Free Our Bills! is a campaign led by a cheeky platypus, just escaping from the portcullis of Parliament. Sign up now, or read on…

Sometimes data being free isn’t good enough - it needs to be released in a properly structured format. If you want to reproduce the text of Bills (proposed new laws in the UK), you can get a reasonably good click-use license and go for it.

However, the PDF or HTML you get is not very intelligible to machines. For example, consider the current version of the controversial Human Fertilisation and Embryology Bill. It contains lots of amendments of an earlier 1990 act, and it is very hard to follow without being able to see how those amendments alter the earlier act. As the data isn’t structured, nobody can easily make a user interface to do this. If the Bill was published in a 21 century way, then lots of people could and would do so. This is just one example - there are lots of other ways the data for bills and amendments could be better structured, and more timely.

It’s an esoteric campaign, but a very important one. Having good quality law is vital to all of us. So please do sign up now, and help get Parliament to publish Bills better!

Share This

Recommendation 9. By Budget 2008, government should commission and publish an independent review of the costs and benefits of the current trading fund charging model for the re-use of public sector information, including the role of the five largest trading funds, the balance of direct versus downstream economic revenue, and the impact on the quality of public sector information.

Today that is published on the BERR website as Models of Public Sector Information Provision via Trading Funds by Professor David Newbery, Professor Lionel Bently and Rufus Pollock, all of Cambridge University.

If you read any of it, please post tidbits and thoughts in the comments below!

Share This

But look at our Government at home. It’s still bureaucratic, still top-down and still old-world. It still thinks it knows best and that it should keep all the information.

Then a reminder of a Bill the Conservatives tried to pass last year about open Government financial information:

That’s why we have introduced a House of Commons bill that will require the government to publish - online and accessible to all - every single item of expenditure over £25,000. It already happens in the US. They call it “Googling Your Tax Dollars”.

And then the new bit - a very general promise about local Government information:

We will require local authorities to publish this information - about the services they provide, council meetings and how councillors vote - online and in a standardised format.

Nothing about the licensing to be used, and short on detail about the scope of the information (will it include bus timetables?). But that’s not surprising in a high level speech - now’s the time to push the Tories to use a proper Open Knowledge Definition conformant license, and to include a good scope of information.

And now’s the time to push anybody you know in the Labour party, so we can get the data now, rather than having to wait until at least 2010…

Share This

At mySociety, we recently released made some travel time maps which help you work out where you should live that would have the quickest commute to your place work. Interactive flash sliders to balance this delicate equation against house prices would definitely be a simple, tangible benefit for many in the UK.

But for some reason, public transport data is closed. So we can’t make this available for anyone, whatever the postcode of their place of work.

The data itself is totally locked up. Tim Howgego gives a detailed reckoning of the status of timetable data for everything from local buses to long distance trains. Both TransportDirect and Transport for London have beautifully integrated together all forms of public transport to help individuals find individual routes from A to B. They are, however, impenetrable consortiums with impossibly hard to negotiate, non-existent licensing terms, even if you have the cash.

Meanwhile, Google are negotiating access to the transit information for their own website, so that they can make money from adverts relating to it. But the data itself stays closed (Tim Howgego again). Tim’s article gives lots of details about this.

The mystery is why the public transport companies and local authorities don’t make the data free. They would not only enable new innovative services like the mySociety house prices / where to live one, but also make more money as more people would know when and how to use their buses and trains.

The answer? It’s totally anodyne. “Local public transport has little motivation to sell itself. And when it does, the decision making process is cluttered by many organisations with slightly different aims.” (Tim Howgego again).

How do we fix that!

I’ll be talking more about the travel time maps, and mySociety’s upcoming Freedom of Information site, at the next Cambridge Open Knowledge Meetup on Wednesday 13th Feburary. Rufus will be talking about CKAN and open knowledge packages, and Jim giving an update on open access in Chemistry. Remember if you’re in London, Cambridge is just 45 minutes from Kings Cross, nearer than some parts of London.

Share This

Last Thursday, Google, along with an incredible list of partners including MySpace, LinkedIn and Bebo, launched an open API for social network applications. It looks great, basically doing what Facebook applications do, but in an open manner. There has been security problems, but I’m hopeful that they are teething trouble, and Google will make sure they are sorted out. And maybe the API will expand into doing more - such as data portability between sites, enabling backup of social networking data, and even providing a standard online identity system. Tim O’Reilly wrote a visionary post on how this should be going.

Meanwhile, today Facebook launched their new advertising platform. This is exactly what will happen as long as corporations continue to control your data and you don’t - they will exploit it. It looks like Facebook’s adverts are fairly nice and benign, but right now we have no way of guaranteeing that. It would be more reassuring if they had proper competition, that you could move to another social networking site and keep your data and friends as easily as you can change telephone provider and keep your old number.

Last month, Mark Zuckerberg, founder and CEO of Facebook, said they were going to make members’ data portable. Can we trust him? Yesterday, somebody made a Facebook group you can join to put a little bit of pressure on Mark.

Pass it on.

Share This

Just as the Open Knowledge Definition protects our public knowledge, we also need something to protect our private data. Here’s what the Gnome people are saying about the idea of an Open Service Definition:

We need a Free Services License, Open Service Definition, Free Terms of Service, or whatever we want to call it. I see more and more people talking about this, even aside from the GNOME Online Desktop conversation. Topics to cover in an Open Service Definition might include ability to export your personal data, your right to own your data’s copyright, etc. There may also be a requirement to use an Affero GPL type of license. This is very open-ended and unclear at the moment.

To me the reason open source works is that multiple parties with competing interests can collaborate on the software. What would make multiple parties interested in collaborating on a service? Probably a fairly radical-sounding set of requirements. But the GPL was pretty radical-sounding too, many years ago.

(from Havoc Pennington’s blog)

And perhaps, in addition to protecting our private data, the same definition could reward services which use and promote open knowledge. Something using data from OpenStreetMap should perhaps be more likely to be branded an Open Service than something which uses Google Maps.

Share This

The obvious question is, what will the Government do with it at all - it is quite clearly a research paper, rather than Government policy. And even if the recommendations are carried out, is the report asking for enough? Would even carrying out all the recommendations it makes gain the biggest value for society from public sector data?

One of the recommendations in the report is “To encourage innovation in the re-use of information by non-commercial users, UK trading funds should, in consultation with OPSI, examine the introduction of non-commercial re-use licences, along the lines of those pioneered by the BBC’s Backstage project and Google Maps” - an interesting take, given the arguments about non-commercial Creative Commons licenses which OKF people have had recently.

More juicy goodness for you to pick over and over-analyze, this way… Read the original PDF, or scribble on the margins of Sam Smith’s annotatable HTML version. See also the Cabinet Office press release.

Leave a comment below if you find something interesting.

Share This

Summary and full report here.

Share This

(You might think this is something that WIPO might have needed to think about before but, as Ben Ivans the National Association of Broadcasters lobbyist pointed out, WIPO’s never been that hot on gathering evidence for any of the treaties it considers — Ivans point was why start now.)

Preamble

Due to limitations of time we will not respond in full to every one of the items listed in the consultation document but will instead provide ‘general’ comments. Furthermore we will focus on the methodological issues and will not deal with substantive issues (questions 19 and following) as we feel unable to do them justice in the time available.

Use Economics

An impact assessment must be able to assimilate the various different effects of policy in different areas into a single value. Without this it is impossible to explicitly compare competing policies in a transparent way and thereby decided which which is preferable. For this purpose economics provides utility and its associated aggregate: social welfare, both often presented in money metric terms.

Thus our main recommendation is that impact assessment for policies related to IP should be carried out using the framework centred on ’social welfare’ which has been developed by economists. The ‘economic’ toolbox is well suited to a holistic analysis which includes on an equal footing all three of the areas, ‘economic’, ’social’ and ‘cultural’ mentioned in the questionnaire.

Aside: in the questionnaire ‘economic’ seems to equal ‘commercial’, that is relating to the income derived from those activities which involve explicit financial transactions executed by profit-seeking agents. Economists as a rule almost never concern themselves with this alone — even when focused on goods which are traded in markets they would measure utility and not just income. Furthermore economists from the very earliest stages of the profession have concerned themselves the valuation and impact of activities for which there are no associated market transactions — for example the provision of public goods and the effect of externalities such as pollution. Of course, it is true that it may be very difficult — or even impossible — to value some things, for example freedom of speech, within this kind of analysis (though those kind of items are difficult to include within any form of policy evaluation).

Examples of Impact Assessments Using this Approach

Traffic Safety

To give an example of how this works in a very simple way consider the situation from the health and safety area related to the installation of traffic calming measures. Suppose that the installation of a particular scheme will cost 1 million pounds and that it is estimated will save an estimated 2 people’s lives over the life-time of the scheme. Using a willingness-to-pay (WTP) approach (a proxy for utility) we would then estimate a value for benefit of saving a life. For example using Highways economic note No. 1: 2004 published by the Department for Transport we would obtain a figure of approximately 1.3 million pounds as the ‘cost’ of a fatality. Comparing two times 1.3 million (2.6 million) to 1 million it is immediately obvious that the scheme yields a net benefit.

Copyright Term

The second example is of copyright term. The trade-off involved in setting the length of copyright is that involved in copyright more generally: there are benefits in the form of the welfare derived from the new works created as a result of longer term but costs because of restricted access to and reuse of works in general (copying without permission is prohibitied and prices rise as a consequence). The basic jobs of an impact assessment here would be to work out the value of works (or, more specifically, the demand system for the set of works as a whole). Using this it would then be estimate the costs and benefits and finally a comparison to see which of the two was the larger.

Summary

An impact assessment in relation to IP policies should:

1. Be firmly grounded in economic theory: a sound understanding of the economic theory, particularly in relation to ‘knowledge’ goods is essential do performing a full and proper assessment.

2. Examine both costs and benefits: costs and benefits should be examined together. For example, with copyright, deadweight losses should be estimated as well as the gains from the production of new works.

3. Have a common yardstick: Every effort should be made to incorporate the impact in diverse areas using one common yardstick, such as money.

4. Be transparent as to assumptions: make clear what assumptions are made and their impact on the results wherever possible.

5. Take account of the wider picture: (general versus partial equilibrium) where possible the assessment should take account of the effects of the policy outside of the immediate groups (industries, individuals) affected. To take the simplest example: a subsidy to industry A will likely increase employment and output in industry A but in doing so may displace employment and output from industry B.

6. Be evidence based: wherever possible evidence should be provided for any valuations obtained. In particular every effort should be made to obtain ‘hard’ empirical data.

7. Open to review: every effort should be made to ensure that all material used in the impact assessment, particularly data, is available for peer-review by other scholars and policy-makers.

Share This

It’s little known, but we also make all that open data available in a separate screen scraping project called parlparse. Rufus first got in contact with me years ago because he was excited by that possibility, of really useful new data being made available. Then he went on to set up KnowledgeForge which is where parlparse is now hosted. Thanks Rufus!

Unfortunately we’ve had relatively little take up. Not many people have made use of the data. To use a bunch of XML files you still need quite a lot of skills, which most people don’t have. So this time, we’re trying something different. Earlier this year mySociety won a grant from the Department for Constitutional Affairs (yay for the DCA!) to build an Application Programming Interface (API) for TheyWorkForYou. It’s now available, have a look at the TheyWorkForYou API page.

Now, API and XML are both three letter acronyms, and it’s not completely clear that one is easier than the other. But one of them is easier - you not only need code to use XML files, you also need somewhere to store data yourself, index it and so on (a non trivial job, for gigabytes of data). To use a rich API, you can just write code. Even just in-browser Javascript, as the TheyWorkForYou API can return Javascript data structures.

Amongst other things, the new TheyWorkForYou API lets you search Hansard, convert a postcode into a constituency and hence an MP, and even extract user contributed data from comments (shhh! don’t tell me, as I don’t like tagging, but you could always add comments with code words and use them as tags for speeches).

I’m hoping some people will use this to make focussed campaigns websites. For example, a site could take your postcode, look up your MP using the API and give you advice on what to write to your MP on a particular issue according to their party. I’m sure you can come up with lots of other ideas. Meanwhile though, content yourself with a fun but silly example. There are more examples at the bottom of the API front page.

The bad side of an API, as Rufus explained earlier this month, is that you don’t get hold of the actual data. In this case, it is all available as data from the parlparse project, apart from the postcode database (which we don’t have the license to republish). And that’s important. Having access to a free API is no security for the future, whereas having the data itself is.

P.S. We have a mySociety event this Friday in Liverpool. Come along, we pay for your transport (via DCA) if you are in the North West.

Share This