2. Is the prosecutor an advocate or a minister of justice?

In adversarial systems, many prosecutorial obligations reflect the prosecutor’s role as a minister of justice. ^{Footnote 5} The precise content of the term “minister of justice” is elusive but, broadly, it entails the idea that the prosecutor “represents the sovereign and must make decisions for society at large—not for any individual client” (Griffin and Yaroshefsky, Reference Griffin and Yaroshefsky2017, p. 304). Accordingly, the prosecutor must not be led solely by the victims’ or police’s interests, or by a single-minded desire to secure convictions (Mwalili, Reference Mwalili1997). ^{Footnote 6} Rather, she must strive towards processes and outcomes that are for the common good. Some illustrative examples lend weight to the idea that the adversarial prosecutor is not simply a combative adversary to the defendant. In Australia and in England and Wales, “public interest” is one of the grounds on which the prosecutor decides whether the case should proceed to trial (Cps.gov.uk, 2018, [3.4]; Cdpp.gov.au, 2021, p. 5). Similarly, under the US Attorneys’ Manual, the prosecutor’s decision to press charges represents a judgment that “the fundamental interests of society” would be served by doing this (Justice.gov, 2018, [9-27.001]). The Indian Supreme Court finds that a prosecutor “is really a minister of justice … [s]he is not a representative of any party.” ^{Footnote 7} The Public Prosecution Service of Canada develops prosecutorial obligations framed by the explicit recognition of the prosecutors’ quasi-judicial role as “ministers of justice” (Ppsc-sppc.gc.ca, 2014, Chapters 2.2, 2.3, 2.9). These examples, which could be multiplied manifold, highlight that prosecutors in adversarial jurisdictions serve the public interest at large and are not expected to be motivated solely by a desire to secure convictions.

While the term “minister of justice” is sometimes used interchangeably with “officer of the court,” these are distinct but overlapping concepts. The latter term includes all lawyers and members of the court’s staff who are professionally obligated to serve the court with integrity. “Minister of justice” refers specifically to the prosecutor and describes her distinctive obligations to serve the public interest. Some of these obligations are shared with other officers of the court; others are not. For example, as officers of the court, all lawyers, including defence counsels, are usually obligated not to mislead the trier of fact with fabricated evidence. However, for the prosecutor, this obligation arises not just from being an officer of the court, but also from being a minister of justice. Further, as ministers of justice, prosecutors bear enhanced responsibilities. For instance, prosecutors in most adversarial jurisdictions are required to disclose material evidence to the defence counsel. This obligation arguably stems from the prosecutor’s minister-of-justice role (Plater, Reference Plater2006). The defence counsel generally does not bear an equally onerous reciprocal obligation to share evidence with the prosecutor. ^{Footnote 8}

Along with being a minister of justice, the typical adversarial prosecutor is also responsible for carrying out “vigorous advocacy” (Civiletti, Reference Civiletti1979, p. 20; Heinze, Reference Heinze, Wright, Levine and Gold2021, p. 136). This obligation arises from the oppositional nature of adversarial systems. Trials in adversarial jurisdictions are popularly understood as contests between equal and opposite adversaries (Bazuaye and Oriakhogba, Reference Bazuaye and Oriakhogba2016, p. 130). A popular, though imperfect, metaphor is that of a football match. The judge is the neutral umpire who ensures that the relevant rules are enforced and the contest is fairly fought (Heinze and Fyfe, Reference Heinze, Fyfe, Ambos, Duff, Roberts, Weigend and Heinze2020, p. 348). ^{Footnote 9} The prosecutor and the defence counsel are the two metaphorical teams. Adversarial logic requires that they compete in opposition to each other, within the rules of the game. According to this logic, it is only when both sides have the opportunity to expose each other’s strengths and weaknesses that the judge can arrive at the optimally reliable outcome (Uviller, Reference Uviller2000, p. 1698). Thus, for an adversarial system to work, it is necessary that there be a contest between the prosecutor and defence counsel.

In adversarial theory, the fact-finding exercise will only yield a reliable outcome where the parties pitted against each other are on a level playing field. To this end, adversarial systems have evidentiary, procedural, and constitutional rules to mitigate the structural “inequality of arms” between the prosecution and the defence. Whether the defendant’s time and resources are outweighed by a particular prosecutor’s office varies. Nonetheless, in all cases, the prosecutor has ready access to the state’s resources, including the police. ^{Footnote 10} The prosecutor also benefits from the social capital associated with being the state’s lawyer (Green, Reference Green1999, p. 609). She can appeal to judges’ “desire to be protected against harm to person or property” (Civiletti, Reference Civiletti1979, p. 20). Further, the judge might be invested in maintaining a smooth working relationship with the prosecutor, particularly where the prosecutor is associated with a particular courtroom rather than a specific case file. This may be offset by the defence counsel also being a repeat player in the courtroom. But, on average, the prosecutor is in a more advantageous position than the defence counsel. ^{Footnote 11} This imbalance is objectionable in principle since it pits two unequal adversaries against each other. However, it is also objectionable on the consequential basis that it distorts the decision-making process and could lead to miscarriages of justice (Brants and Field, Reference Brants and Field2016, p. 267). Accordingly, adversarial systems are designed to uphold several civil liberties, many of them in the form of constitutional safeguards (Heiner, Reference Heiner2016, p. 601). For instance, the burden of proof rests on the prosecution; the prosecution must prove its case beyond reasonable doubt; and so on.

Within adversarial systems, there is an inherent tension between the twin roles of the prosecutor: as a minister of justice and as an advocate. On the one hand, the minister of justice does not represent partisan interests, but works for the common good. On the other, the advocate’s role must be a partisan one, which is developed in opposition to the defendant. These two demands on the adversarial prosecutor can be inconsistent with each other (Sklansky, Reference Sklansky2018, p. 461; Griffin and Yaroshefsky, Reference Griffin and Yaroshefsky2017, p. 313). Yet, few attempts have been made to understand or address this paradox. Formal guidance tends to restate these roles, without identifying their potentially contradictory nature. For example, the Indian Supreme Court finds that

Similarly, in Shaikh, the Supreme Court admonishes the prosecutor, observing that, while the latter “is not supposed to be a persecutor, yet the minimum that was required to be done to fairly present the case of the prosecution was not done.” ^{Footnote 13} In Sud, the Indian Supreme Court holds that the prosecutor is “required to prosecute with detachment on the one hand and yet with vigour on the other.” ^{Footnote 14} These decisions do not acknowledge, much less explore, that the obligations summarized can counter each other.

In some jurisdictions, such as in England and Wales, there is delegated legislation to guide the exercise of prosecutorial discretion. The Code for Crown Prosecutors issued by the Director of Public Prosecutions requires that “prosecutors must always act in the interests of justice and not solely for the purpose of obtaining a conviction” (Cps.gov.uk, 2018, [2.7]). This Code seems to emphasize justice over advocacy—an interpretation with some academic support (Hoyano, Reference Hoyano2019, p. 111). The House of Lords has also described prosecutors as “independent ministers of justice,” although the emphasis on that occasion was on establishing prosecutors’ independence from the executive. ^{Footnote 15} Nonetheless, the advocacy role of the prosecutor cannot be discounted entirely, given that the English and Welsh criminal justice system is an adversarial one. The Code also acknowledges this by recognizing that prosecutors can advocate for a conviction, though this must not be their sole purpose. However, like other formal authorities, the Code does not investigate the tension between the twin prosecutorial roles.

In India, guidance on the prosecutor’s role is not available consistently across all states. Given that published prosecutorial guidance can inspire public confidence in prosecutors, enhance the consistency and fairness of prosecutorial action, and provide a foundation on which to hold prosecutors to account, it may seem curious that more Indian states do not provide such guidance. ^{Footnote 16} However, there are other jurisdictions, such as Cyprus, where prosecutorial guidelines analogous to the Crown Code are not instituted (Kyprianou, Reference Kyprianou2010, pp. 37, 149–50). Within these legal systems, the introduction of prosecutorial guidelines is regarded as an unnecessary and unacceptable interference with prosecutorial discretion, which could get in the way of taking specific characteristics of a defendant or case into account (Kyprianou, Reference Kyprianou2010, pp. 37, 149–50). It is possible that a similar rationale explains the absence of detailed prosecutorial guidance in India. Where such guidance is available in India, it either underemphasizes one of the twin roles or fails to address and resolve the possible contradiction between them. The Himachal Pradesh guidance, for instance, understands the mission of the prosecutor as being to secure high conviction rates (Himachal.nic.in, 2016). While this document does encourage prosecutors to “[e]nsure proper investigation … and secure compensation to the victim,” there is no broader recognition of the prosecutor’s role as a minister of justice beyond this (Himachal.nic.in, 2016). In contrast, the guidelines issued by the state of Odisha cast the prosecutor, in explicit terms, only as a minister of justice (Dppodisha.nic.in, no date). This characterization overlooks the prosecutor’s adversarial role on which the trial is premised, as brought out earlier in this section.

A closer look at the Odisha guidelines reveals that the adversarial role of the prosecutor is recognized impliedly. The guidelines expressly acknowledge that there can be no fair trial in which the prosecutor “acts in a manner as if he was defending the accused” and state that the prosecutor should “try to prove all relevant facts necessary for the proving of the case of [p]rosecution” (Dppodisha.nic.in, no date). At one end of the adversarial extreme, the guidelines mandate that the prosecutor must oppose the release of seized property where such a release would impede a successful prosecution; and where the allegation relates to non-bailable offences, the prosecutor should oppose the grant of bail “to the best of his ability” (Dppodisha.nic.in, no date). ^{Footnote 17} However, as with other guidance, while the prosecutors’ twin roles are reflected in the Odisha guidelines, these guidelines do not provide a more general account of what should happen when the twin prosecutorial roles pull away from each other.

A similar omission characterizes other national and international guidance. For example, the Northern Irish Code for Prosecutors describes the prosecutor’s role at trial to be guided by the “interests of justice” (Ppsni.gov.uk, 2023, [5.1]). However, the same paragraph also requires prosecutors to “prosecute to the full extent of their abilities within the law and rules of evidence and procedure,” emphasizing advocacy for conviction (Ppsni.gov.uk, 2023, [5.1]). What is not acknowledged in this guidance is that prosecuting to the fullest extent might be contrary to the public interest. Similarly, the UN Guidelines on the Role of the Prosecutors require the prosecutor to play an “active role” in instituting prosecutions, particularly—though not exclusively—for crimes committed by public officials (Ohchr.org, 1990, [11], [15]). Simultaneously, these guidelines recognize the prosecutor’s commitment to the public interest (Ohchr.org, 1990, [13b]). In the same vein, the standards adopted by the International Association of Prosecutors (IAP) indicate that a case should be “firmly but fairly” prosecuted, thereby recognizing elements of the prosecutors’ dual roles but not their possible contradiction (Iap-association.org, 1999, [4.2e]). These guidance documents do delineate clear duties under some specific circumstances in which prosecutors might need to prioritize one role over the other, such as by obliging prosecutors to disclose all relevant evidence to the defendant (i.e. prioritizing justice over advocacy when it comes to disclosure). ^{Footnote 18} However, they do not provide a more general understanding of how the prosecutor should resolve possible conflicts between being a minister of justice and an advocate for convictions.

The academic literature does not take us much further. Within the scholarship on prosecutorial discretion, there is a heavy focus on prosecutorial decision-making about whether the defendant should be prosecuted and, if so, what charge or sentence recommendations the prosecutor should pursue. ^{Footnote 19} Extant research also emphasizes and analyzes prosecutors’ decisions to pursue alternatives to the trial such as restorative justice or plea bargaining. ^{Footnote 20} In analyzing prosecutors’ approaches to charging or sentencing the defendant, some scholars are particularly critical of the prosecutor’s role in bringing about wrongful convictions or other miscarriages of justice. ^{Footnote 21} In addition, apart from scrutinizing legal factors, the academic literature assesses the extent to and manner in which discriminatory prosecutorial attitudes, including sexist and racist ones, influence the exercise of prosecutorial discretion. ^{Footnote 22} Finally, existing scholarship evaluates how broader structural factors—such as institutional independence, ^{Footnote 23} systems of accountability, ^{Footnote 24} promotion and rewards systems, ^{Footnote 25} and even broader political conditions ^{Footnote 26} —affect the exercise of prosecutorial discretion. Thus, most studies on prosecutorial discretion focus on either specific prosecutorial decisions or particular influences on the exercise of prosecutorial discretion. Besides the research reviewed later in this section, the academic literature for the most part does not address the potentially contradictory pulls placed by the dual roles of the adversarial prosecutor. ^{Footnote 27} Even where both roles are explicitly recognized in the research, they are seldom acknowledged as being in tension with one another. For example, with respect to Indian prosecutors, Sharma tasks the prosecutor with securing convictions, while simultaneously proposing that she “need not be overwhelmingly concerned with the outcome of the trial,” and that she “is required to present a truthful picture before the court” (Sharma, Reference Sharma1997).

Nonetheless, there are notable exceptions that address the central concern of this paper. Some scholars suggest that the tension between the twin roles should be resolved by choosing one over the other. For example, Fish argues that the prosecutor’s advocacy role should be done away with and adversarial trials should be recast in the image of their inquisitorial counterparts (Fish, Reference Fish2018, p. 1440). His proposal draws specifically upon inquisitorial jurisdictions, including France and Germany. Similarly, Bellin suggests that a “non-combative prosecutorial mindset … may be the most valuable thing American legal systems can import from continental Europe.” ^{Footnote 28} Reasoning that casts the prosecutor only as a non-adversarial minister of justice also underpins obiter dicta from the Delhi High Court. ^{Footnote 29} On the other end of the spectrum, Bennet Gershman argues that

Based on the adversarial theory outlined in this section, neither approach is tenable. Dispensing with the prosecutor’s advocacy function is undesirable for legal systems that are committed to an adversarial evidential, procedural, and constitutional framework, and do not want to overhaul their trials into inquisitorial ones. Perhaps this broader overhaul is desirable. Some scholars have argued that the commitment to adversarial trials is ill-founded and often leads to inaccurate outcomes (Zalman and Grunewald, Reference Zalman and Grunewald2015). Yet, transforming the prosecutor’s role without changing other aspects of the criminal justice system is like trying to fit a square peg into a round hole. At the same time, thinking of the prosecutor only as an advocate could have a high cost in terms of miscarriages of justice, given the usual inequality of arms between the prosecutor and the defendant.

Another approach could be to divide the prosecutor’s office into different units or cadres so that no official is simultaneously expected to be a minister of justice and an advocate (Uviller, Reference Uviller2000, p. 1698; Barkow, Reference Barkow2008). However, this approach appears unwieldy. Imagine a case in which the police secure clinching evidence of the defendant’s guilt halfway through the trial. The prosecutor suspects, but does not know, that this evidence has been obtained under circumstances that might render it inadmissible. The defence counsel is unaware of the doubtful admissibility. Should the advocate-prosecutor use the evidence, since she does not belong to the minister-of-justice unit? Should she request an adjournment to seek advice from the appropriate unit? The first option could lead to a miscarriage of justice, the second to undesirable delays. Proposals to bifurcate the prosecutor’s office would also lead to inefficiency, with a larger pool of prosecutors being drawn into decision-making about the same case. Further, such proposals do not shed light on which of the twin prosecutorial roles should influence prosecutor-victim engagement. We are left to assume that the advocate-prosecutor has no obligations towards the victim. This accounts for neither the victim’s special interest in the case nor the value of the victim’s testimony to the advocate-prosecutor’s case.

Finally, some scholars argue in favour of a differentiated understanding of the prosecutor’s role. Prominent among them is Fred Zacharias, who proposes that

According to Zacharias, a breakdown requiring the prosecutor to perform a non-adversarial role can happen in several ways, such as where the trial is marked by unequal adversaries or the tribunal is biased. Zacharias’s work focuses on the litigative role of the prosecutor. Against this background, his default presumption is that the prosecutor should serve as an advocate unless there is special reason not to do so. In many ways, this is the inverse of the model proposed in this paper. This is because this article is concerned not just with the litigative role of the prosecutor, but also with the advocacy-justice dilemma facing her more generally, including in her interaction with the victim. Nonetheless, Zacharias’s analysis is one of the clearest articulations of how prosecutorial functions relate to the adversarial nature of trials. It provides a rich foundation for the framework developed in the next two sections.

3. A dynamic account of the prosecutor’s role

I argue that a prosecutor’s role is best conceived as a dynamic one. Thus, the tension between the prosecutor’s twin roles should be resolved depending on what power the prosecutor is exercising at any given time, in line with the following principles. These principles are sketched out below but exemplified and developed in the next section.

4. Prosecutors in Delhi trial courts

This section draws from fieldwork conducted in trial courts in Delhi to exemplify and justify the dynamic theory of prosecutorial roles. It introduces the methodological and legal framework underpinning the empirical research, before engaging with the substantive findings and arguments.

The fieldwork was conducted from April to October 2016. Ethical approval to do so was obtained from the University of Oxford. The fieldwork was conducted as part of a research project to evaluate factors associated with adjudication in rape cases. It entailed observation of proceedings in special fast-track courts that have been established to prosecute prominent sexual offences. ^{Footnote 35} With the relevant judge’s permission, one such court was observed in each of Delhi’s six existing trial court complexes, for a period ranging from seven to ten days. Typically, brief simultaneous notes were made during the proceedings, with fuller descriptions being typed out over the following few days. Primary data were also collected through interviews with the participants described in Table 1.

Table 1. Summary of research participants interviewed

^a The Delhi Commission for Women (DCW) is a statutory body established under the Delhi Commission for Women Act 1994 to promote women’s development.

^b This was done for disabled or child victims, as I did not have ethical clearance to speak to them directly.

Usually, semi-structured qualitative interviews were carried out in Hindi or English, although ten interviewees participated using written questionnaires. On average, oral interviews were 39 minutes long.

Interviewees were recruited through a range of methods. For example, some victims or members of victims’ families were approached in court. Others were contacted through lawyers, police officers, or victim-support organizations. Similarly, some prosecutors were approached directly in court; the remainder were introduced to me by judges, defence counsels, or by other prosecutors. Lawyers apart from prosecutors were recruited through personal or professional networks, “snowballing,” victim-support organizations, the Delhi Legal Services Authority (DLSA), and, where they were well known for providing representation in rape cases, through cold-calling. ^{Footnote 36} Judges were interviewed through questionnaires administered by the Delhi High Court on my behalf. Victim-support organizations were located through targeted Internet searches and contacted directly through phone calls or emails.

All interviews were transcribed and, where relevant, translated by me. Collectively, the court observation and interviews generated a voluminous record, which was read over multiple times to identify relevant themes. The arguments developed in this article are based on that thematic analysis. More details of the research process are available elsewhere. ^{Footnote 37}

Having introduced the research methods, a summary of the procedural framework for rape prosecutions in Delhi will help in understanding the following empirical examination. Most commonly, proceedings are initiated when information about the offence is reported to the police under section 154 of the Code of Criminal Procedure 1973 (CrPC). Thereafter, the police investigate the allegations and prepare a report (“charge sheet”) in which they record their determination of whether the evidence discloses the commission of an offence. ^{Footnote 38} The report is submitted to the magistrate, who takes “cognisance” (judicial notice) of the offending where appropriate. ^{Footnote 39} The case is then “committed” (transferred) to a sessions court, where a judge decides which charges it is appropriate to “frame” (lay), if any (sections 209 and 228, CrPC). ^{Footnote 40} The prosecution is based on these charges.

The prosecutor does not decide whether charges should be framed, but she is still entitled to withdraw the prosecution with the judge’s permission. ^{Footnote 41} The Supreme Court has explicitly connected this power of withdrawal to the prosecutor’s role as a “minister of justice,” ruling that proceedings can be withdrawn to ensure “public justice, public order and peace.” ^{Footnote 42} The interests of public justice may transcend “the legal justice of the particular litigation.” ^{Footnote 43} For example, the prosecutor may withdraw proceedings if pursuing them would reignite settled labour disputes. Further, the nature and gravity of the alleged offence are relevant when assessing whether the prosecutor should withdraw proceedings. ^{Footnote 44} However, the prosecutor’s decision to withdraw must be made impartially and independently of governmental control. ^{Footnote 45} The prosecutor must also inform the court of the reasons underlying the withdrawal. ^{Footnote 46} While the withdrawal is subject to judicial permission, this permission can only be denied if the prosecutor has acted improperly, in bad faith, or with ulterior motives; where she has attempted to interfere with the normal course of justice for illegitimate reasons; or where the grant of judicial permission would result in manifest injustice. ^{Footnote 47} Thus, in theory, the prosecutor’s power of withdrawal is an expansive one. In practice, it is frugally exercised inter alia because of the managerialism characterizing the prosecution service. The prosecutorial reluctance to exercise the power of withdrawal is analyzed more fully later in this section.

While decision-making powers in the criminal process rest mainly with judicial officers, the prosecutor nonetheless presents arguments relating to all key decisions, including which charge should be framed, whether the defendant should be granted bail, and what sentence should be imposed on a convicted defendant. For our purposes, it will be useful to investigate which of the twin prosecutorial roles underpins the exercise of prosecutorial discretion in such situations.

Further, in rape cases, the prosecutor is mandatorily assisted by a lawyer from the Delhi Commission for Women (DCW) ^{Footnote 48} and may be further assisted by a private counsel hired by the victim or accessed by the victim through legal aid schemes. ^{Footnote 49} Nonetheless, the focus of this section remains on the public prosecutor, who leads the prosecution and decides the litigation strategy in every case. ^{Footnote 50}

There is no instrument analogous to England and Wales’ Code of Crown Prosecutors or the US Department of Justice Manual applicable in India to provide guidance to prosecutors on how they should exercise their powers; nor is such guidance available through statutes. Appellate court rulings provide some indication of the principles that should guide the discharge of prosecutorial duties. As per these rulings, prosecutors must be guided by the need to convict the guilty, acquit the innocent, maintain law and order, and uphold the rule of law. ^{Footnote 51} They must ensure fairness in the proceedings so that “justice prevails.” ^{Footnote 52} Thus, according to the Supreme Court,

Accordingly, prosecutors must eschew “excessive zeal” to convict suspects, even where the police seek a conviction. ^{Footnote 54} These rulings emphasize important values, such as fairness and institutional independence. The importance of these values is echoed in high court rulings, which create the expectation that prosecutors remain independent and non-partisan actors. ^{Footnote 55} For instance, where the prosecutor argues that the death penalty should be imposed on the defendant, the prosecutor must adduce evidence relating to both mitigating and aggravating sentencing factors before the court. ^{Footnote 56} In such cases, the state must provide a psychiatric and psychological evaluation report that is focused on the defendant as well as other relevant information relating, for example, to the defendant’s socioeconomic background and criminal antecedents, or lack thereof. ^{Footnote 57}

Nonetheless, extant rulings do not explore the prosecutors’ role in adequate detail. For example, at what point does the advocate-prosecutor’s “zeal” to convict suspects become “excessive”? Does the prosecutor’s obligation to act with “fairness” include a minister-of-justice duty to the victim? How should the prosecutor navigate between her potentially contradictory obligations? The following empirical analysis will delve deeper into these abstract concepts, deploying original primary data to exemplify and justify the proposed dynamic model. Since the fieldwork did not cover sentencing processes, the discussion will concentrate on the first four principles of the model developed in the previous section. To recap, the dynamic account of the prosecutor’s role is summarized again:

1. 1. In general, the prosecutor should operate as a minister of justice.

2. 2. When the prosecutor is facilitating the competitive fact-finding exercise on which adversarial systems are based, she should operate as an advocate. Nonetheless, while engaging in advocacy, the minister-of-justice role should continue to constrain the prosecutor. Such constraints should apply where they are necessary to ensure the conduct of legitimate and reliable adversarial proceedings.

3. 3. In her advocacy role, the prosecutor should advocate for the conviction of the defendant.

4. 4. Prosecutor-victim relations should be governed by the prosecutor’s role as a minister of justice.

5. 5. In sentencing proceedings, the prosecutor should act as a minister of justice.

During the fieldwork, the necessity for clear guidance on how prosecutors should navigate between their twin roles manifested in interviews with prosecutors. The following vignette captures the confusion experienced by one prosecutor who fluctuated abruptly between casting himself as a minister of justice and as an advocate. It reflects the need for more scholars and practitioners to explore this paper’s central theme.

Vignette 1: Ambiguity in prosecutors’ roles

The prosecutor in this account was conscious of the simultaneous demands on him as a minister of justice and as an advocate. Yet, while his rhetoric cast him solely as a minister of justice, his practical approach appeared to be motivated by considerations of advocacy alone. He resisted this explanation for his actions but was unable to provide examples to the contrary. This vignette demonstrates how a policy failure to address the potential conflict between prosecutorial roles can result in arbitrary prosecutorial decision-making, with prosecutors unjustifiably choosing one role over the other in practical terms—whether by design or oversight.

Adopting the proposed model would enable the prosecutor in the preceding vignette to prioritize distinct aspects of his role at different times. The prosecutor would recognize that, as a minister of justice, he was bound to disclose relevant evidence to the defence counsel, including evidence of the victim socializing with the defendant shortly after the incident in the first case. However, having done this, it was his duty as an advocate-prosecutor to focus on the defendant’s culpability during the trial. There was no need for him to be confused, ashamed, or defensive while doing so. Similarly, in the second case, he would recognize that, in proceedings relating to the charge, he was not obliged to argue for charges to be framed simply because he was a prosecutor. Since these arguments were not related to oppositional fact-finding at trial, he would ground them in his role as a minister of justice. Thus, he would prioritize the public interest at this stage and consider the risk of wrongful conviction before arguing for charges to be framed. However, without a clear framework for determining which role to prioritize, the prosecutor seemed unable to provide principled justifications for his actions. This article argues that the appropriate framework to guide prosecutors facing similar dilemmas is the proposed dynamic model.

The remainder of this section discusses each principle of this model in turn. It uses the principle to critique the observed operation of prosecutors and draws out why adopting the dynamic framework would lead to better prosecutorial decision-making.

The first principle is that, in general, the prosecutor should operate as a minister of justice. Adopting this principle would be useful for the prosecutor while arguing about preliminary or ancillary matters, such as charge or bail. The following vignette presents a case in which the prosecutor based arguments relating to charge on the threat of negative publicity. The better alternative would have been for her to act as a minister of justice and base them on the public interest, as required under the proposed model.

Vignette 2: The prosecutor as a minister of justice in proceedings relating to charge

As this vignette illustrates, it was rare for a prosecutor to argue that the defendant should be discharged in a rape case, no matter how weak the prosecution evidence. This happened in only one observed case during the entire period of fieldwork. Prosecutors cited a fear of adverse publicity as a major reason for their pro-advocacy stance in proceedings related to the charge, especially given that Delhi has received intensely negative press coverage on the issue of violence against women (Roychowdhury, Reference Roychowdhury2013, p. 282). This adverse reporting has resulted in heightened scrutiny by the Directorate of Prosecution and could lead to internal inquiries of the type described above. One prosecutor interviewed for this study suggested that this scrutiny was a direct result of managerialism in the prosecution service. ^{Footnote 58} Another prosecutor talked about a high volume of paperwork having to be filed where defendants were discharged in rape cases. Against the background of additional paperwork and departmental inquiries, she deemed it more efficient to argue for charges to be framed in all rape cases—especially since prosecutors tend to be overworked in general. Concerns about overwork and substandard employment conditions were echoed by several other prosecutors and have also been articulated in academic literature (Sharma, Reference Sharma2017, p. 96; Bhandari, Reference Bhandari2016, p. 94).

Nonetheless, prosecutors should not approach the framing of charges under the influence of irrelevant considerations, such as press coverage. Further, whether it stems from managerialism, or from fears of negative publicity, or for any another reason, an automatic pro-advocacy stance in relation to charges could lead to innocent people being prosecuted or convicted (Green, Reference Green2019, p. 602). Following a mechanical, pro-advocacy approach to charging can also interfere with the prosecutor’s advocacy during the trial, when she is expected to argue for the conviction of defendants whom she is convinced are innocent—as analyzed later in this section. For all these reasons, a pro-advocacy stance in relation to the framing of charges is undesirable. Adopting the proposed dynamic framework would ensure that the prosecutor operates as a minister of justice while arguing about the charge, rejecting an automatic pro-advocacy orientation. Thus, it is public interest that would determine the prosecutor’s arguments about charge. In some cases, this would entail arguing for the defendant to be charged; in other cases, it would entail recognizing that some or all charges would be inappropriate.

The desirability of giving prosecutors the leeway to frame arguments relating to the charge in the public interest could be countered by references to their corruption, as well as the stereotypical attitudes espoused by many prosecutors in relation to rape victims, or to marginalized groups more generally. ^{Footnote 59} The primary data lend weight to these counterarguments, as also illustrated in the discussion on prosecutors’ discriminatory attitudes towards the end of this section. To give a particularly jarring example, one prosecutor said that he could not understand how it was physically possible to have sex with a woman against her wishes. According to him, two people would need to hold her down while a third one committed the act. On this basis, he concluded that rape complaints were mainly filed for money because of the availability of generous compensation to victims of this crime.

Nonetheless, I argue that, despite evidence of discrimination or corruption, prosecutors should reject an automatic pro-advocacy approach to charging. In part, this is because prosecutors are statutorily limited to making arguments about charges. Only the judge can decide which charges should be framed. These statutory limits placed on the prosecutor’s role should alleviate some concerns about dishonest or biased conduct. Further, apprehensions about such misconduct are not particular to the framing of charges. They are relevant to the discharge of prosecutorial duties more generally. The appropriate remedy would be to create robust systems of accountability to tackle any mala fide prosecutorial action, at any stage.

Having discussed the prosecutor’s default role as a minister of justice, I will now discuss the second and third principles. Under these principles, during competitive fact-finding at trial, the prosecutor should operate as a partisan advocate for the defendant’s conviction. The following vignettes describe prosecutors who inadequately discharged their duty of advocacy during the trial. They justified this approach by suggesting that the cases in question were unfit for trial. This approach seems to reflect public interest concerns, drawing from the prosecutor’s role as a minister of justice. However, under the dynamic model, prosecutors would focus on advocacy during competitive fact-finding. Where the case for conviction is weak, there are alternative courses of action that the prosecutor can explore in her capacity as a minister of justice, as discussed below. A dilution of the prosecutor’s advocacy obligations during adversarial fact-finding is an inappropriate response to weak cases.

Vignette 3: Advocacy during competitive fact-finding

This prosecutor deliberately refrained from engaging in advocacy in promise to marry cases, since he understood these allegations to be generally false. However, his determination of falsity was based on unpersuasive evidence. For example, the existence of a loving relationship between the victim and defendant is compatible with an allegation of rape based on a fraudulent promise of marriage. Such promises are likely more convincing when made by someone who professes to love the victim. Similarly, the circumstances under which the victim and defendant first met are unlikely to be related to whether the defendant later raped her. In fact, during the fieldwork, prosecutorial disbelief in the victim’s credibility often appeared to be prompted by prejudicial attitudes towards women who reject conventional sexual norms, such as by having premarital sex.

Under the dynamic model, even if the prosecutor in the above vignette was unconvinced about proceeding against a particular defendant, half-hearted advocacy during the competitive fact-finding phase would not be an option. Instead, the prosecutor would have alternatives available to him outside the fact-finding process. For example, he could argue in favour of the defendant’s discharge or seek judicial permission to withdraw the prosecution. Even so, in making these arguments, he would have to be guided by the public interest and not by the sort of fallacious reasoning on display in the vignette. If the case were to proceed to the competitive fact-finding phase—for example, if the judge decided to frame rape charges or deny permission for a withdrawal order—the advocate-prosecutor would then be obliged to put forward the strongest possible case for conviction. In adversarial logic, as reflected in the second and third principles of the dynamic model, it is a contest between the prosecutor and the defence counsel that would allow the judge to arrive at an optimally reliable outcome. This contest is impossible without the prosecutor’s advocacy during the fact-finding phase. It follows that such advocacy is obligatory even where the prosecutor lacks a personal belief in the defendant’s guilt, which is not explicable on relevant evidence. This is especially important given the stereotypical attitudes that often shape prosecutors’ beliefs.

Further, there is ongoing debate about whether rape laws should include promise to marry cases (Garg, 2019). Many prosecutors criticized existing laws for being overinclusive and this made them unwilling to advocate for a conviction in promise to marry cases. Nevertheless, legally, the defendant’s conduct in such cases is classifiable as rape. The prosecutor might subjectively regard the law as unjust but, by itself, this is an unacceptable reason to abandon her advocacy obligations under the dynamic model (Young and Sanders, Reference Young and Sanders2004, pp. 200–5).

Like the last vignette, the following vignette describes a prosecutor who refused to advocate for the conviction of a defendant whom she believed to be innocent. Once again, this rejection of advocacy seemed to be grounded in her role as a minister of justice. However, under the dynamic model, the prosecutor would prioritize advocacy in her litigative role. Rather, the weakness of the evidence or undesirability of the prosecution would inform other aspects of her prosecutorial role such as those relating to the framing of charges, as brought out in the discussion below.

Vignette 4: Advocacy during competitive fact-finding

The prosecutor appeared to have concluded that the accusation was false based on an informal conversation with the investigating officer and was willing to forego advocacy on that basis. As with the previous vignette, the prosecutor’s impulse against advocacy was motivated by her subjective beliefs about the prosecution being unjust and not by relevant evidence that formed part of the record. This would be unacceptable under the dynamic model. The proposed model would require that the prosecutor act as a minister of justice in discharging all her functions, except her engagement in competitive fact-finding at trial. Thus, as a minister of justice, the prosecutor would bring the exculpatory evidence to the attention of the court and the defence counsel. Given the belated stage at which she found out about the evidence, she would also consider whether it was appropriate to withdraw the prosecution. ^{Footnote 64} Having taken these steps, if the trial continued, she would proceed to set up the best possible case for the defendant’s conviction. Instead, the prosecutor chose to protract the proceedings while deliberately underperforming in her role as an advocate. This was regarded by her as a justifiable moral compromise. However, this undermined the adversarial premise of the trial while also taking up scarce resources in an overburdened system.

The second principle of the dynamic model also requires that the prosecutor’s advocacy be constrained by minister-of-justice considerations. Where the foundational presumptions of adversarial criminal procedures are undermined, the prosecutor must intervene in line with her role as a minister of justice. In the field, these presumptions were undermined when multiple witnesses were simultaneously examined, compromising the ability of the judge, prosecutor, and defence counsel to discharge their duties effectively. This happened frequently given the resource constraints under which most trial courts operate. ^{Footnote 65} In other instances, witnesses—ranging from the defendant’s landlady to the victim herself—were examined or cross-examined in the defence counsel’s absence. On such occasions, in the absence of an appropriate judicial response, the dynamic model would require the prosecutor to intervene and object. This is because, even though a deviation from adversarial procedures could afford the prosecutor a competitive advantage, her advocacy should be constrained by minister-of-justice considerations. This is of course impossible where the prosecutor is absent herself. For example, one victim’s chief examination commenced with no prosecutor, defendant, and defence counsel in court—only the DCW advocate was present. ^{Footnote 66} Even so, in other cases, it was possible for the prosecutor to intervene and object, and the dynamic model would require her to do so.

Finally, under the fourth principle of the dynamic model, prosecutor-victim relations should be governed by the prosecutor’s role as a minister of justice. Currently, the prosecutor’s approach towards the victim appears underpinned by neither of the twin functions. Prosecutors generally do not engage with the victim, except during her testimony. This attitude can alienate victims from legal proceedings. ^{Footnote 67} It can exacerbate the trauma they experience, first, from having suffered a crime and, second, from having to navigate an unfamiliar and unsupportive legal system (Van Camp, Reference Van Camp2017, pp. 681–3). Getting appropriate information and updates about the case can minimize this trauma and increase victim satisfaction in the criminal justice system (Victimsupport.org.uk, 2017). Conversely, victims’ disengagement from the proceedings can culminate in their refusal to support the prosecution, even where the crime was committed.

Members of civil society who support rape victims were often critical of prosecutors’ attitudes towards victims. One support person—though mischaracterizing the victim as the prosecutor’s client—said:

One lawyer, who frequently appears as a victim’s counsel, contrasted her own approach with that of prosecutors as follows:

These observations were reflected in the experiences of the victims interviewed for this study, who were generally unable to explain the status of the case. Their experience with prosecutors varied, with some never having met the prosecutor.

Many prosecutors were convinced that rape allegations are mostly false, often based on unpersuasive and stereotypical reasoning. For example, one prosecutor said that she could tell from how one victim looked that the accusation was false. She added that, since the victim’s father-in-law (defendant) was the only Hindu person in a Muslim household, he was probably being framed. Another prosecutor said that persons from the middle and upper classes do not commit rape; it is migrant, poor men from rural areas who give Delhi a bad name. ^{Footnote 68} Based on these and other ill-informed and/or discriminatory beliefs, prosecutors often assumed that it was not worthwhile to spend time explaining legal processes or decisions to victims. The dynamic model would reject this approach, requiring instead that the prosecutor should act as a minister of justice in relation to the victim. This includes not only treating the victim with dignity and respect—as all officers of the court are obliged to do—but also acknowledging the victim’s unique interest in the prosecution as someone who has potentially suffered direct damage to her rights and wellbeing. This justice-oriented approach would translate into enhanced interaction with and support for the victim.

Many prosecutors suggested that they could not be expected to engage directly with victims, given their burgeoning workload. They justified their approach based on the presence of other legal points of contact for the victim, such as DCW advocates or victims’ counsels, whether accessed privately or through legal aid. However, none of these lawyers is an appropriate substitute for the prosecutor. There is at best one DCW advocate per court, which means that the case-load for each such advocate is generally at least equal to that of the prosecutor. Privately hired counsels can provide useful explanations to the victim. Nevertheless, many victims find such lawyers unaffordable, if not extortionate. For example, one victim interviewed for this study said she had sold her house and lost her business while trying to afford legal costs, including the fees for the victims’ counsels whom she had hired over the years. While victims can access legal aid through the DLSA, the experience of victims with such advocates was generally negative. One victim recounted how her DLSA lawyer was also representing the defendant, unbeknownst to her. Since he had ensured that the formal documentation was filed by his colleague, she only found out about the conflict of interest belatedly. ^{Footnote 69} She later managed to access another DLSA lawyer, who was able to relay her concerns in court. Nonetheless, while these services are meant to be free, he charged her for them. ^{Footnote 70} She described her experience as follows:

Even where the victim successfully recruits a lawyer, the seriousness with which the prosecutor or judge takes such lawyers varies depending on the latter’s seniority and reputation. ^{Footnote 72} Most victims will not have access to renowned counsels.

The preceding analysis shows why prosecutors should not assume that the victim’s engagement is ensured by the presence of other lawyers in the court. Under the dynamic model, the prosecutor acting as a minister of justice must engage with the victim, providing her with appropriate information and updates. This approach can also collaterally facilitate the prosecutor’s advocacy. For example, many rape victims refuse to support the prosecution as the trial progresses. There can be many reasons for this, including threats from the defendant or pressure from the victim’s family on account of the stigma associated with rape. When one prosecutor was asked how she dealt with such cases, she said:

If the prosecutor keeps an open channel of communication with the victim from the outset, this can prevent the victim’s alienation from the proceedings and give her strength and resources to withstand pressure from third parties while testifying. The successful discharge of the prosecutor’s minister-of-justice role can thus also facilitate effective advocacy on her part.

It could be argued that closer interaction between the victim and the prosecutor could result in creating unrealistic expectations in the victim’s mind. For instance, one prosecutor participant mentioned that, in order to reassure the victim, he would sometimes state that he was her representative, even though he knew this was not true. Misleading the victim in this way could leave her doubly disappointed when the prosecutor’s claims turn out to be unfounded. However, a minister-of-justice approach, while requiring greater prosecutor-victim engagement, would not extend to condoning such a distorted representation of the legal position. The nature and quality of that engagement would still be circumscribed by public interest considerations under the dynamic model.