Special Report No 3/2010: Impact assessments in the EU institutions: do they support decision-making? - Hoofdinhoud

Inhoudsopgave van deze pagina:

Delen

1. Tekst
2. Originele weergave

3. Meer informatie

COUNCIL OFBrussels, 4 November 2010 THE EUROPEAN UNION

15795/10

FIN 541 COMPET 325 INST 464 POLGEN 176 COVER NOTE from:

Mr Vítor CALDEIRA, President of the Court of Auditors date of receipt: 30 September 2010 to: Mr Steven VANACKERE, President of the Council of the European Union Subject: Special Report No 3/2010: Impact assessments in the EU institutions: do they support decision-making?

Sir,

I enclose a copy of special report No. 3/2010 "Impact assessments in the EU institutions: do they support decision-making?" together with the Commission's replies.

The special report, which is shortly to be published, was adopted by the Court at its meeting on 19 May 2010 and is accompanied by the replies from the Commission, which was notified of the preliminary findings on 19 February 2010.

(Complimentary close).

(s.) Vítor CALDEIRA

________________________

Encl.: Special Report No 3/2010: Impact assessments in the EU institutions: do they support

decision-making?EURÓPAI SZÁMVEVSZÉK TRIBUNAL DE CUENTAS EUROPEO IL-QORTI EWROPEA TAL-AWDITURI EVROPSKÝ ÚCETNÍ DVR EUROPESE REKENKAMER DEN EUROPÆISKE REVISIONSRET EUROPEJSKI TRYBUNAL OBRACHUNKOWY EUROPÄISCHER RECHNUNGSHOF TRIBUNAL DE CONTAS EUROPEU EUROOPA KONTROLLIKODA CURTEA DE CONTURI EUROPEAN

O EURÓPSKY DVOR AUDÍTOROV EUROPEAN COURT OF AUDITORS CORTE DEI CONTI EUROPEA EVROPSKO RACUNSKO SODISCE COUR DES COMPTES EUROPÉENNE EIROPAS REVZIJAS PALTA EUROOPAN TILINTARKASTUSTUOMIOISTUIN CÚIRT INIÚCHÓIRÍ NA HEORPA EUROPOS AUDITO RMAI EUROPEISKA REVISIONSRÄTTEN

Special Report No 3/2010

(pursuant to Article 287(4), second subparagraph, TFEU)

Impact Assessments in the EU institutions:

do they support decision making?

together with the Commission's replies

2 TABLE OF CONTENTS

Paragraph

1.1.1. Glossary of terms and abbreviations

1.1.2. Executive Summary I - VI

1.1.3. Introduction 1 - 12

1.1.4. Audit scope and approach 13 - 16

1.1.5. Observations 17 - 81

1.2. Use of impact assessments for decision-making at the Commission, the European Parliament and the Council 17 - 41 Impact assessment helps the Commission to formulate its proposals 19 - 27 The Commission's impact assessment reports are considered by most users at the European Parliament and Council as relevant information in the legislative decision-making process

28 - 33 Amendments to the initial Commission proposal are not subject to additional assessments of potential impacts 34 - 41

1.3. The Commission's procedures for the preparation of impact assessments 42 - 55 Need for more transparency in the selection procedure and targeting of impact assessment work 43 - 46 Consultation is widely used as input for impact assessments, but not on draft reports 47 - 50 Recent improvements in the Commission's quality review of its impact assessment reports 51 - 55

1.4. Content and presentation of the Commission's impact assessment reports 56 - 81 Need to improve further the presentation of how the intervention proposed is to achieve best the intended outcomes

57 - 60 There is room for improvement in the analysis of impacts and the presentation thereof 61 - 71

3

Information on implementation aspects, enforcement costs and administrative burden can be further improved 72 - 81

1.4.1. Conclusions and Recommendations 82 - 87

1.4.2.

1.4.3. Annex I - Comparison of specific elements of the Commission's system with IA systems elsewhere

1.4.4. Annex II - Audit approach and evidence collection methods

1.4.5.

1.4.6. Letters from the Presidents of the European Parliament and the Council of the European Union

1.4.7.

1.4.8.

1.4.9. Commission's replies

4 GLOSSARY OF TERMS AND ABBREVIATIONS

Administrative burden: The administrative and information costs that businesses incur in order to comply with legal obligations.

Annual Policy Strategy (APS): The Commission's Annual Policy Strategy lays down, early in year n-1, political priorities and key initiatives for year n. At the same time, it allocates the corresponding financial and human resources to these priority initiatives. It serves as a basis for a political exchange of views on the Commission's programme with the European Parliament and the Council.

Co-decision: Under the co-decision procedure, the Council shares legislative power with the European Parliament. Both institutions can, however, only act on a proposal by the Commission (which has the sole right of initiative).

Comitology: Committee system which oversees the delegated acts implemented by the European Commission. The committees are composed of representatives of the Member States and have the mandate to regulate certain delegated aspects of the secondary legislation adopted by the Council and, where co-decision applies, the European Parliament. The Commission chairs these meetings and provides the secretariat.

Commission's Legislative and Work Programme (CLWP): In this document, published in November of each year, the Commission presents its planned legislative and other initiatives for the following year. The CLWP does not contain all initiatives to be brought forward in a given year, but identifies policy initiatives of major importance. It thereby operationalises the political priorities and initiatives specified in the APS.

Council Working Party (WP): Working groups specialised in a given policy field at the Council of Ministers consisting of delegations of all Member States.

They prepare the legal acts to be adopted by the relevant Council of Ministers.

Directors and Experts for Better Regulation (DEBR): Expert group consisting of officials in charge of Better Regulation in the EU Member States

5

and other European countries. DEBR meets twice per year and is chaired by the delegation of the incoming Council presidency. The mandate of this group is to exchange ideas and to further develop existing initiatives by the EU and its Member States to reduce bureaucracy and improve legislation.

High Level Group of Independent Stakeholders on Administrative Burdens ("Stoiber Group"): This group advises the Commission with regard to the Action Programme for Reducing Administrative Burden. Dr Edmund Stoiber, former Prime Minister of Bavaria, is chairman of this group. The other members represent important stakeholder organisations. Its mandate expires in 2010. IA: Impact Assessment

Impact Assessment Board (IAB): A body attached to the Commission's Secretariat-General that assesses the quality of each impact assessment report and publishes its opinion thereon. The Board consists of four directors from different DGs and the deputy Secretary-General of the Commission.

Intervention logic: Intervention logic is the conceptual link between an intervention's inputs and the production of its outputs and, subsequently, its impact in terms of results and outcomes.

Mandelkern Group: High-level advisory group that consisted of regulatory experts from the Member States and the European Commission. This group was established by the Public Administration Ministers of the Member States in November 2000 and chaired by the French Dieudonné Mandelkern, a former Member of the Conseil d'Etat. Among other proposals, the group recommended in 2001 that the Commission should develop a tool for assessing the social, economic and environmental impacts of proposed legislation. MEP: Member of European Parliament

6

Rapporteur: Term used in the European Parliament to describe the MEP(s)

who is (are) in charge of a given proposal or report. MEPs in charge of a proposal belonging to political groups different from the group of the official European Parliament rapporteur in charge of the same proposal are called shadow-rapporteurs.

Roadmap: A short document that gives an initial description of a planned Commission initiative, including an indication of the main areas to be assessed in the IA and the planning of IA work. These roadmaps have two functions: they provide an estimated timetable for the proposal and set out how the impact assessment will be taken forward. Roadmaps are published when the Commission Legislative and Work Programme (CLWP) is adopted.

Standard Cost Model (SCM): Method to assess costs incurred to meet information obligations created by legislation. Calculates cost on the basis of the average unit cost of the required administrative activity multiplied by the total number of activities performed per year.

Transposition: In European Union law, a process by which the Member States give force to an EU Directive in national law by adopting appropriate implementing legislation.

7 EXECUTIVE SUMMARY

I. Impact Assessment is one of the cornerstones of the Commission's Better Regulation policy for the improvement and simplification of new and existing legislation. Its purpose is to contribute to the decision-making processes by systematically collecting and analysing information on planned interventions and estimating their likely impact. This should provide the bodies involved in legislative decision-making with a basis on which to decide on the most appropriate way to address the problem identified. II. The audit analysed whether impact assessments supported decision-making in the EU institutions. In particular, it examined the extent to which

• - 

  impact assessments were prepared by the Commission when formulating its proposals and the European Parliament and the Council consulted them during the legislative process;
• - 

  the Commission's procedures for impact assessment appropriately supported the Commission's development of its initiatives; and
• - 

  the content of the Commission's impact assessment reports was appropriate and the presentation of findings was conducive to being taken into account for decision-making. III. The period under examination was 2003-2008 and the audit involved inter alia an international comparison of impact assessment systems, an analysis of a sample of Commission impact assessments, interviews and surveys with people involved in performing, reviewing and using the Commission's impact assessments, both within and outside the Commission. The findings were examined against the relevant inter-institutional agreements, the Commission's guidelines and a set of good practices observed in policy documents and established by the OECD. Throughout the audit, expert groups provided advice and supported the audit work.

8 IV. Better Regulation is a responsibility of all EU institutions involved in the legislative process. On balance, particularly in recent years, the audit has shown that impact assessment has been effective in supporting decision- making within the EU institutions. In particular, it was found that the Commission had put in place a comprehensive impact assessment system since 2002. Impact assessment has become an integral element of the Commission's policy development and has been used by the Commission to design its initiatives better. The Commission's impact assessments are systematically transmitted to the European Parliament and Council to support legislative decision-making and users in both institutions find them helpful when considering the Commission's proposals. However, the Commission's impact assessments were not updated as the legislative procedure progressed and the European Parliament and Council rarely performed impact assessments on their own amendments.

V. The audit identified areas for improvement related to the impact assessment procedures followed by the Commission and the content and presentation of impact assessment reports:

(i) the Commission did not publish a comprehensive overview of the legislative initiatives outside the Commission Legislative and Work Programme (CLWP) selected to undergo an impact assessment or explain why certain initiatives rather than others were selected.

Consultation with stakeholders was used widely for initial input but not carried out on draft IA reports. Recent improvements were noted regarding the Commission's internal quality control of impact assessment work, but the timeliness of the Impact Assessment Board (IAB)

intervention could be improved; and

(ii) the Commission's impact assessment reports generally provided a sound description of the problem at stake and specified the objectives pursued.

These and other mandatory sections of impact assessment reports were found to comply with the Commission's guidelines. However, the main

9

results and messages of IA reports are not always easy to gather and comparing the impacts of the various policy options presented in an IA report is sometimes difficult. Problems with quantifying and monetising impacts can be traced back to the availability of data. Finally, implementation and enforcement costs and the potential administrative burden of proposed legislation were not always sufficiently analysed or quantified. VI. The Court considers that the Commission should give due consideration to the principles of clarity of objectives, simplification, realism, transparency and accountability when designing new interventions and revising existing ones.

The European Parliament, the Council and the Commission may wish to consider the findings and recommendations set out in this report when revising their Inter-institutional agreements on "Better law-making" and a "Common approach to impact assessment".

10 INTRODUCTION

'Better Regulation' in the EU context

• 2. 

  The Commission's 'Better Regulation' policy aims at designing new legislation better and simplifying existing legislation. The 'Better Regulation'

initiative was a response to the need expressed at the Gothenburg and the Laeken European Councils in 2001 to:

• - 

  simplify and improve the European Union's regulatory environment; and
• - 

  consider the effects of proposals in their economic, social and environmental dimensions.
• 3. 

  Since then the European Commission has introduced several measures, including impact assessment (IA), to improve the way it designs interventions (see Box 1).

Box 1 Better Regulation policy instruments

• - 

  simplifying existing legislation;
• - 

  screening and, where applicable, withdrawing pending proposals;
• - 

  monitoring and reducing "administrative burdens";
• - 

  consulting interested parties; and
• - 

  impact assessment (IA).

11

Impact assessment as a tool for Better Regulation

Coverage of the Commission's impact assessment system

• 4. 

  IAs are carried out for legislative proposals and other Commission initiatives. All major legislative, budgetary and policy-defining initiatives with significant impact must undergo an IA. Major policy initiatives are defined as all those presented in the Annual Policy Strategy (APS) or, later, in the Commission's Legislative Work Programme (CLWP), with some clearly defined exceptions. In addition, other significant initiatives can be covered on a case by case basis.
• 5. 

  The Commission's IA system also applies to existing EU legislation when there is a revision or an update of the 'acquis communautaire', for example under the "Simplification Rolling Programme". Finally, in 2009, the scope of the Commission's IA system was extended to implementing rules (or `comitology' decisions). These implementing measures are a significant source of EU legislation, as around 250 comitology committees adopt around 2 600 such measures every year. Approximately 1 000 of them are based on legislation adopted under the co-decision procedure.

12

Commission's impact assessment system coverage is wider than that of comparable national systems

• 6. 

  Fully operational IA systems are in place in only a few Member States.

Moreover, the Commission's IA system has a different and, to some extent, wider scope than other systems in OECD countries (see Annex I):

• - 

  firstly, other IA systems (such as in the USA) address only implementing rules (i.e. regulatory measures which further specify legislation previously adopted by Congress; in the EU context, such measures are comparable to 'comitology decisions'); and
• - 

  secondly, some European countries (such as the Netherlands and Germany) focus their analysis of forthcoming regulation on the assessment of administrative burden. In the case of the Commission, this is only one of several aspects analysed within an IA.

Content of the Commission's impact assessment reports

• 7. 

  The reports resulting from these IAs are expected to identify and assess the problem at stake and the objectives pursued, develop the main options for achieving the policy objective, and analyse their likely economic, environmental and social impacts. They should also analyse potential administrative burdens resulting from the proposed options, assess possible implementation and enforcement problems and specify appropriate monitoring and evaluation arrangements for the intervention or programme proposed.

13

Organisational responsibility, procedural aspects and cost of the Commission's impact assessment system

• 8. 

  Within the Commission, each Directorate-General (DG) is responsible for preparing its IAs in line with the Commission's guidelines. Following the first version in 2002, this guidance material has been updated on three occasions, with the latest update taking place in January 2009. IAs are carried out by the staff responsible for the policy initiative, supported by dedicated `IA support units'. Since 2007, an Impact Assessment Board (IAB) has provided quality support and control for all IAs within the Commission.
• 9. 

  A preliminary draft IA report must accompany the proposal when going through inter-departmental consultation. A final draft IA report will accompany the legislative proposal when it is sent to the College of Commissioners for its final adoption. The final IA report is also sent to the European Parliament and the Council along with the definite proposal and is made available on the 'Europa' website.

Inter-institutional agreements between the European Parliament, the Council and the Commission

• 10. 

  In the Inter-institutional Agreement on better law-making in 2003, the European Parliament, the Council and the Commission agreed "... on the positive contribution of IAs in improving the quality of Community legislation, with particular regard to the scope and substance thereof". In 2005, a further

14

Inter-institutional Agreement was signed between the European Parliament (EP), the Council and the Commission on a "Common approach to IA".

• 11. 

  In 2005 the European Parliament, the Council and the Commission had agreed that the Inter-institutional agreement was to be reconsidered by the end of 2007. This review is still outstanding. As a preparatory step, in 2008, the European Parliament, the Council and the Commission produced 'stock-taking reports' which set out their views on the "Common Approach to IA" since 2005.
• 12. 

  On several occasions, the European Parliament has adopted resolutions on Better Regulation and the use of IA, which, in general, viewed IA as helpful in informing EU decision-making. Similarly, the Council committed itself, in its

15

recent conclusions on Better Regulation, towards continuing to use the Commission's IA reports and the accompanying IAB opinions throughout the negotiating process. The European Parliament and the Council have communicated to the Court their views on topics covered in this report (letters from the Presidents of the European Parliament and the Council are attached to this report).

• a) 

  Figure 1 provides an overview of the role of impact assessment in developing and deciding on the Commission's initiatives and legislative proposals.

Better Regulation in the EU" (2007/2095(INI)); A6-0273/2007 final, 2.7.2007: see explanatory statement and points 5 to 15, 21, 26, 30, 42 to 43.

16

Figure 1 -The role of impact assessment in developing and deciding on the Commission's initiatives and legislative proposals IA provides support for decision- making ...IA provides support for decision- making ...Final legislationFinal legislation

... in European Parliament and Council... in European Parliament and CouncilLegislative decision-making:Legislative decision-making:Appreciation of IA reports, Appreciation of IA reports, discussion of IA reports, impacts of amendments to Commission proposals are assesseddiscussion of IA reports, impacts of amendments to Commission proposals are assessedIA reportIA report Commission legislative proposal / initiativeCommission legislative proposal / initiative

Development of Commission proposal: Development of Commission proposal:

... in the Commission... in the CommissionCoverage of mostCoverage of most relevant initiatives, IA part of policy development, IA as support for decision-makingrelevant initiatives, IA part of policy development, IA as support for decision-making

Procedures for IA in the CommissionProcedures for IA in the CommissionSelection of initiatives to undergo IASelection of initiatives to undergo IAConsultation of interested parties for IA Consultation of interested parties for IA Quality control of IAQuality control of IA

Content and presentation of Commission's IA reportsContent and presentation of Commission's IA reportsDecription of policy problem, objectives and intervention logicDecription of policy problem, objectives and intervention logicComparison of policy options through assessment of impactsComparison of policy options through assessment of impactsImplementation, enforcement, monitoring and evaluation of initiativeImplementation, enforcement, monitoring and evaluation of initiative AUDIT SCOPE AND APPROACH

• 13. 

  The audit assessed whether the Commission's IAs supported decision- making in the EU institutions. In particular, it examined the extent to which:
• - 

  IAs were prepared by the Commission when formulating its proposals and the European Parliament and the Council consulted them during the legislative process;
• - 

  the Commission's procedures for IA appropriately supported the Commission's development of its initiatives; and

17

• - 

  the content of the Commission's IA reports was appropriate and the presentation of findings was conducive to being taken into account for decision-making.
• 14. 

  The period under review was 2003 to 2008. The audit involved inter alia:
• - 

  a comparison of specific elements of the Commission's system with IA
• - 

  systems elsewhere;
• - 

  an analysis of IA production statistics and of a sample of IA reports (scorecard analysis related to five DGs and corresponding to around a quarter of all IAs produced during the period audited); and
• - 

  enquiries and surveys with people involved in performing, reviewing and using the Commission's IAs both within and outside the Commission.

Annex II contains a more detailed description of the methodology used for the audit.

• 15. 

  The findings were examined against the relevant inter-institutional agreements, the Commission's guidelines and a set of good practices observed in policy documents and established by the OECD20. Expert groups 20 OECD, Regulatory Impact Analysis: Best Practice in OECD Countries, 1997; OECD, APEC-OECD Integrated Checklist on Regulatory Reform: Final Draft, 2005; OECD, Indicators of Regulatory Management Systems, 2009.

18

composed of practitioners in the field of IA and Better Regulation were invited by the Court to provide support and advice throughout the audit.

• 16. 

  The analysis of the quality of individual pieces of legislation or the process of law-making as such were not within the scope of the audit. OBSERVATIONS

Use of impact assessments for decision-making at the Commission, the European Parliament and the Council

• 17. 

  IAs contribute to the decision-making processes by systematically collecting and analysing information on planned interventions and estimating their likely impact. The relevant inter-institutional agreements between the European Parliament, the Council and the Commission define that IAs should first of all help the Commission to develop its policy initiatives and legislative proposals. Subsequently, during the legislative procedure, IAs presented by the Commission and, where appropriate, those prepared by the European Parliament and the Council, should provide the information based on which the legislators can decide on the most appropriate way to address the problem identified.
• a) 

  For this purpose, the audit examined:
• - 

  the extent to which the Commission has carried out IAs for its legislative proposals and other initiatives and whether IAs were actively used in the Commission's policy development and its preparation of legislative proposals;
• - 

  whether the Commission's IAs were perceived by users at the European Parliament and the Council as contributing to the legislative decision-

19

making process and whether IAs affected the way in which legislative proposals were discussed in the relevant committees and working parties;

and

• - 

  the extent to which the legislators carried out IAs of their own substantive amendments or requested the Commission to present updates of its IAs in the light of such amendments and whether national bodies provided additional information to inform legislative decision-making at the EU level.

Impact assessment helps the Commission to formulate its proposals

Significant increase in the number of impact assessments carried out since 2003

• 18. 

  During the period audited IAs had strong institutional backing and were increasingly used to provide input to the Commission's development of policy initiatives and legislative proposals22. Overall, 404 IA reports were published by the Commission during the period audited. Since 2003, the number of IAs carried out annually has increased to 121 reports in 2008 (see Figure 2). 22José Manuel Barroso - President of the European Commission: Response to UK Presidency Programme, European Parliament - Plenary Session, Brussels, 23 June 2005; Uniting in peace: the role of Law in the European Union, Jean Monnet lecture, EUI Florence, 31 March 2006; Qualité de la législation européenne : le temps des résultats Parlement Européen, Strasbourg, 4 avril 2006; Presentation of the European Commission's 2007 Work Programme, European Parliament, Strasbourg, 14 November 2006; Alive and kicking: the renewed Lisbon strategy is starting to pay off; Brussels, 5 March 2007; "We are all new Europeans now", Lithuanian Parliament (Seimas), Vilnius, 29 March 2007. 165 Günther Verheugen, 'Better Regulation' Conference of the Slovenian Presidency; Lubljana, 17 April 2008.

20

Figure 2 - Number of impact assessments performed each year

140 121 120Total of 404 IA reports 10093 807267 60 40 2130 20 0 200320042005200620072008

Source: European Commission.

Commission's objective of CLWP coverage attained...

• 19. 

  The selection of initiatives to undergo an IA is a significant decision, in terms of political implications and resources needed. During the period audited the main condition for selection was the inclusion of the initiative in the CLWP, which contains the Commission's major policy initiatives as set out in the Annual Policy Strategy (APS; see Box 2).

Box 2 Set of conditions for selecting initiatives that are to undergo IA

In 2003, the first year of the implementation of the IA system, the Commission decided to select a certain number of proposals that should undergo IA. The Commission based its selection on the importance of the proposals selected in relation to the political priorities, the feasibility for its departments to perform the assessments in the short-term, and the need to maintain a balance between different types of proposal and the involvement of a broad range of departments. 2004 was the first year in which the IA procedure was fully integrated into the programming cycle and the use of IAs

21

was extended to cover a set of proposals listed in the CLWP which were considered to have significant impacts. Since the 2005, IAs have been required for all initiatives set out in the CLWP, with some types of initiatives being exempted.

• 20. 

  From 2003-08, IAs were carried out for 69 % of all CLWP initiatives. Since 2005, in accordance with the criteria specified by the Commission at the time, IAs were undertaken for all relevant items on its CLWP, i.e. those initiatives that were considered to have significant impact.

... and increasing trend to also carry out impact assessments for initiatives outside the CLWP

• 21. 

  In line with its rules and based on a case-by-case assessment, the Commission has also produced IAs for initiatives not included in the CLWP.

In 2008, such IAs accounted for a majority of IAs carried out in that year (see Figure 3).

22

Figure 3 - Breakdown of impact assessments between 2005 and 2008 CLWP vs non-CLWP

6755 % 1014 %45 2048 %

30 % 6286 %45 % 4770 %4852 %54

2005200620072008

Commission's Legislative and Work Programme Non-Commission's Legislative and Work Programme

Source: ECA Analysis of Commission data.

Conducting impact assessments is becoming standard practice in the Commission's policy development

• 22. 

  IA is increasingly becoming part of the policy development culture at the Commission. Over the past few years the Commission has provided extensive training in IA methodology. The way in which IAs are to be produced is set out in Commission IA guidelines issued by the Secretariat-General that are applicable in all departments and services. Since 2002, this material has been updated on three occasions, with the latest update taking place in January 2009.
• 23. 

  Since 2007, an Impact Assessment Board (IAB) supported by the Commission's Secretariat-General (SG) reviews the quality of the IAs carried

23

out by the Directorates-General. In addition, the IA guidelines have recently specified that Directors-General must assume responsibility for the methodological soundness of the IA documents by signing the draft IA report submitted for the IAB quality check. The guidelines also prescribe the use of IA Steering Groups which aim to ensure consistency between policy areas in the IA exercise.

Impact assessments primarily help to shape Commission proposals

• a) 

  The audit found that IAs were not used by the Commission to decide whether to go ahead with a proposal. The decision whether to launch an initiative is generally taken before an impact assessment report is finalised.

The Commission rather uses IA to gather and analyse evidence that, during the policy development process, is used to improve its proposed initiative (see Box 3 for an example).

Box 3 Impact assessment helps to improve the proposed initiative:

"Roaming I"

In the "Roaming I" case, the decision to legislate by means of a Regulation (rather than through a Directive or not to legislate at all) was upheld, but the legislative proposal was adjusted throughout the IA process. In its second phase consultation, the Commission had proposed a regulatory approach called the 'Home Pricing Principle' as a possible means to address the problems in the roaming market. In the IA, which followed the public consultation process, the Commission took on board the view of the majority of respondents, and ultimately proposed an alternative approach (i.e. the European Home Market approach) which provided a better solution to the problems in the roaming market.

24

• 24. 

  The audit also showed that, in particular for recent years, IA is becoming broader in terms of the number of alternative options analysed and that resources are targeted to the assessment of initiatives according to their importance:
• - 

  analysis of an increasingly wide range of policy alternatives: Commission staff in charge of drafting legislative proposals and the related IAs reported that the IA process required them to consider more alternatives than previously when preparing a policy initiative. It was found that, throughout the period audited, the number of alternative options presented in the IA reports increased. In particular, for IAs produced in 2007 and 2008, the Court's analysis showed that the Commission increasingly analysed several feasible regulatory alternatives during its IAs; and
• - 

  prioritisation of resources on legislative proposals and CLWP initiatives:

The Commission considers that more resources should be devoted to the analysis of the most significant initiatives than to the less significant initiatives. In the Commission guidelines this is called the principle of proportionate analysis. Based on estimates provided by the Commission, it can be shown that more time is devoted to those IAs which are undertaken for legislative proposals (as compared to non- legislative initiatives) and which are included in the CLWP (as compared to those which are not) (see Figure 4). This indicates that the Commission focuses its resources in accordance with the pre-defined priorities.

25

Figure 4 - Average staff resources used for the production of impact assessment reports

Legislative

Non LegislativeCLWP item

Non-CLWP item 0246810

Average staff resources per IA in person-months

Note: 52 IA reports.

Source: ECA analysis of Commission data (2009).

Impact assessments are actively used for decision making at the level of the College of Commissioners

• 25. 

  At the Commission, initiatives are designed and impact assessments are prepared within the Directorates-General under the authority of the Commissioner in charge of a given policy field. Subsequently, Commission proposals are adopted by the College of Commissioners and these decisions are generally prepared at weekly meetings of their Private Office staff.

According to Commission Private Office staff who were interviewed, IA reports provide a valuable source of information about proposals put forward by Commissioners other than their own and IAs (and the related IAB opinions) are regularly discussed at the weekly preparatory meetings.

26

The Commission's impact assessment reports are considered by most users at the European Parliament and Council as relevant information in the legislative decision-making process

Commission impact assessment reports are transmitted to the European Parliament and Council together with the policy initiative

• 26. 

  IA reports should be made available to the European Parliament and the Council "fully and freely" so that the legislators can obtain a comprehensive view of the evidence base provided in the IA report32. During the period audited, all IA reports accompanying legislative proposals were forwarded to both institutions.

The Commission's impact assessments rarely discussed at meetings of European Parliament committees and Council working parties

• 27. 

  In the European Parliament and Council, it is recommended practice to discuss the Commission's IA whenever a proposal is submitted. However, current practice observed in both the Parliament and Council falls significantly short of this recommended approach:
• - 

  at the European Parliament, the Commission's IA reports are not systematically presented and discussed at committee meetings. The audit found that the Commission was only invited to present its IAs in exceptional cases (see Box 4 for a "good practice" example). An analysis of over 12 000 public European Parliament Committee documents of the 32 Inter-Institutional agreement on better law-making: in particular points 25 and 29;

Inter-institutional agreement on a common approach to IA (14901/05 JUR, adopted at 29.11.2005): in particular points 5, 6 and 13.

27

2004-09 parliamentary term revealed that only one document made explicit reference to a Commission IA, and

Box 4 "Good practice" example of an impact assessment being presented and discussed in an European Parliamentary Committee working group

At the Internal Market and Consumer Protection Committee, a working group meeting was dedicated to the presentation and discussion of the IA on the Proposed Directive on Consumer Rights. After a short presentation of the IA by a representative of the Commission, several MEPs commented on the substance of the IA report and made suggestions as to what aspects would merit further attention by the Commission. This example demonstrates that IAs can play a role in informing legislators and discussing IA reports with MEPs can provide relevant feed-back to the Commission with regard to the legislative proposal.

• - 

  internal Council guidance suggests discussion of the Commission's IAs at Working Party (WP) level. According to interviews carried out with Council officials, a formal discussion of the IA and where appropriate a Commission presentation is decided on a case-by-case basis (see Box 5 for a "good practice" example). An analysis of the Council public register for the 2004-2009 period showed that only four documents made explicit reference to a Commission IA.

28

Box 5 "Good practice" examples of impact assessments being discussed in Council Working Party

During the period audited, two IAs in the energy policy field were presented in the Council Working Party: the legislative package on the internal market for electricity and gas and the package of implementation measures for the EU's objectives on climate change and renewable energy for 2020. In both cases the discussion took place at the Commission's suggestion. It was not limited to one WP meeting but informed the debate at the Council throughout the legislative decision-making process.

Commission impact assessments provide relevant and informative documentation supporting the proposal to the European Parliament and Council

• 28. 

  Although the Commission's IAs are in almost all cases not formally referenced in European Parliament and Council official documents, users of the Commission's IA reports at both institutions stated that the reports are perceived as providing relevant additional information in support of the Commission's proposal (see Figure 5).

29

Figure 5 - Council Working Party Survey: The Commission's impact assessment role in informing decision-making within the Council

AgreeDisagreeDon't know The Commission's IA plays an important role in informing decision-making within the Council 53%35%12%

Always / Rarely /

Mostly / OftenNeverDon't knowIA has informed decisions or positions taken by the delegation 50%42%

8%

Delegations comment on the substance of the IA during the WP meetings 38%55%

7%

Delegations ask the Commission for further details on its assessment 35%57%

8%

Source : ECA Council Working Party Survey (2009). Users believe that impact assessments had a positive effect on the quality of EU legislation

• 29. 

  During the audit, users interviewed at the European Parliament and the Council generally indicated their support for IA. A large majority of respondents to the Council WP survey (68 %) felt that the IA reports they had reviewed had a positive effect on the quality of the final legal act. This positive view of the Commission's IA system was also corroborated by national experts on Better Regulation surveyed at the DEBR meeting in June 2009. More than 8 out of 10 respondents (85 %) saw IA as a contribution towards the EU policy objective of Better Regulation (see Figure 6).

30

Figure 6 - Surveys: Contribution of impact assessment to 'Better Regulation'

Directors and Experts for `Better Regulation' surveyed:Delegates in Council WP Is the Commission's IA system surveyed:

effectively For those IA reports I have leading to `Better Regulation'?personally analysed, I consider that they have positively contributed to the quality of the final legislative act.

100%

85 %

75%68 %

50%

25%

15 %20 %

12 %

0%

Note: 26 / 90 responses.

Source: ECA Council Working Party Survey/ Directors and Experts for `Better Regulation'

Survey (2009).

Interest groups and civil society organisations take impact assessments into account when contributing to the discussion of legislative proposals

• 30. 

  Public scrutiny of legislative proposals is of the utmost importance in relation to the policy objective of Better Regulation. The Commission's final IA reports are public documents available online to all interested parties once the related policy initiative has been proposed. This is international good practice (see Annex I).

31

• 31. 

  Interviews with stakeholder organisations representing interest groups and civil society showed that their representatives systematically took IAs into account. However, some indicated that they saw the IA reports as a policy justification of the Commission's proposal rather than an independent assessment of its possible impacts. This criticism was more pronounced with regard to IAs undertaken in the early years of the system's operation and has since become less of a concern.

Amendments to the initial Commission proposal are not subject to additional assessments of potential impacts

The Parliament and Council carried out eight impact assessments of amendments to the Commission proposal

• 32. 

  The institutional set-up of the European Union means that legislative decision-making differs from that in most Member States. According to the Treaty, while the European Parliament and the Council act as legislators, the Commission has the sole right of initiative. Under the Lisbon Treaty co-decision has been extended to more policy areas.
• 33. 

  In practice, nearly all Commission proposals are modified (sometimes to a significant extent) by the legislator during the legislative procedure. Therefore, the European Parliament and the Council have agreed that, where the co- decision procedure applies, they may, "... on the basis of jointly defined criteria and procedures, have IAs carried out prior to the adoption of any substantive amendment, either at first reading or at the conciliation stage". This

32

commitment was reaffirmed in the 2005 Inter-institutional agreement on a `Common approach to IA'.

• 34. 

  However, this aspect of the Inter-institutional agreement has only been partially implemented by the European Parliament and the Council. Between 2005 and 2008, the European Parliament adopted 377 acts of secondary legislation and the Council 1946. The audit found that during this period:
• - 

  the European Parliament had carried out IAs for amendments on seven proposals, and
• - 

  the Council had carried out one IA.
• 35. 

  The European Parliament and Council have recently reconfirmed their commitment to carrying out IAs on their own substantive amendments.

33

However, whereas the European Parliament has created the conditions for producing its own IA studies by putting a framework contract for the related procedures in place, this is not yet the case at the Council.

Impact assessments released by the Commission are not updated during the legislative procedure

• 36. 

  In addition to IAs carried out by the European Parliament or the Council, the inter-institutional agreements provide that the legislators can invite the Commission to update its initial IA in light of the amendments adopted by the European Parliament or the Council.
• 37. 

  Updating the Commission's IA report would be most relevant for significant pieces of legislation or in those cases where significant changes to the initial Commission proposal (for instance with regard to the intervention logic or instruments used) had been proposed for adoption.
• 38. 

  However, the audit found that IAs carried out by the Commission have not been updated during the period audited. This was not even the case for the "Services Directive", which was substantially altered during the legislative procedure (see Box 6).

34

Box 6 Example of the Commission's proposals for a "Directive on services in the internal market"

With its legislative proposal, the Commission aimed at reducing barriers to cross- border trade in services within the EU. The initial Commission proposal was presented in March 2004, and the related IA report in January 2004. On 22 March 2005, the European Council considered that the Commission needed to carry out a far reaching revision of its initial proposal to better preserve the European social model. On 5 April 2006 the Commission presented a modified proposal to the Council, including most of the modifications voted by the European Parliament, in accordance to the co-decision procedure in the first reading. On 29 May 2006 the Council approved the amended text. However, no revised IA was presented to reflect the impacts of the changes made to the initial proposal. The revised proposal finally adopted on 12 December 2006 by the European Parliament and Council, as the Directive 2006/123/EC.

Use of national impact assessments by Member States during the legislative procedure: United Kingdom "good practice"

• 39. 

  In addition, IAs of Commission proposals could also be performed by national administrations to build their government's negotiating position at the Council. The audit found, however, that in practice only the United Kingdom seeks systematically to use its national IA system for this purpose (see Box 7).

35

Box 7 - National impact assessments on Commission proposals - international "good practice"

The United Kingdom carries out national IAs on significant Commission initiatives to support the negotiating position of its Permanent Representation. It carries out its own IA, generally using data related to the United Kingdom, and thereby challenges the analysis provided in the Commission's IA. In Germany, the Bundestag has introduced a requirement to assess the implications of proposed EU legislation, but in actual fact this has not yet been implemented. In Poland, the authorities also have the mandate to prepare a national IA in the event of a significant legislative proposal, but this has not yet resulted in a formal IA.

The Commission's procedures for the preparation of impact assessments

• 40. 

  The Commission's procedures for IA determine the way in which IAs provide support for the Commission's internal decision-making. For this purpose, the audit examined whether the Commission's approach ensured:
• - 

  selection of those initiatives to undergo IA which have the most significant impact or are politically the most sensitive;
• - 

  consultation with stakeholders for the most important stages of the IA process; and
• - 

  quality review of its IA work.

36

Need for more transparency in the selection procedure and targeting of impact assessment work

• 41. 

  The Commission specifies, in its guidance material, the criteria according to which its departments (together with the Secretariat General) should determine the initiatives that require an IA. In particular, IAs should be targeted at those initiatives that have the most significant impacts on citizens, businesses and administrations or that are politically sensitive.

The Commission's selection of initiatives for which impact assessments are undertaken is based on a case-by-case analysis

• 42. 

  The decision on whether to carry out an IA is based on a case-by-case analysis. The Commission does not use quantifiable indicators (such as estimated monetised impact) to establish thresholds above which IAs need to be carried out. An approach using such thresholds is followed in some OECD countries (see Annex I). For IAs relating to legislative proposals outside the CLWP, the reasons for the selection of initiatives to be analysed are not made public.

Commission's monthly reporting on legislative activity does not indicate planned impact assessment work

• 43. 

  An overview of all legislative proposals under development within the Commission (forward planning report) is submitted each month to the European Parliament and the Council and is made public on the Europa website. However, this monthly report does not indicate the initiatives for which an IA is to be undertaken. This situation differs from the US system, where such initiatives are clearly highlighted in the overall catalogue of forthcoming regulatory activity. Similarly, in the United Kingdom, government departments

37

are required to publish a list of all planned regulatory proposals for the next three years which are to be accompanied by an IA (see Annex I).

• 44. 

  In the audited period, the roadmaps attached to the CLWP were the only documents that provided an indication for which initiatives IAs are to be carried out. IAs that were not related to initiatives included in the CLWP (i.e. the majority of IAs in 2008) were not visible beforehand outside the Commission.

Consultation is widely used as input for impact assessments, but not on draft reports

• 45. 

  The IA process should be transparent and draw on the expertise and views of others. Public scrutiny is as an effective verification mechanism to ensure that IAs address the most relevant issues, include all feasible policy options and provide a balanced view. Consultations enable the Commission to gather the opinions of interested parties and to take into account various points of view. These consultations should be carried out in accordance with the Commission's own standards in this matter.
• 46. 

  Consultations with stakeholders are widely used by the Commission, and the contributions submitted are increasingly made public on the 'Europa'

website. The audit found that the minimum consultation periods were generally respected.

• 47. 

  Consulting on draft IA reports is useful in ensuring that the analysis is complete, consistent and accurate. In particular, it provides a basis for identifying and quantifying potential costs and benefits, administrative burdens and problems with implementation and enforcement. However, the Court's

and its anticipated gross impact in terms of costs and/or benefits. This plan identifies for which implementing rules proposed by agencies IAs are to be carried out. The plan is made public.

38

analysis indicated that the Commission never consulted on draft IA reports.

Whereas consultations were sometimes used to identify possible policy options early in the process (21 % of cases of IAs reviewed), they virtually never concerned the Commission's preliminary assessment of alternative policy options. In the survey with Council WPs, 72 % of respondents said that an interim draft report should be made available some time before the Commission proposal was published (see Figure 7).

Figure 7 - Council Working Party Survey: An interim impact assessment report should be made available some time before the proposal

'don't know'

16 %'strongly agree'

26 %

'disagree' or 'strongly disagree'

11 %

'agree'

47 %

Notes: 91 responses.

Source: ECA Council Working Party Survey (2009).

• 48. 

  The Court notes that in some OECD countries like Australia, the USA and the United Kingdom draft IA reports are systematically published for

39

information and comment or draft versions of a proposed legal act are discussed with stakeholder organisations (see Annex I).

Recent improvements in the Commission's quality review of its impact assessment reports

Impact Assessment Board reviews quality of impact assessments within the Commission

• 49. 

  Since 2008, the IAB has reviewed practically all draft IA reports. Its quality control covers all aspects dealt with in an IA and all stages of the process, starting with a preliminary review of the roadmap. In cases where the IAB considers the IA report to be of insufficient quality, it can request resubmission.

In 2008, this happened in 43 out of 135 cases. Four of these 43 resubmitted reports even had to be revised and resubmitted a third time.

• 50. 

  According to the Commission staff interviewed in connection with the in- depth case studies, the creation of the IAB as an internal quality review body has put pressure on the DGs to present good quality draft reports. It has also added transparency to the system since all IAB opinions are published on the Europa web-site.

Effectiveness of quality review subject to timely availability of IAB opinion

• 51. 

  Considering that the Commission initiative has to go through inter- departmental consultation, decision-making by the Members of the Commission and translation, the IAB opinion can only have a substantive effect on the final version of the underlying initiative if the IAB review takes place early enough in the process. IAB recommendations are in many cases substantial and, if followed, would imply significant additional IA work and potential changes to the initiative. This in 2007 and 2008 represented a

40

challenge, since in one third of the cases analysed the time between the final IAB opinion and the adoption of the proposal was less than 6 weeks.

Nevertheless, as indicated in the IAB reports examined, the Board's recommendations were followed up at least to some extent in the final IA report.

The IAB does not have a mandate to require that IAs are carried out but advises the SG and the DGs in selecting initiatives for impact assessment

• 52. 

  The IAB does not have a mandate to require DGs to initiate IA work in respect of a particular proposal (see Annex I). Nevertheless, the IAB advises the SG and the DGs in identifying initiatives that should undergo IA. On this basis, in 2008, the SG initially saw a need for 55 additional IAs. After consulting with the DGs, the SG agreed that an IA was necessary in 21 of these cases.
• 53. 

  Moreover, the IAB cannot put the IA report (and the related legislative proposal) on hold: instead, the Secretariat-General may intervene at the inter- departmental consultation stage in relation to either document. However, no intervention specifically addressing the final IA report had been observed in the audited period.

Content and presentation of the Commission's impact assessment reports

• 54. 

  The results of the IA process are summarised in IA reports which should provide information based on which the legislators can decide on the most appropriate way to address the problem identified. Therefore, the Commission's IA reports are expected to provide a comprehensive and comparative assessment of the different feasible policy options. The audit examined whether:

41

• - 

  the IA reports provided a description of the issue to be addressed and explained how the benefits of a planned intervention would be attained;
• - 

  the information provided in the IA reports was presented in a user-friendly way and allowed a comparison of the alternative options analysed; and
• - 

  IAs provided the content that is relevant to policy makers (such as information on potential implementation problems, enforcement issues and administrative burdens).

In this context the Court recognises that any analysis of potential future impacts is necessarily uncertain, incomplete and simplified.

Need to improve further the presentation of how the intervention proposed is to achieve best the intended outcomes

• 55. 

  IA reports should provide a description of the problem at stake, specify relevant policy objectives and set out a range of appropriate options to address the problem identified. Further, they should illustrate how the intended outcomes and results can be achieved with the proposed delivery mechanism.

Impact assessment reports provide a sound description of the problem and specify policy objectives

• 56. 

  Since 2003, the Commission's guidelines have required IAs to identify the problem to be addressed and establish policy objectives corresponding to the problem and its causes. All IA reports reviewed in the Court's analysis contained sections on problem identification and policy objectives. As far as the in-depth cases are concerned, these sections provided an adequate overview of the problem and specified a set of reasonable policy objectives.
• 57. 

  This was corroborated by the enquiries with users of IA reports in the other institutions, Members States and stakeholder organisations. They reported that these sections of the IA reports often enabled them to understand better the Commission's reasons for initiating a proposal. According to the

42

interviewees, these sections were the most read parts of the IA reports. Also, 84 % of respondents to the Council WP survey "agreed" or "strongly agreed" that the problem description enabled them to have a better understanding of the reasoning behind the proposal.

Intervention logic not used to illustrate how expected benefits are to be attained

• 58. 

  In none of the cases reviewed in the Court's analysis did the IA provide an explicit illustration of the intervention logic underlying the initiative. As a consequence, the IA reports do not provide a standardised presentation of how the objectives and expected outcomes of the proposed intervention can be achieved with the intended delivery mechanisms and, with regard to expenditure programmes, the estimated budget.

There is room for improvement in the analysis of impacts and the presentation thereof

• 59. 

  IAs must be easily accessible and understandable to non-expert readers in order to be used in policy making. They should help to identify the policy option and delivery mechanisms which best address the problem, by providing a comparative analysis of the options. This can often be facilitated by providing information in quantitative and monetary terms. IAs should therefore specify the costs and benefits of proposals, how they occur and who is affected. According to the Commission's guidelines, all impacts should be quantified and monetised where possible and appropriate, based on robust methods and reliable data.

43

Main results and messages of impact assessment reports not always easy to gather

• 60. 

  Although the presentation of IA reports follows a common structure set out in the IA guidelines, it is not always easy to gather their main results and messages. According to the survey carried out as part of the audit, this is mainly explained by their length and technical nature, the complexity of the language used (see Figure 8) and by the other weaknesses described below.

Figure 8 - Council Working Party Survey: Main obstacles to an effective use of Commission impact assessments

Lack of summary33 %

Language availability35 %

Use of technical language37 %

Length of the document48 %

Note: 93 responses.

Source: ECA Council Working Party Survey (2009).

Analysing all economic, social and environmental impacts is a challenge 61. The Commission's self-imposed requirement to assess all significant economic, social and environmental impacts (the "three pillars") is ambitious.

The international comparison identified no other system where a similarly comprehensive approach was followed (see Annex I).

• 62. 

  The Court's analysis showed that, in practice, the Commission's IA work was asymmetric between the three pillars and between costs and benefits (see Figure 9). This reflects the fact that not all types of impacts are equally relevant for any particular initiative. According to the survey of the Council WPs, a majority of respondents thought that there was an appropriate balance between

44

the economic, environmental and social impacts of the different policy options (see Figure 10).

Figure 9 - What impacts were assessed - economic, environmental or social (selected DGs; period 2003-08)?

100% 86 %84 %84 %

80% 63 %62 %

s60%

o f

I A45 %

%40%

20%

0%

EconomicEnvironmentalSocial Type of impact

Cost Benefit

Source: ECA `Scorecard analysis' (2003-2008).

Figure 10 - Council Working Party Survey: In the Commission's impact assessment reports there is an appropriate balance between economic, social and environmental impacts

60%58 %

50%

40%

30%25 %

20%17 %

10% 0%0 %0 %

strongly agreeagreedisagreestrongly don't know disagree

Note: 74 responses.

Source: ECA Council Working Party Survey (2009).

45

Direct comparison between alternative options is sometimes difficult

• 63. 

  IA should provide a basis for comparison of alternative options. The Court's analysis found that this was the case for two-thirds of the IA reports examined.

In the remaining IA reports it was difficult to compare alternative options because of a lack of quantified impact analysis, insufficient use of methods to compare and present qualitative evidence and an asymmetry in the depth of analysis between different options.

• 64. 

  Firstly, impacts are often not quantified and monetised to facilitate a comparison of options (see Figure 11). This is also what respondents to the survey reported: nearly half (48 %) "disagreed" or "strongly disagreed" that an appropriate quantification and monetisation of the costs and benefits of the impacts of the different policy options had been achieved (see Figure 12).

Some OECD countries have a stricter requirement to quantify costs and benefits (see Annex I), although quantification is not always achieved (such as for example in the UK).

46

Figure 11 - What impacts were quantified or monetised (selected DGs; period 2003-2008)?

100%

80%

s60%53 %

o f

I A41 %

%40% 22 %23 %

20%12 %12 %

0% EconomicEnvironmentalSocial Type of impact

Cost Benefit

Source: ECA `Scorecard analysis' (2003-2008).

Figure 12 - Council Working Party Survey: In the Commission's impact assessment reports there is an appropriate quantification and monetisation of costs and benefits

60%

50%

40%38 %42 %

30%

20%14 %

10%6 %

0 %

0%

strongly agreeagreedisagreestrongly don't know disagree

Note: 71 responses.

Source: ECA Council Working Party Survey (2009).

47

• 65. 

  In those cases where quantification and monetisation is difficult, a robust analysis of qualitative aspects can help to compare alternative options. A qualitative comparison of options is done in all impact assessments. However, in terms of the methods used, the Court's analysis showed that `multi-criteria analysis' was used in 44 % of IA reports, `sensitivity analysis' in 12 % and `risk analysis' in 11 %. All these three methods feature in the IA guidelines as methods for comparing impacts.
• 66. 

  Finally, a consistent depth of analysis facilitates comparison between options. However, the Court's analysis revealed that, in approximately half of the cases reviewed, the depth of analysis for the different options was not balanced. Significantly more information was presented for a subgroup of the options and often only for the specific option that was later retained for the Commission proposal.

Availability of data for impact assessments remains a problem

• 67. 

  Many of the challenges with the analysis of impacts can be traced back to the problem of data availability. According to interviews with Commission staff, the timely collection of standardised and comparable data poses a particular problem. Differences in the availability and reliability of data between Member States compound this issue, as illustrated in the example below (see Box 8).

Box 8 Example of problems with data quality and availability

In the case of "Equal Opportunities" the IA, report acknowledges a "lack of reliable data on discrimination" and points to the fact that "collecting data on certain grounds of discrimination [...] is not undertaken systematically by the Member States". In order

48

to mitigate these problems the Commission complemented existing data with opinion surveys on people's perception and experience of discrimination.

• 68. 

  During the period audited, increasing use was made of external studies to gather data and this is further encouraged in the 2009 revision of the Commission guidelines. On the other hand, the audit found that internal sources such as the Commission's Joint Research Centre (JRC) or EUROSTAT (the Commission's statistics department) are not actively used to determine the availability of Member State specific data for IA purposes and to provide such data (for example in co-operation with national statistics offices).

The network of bodies like the Committee of Regions, the Economic and Social Committee or the High Level Group for Better Regulation are also not used to that end.

• 69. 

  In its recent conclusions on Better Regulation, the Competitiveness Council invited the Commission to co-operate with Member States at an early stage when gathering data for the preparation of IAs in order to take into account Member State specificities in its subsequent preparatory work.

Information on implementation aspects, enforcement costs and administrative burden can be further improved

• 70. 

  The effectiveness of an intervention and its costs are influenced by the way in which it is implemented and enforced by the Commission, and ultimately by the Member States. Missing information on implementation problems encountered with existing legislation is problematic since knowledge of such aspects is relevant to prevent similar problems after a review. This is why implementation aspects should be analysed in an IA, and information on the transposition and implementation of existing regulatory measures needs to be

49

collected. This information can be provided inter alia through the ex-post evaluation of existing policies and programmes.

Implementation aspects are not always sufficiently analysed in impact assessments

• 71. 

  While implementation aspects are addressed in the Commission IAs, the audit found that they do not, in all cases, give sufficient emphasis to implementation arrangements. The Court's analysis indicates that at least a reference to an implementation plan was provided in no more than approximately half of all reports reviewed. Several cases were found where important implementation aspects had not been adequately analysed in the IA reports. Once the legislative proposal is submitted to the European Parliament and the Council, a more detailed assessment of these aspects is summarised in specific implementation plans. Furthermore, correlation tables are used by the Commission to monitor legal acts through which specific EU provisions have been transposed into national legislation.
• 72. 

  Enquiries with users of IA reports outside the Commission (MEPs, Member State representatives and officials of the European Parliament and Council) indicated that the implications of transposition and implementation were of great relevance during the legislative decision-making process. This is also corroborated by the work of national audit bodies in Member States. The example in Box 9 illustrates possible consequences of an inadequate analysis of implementation aspects during the legislative decision-making process.

50

Box 9 Assessment of implementation aspects: Impact assessments on "Consumer rights"

The IA on "Consumer rights" did not contain a robust analysis of the European harmonisation effect that would ensue in the different Member States from the proposed Directive. In particular, the IA lacked an analysis of the differences between the existing national legislation and the new proposed harmonised rules on consumer rights legislation. This was also noted by the "Consumer Protection Working Group" composed of MEPs from the European Parliament's "Committee on the Internal Market". As a consequence, the initial Commission proposal is expected to be significantly modified during the legislative procedure to better take the specific situation of Member States into account.

• 73. 

  Commission staff in charge of IAs indicated difficulties in assessing such specific aspects for a European Union in which legislation is incorporated into 27 legal systems and implemented and enforced by an even higher number of national and regional bodies. They also pointed to the absence of national IAs for EU initiatives, which could provide a valuable source of information on potential obstacles to effective implementation resulting from specific national contexts and reliable estimates for the potential enforcement costs of a legislative proposal (see paragraph 41).

Ex-post evaluations of existing policies and programmes are not carried out systematically across all legislative areas

• 74. 

  A public intervention (and its actual impact) should be assessed through on-going monitoring and ex-post evaluation to improve further development of

51

interventions. To enable learning and feedback for future initiatives, ex-post evaluations would need to collect relevant information on compliance with legislation and the effectiveness of the rules as compared with the envisaged results initially set out in the IA.

• 75. 

  An understanding of how effectively EU legislation has been transposed and implemented is a significant element to be considered by the Commission whenever putting forward proposals for revised legislation. The audit showed that, during the period audited, the focus of the Commission's ex-post evaluations differed significantly from that of IA. An analysis of the Commission's overview of its 2007 evaluations shows that only 24 % of ex- post evaluations addressed issues related to the review of existing (sectoral or cross-cutting) legislation. This fact was also noted in a recent study.

Enforcement costs and administrative burdens not sufficiently quantified

Enforcement costs

• 76. 

  Drafting enforcement-friendly legislation is the most effective way to prevent excessive enforcement costs75. The Commission quantifies enforcements costs in the IA reports in the cases where they deem them relevant and significant. The Court observed that, in practice, whilst enforcement costs are referred to in many IA reports, quantified estimates of such costs were thoroughly analysed in only few cases. For EU expenditure 75 European Commission, A Europe of Results Applying Community Law; COM(2007) 502, p. 6.

52

programmes, such costs are either incurred at the Commission itself (in the case of direct management) or jointly with the Member States (in the case of shared or decentralised management). The in-depth review of the IAs related to two major EU expenditure programmes (i.e. the IAs on Cohesion and the 7

th RTD Framework Programme (FP7)) showed that in these cases enforcement costs had not been quantified in detail.

Administrative burden of existing legislation: reduction action programme

• 77. 

  In March 2007, under the EU Better Regulation initiative, the Commission and the Heads of State and Government of all the Member States agreed to reduce the administrative burden resulting from existing EU law by 25 % by 2012. The Commission subsequently initiated a systematic measurement of information costs for businesses resulting from EU legislation (see Box 10).

In August 2007, a "High Level Group of Independent Stakeholders on Administrative Burdens" was set up to provide advice to the Commission in this area.

Box 10 Commission's Action Programme aimed at cutting red tape

In October 2009, the Commission finalised its measurement of the administrative burdens that businesses incur in meeting EU legal obligations to provide information on their products or activities, either to public authorities or to private parties. This

53

study estimated the costs imposed by the 72 acts covered by the Action Programme and its 13 priority areas at 123,8 billion euro in 2005. The Commission has identified a total of 486 EU information obligations, and more than 10 000 national obligations which transpose or implement these EU obligations (of which more than 700 go beyond EU legal requirements). Based on this analysis, and in addition to measures that are under its own responsibility, the Commission has initiated a number of legislative proposals to remove or reduce administrative burdens: so far, the European Parliament and the Council have adopted 33 acts (with an estimated reduction of 5,7 billion euro) proposed by the Commission. A further 18 measures that could bring an estimated reduction of 30,7 billion euro are still pending. SCM rarely used to quantify potential administrative burdens of new legislation

• 78. 

  In the context of an IA for new legislation, the Commission's guidelines require that the Standard Cost Model (SCM) is used to quantify administrative burden, where this is expected to be significant. The SCM is a relatively simple and standardised methodology that is applied in many countries (see Annex I). The assessment of such costs is less complex than determining all economic, social and environmental impacts.
• 79. 

  Out of the 39 IA reports in 2008 examined by the Court, the Commission considered assessing administrative burdens in 35 cases. In 14 of these 35 cases the administrative burden was also quantified. However, the SCM was used in only four of these.

54 CONCLUSIONS AND RECOMMENDATIONS

Use of impact assessments for decision-making at the Commission, the European Parliament and the Council

• 80. 

  Since 2002, the Commission has put in place a comprehensive impact assessment system, which, for several of the aspects reviewed by the Court, can be considered as good practice within the EU. The Court also noted improvements in the Commission's impact assessments during the period audited. Particularly in recent years, impact assessments have helped to improve the Commission's ability to formulate its proposals. The audit found evidence that the IA procedures have become an integral element of the policy development process and that IA reports are actively used by decision-makers within the Commission.
• 81. 

  Regulatory quality is the responsibility of all three EU institutions involved in the legislative process. The Commission's IAs are systematically forwarded to the European Parliament and Council and users within both institutions consider them helpful when considering the Commission's legislative proposals. However, IAs are not updated during the legislative procedure as amendments are proposed. Once the initial Commission proposal is amended, neither the Commission, nor the European Parliament or the Council systematically analyse the impact of those amendments. Therefore, the estimated impacts of the final legislative act are not known if there have been substantial amendments.

The Commission's procedures for the preparation of impact assessments

• 82. 

  Transparency lends credibility, and the Commission's approach to impact assessment is strongest where it provides such transparency. Examples are the publication of the CLWP, roadmaps, full version of the final IA report and, since 2007, the IAB opinions. However, this report identifies the following weaknesses with regard to the Commission's procedures for selecting

55

initiatives to undergo IA, for consultation with interested parties and for the quality review of draft IA reports:

• - 

  the Commission did not indicate in advance all the initiatives which were to undergo an IA and, for IAs relating to legislative proposals outside the CLWP, the reasons for the selection of which initiatives were to be analysed were not made public;
• - 

  consultations with stakeholders were not carried out on draft IA reports.

As a result, the potential benefits of public scrutiny before the proposal is finalised (i.e. gathering the views of all parties concerned at an early stage and the increased acceptance of the resulting legislative proposal) have not fully materialised; and

• - 

  the IAB was found to contribute to the quality of IAs. However, in some cases, quality review took place too late in the process.

Recommendation 1: The Commission should enhance the impact assessment process by:

• - 

  providing an overview of all legislative initiatives (including the review of existing legislation) for which it intends to undertake an IA. A reasoned justification should be provided when an IA is not performed;
• - 

  publishing, for information and comment, interim documents (such as roadmaps and a draft version of the IA report); and
• - 

  ensuring that the IAB's quality review of IA work takes place on a timely basis.

56

Content and presentation of the Commission's impact assessment reports

• 83. 

  The Commission has adopted an ambitious approach to IA by aiming to analyse all significant economic, social and environmental impacts in one single assessment. IA reports should contain a description of the problem at stake, analyse all feasible policy options in terms of costs and benefits and provide an assessment of implementation and enforcement issues and an estimate of the administrative burden resulting from proposed legislation. It also sets out a framework for future monitoring and evaluation. This comprehensive assessment compares favourably to other IA systems.
• 84. 

  Overall, the IA reports were found to have complied with the requirements specified in the Commission's guidelines. However, the audit identified a number of weaknesses with regard to content and presentation of IA reports:
• - 

  the main results and messages of IA reports are not always easy to gather, including how an intervention is expected to attain its objectives.

Comparing the impacts of the various policy options presented in an IA report is sometimes difficult;

• - 

  difficulties in quantifying and monetising impacts can be traced back to the availability of data. This is an area in which the Commission does not yet fully exploit either internal capacities or those of Member States;
• - 

  implementation aspects are not always sufficiently analysed in impact assessments. Ex-post evaluations that could provide relevant information on existing policies and programmes for use in the context of an IA are often not available for legislative measures; and
• - 

  potential enforcement costs and administrative burden resulting from European legislation are not always sufficiently quantified. The SCM is rarely used to quantify potential administrative burdens.

57

Recommendation 2: The Commission should enhance the presentation of impact assessment reports and its content by:

• - 

  preparing IA reports in a way that facilitates the comparison of alternative options in terms of their estimated impacts by improving the quantification and monetisation of impacts and the presentation of qualitative analysis;
• - 

  developing a strategy to improve the quality of data available for IA, taking into account the specific situations in individual Member States;
• - 

  putting more emphasis on implementation aspects and making more use of ex-post evaluations of implementation of EU legislation as an input for the IA process; and
• - 

  analysing the enforcement costs of its legislative proposals in more detail and where administrative burdens are quantified consistently using the 'Standard Cost Model'.

Impact assessment in the EU institutions: an effective support for decision- making

• 85. 

  The Court considers that the Commission should give due consideration to the principles of clarity of objectives, simplification, realism, transparency and accountability when designing new and revising existing interventions. On balance, the audit has shown that, particularly in recent years, IA has been effective in supporting decision-making within the EU institutions. The European Parliament, the Council and the Commission may wish to consider the findings and recommendations set out in this report when revising their Inter-institutional agreements on "Better Regulation" and a "Common approach to IA. This will also provide an opportunity to take account of the changes

58

resulting from the Lisbon Treaty that came into force on 1 December 2009, such as the generalised right of initiative for a group of Member States and the role of national parliaments in EU decision making.

This report was adopted by the Court of Auditors in Luxembourg at its meeting of 19 May 2010.

For the Court of Auditors

Vítor Manuel da SILVA CALDEIRA President

59 ANNEX I COMPARISON OF SPECIFIC ELEMENTS OF THE COMMISSION'S SYSTEM WITH IA SYSTEMS ELSEWHERE

European Germany France The Netherlands United Kingdom USA Commission (federal level) (federal level)

Scope of IA system Legislative proposals* Always Always For major regulation For major regulation Always

No Implementing measures* For major regulation Always In some cases For major regulation Always For major regulation Other non-regulatory initiatives Yes

No

No

No Yes

No Clear threshold for applying IA to new regulatory proposals*

No

No

No

No Yes Yes Analysis of impacts Integrated approach to economic, social and environmental impacts Yes Yes

No

No

No

No Cost: quantitative assessment* In some cases Always For major regulation Always Always For major regulation Benefits: quantitative assessment* In some cases In some cases For major regulation Always Always For major regulation Use of Standard Cost Model (SCM) for administrative burden*

Yes Yes Yes Yes Yes

No

Quality control body External to the administration

No Yes (only admin. Yes (only admin. Yes (only admin. Burden)

No Burden) Burden)

No Can prompt IA work

No

No

No

No

No Yes Can block regulatory proposals*

No Yes

No

No Yes Yes Publication of IA work List of forthcoming IAs

No

No

No

No Always Always Interim IA documents

No In some cases

No

No Always Always Final IA reports Always Always

No In some cases Always Always Ex-post comparisons of the actual vs predicted impacts of regulations*

No Yes

No

No Yes Yes

• * 

  Sources: OECD, "Indicators of Regulatory Management Systems", Regulatory Policy Committee, 2009 Report; additional ECA country analysis. TRE009280EN08-09PP-DEC010-10VO_C1-RS-IMPACT_ASSESSEMENT-OR.DOC 19.5.2010 60 ANNEX II AUDIT APPROACH AND EVIDENCE COLLECTION METHODS

Identifying the audit subject and planning the audit fieldwork

• 1. 

  The European Court of Auditors identified in November 2007, in its annual work programme for 2008, the "Better Regulation" programme and impact assessments as

a relevant audit subject

• i. 

  In May and June 2008 a preliminary study was conducted which determined the scope of this audit. The audit fieldwork was carried out between October 2008 and July 2009, based on a specific audit planning memorandum and evidence collection plan.

Overview of evidence collection methods utilised during the audit

• 2. 

  First, the audit team reviewed relevant policy decisions, applicable guidelines of the European institutions (Commission, EP, Council), the OECD, the Member States and countries outside the EU (USA, Australia), previous evaluations of the Commission's IA system

ii, studies and academic literature in the fields of 'Better Regulation' and regulatory impact analysis

iii and participated in conferences related to the subject

iv.

• 3. 

  Moreover, the audit work comprised the following five work packages:
• - 

  a comparison of the Commission's IA system with systems elsewhere (WP 1);
• - 

  a quantitative analysis of the extent to which Commission initiatives during the period audited had been subject to IA, and how these IAs had been used by the European Parliament and the Council during the legislative procedure (WP 2);
• - 

  a 'scorecard' analysis of a sample of IA reports to verify whether the Commission had carried out its assessments in accordance with its own internal procedures and methodological guidance, together with an 'in-depth' review of a sample of IA reports (WP 3);
• - 

  enquiries with people involved in performing, reviewing and using the Commission's IAs both within and outside the Commission (WP 4); and 61
• - 

  surveys among members of Council Working Parties and the Directors and Experts of Better Regulation (WP 5).
• 4. 

  Detailed information on the content of the five work packages is provided hereafter. WP 1:

International comparison Specific aspects of the Commission's IA system were compared to approaches followed elsewhere on the basis of publicly available documentation (such as OECD documentation

v, academic literature and comparative studies by bodies responsible for 'better regulation' in the Member States

vi or co-funded by the European Commission

vii.

Corroborative evidence was gained through country visits to the USA and seven EU Member States (Belgium, Germany, France, the Netherlands, Portugal, Sweden and the United Kingdom). WP 2:

Quantitative analysis An analysis of the coverage of the Commission's IA system as a whole to verify to what extent impact assessments were applied to Commission initiatives published during the period audited and the items contained in the CLWPs. For this purpose, a database of almost 5 000 Commission initiatives was established and analysed.

The audit team also assessed the extent to which the EP and the Council used the Commission's IAs, on the basis of the publicly available documentation of the two Institutions (such as meeting agendas, minutes, references in resolutions, etc.)

Moreover, the audit comprised a comparative analysis of the scope and number of IAs and retrospective evaluations undertaken in 2008. WP 3:

'Scorecard'

analysis and 'in-depth'

review The 'scorecard analysis' was carried out for all IA reports produced by five Directorates-General: Energy and Transport (DG TREN); Regional Policy (DG REGIO); Employment, Social Affaires and Equal Opportunities (DG EMPL); Research; (DG RTD) Information Society and Media (DG INFSO). This corresponds to 115 (out of a total of 404) IA reports published between 2003 and 2008.

The scorecard was established on the basis of the Commission's IA guidelines, and took account of the requirements applicable at the date of publication of the IA report

• viii. 

  The checklists used by the United Kingdom NAO in a similar exercise related to the UK impact assessment system

ix and those used by a previous evaluation of the Commission's IA system

x were taken into consideration for the preparation of the scorecard. Further questions such as certain aspects related to the Impact Assessment Board's quality control were added. 62

In addition, an in-depth review was carried out to observe the complete life-cycle from the legislative proposal to the monitoring and evaluation of policies and programmes. The sample included not only recent cases (in particular to assess the IAB's role), but also reports dating back two years or more.

It included both reports for proposals with and without budgetary implications and IAs related to directly applicable legislative proposals (Decisions and Regulations) and those requiring transposition and implementation in the Member States (Directives)

xi.

Some additional cases from outside the remit of the five DGs in the sample were included in this review to obtain a corroborative view on the situation across the Commission as a whole. WP 4:

Interviews During the audit, around 190 people were interviewed to obtain their views on the Commission's IA system.

These were producers (Commission staff in the policy units responsible, the IA support units in the DGs, and the Secretariat General) and users within the EU Institutions (such as staff from the Commissioners' private offices, MEPs, EP and Council staff, and staff from the Permanent Representations).

They also included representatives from international and national organisations involved in 'Better Regulation' (such as the OECD and agencies in the Member States), other bodies and groups (such as the `High Level Group of National Regulatory Experts' and the 'High Level Expert Group on Administrative Burdens').

In addition, interviews were carried out with representatives of stakeholder organisations: Business Europe, Eurochambers, European Consumers' Organisation, European Environmental Bureau, European Federations for Transport and Environment, European Trade Union Federation, Social Forum, World Wide Fund for Nature (WWF). WP 5:

Surveys Two surveys were performed during this audit: the first, in May/June 2009, among the members of seven Council Working Parties on social questions, information society, energy, transport (maritime transport; land transport; transport, intermodal questions and networks), research, structural actions, and better regulation, concerning their perception and use of the Commission's impact assessments; the second, among the delegations of 25 Member States and Norway at the Directors and Experts of Better Regulation Meeting of 3-5 June 2009 in Sandhamn (Sweden). 63

International experts in the field of "Better Regulation" provided support throughout the audit

Advisory panel

• 5. 

  An advisory panel of experts in the field of 'Better Regulation' and impact assessment in November 2008 and March 2009 was set up to provide support throughout the audit
• xii. 

  The experts' main input was to contribute to the Court's methodology and analysis and endorse the overall structure for presenting the audit observations, conclusions and recommendations set out in this report.
• 6. 

  In particular, the experts provided their input to the specification of robust audit standards against which the observations in this report are made. These standards were developed on the basis of a meta-analysis of documents issued by the "Mandelkern group"

xiii, the OECD

xiv and the Commission

xv and of agreements between the latter and the European Parliament and the Council

xvi to specify a set of conditions (or 'good practices') which illustrate what should be expected from an effective impact assessment system

xvii.

Focus Group

• 7. 

  An important step in this audit was the review of the preliminary audit findings during a series of facilitated focus groups on 8 - 10 July 2009 in Luxembourg with international experts in the field of "Better Regulation" and IA. These focus groups constituted an integral part of the audit process and were designed to provide a forum for the critical review of the Court's preliminary conclusions and possible recommendations. Experts included representatives of the OECD, the World Bank, agencies in charge of IA work (or quality control) in the USA and several EU Member States, national audit bodies, but also academics, evaluators of IA systems and representatives of the 'High Level Expert Group on Administrative Burdens'.

Representatives of the Commission participated as observers. More information on the purpose of focus groups is provided hereafter.

What is a focus group?

Focus groups bring together a group of individuals with a common interest in the form of a 64

are asked so as to trigger a debate amongst the participants. They are particularly well- suited for obtaining a number of views on the same subject. The use of focus groups is common in marketing and, increasingly, in politics and opinion polling in order to elicit responses and reveal new perspectives from a group of people held to be representative of consumers or target groups. The participants need to have an interest in the subject and ideally there should be a mixture of backgrounds. During a focus group session it can be expected that the discussion will stimulate views that, at first, are diverse and even divergent.

One key role of the facilitator is therefore to manage the discussion in such a way that common ground can be found and views begin to converge, although full consensus of views is not the aim of a focus group discussion.

• 8. 

  The experts were asked by external facilitators to scrutinise the audit work: the standards that were applied, the facts and findings reported, the conclusions drawn from this and finally, whether the recommendations would help to mitigate the challenges observed by the audit. Their review significantly contributed to the quality, relevance and legitimacy of the overall outcome of the audit process. A report summarising the proceedings and the outcome of the discussions by the external facilitators was submitted to all participants, and the Commission, in August 2009
• xviii. 

  65

Notes continued ...

i ECA, 2008 Annual Work Programme, p. 3; ECA, Annual Activity Report, p. 35.

ii The Evaluation Partnership, Evaluation of the Commission's Impact Assessment System, Contract SG-02/2006 on behalf of the European Commission, April 2007.

iii Radaelli, C., What does regulatory impact assessment mean in Europe?, AEI-Brookings Joint Center for Regulatory Studies, USA, January 2005; Rowe, G.C., Tools for the control of political and administrative agents: impact assessment and administrative governance in the European Union in Hoffman, H.C.H and Turk, A.H. (eds), EU administrative Governance, Edward Elgar Publishing, p. 452, 2006; Renda, A., Impact assessment in the EU: The state of the art and the art of the state, CEPS Paperbacks, 2006; Mather, G. and Vibert, F., Evaluating Better Regulation: Building the system, City research series, European Policy Forum, London, 2006; Wiener, JB, Better regulation in Europe in Current Legal Problems, 2006, vol. 56; Kirkpatrick, C., Parker, D. et al, Regulatory impact assessment. Towards better regulation? The CRC Series on Competition, Regulation and Development. Edward Elgar Publishing, Inc.

Massachusetts, USA, 2007; Meuwese, A.C.M., Impact assessment in EU lawmaking, Kluwer Law International BV, The Netherlands, 2008; Open Europe, Out of Control?

Measuring a decade of EU regulation, 2009; Hertin, J., et al., The production and use of knowledge in regulatory impact assessment An empirical analysis in Forest Policy and Economics, 2009; Centre for European Policy Studies (CEPS); Policy-making in the EU past achievements and proposals for reform; 2009.

iv International Regulatory Reform Conference, Berlin, November 2008; Zwischenbilanz Nationaler Normenkontrollrat (NKR), Berlin, May 2009.

v OECD, Regulatory Impact Analysis (RIA) Inventory, April 2004; OECD, Regulatory Impact Analysis: Best Practice in OECD Countries, 1997; OECD, APEC-OECD Integrated Checklist on Regulatory Reform: Final Draft, 2005; OECD, Indicators of Regulatory Management Systems, December 2009.

vi DEBR, Report to the Ministers responsible for Public Administration in the EU Member States on the progress of the implementation of the Mandelkern Report's Action Plan on Better Regulation, Athens, 2003; EU Directors of Better Regulation Group, A Comparative Analysis of Regulatory Impact Assessment in Ten EU Countries, May 2004.

vii European Network for Better Regulation (ENBR) (see http://www.enbr.org/home.php);

Evaluating Integrated Impact Assessments (EVIA) (see http://web.fu- berlin.de/ffu/evia/contact.htm). The EVIA project was funded by the Commission under FP6. It was coordinated by the Freie Universität Berlin and the research partners were the Centre for European Economic Research (ZEW), the Institute for European Environmental Policy (IEEP), AVANZI, the Institute for Prospective Technological Studies (JRC-IPTS), the Centre for Regulatory Governance at the University of Exeter and the Institute for Environmental Studies at the Vrije Universiteit Amsterdam.

viii European Commission, Communication on 'Impact assessment' - 2002 IA Guidelines, COM(2002) 276 final; European Commission, 2005/2006 IA Guidelines, SEC(2005) 791 (Unless otherwise stated); European Commission, 2009 IA Guidelines SEC(2009) 92. 66

ix National Audit Office (NAO), Delivering High Quality Impact Assessments, 30.1.2009,

• p. 

  22-23.

x Evaluation Partnership Ltd (TEP), Evaluation of the Commission's Impact Assessment System, April 2007.

xi Council Regulation laying down general provisions on the European Regional Development Fund, the European Social Fund and the Cohesion Fund (SEC(2004) 924); Decision of the European Parliament and of the Council concerning the seventh framework programme of the European Community for research, technological development and demonstration activities (2007-2013) (SEC(2005) 430);

Directive of the European Parliament and of the Council amending Directive 2002/59/EC establishing a Community vessel traffic monitoring and information system (SEC(2005) 1514); Regulation on 'roaming' (SEC(2006) 925); Decision on effort of Member States to reduce their greenhouse gas emissions to meet the Community's greenhouse gas emission reduction commitments up to 2020 (SEC(2008) 85/3);

Proposal for a Directive on implementing the principle of equal treatment between persons irrespective of religion or belief, disability, age or sexual orientation (SEC(2008) 2180 and SEC(2008) 2181/2).

xii The members of the Advisory panel were Erwin de Pue (Director-General of Dienst voor de Administratieve Vereenvoudiging), Gisela Färber (Member of the German 'Normenkontrollrat' and Professor at DHV Speyer); Klaus Jacob (EVIA Project Coordinator at FU Berlin); Claudio Radaelli (University of Exeter); Alberto Alemanno (Law clerk at the European Court of Justice and Professor at HEC Paris) and Jeroen Nijland (Director-General of the Regulatory Reform Group).

xiii Mandelkern Group on Better Regulation, Final Report, 13 November 2001.

xiv

• - 

  OECD (1997) Regulatory Impact Analysis: Best Practice in OECD Countries, Paris; OECD (2005) APEC-OECD Integrated Checklist on Regulatory Reform: Final Draft, OECD, Paris.

xv European Commission, Communication on 'Impact assessment', COM(2002) 276 final;

European Commission, Communication on 'Towards a reinforced culture of consultation and dialogue - General principles and minimum standards for consultation of interested parties by the Commission', COM(2002) 704 final; Information note from the President to the Commission, Enhancing quality support and control for Commission Impact Assessments, The Impact Assessment Board, 14 November 2006; European Commission, 2005/2006 IA Guidelines, SEC(2005) 791; European Commission, 2009 IA Guidelines SEC(2009) 92.

xvi Inter-Institutional agreement on better law-making (2003/C 321/01); Inter-Institutional Common Approach to Impact Assessment (IA), 2005.

xvii These 'good practices' were first discussed at the second meeting of the 'Expert Advisory Panel' on 27 March 2009 and subsequently reviewed by the participants of the 'Focus Group' meeting on 8 to 10 July 2009.

xviii CM International, Meirion Thomas, Report on the Focus Groups held in support of Audit task 08TR2203, August 2009.

Replies of the Commission to the special report of the Court of Auditors

"IMPACT ASSESSMENTS IN THE EU INSTITUTIONS: DO THEY SUPPORT DECISION MAKING?" EXECUTIVE SUMMARY

I. Impact Assessment (IA) is a key tool for ensuring better regulation. In recent years, the Commission's aim has been to ensure that impact assessment becomes firmly embedded in the working culture of the institution and is delivering the expected results. The Court's findings encourage the Commission to continue with its approach and its recommendations will help to strengthen further the effectiveness of the Impact Assessment system. INTRODUCTION

• 5. 

  Second Indent There are few national IA systems which have the same level of ambition and scope as the system that the Commission has put in place.

O BSERVATIONS 26. First indent

Analysis of an increasingly wide range of policy alternatives: The Commission welcomes the finding that the number of alternative options presented in the IA reports increased and believes that the particularly positive development in the quality of the IAs prepared in 2007 and 2008 can be attributed to the successful quality control and support function of the Impact Assessment Board (IAB).

• 40. 

  The Commission recalls that in the Common Approach to Impact Assessment, European Parliament and Council have accepted the responsibility to carry out impact assessments on substantive amendments they make. They can also invite the Commission to do so.
• 41. 

  The Commission agrees that national impact assessments could effectively complement the Commission's impact assessments. They could inform discussions in Council on proposed changes to Commission proposals, and could also help Member States with transposition and enforcement issues.
• 44. 

  The Commission does not use quantifiable indicators to determine whether an IA should be carried out for a number of reasons. First, the integrated approach to IA requires that even when impacts cannot be easily quantified, or are limited in size, they should be analysed if they have important repercussions for specific groups, sectors or regions. Second, it would be difficult to develop ex ante criteria to reflect the diversity of initiatives, both legislative and non-legislative, for which IAs are carried out. Third, it would be difficult to make quantitative thresholds operational, because the necessary data often only become available during the process of doing the IA.
• 49. 

  The Commission is of the view that consulting on the draft IA report is not the only way of ensuring that the analysis is complete, consistent and accurate. Commission services consult and inform stakeholders at different stages of IA work using roadmaps, specific consultation documents/instruments (e.g. results of external studies) or Green or White Papers to ensure that this is the case.
• 50. 

  In the UK and USA (for the latter only for implementing measures) draft IA reports are published alongside a draft proposal. This practice is comparable to the EU situation where the Commission makes a proposal (it does not adopt a law), and the proposal and the IA report are published and transmitted to the legislators (Council and European Parliament). The public is able to engage in the ongoing debate given that both the proposal and impact assessment analysis are public.
• 53. 

  The Commission is aware that further efforts by the Commission services are necessary to ensure proper planning and sufficient time for the IAB scrutiny process.
• 55. 

  The role of the IAB has been further strengthened in the President's communication on the working methods of the Commission 2010-2014 (C(2010)1100 of 10.2.2010), which states that "in principle, the positive assessment of the Impact Assessment Board is required before an inter-service consultation can be launched".
• 60. 

  Impact assessments are a set of logical steps from problem definition to identification of the most appropriate form of policy action, and the Guidelines encourage services to present this logic as clearly as possible using, for example, problem trees, tables, maps and other illustrative techniques.
• 61. 

  The Commission considers that it is not always possible or proportionate to quantify or monetize impacts.
• 62. 

  The Commission is aware that it can facilitate further the use of impact assessments by ensuring that they are clearly presented and accessible. The measures it has taken are described in the Commission's reply to recommendation 2, indent 1).
• 63. 

  The Commission is convinced that an integrated approach to impact assessment, ensuring that all relevant economic, social and environmental impacts in terms of benefits and costs are analysed and presented together in one single document, is the most appropriate way of arriving at a balanced assessment of any potential legislative or non-legislative initiative.
• 64. 

  The Commission stresses that it does not expect the depth of analysis of the three pillars always to be the same in an impact assessment. For example, where there is no environmental impact, then no environmental analysis will be reported in the IA report.

The "asymmetry" as the Court terms it, is therefore normal.

• 68. 

  The principle of proportionate analysis applies to the entire IA process. The guidelines make clear that not all options have to be assessed with the same level of detail and that a screening process can be applied. For example, in some cases the scope of realistic/feasible ('high level') options is limited because of limits to EU competence, existing legislation, results of case law, or legal obligations. Such 'high level' options can be eliminated at an early stage of IA work as long as this is done transparently.
• 70. 

  Given the wide scope of Commission IAs and the fact that information needs are often specific and one-off, it will often remain the case that it is not possible to use bodies such as EUROSTAT for ad hoc solutions for data needs.
• 71. 

  The Commission would welcome more active participation of Member States in the Commission's IA data collection consultation processes, and more active IA work by Member States to feed into and complement the Commission's own IA reports.
• 72. 

  Commission impact assessments address implementation and enforcement issues at a general level during the IA process, examining in particular what has worked in the past, what failed or what requires correction. The concrete operational transposition and enforcement work can only be based on the adopted proposal itself, and not on the impact assessment. The Commission has developed specific instruments for this purpose.
• 73. 

  The Commission recalls that implementation plans are a separate instrument to impact assessment and are not part of the IA system. The IA guidelines do not require impact assessments to include an implementation plan.
• 75. 

  The Commission believes that national IAs for EU initiatives could provide a valuable source of information on how best to implement legislation in specific national contexts, and of reliable estimates for the potential enforcement costs of a legislative proposal. It thus would welcome if Member States were to prepare such impact assessments more systematically.
• 77. 

  The Commission's ex-post evaluation system has traditionally focused on spending programmes, and this is different from the approach to impact assessment. This situation has its origins in the fact that there is a legal obligation to carry out ex-post evaluations only for spending program measures (see Art. 27 of Financial Regulation).

The measures the Commission has already taken to focus its evaluation work increasingly on existing legislation are described in the Commission comments to recommendation 2, third indent.

• 78. 

  The principle of proportionate analysis applies to the assessment of enforcement costs. These need only be assessed if they are relevant, and quantified only if they are significant. A Commission analysis of the 2008 IA cases shows that this was done. As EU rules are in general implemented and enforced by Member State authorities, precise estimates of enforcement costs depend on each national system. Enforcement costs cannot therefore always be assessed comprehensively in a Commission IA. CONCLUSIONS AND RECOMMENDATIONS 83. In the Common Approach to Impact Assessment, Council and European Parliament accepted that it is their responsibility to assess the impacts of the substantive amendments they make. The Commission confirmed in the Second and Third Strategic Reviews of Better Regulation, that it will respond constructively and on a case by case basis to requests from Council and Parliament to expand on aspects of its original impact assessments.

The Commission underlines that its impact assessments, even when not updated by Council and European Parliament, cover essential elements of the final legal act and therefore remain relevant.

• 84. 

  First indent The Commission indicated in the roadmaps attached to its work programme which initiatives would have an IA and explained why for certain work programme proposals no IA was carried out .
• 84. 

  Second indent The Commission is of the view that consulting on the draft IA report is not the only way of ensuring that the analysis is complete, consistent and accurate. Commission services consult and inform stakeholders at different stages of IA work using roadmaps, specific consultation documents/instruments (e.g. results of external studies) or Green or White Papers to ensure that this is the case.

Recommendation 1 First indent The Commission has already identified two ways of enhancing the transparency of the IA planning process for the other institutions, Member States and stakeholders. First, in the 2009 IA Guidelines the Commission has made a commitment to prepare roadmaps for all proposals likely to have significant impacts/policy implications, and not only for the initiatives contained in its annual work programme. If an impact assessment is considered not to be necessary, the roadmap will explain why. As has always been the case, these roadmaps will be published. Second, following comments from the European Parliament in its contribution to the review of the Common Approach, the Commission will publish shortly after the work programme is adopted (normally at the beginning of each year) a full list of up-coming initiatives for which IAs are planned for that year. This will be updated regularly.

Recommendation 1 Second indent The Commission's public consultation on the draft IA guidelines in 2008 indicated that this is an area where a large number of stakeholders saw a need for improvement.

Consequently in the 2009 IA guidelines the guidance and requirements for consultations have been strengthened. As explained above, the Commission will produce roadmaps for all proposals likely to have significant impacts/policy implications. The 2009 IA Guidelines stipulate that Commission services should encourage stakeholders to examine roadmaps and give early feedback on the plans for impact assessments. The new Guidelines also introduced the requirement that stakeholders should be able to comment on a clear problem definition, subsidiarity analysis, description of the possible options and their impacts. The IAB will continue to check systematically the quality of the reporting of the results of the stakeholder consultation. The pro-active use of roadmaps available for all important Commission initiatives and the strengthened requirements for consultation together provide for improved transparency and stakeholder involvement. A consultation on a draft version of the IA report is therefore superfluous.

Recommendation 1 Third indent The Commission is aware that further efforts by the Commission services are necessary to ensure proper planning and sufficient time for the IAB scrutiny process. The 2009 IA Guidelines have introduced new requirements on this. The Guidelines also make clear that sufficient time needs to be foreseen in case the IAB asks for further work or for the revised impact assessment to be resubmitted. In the Communication on the Working Methods of the Commission for 2010-2014 it is explicitly highlighted that the planning of adoptions should take full account of the need to complete the impact assessment process in good time and that in principle the positive assessment of the IAB is required before an inter-service consultation can be launched. 86. First indent The Commission is of the view that comparisons between options are possible and that the idea of intervention logic is intrinsic to the impact assessment process. The Commission is nevertheless aware of the need to continue to improve the presentation of its reasoning, and the 2009 Guidelines encourage services to present the logic of the analysis as clearly as possible using, for example, problem trees, tables, maps and other illustrative techniques.

• 86. 

  Second indent The Commission agrees that difficulties in quantifying and monetising impacts can be traced back to the availability of data and that this is an issue which is a challenge for all advanced IA systems. It recalls that the often case-specific nature of the information needed and the likely administrative burdens limit the scope for creating more permanent data collection structures.

Recommendation 2 First indent.

The 2009 IA guidelines have further strengthened the guidance on how to compare options, in particular by providing more concrete guidance on cost-benefit and multi- criteria analysis. However, quantification is not always feasible or proportionate.

Recommendation 2 Second indent.

The Commission's experience shows that the kind of data needed for IAs is often very specific, and is not readily available from statistical offices or government authorities.

It will nevertheless consider whether practical ways exist to improve data availability and more generally how to encourage Member State authorities to be more active in providing necessary information.

The Commission is already developing cooperation with the Committee of Regions to use their network of consulting local and regional authorities in the framework of IA work.

Recommendation 2 Third indent The revised Impact Assessment Guidelines of January 2009 put increased emphasis on the need to take implementation issues into account when preparing new legislation, while recognizing that impact assessments can address implementation and enforcement issues only at a rather general level.

The Commission identified the need to do ex-post evaluation for legislation in the Communication on "Responding to Strategic Needs: Reinforcing the use of evaluation"

(SEC (2007) 213 of 21.02.2007). Since January 2009, the Commission's impact assessment system and evaluation system have been placed together in the Better Regulation Directorate of the Secretariat-General to ensure that the strategic and operational synergies between the two systems are fully exploited. President Barroso announced his intention in the political guidelines for the new Commission of September 2009 to intensify efforts on ex-post evaluation. The aim is to check whether the Commission's proposals deliver what they promise and to enable the Commission to revise and correct them where they fail to work as expected. Links between impact assessment and ex-post evaluation will be reinforced. The evaluation should be a reality check for the previous impact assessment.

Over time a full ex-post evaluation will become a requirement for the revision of important legislative acts.

Recommendation 2 Fourth indent To ensure further improvements in the analysis of administrative burdens, the Commission has placed the administrative burden experts along side the teams dealing with impact assessment and ex-post evaluation. This will ensure that the expertise which has been developed in implementing the administrative burden programme will be at the direct disposal of the impact assessment system. The Commission will provide a help desk function advising on all questions/issues relating to the assessment of administrative burdens, including appropriate use of the Standard Cost Model.

• 87. 

  The Commission welcomes the Court's finding that IA has been effective in supporting decision-making within the EU institutions particularly in recent years.

The European Parliament and Council have acknowledged in the Common Approach to Impact Assessment that it is their responsibility to assess any substantive amendments they put forward. The Commission launched the review of the Common Approach at technical level in April 2008 to identify concrete ways in which implementation could be approved.

• 22 okt '09
  COM(2009)544 - Actieprogramma ter vermindering van de administratieve lasten in de EU - Sectorale plannen ter vermindering van de lasten en acties in 2009
  
  
  
• 8 okt '08
  COM(2008)614 - Consumentenrechten
  
  
  
• 23 jan '08
  COM(2008)17 - Inspanningen van de lidstaten om hun broeikasgasemissies terug te dringen om aan de verbintenissen van de EG op het gebied van het terugdringen van broeikasgassen tot 2020 te voldoen
  
  
  
• 23 jan '08
  COM(2008)19 - Bevordering van het gebruik van energie uit hernieuwbare bronnen
  
  
  
• 23 jan '08
  COM(2008)16 - Wijziging van richtlijn 2003/87/EG teneinde de regeling voor de handel in broeikasgasemissierechten van de EG te verbeteren en uit te breiden
  
  
  
• 20 dec '07
  COM(2007)842 - Werking van de comités in 2006
  
  
  
• 19 sep '07
  COM(2007)529 - Wijziging van richtlijn 2003/55/EG betreffende gemeenschappelijke regels voor de interne markt voor aardgas
  
  
  
• 19 sep '07
  COM(2007)530 - Agentschap voor de samenwerking tussen energieregelgevers
  
  
  
• 19 sep '07
  COM(2007)531 - Wijziging van Verordening (EG) nr. 1228/2003 betreffende de voorwaarden voor toegang tot het net voor grensoverschrijdende handel in elektriciteit
  
  
  
• 19 sep '07
  COM(2007)532 - Wijziging van Verordening (EG) nr. 1775/2005 betreffende de voorwaarden voor de toegang tot aardgastransmissienetten