Category: Uncategorized

How to reform public services: one person at a time

An extraordinary story is emerging from the public sector. The bad news is that a colossal amount of the money spent on public services is wasted (nothing new there, then). But the good news is that while there is certainly a huge effectiveness issue, despite the endless narratives of doom there is no resource problem. We don’t need to spend more money on the NHS, social care or the elderly, or possibly even housing. In fact, more money may be the last thing the public sector needs.

Getting to this point has been a long learning journey shared by a consultancy, Vanguard, and a number of councils, services and other organisations curious (and brave) enough to attempt something different from the standard industrialised, commodified public service models of today.

The starting point a number of years ago was Vanguard’s use of systems principles to devise a method for improving individual services. The results of creating joined-up services were startling – benefits paid in days rather than weeks, housing repairs done at the time and date specified by the tenant, moving a hospital’s stroke performance from worst to near best, dramatically lowering local crime rates, to name just a few.

But in each case solving the apparent problem revealed bigger ones lurking behind, often outside the individual service’s control. Dealing with people quickly and efficiently in A&E is good, but less so if they shouldn’t be there in the first place. Likewise with petty offenders if they turn up drunk and disorderly again as soon as they exit custody.

In other words, the presenting problem was typically the symptom of another. So services started working with others to tackle the next layer of issues. Learning from housing repairs and benefits helped other separate sub-departments to overhaul housing as a whole. Teams from social services, GPs, physios and others wrought wonders in adult social care, transforming lives with the simplest (and cheapest) of means (better lighting, walk-in baths and showers). And so on.

In all these initiatives, although cost wasn’t the priority and the amount was impossible to predict in advance, overall costs fell as a consequence of providing a service that met people’s needs. The (counterintuitive) lesson is that bad service is much more expensive than good, because of the huge cost of rectifying mistakes and failure. To rephrase the learning (even more counterintuitively): managing cost pushes cost up, while managing value does the opposite.

Why should this be? Like the original NHS, public services were set up to ‘fix’ something: mend a broken leg, pay a benefit. But in a more complex world with ageing populations, the nature of the problems faced has changed. They are no long ‘fix-it’ but ‘help-me’: the need to manage long-term illness or mental health issues, or cope with family break-up, difficult children, poverty and unemployment, for instance.

Using ‘fix-it’ means to solve ‘help-me’ problems is at the heart of the supposed public-service ‘crisis’. Instead of giving people the means to get unstable lives back in order (so they can join the stable majority that makes minimal claim on public services), ‘fix-it’ methods (fragmented, unjoined-up, episodic, delivering centrally specified packages of care) keep them living their instability in an unending loop of public service dependency which never solves their problem. This is the dark secret of public-service waste: the vast quantity of of resource (maybe 50 per cent) that is consumed by the 5 per cent of the population that presents over and over again, often to multiple service doors.

That is a giant step forward in understanding. And it has led to an equally important step in the journey – ‘the most important work we have done in the public sector’, according to Vanguard chairman John Seddon. At a conference in Stoke in October, two councils – Stoke and Bromsgrove & Redditch – outlined a radical ‘whole systems’ approach to public service reform which seeks to use these learnings to guide services not just within traditional sectors but across a whole geography. Bringing together partners including fire, police, health, housing and social services, among others, it aims to understand individuals and families in need in order to fulfill a simple purpose: ‘help me to resolve the problems (that prevent me from living a good life)’.

It’s early days, but operating to purpose and principles, switching off targets, departmental procedures and top-down management structures to do just what is necessary to meet the purpose, pilots in both councils are producing profound results. According to the council chief executives, ‘Citizens previously labelled as “lost” are living good lives, demand on services are going down, and the size of the opportunity to reduce cost is staggering.’ In one pilot in housing (admittedly based on very small numbers), rent arrears, evictions, antisocial behaviour, demands on the NHS, police and justice system were all down compared with control groups; and while the cost to the council increased, overall costs across all the agencies fell. On the basis of the fully costed cases, the overall saving to the public purse in one council would amount to £81.5m over two years.

Just as the failure to solve problems multiplies apparent demand (‘problem families’ are typically well known to police, the NHS, schools and fire as well as social and other council services), the reverse is true: when people get their lives back on track, demand evaporates across all services, including in tantalisingly unexpected areas: stable people are more likely to tackle obesity, smoking and alcohol issues, for example.

Conventional wisdom as incorporated in current service design holds that knowing and understanding individuals can’t be afforded; the lesson of Stoke and Bromsgrove is that knowing and understanding people is the only way that good services can be afforded. As Seddon sums up the learnings so far: ‘Demand is stable (not rising), demands are simple (not complex), we have plenty of capacity (no financial Armageddon), we have an effectiveness problem (not an efficiency problem), and when we provide services that meet needs, we improve. Costs fall dramatically and, most remarkably of all, demand falls. Yes, fewer problems in families and communities – isn’t that what public service ought to be about?’

The real lessons of Nokia

How are the mighty fallen. Once high-flying Nokia has sold its mobile handset operation to Microsoft. Computer maker Hewlett-Packard, a venerable Silicon Valley pioneer, has crashed out of the Dow Jones Industrial Index. It seems to be true, as a recent post on MIT Technology Review notes, that ‘the lifespan of great corporations is getting shorter and shorter’. Back in the 1950s, when a company made it on to the S&P 500, a roll-call of the corporate great and good, it could expect to stay there for 60 years. These days, on average it won’t get out of its teens. But why?

‘Technological disruption could be one big reason’, opines the author – and it’s a view common to the point of cliche. Thus, Kodak’s demise is always attributed to its being overtaken by the switch to digital; Palm, RIM and Nokia were leapfrogged by the iPhone and have never managed to catch up.

Yet let’s think for a minute. With a few exceptions – technology really has disrupted newspapers – blaming technology is no kind of explanation. It simply restates the problem. A more convincing, and down-to-earth, interpretation is that these companies failed to respond to changing markets because something in their internal organisation and culture didn’t let them.

One ex-Nokia manager cites a Finnish newspaper which points the finger squarely at the company’s stock option scheme for top and senior managers. In this narrative the options progressively corroded the culture from within, setting up debilitating competition between groups, people and platforms. Competition exacerbated divisions created when former chairman Jorma Ollila restructured Nokia’s business in 2003 into three separate groups, each with its own budgets and targets. The snag was that market/customer demand had to go through each of these divisions. By 2006 it was apparent that lack of trust and bad blood had caused a breakdown of cooperation between people and divisions that was directly affecting the company’s markets. But by then it was already too late.

In his biography of Steve Jobs, Walter Isaacson describes a strikingly similar situation at Sony in the run-up to the launch of Apple’s iPod and iTunes. By rights, that market should have been wrapped up by Sony, which unlike Apple already had both the technology and the content to make it its own. It failed to act because it was paralysed by rivalries between divisions, something that Jobs had expressly legislated against by insisting on a single balance sheet and P&L. As one Nokia manager described the result: ‘Everyone had two personal targets per year (defined by the hierarchy, top-down); [the result was] you only did your own things that enabled personal bonuses even when the business environment changed dramatically’.

Moreover, says the insider, at operating level Nokia’s work design faithfully mirrored the traditional reductionist thinking evident in remuneration. Complex tasks were systematically broken down into multiple simpler subtasks. ‘But knowledge work is very different from assembly line work… There was no flow in the work because every small decision had be escalated to multiple steering groups (or steering groups of steering groups of steering groups), and decision making wasn’t integrated with work. There could be weeks of waiting for the decision and for the time being, some other work had to be done to keep the “worker utilized”. Obviously, resource planning became a very complex task for the organisation’.

Another side-effect of this way of working became apparent when Nokia’s fortunes changed and it began laying people off. Highly reputed as it was, Nokia’s high-specialisation work design meant that engineers with 10 years of programming experience focused on, say, specific Bluetooth drivers in certain operating systems, and only used to following the specifications they were given, were not easily employable elsewhere.

In this light, the ‘technological disruption’ story looks like a lazy oversimplification. The real moral is more sobering. While rapid technological change means that more companies can ride the wave to the top, it also flatters managers (banking their huge option awards as they do so) into believing it was their own managerial genius that deserves the credit rather than the good fortune of being in the right place at the right time. The receding tide separates the firms that got the more boring but lasting things like reward and work design right from those that didn’t. In other words, the turning technology cycle just reveals how many companies simply weren’t that great, after all.

Facing the facts on 111

No apologies for returning to health this week. On BBC Radio 4, a Face the Facts edition by John Waite perfectly summed up the cul-de-sac we have run ourselves into with the NHS: having given a forensic description of what went wrong with the disastrous 111 number, the programme utterly failed to draw the conclusions of its own story and offered no hint of a way out of the vicious circle of rising demand and rising cost.

111 had its genesis in a 2010 decision to replace NHS Direct, in the official narrative an accepted, well-respected resource for patients, especially at night and at weekends when the rest of the NHS was closed, and for the NHS a means of relieving pressure on GP surgeries and A&E departments. The ‘problem’ was that it was expensive (about £123m a year); so the idea, no doubt underpinned by expensive consultancy advice, was to use technology to make efficiency gains by economising on highly-paid professional medical staff.

In other words, 111 was conceived of as a network of standard call centres of the kind that every customer loves to hate, manned by call handlers with no technical knowledge, reading a script from a screen, with professional expertise ‘on call’ behind. Just as classically, NHS England put contracts out to tender based on unit cost, or cost per call. The benchmark, according to Peter Carter, general secretary of the Royal College of Nursing, was £7 for a non-clinical call and £10 for one that had to be referred on to a clinician. Carter said: ‘The only way they [the contractors] can do it is by compromising the staff mix – they’re trying to do it on the cheap’.

When the system went live in April, the result was as wearily predictable as the 111 computer script. Technology malfunctions were the least of it. Lives have been endangered if not lost because neither patients nor computer programmes can easily distinguish between conditions that are really non-urgent and those that look it but aren’t (as a doctor noted, some conditions are almost impossible to diagnose over the phone but easy and instant for a trained physician in person). In other words, the system is unable to absort the presenting variety – a computer script is an utterly inadequate substitute for trained human judgment.

Meanwhile, as at any script-driven call centre managed for cost, call-handler morale was low and turnover consequently high. Staff felt undertrained and overworked and looked on in dismay as many colleagues treated it as casual work, clocking on for a few hours and then drifting away again. Call handlers left because they found it ‘just a really depressing environment’.

But as ever, cheap actually isn’t cheap: managing costs breeds more costs. Under the previous regime the original incumbent, NHS Direct, had been getting £24 a call, and its first bid for 111 contracts was too high. So it ‘reworked the figures’ to get under the hurdle and was duly awarded 11 areas, the largest provider. Unfortunately, its first calculations were nearer the mark. Calls were taking double the five and seven minutes estimated for unreferred and unreferred-on calls, for which it was paid from £7 to £10. It pulled out of two contracts even before the launch and now wants to give up the rest as financially unviable.

As depressing as the litany of failure is the response to it. NHS England claims that 111 provides a ‘safe, proven, consistent clinical assessment of callers’ symptoms’, and that 96 per cent of calls are answered within 60 seconds. Even critics assume that the issue is teething problems and poor execution. NHS Direct is blamed for ‘getting its sums wrong’, and the government for rolling 111 out too quickly and trying to cut corners on cost. Rising demand is taken as a given. Also taken for granted is that 111 is necessary – so the only solution is to do it better (read: more expensively).

In fact, everything about 111 was wrong from the beginning, including its purpose and starting point. But you wouldn’t know that. Nowhere is there a hint that the 111 story is not about doing things wrong, but doing the wrong thing. Nowhere is there a hint that the cost and time that matter are not cost and length of call, still less the time taken to pick up the phone, but the end-to-end cost of solving the problem so it doesn’t recur. Nowhere is there a hint that the only sensible way forward is to establish the real, as opposed to repeat and knock-on, volume of underlying demand and design a system to meet it, 24/7.

There being, in short, no sign that anyone has learned anything at all, the prospect is that we will do the same thing all over again and expect a different result. Which, as Einstein declared, is one definition of insanity.

Two cheers for Berwick’s report on patient safety in the NHS

There are plenty of things to approve in Don Berwick’s report on patient safety, as well as some to query. The fact that both good and less good are the opposite of what most people think they are says much about the state of public debate on the NHS as well as the difficulty of making the changes that Berwick recommends.

Overall response to the report has been lukewarm. It has been criticised as too general and short on particulars: no headline proposals for tough legal sanctions on individuals or regulation, no minimum staffing levels for wards (a ‘missed opportunity’, according to the nurses). In fact these reflect what’s good about it. Along with other clear principles – constancy of purpose in minimising harm, suspicion of numerical targets, the need to drive out fear and build pride and joy in work, attributing blame to a bad system rather than bad individuals, turning the NHS into a learning organisation – they come straight out of the Deming management handbook. Real cheers, then, for a report that recognises that the NHS is a system, so that change has to be systemic too.

In fact, that was one of Berwick’s most important points: awful though events at Mid Staffordshire were, he noted in a Newsnight interview, the fact that the NHS was a single system meant that when things went wrong there was an opportunity to do improve them system-wide – an impossibility in the fragmented US, for example.

The problem is that many people – inside the NHS as well as out – won’t get the radical implications of the system point. Without being exposed to them in action, they rarely do. We prefer easy solutions to hard ones, and sound-bite-obsessed media don’t help. And for them, the report will indeed seem general and even comforting. Learning organisation, ‘culture change’ – what could be softer focus, less contentious than that?

Paradoxically, one of the warning lights is the choice of patient safety as top priority. Of course, that’s what Berwick was asked to report on, and to be strictly accurate, it is ‘the quality of patient care, especially patient safety’, that the report singles out as the NHS’s most important aim. But safety and high-quality patient care, as Vanguard’s Andy Brogan, a keen and involved student of NHS affairs, points out, are ‘hygiene factors’, something necessary to achieve the purpose but not identical with it. Safety is a priority for air travel, too, but it’s not the purpose, and nor is it for the NHS.

‘When you have to make safety a priority, it tells you that the focus must have been wrong before’, Brogan reasons. ‘We ended up with a safety problem because we weren’t focused on purpose and value. A patient safety focus is not the same as a focus on purpose and value and therefore doesn’t remove the precipitating cause of the problem. It treats the symptom.’

The purpose of the NHS is to allow people to live healthy lives, in our own context. For Brogan, the big story in the NHS, nowhere hinted at in Berwick, is that it has conceived of its purpose from the wrong point of view: it is producer- rather than patient-centric, reducing patients to their conditions and then handing them out more or less standard medical packages – ‘push’ rather than ‘pull’. This is Fordist, industrialised, Model-T medicine – and the surface pressure and busyness conceals huge underlying waste and inefficiency.

‘Almost every improvement effort I see in the NHS assumes that it has a single-loop problem to solve – i.e. we are doing the right things, we just need to learn how to do them better’, says Brogan. ‘But the NHS has a double-loop problem – it is doing the wrong things. When you see people’s demands in the context of the lives they want to lead, many of the condition-shaped interventions are simply the wrong thing to do. Making them safer is just doing the wrong thing righter’.

Thus it’s all very well to call for the NHS to become a learning organisation – who could disagree? – but it can’t do that unless it changes how it measures. Measures can be used either for learning or accountability and control but not both (the argument is here). Broadly speaking, NHS measures are related to activity, not purpose from the patient’s point of view: number of GP or A&E visits, for example, all of which are assumed to be demand needing to be managed, which is done by rationing the available resource to match. No learning is involved; the measures are used for control, not to learn how to improve patients’ lives.

When measures are related to the purpose of enabling people to live healthy lives, on the other hand, the first discovery is that the more the organisation is ‘improved’ to increase throughput and cut costs, the worse it serves its patients. Take limiting GP appointments to 10 or even 8 minutes, or sometimes one problem at a time. This boosts throughput, but the price of failing to deal with the complete issue at first pass is multiplication of backed-up ‘failure demand’ as patients return for further consultations or present themselves at A&E instead. These pressures led to the setting up of first NHS Direct and now the disastrous 111 number – both unwitting factories for amplifying yet more repeat ‘demand’, the epitome of non-learning organisations.

Vanguard’s figures suggest that up to a terrifying 85 per cent of all demand into the NHS is failure demand. While much of it is medically justifiable, it needn’t and wouldn’t happen if it were dealt with properly the first time round. This is the nightmare treadmill that the NHS has to get off and that is the essential subtext to Berwick – small print that is more urgent and alarming than the report’s headlines. This is why the influential Roy Lilley, broadly a Berwick admirer, describes it as ‘the most exasperating and annoying report I’ve ever read.’ While not going that far, and accepting that ‘it’s important not to lose the political will,’ Brogan also expresses his frustration: at some point, he says, ‘someone will have to get up and call a spade a spade’.