Error Types

The Role of Error Analysis in Antitrust Cases and Why Antitrust Cases are Vulnerable to Erroneous Decisions

The Error Type concept has been borrowed by antitrust scholars from the field of statistic sciences, where it is commonly used to estimate the probability of errors in a sample being tested. It is used to determine whether there is a relationship between variables and potential error in certain population from which sample data are drawn.

Error analysis is important because it enables more accurate scrutiny of facts. It is as well important in the field of antitrust to ensure that cases do not lead to a high amount of false judgment whether Type I error or Type II error. But first, what are Type I and Type II errors?

Type I error is defined as the incorrect rejection of a true null hypothesis. It is also known as a false positive or an error of the first kind. For example, a fire alarm goes off indicating a fire breaking out when in fact there is no fire. Or a blood test result that shows a patient has a certain disease when in reality the patient does not have that disease.

In antitrust cases, Type I error represents a false judgment in which the court condemns a conduct that was not anticompetitive. Type I error reflects an over-enforcement or over-regulation.

Type II error is defined as the failure to reject a false null hypothesis.  It is also known as a false negative or an error of the second kind. For example, a fire breaking out and the fire alarm do not go off. Or a blood test fails to indicate that a patient has a certain disease.

In antitrust cases, Type II error represents a false judgment in which the court fails to condemn a conduct that is anticompetitive. Type II error reflects under-enforcement or under-regulation.

Table 1-1 is a matrix illustrating potential outcomes of the Type I and Type II Errors and correct outcome.

lkdnflkdsjflkdsjflkdsjflkdsjf

 

Error Types analysis and its importance in the antitrust field

Error Types are important for competition law enforcement and policy because both Error Types have a negative impact on consumers’ well being and market integration. Type I errors force firms to act precautionary when facing the risk of being falsely accused of anticompetitive violations. As a result, companies will be less inclined to engage in business transactions. This may reflect negativity on promoting businesses, limit availability of products and services to consumers, and harm market integration overall. Similarly, Type II error has its negative impact. It has the potential of harming consumers by failing to detect potential monopoly.

The European Commission always attempts to promote market integration policy and consumers’ protection. Antitrust cases, however, are extremely complex because it may present agreements or conducts that are not clearly anticompetitive. As a consequence of such complexity, antitrust enforcement errors may harm genuine competition. Error Type analysis helps enforcement agencies to avoid erroneous outcomes.

Error Types Analysis for agreements and concerted practices cases studied 

The main objective of this analysis is to illustrate one way to look at the probability whether the CJEU might have erred when examining parties’ conduct under Article 101. In addition, the analysis will assess whether CJEU has an error preference when deciding antitrust violations. Finally, this analysis will help to show a tool that could be useful in looking at trends for the CJEU when deciding antitrust violations.

The first step is to create a table containing cases under examination and relevant variables such as: parties’ relationship; whether vertical or horizontal; type of conduct; whether agreement or concerted practice; the CJEU decision and the AG’s opinion’, and what error types the CJEU or AG apparently prefer.

Once the table is populated with data, the second step is to look at CJEU decisions and AG’s opinions.  If the CJEU decision follows the AG’s opinion, then there is perhaps a lower probability of an error. If the CJEU decision does not follow the AG’s opinion, then there is perhaps a higher probability of an error.

Table 1-2 illustrates the above approach.

 Table 1-2[1]

lkdsjflkejflksjf;kdsjf;jf;lkdjf;lkdsjf

Method to measure the probability of an error

The following formula is created to represent the variables that might be relevant when examining the probability of error committed by the CJEU or AG. It is important to highlight that this formula is for illustrative purposes and has not been examined by our team due to limited resources. Future competition law and Article 101 studied with more resources and time might consider developing empirical data to examine this formula, test its accuracy and potential usefulness.

The proposed formula is:

P = α + βX + βY + βZ

α = (51% more likely than not) to (90% beyond reasonable doubt)

X = Agreement or Concerted Practice

Y = Presumption

Z = Standard proof

β = coefficient

Findings and Conclusions

The table demonstrates that the CJEU error preference is Type I Error. Out of the twelve cases we studied, eight cases which the court preferred to commit Type I Error. That is, the court would rather find an agreement or concerted practice in violation of Article 101 even if there is not anticompetitive object or effect.

On the other hand, the AG   prefers Type II Error. From the twelve cases we studied, there are seven cases in which the AG preferred to commit Type II Error. This means the AG would rather find an agreement or concerted practice not in violation of Article 101 even if there is an actual anticompetitive object or effect.

This is also evidence from the cases were the CJEU and AG disagreed on the outcome. For example, in Consten Grundig the CJEU clearly preferred to commit Type I Error by finding a vertical agreement between electronic manufacture in Germany and distributor in France in violation of Article 101. No facts supported such finding. The AG argued for Type II Error by stating that the agreement did not have any restriction on competition by either object or effect. In fact, there was support that the agreement help the German distribute to enter the French market and made electronics available for France consumers.

 

 


[1] A (Agreement), CP (Concerted Practice), H (Horizontal relationship), N (No violation of Article 101), V (Vertical relationship), Y (violation of Article 101), CJEU (Court of Justice of the European Union), AG (Advocate General)