Microsoft gave British government bodies a whopping 50% discount on corporate prices when in 2004 it negotiated its three-year deal with the Office of Government Commerce (OGC), the UK's public procurement sheriff.
But it forbade the figure from being published. That was how much Microsoft wanted public market share. Bill Gates played a personal part in talks with the then Chancellor Gordon Brown. Gates got a knighthood. Microsoft became the standard for the UK's big e-government push. It suppressed the only viable alternative to its monopoly: free software.
Is it safe to assume then that the deal, and that the new agreement the OGC and Microsoft signed last week, was good for Britain?
Yes, if you take the OGC's word. In September 2006, the OGC issued a rare glimpse of its secretive Microsoft terms when it announced that since the UK had bought its 1.5millionth Microsoft licence, it had broken a threshold that released better prices. The deal would now save the public sector £1m a year, it said.
But the OGC didn't want to talk when in June 2008 Microsoft put prices up 20% because so many public bodies had resigned from the agreement. They had signed up long enough only to upgrade their Microsoft software. Neil Dewar, head of ICT at North Somerset, dropped out of the agreement "two or three years" ago because he "saw no benefit" in it. "We were paying x-per annum and getting nothing for it," he says.
So the new deal includes a penalty for dropping out of the agreement so that it can better perform the trade it was designed to embody: price against customer lock-in. This is crucial for Microsoft now alternatives to its software are viable, as has been demonstrated by open source implementations by Continental organisations such as French Parliament.
The UK is still hefting the weight of its buying power as a monopoly basher. It was only in 1999 that the Gershon Report, on which the OGC was founded, declared "insufficient aggregation" of public purchasing had allowed gas suppliers to charge 140% over corporate prices.
Aggregated buying was a mantra of the late 90s new economy swizzle. The internet made it possible by losing commercial information from its paper moorings. Prices would be exposed in such a world. Monopolies would fall. Competitive markets would thrive. The buyer would be king.
It was in these conditions, and amidst the dotcom bust, when in 2001 Microsoft introduced new, world-wide licence terms that were controversially to raise public sector prices, even as the OGC was first applying its aggregated power to Microsoft.
Barely had the OGC opened for business in 2000 than it had it secured £38m aggregated savings from Vodafone. Hotel fees were cut similarly in 2001. This "demonstrates what can be achieved," said the OGC at the time. But hotels and mobiles were both competitive markets. Microsoft was fighting a US government edict to break up its monopoly.
The big lock-in debate
Critics of Microsoft's 2001 license terms said they would lock customers into buying upgrades they didn't need. Customers branded this a "Windows tax". UK government departments told the National Audit Office they would increase costs and pressure to upgrade.
When the OGC applied its aggregated buying terms to Microsoft in February 2002, it claimed it would save £100m over three years; except it wouldn't, as the NAO revealed in 2003. £74m of the savings were attributed to "efficiency and productivity savings" that could arguably be found using any software.
They were 'Alice in Wonderland' numbers, says Eddie Bleasdale, an open source consultant who has competed fiercely with Microsoft's public privilege. The OGC had estimated that the public cost of Microsoft's 2001 licensing change would be between £40m and £60m. The actual cost saving of the 2002 deal was estimated at between £31m and £36m.
But we can't be sure. The OGC begged the savings estimate from Microsoft because the public sector didn't collect the information. The NAO recommended something be done about it, otherwise the UK wouldn't know if its software agreements were any good. Sir Peter Gershon had declared this very same as a mission of the OGC when he launched it in 1999. A "complete absence" of common measures made it impossible to assess UK procurements, said his report.
Ten years later, on 21 April 2009, the Treasury's Operational Efficiency Programme report condemned the public sector again for having little clue about its IT spend. "This makes it difficult to establish trends, make comparisons and manage down costs. What is not measured well, will not be managed well," it said.
Yet the OEP felt justified in endorsing the OGC's collaborative IT agreements, suggesting more of them could save the public sector £6.1bn. It didn't say where it got the number from. But it did make the first public call in this saga for "much needed transparency".
Collecting and then benchmarking IT spending data would give the UK confidence in its procurements by making sure they were "good value for money" and got "independent scrutiny and cross-questioning", said the OEP. Newham Council, which became Microsoft's flagship and the OGC's template when, in 2004, it signed its own 10-year agreement with Microsoft, had then also promised benchmarks to assess its own and other public IT spending. By the time it came to renegotiate its deal last year as part of wider OGC renegotiations, it had still not developed the measures.
How could Newham negotiate its new Microsoft terms confidently without benchmarks? We don't know because it refused to answer questions about the 2004 deal. Microsoft and the OGC were also tight-lipped. But in an extraordinary turn of events, transparency has been written into the new deal. Corporate lawyers are no longer imposing secrecy on the public purse.
In this spirit, the OGC has already confirmed the talks stalled over the usual price/lock-in trade-off. The OGC wanted bigger discounts. Microsoft wanted guaranteed buying volumes. The OGC got the same discount it agreed in 2002 - a base-level 40%. Microsoft made another 10 points conditional on enough public bodies signing up to the agreement before its year-end.
Standardisation and cost savings
After seven years of aggregation, the basic trade-off remains the same. Public bodies dropped out of the agreement last year for the same reason they complained about Microsoft's 2001 terms: because they didn't want to pay for upgrades they didn't need; yet the UK has agreed terms that give greater rewards to those who continue paying Microsoft's annual fee under the OGC agreement.
The government has operated on the assumption that this situation is favourable. Newham justified its 2004 agreement on the basis that it would be "standardised" on Microsoft, which would make everything simpler and therefore, in the short run, cheaper. Likewise, the OEP called last month for "greater standardisation and simplification of IT systems, desktops, infrastructure and applications across the public sector". It did not specify how this would be done, but it has mostly already happened: councils say they cannot seriously consider open source alternatives to Microsoft because any other systems they buy are either Microsoft also, or designed to integrate with Microsoft software.
"Every business application you buy integrates with Word and Excel. They don't integrate with Open Office yet. It's very difficult to pull away from Microsoft Office," says David Tidey, chair of the BCS Elite Group and assistant chief executive at the Royal Borough of Kensington and Chelsea.
This is evidence of the way Microsoft's desktop monopoly perpetuates itself, but also how it may be broken. Newham delayed its planned rollout of Microsoft Vista in 2007 because many software suppliers hadn't made their software compatible with it. It had to wait before it could make the upgrade it was paying Microsoft £470,830-a-year to get.
Though the deal retains without bettering the UK's price for Microsoft software, it has made some gains that reflect Microsoft's loosening grip on power. When, in February, the Cabinet Office's Open Source Action Plan called for the re-use of software licenses, it nudged Microsoft into agreement with the OGC on that very issue and cleared a bottleneck in the talks. Yet as open standards are the crucial way in which the plan will tackle the proprietary lock-in that sustains Microsoft's monopoly, it is time to reassess the old trade-off and put the Treasury's priorities in context. What exactly did the OEP mean by standardisation?