L7 - Collusion Case Studies

Question 1

What are the 4 key features of Anti-Cartel Enforcement Policies?

Answer 1

1. Deter cartels forming in the firm place
2. Detect Cartels that haven’t been deterred and have formed
3. Punish the detected Cartels
4. Compensate the victims of the Cartels –> the firms arent used to compensate the victims though (too costly for the Competition Agency to administer –> so usually goes into the public purse)
   
   1. Victims have to sue for damages to get their compensation

Large fines with a leniency program can help with the first 3 points

Question 2

What are the two cases that we will be looking at?

Answer 2

• 1. Graphite electrodes cartel: deterring explicit collusion
     * How fines are calculated, asked if the commission policy for issuing fines is in line with a policy promoting deterrence
• 2. District heating cartel: suing for damages
     * How damages are calculated (economics of this calculations)

Question 3

GEC: What is the market?

Answer 3

• Product market: Graphite electrode
  
  • ceramic-moulded column of graphite used in the production of steel –> used to transform old steel into new steel
    
    • The process of using a graphite electrode on top of a furnace to recycle steel accounts for 1/3 of steel production in Europe in the 1990s
• Geographic market: Global (US, Europe, Asia)
  
  • Firms: 9 firms operating in Europe in the 1990s
  • Two global leaders controlled most of the market: SGL and UCAR
  • 4 big players in the US (UCAR, SGL, C/G, SDK) CR4=95%
    
    • A strong fall in demand(slump in steel production) in 1982 reduced the number from 18 overtime –> common theme among cartels is a fall in the demand right before they are set up
• Entry barriers: High
  
  • “complicated and expensive processes necessary for production” No significant entry since 1950

Question 4

GEC: Overview of the case?

Answer 4

• June 1997:
  
  • the investigation started by European Commission (in parallel with others) –> started with Dawn raids of their factories
  • Infringement of Article 81 (now 101) and conduct continued beyond June 1997
  • Firms attempted to obstruct the investigations and gathering of evidence
• July 2001: decision of EC delivered
  
  • 8 of 9 firms fixed prices during 1992-98 in ‘top guy’ meetings in Switzerland
  • Side payments between firms prevented gains from cheating
    
    • E.g. if one cartel made lots of sales it would buy products from the other firms to smooth the profits across all
  • C/G did not attend meetings but received information from cartel
    
    • Used this info to undercut collusive price more than doubling its sales ==> they were deviating from the cartel
• April 2004:Court of First instance decision relating to appeal by 7 firms
• June 2006:
  
  • European Court of Justice (highest court) decision relating to appeal by EC and 2 firms
  • Firms fined €218.8m in total following appeal (9.5% of EEA turnover)

Question 5

GEC: What was the method of calculating the fines?

Answer 5

• Gravity –> how serious
  
  • Cartel’s are considered vary serious ($20 million +)
• Used to increase the fine by 5% per full 6 months the cartel was operating
  
  • Now is 100% per year
• Can add or take from the fine
  
  • • (aggregating circumstances) –> repeating infringements, refusal to cooperate, if the firm is the ringleader of the cartel
  • • (attenuating circumstances) –> passive, just following the cartel or if the infringement has been committed unintentionally or as a result of negligence
• Then it considers the level of the fine –> that cannot exceed 10% of its global turnover –> commission doesn’t want to be in the market to bankrupting firms
• leniency is then applied:
  
  • Did they come forwards to help
  • Ability to pay and Other factors

Question 6

GEC: How were the fines distributed given what the theory outlines?

Answer 6

Table one: result of the Two global players and the outcome of the fines (including appeals)

• DGComp –> Competition wing of the European commission
  
  • CFI –> Court of the first instance
  • ECJ –> European Court of Justice
• They were category 1 as they were considered the worst firms
• Gravity of the situation was high
• 55% signifies the cartel lastest for 5 and 1/2 years
• Aggravating circumstance:
  
  • SGL ONLY –> 25% for obstructing proceedings
  • 50% for being the ringleader
  • 10% for continuing the infringement after the investigation started
• Binding –> capped at 10%
• Leniency
  
  • Got 30% removed as they provide information beyond what the commission asked for (meetings they were not awarded of)
  • Their appeal led to an increase in this appeal

Category 2 firms:

• SDKK has a deterrence multiplier as it was the biggest player out of the Category 2 lot by a lot
  
  • Also had the largest leniency discount as they were the first out of all of them to reveal the information about the cartel
• CGG
  
  • Get attenuating circumstances
    
    • As it didn’t attend the meetings and just followed the cartel

Question 7

GEC: Theory of (Optimal Deterrence)?

Answer 7

• Benefits = (profit of cartel - profits of competition) * duraton

Question 8

GEC: Was the EC approach to leniency in line with deterrence theory?

Answer 8

• Technically one should offer anyone that offers information on the cartel 100% leniency –> as this is a great incentive for them to come forward
  
  • Yet what we saw was that everyone was offered leniency and not up to a 100% which is not inline with the theory

Question 9

GEC: Was the EC approach to C/G’s fine in line with the theory?

Answer 9

Cartel may actually allows someone to undercut them if they are smaller and less symmetrical to the other firms –> allows for a more stable cartel system

Question 10

DHC: How are damages usually calculated?

Answer 10

• ‘but for’ price –> that price that would have occurred if the cartel was not in operation
• damages are the red square
• What is the problem with this calculation?
  
  • Issue is we are only calculating some of the damages
    
    • considering the whole Consumer Surplus we are the whole triangle to the right of the square
      
      • Damage to the consumers that were priced out of the market due to the cartel’s price
      • If the demand curve is even more elastic (flat) this extra damage not considered will be even larger!

Question 11

DHC: What are the 5 methods for calculating the but-for price?

Answer 11

• Before/After method:
  
  • use prices in same market from before or after cartel period
• Yardstick method
  
  • use prices in a different comparable (geographic) market for the same product
• Cost-based method
  
  • use accounting data to calculate average cost + normal return
• Econometric method
  
  • use time-series regression to estimate the overcharge
• Oligopoly model
  
  • develop a theoretical model of competition in the market and
  • use real world data to parameterise the model to quantify the effects

Question 12

Pros and Cons of the methods of calculating the but-for price?

Answer 12

• Simple methods (B/A and Yardstick) dont control for other factors e.g. demand and supply conditions were very different before and after the cartel came into play
• How do we know that prices before and after or in different geographical markets are competitive –> these prices could have been affected/influenced by the cartel and arent wholly accurate
  
  • Also if firm’s in the cartel know damages are going to be calculated using the price before and after the cartel disappears –> they would raise the after price as much as possible to reduce what they would have to pay
• Cost-based method
  
  • issue is it start of with perfect competition for it calculations but we know that the Cartel’s arise in markets with few competitive firms
    
    • seems like a strange way to calculate what the price would have been without the cartel –> market would have always been imperfectly competitive anyways
• Oligopoly model:
  
  • You need to be able to develop a theoretical model of competition in the market and this is always going to need some simplification –> so you may miss out on some of the important factors that are going on in the market

Question 13

DHC: How is the Econometrics Method performed?

Answer 13

• probably the most sophisticated and best method out of the lot (although not flawless)
• The lines represent the point at the cartel entered the market and artificially raised the price
• Beta₂ –> will give us our estimate of the overcharge

Pros of this method –> more sophisticated can control for other factors such as cost and demand /supply conditions

Cons of this method –>

• biggest issue are trying to decide/ find out when the cartel actually started and when it ended –> this could influence the line of the best fit of the competitive price (by including some of the cartel prices as a later start date was included and shifting it up) –> this reduce the estimated overcharge
• Similar issue to the after method –> incentive to raise prices after-the-fact to reduce the damages they would have to pay by also raises the line of best fit due to higher after-period prices

Question 14

DHC: What was the market like?

Answer 14

• Product market:
  
  • District heating pipes Pre-insulated pipes to supply hot water from district heating stations to households
  • Discontinuous (standard way) vs continuous pipe production (cheaper and faster to do with less materials)
• Geographic market: Western Europe (national level)
  
  • Found in Nordic countries, northern Germany, Eastern Europe, Mediterranean
  • Germany largest market (40%) and Denmark second largest (20%)
• Firms: more than 10 firms active in the market
  
  • 4 largest firms: ABB (40%); LR (20%); Tarco (14%); Pan-Isovit (12%)
  • Highly concentrated market where CR4=86 and HHI = 2461
• Customers:
  
  • municipalities (local government)
  • Buy pipes through tender or EU procurement (bid for the job)

Question 15

DHC: Overview of the Case?

Answer 15

• June 1995- investigation started by European Commission
  
  • Tipoff from competitor Powerpipe (Swedish company)
  • The cartel continued 9 months after investigation began
• October 1998 - decision of EC delivered
  
  • The infringement occurred between 1990-1996 (though may have started in 1988)
  • There were 10 cartel members in total, with four core Danish firms
  • The cartel was active in Demark initially (1991), then expanded to Italy/Germany (by 1994) , then EU
  • The behaviour included: price-fixing, market sharing (firms in certain countries would supply certain markets (German companies service German municipalities), delayed innovation & predation (try and knock Powerpipe out the market)
• March 2001 –> decision of Court of First Instance (later upheld by ECJ)
  
  • Fines totalling €92.21m were reduced by €5.1m on appeal
• 2004/5
  
  • Four Danish municipalities sued 3 main Danish firms for damages (ABB, DRI, LR)
  • Settlement in court in which 3 firms paid total of €21m (57% of claimed damages)

Question 16

DHC: What method for determining damages was used?

Answer 16

• The parties used two of the five methods.
  
  • After method –> used by the plaintiffs to calculate their claimed damages (€38m)
    
    • The lowest price obtained in first tender after the cartel period
    • This was 35-40% lower than during the cartel period
      
      • Defendants complained: If prices applied to all customers, firms would have made a loss
      • Prices were low not because of the disassembled of the Cartel but actual from lower demand in general
  • Before method –> used by defendants to calculate alternative damages
    
    • Calculation of damages decreased by 80%!
    • Used price in 1990 despite some evidence indicating that cartel started in 1988
      
      • Probably picked this as prices could have already been collusive and would have reduced the ‘but for’ price and reduced the overcharge and damage payment
  • Cost-based method
    
    • one defendant provided another estimate –> which had nothing to do with prices however
    • Damages 20% of the claimed damages
      
      • Also, delay innovation by using the old discontinued pipes –> which raised the ‘but for’ price

Question 17

DHC: What method for determining damages was not used?

Answer 17

• Yardstick method –> although there was data on another geographical market between 1990-91 before the cartel had gone from Denmark to Germany
  
  • Prices were 15-20% lower in Germany than Denmark
• Econometrics method
  
  • Requires data of a level of detail that is not possible to provide in this case
  • (There appeared to be detailed information on prices, however)
• Oligopoly model
  
  • The defendants ‘successfully’ argued the Cournot model was appropriate(Though this is questionable)
  • The parties chose not to use it due to an erroneous critique of the model
    
    • Said that the decision-makers with not rational in the model (even though they are rational-profit maximisers)
    • And the model assumes the product is homogenous –> it is a simplification of the model, but you can be sold for a Cornot equilibrium for undifferentiated products