September 07, 2011 | Javier Ruiz
How problematic is the Public Data Corporation?
Jonathan Raper from open data company Placr
argued that the PDC is a massive step backwards for the direction we thought data policy was moving under this government (US style marginal cost + tax funding model). This is going to stop on its tracks opening up other key core data areas, such as Companies House. It is not just the impact of the datasets involved in the current deal for the PDC but the general policy u-turn.
argued that it is not that bad, and that a PDC selling high value data (he gave as example OS Mastermap) will not have a real impact on citizens and the general open data transparency drive. He was also very sceptical of the economic possibilities of open data, saying it is overinflated and that most apps challenges produce unsustainable rubbish that VCs would not touch with a bargepole (he has a point though, as apparently only one of the original apps for democracy is still running). He saw no option but to keep some trading fund structure to finance data production.
One added issue is that this has to be read in conjunction with the Making Open Data real consultation, which establishes that open data is now mainly about the performance of public services and enabling choice, as part of the Open Public Services plan (a.k.a. "any qualified provider"). This means that citizens outcomes for health and education will be dependent on the information they receive, thus making data interpretation critical. The question is how can open performance data be useful without a surrounding ecosystem of open data for mapping, core reference, etc.
Maybe this is not going to be such a problem. We would love to hear people's opinions, particularly if you have any examples relating to Land Registry.
Our preliminary view is that that the PDC consultation presents a very narrow set of options that do not fundamentally alter the Trading Fund model.
The option of the PDC as a Data Utility which would provide raw data for free paid by taxes and without any upstream involvement is ditched in one short paragraph as unsustainable and carrying the risk that the market would not provide key value-added data. Maybe they are right, as it would cost too much for government to buy back refined data from the private sector, but we have not seen the evidence behind this decision. There is no consideration of the possibility of providing raw data for free and refined for a price.
Michael Cross in The Telegraph
believes this to be the intention of the Fremium model explored, but I cannot find anything in the document implying that raw data will be given away for free. Public task data would at best be harmonised, and not public task (refined/upstream) data will be sold with less constrains and some improvements in licensing, but more or less as it works now. The Freemium model proposed for the PDC seems less about the above raw/refined traditional PSI split and more about business use cases.
The option of profit maximisation is there but quickly abandoned as not aligned with policies, as almost a sign of goodwill, but they fail to mention that this is not a real option within the EU Public Sector Information reform climate, at least not with public task data.
The biggest problem is that the consultation fails to ask the basic question of what pricing model we want. The responses to the EU
consultation for the review of PSI regulations were overwhelmingly in favour of free (and even more supporting free for non-commercial use, which is not the same as Freemium). The chart below is a simplified summary of the responses from the table in page 35 of the report.
We have not found many people enthusiastic about the proposals in the PDC paper, even in the PSI business community who seem to be the main targets for the pricing and licensing reforms proposed. We are also getting reports of widespread concern about the PDC being a packaging exercise for privatisation, although there is no hard evidence as far as we know.
Open Rights Group have organised a public policy event on the 3rd of October to discuss both consultations. If we think the PDC is really a bad idea we should coordinate some form of response now, or put up with the results afterwards.
The policy event is in the afternoon, before the OKFN London open data gathering, so you can stay for both!