HOW A PRISON GOES BAD

NOTE [added 1 September: I have received quite a few comments privately on this piece; I have summarised some  of general interest at the end]

I have been reading the report of the internal MoJ inquiry into the major riot at Birmingham prison on 16 December 2016, which had been withheld, but then released this week following a Freedom of Information Act application. It has been redacted, but not much.

The report, by a former Director of what was then the National Offender Management Service (NOMS), is a most illuminating document, detailed, balanced, thoughtful. It not only casts light on how Birmingham prison reached the dire state described by the Chief Inspector in his Urgent Notification last week, but vividly illustrates perennial truths about how to run – and how not to run prisons – and what a difficult and important a job that is. Anyone seriously interested in prisons ought to find time to read it.

It comprises first an analysis of what happened on the day of the riot, and second how the prison had deteriorated in the preceding 12 months, to the extent that staff had effectively already lost control well before the riot.

I am not primarily interested in the actual riot, but the detailed, minute by minute account of the loss of control is as gripping as any film – quite terrifying. It sets the scene by describing how staff shortages and high turnover over many months placed staff, inadequately supported by management, under intolerable pressures and lead to disrupted regimes which in turn fueled prisoner boredom and discontent, and how staff gradually ceded control of the wings to prisoners. Then, on the day, an act of defiance by a couple of difficult prisoners gradually escalated, needlessly, as several opportunities to contain things and regain control were lost due to muddle and lack of grip by middle and senior management. Prisoners reacted to indecision by turning to violence, and soon even those who'd have preferred to stay out of it felt obliged to join in.

But the report also records some acts of great courage by staff, in rescuing vulnerable prisoners from danger and holding vital control points despite ferocious assaults by groups of prisoners. One scene sticks in my mind: one wing had been lost, staff were running for safety, but one staff member saw a vulnerable prisoner who he knew would be severely beaten or worse, ran back at great risk to get him and they both then pelted down the wing for safety, closely pursued by a mob. 'The PCO described his good fortune at managing to unlock and secure the gate first time...' I bet he did, but not in those words! Once safe, the prisoner vomited with terror and relief and then would not leave the PCOs side. G4S failed abysmally - but some of their staff past the test alright. I do hope their courage and professionalism has been recognised.

The second part of the report analysis how control was eroded over the preceding year or so and it is hugely illuminating.

First of all it confirms what I have been saying, that the prison which improved markedly after transferring from public to private sector management in 2011. Two things confirm that. The first is research by Alison Liebling and others which showed improvements on 7 dimensions of prisoners' 'quality of life, as defined and measured by them: ‘respect/courtesy’; ‘humanity’; ‘decency’; ‘care for the vulnerable’; ‘staff-prisoner relationships’; ‘fairness’; and ‘personal autonomy’. Even more marked were improvements in the quality of staff life, including relationships with line and senior management – an extraordinary and important achievement, considering the feelings aroused by of the competition in 2011 and the highly contentious move of existing staff to the private sector, the first time this had ever happened and (as I know) long the source of much worry amongst policy makers (1).

This progression is confirmed by NOMS own Prison Rating System, a sophisticated analysis of many aspects of prison operation, over the same period. I reproduce a graph, showing this improvement according to the PRS after privatisation in 2011, up to January 2015, and the precipitous decline thereafter. (Note the steady decline in public sector prisons right across this period).

 Chart 1: ratings of Birmingham and other local prisons, NOMS Prison Rating System 

Thus, two lessons immediately: 1) privatisation was not the issue (in fact, privatisation had initially lead to much improvement, albeit from the low level in which previous public operation had left the prison); and 2) the decline began way back - two and half years ago.

The report doesn't help pinpoint quite why the prison started its rapid decline around January 2017 though there is some suggestion that the drug problem became much more serious around then, and it was just at that time that assault and self harm rates began to accelerate upwards across the whole system, so the cause or causes weren't peculiar to Birmingham.

Staffing levels, and management of staff, are central issues. Mapping the data from this report onto data I previously obtained under FoI, budgeted staffing levels (302 PCOs) have not changed significantly since 2013, and probably not since privatsation in 2011. Neither budgeted staffing levels not the number of posts vacant (about 3%) were out of line with other prisons – on my analysis, the staff:prisoner ratio was considerably better than Doncaster or Thameside, and about the same as Parc.

Chart2: Prisoners per PCOs at Birmingham and comparator contracted prisons

Source: PQs and FoI applications
Note: figures affected by reduced pop after 2016 riot

But the staffing situation was nevertheless very difficult indeed – dangerous, in fact - for several interconnected reasons:

1. High churn rates – resignations were running at 24% a year and recruitment wasn't quite keeping pace: thus a high proportion of staff were very new to the job (2)
2. An astonishing 60 PCOs were off work, on sick leave on training or away for other reasons, thus the prison was trying to operate with 1 PCO in 5 (on very tight budgeted staff levels) not available for work.
3. On top of that, poor work detailing meant that staff deployment was a shambolic affair, with staff were turning up to work to find they weren't detailed, and managers never knew how many staff they'd have available. This lead to regimes activities being needless interrupted and staff time being inefficiently used.

Thus the planned regime simply wasn't deliverable with available staff numbers. There weren't enough staff on landings, prison routines were constantly disrupted and many staff were new, insufficiently trained and scared. As we know from previous breakdowns in privately run jails (3), this quickly become a downward spiral - with new staff appalled at what they are faced with and choosing to quit or, if they stay, adopting passive and defensive behaviours rather than challenging misbehaving prisoners. Staff complained that they were poorly supported: 'new staff....reported feeling ignored and marginalised, and made to feel like a nuisance when they asked for help'.

Management was suprisingly weak in many other ways. Managers insisted staff maintain a 'full regime' despite low staff availability - but it wasn't even clear to staff (or the inspectors) exactly what 'the regime' was supposed to be! Policy on unlocking was also unclear and staff seemed to make it up as they went along.

The consequence of all this was a gradual retreat from authority, a process so familiar in prisons. Prisoners bullied and conditioned staff into unlocking everyone, whether entitled or not, and the IEP scheme, meant to relate privileges to behaviour, was hugely undermined. Unlocking everyone despite dangerously low staff availability became the norm, with staff unable to reconcile safe staffing minima with management's demand that they to keep prisoners unlocked (4). Staff over relied on prisoner volunteers to keep order, and some of these were not averse to using violence to do that job (and possibly some had other agendas of their own). The jail really had been turned over to prisoners to run! Extraordinarily, prisoners who wanted to be unlocked into this unsafe environment were asked to sign that they accepted the risks, an absolute abdication of responsibility by staff (and G4S). Staff seemed to loose control of such a basic thing as movement around the prison. Discipline was effectively abandoned - in one period only 1 in 6 serious assaults were referred to the police, of 655 cases so referred only 1 in 10 lead to a conviction. Violence went unpunished and drug taking went unchallenged. Vulnerable prisoners were not properly protected.

Prisoners' attitudes in such circumstances varies – some relish the lack of control, but others find it threatening and difficult, finding they were unable get the help prisoners need from staff to get things done:

“There as too much freedom,we like freedom but there weren't enough staff to deal with it. You weren't made to do anything, no discipline, no structure ...You couldn't find any staff...they let us get away with anything. Every day was a party. But boredom leads to badness”.

“Boredom leads to badness” - that should be inscribed above the gate of every jail.

The result was soaring rates of drug taking, self harm, and prisoner on prisoner and prisoner on staff violence.

Chart 3:  number of PCOs, assaults, self harm incidents at Birmingham, all per prisoner
Note: PCOs on right hand axis, other data on left axis

The management response was hopelessly inadequate. There are processes to ensure data such as assault and disciplinary rates are regularly reviewed in jails and that Governors/Directors regularly assess the stability of the jail. At Birmingham management assessed the jail in the weeks before the riot as 'low' i.e. 'there is no evidence of increased stability or instability above the norm'. That had been the rating throughout 2016, as every indicator turned glaringly red. The report suggests that management had been conditioned to accept very high levels of violence as the 'new normal'. In plain English, management had completely lost it.

And not only management. Each private prison has a number of MoJ officials – 'Controllers' – permanently working inside the jail to monitor operations. Here the report is more reticent, but it appears to confirms what a number of commentators have remarked on, a tendency of the Controllers to retreat into bureaucratic box ticking role. If they had walked the walk and talked to staff and prisoners, and taken a more common-sense over view of the state of the jail they could not have missed the seriousness of its rapid deterioration. But the role of Controller has always seemed unclear, and it's a difficult job to fill, because you need able experienced Governor grades, but such people would rather be off running their own jail.

What does the report, on a riot 20 months ago, tell us about the run up to the Chief Inspectors extraordinary intervention last week? Difficult, because until the full report on this most recent inspection is published, we don't know what happened in those 20 months. But I think we can already say:

1. Birmingham's problems were not caused by being privately managed.
2. After some years of improvement post privatisation, there was a sudden sharp downturn in early 2016.
3. We don't know why that occurred, but increased availability of drugs may have played a part.
4. Staffing was the main reason why the prison became so unsafe: while there had been no staff cuts, as in many public sector prisons, and absolute numbers of PCOs were comparable with other privately run prisons or better, a huge proportion were not available for work, high churn meant many staff were new and lacking confidence, while poor detailing meant those who were available weren't used to best effect.
5. Consequently the regime wasn't deliverable; and managers insistence on nevertheless high levels of unlocking meant the prison was poorly controlled.
6. The many signs of management weakness in fundamentals– in detailing staff, defining expectations in terms of regimes, managing unlock, dealing with assaults, assessing the stability of the prison, then lack of grip and decisiveness on the day of the riot, and poor adherence to procedures for managing disturbances, are quite damning.
7. Drug taking and violence escalated phenomenally in a short period and were not confronted or punished, self harm also escalated.
8. Staff withdrew from formal control in many ways, and let prisoners run things to a clearly unacceptable and dangerous degree. Control had been lost long before the riot.
9. It appears the Controllers assumed a somewhat bureaucratic role, whch may have stopped them from grasping the urgency of the situation.
   
   

What we still don't know;

1. Why the turning point in early 2017, after years of improvement? Was staff churn always so high, if not what caused it to rise to this level? Which came first – high churn, drugs, violence?
2. Did the Inspectorate misread the situation and overstated management grip in their February 2017 report, so soon after the riot?
3. What happened to the recovery effort praised in that report, in the 20 months since then? Did it begin to make progress and then fail, or did it never gain purchase?
4. Just where is the contract in all this? The case for contracting is that it sets out clearly what standards must be delivered by the operator and provided sanctions if it does not. Were there financial penalties on G4S during 2017 and 2018 (I've FoI'd this and will post the answer here)? If not, was the fault with the contract, or with contract management? (5)
5. Why did both G4S, and the MoJ, failed to understand the serious of the deterioration until the Chief Inspectors intervention last week? In the case of G4S, how is it that the company can make such an outstanding success of some prisons (6), and such a disaster here? Is it funding? Leadership? And  whatever the cause, faced with this obvious failure, over several years, why didn't G4S act? (The other prison operators have serious problems, but with them I don't get the impression that I do here, and at Medway, that management simply didn't register what was happening.)
6. In the case of MoJ, did they not understand that there was a quantum difference between the failures at Birmingham compared to other jails? Did the Controllers under-estimate the seriousness of the deterioration during 2017? How was the Controller's role defined? Was a factor abolition of area managers, who in earlier times would have been in and out of such a deeply troubled prison weekly, and not needed the Chief Inspector to tell them how bad things were? Does the withdrawal of procurement expertise to the centre of MoJ, away from the Prison and Probation Service, help explain why contract management was ineffective, or slow? Are civil servants so overwhelmed by the sheer number of failing prisons - more, on the latest count, than ever before since this system was first set up - that they can no longer respond
7. What had the Minister been told was the state of the prison before last week?

More general thoughts prompted by reading this extraordinary document:

1. MoJ must start publishing staffing data for all prisons, public or private: the budgeted levels of operational staff, number actually in post, churn rates. Sometimes it has said that such data for privately run prisons is commercially confidential – but then they do publish it, in response to PQs or FoI applications (so it can't be commercially confidential). Besides, all the operators know the others' staffing levels!
2. When I worked in the then Prison Department in the '70s and '80s – yes, I am incredibly old! - the great scandals were about staff beating or abusing prisoners – the beating to death of Barry Prosser in the healthcare unit (yes!) of, as it happens, Birmingham jail, in 1980; the beatings that sparked the Hull riot in 1976 and followed after the surrender; the systematic torture of prisoners in the Seg Unit at Scrubs over many years, til exposed in the late 90s; the racist abuse at Brixton and other jails exposed by the CRE around 2000. (Note for Guardian writers and readers: all these vile episodes were committed by the public sector. whose monopoly you are so desperate to restore.) But in recent years, the scandals are rather of prisoner violence towards other prisoners, and towards staff (Medway STC is one exception I can think of). It's the staff who are now running scared. Is that progress? In a sense, maybe: but a pretty perverted sense.
3. One thing that we have got right: a series of Chief Inspectors of Prisons who are informed, methodical, passionate, balanced and fearless.  and one cheer also for the FoI Act, that got this secret report unsecreted.
4. Reading this report, and the profound causes of the malaise at Birmingham, Rory Stewart's insistence that assault rates can be cut by 10% or 25% in 10 prisons in 12 months if chaps just get down to it with a will, looks very misjudged. Recovery isn't just a matter of a super-manager being parachuted in, or windows being re glazed – its a hard grind to recruit and retain the right calibre of staff, properly train and support them, and to restore confidence, respect, relationships; that will take years. In my book, I looked at 4 prisons, 2 public, 2 private, that reached near meltdown around the year 2000: in some cases, they were still in very poor state years later. I dont fancy Stewart's chances.
5. Finally, what on earth can be the benefit of sending more and more prisoners into such hell-holes? Isn't it clear that it can only lead to their becoming addicted to drugs and brutalised, whether as victims or perpetrators? Aren't we again at the position, enunciated so clearly by a previous Tory minister in 1992, that sending offenders to prison in the hope of reforming them is 'merely an expensive way of making bad people worse'? (7).

COMMENTS RECEIVED

1. the Deputy Governor at Birmingham left to go to Altcourse just before the riot - and before that, the Governing Governor had been providing support to Altcourse, which had been in difficulty. This may have had weakened management at Birmingham.
   
2. the mix of prisoners at Birmingham may have got more difficult due to local police decisions to focus on bringing members of violent gangs to justice.
   
3. a wider and interesting point: the emphasis on leadership by the Governing Governor is overdone. As important, if not more so, is the quality of middle management which (this commentator says) is not as good as it needs to be throughout the prison system. Many of the weaknesses at Birmingham described in this report, before the riot and during it, were at middle management level, not right at the top.
   
4. the riot also illustrated failure to train basic grade staff and their immediate supervisors well enough in responding to incidents – better trained staff might have acted at once to contain the situation before it escalated.
   
5. the dividing line, especially in local prisons, between just coping and losing control is very thin. There are other Birmingham that could happen any day.
   
6. is the private sector sometimes too keen to please the client (who after all is a monopoly purchaser – there's nowhere else to go for business)? Do private prisons too readily accept dangerous overcrowding and transfers of difficult prisoners? (Of course there may also be a purely financial interest in taking more.) [From what I hear, this was the (or a) cause of the G4S Olympics fiasco – a very late demand from Government for masses more security staff, which G4S couldn't realistically recruit in the time available, but still they agreed to do it.] As described in my book, much the same has been said of public sector Governors also – that they no longer feel able to 'push back' when the centre makes unreasonable or unsafe demands.
   
7. Finally, an addendum from myself: in light of this article one cannot simply blame 'cuts' for the state of our prisons. There weren't any cuts at Birmingham – and it had functioned fairly well with the same staffing levels for some years after moving to the private sector. Rather, as several Governors have put it to me, the tighter staffing levels imposed by MoJ on public prisons, and which have always been the case at most private prisons, are OK as nothing out of the ordinary happens. When it does, e.g. the arrival of the NPS drugs, or local recruitment problems, there isn't a margin for coping with it.
   
8. NB as I have previously documented, Altcourse is an interesting example of cuts in the private sector; always extremely well funded, its excellent performance seemed to dip after MoJ demanded money back, in the early part of this decade - contracts are little protection against aggressive demands by a monopoly purchaser! -  which had the effect of driving the staff: prisoner ratio up.
   

NOTES

(1) In my time it was thought too risky to attempt without much reducing prisoner numbers during transition, something we never had space to do.

(2) Rates this high were common in privately run prisons in the early 2000s and as recounted in my book, were key in bringing several to crisis point. Public sector rates were then very low – possibly too low. I had been told that changes in the labour market after the 2008 Crash had brought rates right down, to 10% or so: but clearly, not everywhere. The Crash feed through into austerity and big cuts in starting pay in the public sector, raising public sector churn rates and contributing to problems they have been experiencing. One conclusion is that Government should publish staffing levels in all prisons, public or private, and also vacancy and churn rates

(3) Particularly, at Ashfield in the early 2000s, described in my book, in the section 'Four prisons in trouble' – the only other time that Government has intervened to take over a failing privately run prison.

(4) Management's insistence on keeping prisoners unlocked, even though for many, there was nothing for them to do is odd, because it appears there was nothing in the contract to require this.

(5) Curiously, there were significant financial penalties in respect of this prison in 2011-12 (£95k) and 2012-13 (£205k) – the very years when the prison was doing OK! See Commons WA 5.10.16

(6) “Overall, Altcourse was in some key areas bucking the trend when compared to other local prisons. While it still faced significant challenges around safety, the downward trend in violence and anti-social behaviour was highly creditable. This was in no small part due to the energetic and proactive approach taken by the prison. Levels of self-harm, while still high, were also decreasing and there had been a real focus on ensuring men with these vulnerabilities were identified and cared for. This had been supported by a positive staff culture, a good focus on decency, and an excellent regime that was being delivered consistently….. the director and his team were providing strong leadership, enabling a highly positive staff culture and delivering good outcomes in many key areas. Overall, Altcours e showed that a local prison can provide fundamentally decent treatment and conditions for prisoners, despite facing many of the same challenges as the rest of the prison service. There was much here from which others could learn.“

Stick that in your pipe and smoke it, Guardian!

(7) 'Crime, justice and protecting the public'. Home Office, 1990