E L E C T R O S P H E R E    Issue 1.02 - June 1995

Affirmative, Minister

By William Heath

The unstoppable force of new technology has now engaged the immovable traditions of the British civil service. Amid a welter of study groups, committees, white papers and green papers, the future of Whitehall is taking shape. But the process is agonisingly slow.

Britain finally has both ministers and opposition spokespeople who at least understand the arguments about the role of technology within government. Amazingly enough, the Internet is proving a sufficiently powerful force to bring change - or at least talk about change - even to the tradition-bound corridors of British government.

The first UK government Web page (www.open.gov.uk) went on the Internet in late 1994 - and, inevitably, was hacked within minutes, by a student who wished to improve its design. The first minister to multi-cast a speech over the Net was Cabinet Office minister Robert Hughes, discussing the information superhighway at a public consultative meeting in November 1994. A minister interested in civil service reform and in information systems seemed the right man to embrace new channels of communication. (But a minister who confesses to an extra-marital affair, as Mr Hughes did in February, finds he has media of the traditional tabloid format to contend with. His resignation was immediate.) Budget details are published on the Net; and the schools inspectorate, Ofsted, has announced it will place independent reports on schools on the Net, to give parents access to information about school performance.

Now is the lull before the argumentative storm. Labour has new policies on the information superhighway. Think tanks and pressure groups - the Adam Smith Institute, the Social Market Foundation, the Institute for Public Policy Research, Liberty, the Campaign for Freedom Of Information, Demos and Charter 88 - are honing positions on how new technology could, or should, affect the process of British politics. The government itself seems to be in an open, consultative frame of mind while it ponders what to do next. Then, around the end of 1995, the decisions should start to hit the fan.

Three questions lie at the heart of the debate: what value does information technology provide, and how can it be measured; how is the civil service to be managed; who owns and controls information? None of them are easy to answer. Let's start with what should, at first glance, be the easiest: does the taxpayer get good value for the £2.5 billion that the civil service spends each year on information technology?

This may not be the most exciting question to ask about technology and government, but it is still the one foremost in the minds of ministers. For government views technology primarily as a productivity booster, justified entirely on cost. This may seem depressingly unimaginative in the age of the Internet, but what is still more depressing is that government seems to be struggling even to achieve its own narrow goals.

The Foreign Office had its accounts qualified after the disastrous introduction of a new finance system, and had to abandon an overambitious secure-office automation system. The Employment Department wasted almost all of the £48 million it spent on a system called Field, which was designed to help manage the Training and Enterprise Councils. The Ministry of Defence, which has the biggest technology budgets, tends to run into the most spectacular difficulties. Its digital defence communication system, called Boxer, is now running four years late and more than £200 million over budget. Not surprisingly, Parliament's spending watchdog, the National Audit Office, has recently been less than complimentary in its reports on government's technology performance.

But the real worry is that over-spending may be merely the tip of the iceberg. To miss a budgetary goal is one thing, but it is entirely another if the budgetary goals themselves are inaccurate, unrealistic, or both. Yet that is the problem government is now struggling with. The government devotes so much effort to wise budgetary management that it has developed its own acronym: VFM (Value For Money), while Whitehall has its own advisory agency, called the CCTA (which once stood for the Central Computer Telecommunications Agency, but which in these modern and decentralised times is known only by its initials). The CCTA creates tomes on best practice in the use of technology and procurement support - not to mention British and EU purchasing laws.

The typical 50-ish-year-old civil servant, be he or she dutifully obedient or sullenly resentful, has learned through hard experience that following these rules is the key to career survival. Yet in its attempts to analyse the value of technology, bureaucracy can sometimes take leave of reality.

Take, for example, the logistics project which forms the core of the automation plans of the Royal Air Force (RAF). The project will cost about £400 million and will transform the way in which the RAF ships supplies around the globe. To justify the expense, bureaucrats projected savings of up to £800 million under the new system. The snag was that the project started just as Communism was beginning to crumble in the late 1980s and is not scheduled to finish until 2003. By which time any number of new cold wars may have started.

Trying to plan for what everyone accepts is a transformational technology over this sort of timescale leaves bureaucrats in a double bind. If they stick to their planning assumptions, they will end up in 2005 with 1980s logistics processes and technology - and to remind yourself how out of date that might be, remember that a decade ago people were still arguing about whether those cute little personal computer thingies would catch on. But if they admit that their assumptions may be a temporary expedient, they remove the foundations of their justification for the technology. Oh dear, what's a civil servant to do?

In industry, the word that crops up at about this point in the discussion is re-engineering - which brings up the second of the key questions underlying Whitehall's technological debate. How is the civil service to be managed?

Re-engineering is a concept with two underlying meanings. It accepts that the point of technology is dramatic change: "Don't automate, obliterate" was the title of one of the first articles on the subject (by Michael Hammer in the Harvard Business Review). Also, most practitioners of re-engineering argue that the best way to manage amid the radical change which technology brings, is to set clear goals, and then to devolve responsibility for meeting them to the lowest possible point in the organisa- tional hierarchy. That way benefits come not from a single heroic act of planning but from a continual, opportunistic evolution towards accepted goals.

But the civil service has prided itself traditionally on being without goals of its own, and thus amenable to the direction of whatever government happens to be in power. It begs the question of whether the civil service traditions of caution, responsibility and accountability can be reconciled to re-engineering's underlying credo of empowerment and entrepreneurialism.

To the government's credit, the questions are at least now being asked. But the task of answering them - and doing something about the answers - is a huge one, both qualitatively and quantitatively. This is not a small operation: the civil service is the nation's largest employer, with 571,000 people working for it, and administrative costs of £20 billion a year. It recruits some of the brightest minds in the country, but trains them to focus more on rhetoric than results. After all, the highest calling for a civil servant is still to develop a complete range of policy options for ministers, not to take a great idea and run with it.

Add to this traditional rivalries between civil service departments, and you start to understand the special bond of shared suffering and frustration among those who have tried to reform the civil service. It is easier to get rid of a Tory minister - even to get rid of a whole Tory government - than to get rid of the organisational barriers which civil servants have created to protect themselves from each other. Organisational barriers are precisely the sort of thing which traditional civil servants delight in creating and defending. As one insider put it: "There's a lot of talk about inter-departmental co-operation. In reality there is nothing that gives us more pleasure than telling colleagues from other departments to sod off because it's none of their business."

DSS disaster

The failure to reform social security benefits illustrates how far territorialism can go. There are now over 60 different sorts of benefit, each with different rules for entitlement. It is highly unlikely that any civil servant (let alone any claimant) fully understands the complex interactions between them. All staff can do is key claimants' details into the mainframe systems and see what comes out.

Recently reformers tried to simplify unemployment benefits by creating a new job seekers' allowance to replace with a single benefit two previous ones: unemployment benefit (a flat amount, evaluated by the Employment Service, and paid by it through its nationwide network of Job Centres) and income support (a means-tested benefit evaluated by the DSS, and paid by the Employment Service Job Centres on the DSS's behalf). In the ensuing turf battle between the Employment Department and the DSS, proposed simplification became obfuscation - a case study of re-engineering gone wrong. No amount of automation, e-mail or artificial intelligence could transform the dog's dinner of confused responsibilities and unnecessary complexity that resulted into a workable process.

So ministers and senior civil servants are looking for more powerful levers to break down the barriers to change.

The first idea, after some basic improvements in financial management, was to organise bits of government that provide a particular service more clearly under a single manager. In 1988 Margaret Thatcher's government launched a Next Steps programme to do just that. Many areas have since been considered for abolition, privatisation or contracting out. Or they have been transformed into more business-like "executive agencies", under a chief executive reporting directly to a Minister.

For a few, freedom from bureaucratic constraints would be enough to improve services. For instance, the crown publisher, HMSO, quickly took the opportunity given to it by a reorganisation which placed it under its own chief executive, to launch an electronic publishing division, and began using EDI (electronic data interchange) to take orders and to pay bills. Companies House now delivers company data under its own management; the Driver & Vehicle Licensing Agency (DVLA) in Swansea has similar authority to process vehicle and driver licences; and the Ordnance Survey is digitising its maps of Britain for a new all-singing, all-dancing database from which it hopes to make money on its own account.

To give extra impetus to the wheels of change, John Major's government in 1990 introduced market testing - an ersatz competition between government departments and private firms. Bureaucrats are being asked to define what services they need - both for themselves and for the citizens they serve - and then to ask if a private firm could do the job more efficiently.

As a result some computer functions have been contracted out - staff and all - to outside companies. Computer processing of income tax claims went to EDS (although special arrangements were made to keep the tax affairs of the Royal Family, MPs and the secret services in-house). Computing for the Northern Ireland civil service went to ICL's facilities management operation which is known by its initials, CFM; the Home Office's bureau in Bootle to Sema; the MoD's Devizes bureau to Hoskyns. In other cases the in-house team just got better at its job, such as the Employment Department and the Lord Chancellor's Department. It's hard on the staff, who don't know who they're working for from one contract to the next - or indeed if they are working. But the taxpayer does save money - at least over the short term.

Over the long term, two problems threaten to diminish the effectiveness of market testing: if key staff are contracted out along with the services they provide, the civil service loses much of the expertise needed to manage the work. If they are not, an independent contractor may lack the required expertise. Even where outsourcing does work, contracts can create barriers to further change. Jobs are divvied up among different contractors, tasks are specified in legal detail and change starts to involve lawyers as well as managers.

Fragmentation is also a worry with the government's next big idea: the private-finance initiative, in which companies are invited to pay for the installation of the systems government needs and then to make a profit as they are used. Already deals are on offer to install new systems for National Insurance and for Post Office counters. To help prevent private finance and contracting-out from fragmenting civil-service management into privatised fiefdoms every bit as uncommunicative, unco-operative and downright wasteful as anything the civil service could come up with by itself, the government's IT visionaries have come up with a brave new idea.

If only each citizen could be uniquely identified then technology could provide better services more cheaply and flexibly. Tasks could easily be transferred between one department and another, and departments could co-operate more effectively in providing service. Fraud could be eliminated. After all, why have four databases when one will do?

At least that's the theory. In practice, such an administrative simplification raises the thorniest of the three questions underlying the government's information technology debate: who owns and controls information? The much-debated national identity card is only the most visible aspect of this question. The card offers a trade-off between convenience and control. Everybody could save time now spent filling out forms that tell one branch of government what another already knows - or worse, correcting the problems caused when one bureaucrat fails to communicate with another.

But better communication is a double-edged sword. Do you really want the Inland Revenue to know about your unpaid parking fines? Your television licence? Where should Britain draw the line between state efficiency, effective justice and high levels of state service on the one hand - and privacy and liberty on the other? Should government information be used to develop Singapore-like national levels of efficiency and justice? Or is ignorant government better government?

The next stage of debate over these fundamental questions awaits a report on the prospects for smart ID cards from a special inter-departmental committee (called GEN34), which is due in a few month's time. Given the opposition to identity cards, the debate could drag on for years. But there is more to the question of ownership and control of information than personal control over personal information. For whose benefit is information held: government, the state, the department or agency that collates it, or the individual? And if all of the above, how can all of these conflicting interests be balanced?

Meanwhile, the government is quietly making a number of decisions that put ownership of, and benefits from, information into the hands of individual departments. The justification is that, as information is a valuable resource, departments should sell it in order to defray administrative costs that would otherwise be borne by the taxpayer.

Departments and agencies that publish information are, in their increasingly businesslike way, trying to earn higher fees through sale of information. While Newt Gingrich in the United States shovelled government information on to the Internet, the prelapsarian Robert Hughes pointed out that the Central Office of Information was "a bit irritated" when the budget of November 1994 was put on the Internet. "They sell this information," he admitted, "and that is part of their budget strategy."

To make free, or even cheap, the information from the Meteorological Office, the Natural History Museum, or even from Hansard, the record of parliamentary democracy in action, would, Hughes argues, merely subsidise those who make professional use of the information at the expense of taxpayers.

Hansard, which used to cost 12p an issue in 1970, now costs £7.50, over 60 times as much. To make transcripts of debates inexpensively available over the Internet, he adds, would subsidise journalists and lobbyists; to make available legal information would subsidise legal publishers. But to hike the price of government information would encourage precisely the sort of turf-battling, data-hoarding, citizen-spurning behaviour that reforming civil servants are trying to get rid of in the first place. Not to mention making a nonsense of the idea of electronic democracy - a fully (and freely) informed electorate participating in all sorts of government decisions now denied them for lack of information.

For the time being, the first hesitant moonwalks across the lunar lanscape of digital debate are being left to a few brave members of the e-mail account holder/government official demographic - as I discovered the first time I posted to Usenet a year ago. I had been trying to work out the government's phone bill, and investigated whether a parliamentary question would help. Never having done it before, I asked a parliamentary consultant who said such a question "would cost a bit ... weekend for two in Paris, that sort of thing." This seemed odd, and the news thread "MPs for hire" on alt.uk.politics seemed the right place to do my first posting, asking whether the same suggestion had been made to anyone else. This drew an indignant response by e-mail from David Shaw MP (Dover & Deal, whose e-mail carries the slogan 'Why use the Channel Tunnel when you can use Dover's ferries?') who said the practice was quite improper. If there was hard evidence it should be drawn to the attention of the Speaker, but in his experience such allegations were insubstantiable and anyway MPs were too busy to take weekends in Paris.

True, the conventional media brought the story to a climax with the resignation of two MPs over allegations of cash for questions and weekends in Paris. But the Internet had the story first. Perhaps we'd better get on with this electronic democracy before the government sets up another working party.

William Heath (william@kable.demon.co.uk) is a partner in Kable, a firm specialising in public-sector IT research and publications.