1

How KPI of competition authority can distort enforcement targets in competition

investigations: the case of Russian Federation

Svetlana Avdasheva (avdash@hse.ru ), Dina Tsytsulina (dtsytsulina@hse.ru), Svetlana Golovanova

(svgolovanova@hse.ru), Higher School of Economics National Research University

October, 2015

ABSTRACT

Incentives of competition authority may provide strong influence on the standards of evidence

applied in the authorities’ investigation, and also by the courts that resolve competition cases. In the

Russian competition enforcement motivation of the authority to achieve high ratio of the decisions

coming into force results in the distortion of enforcement structure and standards of evidence in

investigations.

Using a unique dataset of the appeals of infringement decisions in 2008-2012 years, we classify the

investigated cases according to their potential impact on competition. The analysis reveals that the

majority of cases would never be investigated under an appropriate understanding of the goals of

antitrust enforcement, restrictions on competition and basic cost-benefit assessments of agency activity.

Our analysis shows that antitrust authorities tend to investigate cases, which require less input in terms

of economic evidence but result in infringement decisions with lower probability of being annulled.

Typical targets of enforcement are cases where harm is an independent evidence of competition law

violation.

Key Words: antitrust enforcement, authorities’ incentives, motivation, competition, harm, Russia.

JEL classification: K21, K42

ACKNOWLEDGEMENTS

The article is a result of the Basic Research Program Project of the Higher School of Economics, National Research

University. The authors cordially thank Yoannis Katsoulacos for his outstanding advice on the collection and design of the dataset,

the strategy for data analysis development, discussion of conceptual framework and policy recommendations, Vadim Novikov for

providing the initial dataset, and discussing the conclusions, Polina Kruchkova, Maxim Markin, Andrey Tzyganov, Vitaliy Dianov,

Alla Varlamova for helpful comments on legal issues in enforcement of Russian antitrust legislation, Marina Kolosnitsyna for

helpful comments and suggestions on the organization of results’ presentation, and Marina Lopatina, Karina Fayzullina, and Yulia

Modina for ideal research assistance in collecting the database. The usual disclaimers are applied.

2

I. INTRODUCTION

Appropriate standards of evidence are sine qua non condition for effective competition enforcement. Under given

substantive norms they emerge as a result of interaction between judges, suspected in violation and competition

authority. The latter is crucial under administrative system of enforcement. Indicators of performance assessment of

competition authority are no less important than welfare standards of enforcement. They are of special importance for

the newborn competition enforcement, without considerable experience in the field. Leading by performance

indicators, competition authority takes enforcement actions, which create interpretation of substantive norms by

market participants and courts. It is especially important when government applies for competition authority specific

set of performance indicators (key performance indicators, KPI) to assess the outcomes of efforts.

Positive analysis of enforcement allows explaining how institutional features, including model of performance

assessment contribute to the development of competition enforcement. Russian competition enforcement, which is

relatively young and still under formation, is of interest in this respect. Extremely large number of antitrust

investigations in Russia presents a challenge for the assessment. On the one hand, many infringement decisions may

be interpreted as an indicator of high enforcement efforts. On the other hand, if investigations proceed under poor

legal and economic standards the impact of enforcement on competition is questionable. Large number of

investigations may indicate even the ineffectiveness of enforcement, and in any case represents miracle that requires

an explanation.

One obvious explanation is that competition authority does not properly understand the goals of enforcement.

Antitrust is aimed to prevent anticompetitive behavior, thus restricting its negative effects on welfare. The distinctive

feature of antitrust legislation is that being welfarist also means process-oriented (Farrell, Katz, 2006). Only actions

that reduce social welfare through restrictions of competition are prosecuted. In this respect antitrust legislation differs

from other legislations which have similar purposes to protect one group of economic agents from harmful actions of

another group such as consumer protection legislation and legislation relating to the regulation of natural monopoly

activities where liability rests on just a finding of harm to others. That is why in most countries the policies are often

governed by different and independent institutions or at least by independent structural divisions of the same institution

with well-defined and distinct sets of responsibilities.

In Russia the legal backgrounds are the same. The fundamental objective of antitrust or more broadly of

competition policy is to protect competition rather than competitors. However, outcomes are different at the end, and

that is what we plan to discuss in this article.

The Federal Antimonopoly Service (the FAS, hereafter) is the authority that controls the application of the

antitrust legislation in Russia. The FAS is responsible for regulating and controlling compliance with antitrust law, as

well as with regulations of natural monopolies, advertising, procurement for the federal government and foreign

investment in strategically important sectors. Thus its functions are defined wider than in other countries. The

resources and field of responsibility of the FAS are constantly increasing. In the summer of 2015, the Russian

Parliament State Duma went one step further by subordinating to the FAS the authority responsible for tariff

regulation, i.e., the Federal Tariff Service. This step among other steps reflects the belief in the superiority of antitrust

authority performing this task in addition to other responsibilities.

At first glance, the goals of antitrust enforcement are defined clear enough. However analyzing the decisions of

antitrust authorities, we find that there are substantial overlap of Russian antitrust enforcement with other areas of

3

economic policies, namely consumer protection and conflict resolution. FAS subdivisions often consider cases, which

are not proper antitrust cases. For example, commercial conflicts causing harm to one of the parties or failure to

comply with natural monopoly regulations are frequent targets of antitrust investigations

This gives us an alternative vision of the fact that according to the Rating Enforcement, Global Competition

Review, in 2013, Russia led the number of investigations in the rating of competition authorities all over the world.

Specifically, the FAS investigates more abuse of dominance cases than all other competition authorities in the world:

in 2013 - 2,635 investigations were opened, 2,212 were cleared.

The contradiction between high number of cases under investigation and consideration, on the one hand, and the

complexity of a typical antitrust case that requires the application of high standards of economic analysis, on the other

hand, has been discussed many times (Girgenson, Numerova, 2012). The FAS resolves this contradiction by

decreasing the quality of the decisions in terms of the economic analysis undertaken. Even the data of the Rating

Enforcement show that decisions of the FAS are ‘cheap’ in terms of the resources spent: the average duration of an

investigation is only 3 months (it is almost 10 times longer for cases in the European Commission).

There are competing explanations of the combination of large-scaled enforcement with modest targets of

investigations and infringement decisions, including lack of experience of competition agencies and commercial

courts, demand for antitrust enforcement to achieve ‘complementary’ policy objectives not specific for antitrust. Lack

of experience that is often indicated as shortage of human capital in relatively young competition authorities (Gal,

2010) is important but does not explain why non-proper antitrust investigations and decisions take larger share in the

investigations over time. An important complementary explanation stresses the importance of procedural rules for

selecting cases for investigation in Russian competition policy and other areas of control and monitoring (Avdasheva,

Kryuchkova, 2014). The legal rules of administrative actions in Russia attribute a high importance to complaints,

which makes the absence of a response to complaints expensive for every official at the FAS. Because of the

importance of complaints, antitrust enforcement is skewed towards cases with large individual effects (where harm to

specific market participants is more important than restrictions on competition). However, effects of motivation of

competition authorities on the selection of cases for investigations is not studied enough though it could be of primary

importance. Specific feature of motivation of competition authorities in Russia is that the latter are motivated on

number of tasks (cases) that are successfully completed by positive as well as negative incentives. Legal rules of

working with citizens’ complaints impose negative incentives on the number (‘no complaint should remain without

answer’) while internal system of motivation of FAS positive incentives (‘higher share of infringement decisions that

came in legal force’) to the number of technically successful decisions.

The goal of this article is to explain the impact of the incentives of authorities concerning antitrust enforcement

on the structure of enforcement and understanding of the substantive norms and welfare standards in Russia using

case-level evidence. We use the unique dataset of the claims to commercial courts to annul the infringement decisions

of the competition authorities from 2008-2012, which represents more than one-third of all the FAS decisions

(collected by the Laboratory of competition and antitrust policy, Institute of Industrial and Market Studies Higher

School of Economics, LCAP dataset hereafter). We also combine quantitative and qualitative analyses to restore the

understanding of antitrust prohibition in the practice of enforcement. We discover at least three important drawbacks

to the practices of judges and the public authorities regarding antitrust legislation. The first shortcoming is the

interpretation of the main goal of antitrust enforcement as prevention of the harm imposed on market participants,

irrespective of the size of the harmed group and the magnitude of harm. The second drawback is the interpretation of

4

any harm imposed by the dominant seller in the contractual relationship as sufficient evidence of abuse of dominance.

The third weakness is the interpretation of any loss or dissatisfaction of the counterparty to the dominant supplier as

evidence of harm.

The content of the cases shows that not only competition authorities but also judges often consider harm is

sufficient evidence of a competition law violation. The probability of a successful claim to annul the infringement

decision of the FAS is significantly lower for cases where the harm imposed is independent evidence of a competition

law violation. In turn, it is an important advantage for competition officers to consider these cases because they are

incentivized by the large number of rapid decisions with a low likelihood of reversal by the courts. Emerging standards

of evidence in antitrust cases concentrate attention on structural features (dominance) and then directly on harm, which

is defined in such a broad way and does not sufficiently consider the restrictions of competition. Significant efforts

and complex changes are necessary to correct the distortions in Russian antitrust enforcement.

We concentrate the research on specific Russian enforcement and use data on national commercial court system,

but research question is important not only for competition enforcement in one transition country. Competition policy

is among the most analysis-intensive directions of economic policy, and analysis of competition becomes more

complex. Category-based enforcement (enforcement of per se prohibition) is displaced by effect-based enforcement

(enforcement of provisions based on rule of reason) that requires large intellectual resources, and also lot of spending.

At the same time many areas of competition policy become more complex. Governments widely accept the idea that

competition protection cannot restrict itself on the enforcement of few simple rules, i.e. do not fix prices and share

markets with your competitors. At the same time organization of competition policy tries to capture synergies of

specialized expertise in the assessment of welfare effects of public intervention. Expertise of antitrust authority might

be successfully applied in consumer protection for instance. In many countries, including OECD and non-OECD,

competition authorities become responsible for enforcement of access to networks in deregulated industries and

consumer protection. Finally, in the framework of effect-based enforcement notion of efficiencies expands. Narrow

understanding of total and consumer welfare as a function of cost, price and quantities are often complemented by

long-term considerations of sustainability of economic development and long-term competitiveness.

These developments might affect organization of public authorities responsible for different areas of competition

policy. First, many tasks become more difficult to fulfill and more demanding of resources. Second, among the tasks

competition authorities should fulfill the difference arises between ‘simple’ and ‘different’ targets. Traditional split

between ‘category-based’ enforcement and ‘effect-based enforcement’ is only one example of the divide. Another is

competition protection vis-à-vis competition promotion: cases from these two groups require different types of

expertise, with different level of ‘difficulty’.

Difference between ‘simple’ and ‘difficult’ enforcement targets is not only in terms of cost but also in terms of

welfare effects, and this difference is not straightforward. Other things being equal, antitrust proceedings in per se

cases are less difficult than those rule-of-reason type of cases because the latter require complex economic evidence

and analysis of specific efficiencies. Competition investigations of large multinationals are more difficult than the

cases against medium-sized companies, at least because larger market participants have higher incentives and higher

budgets to defend their position. In innovative industries, economic analysis is more difficult than in mature ones.

Allocation of scarce resources – including efforts that we consider throughout the paper - of the authority predicts

welfare effects of overall activity. Decisions of authorities differ from those predicted by total welfare maximization.

Policies of prioritization within public authorities try to overcome mismatch between socially desirable resource

5

allocation and resource allocation desirable for the staff in public authority. In spite of close attention to this issue in

the policy debates and important contribution in modelling of prioritization within competition policy in the

framework of multi-tasking (Schinkel et al, 2014), recently there is little or no empirical analysis on the allocation of

resources of competition authority. The aim of the article is to fil this gap.

The article is structured as follows. Section 1 presents brief overview of the role of performance indicators in the

allocation of efforts in public agencies, including competition authorities. Section 2 briefly describes the development

of antitrust legislation and enforcement in Russia. Section 3 describes research strategy and data. Section 4 analyzes

the structure of cases providing the comparison between competition restrictions and the harm imposed. Section 5

discusses the impact of ‘proper’ and ‘non-proper’ antitrust cases on the cost and performance indicators of competition

authorities. Section 6 concludes.

II. PERFORMANCE INDICATORS AND PRIORITIZATION IN ANTITRUST ENFORCEMENT

Motivation in the public sector attracts close attention of researchers (see Dixit, 2002 for survey of the research

agenda). Design of incentives under multiple dimensions of tasks and multiple tasks (Holmstrom, Milgrom. 1991) is

one of the most difficult issues if observability of different dimensions of tasks for principal varies, and efforts on

more observable (but less important or worthless for principal) dimension of task substitute efforts on less observable

but more valuable for principal tasks. Alignment of incentive contract with the achievement of more observable task

crowds out the efforts on less observable one decreasing the utility of principal. This framework traditionally refers

to trade-off between quantity as easily observable and quality that is not. There is an empirical evidence on crowding

out of efforts on quality dimensions after the introduction of performance payments in public sector.

Important part of empirical researches is devoted to the effects of implicitly imposed incentives. Leaver (2009)

shows that unwillingness to make mistake that would be publicly discovered explains decisions of servants in the

agencies responsible for price regulation. The more public officer is exposed to the risk of negative external assessment

the more she refrains from the actions with high risk of making mistake, which are reviews of the regulated rates. The

framework of the empirical part of our research is close to this approach. Bandyopadhyay and McCannon (2014) study

the influence of motivation on the selection of cases by criminal prosecutors. Under the pressures of re-election

criminal prosecutors increase the number of cases brought to trial and prefer to take the cases, which result in

convictions from trial in contrast to guilty pleas and cases with shorter sentence. Both preferences towards ‘no

publicity’ (Leaver, 2009) and towards ‘good publicity’ (Bandyopadhyay and McCannon, 2014) provide distortions in

the structure of activity. These distortions are in the center of our analysis as well.

Empirical analysis of judicial decisions or judicial review of the decisions of competition authority confirms the

natural assumption that difficulty and complexity of cased increase uncertainty of final decisions. However many

other determinants interact with the impact of case complexity. Baye and Wright (2011) show the importance of

experience and training of the US judges, together with the complexity of case, to provide the decision that is not

appealed. Carree et al (2010) in the analysis of European Commission decisions found that probability of appeal on

decisions is explained by the factors that make case ‘difficult’ either because of the high requirements for effect-base

analysis (decisions on abuse of dominance) or because of high incentives of the party to appeal the case (average fine

on violator). In the empirical part of our research we follow the similar strategy.

6

The model on the selection between ‘simple’ and ‘difficult’ cases by the head of competition authority (Schinkel

et al, 2014) is the most similar to our framework. Replacing the notion of reputation by KPI we can interpret the model

as perfect description of Russian competition enforcement. However in contrast to the cited model we analyze the

environment, where difficult cases do not necessarily contribute to the authority’s reputation more than the similar

ones. We consider judicial review is a factor that determines effect of decision on reputation. Decision that is reversed

by commercial court makes negative contribution to reputation (KPI).

Finally, the evidence on the complexity of modern economic analysis as an input for decisions of competition

authority and judges is important for our research, together with the evidence that modern economic theory hardly

provides undisputable implications to decide certain case (Neven, 2006; Connor, 2008; Shinkel, 2008). These

arguments support the hypotheses we test that an extra unit of economic evidence decreases the probability to make

Type I error (wrongful conviction) but not necessarily decreases the probability of the given infringement decision to

be appealed and reversed by the court.

III. DEVELOPMENT OF ANTITRUST ENFORCEMENT IN RUSSIA: BRIEF OVERVIEW

History of Russian competition legislation and enforcement accounts for quarter century. The blueprint for the

competition legislation was European provisions. The content of the main articles of the law is nearly a precise

translation of the relevant rules of the Treaty of Rome: art. 10 of the law prohibits the abuse of a dominant position in

words very close to art. 102 TFEU, whereas art. 11 prohibits agreements, horizontal and vertical, as art. 101 TFEU

does. In 2007 turnover penalties for antitrust violations were introduced, with the threshold close to that in European

competition enforcement (up to 4% of turner in contrast to 10% in EU).

Specific feature of the Russian competition law is that any practice that is listed as illegal for dominant company

or participants in the agreement may be considered as law violation not only if it restricts competition but also if it

imposes harm on the counterparty. Since 2011, legislator confirms that infringements of competition legislation can

not be related to restriction of competition, by introducing fixed cap of monetary penalties on that type of violations.

The FAS may inspect compliance with legal requirements either on their own initiative (ex officio) or on the basis

of complaints received. Recent developments in control and monitoring in Russia attach great importance to

responding to complaints. A special law, ‘On the procedure of considering complaints of citizens of the Russian

Federation’ (2006), requires a responsible authority to consider every complaint and either open an investigation or

provide a reasoned refusal within 30 days. Authorities and public servants are responsible for both decision-making

delays and unjustified refusals to open complaint investigations. Although antitrust authorities are formally entitled to

select among complaints and cannot be compelled to conduct investigations on every complaint received, they are

strongly incentivized to do just that. Citizens and companies can sue authorities and officials for any harm that is

resulted from inaction (Trochev, 2012). Servants` compensation strictly depends on delays in proceedings and

‘unjustified refusals to open inspections’; such delays and refusals reduce quarterly payments, which represent a

significant share of public servants’ total salary.

The FAS has a power for both inspection and investigation, and decision on the infringement, representing a type

of ‘inquisitorial’ system typical for administrative law enforcement in continental model. Violator has a right to appeal

the decision of the FAS in a court (during the period under analysis – commercial courts concentrated on litigations

in the economic area. Costs of access to court are relatively low in Russia (negligible fees, rule of cost indemnification,

7

no restrictions of representation, no time restrictions on introducing new evidence to litigation etc). That is why

decisions of competition authority are very often appealed in Russia. ‘Generalist’ judges in commercial courts resolve

disputes on the decisions. Due to specific rules of FAS motivation judicial review is of primary importance for the

authority.

IV. DATA AND EMPIRICAL APPROACH

The sample covers apparently all the decisions made by commercial courts in the Russian Federation non the

claims to appeal infringement decisions of competition authorities during 2008-2012. The coverage of our dataset

exceeds one-third of all the infringement decisions and half of the infringement decisions concerning agreements

(horizontal and vertical) and concerted practice (Table 1).

Ratio of claims to annul is high not only for the infringement decisions of the authority but also for decisions of

commercial court of the first instance (more than ¾ of the decisions of commercial courts are appealed, Table 1). This

size makes the database a relevant source of information regarding the standards of proof applied by the FAS and the

commercial courts. During the entire period, Russia’s commercial courts provide us with rapid decisions; on average,

it takes less than one year on the case to obtain the final decision.

Table 1 shows that during even a short period commonly accepted by competition authorities and judges,

standards of proof developed. In 2012, a noticeably lower share of FAS infringement decisions was annulled by the

commercial courts; in turn, the higher courts reversed the lowest share of decisions of the first instance courts.

To describe and explain the standards of proof of competition investigations, we combine qualitative and

quantitative analyses. Using the decisions of the commercial court as an observation, we attribute to the observation

quantitative characteristics, including the following:

features of alleged violations (abuse of dominance or agreements and concerted practice, articles 10 and 11

of the law ‘On protection of competition’ Russian Federation, respectively);

features of the alleged violator (has the alleged violator the legal status of a natural monopoly);

indicators of the court decisions (does the court of first instance satisfy or refuse the claim, do the parties

(claimant or the FAS) appeal, does the higher court reverse the decision of the first instance);

duration of the litigation as an indicator of the efforts the parties have made;

qualitative features of the alleged violation. These features, in turn, are divided into several groups.

One group represents the ‘functional’ features of a violation. For example, we indicate separately non-compliance

with the rules on the final service provision by natural monopolies, non-compliance with the rules on interconnection

of competing networks, access to the network by vertically disintegrated competitors, and conflicts between operators

of local networks and their sub-subscribers. By impact on competition the sample is divided into cases where

restriction of competition represents the main evidence of law violation and cases where the harm imposed is

independent and the main evidence of a presumed violation. In cases where the harm imposed is the primary evidence

of violation, we also divide these into cases where the harm to the group is sufficiently large relative to the overall

market demand or supply in contrast with the cases that consider harm for only a small group (to one physical or legal

person in extremis);

We begin with a quantitative description of the structure of infringement decisions to show a ‘typical’ or ‘average’

decision in Russian commercial court. For every group, we describe typical examples of the alleged infringements.

The combination of qualitative and quantitative analyses allows us to assess the structure of cases in terms of ‘harm’

8

and ‘competition restriction’ as a principal component of proof. To explain the structure described, we compare

resources spent by the parties to resolve certain types of cases. The general research hypothesis (specified for empirical

hypotheses further in the text) is that alleged violations that prevail in the structure of antitrust investigations require

fewer resources from the competition agency and at the same time provide higher performance indicators. Hypotheses

of empirical analysis correspond to the role of harm as independent evidence of antitrust violation in evidence-intensity

of cases.

To test the empirical hypotheses we apply simple comparison of the characteristics of groups of cases that affect

cost of competition authority and their performance indicators. Indicators, which are crucial for performance of

competition authorities according to their internal guidelines, are rates of reversals of the infringement decisions.

Indicators of cost include proxies of economic analysis of substantial and organizational nature. Substantial indicators

are, first, quantitative assessment of any variable important for characterize competition and, second, are Guidelines

for market analysis and competition assessment (with legally required application in investigations of dominance)

discussed before judge. Organizational indicators of economic analysis are specialized expertise provided either by

competition authority or by claimant in the court.

V. STRUCTURE OF CASES: RESTRICTIONS ON COMPETITION COMPARED WITH THE HARM

IMPOSED

Table 1 shows that structure of enforcement is skewed towards investigations and decisions regarding the abuse

of dominance in contrast to agreements in the proportion of at least four times. In addition, in large part of decisions

enforcement targets are infringements with no competition concerns. A ban on abuses of an exploitative but not

exclusionary nature, which is a specific feature of European competition rules (Vickers 2008), plays a more important

role in Russia: the predominant part of investigations and decisions on art. 10 addresses the conflicts in which harm

to the consumers or counterparties is independent and often sufficient evidence of law violation.

The latter group of cases is very representative in the Russian competition enforcement and very specific. They

are not related to competition policy issues according to conventional understanding of the targets of antitrust

enforcement. Respectively we call these two groups of cases as ‘proper antitrust’ (PA) and ‘not proper antitrust’

(NPA) cases. Using the description of the infringement from judicial review we identify three main groups of NPAs

in Russian antitrust enforcement (see Appendix 1):

Cases initiated against natural monopolies due to some harm imposed on counterparty without any

evidence that the harm is intentional and was caused by anticompetitive behavior. In cases of this type the alleged

violation may be a technical mistake in a contract term, or a sign of a low quality of service. There are also several

cases where the alleged violation represents contract terms implied by the current law to discipline consumers in

planning demand.

Conflicts on interconnections with sub-subscribers (when one organization connects to a network

through a device located at the premises owned by another organization) are sometimes considered by the antitrust

authority as antitrust cases on abuse of dominance in the form of a restriction of access to a network, by charging

excessive prices, etc. The dominant position of the accused organization is usually reached by a narrow definition

of the market. Thus, it appears that the essential facilities doctrine is applied to non-antitrust cases;

9

Cases where harm is used as an independent proof of violation (without any evidence concerning

actions that restrict competition). The discretionary definition and vague evidence of harm are not specific for

cases against owners of local networks or natural monopolies. It is so in more than ¾ of the appeals submitted

(see Appendix 1 for explanations and examples).

Fig. 1 indicates the structure of all the infringement decisions across different groups regarding different

infringement evidence between ‘restriction of competition issues’ and ‘harm issues’ and also between ‘harm to

consumers as a group’ and ‘harm to one specific consumer’. The typical infringement decision does not correspond

to internationally recognized and accepted understandings of what constitutes a violation of law ‘On protection of

competition’. The distinction between ‘proper antitrust’ and ‘non-proper antitrust’ cases does not imply that we make

any judgment about the welfare effects of antitrust enforcement or about the reasons of the antitrust authority to

investigate a specific case. The border between proper antitrust decisions (PADs hereafter) and non-proper antitrust

decisions (NPADs respectively) is the presence/absence of specific competition considerations in assessing the

violation of the law ‘On protection of competition’.

Figure 1. Structure of decisions by the primary infringement evidence across presumed violations

Source: LCAP database

Recall that division between PAD and NPAD is specified in Russian competition legislation. They are different

in terms of system of penalties applied (turnover-based penalties are introduced for PAD since 2011 but not for

NPAD). However there is no difference in the procedure of analysis presumed by the law. The latter is important

both for incentives of officers in competition authorities and for expected welfare effect of enforcement. Every legally

successful infringement decision, that is not appealed or not reversed by commercial court if appealed, contributes the

same score in KPI of regional subdivision of FAS or individual score of an officer. As a result, efforts are shifted

towards investigations, which require less cost and at the same time result in infringement decisions with a higher

probability of legal success.

0 500 1000 1500 2000 2500 3000 3500

Horizontal agreements (art. 11)

Vertical agreements (art. 11)

Concerted practice (art. 11)

Abuse of dominance (art. 10)

Natural monopolies: access and interconnection of competitors (art. 10)

Natural monopolies: (non) compliance with the rules of final service provision

(art. 10)

Interconnection with sub-subscribers (art.10)

Other abuse of dominance cases (art. 10)

Restriction of competition Harm to the large group of market participants Harm to certain market participants

10

Assumption that investigations on cases devoted to ‘harm only’ in contrast to investigations on the practice with

competition concerns may be surprising. However standards of evidence on harm imposed in the practice of Russian

commercial courts are much lower than the standards of evidence on the impact of competition. Legal studies of post-

Soviet courts stress the low level of compensation for damages as an important problem (see, for example, Hendley

forthcoming) that impedes deterrence. However, there is the other side of the coin. In addition to the low level of

compensation for damages, and especially for moral damages, the unclear difference between compensation for

damages and compensation for moral damages as well, the causal links between the actions of an offender and the

harm of a victim are established using relatively weak standards in Russian legal practice (Maggs et al 2015). The

prevailing legal approach to this issue is to replace fault-based liability with strict liability. First, it was implemented

in consumer protection law (Reich, 1996); however, it was subsequently extended to other fields of Russian civil law

(Brüggemeier, 2011). This tradition also strongly affects administrative law: executive authorities can make

conclusions regarding harm using very rough evidence. The importance of complaints as the drivers of investigations

reinforces this peculiarity of damage verification: in many examples, the statement of the alleged victim is sufficient

for an administrative decision. Enforcement against exploitative actions of a large seller implicitly shares the concept

of the strict liability of the dominant market participant for damage that is allegedly imposed on every counterparty.

This extends a step further from the European concept of dominant position as a basis for the special responsibilities

of the dominant company (Larouche and Schinkel 2013), which are derived from ordo-liberal tradition. Compared

with art. 102 TFEU, the enforcement of art. 10 of the law ‘On protection of competition’ tends to use the concept of

the strict liability of the dominant seller for any harm imposed on every individual counterparty, with a rough

interpretation of the casual links among dominance, the behavior of the dominant company and the harm imposed.

Under these standards to prove the fact of harm imposed in the commercial courts are relatively clear to all the parties

in litigation. It is not the case for cases with competition concerns. From all antitrust cases only enforcement against

explicit collusions (price fixing and market sharing) can rely on direct evidence (that is however difficult and

expensive to collect). In contrast, in large number of antitrust investigations (for example, on vertical restraints,

predation, discrimination etc.) administrative decision should rely on modern competition economics that in many

cases is unable to provide indisputable implications (Neven, 2006; Connor, 2008; Shinkel, 2008). In this framework,

NPADS (non proper antitrust decisions) are both easier to make (less costly) and easier to defend under judicial review

(of higher probability of being legally successful).

Combination of these two features with limited discretion of competition authorities in selecting the target for

investigation (importance of reaction of complaints) and motivation to make legally successful decision explains

increasing share of NPADs in the structure of enforcement. The structure of infringement decisions made by the

Russian antitrust authority implies the limited positive effects of enforcement on welfare. Standards of economic

evidence collection, presentation and discussion in competition analysis implicitly imply that enforcement target is a

practice with large negative effects on consumer surplus. For the case where practice in question is only non-

application of discount to calculate the receipt for electricity or late connection of residential equipment to local

electricity grid (see Appendix) the use of any tools of economic evidence (including analysis of market boundaries,

calculation of market shares and concentration indexes, assessment of stability of concentration etc.) is excessive.

Main recommendation of OECD experts for FAS that is to improve standards of economic analysis under competition

enforcement (2013) also implies that Russian antitrust policy addresses to large companies, large markets and practice

with large welfare effects. The issue is that the structure of competition enforcement is heterogeneous. ‘Large’ and

‘difficult’ cases coexist with ‘small’ and ‘simple’ ones.

11

In the next section we test the hypotheses, which support our framework.

VI. IMPACT OF “PROPER ANTITRUST” AND “NON-PROPER ANTITRUST” CASES ON THE COST AND

PERFORMANCE INDICATORS OF COMPETITION AUTHORITIES: EMPIRICAL HYPOTHESES AND

THE RESULTS OF TESTING

Public officers in competition agencies are motivated on the quantitative indicators of their activity, which are the

number of decisions made and especially the amount of decisions that commercial courts do not reverse. Under

performance indicators of this type, agencies should prefer to take less costly and less disputable cases. The main

hypothesis is that PADs represent the preferable group of decisions both in terms of cost and in terms of legal quality.

As a measure of cost we apply duration of litigation in courts of the first instances (time that is necessary to achieve a

decision). Therefore, we test the following empirical hypotheses:

H1. PAD infringement decisions have higher legal quality in the system of Russian commercial courts.

H2. Infringement decisions that are concentrated on the harm to the larger group have higher legal quality in

Russian commercial courts.

H3. Cost of judicial review of PAD infringement decision is higher.

H4. Cost of judicial review of infringement decision that is concentrated on the harm to the larger group is higher.

To test these hypotheses we use the LCAP database. Variables of PAD (proper antitrust decision) and

Harm_Group (harm discussed in the decision is presumed to be harm for more than one consumer or counterparty)

are the outcomes of qualitative analysis and classification of decisions. These variables reflect the features of

infringement decisions, which are most specific for Russian antitrust policy and at the same time most important in

our analytical framework. In addition we use variables, which reflect types of violation, classified as abuse of

dominance (Art_10), explicit horizontal agreement (Collusion), vertical agreement (Vertical agreement) and concerted

practice (Concerted practice). As we have shown earlier, in the structure of decisions under article 10 (abuse of

dominance) PADs represent large part. In contrast, decisions on agreements and concerted practice are predominantly

PADs.

We use three variables as indicators of legal quality of infringement decisions, and they are:

reversal decision of the commercial court of first instance (takes the value of 1 if claim to annul infringement

decision is satisfied and 0 otherwise);

appeal on the refusal to satisfy claim by the commercial court of the first instance (takes the value of 1 if a

company decides to appeal the refusal to satisfy the claim on annulment and 0 otherwise),

reversal of infringement decision by any higher instance court if a company appeals unfavorable decision of

the first instance (takes the value of 1 of any if higher commercial court reverses infringement decision and

0 otherwise).

Decision of every instance of commercial courts is important. As Table 1 reports, not only large part of the

infringement decisions of competition authorities is appealed. Refusals of the first instance courts to satisfy the claim

on annulment are appealed even more often. It is difficult to distinguish between the impact of appellation, cassation

and Higher Court on the final decision. However, we compress all the instances higher than the first instance in one.

According to the tradition of legal studies (see for instance Baye and Wright, 2011) we consider the fact of appeal as

another indicator of legal quality. In general, companies decide to appeal the infringement decision of competition

12

authority, or, as in our case, of first instance court, only if probability of final reversal or at least of extension of judicial

review1 is high enough. The lower legal quality of decision is the higher probability to reverse it is and higher

incentives to appeal are. That is why the fact that one type of non-reversal decisions of commercial courts is appealed

more often than the other, controlling for other determinants to appeal, including toughness of penalties and remedies

for the company, indicates lower legal quality of decision.

As an indicator of judicial cost of the decision we use the variable of time (in months) spent to achieve decision

in the commercial court. We analyze determinants of the time spent to obtain decision in the first instance as well as

final decision.

We use several groups of control variables in the regression. First group are characteristics of administrative

decision, including monetary penalties (in logarithms) and conduct remedies. The higher toughness of penalties and

remedies is, the higher incentives to appeal are, the higher efforts to present counter-evidence at the stage of appealing

and higher probability of reversal and longer time spent to obtain decision we expect. Second group of control

variables reflects the characteristics of company, including size, experience of antitrust investigations and being in a

Register of the market participants with the market share exceeding 35%. Impact of the size of infringer on the decision

and cost (time) to obtain decision could be strong but uncertain in sign. On the one hand, larger company has more

resources to appeal, to collect and provide evidence. On the other hand, judges can be prejudiced against large

companies, and outcomes of judicial review can be biased respectively. The same can be true with the experience of

being under investigations: on the one hand, it can create learning by doing effect by reducing cost of further claims,

and simultaneously increases the incentives to avoid further investigations and infringement decisions, but at the same

time past infringement decisions serve a stigma and can make court be biased against those companies. Register of

the undertakings with large market share (>35%) is a specific list developed by Russian competition agencies from

1991. Legally being in a register does not mean any presumption of dominance and higher likelihood of infringement

of competition legislation, however judges may be prejudiced against the companies which are in the Register.

We measure size of company by binary variable that takes the value of 1 if company belongs to the list of largest

Russian companies developed by Expert RA (largest Russian rating agency) and 0 otherwise. This is very crude and

imperfect measure of size, but allows to reflect the specific position of the largest companies under antitrust

investigations and appealing the infringement decisions. Experience of antitrust investigations is measured by number

of litigations on the violation of antitrust legislation in which company has participated. Register variable takes the

value 1 if company is in the register of either Central office or regional subdivision of FAS and 0 otherwise.

Finally, we use the indicators of the experience for FAS subdivision and judge in the commercial court of the first

instance. Experience of FAS subdivision is measured by the number of litigations on the claims to appeal the

infringement decision up to the date of given decision. The number of court decision resolved by particular judge,

where FAS participated as one of the parties, indicates his/her experience. Impact of the expertise of parties under

litigation on the outcome provides important information on the positive standards of proof. We should expect that

legal quality of competition authority’s decisions increases with the experience of litigation. Impact of the experience

1 Motive of judicial review extension could be strong in the Russian competition enforcement, because administrative

decision on monetary penalty or remedy does not enter in the legal force since any instance of commercial courts take

‘final non-reversal decision’ that company does not appeal. However relatively short period of case consideration

makes extension less important for the companies. During the period under consideration, the longest litigation lasts

only 46 months in the first instance, and it is ten times longer than average period of litigation. In the international

perspectives, decisions in the Russian commercial courts are very quick.

13

of judge on his/her decision is much less predictable. According to Baye and Wright (2011) experience matters in US

courts. It should be even more so for the Russian commercial courts because of a relatively short period of antitrust

enforcement, and at the same time because of large number of claims for annulment of antitrust infringement decision

to commercial courts. It is sufficient to mention that the largest number of cases resolved by judge is 436 in our sample

that is more than 10 times higher in comparison with the data of Baye and Wright (2011) for the administrative law

and federal district court judges in the US.

The main source of information on the nature of violation, amount of penalties, application of remedies, duration

and outcomes of judicial review in several instances is LCAP database. Data on the size of company is from Expert

RA rating agency. Data on experience of judges and companies is collected from the database of decisions of

commercial courts (kad.arbitr.ru). Descriptive statistics of main variables and correlations is presented in Tables 2

and 3. Correlations between variables, which reflect characteristics of company (size, experience, listing in the register

of market participants) and characteristics of decisions (amount of monetary penalties and the fact of remedies

application) are high but too much. Correlations between the experience of competition authorities and particular

judges in commercial courts are also inevitably high: each of several judges in the commercial court accumulates the

experience of hearings of cases on antitrust violation together with regional competition authority that is defendant

under litigation. We apply probit regression to test the hypotheses H1 and H2 (Table 4, only marginal effects are

reported), OLS regression to test the hypotheses H3 and H4 (Tables 5 and 6).

All four hypotheses are confirmed (Tables 4-6). Infringement decisions where the practice in question is

restriction of competition are reversed more often at the first instance (Table 3, columns 1-3, only marginal effects are

reported). The fact that harm is imposed on the large group of customers or consumers instead of one customer

(consumer) helps companies to win under judicial review in the higher instances (columns 7-9). At the same time

competition issues make judicial review more costly in terms of time spent (Tables 5 and 6). At the first instance the

cases where infringement is about competition are about 1/3 longer than the cases which are concentrated on the harm

issues imposed by large companies.

Division of cases from proper antitrust subsample into subgroups brings one additional observation. Cases, which

take the longer time in the courts, are claims to annul the decisions on concerted practice and vertical agreements.

They substantially differ from the decisions on explicit horizontal agreements (collusions). For the latter the main

evidence is a direct one, and that is evidence on communications between market participants and conspiracy on

market sharing and/or price-fixing. To decide the case from the former group, in contrast, effect-based analysis (for

vertical agreements) and the analysis of indirect evidence on market structure, communication among sellers and

casual links between communication and market prices. Thus, the necessity to employ modern competition economics

makes decisions costly under judicial review.

The importance of this result and its influence on further development of competition enforcement in Russia in

the framework of motivation applied for the authorities cannot be overestimated. At first glance, it is expected that

cases on antitrust violations are difficult to resolve, and under judicial review many infringement decisions could be

annulled. Specifically, the success rate of European Commission in the Court of First Instance under article 81 (on

horizontal agreements) is not much higher than that of Russian competition authorities. Every 4th decision is annulled

(Neven, 2006). Comparable annulment rate in relatively young competition jurisdiction cannot attest national

competition authority as incompetent. However the problem arises in the framework of motivation system applied.

When competition authorities are motivated on legally qualitative and cheap decisions, and there is a subgroup of

14

cases where success can be achieved with lower cost, it is rational to concentrate the efforts exactly on the cases of

that type. Data of Table 1 shows that proper antitrust investigations and decisions are crowded out by non-proper

antitrust decisions. The results obtained predict that the first victim of the motivation would be decisions that need

more complex economic analysis. Trend in the number of investigations and decisions on concerted practice illustrates

it clearly (fig. 2). Afther the large number of cases (more than one hundreds cases are only claimed for review) in

2008-2011 recently FAS suspects tacit collusion in only few cases (less than dozen).

Figure 2. Number of claims in commercial court to annul infringement decisions on concerted practice (LCAP database), number

of FAS investigations on concerted practice (FAS annual report -2013, for 2012-2013) and number of FAS infringement decisions

on concerted practice (FAS annual report-2014, for 2012-2014)

After several dozens of decisions annually before 2012 in recent years competition authorities take only few

investigations on concerted practice.

The issue for economic policy is that a specific group of antitrust decisions that are concentrated around the issue

of harm imposed by dominant seller replaces antitrust under conventional understanding. It unexpectedly happens

under the impact of judicial review of competition authority’s decisions.

One interesting result of our analysis is that success rate of competition authorities under judicial review increases

over time because of increasing experience and changes in the structure of cases but not because of growing consensus

of judges and competition authorities on the standards of evidence. Higher experience of judge increases the

probability to annul the infringement decision under review.

VII. CONCLUDING REMARKS

The analysis of the Russian competition authorities’ decisions appealed in the commercial courts from 2008-2012

shows that the excessive scale of enforcement measured by the number of infringement decisions is explained by the

fact that these cases would never be opened under a correct understanding of the goals of antitrust enforcement.

The majority of cases are not proper antitrust ones. Infringement decisions rest purely on finding a harm to a

particular counterparty (as it is in cases on commercial conflicts), in contrast to the evidence of harm to a

representatively large group of counterparty, and even more to decrease of consumer welfare.

22

54

121

133

29

31

11

23

2 40

20

40

60

80

100

120

140

2008 2009 2010 2011 2012 2013 2014

Number of claims in commercial courts

(LCAP dataset)

Number of investigations on article 11.1

(FAS statistics)

Number of infrignement decisions on

article 11.1. (FAS statitics)

15

The statistics of enforcement and empirical analysis on the determinants of success under judicial review and cost

on the review allow reconstructing the incentives of competition agencies and partially those of judges. Procedural

rules regarding reactions to citizen complaints explains the high number of filing complaints. The orientation on

performance indicators such as the share of complaints obtaining appropriate response and the share of infringement

decisions taking effect (not challenged by the court) explains preferences for ‘easy to decide’ cases. The competition

agencies’ and the commercial courts’ understanding of harm explains the standards of proof applied. All of these

factors explain the large number of cases on abuse of dominance, with the importance of structural analysis and the

vague interpretation of harm.

Because of the motivation of competition officers, large-scale antitrust enforcement may coexist with difficult

competition restrictions and relevant harm to the consumer. In addition, enforcement may have a very low deterrence

effect that causes substantial harm to consumers and social welfare. In-depth analysis of cases investigated by

competition authority does not only confirm that any quantitative performance indicator may be problematic if

principal wants to achieve objective that cannot be reduced to the set of simple tasks. Another important take-off is

that it is dangerous to develop performance indicators based on the outcomes of judicial review of administrative

decisions. It might easily happen that for some accidental reasons decisions with a lower welfare impact have higher

legal quality. Under these circumstances, the increasing power of incentives may decrease deterrent effect of

enforcement.

REFERENCES

1. Avdasheva S., Kryuchkova P. (2014) The ‘reactive’ model of antitrust enforcement: when private interests dictate

enforcement actions – The Russian case. International Review of Law and Economics, in press.

2. Bandyopadhyay, S., & McCannon, B. C. (2014). The effect of the election of prosecutors on criminal trials. Public

Choice, 161(1-2), pp. 141-156.

3. Baye, M. R., Wright, J. D. (2011). Is antitrust too complicated for generalist judges? The impact of economic

complexity & judicial training on appeals. Journal of Law and Economics 54 (1), pp. 1-24

4. Brüggemeier G (2011) European civil liability law outside Europe. The example of the Big Three: China, Brazil,

Russia. Journal of European Tort Law 2(1): 1-22

5. Carree, M., Günster, A., Schinkel, M. (2010). European antitrust policy 1957-2004: An analysis of Commission

decisions. Review of Industrial Organization, 36 (2), pp. 97-131.

6. Connor, J. M. (2008). Forensic economics: an introduction with special emphasis on price fixing. Journal of

Competition Law and Economics, 4(1), pp. 31-59.

7. Dixit, A. (2002). Incentives and Organizations in the Public Sector: An Interpretative Review. Journal of Human

Resources, vol. 37, n. 4, pp. 696-727.

8. Farrell J. and Katz M.L. (2006), The Economics of Welfare Standards in Antitrust. Competition Policy Center,

Institute of Business and Economic Research, UC Berkeley, July 2006,

http://econpapers.repec.org/paper/cdlcompol/qt1tw2d426.htm

9. Farrell, J., Katz, M. (2006), The Economics of Welfare Standards in Antitrust, Competition Policy Center, Institute of

Business and Economic Research, UC Berkeley, http://econpapers.repec.org/paper/cdlcompol/qt1tw2d426.htm.

10. Gal, M. S. (2010). When the going gets tight: institutional solutions when antitrust enforcement resources are scarce.

Loyola University Chicago Law Journal 41 (3), 417-441.

11. Geradin, D. (2009). Perils of antitrust proliferation: The globalization of antitrust and the risks of overregulation of

competitive behavior, The. Chi. J. Int'l L., 10, 189-212.

12. Girgenson I. and Numerova A. (2012) A reform of Russian competition law: It’s a long way from Brussels to Moscow.

Journal of European Competition Law and Practice, 2012, 3 (3), 293-299.

13. Hendley K. (2014) (2014) Bargaining with strangers: explaining the aftermath of auto accidents. In: Everyday Law in

Russia (forthcoming)

14. Holmstrom, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership, and

job design. Journal of Law, Economics, & Organization, 24-52.

16

15. Larouche P., Schinkel MP (2013) Continental drift in the treatment of dominant firms: Article 102 TFEU in contrast

to § 2 Sherman Act. TILEC Discussion Paper 2013-20

16. Leaver, C. (2009). Bureaucratic minimal squawk behavior: Theory and evidence from regulatory agencies. The

American Economic Review, 99 (3), pp. 572-607.

17. Maggs PB, Schwartz O, Burnham W (2015) Law and legal system of the Russian Federation 6th Edition. Juris

Publishing Inc

18. Neven, D. J. (2006). Competition economics and antitrust in Europe. Economic Policy, 21 (48), pp. 742-791.

19. Reich N (1996) Consumer protection in countries of emerging markets: the example of Russia. Journal of Consumer

Policy 19(1): 1-43

20. Schinkel, M. P. (2008). Forensic economics in competition law enforcement. Journal of Competition Law and

Economics, 4(1), pp. 1-30.

21. Schinkel, M. P., Tóth, L., & Tuinstra, J. (2014). Discretionary Authority and Prioritizing in Government

Agencies. Amsterdam Center for Law & Economics Working Paper, (2014-06).

22. Trochev, A. (2012). Suing Russia at home. Problems of Post-Communism, 59 (5), 18-34.

23. Vickers J (2008) Abuse of market power. In: Buccirossi P (ed) Handbook of antitrust economics, Cambridge, MA:

USA, MIT Press, pp 415-432.

17

Appendix 1. ‘Non-proper antitrust decisions’ in the Russian competition enforcement: examples and structure)

The largest portion of cases under art. 10 involves alleged violations by natural monopolies2. The evidence

corresponds well to the data of the FAS; according to the annual reports ‘On Competition and Competition policy in

the Russian Federation’, cases against natural monopolies represent two-thirds of the activity of the FAS. This group

includes large in absolute, not in relative, terms a group of cases where the alleged violation is refusal to provide

interconnections for competitors on fair contract terms (especially in telecommunications) or access to networks for

competing suppliers (especially in electricity).

However, instead of access and/or interconnection issues for competitors, provisions of retail services for final

consumers represent the largest group of cases in both absolute and relative terms. Typical examples (decisions based

on inspection of compliance to service of natural monopolies – mainly regional utilities - to household and commercial

customers) follow3.

Case А35-6556/20124

In 2011, an antitrust investigation against company "Samaraenergo" (regional supplier of electricity) was initiated.

The investigation was based on the complaint of G, and the company was suspected of violating by non-use of the

discount coefficient of 0.7 in calculations of the price for electric energy consumed by each household. The FAS found

the company guilty of abusing a dominant position (part 1 of article 10 of the Law on competition) by imposing harm

on the consumer.

Case А32-5081/2012

In 2012, an antitrust investigation against "NESK-elektroseti" (operator of electricity network) was initiated. The

investigation was based on the complaint of T, A, H and P for the company’s delay in providing conditions and

specifications for the technological interconnection of power receivers that belonged to the group of households. In

the court decision, it is clearly mentioned that this delay does not comply with the “Rules of technological connections

of the power receiving devices of electricity consumers” that was approved by Government Resolution dated

December 27, 2004 No 861. The regional subdivision of the FAS found the company guilty of abusing a dominant

position in the form of infringement on personal interests.

Case № А76-8002/2012 and case № А76-3247/2012

Two antitrust investigations against "Gazprom Mezhregiongas Chelyabinsk" (a regional supplier of gas and a

Gazprom subsidiary) occurred in 2011-2012. The regional subdivision of the FAS considered it a violation of antitrust

law to use the take-or-pay principle that includes penalties for undertaken (case № А76-8002/2012) and overtaken

volumes of gas (case № А76-3247/2012) in long-term gas supply contracts with industrial customers. The antitrust

authority argued that these terms cause losses to gas consumers and can be classified as a violation of paragraph 10

of part 1 of article 10 of the Law on competition. Interestingly, during the period when the alleged violation occurred,

the law ‘On gas’ prescribed certain penalties for under- and overtake, and there was no information that the regional

supplier’s rates exceeded the tariffs established by the regulator.

The common features of the described example and all similar decisions of the FAS is that, first, all alleged

violators are dominant in the regional market of supply to residential and small industrial customers. Second, there is

no evidence of competition restriction, and all the hearings are concentrated on the harm imposed on the customer.

Third, there is no sign that a dominant position in the market creates possibilities to impose harm. Fourth, in the cases

where the final customers are involved, the harm in question occurs to a small number of them (in extremis, on only

one customer). Finally, in many cases, there is no evidence that the harm is intentional. Sometimes the alleged

violation may be a technical mistake in a contract term (А35-6556/2012), and sometimes it may be a sign of a low

quality of service (А32-5081/2012). Sometimes the alleged violation represents contract terms implied by the current

law to discipline consumers (А76-8002/2012, А76-3247/2012) in planning demand.

Consumers complain to the FAS for two reasons. First, in Russia, natural monopolies and their regional

subsidiaries have no specific responsibility for non-compliance with the regulated terms of a contract. Therefore,

consumers should choose between consumer protection and antitrust legislation to enforce the contract terms. In this

context, the advantage of antitrust legislation is opportunity to impose fairly large penalties5. High penalties are applied

2 The Russian law ‘On natural monopolies’ includes a list of activities with the relevant legal status. However, we also include in this group cases against the participants of the industries that are subject to direct price (or mark-up) regulations (for example, wholesale and retail trade of

pharmaceuticals). Here, we use the similarity of the alleged violation. 3 The names of the complainants in the following paragraphs are indicated by first letters. 4 This is the identification number of the case in the commercial court of the Russian Federation. This number indicates the commercial court of

the first instance, a two-digit number – subject of the Russian Federation (for example, 35 – Samara) – then, the number of claims to the commercial court and year. 5 More specifically, type of penalty imposed depend on the fact does violation involve restriction of competition or harm on the counterparty only.

Violator without legal status of natural monopoly if it does not restrict competition (but only imposes harm on counterparty instead) pays fixed penalties from 300 to 1000 thousand RUB (about 6 -20 thousand Euro). For restriction of competition and for violation by natural monopolies

18

rarely, but even a low probability makes compliance easy to enforce. Moreover, if the competition authority prescribes

contract terms in the form of a remedy, non-compliance with the remedies would almost certainly be fined.

A. Abuse of dominance: conflicts on interconnections with sub-subscribers

It is a common situation when one organization (A) connects to a network through a device located at the premises

owned by another organization (B). The parties must agree with one another on the terms of interconnection to the

network and their rights and responsibilities. An outstanding contractual relationship may lead to a conflict that results

in the restriction of network connections. This group of cases (interconnection with sub-subscribers) represents a

sufficient share of claims to annul infringement decisions (more than 10% of all the cases).

Here, we present several typical cases (decisions based on the investigations of complaints by sub-subscribers of

utility services) where one side of this conflict (A) appeals to the antitrust law to strengthen its position. The base

contains 400 cases of this type, including interconnection of electric power lines, heating networks, water supplies

and sanitation networks. These cases are considered by the FAS as antitrust cases involving a dominant position (in

the form of a restriction of access to a network, by charging excessive prices, etc.). The dominant position of the

accused organization (B) is usually reached by a narrow definition of the market. In these cases, it appears that the

essential facilities doctrine is applied to non-antitrust cases.

Case А53-1656/11

The business of M is supplied with electric energy by “Donelektrosbyt” (regional power supplier). The power

receiving devices of the entrepreneur are connected to the electric network through the transforming substation located

on the property of "Azovobuv" (shoe factory) and with the use of the local power network owned by this company.

One day, the company suddenly disconnected the entrepreneur without any prior notification. In this regard, M

initiated an antitrust investigation against "Azovobuv".

The competition authority concluded that the company "Azovobuv" was the only (and correspondingly, dominant)

supplier of electricity on the market within the boundaries of the area covered by its network. The FAS concluded that

"Azovobuv" violated part 1, article 10 of the Law on competition. The company was ordered to restore the connection

within five days from the receipt of the remedy and to resume the supply of electrical energy.

Case А13-3603/2011

This antitrust case was initiated against "Factory "Krasnij tkach" (textile factory). The facilities of the individual

entrepreneur S are connected to heating and electric supply systems through the local network owned by the Factory.

The company and S concluded a contract for the transit of heat and electricity. The company charged S the costs for

transit of thermal and electrical energy. However, the rates were different from the regulated tariffs established by the

authorized state body for the supply of energy by a local provider, and this difference led to conflict.

The antitrust authority defined the geographic markets in boundaries within the territory covered by the networks

owned by "Factory "Krasnij tkach" because rates of alternative suppliers were higher by more than 10% (this

difference explains why S continues to use the intermediation of the factory). The FAS concluded that the Factory had

a dominant position on the market and that it violated part 1 of article 10 of Law on competition. The FAS issued a

remedy and ordered "Factory "Krasnij tkach" to transfer the ‘income received from illegal monopolistic activity’ to

the state budget. Therefore, monetary sanctions were applied to the supply of a sub-subscriber by the rates, which

historically were lower than those in the region.

The common feature of this group of cases is that local networks were defined as relevant antitrust markets.

Automatically, the operator of a local network becomes ‘dominant’ with his own facilities. Then, the approach

described in the previous section is applied: any broadly defined harm is considered an abuse of dominance.

turnover penalties (up to 4% of turnover, generally around 1,5% of turnover for abuse of dominance) are applied. Even fixed and especially turnover penalties for violation of competition legislation exceed penalty threshold for the same action classified according another type of legislation.

19

Table 1. Complaints, investigations and claims for the annulment of competition authorities’ infringement decisions: 2008-2012 2008 2009 2010 2011 2012

Complaints submitted to competition authorities(1) 10704 16959 23046 27063 27347

Investigations opened by FAS(1)

6541 9664 11431 11276 10009

Investigations opened by competition authorities ex officio (% of the number of investigations) (1)

3551 (54.29)

5243 (54.25)

5832 (51.02)

4818 (42.73)

3794 (37.91)

Infringement decisions

Infringement decisions made by the FAS(1) 1045 1731 1979 2625 3216

- on abuse of dominance (art.10) 862 1438 1539 2310 3029

- on horizontal or vertical agreements, concerted practice (art.11)

183 293 440 315 187

Claims for the annulment of FAS decisions

Claims for the annulment of FAS decisions submitted in commercial courts of the first instance (% of the number of decisions in parentheses):

337 (32.25)

648 (37.74)

962 (48.51)

1187 (45.22)

796 (24.75)

Including

- not proper antitrust (% of the number of claims in parentheses):

259 481 720 1068 727

(76.85) (74.23) (74.84) (89.97) (91.33)

- proper antitrust (% of the number of claims in parentheses):

77 167 242 119 69

(23.15) (25.77) (25.16) (10.03) (8.67)

Decisions of the commercial courts

Infringement decisions annulled (completely or partially) in the courts of the first instance (%)

51.34 42.75 41.27 37.91 32.91

Appeals of the decisions of the courts of the first instance (%) 73.29 78.70 84.20 83.99 82.91

Decisions of the court of the first instance, reversed by the higher court, from all the appealed decisions (%)

43.72 39.80 20.12 19.66 17.42

Share of FAS decisions finally annulled (%) 72.61 65.54 53.09 50.12 44.60

Average time final decision takes (in months, mean, standard deviation in parentheses)

9,36 (7.05)

9,83 (7.40)

9,78 (6.80)

10,76 (6.85)

10,21 (6.54)

Source: LCAP database, data of the Federal Antitrust Service RF (1).

20

Table 2. Descriptive statistics of observations (type of infringement, infringement decisions, alleged infringers,

experience of competition authorities and judges in commercial courts)

Mean St. deviation Max Min

PAD (proper antitrust decisions) 0.17 - 1 0

Harm_Group (violation allegedly

imposes harm to large group of

counterparties or customers)

0.17 - 1 0

Penalties (monetary penalties, in

thousand RUB) 2 665 79 568 4 197 813 0

Conduct remedies 0.15 - 1 0

Size (largest companies from Expert

RA 400 list) 0.15 - 1 0

Recidivism (number of infringement

decisions before given claim) 4.13 7.90 73 0

Register (Market share > 35%) 0.23 - 1 0

Experience _FAS (number of cases

under judicial review) 53.03 54.23 277 0

Experience_ judge (number of

resolved decisions where FAS is one

of participants)

54.86 59.14 436 0

Months to take final decision (that is

not appealed by either party) 10.12 6.11 56 1

Months to take decision in the court

of the first instance 4.91 4.73 46 1

21

Table 3. Correlations between characteristics of violations, decisions, company and experience of competition

authorities and judge

Harm_Group Article_10 Collusion Concerted

practice

Vertical

agreement

PAD 0.39*** -0.60*** 0.38*** 0.14*** 0.47***

Harm_Group - -0.22*** 0.24*** 0.08*** 0.07***

Remedies Article_10 Collusion Concerted

practice

Vertical

agreement

Penalties -0.02 0.01 -0.00 -0.00 -0.00

Remedies - 0.17*** -0.03 -0.16*** -0.04***

Size Experience Register

Penalties 0.02 -0.01 -0.01

Remedies 0.01 0.01 0.08***

Experience 0.21*** - 0.05***

Register -0.01 0.05*** -

Experience of

judge

Experience of

FAS 0.41***

Table 4. Marginal effects from a probit model of the decisions on the claims to appeal infringement decision of

competition authorities (standards errors are reported in parentheses)

Dependent variables

REV_CFI APPEAL_CFI REV_CH

Characteristics of violation

1 2 3 4 5 6 7 8 9

PAD 0.27***

(0.08)

0,17**

(0,08)

0,20**

(0.09)

0,25**

(0.13) 0.14 (0.14) 0.10 (0.13)

0.19

(0.14)

0.06

(0.10)

0.13

(0.14)

Harm_Group 0.16*

(0.09) 0.10 (0.09)

0.07

(0.09)

-0.26**

(0.13)

-0.21

(0.14) -0.21 (0.14)

0.36***

(0,13)

0.24**

(0.09)

0.30**

(0,13)

Characteristics of decisions

Ln_Penalties 0.01 (0.02)

-0.01 (0.02)

0.00 (0.02) 0.02

(0.02)

Conduct

remedies

0.13**

(0.07)

-0.23**

(0.10)

-0.23**

(0.10)

0.19*

(0.10)

Characteristics of company

Ln_Size

0.32***

(0.09)

0.31***

(0.09)

-0.42***

(0.14)

-0.43***

(0.14)

0.23*

(0.13)

Ln_Experience -0.05 (0.03) -0.05 (0.04)

-0.27***

(0.05)

-0.24***

(0.05)

0.00

(0.00)

Register (Market share > 35%)

-0.12 (0.07) -0.13*

(0.08)

-0.19*

(0.10)

-0.23**

(0.11)

0.13

(0.11)

Experience of competition authority

Ln_Experience

_FAS

-0.27***

(0.03)

-0.27***

(0.03) -0.06 (0.05)

-0.18***

(0.04)

-0.25***

(0.05)

Experience of the judge in the court of first instance

Ln_Experience_

judge

0.12***

(0.04)

0.11***

(0.04) -0.08 (0.05)

0.02

(0.04)

0.00

(0.05)

Number of

observations 3681 3681 3681 2394 2394 2394 1973 1973 1973

Wald 𝜒2 21.48*** 105.48*** 109.87*** 5.62* 59.55** 68.72*** 15.23*** 48.23*** 57.91***

Pseudo R2 0.01 0.02 0.02 0.00 0.03 0.03 0.01 0.03 0.01

Notes: REV_1 =1 if infringement decision is annulled by the court of first instance, 0 otherwise.

APPEAL_NON REV = 1 if refusal to annul infringement decision by the court of first instance is appealed by the claimant, 0 otherwise.

REV_CH = 1 if infringement decision is annulled by one of higher court instances after refusal to annul by the court of first instance and subsequent appeal to higher instance, 0 otherwise

*coefficient significant at 10%, ** 5%, *** 1%.

22

Table 5. Dependence of time spent to achieve the final decision in the commercial courts: OLS results (t-

statistics are reported in parentheses)

Characteristics of violation

PAD 0.16***

(5.42)

0.17***

(5.53)

0.18***

(5.80)

Harm_Group 0.10***

(3.36)

0.14***

(4.86)

0.14***

(4.99)

0.09**

(3.15)

0.14***

(4.77)

0.07**

(2.35)

0.12***

(3.93)

Article_10 -0.08***

(-3.08) ref. ref ref

Collusion 0.06 (1.53) 0.07* (1.73) 0.07* (1.67)

Concerted

practice

0.14***

(3.83)

0.15***

(4.23)

0.19***

(5.12)

Vertical

agreements -0.09 (-1.26) -0.07 (-1.04) -0.07 (-0.97)

Characteristics of decisions

Ln_Penalties 0.01**

(2.70)

0.01**

(2.75)

0.01**

(2.55)

Conduct

remedies

0.09***

(4.20)

0.09***

(4.28)

0.09***

(4.12)

Characteristics of company

Ln_Size 0.18***

(5.90)

0.18***

(6.29)

Ln_Experience 0.02 (1.51) 0.02 (1.62)

Register

(Market share >

35%)

0.11***

(4.50)

0.11***

(4.49)

Experience of competition authority

Ln_Experience

_FAS

0.04***

(3.25)

0.03***

(3.00)

0.03***

(3.00)

0.04***

(3.23)

0.03***

(3.05)

0.04***

(3.22)

0.03***

(3.25)

Experience of the judge in the court of first instance

Ln_Experience_

judge

0.01

(1.04) 0.01 (0.01) 0.01 (0.01) 0.01 (0.01) 0.00 (0.01) 0.00 (0.01) 0.00 (0.01)

Constant 1.92***

(47.28)

2.02***

(43.78)

1.95***

(48.25)

1.90***

(45.52)

1.92***

(46.37)

1.84***

(43.83)

1.86***

(44.56)

Number of

observations 3890 3890 3890 3890 3890 3890 3890

F-statistics 19.48*** 14.25*** 10.98*** 16.11*** 10.78*** 18.28*** 14.39***

Adj R2 0.02 0.01 0.02 0.02 0.02 0.04 0.04

Notes:

*coefficient significant at 10%, ** 5%, *** 1%

23

Table 6. Dependence of time spent to achieve decision in the commercial court of the first instance: OLS

results (t-statistics are reported in parentheses)

Characteristics of violation

PAD 2.74***

(5.55)

2.63***

(5.29)

2.76***

(5.34)

Harm_Group 0.00 (0.00) 0.51 (1.00) 0.86* (1.66) -0.04 (-0.08) 0.81 (1.53) -0.12 (-0/21) 0.77 (1.45)

Collusion -2.21***

(-5.09) ref -0.04 (-0.08) ref ref

Explicit

horizontal

agreement

0.30 (0.43) 0.23 (0.33) 0.19 (0.26)

Concerted

practice

3.43***

(6.14)

3.32***

(5.84)

3.42***

(5.86)

Vertical

agreements

1.91**

(2.02) 1.67* (1.76) 1,60* (1.66)

Characteristics of decisions

Ln_Penalties 0.22**

(2.38)

0.22**

(2.29)

0.22**

(2.30)

0.21**

(2.20)

Conduct

remedies -0.49 (-1.11) -0.21 (-0.51) -0.40 (-0.97) -0.17 (-0.41)

Characteristics of company

Ln_Size 0.41 (0.80) 0.78 (1.23)

Ln_Experienc

e -0.12 (-0.53) -0.07 (-0.31)

Register

(Market share

> 35%)

0.67 (1.45) 0.63 (1.37)

Experience of competition authority

Ln_Experienc

e _FAS -0.05 (-0.21) -0.12 (-0.60) -0.06 (-0.30) -0.09 (-0.42) -0.10 (-0.48) -0.06 (-0.27) -0.06 (-0.27)

Experience of the judge in the court of first instance

Ln_Experienc

e_ judge 0.09 (0.41) 0.07 (0.34) -0.00 (0.01) 0.04 (0.19) -0.04 (-0.21) 0.04 (0.19) -0.07 (-0.30)

Constant 4.24***

(6.77)

6.60***

(9.64)

4.50***

(7.24)

4.56***

(7.08)

4.75***

(7.40)

4.32***

(6.52)

4.52***

(6.84)

Number of

observations 694 694 694 694 694 694 694

F-statistics 9.11*** 7.86*** 7.65*** 7.17*** 6.39*** 5.16*** 4.93***

Adj R2 0.05 0.04 0.05 0.06 0.06 0.06 0.06

Notes:

*coefficient significant at 10%, ** 5%, *** 1%