overseasStudies of other countries show that funding models for government data subtly affect their usefulnessMichael Cross The Guardian, Thursday December 21 2006 Article historySlowly but relentlessly, the question of how government should encourage people to exploit the vast resources of digital information it holds is creeping above the political horizon. Last week, opposition MPs tabled questions to ministers about what action they plan to take following the Office of Fair Trading's criticisms of the way the public agencies behave in the market for data.
Internationally, too, the issue is gaining momentum. The Organisation for Economic Co-operation and Development (OECD) last week agreed that public-sector information should be on the agenda of its next big ministerial conference on the information economy, in Korea in 2008. That may seem a long way off - but officials says that much work needs to be done to understand the issue.
Technology Guardian's Free Our Data campaign proposes a simple solution: government bodies should make available all data owned or funded by taxpayers, subject to the demands of privacy and security, at the marginal cost of dissemination. In the internet age, this is usually zero. There is a stark contrast between the UK's approach, running parts of government as businesses selling data and the US, where federal data is available for free.
Complex picture
However, an international study carried out for the Office of Fair Trading paints a more complex picture. The study is a rare attempt to perform a consistent analysis of how three different countries' agencies work. They are the US, where federal data is almost all free; Sweden, which has a similar position to Britain's, and Australia, which falls between the two. For each country, the study examines the state mapping, meteorological and companies registration agency. These generate the most commercially exploited public-sector data. But the way they run their businesses varies greatly: Australia's meteorological agency makes a profit of US$6.15m (£3.12m) a year, Sweden's mapping agency US$5.3m. Their US equivalents depend on government funding.
Sweden, despite a freedom of information tradition going back to 1766, is no free data paradise. The study says that several agencies are "very active in commercial sales" and that this has caused problems when an agency competes with the private sector. In an echo of the UK position, it finds government policy unclear: there is no overarching approach to the regulation of fees and charges for the commercial use of government data.
As in Britain, the Swedish government does not set a clear boundary between primary, "unrefined", information and "refined" value-added products. The study reports concerns that Sweden's meteorological agency may enjoy "various advantages" over commercial businesses as a result: the state agency supplies 95% of the market in value-added products. Sweden's mapping agency, too, has come under official scrutiny for failing to separate its accounts between unrefined and refined activities. Government inspectors have reported suspicions of cross-subsidisation - a similar charge made against the British equivalent, Ordnance Survey.
In Australia and the US, by contrast, the study finds that governments prevent public agencies from competing directly with the private sector. There has not been a single complaint from a private operator in either country for the past five years.
There is a price to pay, the international study suggests. In the US, it finds that, while the supply of free data from the US Geological Survey has stimulated private investment, the private sector has now overtaken the government agency, whose maps are now used only as a last resort. Part of the problem may be that an agency forced to provide data for free has no incentive to make its data relevant to the market. The study suggests that, in a free-data environment, regular surveys of users may be necessary to ensure that an agency's products are still needed.
Key assumptions
Overall, however, the international study does not come out in favour of free data - though it warns of risks to competition when public agencies behave as businesses. The study's authors, London Economics, say that the "free data" argument depends on two key assumptions. First, that the government is better than the market at deciding what data consumers want and, second, that the extra taxes needed to support "free" data will not distort the economy. "Neither assumption is likely to be true in practice," they say.
We disagree: the data that we want freed was almost all created to serve government policy, not the consumer. As for the second assumption, we argue that the government might actually save money when public bodies no longer have to market and protect their data, solemnly billing each other for its use. What we need is more evidence, and more discussion, domestically and internationally. That, at least, is beginning to happen.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment