Consultation response on the EMBAG

  • Advocacy

Dear Federal Councillor,
Dear Sir or Madam,

You have invited us to comment in the consultation on the Federal Act on the Use of Electronic Means to Fulfil Official Tasks (EMBAG). We are pleased to take up this opportunity.

Since 2011, Opendata.ch has been working to ensure that more people have access to more data and can use it to generate more knowledge, progress and value. Organised as a non-profit association, Opendata.ch has for years been networking and organising around data policy and data technology issues and, among other things, runs interdisciplinary innovation formats such as hackathons.
Promoting an open data culture at all levels of government is a particular concern of the association, including in its long-standing collaboration with the Confederation.

Opendata.ch welcomes the general thrust of the Act. However, we believe that, particularly in the rapidly changing field of digitalisation, a number of developments already under way must be taken into account even more strongly and more quickly. This is necessary to keep the authorities adequately able to act in future and to take account of the evolving needs of citizens, business and civil society.

In the present version, the articles on open data and open source lack the necessary breadth of impact, the necessary binding force and the indispensable instruments to effectively initiate decisive steps towards successful digitalisation and to measure them continuously. We sorely miss an article on interfaces, whose importance is constantly increasing. Equally sorely, we miss in the preliminary draft the possibility for the authorities to collaborate not only with companies but also with organisations of digital civil society, not least the associations in the field of open source and open data.

Enclosed you will find our detailed proposals for concrete improvements to the preliminary draft. We are confident that you will be able to take them into account in the further process. We can only support the current preliminary draft on this basis, i.e. with more targeted, more binding wording on the topics mentioned and greater consideration of civil society. We thank you for the attention you give to our remarks and ask you to take our concerns into account.

Yours sincerely

Andreas Kellerhals, President Opendata.ch
Hannes Gassert, Vice-President Opendata.ch

Comments on Art. 2 Scope

In the interest of a broader impact, in particular of the provisions on open data, we consider an extension of the scope to be clearly indicated, for example by analogy with the thematically related Freedom of Information Act (FoIA) and Archiving Act (ArchA). At least persons under public or private law, insofar as they perform federal enforcement tasks delegated to them, should also be made subject to the EMBAG.

Comments on  Art. 4 Principles

The wording of this article gives greater weight to the risks than to the opportunities, spelling out the former but not the latter. In the interest of balance, we therefore consider it appropriate to name the most important opportunities as well, such as speeding up and simplifying administrative activity or increasing participation.

Furthermore, in our view, in addition to the obligations of coordination and accessibility, there is no obligation to continuously adapt processes to digitalisation. We therefore propose an additional paragraph: 

4 When drafting acts, ordinances and directives, they shall ensure that electronic means are used efficiently and as a priority.

Anyone who digitalises a suboptimal process merely gets a suboptimal digital process. This can be actively prevented, which we consider necessary. 

Comments on Art. 7 Bearing of costs

From the point of view of open source and open data, but also from a start-up perspective, the question of pre-financing arises here. So that agreements under paragraph 1  are possible not only with established, well-funded players but also with emerging actors in the “civic tech” and “gov tech” field, at least the issue of innovation funding and pre-financing must be addressed here.

In our view, there is therefore an urgent need not only to consider the bearing of operating costs here, but also to include the financing of the necessary innovations at an early stage. In Germany, this is ensured, among other things, by the “Prototype Fund“, financed by the Federal Ministry of Education and Research there. Surprisingly, the Swiss counterpart, on the other hand, is currently financed purely privately.

Here the EMBAG offers an opportunity to open up innovative new possibilities for Switzerland too, including with non-traditional actors. We therefore propose supplementing Art. 7 accordingly, for example with the following new paragraph:

2 The Confederation may provide pre-financing and start-up financing if an arrangement under paragraph 1 can be found.

Comments on Art. 10 Open Source Software (OSS)

We very much welcome the clear regulation of this issue. Anchoring the open source principle at the level of law is correct and important.

However, the current wording clearly does not go far enough. Open source software represents an  outstanding opportunity to decisively strengthen Switzerland’s digital autonomy, to make an important contribution to the development of our digital knowledge society — and to reduce costs, particularly from an economic perspective. The European Commission’s Open Source Strategy also states this clearly.

Based on this assessment, we are convinced that the proposed “may” wording is not sufficient here. In our view, paragraph 1 should be amended as follows:

1 The federal authorities subject to this Act shall make software (..) available:

Exceptions to this principle may be permitted by the competent department if official secrecy, national security or other legitimate overriding factors make this necessary. This amounts to “open by default” in accordance with Art. 11 for software as well. Alternatively, an obligation to examine open source alternatives may be considered, provided it is sufficiently binding.

In order not to compete unnecessarily with the private sector, we propose making paragraph 4 more nuanced and, instead of “cost-covering fees”, requiring rather “remuneration in line with market conditions” if a market exists for comparable services.

Irrespective of the points mentioned above, we consider it necessary to address the topic of procurement here in Art. 10. On the one hand, “open by default” must be reflected in a binding way in the award criteria. On the other hand, practice shows that sufficient attention must always be paid to the aspect of source openness at an early stage in order for the various advantages of OSS, not least the “total cost of ownership”, to take effect. We therefore propose the following new paragraph:

6 Release shall be planned at an early stage in the design, procurement and development of the software.

In conclusion, we are convinced that with this increased binding force, Art. 10 clearly opens up more possibilities for a positive impact for both the authorities and business. The share of value created in Switzerland in software procurement, for example, is likely to increase significantly as a result.

Comments on Art. 11 Open Government Data (OGD)

Anchoring “open by default” in law is a milestone for digital Switzerland. It is likely to create great benefits for business, science and civil society, and also for the authorities themselves.

Here too, however, we would strongly encourage greater binding force. The national open data strategies and the corresponding cross-sectional audit by the Swiss Federal Audit Office (SFAO) have shown that greater binding force is necessary. We therefore propose the following clarification of paragraph 1:

1 The administrative units of the central Federal Administration are obliged to actively make the data (..) available for free re-use. Every person has the right to consult and use these data and to obtain information about their content from the authorities.

At present, many data under the heading of Open Government Data are only made available if a justified request is submitted. This clearly contradicts the spirit of OGD. A clarification as proposed above (“actively”) would prevent this and clearly anchor the requirement of the corresponding strategy in law as well 

In addition, a clear legal right of access must be defined, by analogy with the FoIA. This is the purpose of the second sentence of our proposal for Art. 11 paragraph 1.

Paragraph 3a unnecessarily restricts the free use of government data. Data and information protection are already mentioned in the strategies of 2014 and 2019. No further restrictions are needed, because registers in particular (with the exception of the civil status register) or platforms such as simap.ch contain data that should very much be understood as open government data in principle and should finally be made accessible for re-use. They enable structural insights that are indispensable for a modern society and its political debates.

Also in light of the experience gained with the national open data strategies, paragraph 3b must be supplemented by an obligation to prove any disproportionality. A mere statement that opening up would require too much effort must not suffice. In addition, it must be examined in each individual case whether the effort can be reduced by using new or other technical means in line with the “state of the art”. This is measured against the most efficient practice already established in cantons or communes.

We consider paragraph 5 questionable and propose deleting it. Accuracy and completeness can only ever be proven in practice. A complete prior check and a guarantee of correctness are major obstacles to publication. What is essential, however, is to reliably operate clearly defined channels for feedback on the data, so that the data can be continuously improved. In our view, this is the “state of the art”, as practised for example in transport or with geodata.

In addition, for the same purpose, an obligation could be introduced to provide clear information on accuracy, completeness and plausibility in the metadata.

In addition to binding force, in our view the pace of implementation must also be increased. At the same time, restrictions on “by default” must be kept as minimal as possible! In particular, complex release processes with a large number of stakeholders are to be avoided. The Confederation must be able to take the sole decision on opening up data, including data from third parties with whom it works or which it manages centrally for them

Comments on Art. 13 Standards

The wording “It shall be guided by internationally recognised or widely used standards.” is too weak; mere approximate guidance runs counter to the purpose of standards. Moreover, mere prevalence is not a sufficient criterion for the suitability of a standard for Switzerland and its authorities. We would therefore suggest adding a sentence such as the following:

Wherever possible and appropriate, standards shall be chosen that are freely and openly available and have an open reference implementation.

Comments on Art. 16 Transitional provisions

On paragraph 1: the maximum period should be set at 2 instead of 5 years, particularly with regard to Art. 11. A binding national Open Government Data strategy has existed since 2014; for Switzerland not to fall behind, a faster pace is now required.

On paragraph 2: of course, an obligation to release data already exists before entry into force, namely through the binding national open data strategy mentioned. This obligation has existed since 2014 and the present Act must not release anyone from it. If a wording is necessary here, “before the entry into force of this Act” should be replaced by “before the  year 2014”. An option we could accept is to prioritise the publication of current data, with retroactive publication as a second step.

Comments on Art. 17

In our view, paragraph 2 should be deleted, particularly with regard to Art. 11; given the urgency of clear progress mentioned above, the Act should enter into force upon its adoption.

Comments on the topic of interfaces (API)

In future, interfaces, so-called application programming interfaces (APIs), will be among the most important electronic means of fulfilling official tasks. They allow communication from software to software and are already in use in various places today. The aim in each case is either to give access to specific datasets without releasing the entire database as OGD or, and this is becoming increasingly important, to enable access to functions. This massively facilitates the interplay of applications, which not only makes new applications possible in the first place, but also clearly increases the re-use of these functions and thus their value.

The topic of interfaces is already of great importance today — and this will continue to grow. We therefore consider it urgent to enact the necessary rules in a separate EMBAG article.

These include:

  • An “API by default” principle, analogous to “open by default” in Art. 11 on OGD. Newly developed software must have documented interfaces that, wherever possible, can potentially be used by everyone. A transitional period is to be defined for “legacy software”. 
  • The available interfaces and their metadata are to be published on a central platform analogous to OGD — or, ideally, on the very same platform. This platform must provide “API management” with which access rights and query limits can be assigned per function and per class of functions.
  • The classification of functions (“API endpoints”), e.g. into “public”, “public with authentication”, “shared” (e.g. across federal levels) and “private” (only within the Federal Administration), is carried out by the competent department, as is the corresponding assignment of permissions.
  • The competent department operates a central “single point of orientation” to which proposals for new interfaces, suggestions for improvement etc. can be submitted.
  • The interfaces are to be designed according to open international standards, in line with Art. 13.

The topic of interfaces forms a bridge between the topics of OGD, OSS and standards; from our perspective, clear and binding regulation is necessary. The potential for business and society is enormous. In this regard, we would also like to refer to motions 20.4260, 18.4276 and 18.4238