IT'S PAYBACK time. As the election approaches, the government is demanding returns on an unprecedented period of IT spending. Massive programmes in health (the national programme for IT - nearly pounds 2 billion this year), education (pounds 2.6bn), local government (pounds 3bn), defence (pounds 1.9bn), justice (pounds 1.2bn) and central government departments (pounds 2.7 bn) take total public-sector spending on information technology up to pounds 12bn this year, making the wider public sector the computer industry's biggest customer. According to Kable, a research company, by 2006/7, the public-sector IT market will be worth more than pounds 16 bn - 72 per cent higher than 2001/2.
The result of this huge modernisation expenditure ('no investment without reform') is expected to be the automation of 100,000 civil-service jobs, removing 16.5 per cent from central-government running costs and ploughing more than pounds 20bn back into frontline services such as hospitals, schools and criminal justice.
In the 21st century, for example, the government believes there's no reason why every central department should have its own separate payroll, HR, finance and property management systems, or why the public sector as a whole should have 30,000 back offices collecting and processing information. There are similar opportunities, it believes, for rationalising front offices where officials transact with citizens face to face.
There's no doubt about the spending. There's no doubt also about the scope for improvement. But will the computers that are supposed to link the two actually live up to the government's expectations by improving efficiencies and helping it win the next election? There are two reasons why this is, as they say, a big ask.
The first is history. The spending programme includes a number of huge and complex projects - NPfIT (for the Health Service) merging the Inland Revenue and Customs and Excise computer systems Department for Work and Pensions Criminal Records Bureau a 10-year Defence Information Infrastructure project, with identity cards and the unification of the justice system on the horizon - in which the UK's track record is mixed, to say the least.
Figures quoted by the British Computer Society suggest that just 16 per cent of large UK projects meet all their goals. Other estimates are lower. Hence the steady stream of computer disaster stories: the 18-months-late pounds 400m Child Support Agency computer, for example, or the pounds 318m (so far) system for the magistrates' courts, referred to by the chairman of the Public Accounts Committee as 'one of the worst IT projects I have ever seen'.
When the Audit Commission looked at the well-publicised episode of the Passport Office computer, it found that the new system had raised unit costs by 30 per cent. The National Audit Office is now preparing to take a look at NPfIT.
The second reason for scepticism is method. Although the failure rate of such projects is partly to do with poor project management, more importantly their rationale is flawed from the outset.
When Michael Hammer wrote his celebrated 're-engineering' manifesto, 'Re-engineering Work: Don't Automate, Obliterate', in 1990, the kernel of the idea was the need to recast the way work was organised. He saw, from examples in Japan and elsewhere, that by abandoning a system in which work (processing an order or paying a bill, say) was passed sequentially from one department to another and instead managing it as a seamless process from beginning to end, huge improvements could be made: a bill could be paid (or housing benefit processed) in days rather than the months it traditionally took.
Hammer got many things right. He emphasised, for example, that the 're-engineered' process worked by putting the decision point where the work was carried out. 'The new principle suggests that the people who do the work should make the decisions and that the process itself can have built-in controls,' he wrote.
In other words, re-engineering would reverse the fragmentation that made assembly-line jobs, whether in a factory or local or central government office, so soul-destroying. And by putting decision points in the flow of work it could get rid of suffocating bureaucracy and layers of management control.
Unfortunately, organisations overlooked these humanistic subtleties in favour of Hammer's other prescription: that computers were an essential component of this large-scale reorganisation. The result was the exact opposite of his brave new world of work. Instead of obliterating departmental functions and boundaries and re-engineering work in an end-to-end flow, companies queued up to buy IT 'solutions' that not only preserved the bad old systems in aspic, they actually disabled further improvement by locking in the old methods in. Hence the subsequent finding that the majority of re-engineering attempts failed.
Unfortunately, in the public sector this tendency is only now reaching its apogee in the government's current spending review. Almost all public services are ripe for reshaping and streamlining as a process from the customer's point of view. But the current regime of specifications and targets, completely functionally and activity based, is utterly antithetical to any such project. One ex-civil servant remembers being instructed how to pick up a pencil faster in a desperate attempt to squeeze efficiencies out of the existing system.
In a travesty of 'reform', what is now being proposed is giant automated information-processing factories for mass-producing transactions - procurement, information collecting and processing - which will at the same time institutionalise the waste inherent in existing systems and perpetuate the dehumanisation of separating work from decision-making.
A classic example of the specification cart preceding the improvement horse, thus effectively bolting the latter in its stable, is e-government. The government mandate that all possible public services should be deliverable electronically by the end of 2005 will cost around pounds 7.5 bn. But as departments and local authorities rush to meet the deadline, public takeup of the spanking new e-services is so disappointing that the government has launched a marketing campaign to boost it. A local authority or agency website, after all, is only another way into the shop. By specifying e-government without any idea of what matters to an e-citizen, the government will have simply added an expensive new entry channel to existing ones. Unsurprisingly, Kable estimates that on current trends e-government will never remotely cover its costs, let alone free up resources for the front line.
Mere pen-pushing to specification ('reply to all queries with two working days') and control is indeed demeaning and demoralising work. But that's because of the way the work is designed. Contrary to the political discourse about bureaucrats and time-servers, the truth is that working to improve back-office support for front-line public service is both worthy and essential, work anyone could be proud of.
Improving public services could and should form part of the job of all civil servants. The tragedy of the current regime - assembly-line work driven by targets and specification - amplified by what's now being proposed, is that it will make improvement impossible, alienating public service workers and lifting service levels only on the margin. The truth is that the real 'reform' that's needed is that of the government's approach.
Otherwise today's IT bonanza will benefit the IT and consultancy industries more than those at the front end, whether citizens or those who deliver the services.
The Observer, 26 September 2004