Gordon Brown should take lessons from William Gladstone in 'saving candle ends'

I supposed the Treasury could massage the figures, but of course that does not tackle the real problem

How do you really boost efficiency in service industries? Stand aside from the political significance of Gordon Brown's spending review - intriguing though it was - and this is the stark question that matters most.

How do you really boost efficiency in service industries? Stand aside from the political significance of Gordon Brown's spending review - intriguing though it was - and this is the stark question that matters most.

The only way our society can continue to increase its standard of living is to increase productivity in services. The rest of the debate - whether a particular service should be in the public or private sectors, whether you can get rid of 84,000 jobs through greater efficiency, what is the right form of regulation and so on - all these other issues are secondary.

It was not so until quite recently. For most of the post-war period, living standards rose because of two things. One was favourable demography: a rising workforce relative to the total population. The other was increased productivity in manufacturing industry. Both those advantages are waning. In absolute numbers the workforce is still rising a bit and will do for a few more years, but the numbers of older people is rising faster. And while manufacturing productivity is still climbing, that sector is now less than 20 per cent of the economy. In relative terms, every year that goes by sees manufacturing shrinking still further.

For a number of reasons it has proved much harder to increase productivity in services than in manufacturing. Go to a car factory nowadays and the startling thing is that there seems to be hardly anyone there relative to the size of the building and the equipment. By contrast, go to a City dealing room and the place is crammed with people, noisy ones too.

For years it has been possible to increase manufacturing productivity by around 4 per cent a year. But in the private-sector service industries, it has been a struggle to increase productivity by more than about 1.5 per cent. And in the public sector it has been a struggle to increase productivity at all.

To say that is not to get at Gordon Brown or the British Civil Service, for all countries have the same problem. Far from rising, public-sector productivity, as measured by the Office of National Statistics, seems to be falling.

That is partly because of difficulties of measurement. Next week a review body, headed by Sir Tony Atkinson, will produce its first report on improvements on the way government productivity is measured. It will presumably conclude that the performance is not as bad as it looks. Already the Government has discovered that the NHS has been doing better than the previous calculations suggested.

What is beyond dispute, however, is that government staff numbers have risen by more than the government itself expected. The Institute for Fiscal Studies points out, as a background to the planned 84,000 job cuts, that "civil service numbers have exceeded Treasury plans in every set of annual public spending projections since 1999". It adds: "If the plans published in May 2002 had actually been achieved, there would already be about 45,000 fewer civil servants than there are now."

Looking ahead, the IFS is sceptical. It says that it is hard to compare actual administrative costs with planned ones because some £5bn of admin costs have been reallocated and called "frontline" services. And it notes that "prior to this reclassification, the Treasury has promised every year since 1998 to reduce administrative spending, only to see it rise consistently."

Given this disappointing performance, it is hard to see the Government achieving the 2.5 per cent annual increases in productivity built into its projections. The IFS notes that this would be more than private-sector service industries have achieved, though less than the former nationalised industries managed when they were privatised.

So what should governments do? That will be an immediate question for the next British government after the election. Given the Treasury's failure in the past, it is hard to believe it can succeed in the future. I suppose it could massage the figures, but of course that does not tackle the real problem.

There are only two possible ways forward. One is to hand over the functions to the private sector, and try to improve its own monitoring and specification of contracts so that benefits of greater efficiency really do accrue to the citizens. That is what may happen in the end. But this government is not going to do that, and in any case I'm not sure it is necessarily the best way forward. It is very difficult to specify quality service standards sufficiently precisely to make sure that the taxpayer does get value for money, and managing a web of service providers may be more difficult that managing a not-too-centralised bureaucracy.

The other is to apply the techniques developed by the most efficient private-sector companies to drive productivity upwards. And what are they?

Well, there is no single model, no single "right" way of boosting productivity. Instead there is a constant flux of experiment, of companies looking at the detail of their procedures and figuring out how to do things better.

Sometimes improvement is top-down. Ryanair is a good example of a supremely efficient Stalinist company, simplifying every detail of its working systems again and again to drive down costs. Its model is not so different from that of William Gladstone, still the longest-serving Chancellor (he did two chunks) and famous for his view that public finance should be on the principle of "saving candle ends". Maybe Ryanair's management style is Gladstonian rather than Stalinist.

But in most well-run companies there is a constant give-and-take between top management, middle management and talented people on the service delivery end. A lot of the work is in making information systems work more efficiently so that the person on the end of the phone can do the deal - let's say it is approving a mortgage application - more quickly, with less need to refer up the management tree, and giving the customer a more pleasant experience.

But this only happens if there is a culture that keeps asking two questions. Can we do this more simply? Do we need to do this at all?

The public sector is very bad at this. Let me give an example. Prisoners on remand have, at huge expense, to be escorted through heavy traffic to the courts, often for a two-minute formal statement and often arriving late. Quite often their case is for some reason postponed. Then they have to be escorted back.

The much more rational way of doing things - what one might call the Ryanair approach - would be to reverse the process and move the courts to the prisons. Judges and barristers don't need to be guarded. Indeed you could afford to send them there in chauffeur-driven Bentleys for the price of shipping the defendants around. And they would not do a runner as the more enterprising criminals seem increasingly inclined to do.

Yet, with the main exception of Belmarsh in south-east London, the courts are separate from the prisons. Yet when judges suggest even less radical efficiencies to the Home Office, I am told they get nowhere. That is not the way the public sector is taught to think. Come forward, William Gladstone, we need you now.