Alge D, 'Negotiated Plea Agreements in Cases of Serious Fraud in England and Wales : Coercion, Consensus or Playing the System' (2013) 19(1) Web JCLI

Negotiated Plea Agreements in Cases of Serious and Complex Fraud in England and Wales: A New Conceptualisation of Plea Bargaining?

Daniele Alge

Lecturer, University of Surrey.

© 2013 Daniele Alge
First published in the Web Journal of Current Legal Issues
Citation: Alge, D., "Negotiated Plea Agreements in Cases of Serious and Complex Fraud in England and Wales: A New Conceptualisation of Plea Bargaining?", (2013) 19(1) Web JCLI


This article examines negotiated plea agreements introduced by the Attorney General in 2009 for cases of serious or complex fraud, and the degree to which these differ from plea agreements reached through informal plea bargaining in other types of criminal case. It first considers whether the formally negotiated agreements are a result of coercion being brought to bear on defendants, or of defendants ‘playing the system’ (the two most common criticisms of ordinary plea bargains). It is then argued that an alternative conceptualisation may be more appropriate in serious fraud cases. To this end, approaches to plea bargaining more commonly applied in the United States (consensual, concessions, and contractual models) are considered in light of the current context. It is submitted that whilst these approaches have only limited application to defendants in ordinary criminal cases, they may help explain the dynamic of plea agreements in serious fraud cases. This in turn provides a basis upon which to assess the fairness of negotiated pleas in serious fraud cases, and highlights issues which lie at the core of the plea bargaining debate.


1. Introduction
2. The current legal framework
3. Traditional approaches to plea bargaining: coercion and ‘playing the system’
4. Coercion in serious fraud cases
5. ‘Playing the system’ in serious fraud cases
6. Alternative models of plea bargaining
6.1 Plea bargaining as consensus
6.2 Bazaar or supermarket style justice? Concessions and contractual approaches
7. Conclusions

1. Introduction

Plea bargaining in England and Wales has received renewed levels of attention in recent years in terms of policy, case law and (to an extent) academic research. Since Baldwin and McConville’s ground-breaking study Negotiated Justice in 1977 there had been surprisingly little research on the practice in this jurisdiction. This is in stark contrast to the United States where plea bargaining has remained firmly on the agenda for several decades. The announcement in 2008 of powers to be granted to the Serious Fraud Office (SFO) to negotiate plea agreements in cases of serious fraud has inevitably sparked further interest, albeit primarily in practitioner journals. (1) Informal plea bargains by means of the sentence discount, reduced or lesser charges, and Goodyear [2005] EWCA Crim 888, [2005] 1 WLR 2532 advance indications of sentence have long been a common feature of criminal trials. However, SFO negotiated plea agreements are the first openly regulated and formalised form of plea bargain in England and Wales. Several high profile cases (on which this article will focus) have involved the use of the new powers, most notably R v Innospec [2010] Lloyd’s Rep FC 462, R v BAE Systems Plc [2010] EW Misc 16 (CC) and R v Dougall [2010] EWCA Crim 1048, [2011] 1 Cr App R (S) 37. This in turn has fuelled further academic debate about the nature and role of plea bargaining in such cases (Vamos 2009; Watson 2010), and coincides with new work on plea bargaining in ‘ordinary’ criminal cases (Rauxloh 2012).

As the procedure for negotiating plea agreements in cases of serious fraud is in its infancy, and plea bargaining has previously always taken place behind closed doors in England and Wales, there has been considerable attention paid to the new process, not least by the judiciary. In Innospec, Thomas LJ issued lengthy sentncing remarks which considered the duties of the prosecutor and the way in which the agreement had been reached. (2) Most notably, Thomas LJ stated his opinion (at para. 40) that the fine imposed was “wholly inadequate” but that he felt bound by the terms of the agreement already reached between the SFO, the US Department of Justice, the Securities and Exchange Commission and the US Office of Foreign Assets Control. This in itself demonstrates the need for a closer analysis of the nature of the agreements, and a level of judicial disquiet regarding the potential operation of future plea agreements in fraud, and other, cases. During the consultation stage of the Attorney General’s Guidelines, a sub-Committee of Circuit Judges stated that they were “fundamentally opposed” to the extension of any plea bargaining framework beyond serious or complex fraud cases (Council of HM Circuit Judges June 2008, at para. 6). Ostensibly, plea agreements in serious fraud cases have not been introduced as a precursor to formalised plea agreements in other types of criminal case, but a degree of scepticism about that assurance would be prudent.  When the powers were first announced, Baroness Scotland, the Attorney General at the time, was at great pains to distinguish them from any form of plea ‘bargaining’ (Attorney General’s Office Press Release, March 18th 2009). Only three years on, it is accepted that the SFO agreements are indeed plea ‘bargains’ and the term is used by the SFO in its own online information. (3) The current SFO plea agreements are therefore all the more important, not just in their own right, but because they have the potential to become a blueprint for formalised plea bargaining on a broader scale.

The focus of this article is on the approaches which can be applied to better understand the nature of the agreements themselves; in particular the balance of the parties’ bargaining powers. The US literature on the application of models of bargaining to plea agreements is exponentially more developed than the UK literature, and several distinct approaches can be identified: consensual, concessions and contractual. It is argued that these models may indeed be applicable to the SFO procedure, and that SFO agreements possess different characteristics to those of ordinary plea bargains, which in turn gives rise to a different set of concerns about the justice of bargained for case dispositions. 2. 2>2. The current legal framework

Before exploring the applicability of approaches to, and models of, plea bargaining to negotiated plea agreements, it is first necessary to outline the current legal framework. In ordinary cases (those which do not fall within the SFO definition of serious or complex fraud), (4) there is no formal framework of plea bargaining.

The closest the UK courts came to tackling the issue was the case of R v Turner [1970] 2 QB 321 in which plea bargaining was referred to as “the vexed question of so-called plea bargaining”. The resulting Turner rules state, inter alia, that:

“Counsel must be completely free to do what is his duty, namely to give the accused the best advice he can – if need be in strong terms. This will often include advice that a plea of guilty, showing an element of remorse, is a mitigating factor which may well enable the court to give a lesser sentence than would otherwise be the case.” Turner [1970] 2 Q.B. 321, 326

The appeal in Turner was allowed only on the basis that the appellant had been given the impression that counsel’s views on the outcome of the case on a guilty plea compared with a not guilty plea came directly from the trial judge, not because counsel had exerted any pressure on Turner. The Court of Appeal felt this was “a very extravagant proposition, and one which would only be acceded to in a very extreme case” [1970] 2 QB 321, 325. The judgment in Turner also stated that the judge should never disclose the sentence he was minded to impose, but this element of the rules has been superseded by the Criminal Justice Act 2003. Schedule 3 of the Act provided for advance indications of sentence whereby a defendant may request an indication of the maximum sentence, if he were to plead guilty at that stage. If an indication is given, it is binding on the court. In Goodyear, the Court of Appeal laid down additional guidelines as to how sentence canvassing should work in practice in the Crown Court and held that:

“A judge should never be invited to give an indication on the basis of what would appear to be a ‘plea bargain’. He should not be asked or become involved in discussions linking the acceptability to the prosecution of a particular plea or bases of plea and the sentence which might be imposed and he should not be asked to indicate levels of sentence which he might have in mind depending on possible different pleas.” [2005] EWCA Crim 888, [2006] 1 Cr App R (S) 6 [67)

This denial that advance indication of sentence has anything to do with plea bargaining is difficult to rationalise, other than on the narrowly conceived basis that a response to a request is just that, rather than a bilateral exchange of concessions in the stricter sense of a ‘bargain’. But what is a formal, judicial indication of a light(er) sentence in exchange for a guilty plea, if not a plea bargain?

Plea bargaining is undoubtedly somewhat hidden behind a façade of measures introduced to maximise administrative efficiency and the rhetoric of ‘bringing offenders to justice’. Nonetheless, it is commonplace and is facilitated (informally) by several mechanisms, in addition to Goodyear indications. The lynchpin is the sentence discount, which has been increasingly structured since the first publication of the Sentencing Guidelines Council Reduction in Sentence for a Guilty Plea Guideline in 2004. Every defendant will be made acutely aware of the fact that in exchange for a guilty plea, he can expect up to one third reduction in sentence from that which he would receive following a conviction at trial, with the greatest discounts available to those who plead guilty at the earliest opportunity. Charge bargains are also a common means of case disposition and are facilitated by the exercise of the prosecutor’s discretion in reducing or dropping charges in exchange for a guilty plea. This is fully compatible with the Code of Conduct for Crown Prosecutors. (5) Additionally, bases of plea can be used to agree a less culpable version of events (with the resulting reduction in sentence) in exchange for a guilty plea, in what Darbyshire (2000) refers to as “fact bargaining”.

So, despite plea bargains being a by no means unfamiliar feature of the criminal justice system in England and Wales, SFO plea agreements nonetheless represent a marked departure from the system of plea bargaining which applies in ordinary criminal cases. The Attorney General’s Guidelines provide that discussions may be initiated by either the prosecutor or the defence and should not usually commence until the suspect has been interviewed under caution (para. C2). All discussions are to be recorded in writing, and even the invitation to initiate discussions should be in the form of a letter (para. C4). If an agreement is successfully reached as to pleas and charges, the parties should attempt to present a joint written submission to the court as to sentence, although the Guidelines do make clear that the submission as to sentence is not binding on the court (para. D12). The Guidelines also set out the general principles that the prosecutor must act “openly, fairly and in the interests of justice” and justice is interpreted as the requirement that the plea agreement reflects the seriousness of the offending and allows victims, the court and the public to have confidence in the outcome. They state that “[t]he prosecutor must not agree to a reduced basis of plea which is misleading, untrue, or illogical” (para. B2).

As SFO plea agreements are more clearly structured, regulated and transparent than their informal counterparts, this gives rise to questions about what this means for the traditional view of plea bargaining. Is fraud ‘different’? What is the nature of the agreements? To what extent does the balance of bargaining power shift? The remainder of this article seeks to address these questions.

3. Traditional approaches to plea bargaining: coercion and ‘playing the system’

The two most prevalent criticisms of (informal) plea bargaining are diametrically opposed: (i) that it operates coercively against defendants and (ii) that it allows defendants to ‘play the system’ and evade appropriate convictions and / or sentences.

The groundswell of academic opinion encapsulates the first criticism, that plea bargaining is undesirable because it disadvantages the defendant by robbing him of the genuine choice to exercise his right to trial and places pressures upon him to plead guilty. The sentence discount is regarded not as a benefit, but as a “trial penalty” (Darbyshire 2000) which punishes defendants for wasting the state’s resources if they are convicted following a trial. Additional charge, fact or sentence bargains are viewed as further undue inducements to plead guilty. If a plea bargain is struck, evidence is not tested in open court and the prosecution is not required to prove the case against the defendant. Instead, the defendant is required to make an assessment of the likelihood of conviction at trial, and if the risk of conviction and a higher sentence appears too great, the defendant is likely to opt for a plea bargain. This may seem a rational choice, and if the defendant is indeed factually guilty gives rise to the criticism that it allows the defendant to escape with a lighter punishment than is his just desert. However, this criticism only applies if the defendant was able to make an informed choice, based on full knowledge of the relevant issues, was not unduly influenced by threats of a higher sentence, and received quality legal advice. As several studies suggest, this is not so easy to come by. (6)

Not only do inducements to plead guilty operate unfairly against factually guilty defendants, but Baldwin and McConville (1977), McConville (1998) Hodgson and Belloni (2000), Darbyshire (2000, 2005), Ashworth (2005), and Sanders, Young and Burton (2010) all adopt the position that these pressures are so powerful that even innocent defendants may plead guilty.

The media and policy approach to plea bargaining in England and Wales takes a very different stance. The practice is generally portrayed by the media as a loophole which allows defendants to escape the punishment they deserve, a view which is reinforced by the frequency with which defendants in US crime dramas are able to ‘cut a deal’ and which many members of the public may mistakenly believe to be representative of all types of plea bargain across all jurisdictions. There is a perception (part of a wider phenomenon of public opinion that sentencing is too lenient, see for example Hough and Roberts (1998)) that the sentence discount allows defendants to ‘get away with it’ and to play the system by trying to extort as many concessions as possible from the state. This perception is not limited to the media, nor to the public; it also comes across markedly in the 1993 Royal Commission on Criminal Justice, the 2001 Auld Report, and the Justice for All White Paper, published in 2002.

Despite the gulf between these perspectives on plea bargaining, both sets of objections do have in common a distaste for the betrayal of adversarial principles which plea bargaining entails. The resulting “bazaar atmosphere” (McDonald 1979) in which plea bargaining takes place is seen to reduce the criminal process to an unprincipled series of exchanges of concessions and rights.

4. Coercion in serious fraud cases

When the criticisms outlined above are applied to formal plea agreements in cases of serious fraud, however, it becomes apparent that they are somewhat differently contoured. It appears that there is considerably less scope for the prosecution or court to act coercively and pressurise the defendant into a bargain. Although the stakes may be high for the companies and individuals involved in fraud, bribery or corruption, they are nonetheless unlikely to be as high as the loss of liberty faced by, for example, a defendant charged with his third domestic burglary. The defendant charged with burglary faces a potential sentence of the maximum 14 years in prison, and perhaps nine or ten years even following a guilty plea. When the defendant in an SFO investigation is a company, the potential risk to individuals amounts at most to loss of reputation, employment or earnings. Even where individuals are charged, an immediate custodial sentence is unlikely, following a guilty plea. In Dougall [2010] EWCA Crim 1048, [2011] 1 Cr App R (S) 37 the defendant had taken part in the large scale corruption of foreign officials, but had cooperated fully with the investigation and assisted the SFO in its prosecution of others involved, as part of a plea agreement. In the agreement, the Director of the SFO made the submission to the court that any custodial sentence imposed should be suspended. Dougall was sentenced to 12 months imprisonment, and appealed; his appeal was allowed and the sentence was reduced to the suspended sentence (of 12 months) envisaged by the plea agreement.

The case is particularly significant as the Court of Appeal took the opportunity to comment on the appropriateness of suspending sentences in cases of this nature. In his judgment, Lord Judge CJ was critical of the SFO for having in effect advocated on behalf of the defendant and requesting that the court impose a particular sentence, stating that: “look where we may, in our criminal justice system, agreements between the prosecution and the defence about the sentence to be imposed on a defendant are not countenanced” [2010] EWCA Crim 1048, [2011] 1 Cr App R (S) 37, [23]. The SFO had in fact gone further still in its submission in Dougall, by suggesting that as a general rule, it would not be appropriate to impose immediate custodial sentences on offenders in similar cases:

“Unless a ‘white-collar’ defendant, in an appropriate case, has the prospect of avoiding an immediate custodial sentence by fully co-operating with the authorities the important public interest in him doing so will not be secured. For such a defendant it is the fact of being sent to prison that matters, not the length of the sentence…” [2010] EWCA Crim 1048, [2011] 1 Cr. App. R. (S.) 37, 32

Although the appellate court endorsed the trial judge’s view that a convicted offender in a fraud case was still a ‘common criminal’, undeserving of special treatment, Lord Judge CJ acknowledged that it would be “unrealistic to ignore these considerations” and that it would “normally follow” that a defendant whose sentence was 12 months or less, had cooperated fully with the investigation, and whose case had mitigating factors, should have his sentence suspended [2010] EWCA Crim 1048, [2011] 1 Cr App R (S) 37, [36]. If, even at a serious level of large scale criminality, the worst case scenario for an individual, following a guilty plea, is a suspended sentence, and the prosecutor himself advocates this position, then the argument that plea bargains operate coercively in such cases is considerably weakened. Undoubtedly, the defendant will feel some pressure to plead guilty precisely in order to avoid a custodial sentence, but it is submitted that this pressure is not exerted or likely to be felt in the same manner as in other cases. Incentives to comply and engage with SFO plea bargains arise not so much from the coercive legal pressures prevalent in other types of case, but rather stem from an organisational, business, or perhaps personal rationale. Individuals or companies have a vested interest in cooperating once evidence of fraudulent activity has been discovered. This may be to evade a custodial sentence, but is more often to minimise damage to their companies, reputations, and lifestyles. These are powerful self-interests and may well induce individuals to cooperate, but where liberties are not at stake, the reach of coercive power is much reduced. (7)

It is of course still true that an SFO plea agreement eliminates the need for the prosecution to prove the case in open court, but the defendant in a serious fraud case has far greater resources. He is likely to be educated, informed of the legal and factual issues, aware of the options open to him and represented by qualified, specialist lawyers. This is in stark contrast to the position an ordinary defendant may find himself in. He is likely to possess limited knowledge of the legal issues, perceive his options to be ‘prison’ or ‘less time in prison’ and be represented by an overworked barrister who may have his own interests in encouraging the defendant to plead guilty. (8)

It seems therefore, that the balance of power in an SFO plea agreement is not such that the prosecution or the court is able to coercively pressurise a defendant (whether a company or an individual) into a plea bargain. The one significant exception to this is if a defendant who has started the process of reaching a plea agreement chooses to opt out. The Attorney General’s Guidelines provide that the prosecutor may not use the fact that the defendant has taken part in plea discussions, nor any information disclosed during those discussions, as evidence against that defendant, should the discussions fail. This seems in line with the stated objectives of fairness and justice. However, there are several caveats. If a signed plea agreement has been concluded, this may be used as confession evidence against the defendant, should he choose to opt out of the agreement after that stage. Further, even if there is no signed plea agreement, information provided by the defendant may still be used against him in any prosecution for related charges, as may information gleaned from enquiries made as a result of provision of evidence by the defendant (para. C8). This latter point is particularly disingenuous, as any information could potentially be deemed to be the result of ‘enquiries’. In short, once a defendant has commenced plea discussions with the SFO, it becomes very difficult to backtrack, and in this scenario, the SFO undoubtedly retains the greater bargaining power.

5. ‘Playing the system’ in serious fraud cases

In ordinary criminal cases, the argument that plea bargains allow defendants to ‘play the system’ is unfair and misjudged. The defendant has few bargaining chips and faces considerable pressures to plead guilty and forgo his right to a trial, regardless of the strength of the evidence against him. In serious fraud cases, however, there is greater scope for the argument to carry weight. In lengthy, complex and often international investigations, those accused of fraud, bribery, or corruption have recourse to greater bargaining powers, with the consequence that the outcomes may be seen as unduly lenient.

The plea agreement in the BAE Systems case [2010] EW Misc 16 (CC) in particular, makes a mockery of the provision in the Attorney General’s Guidelines that the plea agreement ought to instil public confidence and not be based on a plea which is misleading, untrue or illogical. BAE Systems had been accused of wide-ranging multi-national bribery over several years, yet pleaded guilty to one charge of failing to keep reasonably accurate accounting records contrary to s.221 of the Companies Act 1985 in relation to its activities in Tanzania, on the basis that it did not admit corruption. The company was fined only £500,000 after having indicated a willingness to pay a £30m penalty as an ex gratia payment to the people of Tanzania, but that any fine would be deducted from this sum. In reaching his sentence, Mr Justice Bean stated that he felt a ‘moral pressure’ to keep the fine low for this reason. BAE Systems had also agreed to pay a $400m fine to the US relating to its activities in Saudi Arabia and Eastern Europe, and as part of the settlement, the SFO agreed that it would not pursue its investigations into those other potential charges. The SFO had also granted an indemnity for all offences committed in the past, whether disclosed or otherwise, which Justice Bean described as “surprising”. Bean J’s sentencing remarks are an exercise in judicial restraint, it seems clear that he felt himself unduly constrained by the terms of the settlement agreement, and makes several references to his lack of power to vary the terms or to pass a sentence which would better reflect the scale of the alleged offending. (9)  Similarly, in Innospec [2010] Lloyd’s Rep FC 462 the court imposed a penalty lower than the sentencing judge felt appropriate, describing the $12.7m fine as an inadequate penalty which did not reflect the scale of the criminality involved. Nonetheless, Thomas LJ stated that he felt bound to do so as the US Federal District Court had already approved the plea agreement and that in the context of the global agreement, it would have been unjust to impose a greater fine (R v lnnospec Limited [42].

The fact that SFO plea agreements may involve international agencies and global settlements can thus create leverage for defendants in a way which will never be a feature of ordinary criminal cases. Similarly, defendants in bribery or corruption investigations are in a position to offer reparations to nations, and again this is a bargaining tool not open to defendants in ordinary cases (not even in ‘ordinary’ lower level fraud cases). Additionally, whilst evidence not being tested in court is conventionally cited as an example of plea bargaining’s disregard for the defendant’s due process rights, in a high profile corruption case, this can work to the distinct advantage of the defendant company. There are no doubt many damaging details of a company’s activities which can be kept out of the public eye by virtue of a plea agreement. To an extent the principle applies to defendants in all criminal cases, but the advantages to a large organisation which wishes to continue trading are undoubtedly far greater. A hugely significant factor which facilitates these outcomes is that they are also in the SFO’s interests; its priorities are not those of the CPS. The SFO has wider economic interests, for example in not forcing a company out of business and punishing innocent employees. As its submissions to the court in Dougall demonstrate, the SFO is keenly aware that unless it can be seen to make deals which are attractive to defendants, it risks losing cooperation, and with that its own effectiveness.

In some senses then, ‘playing the system’ appears to be an accurate way of describing the outcomes of negotiated plea agreements in serious fraud cases. Defendants, particularly in high profile cases, have access to considerable bargaining power and resources with which plea negotiations can be influenced. However, this may in fact be as inaccurate a conceptualisation of plea bargaining as it is in ordinary cases, albeit for different reasons. In serious fraud cases the system itself is also playing, to an even greater extent than in other cases. It derives considerable benefits as a result. In 2010–11 the average SFO investigation incurred costs of £910,000 and took 24 months to conclude (SFO Annual Report and Accounts 2010–2011). Plea agreements which reduce those costs, require shorter investigations, and allow the SFO to demonstrate its ‘success’ in tackling fraud are hugely advantageous to the agency. This is true to some degree of CPS prosecutions, but as the SFO took only 17 cases to trial in 2010–11 and almost all SFO investigations are deemed newsworthy, the scope and scale of the prosecutions undertaken by the agencies are entirely different.

The notion that defendants in serious and complex fraud cases are being allowed to manipulate the system to their own advantage is therefore overly simplistic, and a more nuanced conceptualisation of the plea agreements is necessary. It is submitted that this can be better achieved by using alternative models of plea bargaining as a lens through which to examine the agreements.

6. Alternative models of plea bargaining

The words ‘consensus’, ‘concessions’ and ‘contract’ are used frequently, but inconsistently, within the existing literature on plea bargaining, which has developed in a piecemeal fashion over the past decades. This article takes ‘consensus’ to mean something akin to the dictionary definition of ‘agreement in opinion’. If a plea bargain is arrived at by consensus, the final outcome reflects the defence and prosecution’s agreed assessment of the correct, or just, outcome. This differs from a contractual or concessions model in that although previous literature does not consider whether they are synonymous, it is argued that they both have at their core the principle that rights, benefits and risks can be traded. The final outcome may not be arrived at by consensus as to the ‘right’ outcome, but it is an agreement to which both defence and prosecution have subscribed. (10)

6.1 Plea bargaining as consensus

Plea bargaining has previously been considered in the light of a consensual model of justice by the North American literature (Heumann 1974; Rosset and Cressy 1976; Church 1978; Jacob 1984; Nardulli, Eisenstein and Flemming 1985) and by literature on continental European criminal justice systems (Jung 1997), but not in the UK context. Nardulli, Flemming and Eisenstein considered both concessions and consensus models and described the consensus model as one which stresses the importance of shared understandings in “lubricating the court’s machinery” (1985, p.1107). The authors adopt Rosset and Cressy’s view that:

“Even in the adversary world of law, men who work together and understand each other eventually develop shared conceptions of what are acceptable, right and just ways of dealing with specific kinds of offenses, suspects and defendants. These conceptions form the bases for understandings, agreements, working arrangements and cooperative attitudes.” (Rosset and Cressy 1976, p.90). (11)

When applied to ordinary criminal cases, the overwhelming flaw of this consensual conceptualisation of plea bargaining, if it is to be viewed as a legitimate means of case disposition, is that the consensus never involves the defendant.  The ‘shared conceptions’ are invariably those of the defence and prosecution lawyers and perhaps the judge and other court staff more widely. Lawyers effectively exist in a microcosm in which their concepts of ‘just’ outcomes dominate the delivery of criminal justice, and these concepts may conflict with defendants’ best interests. Some commentators (although significantly, writing primarily of inquisitorial systems) have nonetheless described plea bargaining as a consensual exchange from which defendants can benefit. Jung argues that:

“The notion of criminal law as the ensign of the monopoly of power vested in the state, and as clearly distinct from private law, begins to falter. Elements of negotiation and participation, hitherto restricted to the sphere of private law litigation, are proliferating in all phases of criminal procedure. This indicates a shifting equilibrium between state, society and the individual” (1997, p.116).

Given the power differential between lawyers and defendants, and evidence of negative attitudes expressed towards defendants, it seems unlikely that defendants in ordinary criminal cases can meaningfully engage with a consensual process of plea bargaining. (12) However, some features of a consensual model may be applicable to negotiated pleas in serious fraud cases. As stated earlier, there is a reduced power differential between the defendant and the legal professionals involved in an SFO investigation. As Katz described it in the US context there is less “social distance” (1979, p.431) between the parties, which makes them more likely to exercise their discretion in favour of each other. Further, when there is less at stake for the defendant in terms of sanctions, there is greater scope for genuine agreement. Few defendants would heartily agree to a prison sentence, but a company director may well genuinely agree to his company paying a penalty. Consensus is easier to reach when individual liberties are not affected. Ultimately the result, as with all plea bargains, is that those with bargaining power determine the outcome of the case. In ordinary criminal cases this almost always excludes the defendant (though includes the defence lawyer), but in serious fraud cases the defendant often possesses sufficient power to play a greater role in the outcome. This does not necessarily equate to a genuine consensus though, and certainly may not tally with what victims or society perceive to be the right outcome. The latter issue is particularly problematic in cases where corruption and exploitation of a third world country are involved. Indeed, many sections of the media expressed their outrage at the conclusion of the BAE Systems case, and to some extent following the judgments in Innospec and Dougall. Moreover, any plea negotiation takes place within an adversarial dynamic, regardless of the type of case or defendant. No matter how fully an individual or an organisation cooperates, they are not willingly in that position and it would be naïve to expect genuine consensus. This article therefore adopts the position that the concessions / contractual approach discussed below may be more appropriate.

6.2 Bazaar or supermarket style justice? Concessions and contractual approaches

The concessions model of plea bargaining is well established, and is the predominant means by which plea bargains are viewed within much of the North American literature; the application of a contractual perspective to the concessions model gained momentum during the early 1990s (Scott and Stuntz 1992a, 1992b; Schulhofer 1992). The essence of the model is that the wide range of issues within a criminal case which can give rise to strengths and weaknesses, such as evidentiary flaws or the credibility of witnesses, become tools by which concessions can be extracted, agreements made and deals struck. Significantly, this model does not assume that genuine consensus is possible. The dynamics of plea bargaining within this approach can be viewed either as an unprincipled “bazaar”, or a more regulated and orderly “supermarket” (McDonald 1979, p.386; Nardulli, Flemming and Eisenstein 1985, p.1106).

To clarify what is meant by the bazaar analogy in this instance; it is taken to describe a system of case dispositions in which there is some degree of disorder and inconsistency, scope for haggling, the values of commodities (sentences and charges) fluctuate. The best deals are to be had by those who maintain a good relationship with those with whom they trade frequently. By contrast, Nardulli, Flemming and Eisenstein write that courts may “also operate more like supermarkets in that they are more orderly than the freewheeling concessions model may suggest” (1985, p.1109). In a ‘supermarket’ concessions model of plea bargaining, prices (that is, sentences, charge reductions and outcomes) would be more firmly fixed, with less scope for haggling, but also less scope for uncertainty and ‘bad deals’. The language used during criminal cases echoes that of contractual exchanges: court room actors refer to guilty pleas being ‘offered’ or ‘accepted’, and ‘deals’ being ‘taken’. The practice of the Crown or the defence indicating that it would be willing to offer, consider or accept guilty pleas to certain offences mirrors that of contractual invitations to treat. The use of the stock phrase ‘indicate a willingness’ becomes a device with which to legitimise the haggling.

Scott and Stuntz argued that plea bargains should be viewed in terms of contract theory, rather than due process rights, and that if plea bargains are analysed contractually, they are not necessarily coercive. They reasoned that defendants should have the freedom to contract or exchange entitlements in criminal proceedings and that to deny them that ability (by abolishing plea bargaining) would undermine the value of those entitlements (1992a, p.1913). Easterbrook had similarly written that defendants are entitled to either use or sell their right to trial (1991, p.1975). Scott and Stuntz argued that the elements of contracts which would make them unenforceable (such as duress or unconscionability) did not, as a matter of course, apply to plea bargains. They felt that large sentencing differentials did not equate to guilty pleas entered under duress, but rather that the right to take a case to trial was highly valuable and that the prosecutor was willing to pay a high price for it (1992a, p.1921). They argued that a bargain is only unconscionable if it is a ‘take it or leave it’ offer which does not react to individual preferences (ibid., at p.1924) but that plea bargains involve bargains whereby the terms of the agreement can be individualised. This latter point is particularly resonant in the context of the highly tailored plea agreements entered into by the SFO.

Even in ordinary criminal cases there are indications that the UK system of plea bargaining is becoming increasingly contractual. This is most evident of Goodyear indications which are a clear and enforceable bargain between a defendant and the state. They empower the defendant by providing him with the certainty that the sentence imposed will not exceed a specified maximum. That Goodyear indications are enforceable is not doubted, and R v McDonald [2007] EWCA Crim 1117, [2008] 1 Cr App R (S) 20 shows that the Court of Appeal considers Goodyear indications to be binding to the extent that even if given in error, they override statutory requirements to impose an indeterminate sentence for public protection. There is also a general trend towards increased clarity in rewards for guilty pleas following the Sentencing Guidelines Council’s Guidelines on the issue and the graduated system of sentence discounts developed. Cases such as Attorney-General's Reference No. 44 of 2000 (Robin Peverett) [2001] 1 Cr App R 27 suggest that the Court of Appeal will prioritise the enforceability of promises made to defendants over undue leniency or conflicting statutory sentences. These recent developments have the effect that a guilty defendant has (some) increased benefit from the greater transparency and certainty of the system within which he may exchange his right to trial for other concessions, and it is submitted that a limited contractual analogy is apt.

However, even Scott and Stuntz write that “contract makes the disquiet of critics seem sensible too, since the bargaining dynamic shortchanges the innocent” (1992a, p.1968). A clearer, enforceable exchange of concessions will not only be more attractive to guilty defendants, but also to innocent defendants. A contractual conceptualisation of plea bargaining can only be defensible if it can ensure, in so far as possible, that factually and / or legally innocent defendants are not induced to plead guilty contrary to their own best interests. Baldwin and McConville examined the possibility of a defensible model of plea bargaining but concluded that as the system could not ensure that the defendant’s plea was free and voluntary, or that each case was disposed of according to the evidence, it could not be defended it was “not calculated to avoid injustice” in the way in which a trial was (1979, p. 216). The contractual analogy can also be critiqued on the basis that ordinarily, a defendant has few concessions to offer. In fact, he really only has one; his guilty plea. Other factors such as the charge or basis of plea are subsidiary, and the prosecution has far greater contractual bargaining power.

Whilst the above is true of non-fraud cases (that is, there is some evidence of plea bargains exhibiting contractual characteristics, but defendants do not have genuine contractual freedoms and there remains a risk to innocent defendants), it is argued that SFO plea agreements are different. Defendants in serious fraud cases are more likely to be resource rich, be more adequately represented, and aware of the SFO’s own interests in achieving a settlement. Significantly, they have a greater range of concessions to offer. The larger the organisation and the more widespread the fraudulent practices of which it is accused, the more concessions it may have at its disposal. Corporations can agree to change business practices, to restructure, and to submit to independent reviews of adherence to regulations and good practice. Often this process will already have begun prior to the conclusion of any plea agreement, and can be used as leverage and mitigation. Additionally, payments and other reparations can be made to nations affected by the corruption (even if the corruption is not admitted), as with BAE Systems’ $30m ex gratia payment to Tanzania after pleading guilty to one minor accounting offence. When this outcome is transposed to a hypothetical ordinary defendant, the gulf between SFO plea agreements and ordinary plea bargains becomes all the more apparent. The equivalent would be a defendant accused of multiple high value thefts over a period of years in several villages, who once discovered cooperates with the prosecution by agreeing to plead guilty to one count of (for example) dishonestly retaining a wrongful credit under s.24 of the Theft Act 1968. The defendant would agree to make a reparation (a considerable sum but not one which would cripple him financially) to the people of one of the villages, in addition to promising to mend his ways, but without admitting to any of the thefts. Part of the agreement would be that the alleged thefts in the other villages would not be investigated and the CPS would make a submission to the court that the defendant should be treated leniently. Clearly, this is unthinkable to the point of being laughable. If we leave aside the practical impossibilities of dealing with high volume criminal cases in this way, the real issues are that ordinary defendants simply do not have these options or this degree of leverage over the prosecuting agency.

Returning to the bazaar supermarket analogy, plea bargains in serious and complex fraud cases are perhaps the high street delicatessen of plea bargains. The goods on offer (concessions) are varied and individually tailored, not subject to the standardised going rates of a chain supermarket, nor to the daily fluctuations of a bazaar, but with some expectation that prices (sentences) will not exceed a certain level (a suspended prison sentence). The fact that if a defendant opts out of a plea agreement, information can still be used against him mitigates against complete contractual freedom. But even in conventional contractual relationships, it will often be the case that one party is at an advantage or that there are limitations to the freedom. If a purely contractual approach is adopted where plea bargains are viewed in terms of concessions, not rights, then it is no different to a consumer having to pay a fee upon cancellation of a flight booking.

7. Conclusions

In summary, there is currently little (if any) evidence to suggest that defendants in SFO cases are pressurised into pleading guilty. Further, empirical, work is needed to examine the nature of the plea bargains more closely, but in the high profile cases discussed, the defendants do not appear to exhibit any of the vulnerabilities of ordinary criminal defendants which would render them susceptible to being pressurised into undesirable or unjust plea bargains. There is, on the contrary, the suggestion that large organisations are able to ‘play the system’ by offering to pay penalties, entering into global settlement agreements, and escaping more serious charges and sentences as a result. This, however, is too crude a view of the situation. The SFO is itself subverting the adversarial system; defendants cannot be blamed for the agency’s desire to save money and demonstrate its own efficacy by favouring settlements over prosecutions. When other models of plea bargaining are considered, consensual approaches are unlikely to be founded in the reality of the practice. Consensus to which the defendant is party requires a level of voluntariness which will never be present in criminal proceedings. Contractual perspectives more accurately describe the exchanges of entitlements, based on relative bargaining positions determined by the evidence and other facts of the case, which are a feature of SFO plea agreements.

It is doubtful whether even the application of a full range of contractual safeguards could ever position an ordinary defendant at a level of bargaining power sufficiently close to that of the prosecution. Given the inherent problems of the principal – agent dynamic between the defendant and his barrister it is unlikely that a defendant can achieve equality of bargaining power with his own lawyer, let alone the prosecution. (13) The greatest inducements to plead guilty would invariably be offered to those defendants with the strongest cases, as their right to trial would have greater value. It is questionable whether defendants in this situation could be immune from unconscionable pressure to ‘sell’ their chance of acquittal. However, in SFO cases, defendants and / or the organisations under investigation are in a considerably better position to contract and at times appear to have greater bargaining power than the prosecution. Concluded SFO plea agreements are joint endeavours between the defence and the prosecution. This is demonstrated most clearly in Innospec and Dougall by the SFO advocating on behalf of the defence in putting forward a suggested (lenient) sentence and openly expressing the view that it was in the agency’s interests that criminals in such cases should be kept out of prison.

This begs the question as to whether this degree of cooperation, within the context of what appears to be a contractual relationship, is desirable or just in serious and complex fraud cases. The former Attorney General, Lord Goldsmith, has said that:

“…I don’t see why it should not be possible for experienced prosecutors, who can understand the public interest, together with well advised corporations to be able to reach an agreement and then say to the judge this is what we think and this is why. It is not traditional English criminal sentencing thinking, but I don’t agree with that.” (The Guardian 5th Jan 2011).

This would mean that prosecutors and corporations would be entrusted with the task of reaching just outcomes; Lord Goldsmith implies that the judicial role in overseeing any such agreements should be minimal. If the SFO has a vested interest in reaching an agreement, is willing to assist the corporation in achieving an agreeable outcome in order to do so, and the corporation has considerable bargaining power, then there is no guarantee that the resulting agreement will reflect the wider interests of justice. Far from tackling fraud and corruption, this perpetuates the perception that white collar criminals are less culpable and that multinational organisations are beyond the grasp of the criminal law. Of equal concern is the possibility that the SFO model of plea bargaining could in the future be adopted for other offences, on the basis of its perceived success from the perspective of both prosecutors and policy makers. As the discussion above has demonstrated though, ‘ordinary’ criminal cases are rather different. Even if whole scale plea bargaining were to be formalised in the same way as the SFO procedure, this would not guard against undue pressures to plead guilty in cases where much more is at stake for the individuals involved.

There still exists a real gulf between policy and practice regarding plea bargaining in England and Wales, most evidently when it comes to informal plea bargains, but also regarding the SFO procedure. In Baroness Scotland’s initial announcement of the powers, she stated that the measures were:

“[N]ot about offering discounts, immunity or incentives to fraudsters. It doesn’t require a defendant to assist the prosecution and is careful to avoid a perception of ‘plea bargaining’ associated with the US.” (Attorney General’s Office, Press Release, 18th March 2009)

The key issue here seems to be the avoidance of negative perceptions of plea bargaining, at the expense of genuine transparency surrounding the process. Of course plea agreements offer discounts and incentives to defendants, if they did not, they would not be entered into. That is the very purpose of any plea agreement - leniency in exchange for a guilty plea. And whilst a defendant is not ‘required’ to assist the prosecution, doing so ensures far greater concessions and the likelihood of avoiding a custodial sentence. In order to examine the procedure more closely, this article has taken an exploratory approach in applying conceptualisations of plea bargaining to both informal plea bargains and plea agreements in serious fraud cases. Whilst alternative approaches can not readily be applied to informal plea bargains, a concessions / contractual approach is at minimum a valuable analytical tool with which to examine the nature of SFO plea agreements. At present, although we know the outcomes of cases which have employed the SFO measures, we do not know enough about the nature of the bargaining dynamic. Future empirical research would no doubt result in a deeper understanding of the plea bargaining process which is such a common, but under-researched, feature of the criminal justice system in England and Wales.


Alexander R ‘The Bribery Act 2010 in force: an opportunity to be taken’ (2011) 32 Company Lawyer 353

Alschuler A W ‘The defense attorney's role in plea bargaining’ (1975) 84 Yale Law Journal 1179

Ashworth A. and Redmayne, M. The Criminal Process Oxford: (OUP 2005)

Attorney General’s Office (2008) Press Release: ‘Criminal Justice Measures to Enhance Fraud Prosecutions to be Introduced’ 18th March 2009.

Attorney General’s Office Guidelines on Plea Discussions in Cases of Serious or Complex Fraud (London: Attorney General’s Office 2009)

Baldwin J and McConville M Negotiated Justice: Pressures to Plead Guilty (London: Martin Robertson 1977)

̶̶    ̶ ‘Plea Bargaining and Plea Negotiation in England’ (1979) 13 Law & Society Review 287

Baker N ‘Bargains and the backlash’ (June 2010) International Bar News 24

Belloni F and Hodgson J Criminal Injustice: an evaluation of the criminal justice process in Britain (Basingstoke: Macmillan 2000)

Blumberg A ‘The Practice of Law as Confidence Game: Organizational Cooptation of a Profession’ (1967) 1 Law & Society Review 15

Church T ‘In Defense of Bargain Justice’ (1978) 13 Law and Society Review 509

Council of HM Circuit Judges The Introduction of a Plea Negotiation Framework for Fraud Cases in England and Wales: Observations of the Criminal Sub Committee of the Council of HM Circuit Judges ( London 2008)

Cownie F, Bradney A, and Burton M English Legal System in Context Oxford: (OUP 2007)

Darbyshire P ‘The Mischief of Plea Bargaining and Sentencing’ (2000) Criminal Law Review 895

̶̶  ̶    ̶  Darbyshire on the English Legal System (London: Sweet and Maxwell 2005)

DLA Piper UK LLP  ‘Fraud and corporate crime (February 2012)) In-House Lawyer 197

Easterbrook F ‘Plea bargaining as compromise’ (1991) 101 Yale Law Journal 1969

Feeley M ‘Two Models of the Criminal Justice System: An Organizational Perspective’ (1973) 7 Law & Society Review 407

Heumann M (1974) ‘A Note on Plea Bargaining and Case Pressure’ Law and Society Rev. 9: 515

Hockman S and Medcroft N  ‘Plea negotiation – dead or alive?’ (2011) 1 Criminal Bar Quarterly 4

Hough M and Roberts J Attitudes to Punishment: Findings from the British Crime Survey (Home Office Research Study 179, London: Home Office 1998)

Jacob H Justice in America: Courts, Lawyers, and the Judicial Process (Boston: Little, Brown and Company 1984)

Jung H  ‘Plea Bargaining and Its Repercussions on the Theory of Criminal Procedure.’ (1997) 5 European Journal of Crime, Criminal Law and Criminal Justice 112

Katz J ‘Legality and Equality: Plea Bargaining in the Prosecution of White-Collar and Common Crimes (1979) 13 Law & Society Review 431

McConville M, Hodgson J, Bridges L and Pavlovic A Standing Accused: The Organisation and Practices of Criminal Defence Lawyers in Britain (Oxford: OUP 1994)

McConville M ‘Plea Bargaining: Ethics and Politics’ (1998) 25 Journal of Law and Society 562

McConville M and Mirsky C ‘Guilty Plea Courts: A Social Disciplinary Model of Criminal Justice.’ (1995) 42 Social Problems 216

McDonald W ‘From Plea Negotiation to Coercive Justice: Notes on the Respecification of a Concept’ (1979) 13 Law and Society Review 385

McLauchlan R ‘Plea bargaining differences put directors in bind’ (2010) Lloyd’s List Insurance Day 7

Nardulli P, Flemming R and Eisenstein J ‘Criminal Courts and Bureaucratic Justice: Concessions and Consensus in the Guilty Plea Process’ (1985) 76 The Journal of Criminal Law and Criminology 1103

Rauxloh R Plea Bargaining in National and International Law (Oxford: Routledge 2012)

Rosett A and Cressy DR Justice by Consent: Plea Bargains in the American Courthouse, (New York: J.B. Lippincott Company 1976)

Sanders A, Young R and Burton M Criminal Justice (Oxford: OUP 2010)

Schulhofer SJ ‘Plea Bargaining as Disaster’ (1992) 101 Yale Law Journal 1979

Scott RE and Stuntz WJ ‘Plea Bargaining as Contract’ (1992) 101 Yale Law Journal 1909

̶    ̶   ‘A reply: imperfect bargains, imperfect trials and innocent defendants’ (1992) 101 Yale Law Journal  2011

Serious Fraud Office, Annual Report and Accounts 2010-11, London: TSO

Sommerlad H and Wall D ‘Legally aided clients and their solicitors: qualitative perspectives on quality and legal aid’ (Law Society Research Paper No. 34, London: Law Society 1999)

Stephen F and Garoupa N ‘Why plea bargaining fails to achieve results in so many criminal justice systems: a new framework for assessment’ (2008) 15 Maastricht Journal of European and Comparative Law 319

Vamos N ‘Please don’t call it “plea bargaining”’ (2009) 9 Criminal Law Review 617

Watson D ‘The Attorney General’s guidelines on plea bargaining in serious fraud: obtaining guilty plea fairly?’ (2010) 74 Journal of Criminal Law 77

Zander M and Henderson P  Crown Court Study (London: HMSO. 1993)

(1) See for example Alexander 2011; Baker 2010; DLA Piper UK LLP 2012; Hockman and Medcroft 2011; McLauchlan 2010. The ‘Attorney General’s Guidelines on Plea Discussions in Cases of Serious or Complex Fraud’ (2009) is available at:

(2) Available at:

(3) See for example the SFO Operational Handbook, accessed at:

(4) The SFO deems fraud to be serious or complex if at least two of the following factors are present: the amount concerned is in excess of £500,000; there is a significant international dimension; the case requires specialised knowledge of financial or related matters; the case involves numerous alleged victims; the case involves alleged significant fraud on a public body; the case is likely to be of widespread concern; the alleged misconduct endangered the economic well-being of the UK.

(5) Section 9.1 – 9.6. Available at:

(6) See for example Baldwin and McConville 1977; Zander and Henderson 1993; McConville et al 1994; McConville 1998; Sommerlad and Wall 1999; Cownie, Bradney and Burton 2007 for further discussion.

(7) It should be noted that the Bribery Act 2010, in force since July 2011, provides for a maximum custodial sentence of ten years, and promises to tackle international commercial bribery, but if even the prosecuting authorities express a preference for suspended sentences, there seems little likelihood of sentences of that level in all but the most exceptional of cases.

(8) See Baldwin and McConville 1977; Zander and Henderson 1993; McConville et al 1994; McConville 1998 for further discussion.

(9) R v BAE Systems Unreported, December 21 2010, Sentencing Remarks of Mr Justice Bean, available at:

(10) The literature on contractual models deals largely with applying classical contract theory to plea bargains, and does not expressly consider the relationship between contract theory and concessions more generally (see for example Schulhofer 1992; Scott and Stuntz 1992a, 1992b).

(11) There is a considerable body of literature on lawyers’ working practices which also emphasises the role and significance of shared understandings in creating routinized working practices in both England and Wales and the United States which supports this view, for example Blumberg 1967; Feeley 1973; Alschuler 1975; Jacob 1984; McConville et al. 1994; McConville and Mirsky 1995. However, the emphasis is on routinized working practices rather than the consensual nature of any agreements reached.

(12) See for example McConville et al’s study, which found that lawyers held an institutional assumption of guilt of the clients with whom they dealt (1994, pp. 189 – 193).

(13) See for example Alschuler 1975; McConville et al 1994; Stephen and Garoupa 2008