Thursday, April 19, 2007

Cartel vs DC

2007 maybe the good year for the all the supporter of free competition. Up till now (April 2007), the Directorate of Competition (DC) at EU already fined two major cartel :

1. Elevator and Escalator Industry for price fixing in the Netherlands, Luxemburg, Belgium and Germany.

The convicted firms are amongs the leading firms in the world : Kone, Schindler,ThyssenKrupp, Kone, Mitsubishi. The total fined collected from these convicted firms are the largest fined in EU history. These five firms owns aroung 76 % of the total sales in elevator and escalator worlwide

Based on the previous experience in the vitamin and lysine cartel, no wonder that these firm will be also investigated by the Anti Trust authority in other countries. Rumors said that there is a current on-going investigation in Hongkong also for this industry.

This is a long time investigation for european case. I think the DC already investigated this cartel for more than two years. Luckily there is a leniency program in the EU. As EU start the investigation on 2004-2005, one firms directly use the the Leniency Program.
2. Brewery in Netherlands. The convicted firms are Heineken, Grolsch and Bavaria. Again the same story happens here. After seven years of investigation, there is another cartel member InBev who give more evidence on the existence of the cartel. InBev use the leniency program after he get caught of the same brewey cartel in belgium (belgium brewery cartel).

Just a couple of weeks ago, DC also annouced that they will start the investigation on the possibility of price fixing cartel in the industry that produced bathroom and toilet utilities. I am quite sure that within this years, there will be at least one firm that will report to the DC about this cartel and thus get the leniency.

So what can we learn from this??

1. There is a tendency that the the number of cartel exposed and convicted increases after the leniency program

2. The Leniency program is effective after the Antitrust Authority announce that there will be an investigation on the specific cartel (this sample also supported by the theoretical work by Motta 1999, 2003 on Leniency Program)

3. The announcement effect in fact not only happen in the monetary field. It also happen in the collusive behavior.

4. Full cartel is very rare. Most of the cartel is partial.

The moral of this story :
" Make The Law of Leniency Program in Indonesia"


Berly said...

Hmm.. but if the industry start to form expectation of future leniency with better term they has little incentive to break the cartel now.

The Dreamer said...
This comment has been removed by the author.
The Dreamer said...


I do not get it. How could the LP have positive effect on the cartel stability? Could you please give more detail?

The LP will give more incentive for cartel member to report the cartel existence to AA. Off course, this will be an equilibrium iff the Aa already start the investigation. The main reason is because the cartel member want to minimize the possible fine.

It is true on whether the distributional assumption yuo imposed on the probability on detection with and without LP.

This could be see on the IC and IR constraint of the cartel member with and without LP. By considering these, we could get that the cartel stability certainily negatively correlated with LP

2:00 AM

Berly said...

If the setting is one shot game then you are correct.

But if it is a repeated game (as in almost every case) and the anti-cartel body has credibility problem, then it is another matter.

After certain treshold of cost-benefit of breaking the cartel and expectation of future leniency program, there is a case where leniency program will increase cartel stability.

Economists just love a good paradox :-D

The Dreamer said...
This comment has been removed by the author.
The Dreamer said...
This comment has been removed by the author.
The Dreamer said...

1. I agree that almost all the analysis in cartel stability use supergame. Supergame is the infinitely repeated game. The results in this repeated games shows that LP has negative impact on cartel stability.

So far the analysis is the supergame :D

2. Credibility ----> doesn't necessarily mean that the LP have positive correlation with the cartel as you said. It could mean that the LP doesn't have any effect on the cartel stability.

That's why I mentioned that the LP will likely be effective if the AA already announced that they will investigate one specific cartel.

3. It is true that there could be a multiple equillibria in a certain games.

One of this equilibria involve an in-creadible threat. However, this increadible threat strategy actually is not a dominant strategy. It should be a (weakly) dominated strategy. Therefore, this strategy will never be played if the game arrived at the nodes. That's why this possible nash equilibria can be eliminated using nash refinement (method).

I think, reporting to the AA, thus get a reduced fine is not a (weakly) dominated strategy. This startegy is certainly better of rather than not reporting (thus get full fine) if the AA announced that they will investigate the cartel.

4. "After certain treshold of cost-benefit of breaking the cartel and expectation of future leniency program, there is a case where leniency program will increase cartel stability."

You said that there is case where the LP will increase the cartel stability. Could you please give the intuition for this? (not just a hunch by saying that there-will-be-a-certain-cut-off-etc)Why should the cartel more stabile?How does LP affect the incentive compatibality and individual rationality?

I think the LP will increase the tempation to report the existence of the cartel (in case that they know the AA will investigated them-they will prefer to report and thus get no-fined rather then get a fined). This temptation to deviate the report to AA, off course finally decrease the incentive compatibality of the cartel member. Unless there is anothers assumption on the characteristics of the industry of the cartel (Bos, 2006)

5. After all, the good theory is a theory that could better explain the real economic phenomenone. If this is the standard of argument, the empirical evidence shows that the punished cartel increased huge after the existence of LP (including cartel member that use LP). I do not know about you, but for me, at least it is a good sign that the cartel stability indeed decrease as the LP introduced by the LP. (although we couldn't give a correlation due to selection effect)

While that, the experimental evidence shows the quite the same thing. The price distrubution decrease as the AA introduce the LP. (see Hinloopen, 200?)

6. Paradox ---> analysis that is anti-intuitive (wikipedia explanation).

The intuition is the main difference between the mathematician and economist. The problem is : there is a logically correct intuition and there is logically not-so-correct intuition.I think, most of economics theory should be supported by an logically correct intuition.

Thank you anyway for the discussion Ber. :D

The Dreamer said...

Ber, after all, this is just my opinion :D

please, cmiiw