Deferred prosecution agreements: the way forward?

The UK government is grappling with the most appropriate way to deal with corporate criminal liability in the context of serious fraud and corruption. Civil recovery orders can sometimes appear too lenient, but a full-blown criminal conviction (whether on the back of a negotiated plea agreement or otherwise) can lead to the complete demise of what might otherwise have been a strong, successful business. Could deferred prosecution agreements (DPAs) be the answer? Richard Alderman, the director of the Serious Fraud Office (SFO) certainly seems convinced. 


Civil recovery orders

The option for the SFO to deal with corporate criminality by way of a civil recovery order (CRO) was introduced with some fanfare in 2009. This route was intended to allow for a swift resolution of the matter in appropriate cases, where the defendant had self-reported the conduct in question to the SFO and had satisfied the SFO that it had appropriate systems and controls in place to ensure that the unlawful conduct would not happen again.

However, the CRO route suffered a severe blow in Lord Justice Thomas’ judgment in R v Innospec Ltd [2010]. The pertinent points of Thomas LJ’s lengthy written judgment, which, unusually, was distributed to members of the public in the gallery on the day of sentencing, can be summarised as follows:

  • It will rarely be appropriate for criminal conduct by a company to be dealt with by means of a CRO.
  • Any fine must reflect the true criminality of the offence (for example, a company should not be charged with failing to maintain accurate books 
and records when it is in fact guilty 
of corruption).
  • Corruption is at the top end of serious corporate offending, both in terms of culpability and harm.
  • A fine may be limited to what a company can afford, but no specific agreement on the level of a fine should be reached between the parties. The prosecution and defence should instead assist the court by presenting an appropriate range in accordance with relevant sentencing guidelines.
  • In respect of a global settlement, there is no reason to differentiate the level of financial penalty to be paid in the UK and other countries (in this case, the settlement in the US had taken the lion’s share of the available funds).
  • Imposing an independent compliance monitor is an expensive form of probation that seems unnecessary for a company with new management – the funds allocated for this would generally be better used for fines, confiscation or compensation.
  • Corporate defendants should not be treated differently from any other defendants.

Thomas LJ was critical of the approach taken by the SFO in Innospec, but felt that his hands were tied to accept the settlement reached between the parties by the early publication of the terms of the settlement by the SFO before the case had even reached court.

There are several benefits to corporate defendants being dealt with by way of 
CRO, not least the fact that perpetual debarment from tendering for public contracts under the EU public procurement directive and exclusion from Export Credit Agency and World Bank/multilateral development bank funding will not be triggered. Given the judiciary’s view of CROs however, they may be less readily available in the post-Innospec era.

Agreed pleas

The second option for resolution of corporate criminal liability is a negotiated plea agreement in accordance with the Attorney General’s Guidelines on Plea Discussions. Such agreements are a relatively new development in the UK. The most well-known example to date is probably the BAE Systems settlement in relation to BAE’s failure to keep accurate accounting records in relation to its 
defence contracts in Tanzania. The 
judiciary are struggling to reconcile some of the wider issues, including whether guilty pleas following a plea discussion, presentation of agreed documents and submissions on sentence represent a fair and proper outcome. The courts have made it very clear that they are not in favour of any agreement that restricts their sentencing powers.

An agreed plea gives the defendant a slightly greater degree of control over the process and credit for co-operation. However, certainty of sentence is not guaranteed, as it is open to the judge 
before whom the plea agreement is 
placed to reject it if they do not consider the submissions to be an appropriate sentence on the basis of the agreed facts. In R v Dougall [2010], the suspended sentence agreed between the prosecution and defence in light of Mr Dougall’s 
co-operation was rejected by the court at first instance and Mr Dougall was sentenced to 12 months’ imprisonment. This was overturned on appeal and a suspended sentence given, though the court was very clear that its decision had ‘nothing to do with any sentencing agreement between the prosecution and the defence’. There are currently no published sentencing guidelines available in respect of corporate criminality (excluding corporate manslaughter) or for corruption offences.

Most importantly perhaps, from a defendant’s point of view, a negotiated criminal plea agreement will not protect against the collateral consequences of conviction for a fraud or corruption-related offence, such as confiscation of the proceeds of the unlawful conduct (which will not be limited merely to the profits) and debarment from EU public procurement contracts. Where a company’s clients include governmental bodies, this debarment can be devastating.


At present the third option for resolution of corporate criminal liability is a full investigation and prosecution by the SFO. With the SFO’s budgets under constant pressure, this is often not an option for the regulator. Companies have no incentive to self-report suspected wrongdoing if they believe that they will be dealt with harshly by way of a criminal prosecution. A new board of directors or new shareholders who were not involved in any wrongdoing will not wish to put their positions at risk, knowing that they will be punished for the offences of their forebears. If companies do not self-report, there is a much lower chance, absent the intervention of a whistleblower for example, that the wrongdoing will ever come to light or, more importantly, that lawful and ethical business practices will be properly implemented in the future.


It is clear that the SFO cannot rely solely on its own resources to discover, investigate and prosecute cases of serious fraud and corruption. The judiciary seems adamant that CROs are not an appropriate means of resolution in such cases, and negotiated plea agreements lack the certainty of outcome that is an absolute necessity for businesses deciding whether to engage and co-operate with the SFO. Against that backdrop, plans are forming for the introduction of deferred prosecution agreements in the UK. DPAs have long been a weapon in the arsenal of the US Department of Justice (DoJ). High-profile examples have included KBR, Siemens, BAE and Johnson & Johnson.

Discussions at the Attorney General’s office in relation to DPAs are currently in their infancy and a consultation is expected to be issued next year. In broad terms, DPAs are likely to include deferred prosecution for two to three years in conjunction with an agreement not to engage in unlawful conduct again, to maintain strong policies and procedures to prevent a repetition of such unlawful conduct, to engage a monitor in appropriate cases and to pay a perhaps substantial fine.

There are clear benefits to this approach. Vast amounts of prosecutorial time and money will be saved by the co-operation of the defendant, while the defendant avoids the stigma of a criminal conviction and the very real damage that can be caused by the collateral consequences of conviction. A problem remains to be solved however, in the lack of certainty that arises in a system where it is for the prosecutor to decide what charges to lay and for the judiciary to decide the appropriate sentence. Current debate seems to favour early judicial involvement in the process, lending a greater degree of certainty. This could take the form of a Goodyear indication (ie an advance indication of sentence by the judiciary) prior to the signature of the DPA.

Incorporating this route into the DPA process will not be without problems. The Goodyear indication would necessarily occur some way into the discussions between the parties. If the company is not happy with the indication given, and decides its best course is to pull up the drawbridge on the discussions, there is a chance that it will already have given the SFO a sufficient route map to proceed with a full criminal investigation and prosecution on its own (discussed further below). Perhaps a better method of incorporating a greater degree of certainty would be the publication of clear sentencing guidelines similar to those used in the US, where a points-based system allows companies to assess in advance the likely level of penalty that will be applied to it. This system could be criticised for allowing companies the means to base their decision on whether to report illegality on a simple commercial exercise, rather than a true desire to run an ethical business. The fact is, however, that the commercial outcome of the decision will be a core consideration for any business that finds wrongdoing in its ranks, and providing it with relative certainty of outcome will be a key driver in persuading it to report that wrongdoing.

A key question to be decided in the new DPA framework is what use can be put to information provided in the course of discussions if the discussions fail. Under the current plea discussion system, the court may reject the plea agreement, or a company may decline to plead guilty following a Goodyear indication or the refusal of the court to provide one. If the SFO decides to proceed with a prosecution of the company, where the plea agreement has been signed, the SFO can rely on it as confession evidence. Even where a plea agreement has not been signed, any defence would be likely to have been prejudiced by the forgoing negotiations and discussions with the SFO. Although none of the information provided by a company in the course of discussions could be deployed against it in a subsequent trial, the SFO would be entitled to use their knowledge of the existence of that information (a) to obtain information from another source; (b) to understand and identify the weakness of the company’s position; and thereby (c) to speed up the process of investigation and prosecution. Companies entering into the DPA process will want clarity on this issue before deciding whether to engage.

In a climate in which multi-jurisdictional investigations are ever increasing, the potential provision of information gleaned through the DPA process to third parties, particularly other law enforcement agencies, will also be a concern for companies when making the decision whether to self-report and engage with 
the SFO. Any self-report will no doubt contain sensitive information about the company and others. The company will 
wish to ensure that confidentiality is maintained over the report and any 
related exchanges of information with 
the SFO. No doubt it will want to know what the position will be if a Freedom of Information Act request is made, or a request for mutual legal assistance by 
an overseas law enforcement agency. 
This may be particularly pertinent for 
the directors and employees of the company, for whom no immunity is guaranteed by the DPA process. Consideration should be given to these issues in the design of the DPA process.

While there are several foreseeable wrinkles to be ironed out in the design of the DPA process, it is inevitable that DPAs will be introduced into the UK system in one form or another. This will be a positive move for British businesses in the wake of the Bribery Act 2010 and the likely resultant increase in regulatory action from the SFO. The Bribery Act makes it much easier for the SFO to pursue businesses now that it can pursue 
a charge of ‘failing to prevent’ bribery, instead of having to prove that a controlling mind of the company had engaged in bribery. DPAs will give companies a better chance of coming out of a bribery or serious fraud charge with their business intact and their reputation rather less scathed than it would be were it to be tarnished with a full criminal conviction.