Perhaps we should start by asking why ministers want shared services. It is because they are attracted to the idea of economies of scale.
I wonder if they read the papers. Ford and GM, the pioneers of economies of scale, are going down the pan. Meanwhile Toyota continues to grow. What is the secret of Toyota's success? They have rejected economies of scale; instead they have a completely different way of thinking about the design and management of work. Instead of economies of scale, they have economies of flow.
As some of you will know, my work has been the translation of the Toyota principles for service organisations. In short this means designing and managing your organisations as a system, the benefits are better service at lower costs; everything the minister wants but only achievable if you ignore ministerial edicts.
Lets take a systems view of benefits processing. It starts with understanding demand. When I visited a benefits processing office in London, the manager was pre-occupied with the length of the queues. They were out on the road. His answer was to move the desks further inside the building, to get the people off the street. I told him he would learn more if he understood why people were in the queue. When I visited a benefits processing office in the North, I found mangers had created a ticket system, just like you find in Tesco's delicatessen. Only the first fifty people to take a ticket were allowed to ask a question (and only one), everyone else was only allowed to drop documents off. The manager was obsessed with meeting his target - seeing people within fifteen minutes. I told him he would learn more if he understood why people were in the queue.
What you learn when you study why people are in the queue is that many people have to make a number of visits, as many as seven or eight, to get their benefits sorted. You find people are being asked to bring in things they have brought in before. You notice in these offices they have signs on the wall saying 'if you are badly behaved with our staff, we'll call the cops'. How would you behave if you had to turn up eight times and were asked to bring in things you've brought in before?
After you have learned about demand you would want to learn about capability. How long it actually takes from when someone says 'I want to make a claim' until they have the money or the answer 'no'. You find no one knows. Of course managers know how they are doing on the government targets but these tell you nothing. It is not unusual to discover that the true end-to-end time is as high as an average of 50 days, with an upper limit of 150 days. How does this square with these authorities apparently meeting their government targets? They cheat. Even the beacons cheat. There are many ways to do it. You can stop the clock and end the claim if after 20 days you don't have all the information you need. (Do you think the fact that we don't have all the information is anything to do with the design of the system? Absolutely it is.) You can allocate a nominal income. This means the claimant will come back saying: "I don't earn £1,000, you've got this wrong". But when the claim starts again it doesn't count.
The cheating is even encouraged by the inspectors; they call it sharing best practice. Why do inspectors behave this way? Because their hierarchy is looking for evidence of improvement. Ministers want to crow about their success in improving the public sector, any plausible story will do.
Remember Nick Raynsford's conundrum? When he was minister he stood up at a conference and said how do we square this: The target data shows public services are improving, yet consumer surveys suggest nothing has changed, if anything things are worse. He rationalised. He thought maybe people take a bit of time to notice that things were better. No Nick, the targets are actually making performance worse.
Lets get back to benefits as a system. What are the causes of such long end-to-end times? How does the end-to-end time compare with the time it would take to do the value work? You know how long it takes to process a claim if you have all the information you need? - less than thirty minutes.
The things that make time go longer begin with the overall design: All authorities now have a front office / back office design. The two are connected with document image processors. This has happened because the DWP mandated it. But this design fragments the flow of work. Who comes in with everything you need to establish eligibility and entitlement? Of course no one does. But whatever they do bring in gets processed - people have to, to meet their targets. The document image processor introduces errors; things get wrongly allocated, lost and duplicated.
The consequence of this design is backlogs in the back office. They have sprung up all over the country. To help local authorities out of the problem, the DWP's 'help team' - I call them the 'no help team' - have been encouraging the use of private sector contractors to do 'backlog busting'. We are paying millions to the private sector to bust backlogs caused by the DWP's design. It is madness.
I was working recently with a private sector 'backlog busting' provider. They had taken over benefits processing for a council and we helped them implement the systems solution. As with others the service improved dramatically. More importantly they saw how impressed the council was. Being opportunists they said to me 'we must go and sell this to other councils'. Leaving aside the questions of whether doing it once makes them experts, or who owns the intellectual property for the systems solution (clearly not things that matter to some), I said to them you don't need to go out and find people to sell it to, you could go to everyone you are doing backlog busting for and tell them you have a better answer. You could give them the solution and it would enable them to cope with all of their work and they would no longer need the backlog busting service. "But we make millions from backlog busting" was their reply.
Many of these contracts are based on transaction costs. Most of the transactions being handled are 'failure demand' (demand caused by a failure to do something or do something right for the customer) a type of waste that is peculiar to service organisations. It is not in the interests of the outsource provider to do anything to reduce the volume of transactions. As Taiichi Ohno would have put it: 'we are paying the private sector to eat our muda'.
Why do ministers think the private sector can do it better?
There are now many local authorities that have the systems solution. They pay benefits in days, their customers are really happy, staff are engaged, because everyday they come to work to solve problems, improve their system. These people have learned that the DWP's 'solution' was, quite simply, the wrong thing to do. The front office / back office split means a handover; this means greater reliance on a 'check-list' approach in the back office and this tops the system absorbing variety. Worse, the front office / back office split opens the system up to fraud. You have to hand it to the DWP, they couldn't have invented a worse design.
The front office / back office split paved the way for 'shared services', in simple terms it means sharing back offices. It is doing the wrong thing wronger. In the East of England they have commissioned KPMG to report on the potential for sharing services. KPMG's brief was: 'if we have different IT platforms and different processes, are there advantages from standardisation and sharing?' It was to ask the wrong questions. The question should have been: 'is there a role for the back office in benefits processing?' I'd have told them for nothing, the answer is no. Effective benefits processing is a front office service.
Why do ministers want shared services? Who is blowing in their ears? The consultants, private sector BPO providers and the like, and the IT companies, for all of them have much to gain.
I urge you to defeat this motion. Benefits can and should be improved. But the gains that can be made should be used to give better public services, not greater returns to private sector shareholders. Send a clear signal to the private sector sharks swimming in the shared services pool.
When your members are lobbied by people who have a problem with benefits processing, you don't want them to say 'there is little I can do, the work is being done by people 200 miles away and they are meeting their service level agreement'. People want services that work. Where authorities have implemented the systems solution costs fall and service improves; claimants bring flowers, they cannot believe they don't have to visit again.
Defeat this motion. You don't want stressed-out staff, their unions opposing you, as they are in the revenue service right now. You want your staff at the heart of good service delivery. Only people can absorb variety. Putting them in factories connected to the customer with IT systems can only fail.
Defeat this motion. Vote for economies of flow. Its better for claimants, its better for your staff, its better for your costs, its better for your taxpayers, its better for everyone.
I urge you to defeat this motion.
John Seddon is an occupational psychologist and management thinker.
