Purnell’s Lie Detector – Following the Money

This is an expanded version of an article published at Liberal Conspiracy.

The next stage of my detailed coverage of the scientific evidence behind the DWP’s current ‘lie detector’ trial will be along, as promised, in the next 24 hours but, in the meantime, I’ve received a reply to an FOIA request I put in around four weeks ago, asking for information about the financial side of the trial and, using that information we’re now going to follow the money and tell you exactly how much taxpayers’ money has been spent on the trial, to date, and exactly where this technology is being used to vet benefits claimants.

And after that, we’ll tell you what we’ve uncovered about the contracts and business deals behind this trial and how, if the DWP did decide to roll out this system, nationally, one company could, within the next five years, build an effective monopoly over the processing of welfare benefit claims in the UK worth tens of millions, if not hundreds of millions, of pounds a year.

THE DWP TRIAL – WHEN, WHERE AND HOW MUCH HAS IT COST

We’ve already mentioned the headline figures for the costs of the trial, £700,000 in its first year and a further £1.5 million for the second (current) year of the trial – now we’re in a position to give a detailed breakdown of the figures showing where the trials have been, and are, taking place and how much public money is being spent in each location.

YEAR ONE (2007/8)

In the first year, the technology was trialled directly by the DWP, through a joint-site Jobcentre Plus pilot in Lincoln and Nottingham and by a total 12 local authorities, 5 individual councils and two ‘shared service’ pilots in which a number of neighbouring councils operate a shared system for processing benefits claims.

The table (below) shows where the trials took place and how much each individual trial cost the taxpayer:

TRIAL SITE/LOCATION (2007/8) COST OF TRIAL
DWP/Jobcentre Plus (Lincoln/Nottingham) £240,000
London Borough of Harrow £125,000
City of Edinburgh Council £71,000
Wealden District Council £10,000
DERWENTSIDE SHARED SERVICE (Derwentside, Chester Le Street, Durham, Sedgefield) £72,000
WARWICKSHIRE SHARED SERVICE (Coventry, North Warwickshire And Warwick) £56,000
London Borough of Lambeth £63,000
Birmingham City Council £63,000
TOTAL £700,000

Interestingly, Harrow Council’s trial, which has been put forward  in the press as the ‘flagship’ for the whole programme, appears to have cost £125,000, double the amount (£63,000) that the media have been reporting.

YEAR TWO (2008/9)

Responding to my FOIA request, the DWP have given the following information for the costs of the second year of the trial.

TRIAL SITE/LOCATION (2007/8) COST OF TRIAL
Aberdeen £107,000
Barking and Dagenham £40,000
Basildon £40,000
Bexley £40,000
Bristol £107,000
Bromsgrove £107,000
Bury £108,000
Coventry £91,000
Doncaster £107,000
Eastbourne £15,000
Flintshire £107,000
Glasgow £107,000
Lichfield £70,000
Northamptonshire Shared Service (Northampton, Corby, Kettering and Wellingborough) £107,000
Swindon £40,000
Vale of Glamorgan £25,000
Walsall 108,000
Windsor and Maidenhead, £40,000
Birmingham £50,000
Derwentside shared service (Chester-le-Street, Derwentside, Durham, Sedgefield) £50,000
Edinburgh £106,000
Harrow £50,000
Lambeth £50,000
Warwick £42,000
TOTAL £1,714,000

This data raises some very interesting points.

To begin with, the total cost of year two of the trial, just over £1.7 million, is £200,000 more than the figure given to parliament, last year, by the-then junior minister responsible for the trials, James Plaskitt.

It’s also clear that Jobcentre Plus is not taking part in the second year of the trial. This was confirmed by the BBC, last week, only the day before the DWP released this information to me (four days late) and according to the Beeb, the reason that Jobcentre Plus pulled out of the trial is the high rate of false positives (63%) that the system was generating.

CAPITA’S ROLE IN THE DWP TRIAL – CUE BONO?

As we’ve already disclosed, Capita, a major supplier of outsourced business and IT services to the public sector, is the company behind this trial, alongside its ‘strategic partner’, DigiLog UK Ltd, which supplies both Nemesysco’s LVA technology and training in the use of its own package of interview techniques and practices, which it calls ‘Narrative Integrity Analysis Techniques’. These interview techniques are what the DWP refer to, on its own website, as ‘behavioural analysis’ and ‘conversation management’.

This brings us to a critically important but, as yet, unreported part of this story, relating to the contractual arrangements between Capita, DigiLog UK and Nemesysco as revealed by our investigations.

Capita joined forces with DigiLog UK late in 2004 when it bought the entire shareholding of a insurance claims investigation company called Brownsword Ltd from its previous owners, ISIS Equity Partners, in a deal worth a reported £8.5 million. At the time of sale, David Brownsword, the founder of Brownsword Ltd, was also a director of DigiLog UK.

As part of this deal, Capita also acquired an exclusive 10 year licence to use DigiLog’s ‘Advanced Validation Solutions’ in the public sector – ‘Advanced Validation Solutions’ is the name DigiLog UK gives to its mix of Nemesysco’s technology and its own interview techniques and practices, the latter having been developed by its current CEO, Kerry Furber.

To thicken the plot a little further, Furber developed these interview techniques while working for Highway Insurance as its head of fraud management. Highway Insurance was, of course, the first British insurance company to pilot and introduce Nemesysco’s LVA technology for evaluating insurance claims, working in partnership with DigiLog UK, which supplied the technology.

Highway officially announced that it has rolled out its system in 2002 and, less than a year later (2003), Furber left Highway to join DigiLog UK, becoming the company’s Chief Executive.

DigiLog UK is the official distributor of Nemesysco’s LVA technology for the UK and Ireland and, in the same way that Capita has an exclusive licence to use this technology in the public sector, so DigiLog UK has the exclusive rights to distribute the technology in the UK.

What this means, in practice, is that if the DWP Trial were to be successful, leading to a decision to roll out this system across the entire welfare benefits system, Capita would be sitting pretty as the sole, monopoly, supplier of these system to the whole of the public sector until at least the end of 2014.

If the man from the DWP says ‘yes’, Capita (and its partner, DigiLog) has a five year window in which it can attempt to build a monopoly over the processing of welfare benefits claims.

WHAT’S A MONOPOLY WORTH

What would a monopoly over the processing of welfare benefits actually be worth?

Based on the most recent available figures, the estimated cost of administering the UK’s welfare benefits system is, currently, around £5.7 billion a year, £2.8 billion of which is taken up in staff costs. The remainder, which is actually just over £3 billion (the DWP earns around £120 million a year by providing admin services to third parties) covers everything from the costs of office space and overhead, to IT services, to outsourced contracts and the costs of reorganisations and other short-term costs incurred in the never-ending search for greater efficiency.

Almost 65% of the DWP administrative costs are taken up by the provision of working age benefits such as Jobseeker’s Allowance, Housing and Council Tax Benefit, Incapacity Benefit, etc and last year, the DWP  handed local authorities a total of £541 million in subsidies to help cover the costs of administering Housing Benefit and Council Tax Benefit.

That’s the value of the market over which Capita could gain an effective monopoly, if the DWP buy into this system.

But there’s more, because under the government’s ‘business transformation’ programme for local government, local authorities are increasingly moving from a position in which a range of different suppliers provide various outsourced IT and business services under contract to different departments of the local authority to one in which more and more councils are putting all the eggs in one basket and contracting out all their outsourced services to a single consortium supplier. A typical consortium would consist of only 2-3 different companies; a main contractor, which handles all the business and administrative services and their specialist partners, who provide the council with integrated telecommunications and IT services to support these outsources administrative services.

In that kind of market, a company which establishes a monopoly over a key business process; the administration of welfare benefits, will automatically have a major competitive edge over its rivals when tendering for any of these integrated outsourced services contract because that company, and only that company, has a licence to use Nemesysco’s LVA technology and there is no other system on the market that performs a similar/comparable function.

The £500+ plus million pounds a year that the DWP gives to councils to cover the administration of the Housing and Council Tax Benefits is only a fraction of the total amount of government business that Capita could attract, if it can leverage this technology into the creation of an effective monopoly over the processing of Housing and Council Tax benefits by local authorities.

But, that all depends entirely on the DWP buying into the idea that Nemesysco’s technology actually does the job that its developer (Nemesysco), and UK licensees (DigiLog UK and Capita) claim that it does.

THE STING IN THE TAIL

Although this system is being pitched to the DWP as a combined package consisting of Nemesysco’s technology and DigiLog UK’s ‘Narrative Integrity Analysis Techniques’, the reality is that this latter half of the package, which the DWP refers to on its website as ‘behavioural analysis and conversation management’ is actually of very limited commercial value, despite it being the only part of the overall package that has an solid scientific foundations.

Take away DigiLog’s ‘buzzword bingo’ sales pitch and all you’ve actually got is collection of interview techniques derived from psychological research into witness interviews and witness reliability, most of which is now 15-20 years old and well documented in papers published in several scientific journals. In short, the kind of thing that any of DigiLog/Capita’s main competitors could easily develop within 12 months just by hiring a couple of halfway-decent psychologists to hit the literature and knock out the relevant training packages.

It is only when you add Nemesysco’s LVA technology to the package, due to the exclusive licensing agreements between Nemesysco and DigiLog, and DigiLog and Capita, that you have a unique product that cannot be duplicated by any of Capita’s rivals for at least the next five years. Capita’s shot at a monopoly over, at least, benefits claims processing depends entirely on the DWP buying into technology which has been shown, by those independent scientists and researchers who’ve examined the system in detail, to be utterly lacking in scientific validity and to be no better at actually picking out high risk claimants than simply flipping a coin and picking out people at random.

If the ‘lie detector’ doesn’t work then neither will a play for monopoly over benefit claims processing.