Reflections on TR

In the context of the on-going Probation Review, lets continue with our theme of looking back at how and why it all went wrong. Here's an absolutely fascinating interview given by one of the top officials on the CivilServiceWorld website in April 2014. She didn't stick around for long of course:-

  
Interview: Antonia Romeo, MOJ

The Ministry of Justice’s track record on outsourcing is a shaky one – but in its probation reforms, it’s running one of Whitehall’s most ambitious greenfield contracting-out schemes. Matt Ross meets project chief Antonia Romeo

The last time Civil Service World interviewed a senior Ministry of Justice figure, in March 2013, it wasn’t a comforting experience for either side. The MoJ’s programme to contract out the management of court interpreters had not gone well, with many criminal cases delayed by shortages of skilled linguists. But when presented with the concerns raised by two select committees and the National Audit Office, permanent secretary Ursula Brennan kept giving the same answer: the MoJ had needed to move fast to avoid “further confusion and the belief that we weren’t determined,” she argued. The desire to appear confident about the scheme, it appeared, had trumped basic elements of good practice such as properly assessing the task at hand, acting on the warnings of the MoJ’s own due diligence assessors, or piloting the planned reforms.

Given this history, when CSW returned to the MoJ’s Petty France HQ to talk about the probation outsourcing programme with Antonia Romeo, director-general of criminal justice, it was with both a long list of questions, and a certain amount of scepticism. After all, since last year the MoJ’s mismanagement of the electronic tagging contracts run by Serco and G4S has raised yet more questions over its ability to manage private suppliers. And the probation scheme is far bigger and riskier than either the interpreters or the tagging projects.

The probation scheme does, though, have the advantage that it’s been conceived to introduce new approaches to rehabilitation and increase productivity – allowing probation services to be extended to a new group of offenders – rather than simply to cut costs. Those sentenced to prison terms of less than one year currently receive no support when they’re ejected back into the community: “The point of this programme is to bring into scope 50,000 offenders,” explains Romeo. “This is an affordable way of extending provision to the under-12-month group, where we think we’re likely to have the most impact in terms of reducing reoffending.”

Compressing the quart

The MoJ’s plans involve squeezing these additional 50,000 offenders into the budget that currently funds around 150,000 low- and medium-risk offenders overseen by the probation trusts. But this isn’t about telling service providers to “stick with your old processes and take on an additional 50,000 people,” says Romeo: by changing the roles and expectations of those working with offenders, she believes, the scheme will cut the costs of reducing reoffending. “This work is going to be done in a completely new way,” she says.

Most offenders, she points out, have “a very complex and intractable set of problems that need to be looked at holistically.” Reducing reoffending involves helping people tackle challenges around housing, drug use, skills, literacy, mental health and a host of other issues; it means coordinating public and voluntary sector bodies’ work; and it demands new services, such as one-to-one mentoring, and ‘through the gate’ programmes that prepare offenders to leave prison and support them once they’re free. The MoJ’s plans are built around the idea that ‘community rehabilitation companies’ (CRCs) – contractors that bring together the finance and systems of large firms, and the skills and local knowledge of local charities – will monitor offenders more cheaply, and reduce reoffending more effectively, than our existing probation service.

Probation is currently delivered by 35 probation trusts – each with its own board – whose boundaries align with those of the police forces; however, there will only be 21 CRC areas, with each contractor overseen directly by the MoJ’s National Offender Management Service (NOMS). Given that the programme is ostensibly designed both to mobilise local charities, and to foster cooperation between local public service providers, the probation reforms look remarkably like a classic top-down reorganisation. But Romeo emphasises that the system will vary from place to place: “We’ve got to design a system that works locally, and we spend a lot of time engaging with local partners,” she says. “We’ve got competition teams whose job it is to make sure that we really understand, in running these competitions, what the local issues are. Police and crime commissioners and local authorities are very involved.”

With the contractors looking upwards to NOMS, why should they engage with other local organisations to build those holistic services? “Because they’ll have some of their money at risk against reductions in reoffending,” she replies. “What works is working across the piece with local partners – so they’ll be highly incentivised to do that.” At the “stakeholder events” being run in each area, she adds, “we’re finding that potential bidders really want to understand the local landscape and how those local partnerships work. And we’re asking bidders to say how they’ll sustain and build on local partnerships. We’re not just hoping it’s going to happen; we’re expecting them to tell us in their bids.”

On cash and contracts

In December, the MoJ announced that 30 bidders had passed the ‘pre-qualification questionnaire’ stage – a mix of private companies, consortiums, and probation trusts working towards ‘mutual’ status. At this point in the process, Romeo won’t say much about how the probation budget will be divided – either between the CRCs and the rump National Probation Service (NPS), which will manage high-risk offenders and write court reports; or between the ‘fees for service’ paid to CRCs for delivering their core work, and the funds awarded according to CRCs’ success in reducing reoffending. She does note, though, that “we’re not seeking to take a huge amount out of the overall probation budget over the next few years” – the challenge is to improve productivity to fund a rising caseload, rather than to maintain quality as income declines – and that “the majority of the contract will be on a fee-for-service basis.”

In fact, it sounds as if Romeo hasn’t yet made the final decisions on the risk within the payment-by-results mechanisms: it’s “being set through the competition”, she says, with the MoJ “asking people to tell us how much they’re prepared to put at risk, and what reduction in reoffending” they think they can achieve. The important thing, she says, is that the competition isn’t designed to deliver a set of defined services for the smallest possible sum, but to identify those contractors who’ll provide the best possible service within the set budget. “This is not a price competition; this is a quality competition,” she says. “It’s about getting incremental reductions in reoffending over the long term, to help us live within our means in future years.”

The NPS, meanwhile, will continue to oversee high-risk offenders. As the new system is established, these will be identified using what Romeo calls an “actuarial tool”: a piece of software that “looks at all the indicators and does something rather complicated and determines what level of risk someone is.” When your correspondent expresses scepticism over a machine’s ability to predict a criminal’s chances of reoffending, Romeo emphasises that the final decision will be made by NPS professionals – as will decisions over whether to recategorise offenders once the system is up and running.

Because those deemed low- and medium-risk are handled by the CRCs whilst high-risk offenders are overseen by the NPS, offenders may find their supervision being transferred between CRCs and the NPS – and there are obvious dangers around handing over offenders just as the risk of them reoffending is increasing. However, Romeo denies that such offenders will suddenly be presented with a new set of staff and procedures: “It doesn’t necessarily mean a completely new set of people; it will all depend on what’s best for that particular case,” she says. “They will now be the responsibility of the NPS, but in terms of who’s doing offender management and what are the interventions – that won’t necessarily completely change.”

The relationship between the CRCs and NPS will clearly be crucial here, as will the exchange of data between them. CRCs will, says Romeo, have to make their IT systems compatible with the NPS’s network, enabling them to share information about offenders. However, contractors “will need some support during the first part of the contract; we’re not expecting them to suddenly come in with a whole raft of new IT on day one,” she says. “There will be a period of being supported and using our systems; then a period of them choosing what systems they want to run.” The CRCs and NPS will also be expected to share offices, facilitating joint working and the retention of a strong local branch network. “They will be separate organisations, but there won’t be separate approaches,” Romeo claims.

Getting it right this time

The complexities of the new system are bound to make its introduction challenging – and it’s far from proven that the ministry is capable of successfully managing such a big outsourcing project. After all, four days after the CSW interview in which Ursula Brennan rejected the challenges put to her concerning the courts interpreters project, the Treasury quietly published a response to the Public Accounts Committee report in which it accepted every single recommendation: at that time at least, there were clearly big holes in the ministry’s outsourcing capabilities. So let’s run through some of the things the MoJ got wrong that time.

First off, PAC found that the MoJ “did not have a clear understanding of its requirements under the new system”, and ended up being “driven by bidders’ proposals rather than its actual requirements.” What’s more, it didn’t pilot its scheme before implementing it nationwide. What evidence is there that the new probation system will meet the MoJ’s aim of reducing reoffending? Romeo acknowledges that the ministry hasn’t trialled its final proposals anywhere. “You have to turn on the statute once nationally,” she says. “You can’t provide rehabilitation services to under-12-month cohorts in some areas and not in others – not least because people go in and out of prison, and end up in different areas, so whether they were covered by the statute or not and whether that service provision existed would become impossible to manage.”

Furthermore, she notes, there’s “a timing issue, because the government’s policy is to roll it out by 2015, so we can really start feeling the effects in reductions to reoffending.” There is clearly a political timetable behind the pace at which the MoJ is moving – but Romeo argues that the ministry’s payment-by-results pilot in Peterborough prison has “a lot of similarities” with the probation reforms, providing confidence about the new system. The Peterborough scheme offers the under-12-month cohort intensive support from local charities, with investors rewarded according to their success in driving down reoffending.

Here, says Romeo, the interim results show a 10% reduction in reoffending, compared to an increase of 10% nationally. “The providers have chosen to commit quite a lot of resource to ‘through the gate’ services,” she says. The MoJ’s belief is that, by setting out expectations in contracts and using payment-by-results incentives, it can encourage CRCs handed a much broader caseload of offenders to maintain that focus on rehabilitation.

Taking the con out of contract

Asked about contract management, Romeo emphasises the role of the Crown Commercial Service (CCS) in approving and monitoring its new contracts. In the courts interpreters scheme, PAC said, the MoJ’s own credit rating report recommended that its chosen provider “should not be awarded a contract valued at more than £1m” – but the ministry ignored it. This time, Romeo says, things will be different: “The Cabinet Office has a very clear process for awarding contracts.” The CCS has “brought together all the things departments need to be doing, and obviously we are following those guidelines”: the ministry is currently “building a best practice contract management function, and that will have the deep domain business expertise that is required as well as the commercial expertise and the corporate services support.”

The ministry’s proposed payment-by-results model has also come in for criticism, with the Social Market Foundation arguing that the 3% margin for error initially set out by the ministry would incentivise contractors to let reoffending rates creep up. “One thing is clear: we will not be operating a payment mechanism that incentivises people to do nothing,” Romeo responds. “We published our payment mechanism so that people could give us feedback, and we have looked at what people said in developing the final version.” For example, the MoJ moved from a “binary mechanism” – which required offenders to completely stop breaking the law – to a “hybrid mechanism”, rewarding both complete cessation of offending and a reduction in the number of crimes.

The contracts are expected to run for 7-10 years, but they’ll include a set of penalty clauses and the right for the MoJ to “step in” and take over a failing service. “We’ll have, both within the contract management function and also within the NPS, expertise in probation services,” she points out; the ministry will be equipped to take over if required.

Moving safely in a risky world

As the ministry introduces the new system, Romeo argues, it’s taking every precaution to ensure it’s robust. “To get large transformation programmes working, you’ve got to have really good assurance in place so that you know you’re not believing your own hype,” she says. “We have external, independent assurers telling us if we’re doing the right thing and, before we proceed with any part of the programme, whether it’s sensible and appropriate to do so.” NOMS has a business assurance board designed, she adds, to “give me, the senior responsible officer, the assurance that this is going to work and isn’t taking on any unnecessary risk.”

Equally crucially, the ministry has “a very thorough programme of communication with the trusts themselves; and transition managers who spend all their time talking to trusts about what’s going on. I know that one of the risks in a major programme can be the people at the centre of the programme not understanding how it’s bedding down, and I’m determined to make sure that doesn’t happen. So I personally listen very carefully to what people tell me; and I go out all the time and talk to trusts, and to local authorities, and to police and crime commissioners.”

Romeo is sure that the new system will improve results. “At the moment, we have a very serious and professional group of probation staff working incredibly hard with local partners to reduce reoffending,” she says. “We’re seeking to allow those that move out to the CRCs the freedom and innovation to bring in new and better ways of doing things.” And in that process, she adds, “my job is to make sure that we don’t take any unnecessary risks as we move; that we look very carefully and seek assurance that what we’re doing won’t lead to any reduction in ‘business as usual’ or to any risks.”

Over an intense 50 minutes, Antonia Romeo has answered every question about this wholesale transformation of our probation services – but the risks involved in the reforms remain substantial. Convicted criminals are an unpredictable set of people, and the stakes are high: if contractors take their eyes of the ball and their clientele break the law, people will get hurt. Given the MoJ’s past mistakes in outsourcing schemes, the programme’s success is far from guaranteed.

Yet Romeo is clearly aware of the dangers, and determined both to manage the process as carefully as possible, and to watch carefully for emerging problems. “We need to progress this in a really disciplined and controlled way. We don’t take any risks in moving from one phase to the next,” she says. “My job as senior responsible officer is to make sure we deliver the benefits of the programme. We need to really understand what’s going on – and there are no prizes for not listening.”

Written by Matt Ross on 24 April 2014