Post: 4 December 2014
There’s a good article by Marc Ambasna-Jones published on the City Metric site today, mainly about city-based open data stores:
What has open council data ever done for us?
The article includes some comments from me, as well as from Tom Cheesewright and Richard Speigal. I’ve been quoted fairly and in particular I’m glad the bit about Ordnance Survey derived data was included. (Though I did wince at being called a data “evangelist”. Not a label anyone serious about data should cultivate, IMO.)
Below are the full comments I gave to Marc.
Q. Are taxpayers getting value for money from open data at the moment?
Quantifying the ROI for civic open data is inherently difficult. Open datasets are usually released under licences that don’t require users to apply or ask for permission. That means it’s impossible to accurately track the full extent of use and re-use. Publishers can count downloads, but that fails to capture onward sharing of data within organisations and through supply chains – one of the main mechanisms through which open data can support economic efficiency. We tend to focus on anecdotal examples of open data use, typically in apps developed for consumer markets, but that isn’t the whole picture by any means.
It’s also difficult to distinguish between local authority investment in datasets that they need to produce to support their own public-facing functions, and release of open data. The simplest scenario for civic open data involves a council making available a dataset that they already use internally. In that scenario the investment required should be minimal.
The real investment around civic open data is for development of information portals and APIs. But even there, the business case is usually based on development of new systems to support the council’s internal functions and enhance delivery of their services. The open data aspect, i.e. encouraging businesses and the public to find additional uses for the data – is secondary. I don’t see many local authorities investing significantly in open data for the sake of it. Where civic open data is taking off it’s usually because councils have realised that they can “build in” open data as part of their wider investment in delivery of services online.
Personally I’m comfortable that taxpayers are getting value for money from open data, even if the evidence base is a bit amorphous. It’s hard to isolate the effects of open data on growth and efficiency within a city economy, but that’s equally true of many other policies and inputs. The extent of investment is quite small compared to other types of infrastructure investment that we take for granted, without much scrutiny, as beneficial.
Q. What needs to be done to improve the ROI of local investment in open data sets? More private data? More collaboration on app development?
Even if we cannot accurately measure ROI of local investment in open data, there are several things that I think could be done to improve that ROI:
1. DCLG should reconsider which datasets are prioritised for released under the mandatory local government “transparency code”. At the moment this list is too focused on data that is useful for holding local authorities to account but has little economic potential for re-use. For example I would like to see more emphasis on release of planning records and data from Local Plans.
2. Local authorities should work more closely to agree schemas for open datasets, so that there is consistency in formats and structure. This will make it easier to construct national datasets from local datasets, and encourage the development of apps that can take datasets from different cities.
3. The area in which open data has most economic potential is location intelligence (addressing, geolocation, maps and so on). Local authorities have numerous datasets of this type but are unable to release them as open data because they contain information derived from Ordnance Survey’s detailed mapping and address datasets. We need Government to release those key national datasets as open data so that cities can in turn release the local datasets that derive from them.
4. As you suggest, there are datasets held by private firms (especially utilities) that would stimulate the information economy if released as open data. The problem is how to incentivise those releases. I’ve looked at this issue in a recent article. Government already prevails on many industries to submit data for regulatory and statistical purposes. My own view is that much of that underlying non-personal data should be released for public re-use. This might involve eroding the commercial confidentiality of some large companies, but would greatly benefit SMEs and consumers. (Unfortunately at the moment the direction of travel is the other way, with government agencies reducing the amount of information they collect from industry.)