Voice – Can Defendants have their Side of Story Presented?

The procedural justice theory propositions that providing opportunities for voice is closely linked to perceived fairness of the outcome. If the defendant is allowed to state her perspective or tell her side of story before the decision is made, she is more likely to see the decision as fairly and legitimately made.Footnote ⁷² Two common methods, as O'Hear opines, may help the defendant present her side of story in a plea-led scenario.Footnote ⁷³ First, police ought to be diligent about collecting the defendant's story in their reports.Footnote ⁷⁴ Second, defence counsel is expected to convey the defendant's story before the prosecution makes a plea offer.Footnote ⁷⁵ In the Chinese system of plea leniency, however, both measures are nearly unattainable not least because neither police nor counsel are interested or feel obligated to speak to and for defendants. In a nutshell, the preoccupation with confessions in China's ‘truth-seeking’ criminal process has over time made the police a coercive institution,Footnote ⁷⁶ which likely becomes haughtier in the plea leniency process when under pressure to obtain early guilty pleas. At the same time, the difficulty of Chinese defence lawyers in providing effective legal assistance has hitherto remained a lingering issue that plagues the Chinese criminal justice system.Footnote ⁷⁷ The generally poor quality of legal representation is unlikely to be reversed by the introduction of plea leniency which is devised to downplay the confrontational features of the justice process. Conceived in this way, plea leniency narrows the venue for defendants to have their voice heard by diminishing their already weak rights in speedy criminal proceedings.

The indifference of police to learn the defendant's side of the story is hardly surprising. For decades, policing in China has been dominated by the authorities’ obsession with social control, serving the prevailing socio-political goal of the preservation of stability.Footnote ⁷⁸ This underlying ethos of police power harks back to the Maoist era where a class-dichotomised society presaged the need for ‘political justice’ through heavy-handed policing of ‘class enemies’.Footnote ⁷⁹ With China transforming itself from an egalitarian society under Mao to a stratified society following economic liberalisation, the State has witnessed an unprecedented spike in crime wrought by a constellation of new social problems.Footnote ⁸⁰ As such, concerns over deteriorating public order have propelled the police towards a more demanding social control function by extending their position at the forefront of the fight against crime.Footnote ⁸¹ This is done, as Mike McConville points out, through the making of overly punitive and oppressive laws which operate as ‘a major resource providing legal cover for [police] actions’.Footnote ⁸² In any event, policing in China has shaped a culture of presumption of guilt.Footnote ⁸³ Reflecting on what Herbert Packer termed ‘the crime control model’,Footnote ⁸⁴ it is widely acknowledged that Chinese police prioritise the efficient suppression of criminality in the interests of social order by means of almost exclusive investment in gathering inculpatory evidence.Footnote ⁸⁵ Instead of respecting the defendant's right to the presumption of innocence as notionally incorporated in the CPL,Footnote ⁸⁶ police investigation is characterised by an approach to securing convictions on the basis of ‘substantive truth’.Footnote ⁸⁷ Yet, this search for truth is not performed on grounds of objective evidence, but ‘constructed’ by police to align with the pre-determined guilt of the defendant.Footnote ⁸⁸

Yu Mou, for example, ascribes this ‘construction of guilt’ to the reality that the ‘story-telling’ of criminal facts was monopolised by police.Footnote ⁸⁹ In her observational fieldwork of the Chinese criminal justice system, it found that effort to establish the factual guilt of the accused is intricately interwoven with the misuse of power by police who coerce, induce, and unduly influence the accused to confess to a crime.Footnote ⁹⁰ Where a confession does not conform to the ‘official version of truth’, police were not hesitant to distort, manipulate, and even fabricate the confession to meet the evidential threshold for prosecution.Footnote ⁹¹ Largely, such malpractices were fuelled by a set of feeble procedural rules loosely protecting the defendant's right to remain silent and privilege against self-incrimination, as typically valued in most common law jurisdictions.Footnote ⁹² As disclosed in Mou's research, the abuse of police power was pervasive in standard criminal procedure. Considering this, it may be sensible to assume that police wrongdoings can only be more acute in the simplified and fast-track procedures, given that cases applying plea leniency in these procedures necessitate a more rapid and streamlined process of case deposition. An avid pursuit of confessions or guilty pleas tends to give rise to increased magnitude of coercive or deceptive police practices, as manifested in Western plea bargaining scholarship.Footnote ⁹³ On this point, scholars are pessimistic about any serious commitment to the defendant's voice through police reports in the context of plea negotiations.Footnote ⁹⁴ Similarly, it is reasonable to be doubtful of Chinese police being sensitive to the defendant's story, as almost every legal right in police investigation is either vaguely prescribed or treated as a bargaining chip to be traded for leniency at the sentencing stage.Footnote ⁹⁵

My data on court judgments from twenty courts in Shanghai vindicates this proposition. Across 13,611 criminal judgments on plea leniency cases, it is found that police more often than not overlook, intentionally or inadvertently, the defendant's side of story in their performance of their investigative duties. Figure 1 sets forth the numbers of cases where police included the defendants’ version of facts (Biancheng) in their case dossiers delivered to the prosecution and presented at summary trials. Among 9,543 cases finalised through the simplified procedure where the defendants were sentenced to terms of three years or more, the defendant's arguments appeared only in 113 court judgments. Moreover, there was a significant decline in police attention to what the defendant had to say among less serious cases. Among 4,068 cases heard through the fast-track procedure where a sentence of three years imprisonment or less was usually given, the mention of the defendants’ arguments was only found in 14 cases. Of course, it is likely that in some court reports judges omitted to incorporate the defendants’ argument given that all defendants had admitted they were guilty before the trials commenced. It could also be possible that police failed to record it in their report at the outset. Yet, such a disproportionately high number of judgments where the defendant's voice was missing elucidates, at a minimum, the low interest of police in grasping perspectives different to the official perception of ‘truth’, provided defendants were ever given the chance to speak up in the first place.

Figure 1. The Numbers of Cases Presented with/without the Defendants’ Arguments

If police are apathetic about the defendant's point of view in general, can legal counsel instead play a more useful role in raising defendant voices? Notably, China has made the right to legal representation a focal point in recent legal reforms. In October 2017, the Supreme People's Court and the Ministry of Justice jointly issued the Measures for Carrying out the Pilot Work in Full Coverage of Lawyers’ Defence in Criminal Cases (henceforth, ‘The SPC-MoJ Measures’). In eight provinces and municipalities,Footnote ⁹⁶ it is mandated that every accused should either be assigned a professional defence lawyer or offered counsel assistance from legal aid services.Footnote ⁹⁷ Drawing on the ‘successful experience’ amassed in the test regions, in January 2019 China pronounced the expansion of this practice across the country.Footnote ⁹⁸ This move aims to ensure that counsel appear in every State case, particularly those involving plea leniency where summary trials are meant to take place. Practically, the power to grant legal aid is vested in provincial, municipal or local legal aid centres upon request by the defendant who meet the eligibility criteria.Footnote ⁹⁹ Once assigned to a criminal case, the duty and legal aid lawyer is salaried by local justice bureaus to offer free legal services, including but not limited to counselling and legal representation at trial.Footnote ¹⁰⁰

To be sure, there is a strong rationale behind the national push for this full-coverage legal representation. Since the modernisation of the legal system in the late 1970s, China's inquisitorial criminal justice system has been repetitively accused of lacking adequate protection for defendants to withstand the intrusive force of the State.Footnote ¹⁰¹ On the one hand, the overall rate of legal representation in criminal cases has at all times been below 30 per cent, with affluent locals being relatively more lawyer-equipped than impoverished locals.Footnote ¹⁰² On the other hand, efforts by defence counsel to provide adequate legal assistance have been impeded by the institutional dynamics of Chinese criminal proceedings that inhibit lawyers from acting as an adversarial advocate.Footnote ¹⁰³ The paramount objective of social control has created an overpowering legal apparatus. Relatedly, obtaining confessions at all costs and thus, positing it as the centre of gravity within the exercise of the state power has reduced the role of counsel to symbolic significance. Indeed, numerous studies have empirically proved that Chinese counsel are riddled with various predicaments in their everyday defence work.Footnote ¹⁰⁴ The problems run the gamut from difficulty accessing case files and suspects in detention to the difficulty in collecting evidence for defendants.Footnote ¹⁰⁵ Apart from constituting an action against the best interests of their clients, challenges in hearings against State authorities can also lead to defence lawyers being brandished as ‘foes’ of the State and incurring criminal liability under the CL.Footnote ¹⁰⁶ It comes as little wonder, then, that this has led many commentators to assert that the right to legal representation, though being enshrined as a constitutional norm, has largely remained a construct on paper.Footnote ¹⁰⁷

In many respects, legal counsel walk a thin line in China's criminal process. It is of course undeniable that outspoken lawyers do exist and are devoted to fight for defendants’ rights. Sida Liu and Terence Halliday's work on criminal defence in China identified this group of lawyers as ‘political activists’ and ‘progressive elites’ who uphold and advocated for liberal values in the system while representing their clients.Footnote ¹⁰⁸ But their research also pointed towards even larger legions of ‘pragmatic brokers’ and ‘routine practitioners’ who were usually powerless to challenge the embedded legal culture of criminal justice being disparaging towards defence lawyers.Footnote ¹⁰⁹ This then begs a more pertinent question of whether and to what degree lawyers are willing to convey a defendant's voice in plea leniency. The initiatives of full-scale legal representation are certainly laudable, but evidence showed that the superficial counsel appearance can neither guarantee sufficient legal assistance nor present meaningful voice opportunities to inform plea decisions.Footnote ¹¹⁰ It might be an over-statement to say that counsel play no role at all in plea leniency. However, it seems to be clear that the defendant's right to legal representation is at greater peril during the simplified and fast-track procedures than during the full-scale procedure. For example, one recent study conducted to examine the width and depth of legal representation in plea leniency showed a fairly negligible role of counsel in China's summary criminal proceedings.Footnote ¹¹¹ Based on a questionnaire survey among defence counsel (n = 374) and prosecutors (n = 375) in one selected pilot city (City H in Z Province as anonymised by the researcher), the findings indicated that counsel (including privately retained lawyers and duty lawyers) had found themselves embroiled in the plight of only being able to help defendants with the interpretation of law while ironically, encouraging and witnessing defendants’ acceptance of plea agreements.Footnote ¹¹² A particular discovery of this research illustrated the gulf between the perceived function and the actual role of counsel in representing their clients in plea leniency cases.Footnote ¹¹³ While the majority of interviewed lawyers agreed that defence counsel ought to ‘comprehensively act in their clients’ best interests’, about 89 per cent of respondents conceded that their role was, on the practical front, limited to only seeking lenient sentencing outcomes.Footnote ¹¹⁴

My analysis of the court judgments from Shanghai attests to this observation. Out of the dataset, I compiled a list of 5,748 cases in which defendants were legally represented during the summary trials. The absence of counsel appearance in almost two-thirds of trials is not really startling – for cases to be proceeded by way of simplified and fast-track court proceedings, most defendants must have pled guilty at the investigatory or prosecutorial stage where counsel appearance was mandatory. This has made counsel appearance at trial optional rather than necessary. However, what appears to be striking is the ineffectiveness of counsel attending the trial to operate as the guarantor of defendant voice in any impactful way. The Guiding Opinions explicitly designates the summary trial process as essential for assessing legality, truthfulness, and voluntariness of the defendant's guilty plea.Footnote ¹¹⁵ Consistent with prior findings, nevertheless, it found that counsel were almost always reluctant to question the official narrative of criminal facts and prosecutorial evidence in relation to the defendant's guilt. More frequently, counsel focused their defence predominately on reinstating the defendant's early guilty pleas and leveraging it to seek sentencing concessions. Figure 2 provides an overview of cases in which counsel either raised no objection to or disputed the prosecution's case. It indicates that in almost every case the counsel opted not to contest the validity of guilty pleas presented by the prosecution (5,734). Only 19 cases’ counsel opposed accepting the laid-down charges, the criminal facts, or the prosecution evidence.

Figure 2. The Percentage of Counsels Raising No Objection to / Challenging the Prosecution's Case

The reasons that counsel showed no genuine intention to confront the prosecution may in part be due to the fact that counsel spotted no loopholes in the State cases, since most plea leniency matters typically involve relatively minor offences with simple facts and clear evidence.Footnote ¹¹⁶ It was more likely, though, that the undesirable standing of counsel in the criminal justice system refrained them from challenging the ‘authority’ of law enforcement agencies. As counsel have long been marginalised or even oppressed in the system, to challenge the prosecution case, especially after defendants have allegedly admitted guilty, is to ‘touch the tiger's buttocks’ that may backfire on argumentative or aggressive counsel.Footnote ¹¹⁷ More pragmatically, avoiding hostility seems to be a realistic strategy, particularly for those representing defendants for the first time at summary trials, thereby affording them with very little knowledge about how and why their clients entered a guilty plea in the first place. Under these circumstances, it is not uncommon that counsel only grasp at a sense of self-worth through arguments for lenient sentences. In all 5,748 cases, counsel put forth discounted punishment on the basis that defendants had pled guilty at the stage of investigation or prosecution. Instead of raising this as a new claim, counsel's request for leniency in sentencing served as a reminder to procuratorates to keep their early promises on lighter punishment as part of the plea deal. In addition, just like what Liebman revealed in his study of criminal judgments from Henan,Footnote ¹¹⁸ the fact that the defendant had actively paid compensation or restitution was relied heavily upon by counsel to convince the judge a lenient sentence should be granted. Figure 3 illustrates a variety of reasons commonly argued by counsel to request light penalties. Other than confessions and compensation, the other factors to which counsel endeavoured to draw to the judge's heed include the defendant being a first-time/occasional offender and the victim's forgiveness towards the defendant. Typically, these reasons were presented in tandem in the hopes of carrying greater persuasive weight before the sentencing judge.

Figure 3. The Reasons Argued by Counsel for Lenient Sentences

Of course, counsel's focus on sentencing concessions should not be taken by surprise given China's prevailing culture of over-relying on the prosecution's view of the facts to determine the truth during criminal trials.Footnote ¹¹⁹ This means that the most feasible time for counsel to alter authorities’ narrative lies in the prosecutorial stage where the case is yet to be finalised through guilty pleas and filed for court hearing. On a side note, judicial favouritism towards the prosecutorial evidence which are mainly comprised of case dossiers and written statements has invoked emerging concerns about the fairness of trial. To reverse the pivot of the criminal process, the Government initiated a ‘trial-centeredness’ campaign in 2014 to promote the long-absent elements of adversarialism characterised by live witnesses, cross-examination, and court debate.Footnote ¹²⁰ It is argued that the ideal of trial-centeredness should also run through the process of plea leniency,Footnote ¹²¹ mandating the judge to gauge an overall apprehension of the guilty plea as to its legality and voluntariness. The above data, however, tell a different story. While the data exhibited the limited role of counsel in offering alternate accounts for judges to acquire a full picture of events in court proceedings, a general pattern of lawyering in plea leniency surfaces. Counsel were not accorded greater power to act more freely and assertively as the defendant's agent when facing legal authorities in plea leniency schemes. With the criminal process being compressed to settle criminal matters in a time-saving manner, counsel carry a less adversarial function as a result. Thus, a common thread through the findings of the present research and several studies alike is that counsel appearance, while being promoted as a precondition to an informed guilty plea, amounted to a mere procedural formality as opposed to a genuine form of procedural protection.Footnote ¹²²

Neutrality – Can Defendants Access Consistent and Principled Decisions?

In addition to voice, the perceived fairness of a procedure depends on decision-making being neutral, consistent, rule-based, and without partiality.Footnote ¹²³ This entailed authorities being upfront and explaining to defendants the reasoning behind their decisions in a transparent manner. To this end, Tyler and others have identified objective criteria in place of personal views as an important means by which a decisionmaker can establish her neutrality.Footnote ¹²⁴ Within the plea-bargaining context for example, explanations need to be predominately conveyed by prosecutors – though counsel may take on a similar role – to help defendants enter into an informed plea.Footnote ¹²⁵ Following explanations, authorities were supposed to equally and consistently apply rules across people in like situations.

Plea leniency in China has, unfortunately, failed in both aspects. As a requirement, the plea leniency ordinances – including the 2016 Plea Leniency Pilot Project Decision, the 2018 CPL and the 2019 Guiding Opinions – have all demanded the explication of the nature and procedure of plea leniency by police and procuratorates before they make any plea offers.Footnote ¹²⁶ Primarily, this statutory requirement involved an unequivocal explanation of the defendants’ procedural rights and their ‘free choice’ to confess and receive lenient sentences accordingly.Footnote ¹²⁷ In official terms, advising the defendant of their right to select court proceedings (that is, the summary procedure or the ordinary procedure) formed an integral part of plea leniency because it lay in close association with voluntariness of guilty pleas.Footnote ¹²⁸ Conceptually and operationally, voluntariness bolstered the legitimacy of plea-based criminal justice, as evinced in the common law jurisprudence of plea bargaining (especially in the US).Footnote ¹²⁹ Not only was it a key determinant of legally binding guilty pleas, but plea voluntariness was a critical prerequisite to the circumvention of court hearings, and a waiver of a spate of fundamental rights conferred on the accused.

Oftentimes though, plea voluntariness is coalesced with the right to legal representation and the quality of legal assistance. Despite the argument that the choice situation imposed by plea bargaining on the accused is ‘intrinsically coercive’,Footnote ¹³⁰ it is assumed that effective counsel assistance at the critical pre-trial stages of prosecution may somewhat sustain the volitional power of defendants.Footnote ¹³¹ However, the over-dependence on counsel's diligent work overlooks the duty of explanation that legal authorities ought to discharge in parallel. More likely than not, a guilty plea will not be ‘knowing, intelligent and voluntary’ if legal authorities do not energetically engage in providing information and expounding on reasoning – which together, act as evidence of even-handedness and objectivity in the bargaining context.Footnote ¹³²

Suffice to say, the idea of legal apparatus being open, transparent, and communicative is particularly important in China's plea leniency programs. As discussed earlier, defence counsel in China were overly incompetent. It inevitably places obligations on legal authorities to guarantee defendants' voluntariness of plea; no innocent person should be forced or induced to admit guilt. Empirical studies, however, have illustrated results contrary to such commitment. Based on his examination of plea voluntariness in two districts of City A and B, Xin Zhou suggested that in two respects, police and procuratorates had largely left defendants in the cold when presenting plea agreements and accordingly, making plea decisions.Footnote ¹³³ First, there was significant variation in timeframes for informing defendants of their right to plead guilty. In both districts, police were found to timely advise defendants of their relevant entitlements upon the first questioning, as stipulated in the CPL.Footnote ¹³⁴ However, after the case was transferred from police to procuratorates for formal prosecution, the prosecuting agency in City A did not commence plea leniency conversations at the earliest opportunity until the indictment was presented to the trial court – that is, almost at the end of the prosecutorial period.Footnote ¹³⁵ Contrarily, its counterpart in City B initiated plea leniency conversations normally within three days of receiving the case dossiers from the police.Footnote ¹³⁶ Regardless, the findings revealed that legal authorities in both regions enjoyed exclusive discretion to opt for a time to initiate the plea leniency process as they thought fit when meeting with defendants.

Second, with regards to the content of authority-defendant communications, Zhou's research unveils a common practice of State authorities. That is, police and procuratorates tended to confine their interactions with defendants to providing information about the procedural rules and legal conditions of plea leniency.Footnote ¹³⁷ At the same time, they were reluctant to disclose more particulars of the case in question.Footnote ¹³⁸ Interestingly, many interviewees in this study happened to see the delivery of the ‘plea leniency agreement’ (Renzui Renfa Jujieshu) as an alternative to their duty of explanation.Footnote ¹³⁹ Equivalent to a standard form contract in substance, the plea leniency agreement was said to contain all the necessary details that defendants were required to read through. Procedurally speaking, signing off on this agreement meant that defendants were agreeable to the laying charges and proposed punishments as a package deal offered by procuratorates. Still, this paper-based informing model does not fit well with what Tyler deems ‘openness and objectivity’ of neutrality, simply because defendants were not given genuine conversational opportunity to attain an all-round vision of their cases which was imperative to a knowing and intelligent guilty plea. This was exacerbated by the absence or difficulty of counsel to attain the criminal facts in sufficient detail on the defendant's behalf. Realistically, to the best of the defendant's belief, admitting guilt and accepting a discounted sentence turned out to be a rational choice insofar as little was known about the strength of the prosecution case. Here, defendants were faced with a paradoxical situation – their right to access evidence (whether that be for or against them) was impaired under the plea leniency process, which would otherwise be better protected in some form of judicial oversight should their cases proceed via standard procedure. In this light, the lack of a neutral process in presenting plea deals occasioned a risk of defendants involuntarily entering in guilty pleas, rendering plea-based convictions unsafe or unsatisfactory and open to appeals.

The foregoing analysis only tackled one aspect of neutrality with reference to plea leniency practices. On a further practical note, procedural justice ideas of neutrality only addressed the requirement that decisions of legal authorities be made through the consistent application of rules and consideration of facts.Footnote ¹⁴⁰ When transplanted into plea arrangements, they denoted that plea decision-making should conform to established standard criteria; and legal agencies need to minimise discrepancy by ‘treating like cases alike’. These norms, however, do not seem to be respected nor were they upheld in China's simplified and fast-track procedures. Across the spectrum of plea leniency practice, divergence was perceptible across different cases and defendants.

One exemplar of this incongruency appertains to the selective application of plea leniency for different types of crime. Unlike plea bargaining in Anglo-American countries where plea deals are almost equally struck across misdemeanor and felony cases,Footnote ¹⁴¹ plea leniency in China has been susceptible to discretionary application. Specifically, this programme had devoted its resources disproportionately to minor offences in contrast to more serious offences. In 13,611 court reports, it is found that three broader categories – crime endangering public security, crime infringing property rights, and crime impeding social order management – have topped the list of plea leniency cases in Shanghai. When it comes to individual offences, the charges of theft, dangerous driving, picking quarrels and provoking trouble stood out in each category. Figure 4 and Figure 5 represent the ten most common criminal offences processed under plea leniency in Shanghai. The data showed that the crimes dealt with by the simplified procedure and the fast-track procedure are mostly overlapping despite the fact that these two procedures were created with their own respective targets.

Figure 4. Ten Most Common Criminal Offences Processed through the Simplified Procedure

Figure 5. Ten Most Common Criminal Offences Processed through the Fast-track Procedure

Here, the data provides a window into the routine practice of plea leniency that is both over- and under-inclusive. On the one hand, these commonly recorded perpetrations almost all fall within a general category of minor offences. On the other hand, serious crimes such as violent and sexual offending were generally left out. This is reflected in statistics which indicated that over the last year, Shanghai courts only adjudicated 27 robbery cases, 10 arson cases and 5 murder cases (out of 9,543 cases) through the simplified procedure. In the fast-track trials, the gravest offence heard was rape (n = 1) (out of 4,068 cases). This is not to deny that written judgments may not have yielded the full picture of practical divergences. For example, picking quarrels and provoking trouble as a ‘pocket crime’ embraced a great mass of offences, ranging from willfully attacking persons with serious circumstances to creating a disturbance in a public place that caused serious disorder.Footnote ¹⁴² This surely necessitated further on-the-ground observations to explore variations in documented crimes across different plea leniency processes. Regardless, the dominant focus of plea leniency on minor offences was statistically conspicuous, as manifested not only in the public records collected for this study but also scholarly findings in the same line of research.Footnote ¹⁴³

Indeed, the under-representation of serious crimes in plea leniency resonates with policy considerations. The Guiding Opinions, for example, sanctioned a binary model by impelling legal authorities to dispose of minor criminal cases as simply, quickly, and leniently as possible.Footnote ¹⁴⁴ The dominant tenor was that plea leniency should be applicable especially to ‘first-time offenders, casual offenders, crimes of negligence, or juvenile crimes where the social harm is not great’.Footnote ¹⁴⁵ By contrast, police and procuratorates should cautiously apply plea leniency in cases of crimes that ‘seriously endanger national security or public safety, serious violent crimes, as well as major and sensitive that have drawn broad public concern’.Footnote ¹⁴⁶ Even so, leniency ought to be entertained prudently in serious and/or high-profile cases, avoiding the risk of ‘going against the public's sense of fairness and justice’.Footnote ¹⁴⁷ Certainly, there was nothing wrong with nuanced administration of criminal justice in light of special circumstances of individual cases. The deliberate preclusion of serious offences, however, indicated a pronounced departure from a neutral process that is rooted in impartiality and uniformity. Depriving the right of certain groups of defendants to enter in a plea leniency deal was a parity violation, as if guilty pleas of serious offenders carry less weight than those of minor offenders. A more troubling issue perhaps was the dearth of a national yardstick to guide the classification of minor and serious cases around their intricacies. In many localities, authorities were driven to adopt standards tailored to local conditions;Footnote ¹⁴⁸ but what is regarded as a serious crime, eg, the national security offence, can be interpreted in divergent ways across Beijing, Guangzhou and Xinjiang. As such, the existence of regional variations in plea leniency was a far cry from the principled criteria of practice which underpinned the procedural justice ideal of neutrality.

But even with crimes managed by means of plea leniency, cases were hardly treated alike. Mainly, defendants who go through plea leniency have experienced sentencing reductions unevenly applied across time and region. In an extensive quantitative study of sentencing outcomes regarding the particular crime of dangerous (drunk) driving, the researcher found that the scale, extent and mode of leniency granted on penalties varied enormously over time and from area to area. In some locations (Beijing, Shanghai, and Qingdao) judges were more inclined to reduce the length of imprisonment as a concession for guilty pleas.Footnote ¹⁴⁹ In others (Fuzhou and Xiamen), however, judges were more inclined to impose probation in place of short-term incarceration for most minor offences.Footnote ¹⁵⁰ Still, certain regions (Hangzhou) witnessed a static pattern of sentencing despite the traction of plea leniency gained in the local criminal process.Footnote ¹⁵¹ Across China, the degree to which sentence mitigation was effectuated on the ground was particularly telling to the absence of parity in the plea leniency process. This research showed that while some cities experienced a high level of leniency by more than 60 per cent (eg, the amount of reduction in sentence length), others only awarded less than 5 per cent leniency as consideration for guilty pleas.Footnote ¹⁵²

Respect – Are Defendants Treated with more Dignity?

At the operational level, voice and neutrality are connected with a more technical aspect of decision-making. In the procedural justice model, however, the quality of interpersonal treatment in the decision-making process mattered as much as how the decision was made. It has been established that legitimacy and perceived fairness of a procedure also developed out of the provision of respectful treatment of people, including having their rights acknowledged and their needs considered.Footnote ¹⁵³ As such, legal authorities in a plea-seeking context may undertake a series of measures to advance perceptions of respect. For example, they should treat defendants with dignity by discouraging harsh treatment such as prolonged detention, physical coercion, and various other forms of excessive force.Footnote ¹⁵⁴ While dignity has various meanings and functions depending on the explanatory context, it pertained generally to the basic rights of a person to be valued and respected for their own sake, and to be treated justly and decently. To this end, it is then imperative to provide defendants with a threat-free environment in order to avoid the extraction of involuntary guilty pleas through coercive and forcible measures. Concomitantly, legal authorities should cling to the presumption of innocence until a guilty plea was voluntarily entered. This means that the correlated rights of defendants need to be upheld, especially those central to the protection of defendants from unwarranted incursions on their liberties by the State.

At any rate, there was little evidence that plea leniency operates under the aegis of refraining from repressive practices. Over the past decades, China's image as a punitive or ‘carceral’ state has been largely tied with its stunningly yet steadily high rate of incarceration and conviction. Unlike the adversarial system where bail is a prime facie right of the suspect, pre-arrest detention (Juliu) and arrest (Daibu) have long been the norm in the Chinese pre-trial process.Footnote ¹⁵⁵ Substantial empirical research showed that the rate of arrest approval nationwide has never fallen below 80 per cent despite different natures of suspected offences.Footnote ¹⁵⁶ Based on court judgments in one of the Chinese cities during 2012–2013, Moulin Xiong's study illuminated that on average the length of incarceration pending trial is 174.8 days.Footnote ¹⁵⁷ While lengthy custody applied generally to serious offences including intentional assault, drug trafficking, and robbery, minor offences such as property perpetrations attracted an analogous amount of confinement.Footnote ¹⁵⁸ Just like pre-trial detention, the conviction rate in China was consistently and persistently high. Nearly 100 per cent conviction rates at any given time have intensified the aspiration of the Chinese criminal justice system towards purported ‘substantive justice’ at the expense of due process and individual rights.Footnote ¹⁵⁹ What was equally striking was the high use of imprisonment following convictions. Over time, China has been arguably secondary to the US in terms of its prison population, despite the accurate number of incarcerated people remaining a national secret yet to be disclosed to the public.Footnote ¹⁶⁰

The State's cultural disposition to deprive liberty in criminal justice practices, realistically speaking, is unlikely to be materially reshaped by merely putting plea leniency into use. A retreat from overzealous pre-trial custody requires a paradigm shift rather than a procedural innovation in the system. Of course, the Guiding Opinions urges the procuratorates – when handling plea leniency cases – to cautiously approve arrest should the defendant be deemed to represent a minimal level of ‘social dangerousness’.Footnote ¹⁶¹ In 2021, the People's Supreme Court published its annual work report which highlighted a considerable decrease in the pre-trial detention rate from 96.8 per cent in 2000 to 53 per cent in 2020.Footnote ¹⁶² While this marked decline remains unexplained and should be taken cautiously against concerns over the accuracy of Chinese official data, it simply does not tally with the public records in the present study. Across all cases going through the simplified and fast-track procedures, only 1,332 cases (out of 13,611 cases) handled by the Shanghai courts appeared to involve the use of ‘guaranteed release pending trial’ (Qubao Houshen) – a non-custodial measure which allows defendants to return home and attend the court hearing at a designated date.Footnote ¹⁶³ Because not every court judicial document examined revealed details on the pre-trial status of suspects (in custody or released pending trial),Footnote ¹⁶⁴ it was reasonable to surmise that the actual number of defendants granted partial freedom on the Chinese version of bail would be somewhat higher than that reflected in the observed cases. In any event, however, they represented a significant departure from the figure (53 per cent) presented by the authorities. Figure 6 and Figure 7 include five common types of crime under which defendants were most likely to be detained and granted guaranteed release pending trial.

Figure 6. Five Common Types of Crime under which Defendants were Detained prior to Trial

Figure 7. Five Common Types of Crime under which Defendants were Released Pending Trial

The statistical overlap between these two sets of crime is evident. An individual charged with theft (22.3 per cent vs 13.5 per cent), fraud (10.4 per cent vs 9.8 per cent), dangerous driving (12.5 per cent vs 28 per cent) or picking quarrels and provoking trouble (9.1 per cent vs 7.9 per cent) was likely to end up with either an arrest or guaranteed release pending trial, depending in large part on how police exercise their discretion in a specific case. At the outset, this blurred distinction between liberty and captivity is demonstrative of an irregular and incongruent practice of police coercive powers in like circumstances, even within the same jurisdictional region. More ominously, it suggested scanty efforts by the State to ensure that defendants travelling through the plea leniency process were treated more respectfully than in ordinary criminal proceedings. That said, the finding that pre-trial detention prevails over the simplified and fast-track procedures does not represent a rupture of China's preeminent ideology of what Susan Trevaskes incisively terms ‘heavy penaltylism’ (Zhongxing Zhuyi).Footnote ¹⁶⁵ On this view, pre-trial detention was not only an approach to serving precautionary purposes, but also ‘a deliberate tool of punishment to deter potential criminals’.Footnote ¹⁶⁶ Indeed, the pervasiveness of harshness in China's criminal justice was not invariable and always omnipresent. There was growing evidence that the notion of penal moderation has in some way begun to have a bearing on run-of-the-mill law enforcement in recent decades.Footnote ¹⁶⁷ But the steadfast practice of resorting to incarceration as the primary way of policing laid bare flaws in the plea leniency system wherein the presumption of innocence and the principle of ultima ratio appeared to be circumvented ab initio. It dishonoured the defendant's basic right to liberty and other associated rights that would otherwise be more likely enjoyed had the defendant been released prior to trial (eg, the right to counsel). Worse, the overuse of pre-trial detention engendered a generally accusatory environment ripe for forced confessions. As the data suggested, this trend remained largely intact even during the Covid-19 pandemic. What then is required for more liberal treatment of accused is all but the willingness of these criminal justice agencies to refrain from themselves indulging in predetermining guilt, and to act genuinely as neutral and respectful enterprises that champion procedural fairness.

Needless to say, this discursive shift was arduous not least because the practice of legal authorities displayed a deeply entrenched nature of paternalism or despotism. Yu Mou's inquiry into the tactics used by procuratorates to procure guilty pleas disclosed a systematic use of duress, threat, pressure, or persistent importunity that dominated plea leniency meetings with criminal suspects.Footnote ¹⁶⁸ In the eyes of prosecuting officers, the defendant was in an inferior position to the State party; hence her deference to authority was not only expected but also required.Footnote ¹⁶⁹ Therefore, any confrontational behaviour or attitude from the defendant was readily depicted as ‘a wish to escape punishment by not telling the truth’, and to be dealt with in a manner described as ‘iron-fisted’.Footnote ¹⁷⁰ Antithetical to norms of courteous interactions with the accused as encouraged in Tyler's procedural justice formula, Chinese procuratorates, according to Mou, were committed to overbearing and other paternalist strategies – often exemplified by ‘bitter tones, foul language, intimidating body language and slamming furniture’ – to obtain guilty pleas.Footnote ¹⁷¹ While ordinarily, procuratorates were not engaged in physical violence, their approach to defendants stood opposite to respect and thoughtfulness. Therefore, if at the macro level the institutional impediments to respectful treatment of defendants were axiomatic (eg, the overwhelming use of pre-trial detention), handling defendants with respect at the micro level can only be more far-fetched. All things being equal, the problem was not that procuratorates (or by extension all other legal authorities) have discretion to calibrate their conduct in plea leniency. Rather, the problem at issue was that procuratorates were the only one with discretion in this trial-avoiding and supervision-free process.

Trustworthiness – Are Defendants’ Interests and Needs Considered?

The procedural justice theory posits that respect with which the accused is afforded is the most important antecedent of both procedural justice assessments and judgements about the trustworthiness of motives of legal authorities.Footnote ¹⁷² In general, trustworthiness is symbiotically associated with respect. It refers to the degree to which an individual perceives that an authority is concerned about her well-being and acts to serve her best interests through sincerely helpful, respectful, and caring actions.Footnote ¹⁷³ Thus, perceived trustworthiness is enhanced when legal authorities demonstrate that they are motivated to take into account and accommodate the interests of defendants in the decision-making process.Footnote ¹⁷⁴ Translated in the plea leniency context, this can be illustrated in express efforts of authorities to address any claims asserted by defendants in support of more lenient treatment. Specifically, the defendant's ultimate hope to receive a less severe sentence in return for her guilty plea ought to be sensibly responded in order to reflect the spirit and crux of plea leniency.

Yet, the limited amount of scholarship on everyday practice in the plea leniency system makes any assessment of trustworthiness difficult. As discussed earlier, defendants generally lack opportunities to voice their position when interacting with legal authorities. Due to neglect, indifference, or venality of police, procuratorates, and counsel, defendants are denied a genuine conduit to make their positions articulate and apprehensible. While it may be convenient to attribute the quandary of trust-building to the inadequate functions of legal stakeholders, the greater force at play lies in the unequal power structure that undergirds plea leniency, which also more broadly defines the workings of China's criminal justice system.

At a glance, the plea leniency model in China bears some practical resemblance to the Anglo-American version of plea bargaining. Both practices are aimed at offering leniency in one way or another to the accused who admit guilt whilst accelerating justice delivery. Therefore, reaching a plea agreement with the prosecution in theory guarantees the defendant a reduced sentence and a much quicker turnaround time. And yet, likening plea leniency to plea bargaining seems cursory and overlooks the characteristic feature of the former which sets it apart from the latter. Fundamentally, the Chinese model of plea leniency is configured solely as one of granting leniency to those who show cooperation, rather than as a transactional exchange of leniency for a light sentence. If plea bargaining allows the defendant to negotiate charging and sentencing concessions to a certain degree,Footnote ¹⁷⁵ there is no deviation from the appropriate charges in the exercise of plea leniency. Put differently, plea leniency forbids charge bargaining and only affords defendants nominal leeway to discuss lenient sentencing options.Footnote ¹⁷⁶ In this respect, most scholars view plea leniency as a unilateral and illiberal procedure – the legal authorities dictate what charges to file based on already ‘clear facts and sufficient evidence’ and then gauge a narrow discounted range should the defendant plead guilty.Footnote ¹⁷⁷ From an official standpoint, this arrangement is justified – it prevents plea leniency from functioning as ‘a corrupt program where justice can be negotiated’.Footnote ¹⁷⁸ To maintain the dominant position of legal authorities in steering guilty pleas is to preserve the state authority in discovering ‘objective truth’ which professedly sustains the integrity of justice. Thus, unlike plea bargaining which turns on an adversarial orientation in the criminal process, plea leniency conforms to a larger culture of criminal justice in China – the State exercises monopoly over investigation, prosecution, and adjudication to ‘justly’, swiftly, and severely, if necessary, punish crime.

An authority-orchestrated criminal process, however, marginalises the role of defendants. Calls for lenient treatment are then answerable only when and unless they commensurate with the State's higher purpose of crime control. This is consistent with several findings in the nascent literature on plea leniency – guilty pleas do not necessarily foreshadow lenient sentences as anticipated. In an empirical study based on 6,876 judgment documents on the crime of intentional injury across the country, Fang Wang and Liang Guo found that the act of confession did not contribute to the overall leniency in punishment due to, in principle, a considerably high degree of discretion enjoyed by Chinese courts in determining the appropriate sentence.Footnote ¹⁷⁹ Though a set of nationwide guidelines for calculating sentences have been introduced since 2008,Footnote ¹⁸⁰ criminal sanctions are susceptible to the illogical sentencing practices of judges in accordance with the particulars of the case concerned.Footnote ¹⁸¹ It is important to note that the data examined in this study only covered the first year of plea leniency (ie, 2017) and did not abundantly capture the marked changes to this very program which factually shifted the sentencing power from courts to procuratorates.Footnote ¹⁸² Such shortcomings were overcome in a more recent study using a disparate research setting but generating a same conclusion. Drawing upon about 20,000 driving under influence cases in Fujian Province between 1 June 2013 and 31 January 2019, Yuhao Wu identified an unexpected intensification of penal severity as measured by probation decisions and the length of sentences.Footnote ¹⁸³ Specifically, his findings recorded a steady decrease in the probation ratio following the induction of plea leniency. As a result, instead of receiving softer criminal penalties, the defendants who had pled guilty were more likely to be sentenced to jail.Footnote ¹⁸⁴ Although their declared sentences appeared to be shorter than before, the term they actually served in jail increased by 0.41 months.Footnote ¹⁸⁵

The selected data of court judgments in Shanghai also point to the prevalence of imprisonment sentences in plea leniency-initiated summary trials. The descriptive statistics show that probation in all 13,611 cases is not under-represented (n = 6,480). But meting out incarceration for both serious and minor offences manifest itself as a more palpable pattern of sentencing in China's plea leniency programs. At this juncture, the recurrent imposition of jail sentences in the simplified procedure might be comprehensible, given that the majority of offences falling through this process were those legislatively punishable by imprisonment of three years and more. However, the lower proportion of probation vis-à-vis imprisonment in the fast-track procedure where non-violent and property-related offences constituted the largest pool of processed offences reaffirms the prior finding that guilty pleas and lenient sentences are not mutually interdependent. It is true that this statement is but a numerical observation of the raw data and does not involve the difference-in-differences estimation comparing the status quo to the past. Additionally, in contrast to the striking rate of pre-trial detention, imprisonment sentences were less widely applied where plea leniency is in action. What is problematic, though, is the State's continuing penchant for imprisonment as a means of carceral control over defendants deserving of light punishment because they accepted responsibility through guilty pleas and were unlikely to pose further risk to society. Table 8 and Table 9 list two primary risk factors (recidivism and the severity of crime) associated with the offenders who were sentenced to imprisonment.Footnote ¹⁸⁶ Both tables indicate that the risk levels of defendants bore little relevance to final sentencing decisions between incarceration and probation.

Table 8. Cases Involving Repeat and First-time Offending in which Imprisonment Sentences were ImposedFootnote ¹⁸⁷

Table 9. Three Top Common Types of Crime under which Offenders Received Imprisonment Sentences

As Table 8 shows, despite the finding that first-time offenders were over-represented in plea leniency cases, jail sentences were preferentially selected by judges. This sentencing paradigm was apparent in serious offences but particularly visible in the processing of minor offences. Table 9 presents the top three types of crimes – theft, dangerous driving, and picking quarrels and provoking trouble – where imprisonment sentences were most commonly awarded. Amongst these crimes, defendants who confessed to stealing money or property were more likely to be sentenced to prison. The value of the stolen money and property (2,000RMB) seems to be more a jurisdictional matter to determine the court procedure through which the theft cases are heard, rather than a way in which to differentiate the degree of punishment. In both simplified and fast-track proceedings, imprisonment sentences were discretionarily applied to the circumstances whereby the loss inflicted by stealing is either more or less than the prescribed value. These findings can be better understood in conjunction with the reading of a most recent study on the sentencing outcomes for stealing. Based on 1,000 copies of judges’ sentencing decisions, Wanting Liu discovered that a guilty plea was not the most decisive factor determining the use of incarceration or probation on offenders convicted of stealing.Footnote ¹⁸⁹ Rather, factors like the length of pre-trial detention, the forgiveness of victims, and the amount of returned property all play more crucial roles in the sentencing decision-making process, which tended to be considered in ways that reflect an ‘instinctive synthesis’ approach to punishment.Footnote ¹⁹⁰

By the same token, defendants pleading guilty to dangerous driving were easily subjected to incarceration. And yet, the overwhelming majority of defendants charged with dangerous driving were convicted of drunk driving, which was neither inchoate nor preparatory in nature but an act of negligence remote from causing substantial consequences. This disconnect between the severity of crime and the incarceration imposed is further exemplified in the sentencing of defendants pleading guilty to picking quarrels and provoking trouble. As previously noted, picking quarrels and provoking trouble is defined as deliberate obfuscation of language, extending its reach to street-level wrongdoings that disturb public order in various and a wide array of ways.Footnote ¹⁹¹ While almost any hooliganism-like action frowned upon by the authorities can be fitted into this offence, defendants caught up under the charge of picking quarrels and provoking trouble were equally exposed to imprisonment sentences irrespective of the magnitude of resultant injuries or damages.

In sum, it is far from clear that the legal apparatus has addressed the ultimate interest of defendants by imposing less severe punishment in exchange for their guilty pleas. Certainly, it is a priori unreasonable to reject the existence of the leniency consideration in every single case. But the dearth of a coordinated and definitive obligation that requires legal authorities to put lenient justice into effect has hindered the cultivation of an institutional culture benign to confessing defendants. This was engineered by the unsymmetrical power framework of the criminal justice system at large, and then compounded by the fact that plea leniency was designed also to be state-manipulated with no genuine bargaining power accorded to the accused. Absent real efforts of legal authorities to show goodwill and sensitivity in their handling of cases, plea leniency can only serve the interests of the State in lieu of the accused. There is little doubt that when the State does not concern itself with individual welfare under plea leniency, the perception of defendants about plea leniency is likely to be negatively impacted, which further jeopardises the legitimacy of this new practice for what it has promised but failed to fulfil.