Economics Workshop: Day 2 Better Regulation Executive

Nick CraftsUniversity of Warwick

June 2006

Objectives for the day

To • understand the basics of regulatory impact

assessment• understand how good regulatory design can cope

with uncertainty, informational imperfections, and minimise distortion of incentives

• rationale for and implementation of RPI-X regulation

• the link between regulation and productivity growth

Regulatory Impact Assessment

COST-BENEFIT ANALYSIS

Analysis that quantifies costs and benefits, including items that the market does not value properly

Used for capital projects and procurement decisions policy proposals, including regulatory proposals

environmental standards, health & safety, business regulation

[Cost-effectiveness analysis when benefits are hard to quantify, or externally specified]

THE PROCESS

1. Justify action and set objectives

Identify the market failure or the socially-undesirable outcome that calls for regulatory intervention. Examples

• discharge of pollutants in atmosphere (an externality) reduces air quality: the objective is to reduce pollutant levels by amount x

• congestion externality causes traffic jams in Central London: the objective is to reduce peak-time traffic by 20%

THE PROCESS

2. Identify all options

• prescriptive regulation (quotas, speed limits)• provide incentives to change behaviour (taxes and

subsidies)• create arrangements or institutions to change

outcome (tradable permits)• provide information/educate to alter behaviour

(public campaign on dangers of excessive salt, smoking)

• no intervention

THE PROCESS

3. Evaluate costs and benefits of each option

• Evaluate direct policy costs & administrative costs • Identify benefits and evaluate them as far as possible.

Include externalities (esp. environmental ones), consumers’ surplus, etc.

• Some benefits (e.g. prevented fatality) are hard to evaluate. Can infer prices from revealed preferences. If not, can use stated preference through contingent valuation: Willingness to Pay (WTP) or Willingness to Accept (WTA)

• Identify unintended consequences and cost them too • Where relevant, identify distributional implications

What if benefits and costs are uncertain?

Risk assessment and management is important. • Identify sources of uncertainty. Not enough to

look at most likely scenario: evaluate costs and benefits for range of scenarios. Recognise ‘optimism bias’.

• Use pilot programmes to learn more about the true costs and benefits

• If possible at reasonable cost, transfer risk to party best placed to control it: outsourcing of technology-related risk to private sector

THE PROCESS

4. Develop and implement solutions

Good regulatory design must also consider • regulator’s ability to monitor and verify choices• ease of ensuring enforcement• design of robust targets (avoid targets whose

achievement might run counter to objectives)• proportionality, accountability, consistency

THE PROCESS

5. Evaluation

• Review costs and benefits of previous regulation periodically to assess its usefulness

• If necessary, use sunset clauses in regulation to force evaluation at later date, in light of new information

• Likewise, reassess any ‘no intervention’ decision in the light of new information and developments

Travel Costs on Motorways, 1993 (£m)(Plowden & Hillman, 1996)

Travel Time

Fuel Non-fuel Operating

Accident Road User Costs

Speed Limit

70 mph 5225 1326 7404 659 14614

60 mph 5696 1252 7326 487 14761

50 mph 6640 1161 7240 393 15434

40 mph 8250 1097 7201 353 16901

Group Work: Impact Assessment

For each category of regulation below, identify the social costs and benefits, including any likely ‘unintended consequences’.

1. Compulsory identity cards2. Legislation to keep pubs smoke free3. Green Belt as policy for land-use planning4. Regulating the introduction of new drugs

Information and Incentives

An overview

Regulation amounts to state-imposed limitation on individual discretion, usually supported by the threat of sanctions (stick) or by provision of appropriate incentives (carrot)

Information

Individual choices depend on information too. Two relevant aspects. Information tends to be• imperfect (we do not know everything)• de-centralised (we differ & know more about

ourselves)

The questions • how does information affect case for regulation?• how does information affect scope of regulation?• how does regulation distort information and

incentives?

Imperfect Information

For the class of decisions where• individuals’ information is imperfect AND • the state could better informed, state regulation can correct for individuals’ ignorance and

protect their interests

Examples • product safety regulation• health and safety regulation

Why might the state be better informed?

• Individuals cannot easily assess safety aspects of poor product design

• Employees cannot always assess riskiness of work environment, especially if damage comes with a lag (asbestos exposure, coal dust)

Here the state can be better informed (commission scientific studies) and regulate if necessary

Is statutory regulation necessary?

Regulation not necessary if markets create incentives for firms to protect consumer / employee interests. For instance,

• Reputational concerns may persuade firms to maintain product quality / work-place safety

• Risk of legal actions helps too However, these mechanism are less effective when firms

are small or new

Markets can be informationally efficient..

• There are many contexts in which individuals or firms know more about themselves than others: information is de-centralised

• Markets can work with de-centralised information: individuals choices are based on private information but prices convey the essential bits of information to everyone.

• Regulatory control, as in a command economy, requires centralisation of information. This informational constraint makes it harder to regulate.

... while efficient regulation is hard to achieve

• When firms / individuals have unequal compliance costs, it is economically efficient to impose unequal standards (e.g. ask ‘dirty’ firms to do more abatement)

• but lack of information about compliance costs make it harder to tailor-make regulation

• so that the same regulation may pose too much burden on some and not enough on others (identity cards, for instance)

and getting information is tricky

• In principle, the government could try and gather more information

• but regulation distorts incentives for providing information

• for example, all regulated firms would like to argue that their costs are high

• Of course, some regulatory instruments are better able to cope with informational constraints (carbon trading arrangements, for instance).

Regulation and Incentives

• Individuals choice also depend on incentives• Markets provide sharp (‘high-powered’)

incentives, both carrot and stick. (e.g., profits, risk of bankruptcy, promotion vs. sacked)

• It is not easy to fine-tune the regulatory stick: road safety is only crudely regulated through speed.

• People invest a lot in avoiding detection: better monitoring technology helps but cannot always solve the problem

• If penalties are not proportional to violation, it may create perverse incentives

Regulation changes behaviour

Regulation can have perverse effects

Examples• Safety devices like seatbelts and airbags may have

a ‘lulling effect’, lower effort in safety and even increase risk levels

• Employment regulations that protect workers from being fired reduce incentives to hire them

• Complaints procedures that reduce work effort• Penalties for lateness that slow down transport

Future of Regulation

There is good regulation..

Some regulation provides the framework of civil society• regulations to protect private property: essential spur

to investment• regulation to protect Intellectual Property Rights:

provide incentives for R&D and innovation• regulations against insider-dealing: allow capital

markets to exist

..and bad regulation

Other regulation stifle growth• price regulation inhibits investment• labour-market regulations create inflexibilities: Euro-

sclerosis• regulations that make it hard to set up / wind up

business make the economy less responsive to social change

A broad correlation...

• High regulation, especially in developing countries, seems to stifle growth

• But a cautionary note: growth is not an end in itself• If the aim is greater welfare, some forms of regulation

increase welfare directly, even if they lower growth rates on the margin

The broad trend

The last two decades have seen a trend towards regulatory reform

• Economic regulation is down: state monopolies have been replaced by privatised firms, with lighter regulation overall; firms’ entry and exit has become easier

• Social regulation is up: not surprising as richer societies invest more in health and safety, environmental regulation

The Context

• Globalisation (mobile capital)

• Technological change and new information makes it important to remove impediments to innovation

• Economic theory (asymmetric information)

Why does regulation persist?

• Some regulation corrects persistent market / information failures, so needs to persist

• Political economy: those who would lose from deregulation can lobby more effectively than those who gain from deregulation (similarity with import restrictions)

• In many sectors deregulation has led to reduction in prices and profits (airlines, utilities, telecom): we should hardly expect business to support such deregulation.

Policy Conclusions

• Free markets are usually efficient: the invisible hand works

• However, markets are not always efficient. Market failures make a potential case for government intervention to improve efficiency: when the invisible hand does not work, the government can lend a helping hand

• Beware the risk of government failure: the helping hand may hurt rather than help (heavy-handed intervention)

• Informational problems affect both private decision-making and public interventions: regulation may have perverse effects (fumbling hand)

• Further, the helping hand may become self-serving (the grabbing hand of a predatory state)

Good regulation combines economic theory with practical understanding

RPI – X Regulation

Regulation of Privatized Industries

• Regulate to prevent abuse of market power in cases of natural monopoly or high entry barriers

• British approach based on price capping whereas traditional American version used rate of return

• RPI – X aims to stimulate cost reduction as well as preventing high price/cost margins

Productivity and Prices

• In a competitive market p=mc=LAC (including a normal rate of return on capital)

• In the absence of productivity growth, over time output prices would rise at the same rate as (the weighted average) of input prices

• More generally, output prices in a competitive market rise at the rate of growth of input prices minus productivity (TFP) growth

Price Capping

• If prices are allowed to rise at RPI-X, the industry will be able to maintain normal profits providing it achieves TFP growth equal to the national average, TFPUK, + X

• If TFP growth is greater than (less than) TFPUK + X, the formula implies supernormal (subnormal) profits

RPI-X vs Rate of Return: Key Arguments

RPI-X

Strong incentives for cost reduction and innovation

BUT

Quality may suffer

Prices exceed costs on average

RATE OF RETURN

Prices stay in line with costs

Quality ‘assured’

BUT

Likelihood of excessive investment

Weak incentives for productivity improvement

The Averch-Johnson Problem• Rate of return regulation which allows the firm

to earn a rate of return above the cost of capital encourages the firm to accumulate an excessive capital stock

• The equity value of the firm will be proportional to the capital stock and, provided the allowed rate of return is above the cost of capital, the share price will be positively related to the capital stock

• Have to rely on the regulator to identify unnecessary projects and disallow them from the cost base

The Water Industry• Regulation is a hybrid

• It’s the Averch-Johnson problem that we should fear in the long-run ….. especially given populist pressures

• Excessive capital stock encouraged by

– de-luxe quality directives

– ensuring no supply interruptions

– no peak load pricing

Key question: is marginal benefit (willingness to pay) less than long-run marginal cost?

RPI-X Regulation: Further Points• Optimal length of price reviews trades off gains from

cost reduction against losses from excessive prices …. is shorter the lower is the sensitivity of costs to cost-reducing effort and the higher is price elasticity of demand

• Mitigating ‘regulatory risk’ with sunk costs through credible commitment by regulator is desirable

• Uncertainty lowers advantages of price caps if need to ensure non-negative profits

• Setting X well requires good way of estimating potential for productivity improvement

Managerial Effort and Productivity Growth

• Implementing productivity improvements/cost reductions requires managerial effort, i.e. has disutility for managers

• Monitoring managers in context of asymmetric information encounters free rider problems in private sector and may offer no reward in public sector

• Competition is antidote to agency problem

Privatization and Managerial Effort

• Asymmetric information does not go away

• Private shareholders may improve monitoring/incentivizing of managers

• Competition may increase

• Regulator may have to try and compensate for weaknesses of shareholders and/or competition

Privatization and Productivity Performance (Green and Haskel, 2004)

• TFP growth raised by the privatization process not by private ownership per se

• Productivity growth increased in some cases as X factor made more demanding (e.g. water)

• Regulation central to quality implications of RPI-X incentive structure

• Overall picture is dominated by levels effect of eliminating inefficiency

Total factor productivity in the UK public sector (annual rate of increase, %)

72/3-78/9 78/9-86/7 86/7-99/00

British Airways +3.0 +3.3 +4.2 Privatised 1987

72/3-78/9 78/9-86/7 86/7-93/4

British Coal -2.8 +0.1 +9.0 Privatised 1994

72/3-78/9 78/9-86/7 86/7-94/5

British Gas +8.2 +2.0 +1.5 Privatised 1986

72/3-78/9 78/9-88/9 88/9-97/8

British Steel -5.0 +3.8 +1.8 Privatised 1988

72/3-78/9 78/9-84/5 84/5-94/5

British Telecom +0.6 +3.2 +3.0 Privatised 1984

Source: O'Mahony (1998)

Comparative Productivity: electricity, gas and water sectors, 1979-95 (UK = 100)

0

100

200

300

400

500

1979 1989 1995

USA France Germany Japan

0

50

100

150

200

250

1979 1989 1995

USA France Germany Japan

Total Factor Productivity Labour Factor Productivity

Regulated Prices in the UK

020406080

100120140160180

BT Gas Transportation Water

REC Distribution NGC Transmission

Conclusions• Productivity performance in privatized

utilities may be affected by the incentive structures of the regulatory framework

• In practice, not clear that RPI-X has generally been a strong driver of TFP growth

• Introducing competition where possible delivers stronger incentives to improve productivity

GROUP WORK

1. When would you expect regulation to have a powerful impact on the productivity performance of a privatized business?

2. How should a regulator decide the precise value of ‘X’ at a price review?

Regulation and UK Productivity Performance

Costs and Benefits of Regulation

• Regulation that corrects market failures provides gains from a more efficient allocation of resources

• Regulation also incurs costs so it is relevant to ask how benefits compare with costs

• The costs of regulation may be felt in terms of lower GDP per person

Questions

• In what ways can regulation affect productivity outcomes?

• How good are measures of regulation?

• Is the UK lightly regulated?

• Does regulation actually have a big impact on labour productivity growth?

Regulation and Productivity

• Compliance costs have direct productivity implication

• Additional adverse impacts if disincentives to investment and to innovation

• May create barriers to entry that reduce competition

• Impact has not been well quantified

Compliance Costs

‘Administrative Costs’ = 3 to 4% GDP and ‘Policy Costs’ = 7 to 8% GDP (BRTF, 2005)

• Direct measurement effect will be to reduce measured TFP by an equivalent amount of productive resources diverted away from producing output

• No time series evidence on compliance costs but difficult to believe these direct effects have reduced annual TFP growth much in recent past

Regulation as a ‘Tax’

• Investment and innovation are key determinants of labour productivity growth

• Appropriable returns underpin incentives to investment and to innovate

• Regulation may reduce net present value of projects

• For example, employment protection and ICT expenditures (Gust and Marquez, 2004)

Regulation as Barrier to Entry

• For example, costs of setting up new business, licensing rules, planning restrictions

• Empirical evidence of cross-country comparisons shows tighter regulation reduces entry and raises price-cost mark-ups (Cincera and Galgan, 2005; Griffith et al., 2006)

• Retailing productivity growth example of regulatory barriers having seriously adverse impact in Europe compared with US (McGuckin et al., 2005) in ICT era

Competition and Productivity Growth

• Absence of competition allows managers to be sleepy if ineffective control/monitoring by shareholders

• Competition is strongly positive for productivity outcomes in UK firms without dominant shareholder (Nickell et al., 1997)

• Competition promotes better management practices (Bloom and van Reenen, 2006)

• Patenting performance of UK firms suggests inverted U-shaped relationship with price-cost margin which peaks at about 20% (Aghion et al., 2005)

Policy Impact on Rate of Technology Adoption

Competition Policy PositiveNegative

Industrial Policy Positive Negative

Maximizing Agency Problems

Firm Type

Maximizing FirmsCompetition Policy lowers expected profit from innovation Industrial Policy raises expected profit from innovation

Agency Problem FirmsCompetition Policy cuts rents and raises cost-reducing effort Industrial Policy pays subsidies and lowers cost-reducing effort

Regulation and the Growth Rate

• If regulation is a disincentive to investment and innovation, they will be lower as a result

• Endogenous growth models predict that the rate of growth will be adversely affected

• This would be the most serious consequence of excessive regulation rather than the diversion of resources through conventional compliance costs

Measuring Regulation

• Evidence on compliance costs quite limited

• Investigators looking at relationship between regulation and productivity performance have used indices constructed by OECD, World Bank and surveys of expert opinion conducted by IMD, World Economic Forum etc.

• Both product market and labour market indices available

BUT how good are they?

Measures of Regulation

• ‘Subjective’ vs ‘Objective’

• Comprehensive?

• Take account of enforcement and litigation?

• Include ‘extraneous’ aspects?

IMD Survey Questions

• Business Regulations:

“Regulation intensity does not restrain the ability of companies to compete”

• Labour Regulations:

“Labour regulations do not generally hinder business activities”

OECD Regulation Indices

• Product Market Regulation (Conway et al., 2005): index designed to reflect the extent to which the regulatory environment is conducive to competition including indicators of state control, barriers to entrepreneurship

• Employment Protection (OECD, 2004): index designed to reflect legislation as employer-borne tax on employment adjustment including difficulty of dismissal and extent of severance pay

Is the UK Lightly Regulated?

• OECD measures say yes

• Subjective indicators more equivocal, cf. IMD scores where UK has been slipping down the league

• Overall, within OECD UK closer to ‘relatively liberal’ group including Australia, Canada, Denmark, Ireland, USA than the ‘relatively strict’ group including France, Germany, Greece, Italy, Portugal and Spain

Business Regulations, 2005 (0-10) (Source: IMD, 2005)

FinlandDenmarkNorwayAustriaSwedenAustraliaSwitzerlandCanadaPortugalJapanUSANew ZealandIrelandNetherlandsSpainGreeceUKGermanyFranceItalyBelgium

0 1 2 3 4 5 6 7

Source: IMD, 2005

Labour Regulations, 2005 (0-10)France

Germany Belgium

PortugalSpain

GreeceNetherlands

Italy

Denmark Switzerland

USACanada

JapanAustria

UKIrelandFinland

NorwayNew Zealand

Australia

Sweden

0 2 4 6 8 10

Source: Conway et al., (2005)

Product Market Regulation (0-10)

1998 2003

France 4.17 2.83

Germany 3.17 2.33

Italy 4.67 3.17

Spain 3.83 2.67

UK 1.83 1.50

USA 2.17 1.67

Product Market Regulation and Productivity Growth

• Regulation that creates barriers to entry raises mark-ups and reduces innovation, investment and productivity growth (Griffith and Harrison, 2004; Griffith et al., 2006)

• At the macro level de-regulation has been associated with better TFP growth (Nicoletti and Scarpetta, 2003)

• Product market regulation is negatively correlated with the contribution of ICT-using services to aggregate productivity growth (Nicoletti & Scarpetta, 2005)

• UK shows up well on OECD measures compared with other European countries

Multifactor productivity acceleration and product market regulation

UKUS

Ireland

Australia

SwedenDenmark

GermanyFinland

Greece

Netherlands

Japan

Portugal

Spain

BelgiumFrance

Italy

Canada

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

0.3 0.8 1.3 1.8 2.3 2.8 3.3 3.8

Difference in average MFP growth rate between 1990-2000 and 1980-1990Adjusted for hours worked

Product market regulation, inward oriented, 1998

Correlation coefficient -0.51t-statistic -2.29

Source: Nicoletti & Scarpetta (2005)

Regulation and the contribution of ICT-using services to aggregate productivity growth

-0.2

0.3

0.8

1.3

0 0.5 1 1.5 2 2.5 3 3.5

GBR

USA

MEX

AUS

IRL

SWE

CAN

AUTKORNOR

DNKNLD

JPNDEU

ESPFRA

ITABEL

FINCHE

ICT using services, 1996-2001

Product market regulation (inward-oriented), 1998

Correlation coefficient: -0.62t-statistic: -3.35

Source: Nicoletti & Scarpetta (2005)

Source: GGDC, 2004

Retail Trade: Labour Productivity Growth (% per year)

1990-5 1995-2001

US 2.0 6.5

EU 1.7 1.3

Germany 2.8 0.7

UK 1.2 3.7

France 2.1 1.9

Italy 1.3 1.1

ICT Expenditure and Employment Protection Legislation

• Are inversely correlated

• Firing costs delay adoption of ICT … but do not generally deter investment

• Effective use of ICT often involves upgrading labour force skills and re-organization, i.e. labour turnover

Nickell (2005)

Employment Protection Index (0-10)

1980 1990 1998 2003

France 6.50 7.05 7.00 7.00

Germany 8.25 7.60 6.50 5.60

Italy 10.00 9.45 7.50 4.85

Spain 9.55 8.70 7.00 7.50

UK 1.75 1.75 1.75 1.75

USA 0.30 0.50 0.50 0.50

IT Expenditures and Employment Protection Legislation

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

2.8

0.5 1 1.5 2 2.5 3 3.5

Employment Protection Legislation, Index, 1998

•AU

•US

•UK

•CA

•FI

•SE

•NE •NO

•IT•SP

•JA

•GE

•FR

Correlation = -0.72

IT E

xpen

dit

ure

s (%

of

GD

P),

199

9

Reduction in PMR and UK Productivity Performance

• Nicoletti and Scarpetta (2003) results imply UK has had modest TFP growth advantage over France and Germany in the past 20 years

• This is reflected in decline in TFP (but not other) component of labour productivity gap.

Sources: Broadberry & O'Mahony (2006); Crafts & O'Mahony (2001)

A Decomposition of UK Labour Productivity Gap (percentage points)

France/UK Germany/UK

1979

Labour Productivity Gap 31 30

Labour Quality 6 5

Physical Capital 17 9

TFP 8 16

2000

Labour Productivity Gap 21 17

Labour Quality 4 4

Physical Capital 17 12

TFP 0 1

Note: In 1979 Germany is West Germany only.

Source: Nicoletti and Scarpetta (2005)

Change in TFP Growth over 10 years from Adopting Best Regulatory Practice (% points)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

France Germany Italy Spain UK

Conclusions

• Regulation does have implications for TFP

• In particular, this is true of regulation that inhibits competition

• Administrative costs of compliance are not the key issue