1.Can the Union act? What is the legal basis and competence of the Unions’ intended action?

1.1 Which article(s) of the Treaty are used to support the legislative proposal or policy initiative?

The legal basis for the adoption of the implementing act is the Directive (EU) 2019/1024 (Open Data Directive), and in particular article 14 thereof. The Treaty basis for the adoption of the Open Data Directive was Article 114 of the Treaty on the Functioning of the European Union (TFEU).

1.2 Is the Union competence represented by this Treaty article exclusive, shared or supporting in nature?

In the case of the European Data Strategy, within the Commission 2019-2024 priority ‘Europe fit for the Digital Age’, the Union’s competence is shared.

2.Subsidiarity Principle: Why should the EU act?

2.1Does the proposal fulfil the procedural requirements of Protocol No. 2 1 :

-Has there been a wide consultation before proposing the act?

Yes, the proposal has been consulted and discussed with the stakeholders, in particular via the following measures:

·Series of 10 sectoral and horizontal workshops on the future Common EU data spaces, second half of 2019

·Public online consultation on the EU Data Strategy, February-May 2020

·Feedback to the Inception Impact Assessment; July-August 2020

·Focus groups within the Impact Assessment Support Study, 2020

·Online workshops on the policy recommendations of the IA Assessment Study, July and September 2020

·Meetings of the Open Data Committee, October 2019 and May 2020

-Is there a detailed statement with qualitative and, where possible, quantitative indicators allowing an appraisal of whether the action can best be achieved at Union level?

The explanatory memorandum and the impact assessment contains a section on the principle of subsidiarity (Chapter 3), as explained in section 2.2 below.

2.2Does the explanatory memorandum (and any impact assessment) accompanying the Commission’s proposal contain an adequate justification regarding the conformity with the principle of subsidiarity?

The adoption of the Implementing Act is a fulfilment of the legal requirement placed on the Commission by the European Parliament and the Council. According to Article 14(1) of the Open Data Directive, the Commission shall adopt implementing acts laying down a list of specific high-value datasets belonging to the categories set out in Annex I to the Directive.

The Implementing Act respects the subsidiarity requirements applicable to the basic act (Open Data Directive). Pursuant to its Article 1(3), the Directive builds on, and is without prejudice to, national (and union) access regimes. This means that data excluded or restricted from access by virtue of national (or Union) law will not be taken into account for the purpose of the Implementing Act.

The implementing act is adopted in accordance with the examination procedure provided in Article 5 of Regulation (EU) No 182/2011. A Committee on open data and the re-use of PSI, representing the Member States, assists the Commission in the adoption of the Implementing Act.

Furthermore, for identifying specific high-value datasets and according to article 14(2) of the Directive, the Commission carried out appropriate consultations, including at expert level, conducted an Impact Assessment and ensured complementarity with existing legal acts.

2.3Based on the answers to the questions below, can the objectives of the proposed action be achieved sufficiently by the Member States acting alone (necessity for EU action)?

No, the definition of an EU-wide list of datasets within the themes indicated by Annex I to the Open Data Directive would be impossible for the Member States to successfully accomplish when acting alone. The very objective of identifying a common data layer across the EU, with the corresponding common arrangements for publication and re-use (as mandated by article 14 of the Open Data Directive) calls for an EU-level action.

(a)Are there significant/appreciable transnational/cross-border aspects to the problems being tackled? Have these been quantified?

The aim of the European Strategy for Data 2 , of which the High Value Datasets initiative is an important part, is to create a single European data space – a genuine single market for data.

The value of the data economy, which measures the overall impacts of the Data Market on the economy as a whole, exceeded the threshold of 400 Billion Euro in 2019. The total direct economic value of public sector information for the EU28 was estimated at 52 billion EUR.

The data value chains in the EU are already structured largely in a cross-border manner, with data holders, data processors, data enrichers and final data users scattered across various Member States. The strengthening of the EU data space would not be possible without concerted EU action.

(b)Would national action or the absence of the EU level action conflict with core objectives of the Treaty 3 or significantly damage the interests of other Member States?

The absence of EU action might negatively affect two key objectives of the treaty: promoting scientific and technological progress and enhancing economic, social and territorial cohesion and solidarity among EU countries. Action by Member States alone would not achieve the scale needed for the successful development of the data technologies (e.g. Artificial Intelligence). It would also limit the possibilities for scaling up of SMEs and startups active in the EU data economy.

(c)To what extent do Member States have the ability or possibility to enact appropriate measures?

The removal of the remaining obstacles to an open re-use of public sector information and simultaneously aligning the legal framework to the evolving digital socio-economic environment cannot be achieved by Member States alone. Diverging national legal solutions would likely compromise the growing tendency towards cross-border re-use, whereas the different levels of open data maturity 4 across EU Member States would persist or deepen, having a negative effect on the homogeneity and competitiveness within the Digital Single Market.

(d)How does the problem and its causes (e.g. negative externalities, spill-over effects) vary across the national, regional and local levels of the EU?

There are regional and local variations in the accessibility, quality and quantity of public sector information available for reuse. This is due to a multitude of factors, including the policy and legal traditions (e.g. historically high public sector transparency in the Nordic countries) and the digitisation levels in the public sector (varying substantially across the EU, as shown by the DESI reports). Nevertheless, the overall trend is supportive to the further opening of public sector resources for innovative reuse (see: 2019 and 2020 Open Data Maturity reports of the European Data Portal).

(e)Is the problem widespread across the EU or limited to a few Member States?

The problems identified in the Impact Assessment are widespread and not specific to a few Member States. Despite notable improvements, the EU-wide nature of the problems has been observed since the inception of the Open Data policy and the introduction of first EU legislative measures (Directive 2003/98/EC).

(f)Are Member States overstretched in achieving the objectives of the planned measure?

No. While the public sector of the Member States will be the main stakeholder group burdened with costs related to the measure in question (as opposed to e.g. SMEs or researchers), the achievement of the objectives will not impose undue burdens. The impact assessment calculates that the direct effect of the chosen policy package is EUR 75 billion, representing an incremental impact of EUR 6 billion as compared to the baseline in 2028. On the other hand, one off costs (such as API set up) range from EUR 24.9million, (low estimate) to EUR 435.9 million (high estimate), with a medium costs expected to be around EUR 122.3 million. Recurrent costs (such as loss of revenue and resources needed to increase data quality up until 2028) amount to the average estimate of EUR 473.6 million. The overall cost/benefit ratio could thus be roughly represented as 1:6.

It is worth noting that part of the costs can be offset by funding opportunities from the Digital Europe Programme, specifically designed to alleviate burden related to the publishing of HVDs.

In addition, the Implementing Act will not be applicable until 2022 while the requirement to allow the free re-use of high value datasets can be delayed by the Member States up to 2 years following the entry into force of the Implementing Act.

(g)How do the views/preferred courses of action of national, regional and local authorities differ across the EU?

The views of different actors differ according to the administrative structure of the Member State (e.g. where local or regional authorities are given responsibility over a given high value dataset, they are often more apprehensive about the possible costs associated with making it freely available). National authorities are broadly more supportive to the measures announced This was illustrated by the results of the public online consultation, where a proportion of 82.2% of respondents considered that the establishment of a list of high-value datasets is a good way to ensure that public sector data has a positive impact on the EU’s economy and society. This figure increases to 90% when considering public authorities respondents only.

Local authorities are also generally supportive but this is due to the fact that the Implementing Act mostly targets large national (or regional in some cases) data holders while local authorities would likely be among the net beneficiaries of the proposed initiative.

2.4Based on the answer to the questions below, can the objectives of the proposed action be better achieved at Union level by reason of scale or effects of that action (EU added value)?

Action at EU level is best suited to guarantee that public data of comparable thematic scope are available for re-use across Member State borders under similar legal and technical conditions so as to facilitate the offering of services based on data sourced from different EU countries or for applying a data-based business model tested in one Member State seamlessly to another.

(a)Are there clear benefits from EU level action?

Yes, the benefits are particularly relevant in the case of ensuring a certain level of interoperability between the datasets in question (legal, technical and semantic) allowing for the data to be usable within the context of the Common European Data Spaces (as explained in the EU Data Strategy of February 2020).

(b)Are there economies of scale? Can the objectives be met more efficiently at EU level (larger benefits per unit cost)? Will the functioning of the internal market be improved?

Yes, data economy can contribute with as much as 4% to the EU’s GDP. The Free Flow of Data Regulation (EU) 2018/1807, the upcoming initiative on the governance of the EU data spaces and the results of the evaluation of the Directive 2003/98/EC on the re-use of public sector information all show that the functioning of the internal market is not only improved but greatly depends on the introduction of facilitating measures on EU level.

(c)What are the benefits in replacing different national policies and rules with a more homogenous policy approach?

Many data-dependent technologies require the processing of large and possibly homogenous datasets. This has allowed the quick progress of US and Chinese social networks or payment platforms. The support towards the development of the data markets by individual Member States could never reach the necessary scale to noticeably empower EU companies to successfully compete on the global data market.

(d)Do the benefits of EU-level action outweigh the loss of competence of the Member States and the local and regional authorities (beyond the costs and benefits of acting at national, regional and local levels)?

Yes, while the local or regional data holders may in some cases need to redefine the financing of some of their operations (they will no longer be able to charge for the re-use of data), the overall data market in Europe will benefit from easier availability of datasets with a great potential for the creation of commercial applications. This will indirectly lead to higher socio-economic benefits across the entire EU (more innovation and research, higher number of start-ups, easier scaling up for SMEs, provision of EU-wide data services, etc.)

(e)Will there be improved legal clarity for those having to implement the legislation?

The legislation is very focused and limited by the rules of the Open Data Directive. Given that open data legislation has been applied in the EU since 2003, and given that the Member States are closely associated to the adoption of the measures (Open Data Committee and Public Sector Information Expert Group), we do not expect legal uncertainties to arise, apart from the specific case of personal data protection which is however outside of the scope of this initiative.

3. Proportionality: How the EU should act

3.1 Does the explanatory memorandum (and any impact assessment) accompanying the Commission’s proposal contain an adequate justification regarding the proportionality of the proposal and a statement allowing appraisal of the compliance of the proposal with the principle of proportionality?

Ensuring the proportionality of the proposed measure was one of the main purposes of the Impact Assessment. The final policy choice supported by the Impact Assessment (and guided by the results of an independent study) balanced out the costs and benefits for each of the six thematic areas of intervention. This allowed for the adoption of what could be considered a ‘high intensity’ scenario for those areas where the compliance burden linked to the new measure would be easy to manage whereas choosing a more cautious approach in other areas (in which the scope of the dataset and the accompanying publishing arrangements are not as extensive as pleaded for by the re-users community).

For the reasons of limiting compliance burden, it was decided not to include in the list those datasets that are held by public undertakings (despite the fact that, as data newly included in the scope of the Open Data Directive, they could also be taken under consideration for the Implementing Act).

3.2Based on the answers to the questions below and information available from any impact assessment, the explanatory memorandum or other sources, is the proposed action an appropriate way to achieve the intended objectives?

Yes, the action is appropriate given its measured, limited scope and the policy context (e.g. the post COVID recovery plan includes reference to high value dataset, EU Data strategy and the setting up of EU data spaces as well as the Green Deal initiative requiring better availability of environmental data). The intervention is gradual, allowing the Commission to extend the list of data themes and the corresponding datasets via future delegated and implementing acts in line with the socio-economic developments.

(a)Is the initiative limited to those aspects that Member States cannot achieve satisfactorily on their own, and where the Union can do better?

Indeed. The initiative leaves it to the Member States to decide which data should be produced at all and which data should be excluded from access (due to national security, privacy or other reasons). The initiative only regulates the modalities of re-use of already accessible data and the new rules put in place are limited to those aspects which are necessary to ensure the realisation of socioeconomic potential of these datasets on an EU scale.

(b)Is the form of Union action (choice of instrument) justified, as simple as possible, and coherent with the satisfactory achievement of, and ensuring compliance with the objectives pursued (e.g. choice between regulation, (framework) directive, recommendation, or alternative regulatory methods such as co-legislation, etc.)?

Yes, as an implementing act with a limited thematic scope well framed by the basic act, the choice of a regulation and the relatively generous delay in the applicability of the new provisions will ensure a uniform implementation while leaving sufficient flexibility to find the necessary technical and financial arrangements.

(c)Does the Union action leave as much scope for national decision as possible while achieving satisfactorily the objectives set? (e.g. is it possible to limit the European action to minimum standards or use a less stringent policy instrument og approach?)

Yes, the initiative only concerns the reuse of specific datasets with the highest potential for reuse. All other publicly held datasets are governed by the rules of the Open Data Directive which leaves a much higher degree of flexibility in terms of specific transposition measures on the Member State level.

(d)Does the initiative create financial or administrative cost for the Union, national governments, regional or local authorities, economic operators or citizens? Are these costs commensurate with the objective to be achieved?

The initiative creates costs on the national level, exclusively affecting the public administration. At the same time the citizens, economic operators and the public sector in its capacity of data user are among the beneficiaries of the initiative.

The overall assessment of the costs and benefits has demonstrated a clear socio-economic justification of the measures proposed.

(e)While respecting the Union law, have special circumstances applying in individual Member States been taken into account?

Yes, the most notable burden linked to the initiative is the necessity to allow the reuse of data free of charge. In some Member States the data in scope of the initiative are currently sold. This is why the measure allows individual Member States, based on their own assessment and the national circumstances to defer the application of the free availability requirements for up to two years.