What lessons should be drawn from the string of disasters suddenly befalling Gordon Brown and his unlucky successor as Chancellor, disasters that have demolished in two months the reputation for economic competence that the previous Blair-Brown Government took ten years to build up? I can think of four. The first is about proportionality. The second is about responsibility. The third is about focus and priorities. The fourth is about politics and a fatal flaw in Gordon Brown's personality that many people suspected all along.

Starting with proportionality, it is knockabout fun to ridicule government computer schemes, and the unions are having a field day denouncing civil service staff cuts, but we must not throw the baby out with the bathwater. While the loss of the HMRC discs was certainly a serious mishap, it does not mean that the Government was wrong to try to save public money by eliminating 30,000 jobs through the merger of the Inland Revenue with Customs & Excise.

Still less does it demonstrate that the whole programme of “e-government”, which has computerised tax forms, driving licences, passports, health information and criminal records is fundamentally misconceived. Many of these bureaucracies are now working more smoothly and dealing with customers more quickly than in the days before computerisation. They are generally more efficient and easier to deal with than similar agencies in America, Germany or France.

Moreover, they now compare quite well with comparable private sector bureaucracies such as telephone companies, utilities, banks or airlines. A phone call to the HMRC or the Passport Agency is probably as likely to be answered promptly — and by an employee who is both courteous and can offer a useful answer — as a similar call to a utility or bank. Of course, the government bureaucracies are unpopular, but so are the telephone companies and banks. Large bureaucracies that handle millions of inquiries every year will always make mistakes, simply by the laws of statistics. What they need is inbuilt systems to prevent inevitable mistakes spinning out of control and becoming big disasters. This is the key not only to the HMRC fiasco, but also to the far greater disaster of Northern Rock.

The real outrage at HMRC was not the decision by “a junior official” to post a disc with 25 million tax records through the unprotected internal mail; it was his ability to access and copy these records without going through an elaborate security procedure. It is not good enough to say, as Alistair Darling did on Tuesday, that this official broke management guidelines. The design of the HMRC computer system should have made it physically impossible for him to do so. In the same way, the design of Britain's financial regulation should have made it legally impossible for Northern Rock to get itself into such an overextended position that it needed a £30 billion taxpayer bailout to avert a collapse in which depositors would have lost their savings.

This brings us to responsibility. A junior official at HMRC may have been directly culpable in the case of the missing discs, but true responsibility is clearly located farther up the hierarchy. What ought to be challenged is not the principle of computerising information, nor the decision to merge Revenue and Customs, nor the cuts in staff. The obvious problem lay in the way that HMRC computers were designed and managed, which would seem to pin the blame primarily on the computer boffins, many of them working for private consultants, rather than civil servants themselves.

But in reality the chain of responsibility must be followed farther. Just as the FSA and the Bank of England were regulating Northern Rock within a system designed by Gordon Brown in 1998 to satisfy the criteria that he considered most important, computer consultants design systems to achieve objectives ultimately specified by ministers. The question therefore is how much importance ministers attached to security and how this was defined.

The key question in allocating responsibility, therefore, is about priorities and focus. Priorities, because the need for data to be secure in the HMRC computer had to be traded off against other desirable features such as convenience and cost. Focus, because security could never be imposed as an absolute requirement, any more than cost or convenience. It was therefore necessary for somebody to decide which aspects of security were particularly important and which could be more readily compromised.

This issue of focus relates to a worrying general trend in British public administration — and one that seems to be accelerating under Mr Brown. Britain's data protection regime will prevent you finding out when you last paid your own utility bill unless you can remember an obscure password, yet it fails to prevent 25 million bank accounts being disclosed. This happens, at least in part, because Britain's data protection laws are so broad and comprehensive that they lose the distinction between important security requirements and pettifogging bureaucracy. The upshot is that the data security laws are widely flouted and despised.

As the Government's information commissioner, Richard Thomas, himself said in a recent interview: “Sometimes data protection gets dismissed as something which is rather bureaucratic, it stops you sorting out your granny's electricity bills. People grumble about data protection, but boy is it important in this new age.”

The point missed by Mr Thomas — just as it is missed by the panjandrums of health and safety and racial equality and employment rights and consumer protection and a host of other desirable regulatory objectives — is that people and social institutions, including even bureaucracies such as the HMRC or the FSA, will only accept a certain amount of regulation and government interference. If attempts to interfere with everyday life pass beyond this threshold, then intelligent regulation can quickly degenerate into mindless box-ticking, which only distracts attention from real dangers.

It is therefore essential for regulators to concentrate their attention on the big issues that are really important — protection against wholesale data losses of the kind at HMRC, protecting the taxpayer against enormous losses of the kind now threatened by Northern Rock — while leaving individuals and organisations as free as possible to get on with their lives.

If this is true of regulators, how much more true is it of the politicians who direct them. This is where we come to Gordon Brown's fatal flaw. Instead of making big decisions about the essential functions of government — foreign policy, law and order, energy and environment — Mr Brown seems to have an irresistible urge for fidgety interventions in every aspect of everyday life — banning plastic bags, creating tax subsidies for cycle commuters, introducing identity cards, saving jobs at Northern Rock.

The more Mr Brown engages in such micro-management, the more he will be sucked into one fiasco after another. That way lies ridicule and political defeat.

Times Recommends

• Fundamentalist brew: faith and power
• Big guns wheeled out for the marginals
• Wanted: new merchants of light