Capacity Building On Regulatory Review System Design Module

1
Capacity Building On Regulatory ReviewSystem
Design ModuleDesigning a RIA system that works
effectively21-25 March 2011

• Rod Bogaards
• Productivity Commission
• Australian Government

2
Some disclaimers!

• These are my own personal views although
  reference is made to Productivity Commission work
• While the Australian RIA system is highly
  regarded by international standards, I will focus
  on some of its pitfalls no RIA system is
  perfect!
• How can we improve or enhance RIA institutional
  and analytical regulatory settings - that are
  already reasonably robust - to generate better
  and more consistent regulatory outcomes?

3
Every political party believes in the idea of
better regulation. And yet every political party,
once in government, fails to achieve better
regulation.

• Boyfield (2007)

4
Foundations of an effective RIA system

• Adequacy criteria
• Gate-keeper
• Transparency and accountability

5
Objectives of Regulatory Impact Analysis

• RIA aims to informs political decision makers on
  whether to regulate and how to regulate
• It aims to communicate information about the
  expected effects of regulatory proposals that can
  be used by decision makers

6
Good regulation addresses market failures

• While allowing for
• greater consumer choice, convenience and lower
  prices
• enhanced flexibility and competitiveness
• increased productivity
• greater transparency, accountability and public
  confidence in governance

7
Good regulation checklist

• Minimum necessary to achieve objectives
• Not unduly prescriptive
• Accessible, transparent and accountable
• Integrated and consistent with other regulations

8
Good regulation checklist

• Communicated effectively
• Mindful of the compliance burden imposed and
• Enforceable

9
What is Regulatory Impact Analysis?

• RIA processes follow a formalised sequence of
  steps to establish whether particular regulatory
  proposals would be in the communitys interest -
  by delivering a net benefit to the community
• RIA processes do this by taking a community-wide
  perspective of their effects, to ensure that the
  benefits to society of a regulation (broadly
  conceived) are greater than the costs (also
  broadly conceived)

10
7 Steps in a Regulation Impact Statement

1. Identifies the problem the regulation seeks to
   address
2. Outlines the objectives of government action
3. Identifies a range of feasible options for
   addressing the problem
4. Assesses the costs and benefits of the feasible
   options
5. Documents community consultation
6. Proposes a recommended option
7. Outlines implementation and review mechanisms

11
Practical operation principles

• All seven elements of RIS adequately addressed
• Benefits should outweigh costs and (usually) net
  benefits maximised
• Analysis commensurate with size of likely impacts
• Objective, balanced public statement, not
  advocates document

12
The adoption of RIA processes is now widespread

• RIA have proved popular with governments trying
  to improve the quality of regulation
• in 1980, only 2 or 3 countries were using RIA
• in 2000, 14 of 28 OECD countries had adopted RIA
  programs
• in 2010, all 31 OECD countries and many non-OECD
  countries had mandated the use of RIA

13
Trend in RIA adoption across OECD countries
(1974-2010)
14
Australia was an early adopter of RIA

• In 1985, Australia was one of only eight OECD
  countries with formal requirements for regulatory
  impact analysis
• In 1995, RIA was extended to cover the
  development of all national standards
• In 2009, all Australian states and territories
  had their own individual RIA processes to cover
  state government regulations

15
What should a good RIA process deliver?

• A well designed and implemented RIA process can
  improve the quality of new regulations
• by demonstrating that a non-regulatory option is
  a better solution to the problem (or that the
  status quo is preferable)
• by identifying the most effective and efficient
  design elements to build into the new regulation,
  thereby increasing the benefits and/or reducing
  the costs

16
What should a good RIA process deliver?

• A well designed and implemented RIA process can
  improve the quality of new regulations
• by building stakeholder support for proposals
• by reducing the risk of over-regulation
• by influencing over time the regulatory culture
  within government agencies

17
Is RIA quality very good? OECD view

• The OECD (2009) has identified a number of issues
  with the performance of RIA
• not well integrated in the decision-making
  process
• not applied to the most significant regulatory
  proposals
• fails to fully assess the costs and benefits of
  proposals
• no serious consideration of feasible alternatives

18
Findings of the OECD EU 15 Country Reviews

• The reviews revealed major weaknesses across most
  countries
• Relative lack of integration of RIA in the policy
  process, assessments are done too late,
  consultation is often not robust and oversight
  needs more teeth methodological challenges
• Late timing of impact assessments is a widespread
  issue
• Setting up an oversight function in a single
  central unit has been difficult for many
  countries
• In order to force change a few countries are
  making RIA a legal requirement with sanctions
  for non compliance

19
What is the quality of RIA? Has it improved
regulatory outcomes? US experience

• A US study said most agencies failed to quantify
  costs and benefits, discuss alternatives, use
  consistent analytical assumptions, report their
  results clearly or make them accessible (Hahn
  2000)
• A more recent US study said although there is
  some evidence economic analysis can improve the
  benefit-cost ratio of regulations, there is
  insufficient evidence that economic analysis of
  regulatory decisions has actually had any
  substantial impact (Hahn and Tetlock 2007)

20
What is the quality of RIA? Has it improved
regulatory outcomes? UK experience

• A study of the UK RIA process also found the
  standard of RIA to be poor, a bureaucratic
  sham, treated as a bolt-on extra designed to
  justify a regulation rather than being used to
  shape and inform policy formulation (Boyfield
  2007)
• A study of UK Government RIAs by the British
  Chamber of Commerce concluded, we question
  whether RIAs have changed the political reality
  that ministers introduce regulations because they
  want them as distinct from being able to justify
  them (Ambler et al 2007)

21
What is the quality of RIA? Has it improved
regulatory outcomes? EU experience

• A study of RIA processes and practices in the
  European Union and all Member States found
• In almost all cases we have examined, there is
  a large gap between requirements set out in
  official documents and actual Impact Assessment
  in practice typically assessments are narrow,
  partial and done at a late stage. In many
  countries, a large share of proposals is not
  formally assessed or is assessed with a tick box
  mentality (Jacob et al 2008)

22
How effective has RIA been in improving
regulation? Australian experience

• In Australia the results of RIA requirements have
  also been somewhat mixed
• 21 years after the RIA process was established,
  the 2006 Taskforce on Reducing Regulatory Burdens
  on Business pointed out the dramatic growth in
  the volume of regulation
• Quality of regulation was also at issue as
  complaints about growing regulatory burdens led
  to the Regulation Taskforce inquiry

23
Findings of the 2006 Regulation Taskforce

• Too much regulation imposing an unnecessary cost
  on business due to
• unclear or questionable objectives
• failure to target the regulation too blunt or
  disproportionate to the size of the problem
• undue prescription
• excessive reporting or paperwork requirements
• overlap, duplication or inconsistency with other
  regulation, either within or between
  jurisdictions
• poorly expressed and confusing terms
• unwarranted differentiation from international
  standards

24
How did this happen?

• Rising phenomenon of risk aversion in society
• Over reliance by governments on the development
  of regulatory solutions regulate first, ask
  questions later culture
• The requirements for good regulatory process had
  not been effectively discharged
• Taskforce said all of the above

25
Beginning of a new phase of regulatory reform

• Unless the underlying reasons for regulatory
  failures are addressed the regulatory problems
  will simply re-emerge
• The Regulation Taskforce recommended six
  principles of good regulatory process which were
  endorsed by the Australian Government

26
6 Principles of good regulatory process

1. Governments should not act to address problems
   through regulation unless a case for action has
   been clearly established
2. A range of feasible policy options including
   self-regulatory and co-regulatory approaches
   need to be assessed within a cost-benefit
   framework

27
6 Principles of good regulatory process

• Only the option that generates the greatest net
  benefit for the community (taking into account
  economic, social, environmental and equity
  impacts) should be adopted
• Effective guidance should be provided to
  regulators and regulated parties to ensure that
  the policy intent of the regulation is clear, as
  well as what is needed to be compliant

28
6 Principles of good regulatory process

• Mechanisms such as sunset clauses or periodic
  reviews need to be built in to legislation to
  ensure that regulation remains relevant and
  effective over time
• There needs to be effective consultation with
  regulated parties at the key stages of regulation
  making and administration

29
Changes to the RIA process in response to
Regulation Taskforce recommendations

• Preliminary assessments for all regulatory
  proposals to determine the level of regulatory
  impact analysis required
• A formal assessment of business compliance costs,
  using the Business Cost Calculator (or an
  approved equivalent), and greater use of
  cost-benefit analysis generally

30
Changes to the RIA process in response to
Regulation Taskforce recommendations

• A clear statement in the RIS adequacy criteria
  that the RIS should demonstrate that
• the benefits of the proposal to the community
  outweigh the costs
• the preferred option has the greatest net benefit
  for the community, taking into account all the
  impacts
• A whole-of-government consultation policy

31
Changes to the RIA process in response to
Regulation Taskforce recommendations

• Improved gate-keeping arrangements that prevent a
  regulatory proposal from proceeding to Cabinet
  (or other decision maker) unless an adequate RIS
  has been prepared
• Post-implementation reviews are required within
  two years of implementation when a proposal
  proceeds to a decision maker without an adequate
  RIS

32
Changes to the RIA process in response to
Regulation Taskforce recommendations

• The existing Office of Regulation Review was
  renamed the Office of Best Practice Regulation
  (OBPR) to reflect a new focus to assist
  agencies to develop regulatory best practice
• A specialised cost-benefit analysis unit was
  created in the OBPR to provide advice and support
  to agencies preparing RIS

33
Regulation Taskforce had high expectations that
it would create a better RIA system

• In the Taskforces view, the single most
  important way of strengthening compliance with
  the principles of good process would be for
  government to adhere to a rule that regulatory
  proposals that fail to meet the RIS requirements
  will not be permitted to proceed to consideration
  by Cabinet or other relevant decision maker,
  except in specially defined circumstances.
  Together with stronger analytical requirements,
  this would ensure the processes were taken more
  seriously and at an earlier stage

34
RIA compliance rate following Regulation
Taskforce reforms

• Compliance with Australian Government best
  practice regulation requirements is measured by
  compliance with the requirement to prepare
  adequate RIS (and BCC reports)
• To be assessed as adequate, a RIS must contain a
  degree of detail and depth of analysis that is
  commensurate with the size of the potential
  impacts of the proposal
• The OBPR used the criteria in the BPR Handbook
  (2007) to assess whether each element of a RIS is
  adequate

35
RIA compliance rate following Regulation
Taskforce reforms
Type of RIA 2007-08 2008-09 2009-10
RIS 43/48 (90) 45/53 (85) 63/75 (84)
BCC 7/7 (100) 4/4 (100) 2/2 (100)
Exceptional Circumstances 3 6 4
Preliminary Assessments 753 662 823
36
What does a high RIS compliance rate mean?

• Does it mean the gate-keeper is doing a good job?
  
• A low compliance rate from a failure to conduct
  RISs indicates the RIS process is being evaded
• But a high compliance rate does not necessarily
  tell us much about the quality of the RIS
  supporting a regulatory proposal

37
Role of the Australian gate-keeper

• Independent (but within the Department of Finance
  Deregulation)
• Advises on whether a RIS is required
• Examine/advise on adequacy of RISs
• Provide training and advice on RIS process
• Reports annually on RIS compliance

38
Are gate-keepers prone to Type I errors?
Pass Fail
Adequate RIS Correct Decision Type II Error or false negative
Inadequate RIS Type I Error or false positive Correct Decision
39
Reasons why gate-keepers could be prone to Type I
errors

• Political pressure - from government departments
  or Ministers offices to pass a RIS can be
  intense
• And as the penalties for non-compliance were
  raised in 2007 since the gate-keeper could
  effectively veto regulatory proposals, preventing
  them from going to decision makers - the pressure
  to pass RISs may have increased even more

40
Reasons why gate-keepers could be prone to Type I
errors

• Insufficient transparency publication of RIA
  (and the OBPRs adequacy assessment) potentially
  occurred up to 18 months after a decision to
  proceed with a regulatory proposal had been made
  by government
• this may have provided the gate-keeper with poor
  incentives to perform its role and enforce the
  RIA process properly

41
How does Australias 2007 RIA model rate with
respect to the key drivers?

• Strong adequacy criteria
• Strong gate-keeper
• Weak transparency and accountability

42
But did the 2007 changes lead to better RIA
system?

• The existence of robust adequacy criteria and
  gate-keeper are necessary criteria for an
  effective RIA system but they are not
  sufficient criteria
• The ultimate effectiveness of any RIA system
  depends on the level of transparency and
  accountability in the system

43
A case study of a proposed regulation

• Lets examine a regulatory proposal and see how
  the RIS for the proposal engages with the policy
  development process

44
Productivity Commissions 2009 Annual Review of
Regulatory Burdens on Business

• The Productivity Commission proposed a
  strengthening of the best practice regulation
  requirements and made a number of key
  recommendations
• Industry submissions commented on the ongoing
  failure of government to properly assess the
  risks, costs, and benefits of regulation and
  called for further scrutiny of the process by
  which regulation is developed

45
Productivity Commissions 2009 Annual Review of
Regulatory Burdens on Business

• The concerns raised suggested there remains scope
  for the Australian Government to do more to
  improve regulatory outcomes, through
• more systematic and rigorous application of the
  existing best practice regulatory process
• refinements and enhancements to aspects of the
  design and implementation of this process
• increasing the transparency and accountability of
  the current regulatory process

46
Productivity Commissions 2009 Annual Review of
Regulatory Burdens on Business

• The Commission recommended the Australian
  Government should develop a central online
  register of regulatory impact analysis that would
  include
• RISs (and importantly, the OBPRs adequacy
  assessments) published at the time government
  decisions are made public
• post-implementation reviews (and the OBPRs
  adequacy assessments) published at the time these
  reviews are made public

47
Productivity Commissions 2009 Annual Review of
Regulatory Burdens on Business

• The Commission also recommended the Australian
  Government
• incorporate a consultation RIS in its
  regulation making process
• establish an independent public review of the
  current best practice regulation requirements

48
OECDs 2010 Review of Regulatory Reform in
Australia

• The OECD said Australia rates highly among OECD
  countries on the design and performance of its
  RIS process
• But said there are some potential improvements
  that could strengthen the Australian Government
  RIS process

49
Requirements for RIA processes used by central
governments2005 and 2008 (RIA 1)
Note This figure summarises information about
the existence of key elements of RIA processes in
OECD member countries. It does not offer
information on the quality of specific RIAs.
Questions Weights
a) Is regulatory impact analysis (RIA) carried out before new regulation is adopted? a) if no0, in some cases1, always2
c) Is a government body outside the ministry sponsoring the regulation responsible for reviewing the quality of the RIA? c) if yes, weight3
e) Is there a clear "thresholdfor applying RIA to new regulatory proposals? e) if yes, weight2
h(i) Is RIA required by law or by a similarly strictly binding administrative instrument? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
h(ii) Is RIA required for draft primary laws? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
h(iii) Is RIA required for draft subordinate regulations? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
h(iv) Are regulators required to identify the costs of new regulation? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
If yes h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
Is the impact analysis required to include the quantification of the costs? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
h(v) Are regulators required to identify the benefits of new regulation? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
If yes h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
Is the impact analysis required to include the quantification of the benefits? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
h(vi) Does the RIA require regulators to demonstrate that the benefits of new regulation justify the costs? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
h(vii) Are RIA documents required to be publicly released for consultation with the general public? h(i) to h(vii) if no0, in other selected cases1, only for major regulation1, always2
k) Are ex post comparisons of the actual vs predicted impacts of regulations made? k) if yes, weight1
l) Is there an assessment of the effectiveness of RIA in leading to modifications of initial regulatory proposals undertaken? l) if yes, weight1
Source Question 10 / OECD Regulatory Management
Systems Indicators Survey 2005 and 2008.
www.oecd.org/regreform/indicators
50
Extent of RIA processes2005 and 2008 (RIA 2)
Note This graph summarises information about the
extent of RIA processes in OECD member countries.
It does not gauge whether these processes have
been effective.
Questions Weights
d(ix) Is the RIA required to include assessments of other specific impacts Impacts on the budget, impacts on competition, impacts on market openness, impacts on small businesses, impact on specific regional areas, impact on specific social groups, impact on other groups (not for profit sector including charities), impact on the public sector d(ix), Impacts on the budget, competition, market openness, small businesses, specific regional areas, specific social groups, the public sector if no0, in other selected cases1, only for major regulation1, always2 d(ix), Impact on other groups (charities, not for profit sector) if no0, in other selected cases0.5, only for major regulation0.5, always1
e) Is risk assessment required when preparing a RIA? e) if no0, in other selected cases1, only for major regulation1 always1
For all regulation, for health and safety regulation, for environmental regulation? if yes, weight1
If yes
Does the risk assessment require quantitative modelling?
f(i) Does the RIA require regulators to explicitly consider compliance and enforcement issues when preparing new regulation? f(i) if yes, weight1
f(ii) Are reports prepared on the level of compliance with the above RIA requirements? f(ii) if no0, ad hoc basis1, regularly2
f(iii) Are these reports published? f(iii) if yes, weight2
Source Question 10 / OECD Regulatory Management
Systems Indicators Survey 2005 and 2008.
www.oecd.org/regreform/indicators
51
OECDs 2010 recommendations to improve
Australias RIS process

• Ministerial certification of each RIS to
  establish greater accountability at Ministerial
  level
• RIS training could be extended to Ministerial
  offices
• OBPR should report on compliance by government
  agencies with the obligation to quantify costs
  and benefits of regulatory proposals
• the RIS should be published in draft format for
  public consultation a two stage approach
• the Australian National Audit Office should
  periodically review the quality of RISs Sed
  quis custodiet ipsos custodes?

52
Recent changes to the Best Practice Regulation
Handbook Some positive

• Development of a central online RIS register (but
  only includes a Yes adequacy assessment by OBPR
  but no explanation of why Yes)
• Earlier signalling of non-compliance with the RIS
  process (previously took up to 18 months for the
  public to find out)
• Post-implementation reviews to be published on
  the central online RIS register (but adequacy
  assessment by OBPR not reported online)

53
Recent changes to the Best Practice Regulation
Handbook Some questionable

• That no explicit comparison of options will be
  required in a RIS
• That Cabinet may direct government agencies
  regarding which options to examine in a RIS
• That election commitments to introduce
  regulations will only be subject to a modified
  RIS

54
Recent changes to the Best Practice Regulation
Handbook Some questionable

• That a RIS no longer has to demonstrate that the
  preferred option has the greatest net benefit for
  the community, taking into account all the
  impacts (economic, social, environmental and
  equity)
• That a RIS can be modified after the decision
  makers consideration but prior to publication to
  allow the inclusion of an option not considered
  in the original RIS but without the public
  knowing

55
Recent comments by the Business Council of
Australia on 2010 RIA model

• changes to the RIS process may inhibit the
  independence and robustness of the policy
  analysis and advice received by Ministers from
  the bureaucracy, and this will have major
  implications for the outcomes of policy and its
  effectiveness
• These reforms may also reduce the effectiveness
  of the OBPR as an assessor of RISs narrowing
  the requirements for analysis in RISs means that
  the OBPR will have less ability to assess them as
  being non-compliant with the process ... further
  diminishing the power of the OBPR to reject RISs

56
How does Australias 2010 RIA model rate compared
to the previous model?
2007 RIA model 2010 RIA model
Adequacy criteria Strong Weaker?
Gate-keeper Strong Strong
Transparency and accountability Weak Weaker?
57
A perverse outcome of Australias strong
gate-keeper model?

• Giving the gate-keeper the power to veto
  regulatory proposals in 2007 - if the RIS process
  was not followed may have led to increased
  pressure on the gate-keeper to pass RISs
• Gate-keepers do not need a veto power to be
  effective so long as their adequacy assessments
  are made transparent in a timely manner

58
How can we improve the effectiveness of the RIA
system?

• Transparency and accountability is the most
  important driver of an effective RIA system
• A RIA system must take account of, and improve,
  the incentives that face Ministers, departments
  and gate-keepers
• This requires a significant effort to improve
  transparency and accountability the
  inconvenient driver of an effective RIA system!

59
Foundations of an effective RIA system
60
Future work by the Commission that could impact
on Australian RIS processes

• In 2012, the Commission will undertake a
  benchmarking study of the efficiency and quality
  of both COAG and all Australian jurisdictional
  RIS processes

61
Opportunities for Malaysia ...

• Design opportunities would extending the
  requirements for RIA to Malaysia improve the
  regulation-making system?
• Governance opportunities how to build oversight
  and quality control within the administration?
• Training opportunities providing a strong
  support function within government help desks
• Experience suggests that RIA is always a work in
  progress, governments must always return to
  strengthen and support the process