Professor John Seddon: Single Universal Credit

To: Rt. Hon. Iain Duncan Smith MP and Lord Freud

31st January 2011

Dear Sirs,

I write this as an open letter as it marks the beginning of a campaign to halt the current programme of reorganisation associated with the Single Universal Credit and embark on a better course. I am confident that your current proposals will drive costs up and worsen service; I also have strong evidence that an alternative approach will deliver better service at much lower costs.

This campaign was born at a conference organised by Hbinfo Ltd, a private-sector supplier of training and support to housing benefits services, in London on 24th January. I made the case outlined above and the audience – 150 people from local authority benefits services all over the country – felt moved to act. I agreed to make a start with this letter to you.

Let us be clear. This campaign is not about the concept of the Single Universal Credit (SUC) as such. It is about the design and implementation of its delivery. In this respect we should begin by underlining that there is much in the coalition’s approach to public-sector reform that we support. Getting rid of targets and central directives, abolishing the Audit Commission and encouraging the development of locally relevant services are all positive moves. But the weakness in the coalition’s approach is the continuing unquestioning faith in economies of scale. The organisational proposals for the SUC fall straight into this dangerous trap.

It would appear to be a simple proposition: take costs out of service provision by putting the provision online and/or in a call centre. That these ‘channels’ represent cheaper transactions is not in doubt. But it does not follow that overall the service will be cheaper to deliver. The crucial factor is the complexity of the service. When what is being delivered is simple and unvarying, moving it to telephone or internet channels may be effective. When it is complex and variable, however, it is an expensive mistake, driving costs up and the quality of service down. We can show many examples to prove this, of which the most relevant is housing benefits, of which more later. Housing benefits are not simple and unvarying; even less so will be the SUC.

The idea that services are merely ‘transactions’ is central to what I call command-and-control management. It is a purely quantitative approach that leads service managers to focus on solving the following equation: How much work is coming in? How many people do I have? And how long do they take to do things? All work is treated alike as activity to be managed. I refer to this as their ‘core paradigm’ (with apologies for consultant-speak).

For the consequences of such an approach we need look no further than the 2005 edict directing local authorities to introduce call centres. In the cause of reducing transaction costs, ‘telephone work’ was removed from service departments to call centres. However, access is not the same as service. The separation of telephone work from service led to the fragmentation of work – and, in every case that we (Vanguard) have studied, a sharp rise in demand above the volumes anticipated in the plan. That should have been a signal, as should have been a similar demand spike on the subsequent division of work between front and back offices. To minds trained to the core paradigm, however, more work automatically requires more resource to handle it. So managers specialise the work, move services to ‘lower-cost’ channels, put more know-how into computers (scripts) and, of course, hire more staff, all of which adds cost but little else.

In fact, the extra demand is what we call failure demand – unnecessary contacts resulting from a failure to carry out a service adequately at the first contact. In these industrialised (call centres / back offices) designs, failure demand is always shockingly high, sometimes up to 60 per cent or more of the total; even more shocking is that managers are unaware of it. Reports of lower transaction costs, although true, are irrelevant and, worse, actively misleading, since they disguise the larger truth that overall costs are remorselessly going up.

Failure demand is an easy idea to understand. To conventionally-minded (cost-focused) managers it represents an obvious opportunity to cut costs. It was no surprise that under the Cabinet Office’s preferred title of ‘avoidable contact’ failure demand itself became a local-authority performance target, reflecting the belief that mandating people to monitor and report it would cause them to tackle its causes. In fact, it encouraged them to underreport it. The irony is that the causes of failure demand are to be found in adherence to central directives about method imposed on local authorities by your predecessors and their civil servants – a straitjacket that effectively prevented them from redesigning processes, the only way to root out the causes of failure demand.

In services outcomes are not automatic. Things will always go wrong. In analysing failure demand it is essential to establish which types are predictable and which are not. Only the predictable is preventable (a preferable term to ‘avoidable’), and the only way to prevent it is to redesign the service. I shall return to this.

Consider the evidence against the industrialisation of service based on economies of scale that you are proposing for the SUC.

HMRC, supposedly the flagship of public-sector reform, is now several years into its restructuring programme. Yet the organisation admits that 4.3m PAYE taxpayers have paid too much, 1.4m too little and in the case of 18m it cannot be sure. The tax gap is rising, and accountants report service levels that are still on the decline. Meanwhile HMRC call centres have experienced a 100 per cent increase in call attempts (a sure sign of failure demand), and only answer one-third of their calls while the number of callers receiving engaged tones has risen from 7m to 39m.
Similar shortcomings are evident at DWP, another exponent of service industrialisation. According to the NAO, underpayments to those entitled to them total £800m, overpayments £1.1bn, and there remains a clear imperative for improvement.
A study by Advice UK showed that as much as 60 per of the demand occurring in advice centres comes from citizens, often the most vulnerable in society, trying to rectify mistakes or find ways round the failure of DWP and HMRC to provide a proper service. Costs imposed on advice centres by these failures are put at £500m a year. Costs to DWP/HMRC of re-work and legal appeals (most of which are won by the appellant) will be much higher.

The core paradigm, which you are now proposing to replicate for the SUC, contains three major flaws.

It treats all demand as ‘work to be done’ ignoring, inter alia, the importance of understanding failure demand.
It assumes that individuals should be held accountable for their work. In fact 95 per cent of individual performance variation can be shown to be attributable to the way the work is designed and managed – the system – and is thus the responsibility of management. Holding individuals ‘accountable’ results in sweatshop conditions (DWP call centres are described as ‘living hell’ by staff who have just carried out a two-day strike, while HMRC is disfigured by a long-running dispute on performance measurement) and disastrous staff relations.
It fails to absorb the variety of service demand. By ignoring the nature of demand and measuring and managing activity, as above, it ensures that costs will rise and service worsen. In both these organisations, work has been standardised on the assumption that standardisation leads to lower costs; in fact, along with specialisation, it is a primary cause of failure demand and thus raises costs.

HMRC and DWP managers have been led down this path by a mistaken interpretation of ‘lean’. What now goes under that name originated at the Japanese carmaker Toyota as a sophisticated means of solving the problems associated with manufacturing cars at the rate of customer demand. Self-evidently, the problems involved in calculating benefits or tax amounts for individuals or companies bear little resemblance to those involved in manufacturing cars at a particular speed. How did Taiichi Ohno (the architect of the Toyota Production System) train his managers? Did he train them in a toolkit like a mechanic and tell them to use it to solve whatever problems they ran up against? Quite the reverse. He had them study their system to discover their real problems and then devise specific techniques to deal with each one. ‘Lean’ has only served to worsen performance at HMRC and DWP by reinforcing the assumption that the problem should be made to fit the tools rather than the other way round.

As a ‘product’, the SUC is unique. But we can learn much about how to deliver it from experience in housing benefits. I first learned about housing benefits about eight years ago. Mark Radford, a director at Swale Borough Council, had read one of my books and called me. He had dutifully implemented the DWP’s guidance on how housing benefits services should be designed and managed, incorporating a front and back office with associated targets. The result was such

chronic backlogs that Swale was rated the worst benefits service in the country. Mark was receiving advice from the DWP’s help team which focused on adding resource to the back office to bust the backlog (the core paradigm). Solving the same problem, swathes of back-office work were outsourced by local authorities – an economic folly, since the contracts being based on transaction volumes contain no incentives for reducing them.

Reading the book, Mark realised that the problem was not in the back office but the front. As with housing benefits across the country, the focus on seeing people quickly led to fragmented input, increased overall processing times, errors, fraud, and loss of documents, to the point where ‘customers’ had to visit up to 10 times to get their problems solved.

Today Swale, and many other local authorities that have studied their housing benefits service as a system and redesigned it, deliver a high-quality service in days and have realised 30 to 50 per cent efficiency gains as a consequence. These gains represent the kind of results you should expect from taking the same approach to taxation and benefits.

The essence of the new design is i) to provide the necessary expertise at the first point of contact to satisfy all of the predictable ‘value’ demand (I have a claim, I have a change of circumstances); ii) to allow agents to ‘pull’ expertise for the less predictable demand, using measures that relate to the purpose (right money to right people a.s.a.p.); and iii) to switch management’s focus from managing activity to managing the whole system’s achievement of purpose. Using these joined-up principles, benefits offices have subsequently learned another lesson. People’s needs and problems come in a variety of interlocking forms and guises; solving them all at first point of contact offers huge potential (if usually invisible) cost savings since it reduces demand on other services. In contrast with factory designs such as HMRC and DWP, morale in services organised on these lines is invariably high because people are intrinsically motivated, illustrating once again the old saying that the best way to get people to do a good job is to give them a good job to do.

Under current proposals, the SUC will take seven years to deliver. They foresee massive investment in IT and standardisation. This is a hopeless formula for absorbing variety and will – as it always does – generate massive amounts of failure demand, citizen dissatisfaction and cost. By contrast, if what constitutes the Single Universal Credit could be defined today, housing-benefits offices redesigned along the lines described above could provide it quickly and efficiently tomorrow – guaranteed. Furthermore, to ensure that the Credit is fit for purpose, these housing-benefits offices should be used to develop the rules, taking risk out of the solution.

In the light of all the above, we therefore ask you to think again.

Yours faithfully

Professor John Seddon
Vanguard Consulting Ltd

Vanguard

Villiers House

1 Nelson Street

Buckingham

MK18 1BU

Tel: 01280 822255