(No.189) Open Government Data for citizens, by citizens, about citizens?

Go to Report
Status: 
Accepted
Workshop Theme: 
Emerging Issues
Concise Description of Workshop: 

Increasingly, governments around the world are developing open government data policies that promise to bring a new era of government transparency coupled with a wave of economic development based on the exploitation of these digital resources. This is linked to the rise of "big data". However, questions remain as to the where and how value is created. Much of open government data has been traditionally associated with national registers, maps, weather, etc -- but current initiatives are increasingly looking at data around citizens and public services. There has been some discussion about privacy risks, but not much on the fundamental relations of beween citizens, government and private companies. Besides privacy we will look at wider issues: whether governments have the right to create "value" out of citizens data, how should this be governed, whether citizens should be expected to provide data in exchange for public services. We will also look at the impact of core reference data, for example, should governments provide truly free geolocation services and mapping for their open data initiatives, or leave these to private companies that will provide "free" services paid for with citizens data. We will also look at how these open data policies in many cases sit alongside repressive policies that allow for surveillance of citizens' Internet use, and whether these place limitations on the promises of open data. As a wider issue we will look at the apparent lack of connectedness in these public initiatives, and how the apparently disparate aims of providing openness, security and economic development could be integrated into a comprehensive public interest led Public Data Policy.

Organiser(s) Name: 

Jim Killock (Open Rights Group)
Desiree Miloshevic (Affilias)
India/Chennai Chapter of Internet Society

Previous Workshop(s): 

Yes

Submitted Workshop Panelists: 

CONFIRMED SPEAKERS AWAITING PANELLIST PROCESSING
Desiree Miloshevic (Afilias) 
Javier Ruiz (Open Rights Group, UK) 
Al Kags (Open Institute, Africa)
Lennart Huizing (Green Party, Netherlands)
Dominique Lazanski (Open Data User Group, Cabinet Office Ministry, UK)
Andrea Beccali (IFIL)

Name of Remote Moderator(s): 
Jim Killock
Gender Report Card
Please estimate the overall number of women participants present at the session: 
There were very few women participants
To what extent did the session discuss gender equality and/or women's empowerment?: 
It was not seen as related to the session theme and was not raised
Report
Reported by: 
Javier Ruiz
A brief substantive summary and the main issues that were raised: 

 
Javier Ruiz
 
Our starting point is the restrictive Public Sector Information reuse framework in Europe, while in the US we have simply a lack of intellectual property restrictions in federal information. The US has seen  huge growth and benefits and Europe wants to follow this course promoting open data.
 
 Open data takes things one step further with proactive release and removal of barriers to reuse. This means free data, using open technical formats that enable replication, open licensing, etc. 
This means not having to sign a contract with Government and explain how you are going to use the data, what you are going to do or not do and pay a small amount.  
Of course open data is important in itself for many people, but also it is seen as a building block for what is called Open Government, which is the idea that you can actually transform the fundamental relationship between citizens and the state through information and engagement. 
We can identify several types of data, each with its own issues around governance and potential implications.
Core reference data covers things like maps, weather, registers, etc.  These datasets normally have high economic value, although in some countries may not exist at all.  The question is should the Government be providing this mapping for free? If you want to map public toilets - which is one of the things that has been done in many countries for people with disabilities - you need a map.  You can use Google but you need to give up your own data. We think this data should be opened.
Functional data is where a lot of the direct impact can come from open data. Things like the micro-statistics that government departments use daily, such as levels of pollution, etc.  All that type of really useful bits of information, all that can go out and be put too very good use.  Here the main thing is how do you govern the processes and make sure this data gets out. 
Public services data, here we have two types of data: on the one hand you have performance data which is what many politicians will talk about. For example the mortality rates of a particular doctor or department.  Should you be able to choose which doctor you use on the basis of previous experience of patients.  Of course we believe that, but we think is important is to distinguish this type of accountability from the political accountability of the elected representatives.  It is all very good to say this doctor is responsible but who is responsible for creating the policy framework that put that doctor there.? 
The other question with public services - one of the biggest issues in terms of open data - is the personal data of public service users.  In the UK there are plans to share medical records, welfare data of many types with private companies in order to improve services. But that data is actually private data from citizens. We don't want to stifle innovation but we really have to be careful about this type of data.  
Then something that is important to see as well is public cultural information, this is something that is increasingly relevant.  All the text mining of public libraries can be really useful not just for Google to develop translation services, but for everyone else in society.  There are lots of issues around whether it is public domain or belongs to the state or libraries. 
Public accountability information, such as Government meetings is again public data and we also think that should be part of the framework. 
 
Al Kags
I will focus my conversation on what the experience has been in Kenya.  The Open Institute is an African organisation working with governments in Africa and the developing countries to promote the opening up of data by Government to promote also the participation of citizens.  
The first point I wanted to talk about was the question of motivation.  In Kenya  and in a lot of other countries while governments say that they would like to be as transparent as possible it is not a very popular thing, and the leaders are afraid of what that might mean.  In Kenya we took the view that that instead of pushing the question of accountability and transparency we would instead push the question of prosperity. Thus, open data will give employment, will help young people who are developers to develop obligations, help young entrepreneurs to find solutions that will better their lives and that sort of thing.  Which in Kenya turned out to be a very popular argument for the leadership and which we are finding also among other governments in the region to be a popular way into going about it.  
Once you have got governments to agree to publish the data in principle then you go to the technocrats and start grappling with the hows.  In Kenya we take fairly utilitarian approach and I have become a champion of the utilitarian approach where you deliver in bits.  You take what you can get now and you start with that and then you build on it as you go.  I have seen a number of open initiatives that are in danger of being still born because they try and get everything perfect. 
You can try and ensure that the platforms that you use are open, that the licensing regime is in place and you have a certain number of important data sets in a certain kind of level of quality.  In Kenya we went and looked for whatever was readily available.
We found health data and education data and within eight weeks we launched an open data portal and then we continue pushing more data on to it.  The reason that this is important is because when you do these things incrementally you also show the technocrats who sometimes are also afraid of what open data means and what this openness means and who have grown up, many have been working in the civil service for a long time and who have come to believe that governments are supposed to hold data secrets, it is not supposed to be published.  
You have to demonstrate that there is no threat to publishing the data, and really really hope that the media does not immediately find a scandal.  
 The third aspect of it is once you have published the data you have to really focus on building the ecosystem of data producers, the Government civil society, etc.  Media has access to a lot of data that we have accumulated over many years to ensure that they begin to publish that data and the academia who also have over the years gathered data from different sources and they also publish it.  
The second echo system is of infomediaries.  Quite a lot of the data is not in a format that allows the citizen to engage because you are providing the data in its rawest possible format.  The best way to get the citizen to have access and have an understanding of that data is by dealing with the info media, which tends to be the civil society organisation that works with the citizens that build the capacity with community groups and so on.  The media that are going to tell interesting data stories out of it, and then the developers who build apps that enable the citizen to then interact with it. 
Once you have done this, the innovation is ongoing, the users are working, their media study stories and that sort of thing, then we look at the legal frameworks, which are benchmarked globally but I champion that they must be focused on the local situation. 
In Kenya we have the constitution, that guarantees the citizens a right to information.  Number two, we have a cabinet paper that we did at the time of the launch that tells the public sector how to publish data and that they need to look for the data to publish it on a regular basis.  
We are working on a freedom of information bill and one of the most popular aspects of the freedom of information bill in my view that is currently under review for Parliament is that it proposes Government proactivity in publishing data.  
 
Andrea Beccali
Al talked about ecosystems and intermediaries where you mentioned the media, which has an important job, but I want to also include libraries.  Just to give you an idea from a recent survey there are around 320,000 libraries worldwide, and 73 per cent of these libraries are in developing and transition countries. We think they can play an extremely important role in accessing data.  
When we looked at the Open Government Partnership and we look at the numbers of countries that actually in their plans spoke about the ecosystem; only three countries, including the UK, Ukraine and Tanzania have conceded how a citizen can access those data. Only Ukraine mentioned libraries as an important partner in this.  But we think that libraries are often perceived as a building where you find books and that is true, it is still like that but actually there is much more potential inside them and we think that in the open data they can play an extremely important role, particularly when you look at cost effectiveness of providing access points to the citizens and also when you ensure that everybody has access to these resources. 
 
In Romania the Minitry of Agriculture produces an online application to make subsidy quicker for rural areas and they release the data for the rural areas, all the land that wasn't used and they wanted to use.  17,000 farmers were reached through libraries and got subsidised land and made the whole project work. 
 There is another important role in libraries and data mining.  It may be trivial but libraries have been collecting data all the time.  Data is about who comes to the library to read what and to do what basically and user is data is shared among libraries but is not always used by Government. 
Another important aspect is the role of Parliamentary libraries. All Parliaments worldwide have a library that stores all proceedings and documents and laws and draft laws and bills for the activity of the Parliament. They all have this information and could be put as open data.
 
Lennart Huizing
 
The Dutch Government has for the last couple of years being looking at development aid and actually the new Government that has just been installed this week has announced another cut back of 20 per cent of the budget for development aid.  
One of the reasons for that is I think that the lack of accountability, the lack of clear results that have led to some reevaluation of what the Dutch Government would want to get out there in terms of information for the public, more accountability and transparency.
Two years ago the Green Party in the Dutch Parliament had a motion to release all the data generated on projects funded by the Dutch Government and one year ago they went ahead and released that data.  I think that it is quite a limited set of data in the sense that it is very interesting to see what which company, which organisation gets what amount of money. 
 It is not particularly interesting for the general public but it is very useful for experts to figure out what works and what does not work.  It also has the side effect of being giving some level of accountability for foreign governments spending.  Finally we get to see how much money is actually going and into other countries but there is no clear information about the results that have been generated with it, so it is also very limited.
A year ago I wrote an app by request of one of the development aid organisations in the Netherlands using this information to try to get it to the general public.  You can find more value by combining the data sources and we have now included a lot of data from the World Bank.  
What we found is that the main problem in this data is two‑fold. 
Somebody needs to clean up this data because it is completely unusable for anybody outside of the Government involved, with heavy use of jargon. I come from a back ground in market research where we generally see the same types of questions and of course there we are working to give the information in a very useful way to our clients.  But still you know if you have been in market research for a while you know that you can torture data to make it say almost anything.  
I think that this is a very huge risk in releasing this data.  It is not even, you can't even use it to triangulate on a specific person.  But you can use it to triangulate on a certain politically sensitive project and foreign governments will have the opportunity maybe to locate projects that are not agreeable to them, and I think we have seen here in Azerbaijan, for instance what that might mean.  I am not saying ‑‑ whatever.  Let us not go there.
 One last point would be that open standards should be prevailing in terms of how do you release this data because if you don't use open standards, and I know there has been a movement where people wanted to be technologically neutral, the way that the data is being released which might be meaning that it should be used by any close standard as well which might not be what we want. 
 
Dominique Lazinski
I sit on the Open Data User Group in the UK, which is a volunteer body that sits within the Cabinet Office.  There are 14 of us and from Government and civil society which is my remit and a variety of other big businesses, small businesses, businesses collecting data, businesses start up… so it is quite interesting.
Our role is advising Government on what data should be open and I was particularly interested to hear Al because we have what I think is a rather cumbersome process of people that need to submit requests to data.gov.uk, which is our data portal in the UK.   A number of data sets have been opened already I think at the last count there was over 8,000.  There is a work stream within Government to release more data but we are not proactive, the Government is not proactively releasing data, which is why I think Kenya is well ahead of us on that front.
The UK Government has particular needs for having data to be open, with the remit of economic growth and development, so they are really interested in prosperity but they really want to make sure that they can also look at the value that society will get.  Understanding the economic growth that the UK will get, based on what may be open and projecting this, is a really hard thing to do.  
We are now looking at trading funds which hold a lot of the core reference data mentioned earlier, including the postal address file and mapping data and I hope we can talk a little bit more about that.  
There is a higher strategy  that is being done also within the Data Strategy Board, which looks at sort of the general policy around opening data, going forward an what that means.  I am liaising with civil society and I have been working with big and small charities as well as local groups who may deal with issues around housing or child abuse or anything like that.  I am there to champion all those causes and it is very hard to capture absolutely everybody in civil society in England and Wales.

Conclusions and further comments: 

 
Javier Ruiz stated that although we tend to talk about public sector data, increasingly there is of course private sector data.  We need to start asking questions what the public interest lies, whether it is just with the state, not just in the provision of public services by private companies, but also things like private mapping and other infrastructure.  Infrastructure and transport are two of the areas where the role of private companies would need to be questioned in the next few years.  
Javier also raised that the benefits that they want to achieve in the UK from transparency are improved governance, better public services and economic development, and we think that you can integrate all these aspects although in many cases governments tend to focus on one or the other. We need an open data policy that sits alongside a wider data policy framework that is consistent with data protection. 
Al Kags however disagreed with the view that when you focus on development you might not focus or you might not address yourself to accountability.  I think the whole question is so long as you get open accountability happens as a default, which is why we don't talk about accountability, we talk it down because of the fact that it ends up being a default byproduct.  
Andrea disagreed: I think that you assume that once you put data outside then people will go there and look.  I don't think it is a direct link.  There are many people that have no clue about than and maybe they heard about the media but they don't have the interest and so it is what I think that here libraries can provide this part it actually can make this link work perfectly and I have an example that I think can demonstrate that.  
Al also said that there has been a lot of pressure by civil society to get the Kenyan Government to publish its data.  There has also been a lot of pressure by civil society and citizens to get private sector to publish the data.  I think one sector that is yet to really get open is a civil society itself. 
Lennart reacted to this proposal for civil society to open up: We have now in the Netherlands eight organisations that have released the same type of data in the same format, open standard and that will really help us to get further into the data. So we can show people what is the aid doing on a very low level, increasing of course the risk of politically sensitive projects.  We now have the data being published once every quarter.  I think that is not enough.  We need to go beyond that and try to have as much realtime information that we can get our hands on as long as we keep in mind all the risks involved, that you have mentioned.  And we have to consider the risks involved with releasing data that maybe is not personal, that might have been anonymised or might have been aggregated to a higher level but it might still be used to triangulate on any one person.