James McMurray's Blog

Rust, Linux and other curiosities

Why governments should adopt and invest in FOSS

This post covers why I believe local and national governments should adopt and invest in Free and Open Source Software (FOSS).

This has been in the news recently due to the city of Munich renewing it's LiMux Linux distribution thanks to an agreement between the local SDP and Green politicians, and the efforts of the Public Money, Public Code campaign.

Opinions expressed are solely my own and do not express the views or opinions of my employer.


LiMux is a project by the German city of Munich to migrate its desktop infrastructure to FOSS solutions, specifically to the LiMux Linux distribution (based on Ubuntu) and LibreOffice (initially to OpenOffice). The project was started in 2004, and migration began in 2006.

By the end of 2013, 15,000 desktops had been migrated successfully. However, in September 2016 Microsoft announced the opening of a new headquarters in Munich, and then in 2017 the city council decided to switch back to Microsoft Windows and Office. The timing of this decision led to many accusations of effective bribery, using the potential office opening to pressure the council in to using Microsoft products (note that there is no evidence of a kick-back scheme like practiced by Microsoft Hungary).

Fortunately this decision has now been reversed, and LiMux development and roll-out will continue.

Other cities

Munich is not the only city choosing Free Software solutions, many others are beginning to see the potential benefits too. In Catalonia, Barcelona's Digital Transformation programme supports Free Software. From first-hand experience this means that the government transport / road organisation use LibreOffice for all documents - so local driving schools, etc. do not need a copy of Microsoft Office to handle the necessary digital paperwork.

In Valencia, the local government migrated its administration and schools to LibreOffice saving 1.5 million euros per year, and saving students and other end-users from having to purchase their own copies of Microsoft Office.

However, it is not always straightforward. Munich faced a lot of push-back in some areas, and the city of Barcelona still pays almost 5 million euros in license costs to Microsoft.

The main issue here is not so much the direct cost to the city itself (which is substantial, but not critical compared to other costs - especially since any migration will incur a significant short-term cost), but more that the cost must be paid over and over again by all users separately. Barcelona continues to pay millions to Microsoft every year, but so will other cities in Spain and across Europe, and then so will all users of those services that may have to use a copy of Microsoft Office for compatibility reasons (i.e. students and small businesses).

The same cost is paid over and over again to a foreign company that will not hire local developers or invest in the local economy, and uses vendor lock-in to eventually control all of the critical infrastructure.

Benefits of FOSS

The most commonly cited benefit of Free Software is certainly the cost savings vs. proprietary licensing. Whilst this can be a benefit in some circumstances, I think it overlooks many greater far-reaching benefits, and may not actually manifest itself in the short-term due to the additional costs of the actual migration (i.e. costs of deployment and training).

Shared investment

The main benefit of using FOSS solutions is that any investment in the development and improvement of the software also benefits all other users of the software.

That is, whilst the city of Munich might need some improvements in handling templates, other cities might need other improvements - however they can all benefit from each others' investment, with no need to pay again just to keep a license.

This helps to build up a commonwealth of high-quality, well-maintained software which anyone can use. There is no central, unique owner of the projects (any project could be forked if necessary), so there is no company that can demand payments just to allow you to keep using the existing software.

Any investment in development made is truly an investment, leading directly to project improvements, rather than just money sent to a foreign company whilst the customers have no ownership of the actual product, allowing the company to charge even more in the future just for the right to keep using it (usually this is done by giving discounts for the early migration which then result in higher prices later once the customer is locked in).

Note that this initial investment might be higher than the current costs of existing proprietary software licensing, when considering the migration and training costs and possible development costs. However, unlike those licensing costs it is a one-time cost, and the customer retains some ownership over the resulting product (i.e. there is not a license that can be revoked).

Freedom for development

There is also more freedom in the development process. The fact that the products aren't owned by a company means that contributions can be made by anyone, anywhere.

Users of the software (such as the cities mentioned above) could fund local co-operatives of developers to add desired features and maintain the projects, instead of being forced to fund a foreign corporation with no local investment.

In the long-term this might also help to break up some of the monopolistic Big Tech companies, and result in a freer society and business environment for everyone.

National security

In Europe especially, the current dependence on proprietary software often means a dependence on foreign corporations which operate in co-operation with their country's intelligence service. As revealed by Edward Snowden, Microsoft provided backdoor access to encrypted messages to the NSA, CIA and FBI as part of the PRISM programme.

It is indefensible that such a company can run the critical infrastructure of the vast majority of local and central government administrations, as well as personal computers. Especially when the NSA has been proven to intercept communications of supposedly allied nations.

With FOSS software there is no central company to be pressured by intelligence services or build in backdoors. All of the code can be scrutinised by developers and users. Linus's Law states:

Given enough eyeballs, all bugs are shallow.

The same concept applies to introducing backdoors to access private data. It might be possible to pressure or bribe one company to do so, but it isn't possible to do the same to tens of thousands of independent users and developers.

Bigger picture

The effects of co-ordinated FOSS adoption are much more far-reaching and positive than solely saving the costs of current software licenses.

The adoption of FOSS and investment by many governments (at all levels) would help to create a commonwealth of quality Free Software for all citizens and government bodies to benefit from.

It would also allow development of these projects to be done from anywhere, allowing users to fund local developers or co-operatives. This would help to break up the massive, monopolistic Big Tech companies, and could also be used as a way of bringing investment and jobs to parts of the country/region that require it.

This would also be highly beneficial to developers, as they would have more control over their own work (as they are not tied to the few employers that own these popular products), and would also likely be able to negotiate a larger portion of the compensation directly since there would not be the overhead of salespeople and lawyers, etc. present in current large software companies.

All Free Software users would benefit from the greater usage and investment - GNU/Linux users could expect better hardware support for example, as it becomes more commonplace.

How to get there

Whilst the benefits of adopting FOSS are clear, the cases in Munich and Hungary show that it will not be an easy path to greater adoption. We cannot just rely on central government eventually carry out the adoption as the few representatives are easily pressured by the beneficiaries of the status quo, and are serving a much larger political platform.

Ultimately, I think the solution lies in applying "Linus's Law" to politics itself. It is no coincidence that the successes in FOSS adoption so far have been concentrated in local government - in specific city councils and regional administrations, because it is easier for engaged groups of citizens to have a direct effect in local government.

At a national level, it might be possible for a large software company to pressure a few political leaders and policy makers, however, it is much harder to do the same to a whole board of a dozen or more local councillors (especially across the multitude of different counties and regional administrations). There is "safety in numbers" as it becomes unfeasible to pressure and manipulate large groups of politically-engaged citizens.

It it is the responsibility of every citizen in a democratic society not only to inform themselves and vote in national elections, but also to really partake in the political system: by joining a political party, taking part in local and regional elections, and ensuring that the democratic standards are upheld both inside the party and in political institutions.

This should have support from across the political spectrum, as guaranteeing the security and privacy of citizens' data whilst also cutting out the middlemen of massive foreign corporations, shouldn't be a controversial policy.

Efforts in local politics have already proven successful, such as the LiMux project and others mentioned previously. So if you agree with the points raised here I hope you will join whichever local political party you agree most with, and help to bring about FOSS adoption by your government.


Greater FOSS adoption by our governments benefits the administrations themselves by being free from repeated software licensing costs.

It would benefit us as citizens by helping to ensure that our data is kept secure and private from foreign intelligence services. It would also benefit us as FOSS users by providing a greater user base, ultimately leading to better hardware and software support.

Finally, it would benefit us as developers by providing greater freedom in the possibilities of employment (since many different employers could work on contracts for FOSS improvement), leading to a better geographical distribution of opportunities and not being subject to bad corporate practices such as non-compete agreements and anti-poaching agreements between large software companies.

If you agree with the points in this post, please consider: