An Interview With Arjen Kamphuis

In 2002 Arjen Kamphuis co-authored a parliament motion to mandate open standards for all government IT in the Netherlands. The motion was unanimously accepted and, in 2007, became policy. The Netherlands thus became the first western country to make the use of open standards in public sector IT mandatory. Arjen is now working to export this set of policies to other European countries with the help of local political parties and business partners.

Arjen discussed his experience of lobbying for this policy change and some other questions related to his work as a consultant on IT strategy and the implications of nanotechnology and biotechnology in an email interview with the Centre for Internet and Society.

The Centre for Internet and Society: What is the Dutch government's policy on FOSS and Open Standards specifically and intellectual property rights in general? Provide some history, name the main lobbying factions in the Netherlands and their policy positions. What was your role in the formulation of these policies?

Arjen Kamphuis: The national action plan 'The Netherlands in Open Connection' is the government's answer to a unanimous vote in parliament in November 2002. The parliament stated that the market for desktop software was not functioning as it should and that significant vendor lock-in effects were harming both individual citizens and society as a whole. It requested maximum efforts from the government to change this situation. The suggested method for changing was mandating open standards in all public sector IT and actively supporting the adoption of open source software wherever functionally and technically feasible. 

I was one of the people who got this process started by contacting a member of parliament from the Green Party. This was triggered by my inability to access the website of the national railway on 1 January 2002. The website had been redesigned and only allowed access to visiters with Internet Explorer. As a Linux user, I had previously had comparable problems with local government websites and electronic tax forms (usage of which was mandatory for small businesses like my consulting start-up).

After the unanimous vote in parliament, several people in the Dutch open source community, including me, kept the pressure on the government by monitoring major procurements and writing questions for the Members of the European Parliament (MEPs) to ask the government. In 2004 this led to a breakthrough when the Justice Ministry ran a project to procure 147 million euros' worth of desktop software without going through a proper multi-vendor selection process. They only talked to one vendor, and that is against European Union regulations. Since some of the civil servants working on this project were gagged, we can conclude that some people were aware they were breaking the law, yet went ahead anyway.

When the news broke we made sure the MEPs were armed with the proper questions the next day, and the contract was dropped. In reply to questions asked to the government by the MEPs, the responsible ministers admitted that the government was very dependent on Microsoft for basic functioning of its office environments; that this was a problem; and that the government would take active steps to remedy this situation by moving forward with the requests made in 2002 by parliament.

Two-and-a-half years and an election later, a new under-Minister for Economic Affairs, Frank Heemskerk, took up the challenge and promised a comprehensive policy. I gave input for this plan in mid-2007 and it was formally published and adopted later that year as a national policy for all government and public-sector (i.e. tax funded) organisations.

The policy has three objectives:

• improving interoperability between public sector organisations;
• lowering the vendor-dependence of the public sector;
• improving the functioning of the software market and supporting the Dutch knowledge economy 

Some of the practical measures are the mandating of the use of open standards in all public sector organisations. Whenever software is procured, open source should be considered and preferred whenever functionally adequate. These two very basic rules change the entire market for IT in the Dutch public sector (40% of the entire market) and is having a profound effect on the way software vendors offer their products as well as the negotiating power of the client organisations.

I continue to advise both the decision makers and the civil servants overseeing the implementation of the policy.

CIS: What is the current status on the implementation of these policies?

AK: After a slow start the government organisation that is responsable for overseeing the implementation is now up and running. The basic problem is lack of awareness about both the practical value that open standards and open source software can contribute and the underlying political reasons for making it the preferred option for government information processing.

Thus a lot of the work for the next few years will be communicating these ideas to civil servants (be they IT professionals or managers who have other jobs). The policy helps a lot because it puts some serious weight behind the whole process. The fact that government organisations have to support Open Document Format for instance significantly heightens their interest in the technical subject matter!

So the policy gives the drive needed to get things moving and now it is up to us to communicate the how and the why in a way that is understandable for people who are new to these concepts.  I have no doubt it will be a long process, we have over 20 years of proprietary legacy built up in our public institutions. Replacing those systems with open alternatives will take many years. All the greater a reason to proceed with some urgency. 

The complete policy document has been translated into English and released under Creative Commons Licence:
http://appz.ez.nl/publicaties/pdfs/07ET15.pdf

In December 2007 I gave a talk in Berlin. Here a summary, slides and video are available:
http://events.ccc.de/congress/2007/Fahrplan/events/2387.en.html 

CIS: What can a country like India learn from the Dutch government's experience in eGovernance and ICT in Education?

AK: I am not familiar with the Indian political process but these are some of my lessons learned: 

- The government will not do anything unless constant and significant pressure is applied by citizens. Politicians and civil servants only act if the pain of acting is less than the pain of not acting. Change is achieved by citizens standing up and working on these problems without guarantee of any reward or even achieving any results (it took us five years to get from a unanimous vote in parliament to an actual policy). 

- Big IT companies may be your friend or your enemy. But even if they are your friends they generally will not be at the forefront of political action that could be seen as controversial. Once policies are pushed beyond the controversial stage and have been adopted as official policy some of them will support it. Others, with much to lose, will fight you and the policy every step of the way. The more money or loss of market share is involved the more radical the methods that are employed. Massive lobbying, applying political pressure through foreign governments, bribery and all kinds of other activities are well-funded, well organised and very common. 

- In moving forward with these policies it's the lack of knowledge and vision with the the management of institutions that is by far the biggest bottleneck. Without a clear policy from the top it is impossible to get things moving in most organisations.

- Another big problem in switching over local governments and other smaller organisations is the fact that many of the advantages of such a switch is national and/or macro-economic in nature while the initial cost and risk is micro-economic in nature. Hence again the need for a national policy. 

- The funding required to make significant improvements is often not that large compared to the existing operational budgets. Investing in the smart use of IT in education for instance is something that can pay for itself very quickly. This is generally also true for adoption of open source and open standards in general. By just reducing the yearly spend on software licences by 1% the entire government program can be funded.

- Simply stopping the procurement of new licences (while continuing the use of those already paid for) can often free up enough money to finance a migration process. This has been the case in the city of Amsterdam and the French Gendarmes.

- The actual value of better government services or education is hard to quantify in monetary terms. How do we value improved responsiveness, transparency, national sovereignty in information processing and supporting local service companies instead of foreign software companies?

- IT education should focus on understanding methods and principles, not products. The product life-cycle is 18-36 months, the educational process takes many years and the length of a career is decades. Any education with a focus on products leads to knowledge that is irrelevant by the time the degree is finished. Teach people to drive a car, not just a Volkswagen or Tata.

- The cost of physical books per student per year in the Netherlands is now greater that the cost of a laptop. This is insane since the content of those books is generally written by teachers who get paid very little for it. Using the funds to pay those teachers instad of the publishers and releasing the content under a free licence will free up resources to develop better educational programs and provide all students with computational tools to use them. All without increasing the total cost compared to our current situation. The financial numbers will be different for India but the basic principle is the same and works even better given the larger scale of India. The cost of producing and distributing electronic educational content will drop practically to zero when compared to physical on a per-student basis. Using funds to support teachers in the use of e-learning with open content is the way forward. 

CIS: How can a local support environment for open technologies be created? Can local SMEs ever substitute for the transnational proprietary giants?

AK: Whether SMEs can supplant multinationals depends on the product being replaced. CPU manufacturing requires a very high upfront investment in R&D and manufacturing capability. This is usually far beyond any but a handful of companies. With software development and services things are very different. Software development only requires a human with programming skills, a good idea and a computer. The Free Software Movement has shown clearly that distributed methods of software development can lead to high quality products with excellent local support systems. Local organisations (or communities that are not even organisations) can often understand local needs and respond to local changes much better, faster and cheaper than large, lumbering corporations. If local organisations work together globally to share knowledge (and code) for those parts they all need they can beat any centralised system.

What many senior business and government leaders are struggling with is the realisation that many of the 'truths' they have learned while studying economics or business management or some such subject turn out to be empirically incorrect. For example: it has become clear there is no causal relationship between the cost of software and its quality or utility. This must be a fact that is difficult to truly understand and accept if you have been brought up believing the gospel of the Anglo-Saxon economic worldview. The current economic crisis is a great help in questioning some of those beliefs and opens up room for new ideas about economic vs. societal value of technology and its relationship to businesses trying to earn a living.

CIS: Could you tell us about the Dutch government's rollback on electronic voting machines? What is your opinion on the use of electronic voting machines in the upcoming elections in India?

AK: From the mid '80s onward, voting computers were introduced in the Netherlands. By 2006, the vast majority of all elections were being performed by proprietary computer systems. Citizens would press a button and then go home to watch TV. Some software that no-one could control, monitor or properly audit would spit out a result and that would be it -- new government. Only a handful of engineers (all working for the companies that made the voting computers) actually knew what the software did and could make the computer system say anything they wanted. 

When the city of Amsterdam (the last holdout using paper ballots) announced in 2006 that it was moving to voting computers, a group of activists organised a campaign to ban voting computers. We felt that the very nature of democracy was under attack by running the election process in a way that makes it impossible for ordinary citizens to check the validity of the election. It also makes fraud a lot harder to detect. Detectability of fraud is the one of the primary properties any election process should have. We all know election fraud is also possible with non-electronic means but keeping it a secret is much harder in such cases (as we saw in the US and Zimbabwean election over the last years). There was a actual case of suspected voter fraud in a Dutch municipal election and the judge concluded that while the fraud seemed likely it could not be proven. Regrettably for the suspected council member the fraud could also not be disproven. This shows very clearly that such a method is wholly unsuitable for application in real democratic processes. 

Through lots of media attention, a few spectacular hacks showing the technical insecurity of the systems, and legal pressure, we forced the government in 2007 to reverse the approval of the voting computers and go back to an all-paper balloting system. This reversal is part of a global backlash against electronic voting systems. Comparable changes have been going on in many US states and all over Europe. 

I think India should have voting process that can be understood and monitored by its citizens. This understanding and monitoring should be possible without requiring advanced degrees in computer science, software engineering and electronics. The only way to have such a process is when there is a paper ballot involved. Such a ballot could be printed by a computer to increase the ease of use but all-electronic solutions are ruled out by the basic demands of what a democracy is.

India should move to either all paper systems or voting computer backed-up by a voter-verified paper trail. 

Are more extensive telling of the tale can be found here: http://wijvertrouwenstemcomputersniet.nl/English

This is a link to the Berlin CCC conference of Rop Gongrijp's 2007 presentation (with video): http://events.ccc.de/congress/2007/Fahrplan/events/2342.en.html 

 http://www.blackboxvoting.org has a wealth of information on this subject.

CIS: What are the services provided by Gendo? Could you describe some of the projects that you have undertaken?

AK: My company (gendo.nl) also provides consulting services in the area of IT strategy, development of open IT architectures and implementing those in mixed open source/proprietary environments. We are currently advising both national and local government organisations in the implementation of policies and plans to move to open standards and open source software. We are also involved in projects where we do the actual development and implementation of new systems to enable innovation and lessen the dependance of our client on proprietary systems. Currently we are involved with a healthcare organisation where we are assisting in re-architecting their entire IT environment to allow service innovation, lower cost and increase information security. 

We have also been involved in information security work and other auditing in the financial services and government sector. Here our activities focus on the grey area between technology and process. 

Outside the field of IT we also do other consulting work such as scenario planning and strategic future studies, mostly for large corporate clients. Most of the big Anglo-Dutch multinationals such as Shell or Unilever are on our client list. We also have a large number of clients in the financial services and insurance sector.

For all of these clients we organise presentations and brainstorming sessions, often preceded by research. This helps the leaders in those organisations think about the nature of rapid, technology-driven changes in their markets and the world in general. These insights are then translated into new products, services and ways of delivering them. 

Forgive me if this all sounds a bit vague but with many of these clients there is some confidentiality agreement involved.

CIS: Could you tell us more about yourself? Maybe you would like to share some formative experiences.

AK: Writing my first paper on black holes at age 11 showed me that grown-ups usually also don't know what is going on in the universe either. Despite rumours to the contrary parents, teachers, senior managers and politicians are not all-knowing and are stumbling about just like most two-year-olds where complex issues are concerned.

Over the last quarter century I've had this intuition reconfirmed again and again. In a world that is changing faster and faster experience becomes obsolete rather quickly and wisdom is no longer the sole purview of older, more senior, people. We need young smart-asses who have not yet learned what is impossible, so they go out there and do it.

-----