Richard Burgon,  Shadow Justice Minister, had a field day with his debate on prisons and probation on 14 May, here.  His advocacy of a return to monopoly services had both wind and tide with it, what with the disaster of botched probation privatisation plain for all to see, and a timely bit of analysis showing the levels of violence are higher in private than in public prisons.

But what really made his day was that David Gauke had nothing to say in defence of the private sector, other than bland platitudes about innovation and some favourable inspection reports. Granted it was difficult to use the argument that, thanks to Grayling's cuts, the public sector is as fucked up as the private sector (the public sector 'flagship' of Berwyn springs to mind).  Difficult, too, to run the argument that the private sector is ony more violent because (even) more overcrowded.

But it is astonishing that he made no use of the Cambridge  research that has shown the private sector stronger on 'respect' though weaker on 'control' (shorthand for a much more nuanced report), or improving dramatically at Birmingham in the first years after G4S took over; and that he failed also to point out what a scandal the public sector  was when last a monopoly. It is also beyond belief that the civil servants appeared not to have noticed previously that private prisons have a higher level of violence, or to have any response to those figures, which they themselves had just put together.

As I pointed out in my book, one of the most astonishing, and unpardonable, things about privatisation has been Governments (Labour and Tory) complete lack of interest either in evaluating the results, or making the case publicly. The privatisers seem to think it unnecessary to make their case, even though polls showed as far back as 2011 that the public did not support privately run prisons. They now reap the reward.

There is a kind of institutional studidity in Government nowadays, on this and many other matters. The MoJ sees to have dispensed with its research unit altogether. That made Gauke's attempt to contrast the Governments' supposedly evidence based approach with Labour's ideological bent doubly painful - because on this occasion, Labour had the evidence, the Government had none. It is not Labour that will kill off outsourcing, one of the biggest policy experiments in UK history: it has already been killed off by the ignorance and incompetence of Tory Ministers.

Not that Labour emerge with any credit. Burgon, like Khan before him, has never shown any real interest in justice, beyond a weapon to beat Government with on cuts and privatisation. In particular, Labour have been way behind all other parties in committing to sentencing reform. Everything Labour has to say about the importance of rehabilitation has been said dozens of times before, by every single Minister right back to Michael Howard.  Labour, too, are bereft of any real ideas (1), other than restoring the monopoly power of their union friends, the Prison Officers' Association .  And there is very good evidence as to where that leads.

(1) Has there been any significant new idea in the past quarter century - since Labour introduced 'what works' evidence-based interventions in 1997 (but sadly not evidence-based sentencing, on which Government continues down the decades to do what it knows does not work).

Grayling with mouth wide open, foot sure to follow.

Gauke's announcement today of renationalisation of the probation service was the right decision and will be welcomed by everyone (except 'Reform', the PR people for outsourcers, I notice). Indeed, it was inevitable, the Balkanised structure Grayling set up had to be replaced by a unifed service, and that could only be publicy run.

Many questions remain about the new structure, not least cost.

Much has been made of the figure put out by the National Audit Office of £500m, but that was the cost in excess of what the Grayling reforms should have cost Similarly the Public Accounts Committee). But the Grayling structure could never have worked, that is now clear.

The important figure is the cost of the new  nationalised service compared to what it would have cost,  had Grayling been strangled at birth, and the old public system continued. That, together with set up and transitional costs (which may run into many tens of millions),  would give us the true 'cost of Grayling'. We won't know that figure for some time, not least because MoJ is now consulting on the new structure. The calculation is complex, for example Grayling extended probation supervision to short term prisoners, so extra workload, but then total court ordered supervision of offenders has fallen - and Gauke wants to do away with short terms sentences altogether.

What is certain is that the new arrangements will cost a shedload of money more than MoJ has available, since their spending plans were based on Grayling's reforms working, and part of that was to get cost out of the old public sector structure. The discussions with HMT must have been interesting, as MoJ was struggling with a £1bn overspend even before this volte face.

If one adds to this the cost of Grayling's bizarre contracts for ferry services with companies who don't in point of fact run ferries, which were then cancelled at huge expense, the successful claim by Eurotunnel, the forthcoming claim by rival ferry companies, the legal actions by Arriva and Stagecoach relating to Grayling's handling of rail franchise contratcs,  to name only the ones we know about, Grayling seems to be one of the most expensive idiots in politics (a closely fought field). It looks like the all up costs of Grayling to the British state could be as much as £500,000 for each  day in office.

(I refer only to financial costs. There are then the human costs - staff made redundant, staff public and private sector alike over-worked, offenders not properly supervised, left homeless, self harming and being assaulted in prisons as a result of those cuts....)

What this saga reveals is that the system of accountability on which Parliament relies isn't fit for purpose. Because if you keep moving quickly enough from one Department to another, it's your hapless successors who have to clear up the mess you dumped on them, and answer for it.

The Public Accounts Committee should innovate - they are always demanding that the public setcor innovate, now's their chance to do so. Hold an inquiry into the consequences for public spending of Christopher Stephen Grayling.

 In 2013 the Government announced that it had just discovered that G4S and SERCO has been billing excessively for their contracts for electronic monitoring of offenders.  The companies were made to pay back some £200m.  It has been established that amongst other things, the companies were billing for work they had not done.

The Government referrred them to the Serious Fraud Office. 6 years later, the SFO are still mulling over the case.  Granted, it is one of the most complex in the entire history of the SFO, featuring as it does no fewer than one victim, no fewer than 2 companies, no fewer than 2 contracts and no fewer than one jurisdiction.   Obviously, that must takes many years to unpick!

Meanwhile, here is an interesting graph. It is interesting because as a general rule, if volumes of work done under contract rise, in this case getting towards doubling, the unit cost - the cost per item - can be expected to fall. This is because fixed costs are spread over more items. This is a basic  rule of intelligent contracting.

But in this case, the unit cost actually increased as volumes increased. That would be bizarre whether or not the companies billed for work not done.  And suggests that there was something very wrong indeed with the contracts themselves.

Source: National Audit Office, 'The Ministry of Justice's electronic monitoring contracts', 2013