Making Open Data Real (1 of 2)
5th September 2011
Governments collect and maintain large datasets. Instances of such Public Sector Information range from geographic data to census records to real time traffic flows. In recent years there has also been a move to provide open and unrestricted access to Public Sector Information by publishing it online as Open Data.
The Government has recently announced its Making Open Data Real consultation on Open Data. All of our main political parties have policies which are broadly in favour of releasing Public Sector Information as Open Data and so it might appear that there isn’t anything contentious in this.
Hee hee.
This, of course, is not the case and to reflect this the consultation contains 26 questions across 7 of its sections. There is a debate about whether open data should all be free to use, or if some of it should be charged for. There is a debate about the formats that we will produce open data in. Then there is the question of what the definition of open government data is and which organisations it applies to. And there are questions about whether we should collect more or less data.
I thought it might be useful to look at a couple of these questions and describe what I see to be the arguments around them. In the next post I’ll look at which organizations should be covered by open data legislation Below I’m going to consider whether should we charge for access to some public datasets.
For
The main argument in favour of charging for some datasets is that there is a cost in collecting them and making them available. If there is a possible commercial product that can be made out of it then it is only fair to the taxpayer that a charge is made to whoever accesses that data.
In the UK this argument has been made most strongly by public sector organisations which are also Trading Funds. These are public bodies which generate over 50% of their income by selling goods and services.
Examples of UK Trading Funds are Ordnance Survey and the Meteorological Office. Trading Funds are a relatively new creation, the first ones being set up in the early 1990s.
In the case of Ordnance Survey a significant amount of the goods and services it sells are those that it sells to other public bodies. Until last year these were managed through individual agreements, but that has been simplified by the introduction of the Public Sector Mapping Agreement which covers central and local government, health and emergency responders.
Last year Ordnance Survey also made a number of its less detailed mapping layers available as open data through its OS OpenData™ product using the new Open Government License. It still generates income from its more detailed products.
Against
In opposition to this it can be argued that as public bodies should only collect data for the public good it should be paid for out of general taxes and published in a way that makes it available to be freely used by anybody.
Also, not all uses of open data will be for commercial gain and making it freely available means anybody can use it. This also stimulates economic activity as more people and organisations will do things with it than if they had to pay to look at it first.
The financial argument is that the tax returns this generates will outweigh any cost recovery that individual Trading Funds may benefit from by charging.
There is also the argument that there is a cost to administering the current agreements and payments which would largely disappear if the data was made available free of charge.
The economic potential of open data is a difficult thing to try and measure. There was an Office of Fair Trading report on the Current Use of Public Information from 2006 which estimated the UK market for Public Sector Information of £500m could be doubled if more liberal open data policies were followed.
There are EU studies which put the value of Public Sector Information to be currently in the tens of billions of Euros, but they vary quite widely in their evaluations.
One thing that is worth noting is that if we change things and require Trading Funds to publish for free the data they currently charge for it would have an uneven effect on government finances.
Put simply, organisations such as Ordnance Survey would lose a lot of money while the economic benefits would come in through general taxation to the Treasury. That would need to be balanced out in some way.
In trying to deliver a clear explanation I’ve simplified the details and the arguments somewhat. Anybody who wants to get deeper into the issues discussed might want to look at the documents listed on this page from Linked Gov or download this PDF of the proceedings of an international open data workshop which requires you to provide an email address.
Next week I’ll post about the questions around which organisations ought to be covered by any open data legislation.
