Regulatory Impact Analysis Suyash Rai, Carnegie India November 21, - - PowerPoint PPT Presentation

Regulatory Impact Analysis

Suyash Rai, Carnegie India November 21, 2019

1

Outline

• 1. Introduction and rationale
• 2. Making Regulatory Impact Analysis useful
• 3. Steps in Regulatory Impact Analysis

2

Introduction

Regulatory Impact Analysis

• Regulation can be beneficial, but it also imposes costs and distorts

investment and innovation

3

Regulatory Impact Analysis

• Regulation can be beneficial, but it also imposes costs and distorts

investment and innovation

• RIA is a tool to examine and measure the likely effects of new or

existing regulations, and assist decision makers in taking correct decisions

• Not just a product, but also a process

◮ evaluation of the problem ◮ the objective and intended effect of regulation ◮ consideration of alternative options ◮ assessment of all their impacts ◮ compliance strategies ◮ monitoring and evaluation system

3

The Rationale for RIA

• For Parliaments and Presidents: A mechanism to hold regulatory

agencies accountable

◮ Mitigate the principal-agent problems of delegation ◮ Make regulatory agencies engage with and be responsive

to their stakeholders

◮ Place on-going constraints on regulatory agencies

4

The Rationale for RIA

• For Parliaments and Presidents: A mechanism to hold regulatory

agencies accountable

◮ Mitigate the principal-agent problems of delegation ◮ Make regulatory agencies engage with and be responsive

to their stakeholders

◮ Place on-going constraints on regulatory agencies

• For regulatory agencies: A mechanism to make better regulations

◮ Induce optimisation: alternatives are compared using a

comprehensive analysis, and the best one chosen

◮ Give grounds to resist pressures from interest groups

4

The Rationale for RIA

• For Parliaments and Presidents: A mechanism to hold regulatory

agencies accountable

◮ Mitigate the principal-agent problems of delegation ◮ Make regulatory agencies engage with and be responsive

to their stakeholders

◮ Place on-going constraints on regulatory agencies

• For regulatory agencies: A mechanism to make better regulations

◮ Induce optimisation: alternatives are compared using a

comprehensive analysis, and the best one chosen

◮ Give grounds to resist pressures from interest groups

• Economic rationale: Regulatory quality has an effect on economic

performance (Jalilian et al,2007)

• Governance rationale: regulatory transparency, fairness and access

to regulation (Hahn and Tetlock, 2008)

4

The Rationale for RIA

• Such analysis does not come naturally to humans.
• We should be wary of our ability to reason, especially about public

policy

◮ Reasoning primarily serves a social function (Mercier and

Sperber, 2011)

◮ Confirmation bias: we use reasoning to find evidence to

support the position we intuitively hold (Haidt, 2012)

◮ High bar for evidence that goes against our intuitive views,

but happy to accept weak supportive evidence (Ditto, Pizarro and Tannenbaum, 2009)

• RIA can enable good reasoning and use of evidence

5

Making RIA useful

Effectiveness of RIA

• Many countries have adopted RIA, but huge differences in

implementation outcomes

6

Effectiveness of RIA

• Many countries have adopted RIA, but huge differences in

implementation outcomes

• Ways of using RIA

◮ Exploratory: choose the best regulation ◮ Confirmatory: provide evidence for a given conclusion ◮ Reputational: signal credibility by conducting RIA in

name only

• Specific use depends on the context

6

Making RIA effective

• 1. Clarity of objectives and principles of regulation
• 2. Design regulatory agencies to encourage sound use of RIA
• 3. Strengthen external accountability mechanisms

7

Clarity of objectives and principles of regulation

• RIAs force regulators to state the moral preferences implicit in

regulations

8

Clarity of objectives and principles of regulation

• RIAs force regulators to state the moral preferences implicit in

regulations

• Important to consider what objectives and principles should be

◮ given in the primary law, ◮ prescribed by the government, or ◮ decided by the regulator

• Examples:

◮ weightage of impact on low-income households ◮ inter-generational equity

8

Clarity of objectives and principles of regulation

• RIAs force regulators to state the moral preferences implicit in

regulations

• Important to consider what objectives and principles should be

◮ given in the primary law, ◮ prescribed by the government, or ◮ decided by the regulator

• Examples:

◮ weightage of impact on low-income households ◮ inter-generational equity

• Objectives, trade-offs, non-negotiables, and RIA methodological

principles: most of the existing models leave too much to the expert regulators

• Sometimes politicians have the temptation to allow regulators to

make the decision and then take the blame

8

Institutional Design of Agencies

• Governance

◮ Board structure, not all-powerful Chairperson ◮ Emphasis on diversity of boards, with a few non-experts ◮ Emphasis on consensus, with possibility of majority

decision-making - force them to reason (eg. Jury)

9

Institutional Design of Agencies

• Governance

◮ Board structure, not all-powerful Chairperson ◮ Emphasis on diversity of boards, with a few non-experts ◮ Emphasis on consensus, with possibility of majority

decision-making - force them to reason (eg. Jury)

• Regulation-making process

◮ Regulation-making process to be structured around RIA

(not just delegated to staff)

◮ Transparent and communicative consultation process

9

Institutional Design of Agencies

• Governance

◮ Board structure, not all-powerful Chairperson ◮ Emphasis on diversity of boards, with a few non-experts ◮ Emphasis on consensus, with possibility of majority

decision-making - force them to reason (eg. Jury)

• Regulation-making process

◮ Regulation-making process to be structured around RIA

(not just delegated to staff)

◮ Transparent and communicative consultation process

• Data and analysis

◮ Investment in information collection and research systems ◮ Information sharing across agencies ◮ Build analytical capabilities

9

External accountability

• The Parliament and/or the Presidency: analytical

capabilities to review RIAs coming out of regulatory agencies

• Civil society consumer groups: generally limited to the

chattering classes (eg. net neutrality in India)

• Other regulatory innovations, such as freedom of

innovation laws, are associated with strong adoption of RIA

10

Using RIA

• "RIA’s most important contribution to the quality of

decisions is not the precision of the calculations used, but the action of analyzing, questioning, understanding real-world impacts and exploring assumptions".

• RIA should not be seen as an exact science that can be

exactly replicated everywhere if the "right" methods are used

11

Steps in Regulatory Impact Analysis

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory options
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of choice

among options to the decision-maker

12

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory options
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of choice

among options to the decision-maker

13

Problem definition and the need for regulation

• Problem definition: converting a vague problem into a

precise statement of objective(s)

• The need for regulatory action:

◮ address a significant market failure: eg. market power,

information asymmetry, negative externality

◮ promote certain features of a market: eg. neutrality of

the internet; expansion into remote areas; efficient spectrum management

• The legal power or mandate to take regulatory action.

14

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory
• ptions
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of choice

among options to the decision-maker

15

Understand the baseline scenario and identify the regulatory options

• Baseline scenario: what will happen to the problem

without any additional regulation?

◮ Establish an optimised scenario ◮ Consider market forces within existing legal and

regulatory environment

◮ Assume reasonable innovation and progress

• Types of regulatory option

◮ Do nothing: baseline scenario ◮ Command and control ◮ Nudge towards self-regulation ◮ Incentive-based (eg. performance-linked penalties) ◮ Market-based (eg. competition; disclosures)

16

Baseline scenario and regulatory options

• Example: spectrum management
• Composite option

◮ Dynamic Spectrum Access and Database on Spectrum

Usage

◮ Incentive auctions ◮ Monetary penalties for license breaches

17

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory options
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of choice

among options to the decision-maker

18

Identify the Impacts of Regulatory Options

• Impact on whom:

◮ Within the territorial boundaries of India ◮ Include as many stakeholders as feasible: probability of

recommending the best option increases, but cost of RIA also increases

◮ Comprehensive identification of consequences requires an

• pen consultation process
• Consider direct and indirect impacts for all alternatives,

along with their expected timing

• List measurement indicators for each impact category
• Sources of information: Research; Consultations
• Difficulties: information availability, market response

19

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory options
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of choice

among options to the decision-maker

20

Choose a method to analyse options

• Cost-benefit analysis
• Cost-effectiveness analysis
• Multi Criteria Analysis
• Break-even analysis

21

Choose a method to analyse options

Methodology Suitable when Cost- benefit analysis Monetise costs and benefits for regulatory alternatives, and recommend the option with highest net present value When monetisation of benefits and costs is feasible, and the decision to regulate has not yet been taken, and the best regulation is to be found out 22

Choose a method to analyse options

Methodology Suitable when Cost- benefit analysis Monetise costs and benefits for regulatory alternatives, and recommend the option with highest net present value When monetisation of benefits and costs is feasible, and the decision to regulate has not yet been taken, and the best regulation is to be found out Cost- effectiveness analysis Quantify the benefits (no monetisation), and monetise the costs. Recommend the lowest cost alternative When monetisation of benefits is not feasible, and the decision to regulate has been taken 22

Choose a method to analyse options

Methodology Suitable when Cost- benefit analysis Monetise costs and benefits for regulatory alternatives, and recommend the option with highest net present value When monetisation of benefits and costs is feasible, and the decision to regulate has not yet been taken, and the best regulation is to be found out Cost- effectiveness analysis Quantify the benefits (no monetisation), and monetise the costs. Recommend the lowest cost alternative When monetisation of benefits is not feasible, and the decision to regulate has been taken Multi- criteria analysis Identify criteria for achievement

• f objective. Assign weights to

criteria based on relative importance (use expert surveys). Recommend option with the highest score When quantification of at least some of the major regulatory impacts is not feasible 22

Choose a method to analyse options

Methodology Suitable when Cost- benefit analysis Monetise costs and benefits for regulatory alternatives, and recommend the option with highest net present value When monetisation of benefits and costs is feasible, and the decision to regulate has not yet been taken, and the best regulation is to be found out Cost- effectiveness analysis Quantify the benefits (no monetisation), and monetise the costs. Recommend the lowest cost alternative When monetisation of benefits is not feasible, and the decision to regulate has been taken Multi- criteria analysis Identify criteria for achievement

• f objective. Assign weights to

criteria based on relative importance (use expert surveys). Recommend option with the highest score When quantification of at least some of the major regulatory impacts is not feasible Break- even analysis Estimate the costs of regulation, and evaluate effectiveness

• f regulation in generating

benefits to justify the costs Costs known, but effectiveness

• f regulation in generating the

benefit is debatable 22

CBA: Quantify and Monetize the Benefits and Costs

• Types of impact:

◮ Relatively easy to monetise: eg. compliance

costs/savings

◮ Can be quantified, but difficult to monetise: eg.

inconvenience/discomfort

◮ Very difficult to quantify: eg. values such as privacy

• Ideal: monetisation based on market prices
• Market prices not available/reliable

◮ ’externalities’ or unmarketed spillover effects ◮ goods affected by taxes and subsidies ◮ goods subject to import or export restrictions ◮ labour inputs in the presence of unemployment

23

CBA: Discount the future benefits and costs

• Social discount rate is to be used:

◮ Displaced investment: social cost of capital (SOC) ◮ Decreased consumption: social rate of time preference

(SRTP)

• Social discount rate: weighted average of SOC and

SRTP, based on what is being discounted

24

Case study: Ofcom CBA

Table 1: Option 2 (require VoIP providers to provide calls to fixed line and mobile for emergency calls): Costs, benefits and net benefits, 25% and 40% VoIP take-up by 2011/12

25% VoIP growth 40% VoIP growth Costs (££) 10,204,664 10,204,664 Benefits (££) 23,833,805 33,851,698 Net benefits (££) 13,634,141 23,647,034

• Partial CBA since non-ambulance emergency services not

considered; other qualitative benefits

• Conservative approach to estimating the benefits
• Benefits >Costs in this example

25

Cost-effectiveness analysis (CEA)

• Benefits may be quantified but not monetised; Costs to

be valued

• CEA involves finding out the lowest cost/ most efficient
• ption that achieves a regulatory objective
• Example: Which option will help create the most jobs for

every Rupee spent?1

• Does not help ascertain whether regulation is needed

1https://ec.europa.eu/info/sites/info/files/file_import/

better-regulation-toolbox-57_en_0.pdf

26

Break-even analysis

• Useful when effectiveness of regulations is uncertain while

the type of benefits is clear

• Break–even analysis involves:

◮ Estimating the costs of the regulation ◮ Evaluation of regulations on their effectiveness in

generating benefits that justify costs

◮ Judgments by policy-makers on the effectiveness and size

• f benefits
• Example: Helmets and lives saved

27

Multi-criteria analysis (MCA)

• Useful in cases with diverse quantified impacts and/ or

qualitative impacts2

• Steps involved:

◮ Identify criteria that indicate achievement of a policy

• bjective

◮ Rank the criteria (assign weights) in terms of their

relative importance

• Ranking is generally done using a survey of experts
• Several techniques within MCA: Outranking,

non-compensatory, Analytic Hierarchy Process, Linear Additive, etc.

2https://ec.europa.eu/info/sites/info/files/file_import/

better-regulation-toolbox-57_en_0.pdf

28

Multi-criteria analysis (MCA)

• Steps involved:

◮ Assign scores to each regulatory option for each criterion ◮ Calculate weighted scores to compare options

• Ensures that qualitative analysis is weighed appropriately
• Not to be used as an alternative, rather as a supplement

to more rigorous quantitative analysis where feasible

29

Example: MCA

Weightage Fluoride regulation Advertising campaign Free dentist visits Effectiveness in improving dental health 4 5 (20) 3 (12) 3 (12) Ability to address existing dental problems 2 0 (0) 1 (2) 5 (10) Ability to improve dental health of the poorest groups 2 4 (8) 2 (4) 5 (10) Ability to improve health in all regions 1 5 (5) 5 (5) 3 (3) Cost (higher score for lower cost) 4 5 (20) 4 (16) 2 (8) Score 53 39 43 30

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory options
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of choice

among options to the decision-maker

31

Sensitivity analysis

• Uncertainty in estimates needs to be addressed
• Investigate the influence of key variables
• Possible approaches:

◮ Worst/ best case scenario analysis — using most

conservative and least conservative values for variables

◮ Partial sensitivity analysis for key risk factors ◮ Monte Carlo sensitivity analysis — complex; uses

probability distributions

32

Case study: Ofcom CBA

• Ofcom did a sensitivity analysis of reported costs
• Considered maximum possible costs that could be

incurred by providers under Option 2

• Also did a sensitivity test to consider how industry

compliance costs would vary with new entry.

33

Case study: Ofcom CBA

Table 2: Net Present Value (NPV) of grossed-up industry costs (2007/08 – 2011/12) (2007/08 prices) for Option 2

2007/08 2008/09 2009/10 2010/11 2011/12 Total (NPV) Set-up costs (£) 7,054,516 7,054,516 Ongoing costs (£) 674,105 651,310 629,285 608,005 587,444 3,150,148 Total indus- try costs (£) 7,728,622 651,310 629,285 608,005 587,444 10,204,664

34

Case study: Ofcom CBA

Table 3: NPV of grossed up industry costs (2007/08–2011/12) (2007/08 prices) (sensitivity analysis) for Option 2

2007/08 2008/09 2009/10 2010/11 2011/12 Total (NPV) Set-up costs (£) 17,703,722 17,703,722 Ongoing costs (£) 1,078,568 1,042,095 1,006,855 972,807 939,910 5,040,234 Total indus- try costs (£) 18,782,290 1,042,095 1,006,855 972,807 939,910 22,743,956

35

Distributional and cumulative impacts

Consider impacts on:

• Different states/regions
• Small and medium enterprises
• Disadvantaged sections of society (Lower income groups,

PWDs)

• Gender impact
• Future generations

36

Steps in Regulatory Impact Analysis

• 1. Problem definition and the need for regulatory action
• 2. Understand the baseline scenario and design the regulatory options
• 3. Identify the impacts of regulatory options
• 4. Choose a method and analyse the options
• 5. Special tests: sensitivity analysis; distributional impact
• 6. Present the recommendations along with the criteria of

choice among options to the decision-maker

37

Presenting recommendations

• Results of RIA need to be communicated in a clear and

understandable form

• Use visual aids to present the analysis:

◮ Problem trees: root causes -> core problem -> effects ◮ Objective trees: regulation -> layers of objectives ◮ Intervention logic diagrams: regulation -> results ->

regulation’s purpose -> overall objective of regulator

• Publishing results for comments — helps improve the analysis and

ensures transparency

• Develop a standard template that outlines each step taken in the

RIA process

38