official, “if we wade into [situations] without the support or enthusiasm of the states involved, it's not likely to succeed.”Footnote 75 Practical and procedural factors may also run together; if the OTP cannot secure cooperation from key states during a preliminary examination, it may lack the information necessary to satisfy the pretrial chamber's standards for a full investigation. Absent state support for an investigation, the prosecutor faces a difficult choice. She could seek to authorize an investigation nonetheless and insist that member states cooperate, perhaps even by publicly chiding them for failing to provide needed information and support. Alternatively, the prosecutor could simply wait until cooperation materializes, shifting her attention in the meantime to situations where an investigation is more feasible.
The OTP has interpreted the Rome Statute in ways that accentuate the prosecutor's already heavy reliance on state support during preliminary examinations and may render proprio motu investigations even less likely. First, the OTP has concluded that it “cannot invoke the forms of cooperation specified in Part 9 of the Statute from States” at the preliminary examination stage.Footnote 76 This interpretation of the Statute's provisions is surprisingly weak. Article 86 obliges member states to “cooperate fully with the Court in its investigation and prosecution of crimes within the jurisdiction of the Court.”Footnote 77 The use of the word “investigation” does raise the question of whether member state cooperation obligations apply while the prosecutor is deciding whether to fully investigate. Several observers see a narrow reading as consistent with broader state concerns during the negotiations.Footnote 78 For states, the reassurance that is provided by the pretrial chamber's role in authorizing proprio motu investigations might be reduced if the prosecutor can make legally binding demands of states during preliminary examinations. A broader interpretation is also plausible, however. Member states with no obligation to cooperate during preliminary examinations could withhold vital information with the express intent of preventing a full investigation, an outcome difficult to reconcile with the object and purpose of the Statute.Footnote 79
The OTP has also downplayed its own investigative authority during the preliminary examination phase. The Statute provides that the prosecutor may “take testimony at the seat of the court” during preliminary examinations. One commentator has noted the “exceptional” nature of the authority to take testimony “at a stage in which no investigation is open.”Footnote 80 To date, the OTP has not acknowledged using this authority during its preliminary examinations and has instead stated that information during examinations “is largely obtained from external sources rather than the Office's own evidence gathering powers. …”Footnote 81 Moreover, the OTP has concluded that as a statutory matter it “does not enjoy investigative powers at the preliminary examination stage.”Footnote 82 These interpretations on state obligations and the OTP's investigative authority put the prosecutor in a weak position during preliminary examinations and may, as a consequence, increase the practical obstacles to launching proprio motu investigations.
Political: Deferring to State Preferences
A related but distinct factor may also help explain the pattern of preliminary examination outcomes: the prosecutor may have decided to defer in most cases to the judgments of states about when the Court can play a useful role. The OTP has insisted that the interests of states will not influence its situation selection, but the pattern described above requires consideration of that possibility. The dramatic differences in the outcomes and lengths of preliminary examinations “give rise to the impression that the Prosecutor has been influenced by non-legal factors.”Footnote 83
There are several different ways in which state political preferences might bear on the OTP decisions regarding preliminary examinations. In a case of violence involving nonstate actors on the territory of an ICC member, the prosecutor may decide that a referral from that government is essential to its investigation having the political legitimacy it requires. Viewed from this perspective, a referral is a signal to the Court that an investigation would, in the judgment of the government, serve a useful purpose and not unduly disrupt ongoing political and social processes. The prosecutor, however, has several times indicated that the office cannot incorporate these considerations into its decision-making and that doing so is the responsibility of the Security Council. The OTP has insisted, for example, that the prosecutor cannot “assume the role of a mediator in political negotiations: such an outcome would run contrary to the explicit judicial functions of the Office and the Court as a whole.”Footnote 84 By privileging referral situations, however, the prosecutor may have identified a way to take those concerns into account without doing so in a formal sense.
Political considerations could also impact the OTP in a much broader sense. The office might consider, for example, the preferences of states with strong interests in a given situation. This could particularly be the case when those states are major or regional powers whose support the prosecutor needs in a variety of contexts. The Court was born with weak support—if not active opposition—from several major powers, including the United States, Russia, China, and India. In order to build a relationship with these and other important states, the prosecutor may seek to avoid situations where major powers do not desire Court involvement and to engage with situations where they do. For the prosecutor, prioritizing referral situations for investigations may indirectly serve this broader strategic purpose. If they are so inclined, powerful states can employ political, economic, and diplomatic pressure to influence whether and when other states refer situations to the Court. A prosecutor that privileges referrals may be indirectly rewarding those efforts. To consider whether and how these factors have played out in practice, this article now turns to a more detailed examination of the longstanding preliminary examination in Afghanistan.
III. The Afghanistan Situation
Since at least late 2001, a state of armed conflict has existed in Afghanistan. That conflict has passed through several distinct phases and has involved multiple actors, including the Afghan government, Taliban and other antigovernment forces (based both in Afghanistan and Pakistan), and forces from a UN-authorized multinational coalition. Throughout its evolution, the conflict has produced significant civilian casualties. The United Nations estimates that almost 25,000 civilians have been killed since 2009, when it began regular reporting.Footnote 85 Afghanistan acceded to the Rome Statute in February 2003, and the Statute came into force for the country in May of that year.Footnote 86 Given the high civilian toll and the Court's broad jurisdiction, Afghanistan constitutes a notable case of ICC inaction in the face of large-scale crimes. It merits closer examination not because it is representative of all situations, but because it may make more visible possible determinants of OTP decision-making.Footnote 87
ICC Involvement
From the time the ICC opened in July 2002 through mid-2009, Court officials released little information about the conflict in Afghanistan or potential Court involvement. In September 2009, then prosecutor Luis Moreno-Ocampo first offered detail by indicating that the office was considering a range of potential crimes, including intentional killing of civilians and torture. He also described the inquiry as “exceedingly complex.”Footnote 88 In December 2011, as part of its first comprehensive report on preliminary examinations, the OTP classified Afghanistan as being in Phase II of the examination process, during which the office focuses on whether relevant crimes have been committed. The office cited allegations of deliberate targeting of civilians, torture, attacks on humanitarian targets and other protected objects, and recruitment of child soldiers but described the process of verifying allegations as “challenging and time-consuming.”Footnote 89 A year later, the OTP's update on Afghanistan added several other categories of alleged crimes but repeated verbatim much of the 2011 analysis, including its description of the inquiry as particularly difficult. It also noted that resource constraints had significantly complicated its effort to verify allegations of crimes.Footnote 90
The OTP's 2013 and 2014 assessments of the Afghanistan situation provided much more detail. They determined that there exists “a reasonable basis to believe that crimes within the Court's jurisdiction have been committed … ” and moved the examination from Phase II to Phase III, during which the office addresses admissibility issues.Footnote 91 The OTP provided additional legal analysis of the conflict, including a determination that the conflict should be considered internal rather than international in nature.Footnote 92 The office also refined considerably the categories of alleged criminal activity under examination. The OTP found no reasonable basis to believe that pro-government forces had committed crimes against humanity or that they had intentionally targeted civilians. The office therefore left open only a few categories of crimes, including a range of potential crimes by Taliban forces and certain detention practices by the Afghan government and international forces. In its 2015 report, the OTP specifically identified U.S. detention practices as an area of interest and included brief analysis of how the United States had addressed those practices through internal investigations.Footnote 93 The next year, the OTP expanded its focus to include certain U.S. detention practices outside of Afghanistan (but related to the conflict) and indicated that its decision on whether to launch a full investigation was “imminent.”Footnote 94 But to date, no decision has been made.
State Cooperation
The Afghan government and states with forces in the country have given no indication that they support an ICC investigation. At least initially, the knowledge of Afghan officials about the ICC's potential jurisdiction was limited. According to the former prosecutor, a senior Afghan ambassador at first was not even aware that the country was a party to the Rome Statute.Footnote 95 For their part, foreign governments involved in Afghanistan have focused almost entirely on attempting to rebuild and reform the country's institutions.Footnote 96 No significant donor state or troop contributor has expressed support for an ICC role in the country or substantively addressed accountability for serious crimes. “The United States and most of its allies have been largely silent on the law. In fact, it appears that a desire for a quick exit by NATO has stifled all discussion of the critical need to link reconciliation with accountability and to tackle Afghanistan's longstanding culture of impunity.”Footnote 97
The evident reluctance of both the Afghan government and its international partners to support an ICC role has manifested itself in limited and slow cooperation with the OTP. When the prosecutor first discussed the Afghanistan examination he indicated that he would “welcome” information from any involved states about potential crimes committed in the country. According to former OTP officials, the office sent at least one formal request for information to states with forces in the country. These officials indicate that its requests were mostly met with silence.Footnote 98 In its formal reports, the office has characterized state cooperation carefully but has referred to the lack of state enthusiasm. In its 2011 report, the OTP noted only that it is in contact with the Afghan government. The next year, it reported that “several requests for information sent by the Office the past two years to various States, including the Government of Afghanistan and States with troops in Afghanistan, have been dismissed or remain pending.” The office did note that five states had replied to formal requests for information, although it did not characterize the quality of those responses.Footnote 99 In its most recent update, the office noted that it was still waiting for requested information from the Afghan government regarding national proceedings. It indicated that it had emphasized to relevant states that “effective cooperation is of the utmost importance for the work of the Office in this situation.”Footnote 100 In keeping with its interpretation of the Statute, the OTP has never suggested that member states who do not provide information to the office are violating their legal obligations.
Interactions with the United States
The United States has played a critical role in Afghanistan since early 2002 and continues to have the largest foreign troop presence in the country. U.S. forces have engaged in a broad range of military action to root out Taliban and Al Qaeda forces in the country and to support the Afghan government. The central U.S. role poses a potential diplomatic challenge for the prosecutor. Since early 2005, the United States has moved gradually but steadily from hostility toward the Court to support for most of its investigations. That shift has been important for the Court, and senior officials have publicly and privately welcomed the improved relationship.Footnote 101 However, there are indications that deeper ICC involvement in Afghanistan could complicate that relationship.
Several factors appear to be at play. The United States has devoted significant diplomatic resources to encouraging the stable development of Afghan politics. A Court investigation could complicate that process if it scrutinized the actions of former warlords and militia commanders now serving in government and their backers.Footnote 102 Likely as important from Washington's perspective, an ICC investigation could bring unwelcome scrutiny to certain U.S. practices. Independent observers and the Afghan government have complained repeatedly about targeting practices by international forces and the high level of civilian casualties. Human rights groups have also documented a series of alleged abuses by U.S. personnel on Afghan soil, primarily related to the detention and interrogation of individuals believed to be associated with the Taliban and Al Qaeda.Footnote 103
Since 2013, there has been an ongoing dialogue between senior OTP staff and U.S. officials about Afghanistan that is not reflected in official OTP reports.Footnote 104 As the prosecutor's office prepared the Afghanistan section of its annual report on preliminary examinations in 2013, OTP officials sent the United States material they were considering including in their new report.Footnote 105 That material directly referenced allegations of American abuse of detainees, mostly between 2003 and 2006. Receipt of that material created alarm in Washington and prompted the United States to dispatch several senior officials to The Hague, where they discouraged the OTP from including the material in their draft report.Footnote 106 The report issued by the OTP several months later did not directly reference allegations against U.S. forces. A year later, however, the OTP did specifically identify possible crimes by U.S. forces as an area of interest. That clarification in turn produced expressions of concern and disappointment from U.S. officials, who argued that the Court cannot exercise jurisdiction over U.S. nationals.Footnote 107 These interactions provide a glimpse of the informal pressures the OTP can face as it considers whether to convert a preliminary examination into a full investigation.
“Slow-Walking” an Investigation?
The very high civilian death toll in Afghanistan all but rules out gravity as an explanation for the Court's failure to open an investigation. Given the scale of the alleged crimes, it does not appear that procedural requirements have restrained the prosecutor from seeking a full investigation. While judges have carefully reviewed past prosecution requests to open proprio motu investigations, it is unlikely that a panel of judges would refuse to authorize an investigation in the very bloody Afghanistan conflict. Moreover, the existence of a state of armed conflict in Afghanistan significantly expands the range of crimes the prosecutor could pursue. With the broad category of war crimes available, the difficult question of organizational policy that troubled Judge Kaul as he considered crimes against humanity in the Kenya situation would not be determinative.Footnote 108 Practical and political factors are therefore the most likely factors in the decision to delay a full investigation.
The practical obstacles to an Afghanistan investigation are formidable, and former OTP officials have cited these concerns as a reason for the office's delay.Footnote 109 The difficult security environment means that conducting an investigation would pose significant challenges for the OTP in terms of the safety of both its staff and witnesses. The situation in Afghanistan may also be less susceptible to investigation from outside than was the situation in Sudan, where the OTP was able to collect significant testimony from a large and mostly confined refugee population in neighboring Chad. Yet the Court's quick decision to launch an investigation in Libya—in the midst of intense conflict and without the support of the then government—stands as an important precedent. In the right political context, the OTP has proved willing to proceed with investigations even in the face of daunting practical obstacles. It is difficult to avoid the conclusion that political factors have at least contributed to the prosecutor's slow pace in Afghanistan.Footnote 110
Conclusion
The ICC's situation selection has emerged as a critical issue for the Court's legitimacy and has produced sharp critiques from both certain governments and outside observers. This article has identified the preliminary examination process as a critical piece of the situation-selection puzzle. It has demonstrated that the Court is much less likely to initiate investigations in proprio motu situations than when there are referrals. It has explored a range of possible explanations for this pattern, including substantive, procedural, practical, and political factors. A precise weighting of these factors is not possible, but the available evidence suggests that substantive and procedural factors do not fully account for the disparity and that practical and political factors may have discouraged the OTP from at least certain proprio motu investigations. The gap between the Court's claim of evenhanded treatment of preliminary examinations and its record may grow. The difficulties that have surrounded the Court's Kenya investigation—one of only two investigations launched without either a Council referral or territorial state request—will likely serve as a cautionary experience for the Court.Footnote 111 The prosecutor's record thus far stands as a notable pattern for the study of the Court and, more broadly, for inquiries into how formally independent international officials manage their relationships with states.
For the Court, this pattern forms part of a broader policy challenge: proving itself effective while also maintaining an appropriate distance from state power and influence. The OTP's unwillingness to acknowledge an evident preference for referrals suggests sensitivity about the office's independence and the perception that it might be yielding to state interests; a Court that responds primarily to state requests may appear to be less a pillar of global justice than a provider of judicial services to member states and to the Security Council. For the prosecutor, the most ambitious alternative to the current practice would be pursuing more full investigations absent a referral. To do so, however, the OTP would likely need to bolster its investigative capacity during preliminary examinations and perhaps even revisit its interpretation of its own authority and state obligations during this phase. More robust preliminary examinations would help the prosecutor proceed in an environment of state passivity or opposition and provide the type of evidence that the judges have required before authorizing investigations. Seeking to launch more proprio motu investigations would likely increase political friction between the Court and a number of states. If the prosecutor pursues this route, there is a high probability that several investigations will fail—and the distinct possibility that some states will seek to undermine the Court.
The Court may be better served simply by showing greater openness about its limitations. Explicitly acknowledging a preference for referral situations need not exclude the OTP from relevance in nonreferral contexts. Through its embrace of positive complementarity, the OTP has recognized that a preliminary examination can be more than a mere precursor to an investigation. As Christopher Keith Hall has written, “[t]he Court was not to be simply a passive institution, simply waiting for the Security Council, States parties and Article 12 (3) States or others to seize the Court.”Footnote 112 The OTP has noted that the preliminary examination phase “offers a first opportunity for the Office to act as a catalyst for national proceedings” and that the office can “better identify the steps required to meet national obligations to investigate and prosecute serious crimes.”Footnote 113
Acknowledging a reluctance to launch proprio motu investigations would inevitably diminish the prosecutor's implicit threat to investigate situations on her own. But the Court's catalytic potential does not rest entirely on that possibility, and the OTP could still generate public awareness, share expertise, and, in certain circumstances, assist national judiciaries. During several long-running preliminary examinations, including in Colombia and Nigeria, the Court has interacted repeatedly with national authorities to monitor and encourage accountability. Whether and when those interactions are effective—and to what degree effectiveness hinges on the perceived threat of a full investigation—deserves greater attention.
