A PYMNTS Company

Devising Pro-Competitive Rebates

 |  April 17, 2013

Apr 18, 2013

CPI Europe Column edited by Anna Tzanaki (Competition Policy International) presents:

Devising Pro-Competitive Rebates by Hans Zenger (Charles River Associates)

(Click here for a PDF version of the article.)

Introduction by Europe Column editor Anna Tzanaki

In this month’s Europe Column, we are delighted to host Hans Zenger, winner of the 2013 Antitrust Award in the category of economics (for academic articles). This shorter paper revisits the thorny issue of loyalty rebates and the potential of foreclosure under EU competition law. It analyses the transition from a per se prohibition of loyalty rebates to a more sophisticated two-stage antitrust analysis under an effects-based approach. However, despite the shifting policy guidance and EU case law, the author recognizes the grounds for legal and economic uncertainty in assessing rebate schemes, hence the dominant firms’ reluctance to go beyond the traditionally set path. Most valuably, this paper provides some practical advice under simple rules of thumb on how dominant players can effectively apply pro-competitive rebates and compete rigorously with rivals in the market while avoiding regulatory risks.

We hope you delve into the discussion and enjoy reading!

The traditional per se prohibition of loyalty rebates under Article 102 TFEU had been a source of concern to almost anyone but the most orthodox proponents of orchestrated markets. Faced with a restrictive and form-based case law, firms that had gained a lead in the market usually saw no other option than to refrain from competing in rebates. Against this background, the Commission’s more recent commitment to assess loyalty rebates under an effects-based approach is arguably one of the more significant improvements of EU competition policy in the past decade.

In its 2009 Guidance Paper, the Commission sets out a two-step procedure to evaluate the competitive implications of loyalty rebates. In a first step, the “as-efficient-competitor test” is used to assess whether rivals effectively had to compete with below cost prices as a result of the dominant firm’s rebate scheme. If this test is failed (i.e., if some pricing below cost is found), the Commission ascertains in a second step whether this conduct was likely to harm competition and consumers. Such anti-competitive foreclosure can for instance arise if a dominant firm forecloses such a large part of the market that rivals are impaired in their ability to remain a viable force in the market.


The new policy has appreciably expanded the scope for dominant firms to engage in competition through loyalty rebates. Yet, four years after the publication of the Guidance Paper, the practical implications still remain limited in the eyes of many observers. Indeed, dominant firms often continue to shy away from making use of the possibilities opened up by the Guidance Paper. Arguably, this reluctance to use pro-competitive loyalty rebates mainly has two reasons.

First, with the Intel appeal still pending in court, legal certainty is often perceived to be low. In view of the multitude of unfavorable precedents, many firms are afraid that competition authorities will ultimately use the traditional case law as leverage to press ahead with charges that cannot be sustained on the basis of substance and merit. Second, there is a widespread concern that estimating the contestable part of demand (a key ingredient of the as-efficient-competitor test) is not straightforward in a compliance exercise. Indeed, what is found as “contestable” ex post when competition authorities carry out their investigation may turn out to differ from the judgment the firm itself formed ex ante when conducting its self-assessment.

The uncertainty surrounding self-assessments in a pricing context is well-known. In the face of an interventionist case law, firms understandably suspect that any margin of error will eventually be used against them. Yet, when properly approached, a lot can be done to find a proper balance between the desire to compete rigorously and the regulatory risks inherent in compliance assessments. In what follows, I will outline a number of simple rules of thumb that can be effective to steer clear of the risk of violating the as-efficient-competitor test while maintaining a competitive approach to the market.

1. Compete head-to-head when rivals can contest all units

In a large number of instances, the perceived difficulty of estimating the contestable share of demand is actually a non-issue. Specifically, this is the case when smaller competitors can compete for all units of different customers. For instance, in Coca-Cola, the well-known soft drink producer competed with Pepsi for the supply of restaurants and concert venues. With regard to such outlets, Pepsi is clearly in a position to fully contest each individual customer (as evidenced by the fact that Pepsi is the exclusive supplier for a large number of such venues). With full contestability, however, the Guidance Paper test merely requires that the average price for all units a customer purchases remains above cost, a condition that is typically not difficult to verify. Compliance with the as-efficient-competitor test is therefore straightforward in such cases—irrespective of the type of rebate a dominant firm chooses to grant.

For instance, the above logic was successfully applied in CRV Holding, a Dutch rebates case on the market for bull sperm. In this case, a Dutch appeals court overturned the original prohibition decision of the NMa, despite the fact that CRV held a market share of 90% and had used a variety of loyalty schemes (including retroactive rebates and exclusive dealing incentives). While competitors were small in size, they were in a position to effectively compete for the entirety of an individual customer’s requirements. Hence, there was no sense in which CRV could have leveraged market power from a non-contestable part of demand into contestable sales.

2. Use incremental rebates for must-have products

When a dominant producer sells a “must-have” product, things can quickly become more complex. In such cases, more care has to be taken to structure rebates appropriately. The simplest way of avoiding a violation of the as-efficient-competitor test in such situations is to grant incremental rather than retroactive rebates. Most notably, the use of incremental rebates again avoids the need to estimate the size of the contestable share of demand. Compliance can be ensured simply by keeping the lowest applicable price in the scheme above cost.

3. Use multiple thresholds for must-have products when retroactive rebates are granted

There can be numerous legitimate reasons why a dominant firm would nonetheless want to apply retroactive rebates or other more complex incentive schemes. In some instances, this is simply due to the structure of internal accounting systems. For example, enterprise software is sometimes constructed around the use of retroactive rebates (an indication that such rebates are far more common in normal markets than the case law would have us believe). Moreover, customers often prefer retroactive rebates because they are expressed in terms of average prices (the figure buyers are ultimately interested in). Perhaps most importantly, more complex rebate schemes tend to permit firms to compete more effectively for incremental sales. (More details on the pro-competitive reasons why firms use retroactive rebates can be found here.)

If the producer of a “must-have” product thus decides to apply retroactive rebates despite the increased risk they involve, it is crucial to structure the scheme intelligently to minimize the likelihood of causing foreclosure. The simplest way of addressing the uncertainty surrounding the measurement of contestable shares in this case is to apply multiple rebate thresholds. Indeed, the larger the extent to which retroactive rebates are spread across different quantity thresholds, the lower the likelihood that measurement errors can turn pro-competitive rebates into an instance of alleged foreclosure. Since multiple thresholds smooth the “cliffs” in loyalty schemes, they permit a construction of rebates that ascertains above-cost pricing even if one allows for appreciable margins of error in measurement.

4. Compete for each customer

It is also important to recognize that individualized rebates are generally in line with the as-efficient-competitor test. Whereas old case law, such as Hoffmann-La Roche and Michelin I, typically classified targeted rebates as illegal price discrimination, more recent case practice, such as Tomra and Intel, has abandoned advancing this allegation. In line with this new approach, the Court recently found in Post Danmark that price discrimination does not in itself cause exclusionary harm to competition. This altered perspective on individualization is consistent with an effects-based approach toward rebates, since individualized prices are instrumental for firms to compete effectively for contestable units with their rivals.

5. Avoid punishments and threats

Finally, great care should be taken to avoid threats and punishments if customers shift orders to rival suppliers. When conducting the as-efficient-competitor test, competition authorities are likely to take account of any punishments that customers would be subject to. Indeed, from an economic perspective, properly derived counterfactual prices should consider the actual price changes that a customer would face when it reduces demand—including possible withdrawals of bonuses and discounts as a punishment for dealing with rivals.


Although the Guidance Paper approach toward loyalty rebates has substantially expanded the scope for applying pro-competitive rebates, many dominant firms are still reluctant to compete intensely with loyalty rebates. However, when properly applied, a number of simple rules of thumb permit devising loyalty rebates in such a way as to minimize the risk of violating the as-efficient-competitor test, while maintaining a competitive approach toward pricing. For many dominant firms, there may still be significant scope to find a more competitive balance between commercial desires and regulatory obligations.