MiCRA CALCULATING ANTITRUST FINES & DAMAGES

THEORETICAL UNDERPINNINGS & PRACTICAL APPROACHES

Presentation to Commissioners Malaysia Competition Commission

byRughvir (Shyam) Khemani, PhD (LSE)

Microeconomic Consulting and Research Associates

(www.micradc.com) andFormer Advisor, Competition Policy

The World Bank Group, Washington D.C., USA

Kuala Lampur, Malaysia, 8-9 June 2013

1

MiCRA

Effective Administration of Competition Law and Policy

The effective administration of competition law and policy requires balancing of

EnforcementFostering ComplianceEnforcement-> Investigation, Prosecution,

Adjudication, Judgment, Imposition of Fines & PenaltiesCompliance-> Research & Policy Analysis, Market

Studies, Publications & Speeches, Meetings with Stakeholders, Advocacy

Effective enforcement (including appropriate levels of fines and sanctions) ->Compliance

Compliance->Lowers administrative & enforcement costs, legal and economic uncertainty….

2

MiCRA

Definitions

Fine—“a sum imposed as punishment for an offense”Penalty—“the suffering in person, rights, or property that is annexed by law or judicial decision to the commission of a crime or public offense”Sanctions—“the detriment, loss of reward, or coercive intervention annexed to a violation of a law as a means of enforcing the law”Damages—“compensation in money imposed by law for loss or injury”Remedies—“the legal means to recover a right or to prevent or obtain redress for a wrong/something that corrects or counteracts”

 3

MiCRA Powers to Impose Fines and Other

Remedial MeasuresSection 35 MyCC can apply Interim MeasuresSection 61 MyCC can impose penalties:

Corporate body fine of >5 million ringgit for initial offense; >10 million ringgit for 2nd & subsequent offense(s)

Non-corporate person(s) >1 million ringgit and/or imprisonment of 5 years; 2nd and subsequent offense(s) > 2 million ringgit, and/or 5 years imprisonmentSection 62:Compounding of offencesMyCC can impose fines on enterprises up to 10% of world-wide turnoverSection 64: Rights of private action (by persons directly/indirectly affected)

4

MiCRA

Characteristics of Illegal Price-fixing

5

1. Higher prices

2. Lower variance

3. Price increases gradually to prevent detection

4. Price falls after detection, with lag to reduce estimate of damages

MiCRA

Example of Price-fixing: Level and Variance

6Source: Abrantes-Metz, Froeb, Geweke and Taylor

MiCRA

Empirical Estimates of Cartel Pricing

Regression meta-analysis concludes that increase in price due to cartel is between 20% and 30%

7Source: Connor 2005

MiCRA

Case Study: European Cement

BACKGROUND: MiCRA retained by participant in collapsed cartel to appeal penalty based on agency’s estimate of price effect of cartel

ISSUE: Can fall in price after cartel collapse serve as estimate of effect of cartel on prices?

8

MiCRA

European Cement

•Price falls by almost 50% with collapse of cartel•Implies cartel raised prices by € 30/ton

9

Laspeyres Price Indices for Selected Cements, 2000-2003

0.000

0.200

0.400

0.600

0.800

1.000

1.200

2000 2001 2002 2003Year

Pri

ce I

nd

ex

Agreement Area

MiCRA

European Cement

Cartel agreement was to maintain assigned market shares.

Assigned shares based on capacity Induces massive excess capacityPrices collapse with collapse of conspiracy

to unsustainably low levelsFall in price with collapse of conspiracy

overstates price effect of conspiracy

10

MiCRA

Compare to Margins in Other Countries

11

Cement Industry Variable Margins 1992 – 2001

YearEuropean

Variable Margin

United States Variable Margin

1992 .584 .5391993 .588 .5351994 .598 .5821995 .585 .5821996 .596 .6081997 .593 .6101998 .629 .6251999 .649 .6262000 .666 .6222001 .672 .622

Weighted Average .615 .601

• Implies prices 3.5% higher due to cartel• Approximately € 2.11/ton

MiCRA

Assessing DamagesAssessing size of ‘consumer harm’ arising from

anticompetitive conduct influences size of fine to be imposed

Key issue in commercial disputes estimate compensation

General approach: compare actual outcomes(prices) with what would have been absent anticompetitive conduct (‘but for’ analysis)

Econometric analysis controlling for main factors affecting prices viz., changes in costs, demand and customer mix….’before and after’ study of market and firm pricing-output behavior….

12

MiCRA

Vitamin Case

Vitamins 1999: Hoffmann-La Roche and other firms pled guilty to operating a world-wide cartel over previous decade for main vitamins (especially vitamins A & E)Investigations & prosecutions in EU and other jurisdictionsDocuments indicate price levels and changes pre-post cartel period, gradual, systematic price increases and rapid steep decreases post US investigation..Differentials facilitated estimates of ‘over-charge’ US fine $500 million; EC fine euros 462 million

13

MiCRA GOOGLE On-line Mapping

2012 France’s Commercial Tribunal of Paris (CTP) found Google abused its dominant position fined euros 500,000.

Relevant market: “online mapping allowing for the geolocalisation of sales points on company websites”

CTP held Google dominant (de facto monopoly) in France in search engine market.

Allowed for dominance in connected online mapping market free

Disadvantaged Bottin, a French company offering online mapping for annual subscription fee

14

MiCRA GOOGLE online mapping-continued

Free’ service did not allow for covering of costs e.g. acquiring geographic/aerial rights necessary for mapping charged by 3rd parties

Google pricing strategy exclusionary, drove all competitors (e.g. Maporama) out of the market

Google strategy maximized its advertising revenue to detriment of competitors that needed to charge fees for online mapping service

CTP rejected Google defense:Predatory pricing conditions as per EC guidelines

not provenGoogle was sacrificing short-term profits….and

other arguments…..15

MiCRA

Polypropylene CarpetPolypropylene carpet low grade floor

covering used in low-income houses/officesMid-1990s US DoJ investigated alleged price

collusion: firms’ prices which had previously increased declined rapidly

Econometric analysis and actual prices pre-investigation > post-investigation prices

During alleged cartel period, cost declines of principal inputs not passed thru’ rapidly suggesting collusion; but cost increases post investigation also not passed thru rapidly

Parties argued this as evidence of competitive pressures-fear of losing market share…. 16

MiCRA

Polypropylene Carpet--continuedIn ‘oligopolistic’ markets, prices tend to increase

when costs increase but tend to be ‘sticky’ downward when costs decline

Case settled before adjudication, empirical analysis, evidentiary issues unresolved

Some Complexities:Econometric analysis data intensive, require well-

specified models, inclusion of all relevant variables‘But for’ analysis needs to consider ‘structural

changes’ between pre-post anti-competitive periods

‘Pass thru’ calculations: final vs. intermediate purchasers, inter-connected markets….

17

MiCRA

Pass-Through Rate

Direct and indirect purchasersIllinois Brick in US vs. EU approach

Theory: Pass-through rate depends onCompetition – oligopoly – monopolyFirm-specific vs. industry-wide cost increase

Shape of demand curveSlope of marginal cost curveConduct parameter

18

MiCRA

Pass-Through Rate

Under competition, pass-through rate depends on the supply elasticity and demand elasticity

19

Price

PT = 1

Quantity

Infinitely elastic supply curve

∆P = ∆C

Price

∆P < ∆C

PT < 1

Quantity

Upward-sloping supply curve

MiCRA

∆P = 1/2∆C∆P

Pass-Through Rate

Under monopoly, pass-through rate depends on the convexity of the demand curve

20

Price

∆C

Linear demand curve

PT = 1/2

QuantityMR

Constant elasticity demand curve

PT = 2

Quantity

Price

∆P = 2∆C

∆C

MiCRA

Implementation IssuesOptimal penalties

Theory: harm divided by probability of detection Financial penalties vs. jail time

Level of fines: cost recovery + punitive + deterrent Fines: Too low license fee to commit infractions?

Proportionate, deterrentFactors to consider: Magnitude of price

differentials, profits earned, time duration of anticompetitive conduct, size of market affected, nature and type of customers (individual consumers vs. business firms, income group, etc.), nature of product (staple vs. other), importance in budget/cost…..

Amnesty-Leniency

21