IMPACT ASSESSMENT IN THE EUROPEAN UNION

Andrea Renda, Senior Research Fellow, CEPS Bruges, 31 March-1 April 2011

2

INTRODUCTION: IMPACT ASSESSMENT

IA is seen as a useful tool in support of more efficient, effective, transparent and accountable policymaking

Internationally sponsored (OECD) and currently adopted in many EU countries and at EU level, within broader regulatory reform programmes

The focus and depth of analysis change remarkably from country to country

IA requires resources and transparency of regulatory processes: in many Civil Law countries it has failed so far

2 2

3

INTRODUCTION: IMPACT ASSESSMENT

Impact assessment is:

A tool to analyse the flow of new legislation

Aimed at

transparency: the rationale behind the legislative choices is explained to other institutions and external stakeholders

efficiency: ex ante cost-benefit of legislation can foster more efficient rules

accountability: proposing administrations and/or political decision-makers engage in “evidence-based policymaking”: they take responsibility against evidence provided by the IA document. 3 3

4

IA: MAIN STEPS

4

Analysis of status quo

Identification of need for regulation

Analysis of alternative policy options

Consultation

Collection of information

Identification of preferred option

Detailed cost-benefit analysis

Input to drafting 4

5

IMPACT ASSESSMENT: THE ORIGINS

Regulatory impact analysis (RIA) was introduced for the first time in 1981 in the United States, during the Reagan Administration

This was the result of the promise to bring regulatory relief from red tape to businesses during the electoral campaign

First years testified that:

RIA is costly

RIA must be coupled with public administration reform

RIA desperately needs consultation and oversight 5 5

6

RIA IN THE US (1)

6

1981: Reagan administration introduces RIA (EO 12,291) Does not apply to independent agencies (e.g. FTC, FCC, SEC) Estimated yearly saving: $10 billion

1985: The grand experiment Yearly OMB Report on the costs and benefits of Federal regulation Council of Competitiveness replaces Task Force on Regulatory Relief

1993: Clinton launches the NPR (EO 12,866) Eliminate 16,000 and modify 31,000 pages in the Federal Code Threshold for RIA: only “significant regulatory actions” (> 100M million

USD)

2002: RIA under George W. Bush (EO 13,258) Removal of Vice-President’s role in solving controversies between OIRA and

proposing agencies OIRA Prompt letters: from “consultant” to “adversarial gatekeeper”

6

7

RIA IN THE US (2)

7

OIRA To Congress

Agency

Better RIA

To Congress OIRA

Yes

no

Yes no

Agency

Preliminary RIA

Final RIA

Draft regulation

Consultation

7

8

RIA IN THE US (3)

8

In the US, RIA is based on net benefits calculation…

… but only government agencies are obliged to carry out RIA, i.e. RIA is mandatory only for most significant secondary legislation

Oversight is key: OIRA, GAO, CBO all contribute to quality assurance

The use of net benefits criteria as the sole pillar of analysis is not generally accepted in Congress, independent agencies, and in common law adjudication

8

9

REFLECTION ON THE US RIA SYSTEM

9

The US experience is a polar star in the landscape of regulatory reform, but

The US is a presidential democracy: RIA works as a way to control bureaucracy and fulfil electoral promises

Countries that have undergone no public management reform will have different needs when it comes to RIA

The scope of the US RIA system is limited to secondary legislation. This has been frequently overlooked in adopting RIA in other countries

The US RIA does not generally touch on distributional issues: when RIA is extended to primary legislation, this becomes an outstanding problem 9

10

IMPACT ASSESSMENT IN THE EU

10

1986: Business Impact Assessment System (BIAs) SMEs Task Force at the Commission Since 1989, competence of DG XXIII (now DG Enterprise)

2002: communication on Impact assessment Preliminary assessment + Extended Impact Assessment

2003: Inter-institutional agreement on better lawmaking Reinforced in 2005 with the “Common Approach to IA”

2005: Re-launch of the IA system (“growth and jobs”) New IA guidelines, more emphasis on economic analysis

2007: Appointment of the Impact Assessment Board 5 DGs involved, rooted in the SG

2010: Communication on smart regulation

10

11

IMPACT ASSESSMENT IN THE EU

11 11

12 12

NUMBER OF IAS PER YEAR

12 12

13

NUMBER OF IAS PER DG

13 13

14

IMPACT ASSESSMENT IN THE EU

Proportionality principle: criteria are (i) the significance of likely impacts, (ii) political importance and (iii) the situation in the context of policy development Communications which give broad policy orientations Non-legislative initiatives/white papers which set out

commitments for future action “Cross-cutting” legislative action “Narrow” legislative action in a particular field or sector, and

unlikely to have significant impacts beyond the immediate policy area.

Expenditure programmes Comitology decisions

14 14

15

PLATEAU-ING?

15

Very high initial expectations: it “should guarantee that we know the full costs and benefits of future legislation” (Verheugen, 2005)

Some bad episodes (REACH, Roaming regulation, Services Directive, 2006 Communication on the telecoms review, etc.)

Degree of quantification is still quite low

Lack of skills in some Commission DGs

Growing emphasis on “accounting” methods such as the Standard Cost Model

Still insufficient focus on implementation, compliance and enforcement phases of legislation

15

16 16

2007-2009: TOWARDS SUCCESS?

16

The appointment of the IAB seems to have led to an improvement of the overall quality of the IAs

Still, there are probably too many IAs, or resources are sometimes inefficiently allocated

Attention to admin burdens led to more quantification of costs (but not always benefits) in Commission IAs

Parliament and Council still have to be adequately involved

Member States are still mostly idle on the impact assessment side (but very active on admin burdens – not the same!)

Can we check this empirically?

16

17 17

ASSESSMENT OF IMPACTS

17 17

18 18

ASSESSMENT OF COSTS

18 18

19 19

ASSESSMENT OF BENEFITS

19 19

20 20

ASSESSMENT OF ALTERNATIVES

20 20

21 21

SOPHISTICATION?

21 21

22

REASON TO ACT, 2003-2009

23

OBJECTIVES AND INDICATORS, EX ANTE

24

OBJECTIVES AND INDICATORS, EX ANTE

25

OBJECTIVES AND INDICATORS, EX ANTE

26

EU: CLOSING THE POLICY CYCLE

The second Barroso Commission is trying to close the policy cycle: No ex ante impact assessment without a preliminary ex

post evaluation: will this change anything?

Multi-level governance: national Parliaments involved in ex post evaluation?

Emphasis on the definition of indicators for M&E in ex ante impact assessment documents

Emphasis on linking the definition of general, specific, operational and SMART objectives with M&E indicators

27

THE “SMART REGULATION” COMMUNICATION

“Evaluating the effectiveness and efficiency of EU legislation will improve the quality of policy-making and help to identify

new opportunities to simplify legislation and reduce administrative burdens …

… the Commission will … complement evaluation of individual pieces of legislation with … ‘fitness checks’ [that] assess if the regulatory framework for a policy area is fit for purpose and,

if not, what should be changed …

The aim will be to identify excessive burdens, inconsistencies and obsolete or ineffective measures to help to identify the

cumulative impact of legislation”

COM (2010) 543 11 October 2010

28

THE SMART REGULATION COMMUNICATION

Future steps Strengthen the ex ante consideration of transposition,

implementation, enforcement + use implementation plans

Close the cycle + review the evaluation guidelines

Publish planned evaluations on an ad hoc website

Complete “fitness checks” in environment, transport, employment/social policy and industrial policy, and identify new areas for 2011

Finalize the admin burdens reduction programme

Improve consultation on simplification and involve MS

Expand the mandate of the Stoiber Group

29

OUTSTANDING ISSUES

Efficient selection of proposals

Proportionality of analysis

Oversight

Transparency and accountability

Role of the Standard Cost Model

Efficient use of IA resources

Inter-institutional challenges

Convergence with national IA systems?

29 29

30

BETTER REGULATION: LOST IN…?

Formulation

Drafting/elaboration

Translation (!)

Consultation

Co-decision

Transposition

Implementation

Evaluation 30 30

31 31

EU IA: LESSONS LEARNED

Successful implementation of IA takes time

The design of the system is essential for its success.

First years led to some disappointment among stakeholders.

Demand for quality assurance led to the appointment of the IAB

Additional gatekeepers are entering the stage, from the EP to the ECA and the Stoiber group

RIA needs “sense of ownership”

Several DGs have yet to be fully involved in the process

MARKT, ENTR are working on their version of the IA guidelines

DG EMPL specific guidance for assessing social impacts 31 31

Annex : IA IN EU COUNTRIES

Andrea Renda, Senior Research Fellow, CEPS Bruges, 31 March 2011

33

CIVIL LAW COUNTRIES

RIA, where existent, serves widely different purposes and has different scope and depth

Often the scope is too ambitious or misunderstood (e.g. Italy)

Every paradigm shift (à la Kuhn) needs a cultural shift, and this is lacking in many countries

A “better regulation” community is slowly emerging

Parliamentary democracies are less prone to RIA

Administrative procedures, decision styles, agencies’ knowledge differ widely

RIA gives results in the long-term (next government?)

New mavericks?

Ireland, Sweden, France, the Netherlands… 33

34

GROUPING EU COUNTRIES (1)

Source: OECD, Government at a glance, 2009

35

GROUPING EU COUNTRIES (2)

Pioneers CEE countries with RIA

development

Late adopters Red-tape-dominated

Laggards

United Kingdom

Czech Republic, Estonia, Hungary, Poland, Slovakia, Slovenia

France, Ireland, The Netherlands, Sweden, Belgium (Flanders)

Austria, Belgium (Central gov), Denmark, Finland, Germany

Bulgaria, Greece, Italy, Lithuania, Latvia, Luxembourg, Malta, Portugal, Romania, Spain Cyprus

Recent research projects (ENBR, EVIA, but also OECD EU15) have highlighted a huge adoption-implementation gap, which seems to be hard to fill in the near future

36

LESSONS LEARNED

Political commitment remains essential

Oversight and coordination

Hierarchy

Gradual introduction or big bang?

Need to exploit the links with the SCM

37

CONCLUSION

RIA still faces difficulties outside the US: the European Commission is perhaps the only meaningful example of a successful RIA system

How do we solve the impasse?

Exploit the link with the Standard Cost Model at EU and national level (see below)

“First, make it good!”: too often RIA is grounded in very superficial economics, which eventually leads it to lose credibility with the general audience and academics

38

Reviewing the stock: available options

Process re-engineering De-legislation Codification Recasting Consolidation Review and sunset clauses Staged repeal Scrap and build Guillotine…

38

39

The Standard Cost Model

The SCM is being adopted in many countries to measure/reduce administrative burdens on businesses

Many countries have adopted the SCM on a macro basis – i.e. for a full baseline measurement…

Virtually all EU member states and the EU itself; Australia (Victoria)

Other countries have used it merely to scrutinise specific areas (e.g. business licensing)

The SCM promises significant welfare gains The Netherlands: 1.5% of GDP UK BRTF (2005): +£16 billion (+1%) Key assumption: all resources saved are allocated to a more

productive use

39

40

The Standard Cost Model

40

Direct financial costs Substantive compliance costs Administrative costs

Compliance costs

Monetary transfers to public administration

(taxes, duties, etc.)

Modification of productive process or

output

41

The Standard Cost Model

41

Administrative costs

Business administration costs

Administrative costs from central government regulation

Administrative activities that businesses may

continue if the regulations were removed

(‘Business as usual’)

Administrative activities that businesses only

conduct because regulation requires it, i.e. administrative burdens

The SCM measures all

costs from regulation, but

aims at reducing only

administrative burdens…

42

SCM: initial phases

Phase 0 (start-up) Identification of regulations and demarcation Start-up meetings/pre-consultation (also for ex ante

assessment of administrative burdens) Identification of appropriate methodology

Phase 1 (preparatory phase) Regulations are broken down into IOs, DRs,

administrative activities. Population and segments (e.g. firm size) Classification of origin (ABC or more detailed) Cost parameters Interview guide

42

43

IOs, DRs, administrative activities, etc.

43

Regulation

Information obligation 1

Data requirement 1

Information obligation 2

Information obligation n

Data requirement 2

Data requirement n

Activity 1

Activity 2

Activity n

Internal costs •Hourly rate •Time •Overhead

External costs •Hourly rate •Time

Acquisitions (Monetary value)

44

SCM: completing the baseline

Phase 2 (analysis) Data collection through empirical techniques

Phone Interviews Face-to-face interviews Business panels/focus groups Expert assessment

Standardisation of time and resource consumption per business per segment

Identification of “normally efficient business”

Phase 3 (extrapolation/reporting)

Validation and extrapolation to national/regional level Reporting or integration with RIA

44

45

Example: the EU SCM In 2007, the EU launched an ambitious action programme on

measuring and reducing ABs Expected reduction: 25% of GDP Expected impact: +1.4% of GDP

This is the first example of application of the SCM in a multi-level governance framework

The SCM was also applied to the EU IA model, bridging ewx ante and ex post analysis of legislation

All Member States took it on board Much easier to communicate

Does not require public management reform 45

46

Interim results

46

Results from the first round of

the EU measurement in 13 priority areas

47