Free Our Data: the blog

A Guardian Technology campaign for free public access to data about the UK and its citizens


Why privatising Ordnance Survey (and other trading funds) would be the worst possible outcome (updated)

A story in today’s Sunday Times, ahead of Monday’s Pre-Budget Report, suggests that Ordnance Survey, the Forestry Commission, Land Registry and some other trading funds will be privatised:

A string of state-owned household names including the Met Office, mapmaker Ordnance Survey and the Forestry Commission, are being prepared for sale by the government in the next two years to raise cash for the stretched public purse.

Alistair Darling, the chancellor, is thought to have drawn up a list of 10 companies to offload, including the Queen Elizabeth II Conference Centre in Westminster. He will outline the programme in the prebudget report tomorrow alongside details of a Whitehall efficiency drive.

I’m inclined not to believe this report – sunday newspaper journalism is not always what it appears to be; the phrase “managing expectations” has to be borne in mind (talk up a scary prospect; something less scary but still surprising then seems acceptable). But let’s treat it as correct on its face.

The Free Our Data campaign has always opposed privatisation of any of the government’s data collection agencies. We have said consistently that we admire the OS’s ability as a data collection agency; our argument is with its fiscal model.

If we begin with the Ordnance Survey, here are the reasons why a privatised OS would be far worse than the current one.

  1. as a commercial enterprise, it would not have the commitment to map all of the UK that the national mapping agency must have. If that commitment is forced on it by its articles of association, it would be less competitive than rivals, meaning it would be less competitive and so be vulnerable to takeover or worse
  2. valuing the asset would be almost impossible: what’s the value of the MasterMap and its geographic data? OS has consistently refused to put a price on this*, which has led to its accounts not being accepted by the National Audit Office. What you can’t value, you can’t sell
  3. OS presently updates its database with data provided for free by local authorities and others. This is causing some friction. If OS were a private company, it would have to pay a “market rate” for that data. The arguments about what that rates was would make everything that’s going on now look like a tea party
  4. OS does important defence-related work; are you going to give that to a private company whose shareholders you don’t know about (even if government keeps a golden share)?
  5. privatisations of government properties don’t get good value: the instance of DERA (now Qinetiq) was found to be exceptionally bad value to government, though those who completed it did OK
  6. any asset sale now will not realise the sort of value that might expected: this is a terrible climate in which to try to float a company
  7. financial advisers will tell you anything at present to try to get a selloff – but that doesn’t mean they’re right: look how well they did with mortgage securitisation
  8. most of all, OS doesn’t cost the Treasury anything. It generates a 5% return on its revenues – which is better than you’d get from a bank.

Similar arguments apply to all the other trading funds, at a rough guess. Even the Forestry Commission probably has a valid reason for being part of government.

We’ll wait to see what’s really in the Pre-Budget Report, though.

Update: the PBR gives absolutely no indication of privatisation – quite the opposite, in fact.

* from the latest accounts for 2007/8:

The geographic data (‘the data’ – referred to as the National Geographic Database in previous financial statements) is the term used to describe the suite of geographic datasets that Ordnance Survey collects, develops and maintains to represent as digital and paper products which generate revenue.

The data is an internally generated intangible asset per Financial Reporting Standard 10 and as such can only be capitalised where there is a readily ascertainable market value evidenced by an active market for similar assets. Since the data is unique and has never changed ownership, we consider that no market value can be attached.

(back to text)

    The following posts may be related...(the database guesses):