The Public Administration Select Committee's publication of its "rip-off"-Britain report on IT produced a piece...
of theatre that will give the government's IT strategy a helpful shove on its way.
Bernard Jenkin MP, chairman of the committee, stuck his neck out on publication of findings from the MPs' IT inquiry, titled "Government and IT - A Recipe for Rip-Offs". He issued a statement implying the government IT oligopoly is a cartel, a word that quickly echoed around the media. He used it to make an urgent call for an independent investigation to stand up the accusation.
Jenkin said witnesses to the committee had called the current domination by major suppliers a cartel. In fact, there was one witness who called it a cartel - Martin Rice, CEO of the now-defunct SME Erudine. Robert Halfon, MP for Harlow, asked Rice when giving evidence to the committee on 15 March if there was an "artificial" cartel. Rice didn't like to say. Jenkin stepped in. Is there a cartel, he pressed. Yes, said Rice.
The MPs were only saying in public what people have said in private about the potential for collusion in government IT for years. The concentration of IT suppliers has meanwhile coincided with repeated public IT failures and wasted billions of pounds, a fact that is uncontested. The Public Administration Select Committee (PASC) statement has forced the question to be taken seriously, while its report has challenged government to investigate the charge.
The government IT strategy rests on the premise that there is an IT oligopoly that must be broken. This is true if 80% of government IT is indeed delivered by a small number of suppliers- and both the question and the number rest less on hard data than circumstantial evidence. The question for any investigation then becomes whether there is, to cite the textbook definition, a collusive or non-collusive oligopoly.
Only a formal agreement to collude would justify the accusation that the major IT suppliers are running a cartel - and clearly that is not the case. Even the committee's own analysis implied there may be no realistic proposition of such an accusation being stood up at all.
Its central premise was that the point cannot be proved because government benchmarking of IT spending is atrocious and everything is anyway in the hands of those suppliers accused of ripping off taxpayers by over-charging for work that isn't benchmarked. Cabinet Office minister Francis Maude "shocked" the committee by admitting the new government had to ask suppliers how much money they were earning from its contracts.
Hence, said the committee report, the information on IT spending should be gathered. Whitehall has said this repeatedly since the Office of Government Commerce first promised to do so in 1999. It would be an uncharacteristically guileless oligopoly that didn't clean up its act if ever government showed any real sign of getting a grip. Equally, wild accusations of collusion can be thrown around without any fear of their being disproved.
Large suppliers contacted by Computer Weekly had almost nothing to say about the PASC findings at all. Accenture, BT, Capita and Serco said the report was not applicable to them. Atos Origin refused to comment. IBM and CSC made no response. Fujitsu and Logica made vague statements about the good work they did for government. Only Capgemini said it would talk.
HP, the UK's largest supplier of government IT, said trade association Intellect would speak on its and other suppliers' behalf. Cynics might highlight the irony of saying, "We aren't in collusion, but we'll ask one body to talk for all of us."
Intellect said the talk of a cartel was "outrageous" and made a statement of support for any investigation into such allegations - which it believes would anyway be "a waste of public money". Only Northgate welcomed the report unreservedly, insofar as it presented an opportunity to promote its latest marketing message: "shared value outsourcing".
Putting these statements aside - and passing over the almost symbiotic relationship large departments had with large suppliers in days when large projects were the mode - the track record of public IT disasters is long enough to justify the PASC report's assertion that the entrenched IT establishment is no longer tenable. It ventured instead an elaborate vindication of government policy - so persuasively that no supplier would challenge it in public debate.
It is said the reason government has, despite 10 years or more of good intentions, never got a grip of its IT spend, is because departments, agencies and authorities never wanted to co-operate with a central initiative. Their data about their deals was their business.
The PASC analysis noted government has outsourced thinking about IT along with the business of IT. How can government now cut costs if it doesn't know what it's spending in the first place, the report asked.
The MPs' proposed solution would side-step inter-agency and public-private tensions with the simple requirement that there be full transparency on contracts and a presumption built into public systems for the routine publication of all data. Systems would also be built with an expectation of inter-operation.
These principles would provide the means to open public systems up to all-comers, creating an environment where private companies - in particular innovative SMEs - might more readily provide public services by plugging their systems in, and government systems might easily be attached in turn to other people's services.
The proposal then becomes a vision of universal empowerment that works if you believe that civil servants, corporations, SMEs, charities and everyday people all pull their sleeves up and muck in to the digital common good. What critics fear is that it would really mean an unsightly splurge of personal data. You can't simply open up the provision of nursing home applications by allowing anyone to grab a home's data, residents and all, and turn it to whatever innovative means suit their own ends - that is unless you engage an almighty bureaucracy to pick over what data and what derived data in what systems could and couldn't be opened up; or unless you don't open the systems at all, in which case you are back to lumbering, restrictive outsourcing contracts.
This horror story makes the most "ambitious" part of the PASC proposal the most important. This is the idea that people be given control over their own personal data, which they store with digital agents that converse with public and private systems on their behalf. As well as keeping personal data safely out of the way of the turmoil promised by the Big Society mash-up, it would provide a conduit for their collaboration in the reforms. More unhappy nursing residents might, for example, allow the infrequency of their bed baths to become public knowledge.
The trouble for government is these reforms can't happen without legal reform, just as the agile, iterative systems development methods that are widely believed to be part of the way to avoid expensive IT disasters cannot fully be introduced without legal reform, especially in EU purchasing rules. The old ways have been preserved in data protection and procurement law. The PASC wants the Cabinet Office to get its skates on and, for example, inject agile methods into reforms of European procurement rules it's negotiating in Brussels.
The committee also wants government departments to take control of affairs again after a decade of often knee-jerk outsourcing - to become the intelligent customer. Departments should in other words orchestrate their IT affairs more competently, especially as public services begin to allow genuine public participation in cyberspace. PASC said it was "urgent" that government regained IT skills it lost when it sold its IT off to the oligopoly.
The report was ultimately an attempt to prise public IT projects from the grasp of a misguided perception, in which they are indistinguishable from public infrastructure projects like bridges, and in which visionary megalomaniacs think they are creating a page in history to be indexed alongside Brunel.
It is dawning on the industry instead that since IT systems are an attempt to codify a complex, rapidly changing world, and their stakeholders are as innumerable as the differences in any single citizen's circumstances, then the hierarchical methods and the monolithic systems they produce are wholly inappropriate. The great project is not a single bridge but a network of bridges all using the same standards of interoperation so that one route really can feed into the other.
So the PASC report proposes government as "an open...platform around which others can innovate and improve", which means an open system "built on the principles of open data, open standards and open source".
The fact that the oligopolists don't make public calls in favour of this is, for many critics, reason enough for government to investigate their business practices. The proprietary systems they have typically built suit their commercial interests.
Yet the PASC pulled its punches over the commercial implications of its reforms in a way that questions its bold talk of further investigations.
"Intellectual property rights" over government systems should be "opened up to competitors", it said in a curiously worded rendition of the idea that the public should own public code - in other words, that systems should be open sourced. But on the other hand, it said, "legacy systems" could simply be replaced with new ones that didn't cost so much to maintain. Thus, it was implied, perhaps open source wouldn't be necessary after all.
The government has historically allowed suppliers to retain ownership of public software. If PASC now wants the public to own public code, then the government cannot permit that software - the codification of the public good - to be withheld from public scrutiny and open collaboration, while simultaneously calling for transparency on data, contracts and as much other public information "as possible". Not unless the whole scheme is less about universal empowerment than privatisation. Then it would make complete sense to open one but not the other.