The Open Government Data Initiative: Risks and Rewards

Mel Shefford

The opportunities and risks involved in exploiting consumers’ personal data are the subject of much coverage at the moment. But what about the commercial potential of the vast data sets being made available by the public sector? Datonomy shares some observations.

Some of the Datonomy crew attended a Westminster eForum conference earlier in the month which had as its theme “policy priorities for user data”. Datonomy was particularly interested to hear one of the speakers, Professor Nigel Shadbolt, talk about the open government data initiative. The initiative is supported by a new independent organisation which is funded by the Technology Strategy Board, the Open Data Institute, which the Professor chairs with Sir Tim Berners-Lee as president.

The initiative – which involves publishing anonymised government data – was announced in November 2011 by George Osbourne and can be seen as part of a general drive for transparency in government which in recent years has seen the public procurement process laid bare and the passing of freedom of information legislation.

So far, nearly 9000 data sets have been published on the website. This Datonomist discovered all sorts of information there, ranging from the location of bus stops in Sunderland and alcohol-related statistics, to the intriguing “no crime” data set which shows what percentage, by police force, of reported offences were actually recorded as offences by police.

So why do we care about bus stops in Sunderland?

Open government data is a good idea for several reasons. Firstly, it encourages government accountability (for example, in relation to crime rates or public spending). Obviously the government isn’t always keen on releasing some kinds of information, but Francis Maude, the Cabinet Minister spearheading the initiative, has challenged government departments to be as transparent as possible, saying “I don’t have any doubt that giving our Press a lot of data to pore over will at times be uncomfortable for us in Government. But that’s the whole point. A closed door culture encourages complacency at best and at worst corruption.”

It’s not just about accountability though: analysis of multiple data sets can lead to improved procedures and decision making, and in turn, reduced costs – just last week, Margaret Hodge claimed that the government could save £33 billion by making better use of big data. For example, the NHS could analyse data which shows which hospitals are fighting an infection outbreak well and how they are doing so as a way of trying to control the infection. On a more global scale, data analysis can really yield results: earlier this month, Sir Tim Berners-Lee quoted an example of how data on the price and availability of medicines revealed that some governments were being charged up to 25% more for the same medicines, and this enabled some of the governments to pressurize pharmaceutical companies to reduce their prices.

Government data can also be used to help empower individuals by allowing them easy access to all kinds of data (such as in relation to local healthcare facilities and crime statistics). This ties in nicely with the government’s “midata” initiative, which is designed to allow consumers to gain access to data that is held in relation to them by businesses (such as on their spending or usage patterns).

In addition, it is suggested that the initiative can help drive economic growth because a significant amount of the data is licensed on an open government licence and can therefore be used by third parties (for potentially commercial applications). At a simple level, the bus stop data in Sunderland could be used in a third party app (which could involve advertising) to show the good folk of Sunderland their nearest bus stop. There are also more sophisticated examples, such as the Mapumental website which allows a user to input the postcode of their work location and how long they are prepared to commute. The website then uses government travel data to show the user where they could live. It can also interact with property rental and sale websites to show the user where they could afford to live based on their desired commute time (with somewhat depressing results for this Datonomist!).

So what are the risks in using government data?

It’s worth noting that government data is not risk free. As with any data, the data sets could be inaccurate, incomplete or unreliable. Analysis and extraction may also be difficult if non-transparent file formats are used, such as PDFs.

However, the more significant risk – from Datonomy’s perspective – is that government data could be used to obtain personal data if it is not properly anonymised. For example, a crime map which shows only one reported crime in a street could be used to provide information on a known individual. The Information Commissioner is very much alive to this concern, having published guidance on crime mapping in February 2012 and a new code of practice on anonymisation earlier this week.

However, whilst these risks should be taken seriously, they should not detract from the real value that can be obtained from using government data. In the initiative’s infancy, there are already many interesting examples of how it is being used by the public sector, individuals and businesses, and Datonomy looks forward to it being used in increasingly innovative and commercial ways.


Leave a Reply

Your email address will not be published. Required fields are marked *