Quantitative Techniques for Competition and Antitrust Analysis_11 ppt

6.2.4.2 Identification of Pricing and Demand Equations in Differentiated Markets In a fashion entirely analogous to the homogeneous products case, the identification of conduct generally

338 6 Identification of Conduct

econometric model, but it can also be helpful when collecting other evidence in agiven case (e.g., documentary evidence) On the other hand, such an observationmay concern us since we noted earlier that on occasion cartels have often resulted

in relatively less variation in prices, perhaps because of stability concerns As Corts(1999) noted, a different model of collusion would have different implications forobserved collusive prices

6.2.4.2 Identification of Pricing and Demand Equations in Differentiated Markets

In a fashion entirely analogous to the homogeneous products case, the identification

of conduct generally requires that the parameters of the demand and pricing tions are identified Even if demand rotation can also be used to identify conduct

equa-in differentiated equa-industries equa-in the same way as is done for homogeneous products,demand does need to be estimated to confirm or validate assumptions This presents

a challenge because a differentiated product industry has one demand curve and onepricing function for each of the products being sold In contrast, in the homogeneousproduct case, there is only one market demand and one market supply curve thatneed to be estimated Now we will need to estimate as many demand functions asthere are products and also as many pricing equations as there are products Iden-tification naturally becomes more difficult in this case and some restrictions willhave to be imposed in order to make the analysis tractable We discuss differentiatedproduct demand estimation extensively in chapter 9

A general principle for identification of any linear system of equations is that thenumber of parameter restrictions on each equation should be equal to, or greaterthan, the number of endogenous variables included in the equation A normalizationrestriction is always imposed in the specification of any equation so in practicethe number of additional restrictions must equal or be more than the number ofendogenous variables less one.49 This is equivalent to saying that the restrictionsmust be equal to or more than the number of endogenous variables on the “right-hand side” of any given equation The total number of endogenous variables is alsothe number of equations in the structural model This general principle is known asthe “order condition” and is a necessary condition for identification in systems oflinear equations It may, however, not be sufficient in some cases Previously, weencountered the basic supply-and-demand two-equation system, where we had twostructural equations with two endogenous variables: price and quantity In that case

we needed the normalization restrictions and then at least one parameter restrictionfor each equation for identification We obtained the parameter restrictions fromtheory: variables that shifted supply but not demand were needed in the equations

to identify the demand equation and vice versa (these exclusion restrictions areimposed by restricting values of the parameters to zero) A more technical discussion

49 The normalization restriction is usually imposed implicitly by not placing a parameter on whichever one of the endogenous variables is placed on the left-hand side of an equation.

6.2 Directly Identifying the Nature of Competition 339

6.2.4.3 Identification of Conduct: An Empirical Example

When conduct is unknown, we will want to assess the extent to which firms take intoaccount the consequences of pricing decisions on other products when they priceone particular good In this case, one strategy is to estimate the reduced form ofthe structural equations and retrieve the unknown structural parameters by using thecorrespondence between reduced-form and structural parameters derived from thegeneral structural specification Assuming that the demand parameters are identifiedand marginal costs are constant, we will need enough demand shifters excluded from

a pricing equation to be able to identify the conduct parameters (see Nevo 1998) Inparticular, we will need as many exogenous demand shifters in the demand equation

as there are products produced by the firm Although identification of conduct istherefore technically possible, in practice it may well be difficult to come up with asufficient number of exogenous demand and cost shifters

An early and important example of an attempt to identify empirically the nature

of competition in a differentiated product market is provided by Bresnahan’s (1987)study of the U.S car industry in the years 1953–57 Bresnahan considers the pricesand number of cars sold in the United States during those years and attempts toexplain why in 1955 prices dropped significantly and sales rose sharply In particular,

he tests whether this episode marks a temporary change of conduct by the firms from

a coordinated industry to a competitive one The data that Bresnahan (1987) is trying

to explain are presented in table 6.7 The important feature of the data to notice isthat it is apparent that 1955 was an atypical year with low prices and high quantities.Real prices fell by 5%, quantity increased by 38%, and revenues increased by 32%

To begin to build a model we must specify demand Bresnahan specifies demandfunctions where each product’s demand depends on the two neighboring products in

340 6 Identification of Conduct

terms of quality: the immediately lower-quality and the immediately higher-qualityproduct He motivates his demand equation using a particular underlying discretechoice model of demand but ultimately his demand function takes the form,

in quality, x In this rather restrictive demand model there is only a single parameter

to estimate, ı

To build the pricing equations, he assumes a cost function where marginal costsare constant in quantity produced but increasing in the quality of the products sothat xj >xi >xhfor products j , i , and h These assumptions imply that the wholestructure can be considered as a particular example of a model where demand islinear in price and marginal costs are constant in output By writing a linear-in-parameters demand equation, where qiD ˛i 0C ˛i ipiC ˛ijpjC ˛ihph, we can seethat for fixed values of the quality indices, xi, xj, and xh, the analysis of a pricinggame using Bresnahan’s demand model can be incorporated into the theoreticalstructure we developed above for the linear demand model where the parameters inthe equation are in fact functions of data and a single underlying parameter (Moreprecisely, we studied the linear demand model with two products above and we willstudy the general model in chapter 8.) Specifically, the linear demand parametersare of the form,

˛i i D
ı

1

xi
xh

:Bresnahan estimates the system of equations by assuming first that there is Nashcompetition so that the matrix
describes the actual ownership structure of products(i.e., there is no collusion) Subsequently, he estimates the same model for a cartel

by setting all the elements of the
matrix to 1 so that profits are maximized forthe entire industry He can then use a well-known model comparison test called theCox test to test the relative explanatory power of the two specifications.50Bresnahan

50 We have shown that the two models Bresnahan writes down are nested within a single family of models

so that we can follow standard testing approaches to distinguish between the models In Bresnahan’s case he chooses to use the Cox test, but in general economic models can be tested between formally irrespective of whether the models are nested or nonnested (see, for example, Vuong 1989).

rendition of figure 2 in Bresnahan (1987) (a) Under competition, products with close tutes produced by rivals get very low markups over MC (b) Under collusion, close substitutesproduced by rivals get much higher markups over MC.

substi-concludes that the cartel specification explains the years 1954 and 1956 while Nashcompetition model explains the data from 1955 best From this, he concludes that

1955 amounted to a temporary breakdown of coordination in the industry

Intuitively, Bresnahan is testing the extent to which close substitutes are straining each other If the firm maximizes profits of the two products jointly, therewill be less competitive pressure than in the case where the firm wants to maximizeprofits on one of the products only and therefore ignores the negative consequences

con-of lower prices on the sales con-of the close substitute product Thus, in figure 6.5, ifclose substitute products 2 and 3 are owned by rivals, then they will have a lowmarkup under competition but far higher markups under collusion

Given his assumptions about costs and the nature of demand, Bresnahan findsthat the explanation for the drop in price during 1955 is the increase in the level ofcompetition of close substitutes in the car market

The demand shifters that helped identify the parameter estimates are presented

in table 6.8 as well as the accounting profits of the industry The accounting profits,however, are not consistent with Bresnahan’s theory, as he notes If firms are coor-dinating in the years 1954 and 1956, industry profits should be higher than in 1955when they revert to competition Bresnahan’s response is that accounting profits arenot representative of economic profits and are not to be relied upon We must there-fore make a decision in this case about whether to believe the accounting measures

of profitability or the econometric analysis In other cases, one might hope each type

of evidence allows us to build toward a coherent single story

6.3 Conclusions

 Structural indicators such as market shares and concentration levels are stillcommonly used for a first assessment of industry conduct and performance,

342 6 Identification of Conduct

Per capita

effects- Developments in static economic theory and the availability of data haveshown that causality between market concentration and industry profitabilitycannot be easily inferred However, economic theories built on dynamic mod-els do frequently have a flavor of considerable commonality with the olderSCP literature For example, Sutton (1991, 1998) emphasizes that prices areindeed expected to be a function of market structure in two-stage games whereentry decisions are made at the first stage and then active firms compete insome way (on prices or quantities) or collude at a second stage

 The broad lesson of game theory is that quite detailed elements of the petitive environment can matter for a substantial competition analysis Thegeneral approach of undertaking a detailed market analysis aims at directlyidentifying the nature of competition on the ground and therefore the likelyeffects of any merger or alleged anticompetitive behavior

com- Technically, the question of identification involves asking the question ofwhether two models of behavior can be told apart from one another on thebasis of data The hard question in identification is to establish exactly whichdata variation will be helpful in moving us to a position where we are able totell apart some of our various models The academic analysis of identificationtends to take place within the context of econometric models, but the lessons ofsuch exercises typically move directly across to inform the kinds of evidencethat competition authorities should look for more generally such as evidencefrom company documents

6.4 Annex: Identification of Conduct in Differentiated Markets 343

 The degree to which firms are reactive to changes in demand conditions inthe market can provide direct evidence of the extent of a firm’s market power.Formal econometric models can use the methods involving the estimation

of conduct parameters in structural models to determine whether the tions of firms to changes in prices are consistent with competitive, competingoligopoly, or collusive settings However, the more general lesson is thatchanges in the demand elasticity can provide useful data variation to identifyconduct For example, we might (at least conceivably) find documentary evi-dence suggesting that firms’ pricing reactions accommodate prices in a fash-ion consistent with a firm’s internal estimates of market demand sensitivities(rather than firm demand sensitivities)

reac- We examined identification results for both homogeneous product markets andalso subsequently differentiated products markets Analysis of identification

in the former case suggests that demand rotators are the key to tion In the differentiated product case, the results suggest that (i) examiningthe markups of close-substitute but competing products may be useful and(ii) examining the intensity with which demand and cost shocks to neighboringproducts are accommodated may sometimes be helpful when understandingthe extent of coordination in a market

identifica- In examining the likelihood of collusion, one must assess whether the sary conditions for collusion exist Following Stigler (1964), those are agree-ment, monitoring, and enforcement The assessment of each of these con-ditions will typically involve a considerable amount of qualitative evidencealthough a considerable amount of quantitative evidence can be brought tobear to answer subquestions within each of the three conditions For exam-ple, the European Commission examined the extent to which transactionprices were predictable given list prices to examine market transparency inthe Sony–BMG case

neces- In addition to qualitative analysis of the factors which can affect the likelihood

of collusion, it is sometimes possible and certainly desirable to develop anunderstanding of the incentives to compete, collude, and also to defect fromcollusive environments

6.4 Annex: Identification of Conduct in Differentiated Markets

In this annex we follow Davis (2006d), who provides a technical discussion ofidentification of (i) pricing and demand equations in differentiated product marketsand (ii) firm conduct in such markets In particular, we specify in more detail ourexample of a market with two firms and two differentiated products Define the

10 0

0 0 2

# "

w1t

w2 t

#C

"

u1t

u2t

#:Similarly, suppose that demand shifters depend on some variables x such as income

or population size which affect the level of demand for each of the products:

"

˛01t

˛02t

#D

"

"1t

"2t

#:

Then linear demand functions for the two products can be written as

"

q1

q2

#D

"

˛01

˛02

#C

266

2664

2664

w1t

w2 t

x1t

x2 t

377

5D

2664

or, more compactly in matrix form,

AytC C xt D vt;where the vector of error terms is in fact a combination of the cost and demandshocks of the different products,

2664

5D

2664

2664

5:

Following our usual approach, this structural model can also be written as a form model:

reduced-y D
A1C x C v D ˘ x C v :

6.4 Annex: Identification of Conduct in Differentiated Markets 345The normalization restrictions are reflected in the fact that every equation has a 1for one of the endogenous variables This sets the scale of the parameters in thereduced form so that the solution is unique If we did not have any normalizationrestrictions, the parameter matrix ˘ could be equal to
A1C or equivalently (interms of observables) equal to
.2A/12C

In our structural system we have four equations and four endogenous variables.Our necessary condition for identification is therefore that we have at least threeparameter restrictions per equation besides the normalization restriction In gen-eral, in a system of demand and pricing equations with J products, we have 2Jendogenous variables This means that we will need least 2J
1 restrictions in eachequation besides the normalization restriction imposed by design

There are exclusion restrictions that are imposed on the parameters that come fromthe specification of the model First, we have exclusions in the matrix A which arederived from the first-order conditions Any row of matrix A will have 2J elements,where J is the total number of goods There will be an element for each price andone for each quantity of all goods But each pricing equation will have at most onequantity variable in, so that for every equation we get J
1 exclusion restrictionsimmediately from setting the coefficients on other good’s quantities to 0

Second, the ownership structure will provide exclusion restrictions for manymodels Specifically, in the pricing equations, there will only be Ji parameters

in the row, where Ji DPJ

j D1
ij is the total number of products owned by firm i(or, under the collusive model, the total number of products taken into account infirm i ’s profit-maximization decision) The implication is that we will have J
Jirestrictions

Third, in each of the demand equations in matrix A, we also have J
1 exclusionrestrictions as only one quantity enters each demand equation (together with all Jprices); the parameters for the other J
1 quantities can be set to 0

Fourth, we have exclusion restrictions in matrix C which come from the existence

of demand and cost shifters Demand shifters only affect prices through a change

in the quantities demanded and do not independently affect the pricing equation.Similarly, cost shifters play no direct role in determining a consumer’s demand for

a product; they would only affect quantity demanded through their effect on prices.Those cost and demand restrictions are represented by the zeros in the C matrix.Define kDas the total number of demand shifters and kCas the total number of costshifters For each of the pricing equations in C we have kDexclusion restrictionsbecause none of the demand shifters affect the pricing equation directly Similarly,for each of the demand equations we have kCexclusion restrictions since none ofthe cost shifters enter the demand equations

Additionally, even though any row in matrix C will have as many elements as thereare exogenous cost variables and demand shifters, there will only be as many newparameters in a pricing equation as there are cost shifters in that product’s pricing

In that case, since prices are set to maximize joint profits for the firm, their pricingequations will be interdependent for that reason Theory predicts that the way thedemand of product j affects product i ’s pricing equation is not independent of theway the demand of product i affects product j ’s pricing equation This gives rise topotential cross-equation restrictions For example, the matrix A we wrote down has

a total of sixteen elements but in fact it has only four structural parameters We couldimpose that the reduced-form parameters satisfy some of the underlying structural(theoretical) relations For instance, the first elements of rows 1 and 3 are the sameparameter with opposite signs This could be imposed when determining whetherthe structural parameters are in fact identified from estimates of the reduced-formparameters The more concentrated the ownership of the products in the market themore cross-equation restrictions we will have, but the fewer exclusion restrictions

we will have since we will have fewer zero elements of
In addition, we willneed more exclusion restrictions in each pricing equation to identify all the demandparameters that will be included

Damage Estimation

The estimation of damages has been one field within antitrust economics wherequantitative analysis has been used profusely Most of the work has been done incountries where courts set fines or award compensation payments that are based onthe estimated damages caused by infringing firms Effective deterrence using fines,

as distinct from, say, criminal conviction of individuals, requires that imposed fines

be at least as high as the expected additional profits of firms that would emanatefrom the behavior to be deterred Expected profits can be difficult to measure and incartel cases they are currently often approximated by the damages caused to affectedcustomers This chapter describes the issues investigators confront in estimating thedamages caused by the exercise of market power by cartels We also briefly discussdamage calculations from abuses by a single firm

7.1 Quantifying Damages of a Cartel

A presumption of antitrust law is that cartels are bad for consumers Both antitrustagencies and customers see that cartels increase prices and reduce the supply avail-able on the market For this reason, cartels are illegal in most jurisdictions Forexample, the Sherman Act in the United States, Article 81 in the EU and Chapter 1

of the Competition Act (1998) in the United Kingdom each prohibit firms fromcoordinating in order to reduce competition Nonetheless, because cartels that workcan be very profitable there is a temptation to collude when the conditions in themarket make it possible Illegality per se is not enough of a deterrent when it isnot accompanied by at least the potential for a punishment that will hopefully wipeout the expected benefits of participating in a cartel Cartels are increasingly pun-ished with substantial fines and in some jurisdictions including the United Statesand the United Kingdom some cartel behavior is a criminal offense.1For a fine to

1 Section 188 of the U.K Enterprise Act 2002 introduced a criminal offense for collusion in the United Kingdom It says, for example, that an individual is guilty of an offense if he “dishonestly agrees with one

or more other persons” to, in particular, directly or indirectly fix prices Note that the word “dishonestly” qualifies the word “agrees” so that not all agreements to fix prices are immediately dishonest and hence not all cartel offenses are criminal offenses The term dishonest is frequently used under other parts

of criminal law and so has clear legal status relating both to whether a person’s actions were honest

348 7 Damage Estimation

be an effective deterrent, its expected value should be linked to the expected gainsextracted by the cartel Private enforcement, which is common in the United Statesand which is developing in Europe, comes with compensation payments for theaffected customers.2 In the United States those payments are normally linked tothe damages suffered by these customers It becomes necessary in those cases toassess and quantify the impact of a cartel and to calculate the profit it generatedfor the firms and the harm it caused to customers downstream The next sectiondiscusses the effect of a cartel and the following section proceeds to explain thedifferent techniques used to quantify damages The pass-on defense is discussedand finally the issue of determining the duration of the cartel is presented in moredetail.3

According to the economic theory traditionally relied upon as an underlying rationale

to impose sanctions against cartel members, cartels have two effects on welfare: firstthey decrease the total welfare generated by the market and second they redistributerent from consumers to the firms The damages caused by a cartel are in principle thetotal welfare loss experienced by the customers due to the combination of those twofactors In fact, damages are in practice defined in a more restricted way and usuallyrefer to the overcharge that the customers must pay for their purchases, which isonly part of the loss suffered by consumers

7.1.1.1 Welfare Effects of a Cartel

When firms form a cartel, they coordinate to increase, perhaps even maximize, jointprofits If firms successfully maximize joint profits, then a cartel price can be approx-imated by that of a monopolist setting total production at the level where aggregatemarginal revenue equals cartel marginal cost Compared with a competitive marketwhere prices are set close to marginal costs, this reduces the quantity and raises theprice Because prices are higher in a cartel, firms are able to appropriate some of theconsumer surplus that would go to consumers in competitive markets In addition

according to the standards of most people but also whether the individuals believed such actions were honest The latter might be informed, for example, by evidence of, say, secretly held meetings or seeking

to hide collusive behavior so these may distinguish criminal from civil cartel behavior In the United States there have been criminal sanctions for cartel behavior since 1890 The United Kingdom’s first criminal sanctions were handed down in June 2008 in the “marine hose” cartel Marine hoses are a type

of flexible pipe used to transport oil from storage to tankers Three individuals received between two and three years each out of a maximum sentence of five years’ imprisonment In jurisdictions with both criminal and civil penalties, enforcement will generally proceed in parallel as criminal and civil sanctions are not a substitute for each other.

2 For example, the United Kingdom has some scope for limited private actions and the EU is currently consulting on the appropriate scale of private actions.

3 A nontechnical discussion of issues relevant to the estimation of damages can be found in Ashurst (2004).

7.1 Quantifying Damages of a Cartel 349

Price

DS

the decrease in the aggregate quantity produced causes total welfare to decrease andgenerates deadweight loss The consequences of a cartel on an otherwise competi-tive market are illustrated in figure 7.1 The area indicated by A represents the renttransfer from consumers to producers Consumers pay P1instead of P0and theypurchase only Q1 compared with a higher Q0under competition Area B repre-sents the net welfare loss, known as deadweight loss This is consumer welfare that

is eliminated due to the restriction in output and not captured by the cartel.The total welfare loss generated by the cartel is represented by area B The totaldamage to the consumer is represented by areas A C B The benefit of the cartel

to the firm is represented by A Although the total consumer loss is represented

by A C B, the loss of area B is generally ignored when calculating damages toconsumers Although in principle we would like to estimate both, damages aregenerally defined as the illegal appropriation of profits by the firms represented

by the area A For practical purposes we assume that the firm’s illicit profit andthe damages to consumers are equivalent and this amount is commonly called theovercharge The overcharge on a given unit is the difference between P1 and P0.The total overcharge is Q1.P1
P0/ Such an approximation will often not be toobad if the deadweight loss effects associated with area B are small relative to thesize of the transfer from consumers to firms associated with area A (But see thediscussion of Harberger triangles in chapter 1.)

7.1.1.2 Direct and Indirect Damages

Many cartels are among firms that provide inputs to firms downstream, which thensell on to final customers To understand the consequences of such a situation,consider the case of a downstream firm being the customer of the cartelized industry,

so that the cartel’s price is (or affects) the marginal cost of the downstream firms

 The second element describes the lost margin on units no longer sold underthe cartel Without the cartel we would have sold an extra q0
q1/ units andearned a margin p0
cComp/ on them This “output” effect is seldom takeninto account in damages calculations.

 The third element is the increase in profits earned by charging a higher stream price and captures the pass-through of the cost increase by the cartel

down-to downstream cusdown-tomers This is called the pass-on effect and it attenuates

the damage suffered by the downstream firm It is also called the indirecteffect on the final consumers because it measures the overcharge or damagessuffered by those final consumers rather than the actual customer of the cartel,which is the downstream firm The treatment of the indirect effect both in thecalculation of damages to the intermediate firms or in calculation of potentialdamages to the final consumer is determined by the legal framework.Formally, this downstream firm’s profits under the cartel can be expressed as follows:

1D p1
cCartel/q1;where the superscript “1” indicates prices, quantities, and profits of the down-stream firm under a cartel regime Under competition in the upstream market, thedownstream firm’s profits will be

0D p0
cComp/q0;where the superscript “0” indicates prices, quantities, and profits of the intermediatefirm under competition The difference between the two downstream profits is

0
1D p0
cComp/q0
p1
cCartel/q1:With some algebra manipulation we get an expression for the difference in profitsinvolving three terms corresponding to the bullet points above:

  0
1

D p0
cComp/q0
p1
cCartel/q1C q1.cComp
cComp/ C q1.p0
p0//

D
q1.cComp
cCartel/ C q0
q1/p0
q0
q1/cCompC q1.p0
p1/

7.1 Quantifying Damages of a Cartel 351

7.1.1.3 Empirical Issues

Calculating the damages of a cartel could be important to establish the appropriatelevel of compensation to give to the victims of the cartel or to estimate the illegalprofits of the cartelized industry, the gains from colluding, for the purpose of impos-ing an appropriate fine In either case, quantifying damages presents some importantconceptual and empirical challenges

To start with, one must define the concept being quantified In many cases, ages are defined to be the overcharge to the direct customer of the cartelized firms.That damage will be a lower bound to the true damage of the cartel at any point intime since the reduction in quantities and consequent deadweight loss is ignored.Second, damage calculations can become subject to some very complex issues if

dam-we take into account the potential dynamic effects Dynamic effects might increasedamages if competition would have had positive consequences for quality or inno-vation On the other hand, if high profits would have involved increased spending

on product quality or R&D, then, at least in principle, damages might be reducedalthough one may find it appropriate to consider the incentives to innovate in acartelized environment Due to the complexity of incorporating dynamic effectsand their usually speculative nature, such effects are generally ignored in damagecalculations although one obviously can debate the merits and disadvantages ofdoing so Generally, the policy stance in most jurisdictions reflects an expectationthat cartels will harm consumers in the longer term One should keep in mind thatsuch dynamic negative effects can occur and in those industries where they are likely

to be very important they should serve to aggravate the harm estimated to be caused

by the cartel

The treatment of the pass-on effect on the quantification of total damages or ofthe potential damage to claimants is generally defined by the legal framework Is thepass-on effect allowed to attenuate the potential damage claims of the intermediatefirm? Can final consumers claim damages? The answers to these questions helpdefine the appropriate theoretical framework in which the damage calculation takesplace and clearly these answers need to be understood by the economic analystbefore a quantification exercise is undertaken

The most important and difficult part of damage estimation is the actual tification of the overcharge Calculating the amount of the price increase due to thecartel requires the analyst to estimate what the price would have been in the event of

quan-a competitive mquan-arket upstrequan-am Severquan-al techniques quan-are quan-avquan-ailquan-able to construct whquan-at isreferred to as the “but for” prices—the prices that would have prevailed had the car-tel had not existed Unfortunately, the “but for” prices posit a counterfactual worldsince the world without the cartel simply did not happen Such a situation is notunfamiliar in the competition policy world—mergers must similarly be evaluatedbefore they have happened—but counterfactual situations always involve both esti-

Because damages may occur over an extended period of time, the calculation willhave to be translated into real terms so that the penalty is equivalent in value to thedamage inflicted Whether claimants are allowed to recover interest in the event of

a private claim is also a legal issue that needs to be clarified by the analyst.Each of the issues mentioned above will typically need to be addressed by theeconomist in a damage estimation exercise In the next section we discuss thequantification of the direct damage

Quantifying damages involves estimating the price that would have occurred absentthe cartel during the period of the cartel Clearly, the price we need is not and neverwill be observable so that the exercise will always rely on assumptions and a certaindegree of speculation Such is the nature of forecasting Different methods will rely

on different assumptions and it is important that the investigator is not only aware

of the assumptions but also explicitly states what they are The reasonableness ofparticular assumptions, and hence the best method, may well depend on the particularcircumstances of the case However, when a cartel clearly succeeded in raising prices,the effect of the cartel should be apparent using more than just one method as long

as those methods are correctly applied In practice, conscientious economic expertswill sometimes need to build an estimation framework that combines elements ofthe different methodologies Doing so will sometimes help to ensure that all theavailable data that are informative for the estimation of the “but for” prices are used

As with any econometric exercise, it will be important to test the robustness of theresult to small changes in specification and, as with any other kind of evidence, noeconometric exercise will be completely robust

7.1 Quantifying Damages of a Cartel 353The exercise of quantifying damages must be supported by an in-depth qualitativeanalysis of the industry, which should help provide the justification for the method-ology and specification chosen To carry weight, any econometric results will need

to be plausible given the known facts about the industry

7.1.2.1 Using a Model of Competition

Given an economic model relating pricing to industry structure, it will be possible toanalytically derive the effect of moving from competition to a cartel on prices Forexample, under perfect competition with no fixed costs the price will be equal or close

to marginal cost The overcharge of a cartel forming in that market would then bethe difference between the price observed during the cartel and the marginal cost ofthe industry The cartel price is observed and the competitive price can theoretically

be calculated if we have information on costs Note that if costs change during thecartel period, the prices that would have prevailed under competition during the time

of the cartel also change

To make these observations concrete, let us review our simplest pricing equationsunder conditions of competition and also under a cartel If we assume marginal costsare ctand the following linear inverse demand equation, ptD at
bQt, then profitmaximization by a cartel will involve setting marginal revenue equal to marginalcost:

MRt.Q/ D ct () at
2bQt D ct () QtD at
ct

2b :Substituting this cartel output choice into the demand function, we obtain the pricesunder a cartel:

ptD at
bQtD at
b

at
ct2b



D 12atC 12ct:Under perfect competition the price will be pt D ct and the equilibrium quantitywill be such that pt D at
bQt The overcharge per unit in this case will be thedifference between the prices and the marginal cost:

Overcharge per unit D pCartel
pComp

D 12atC 12ct
ct:

In many cases, oligopolistic competition such as Cournot may provide a morerealistic “but for” scenario instead of perfect competition Obviously, the “but for”prices for Cournot or for other oligopolistic models can each be analytically derivedand doing so provides the specification of the pricing equation However, the model

is further complicated by the fact that the quantity produced by both the colludingfirms and also the equilibrium price that would prevail absent the cartel will each

be sensitive to changes in demand since firms explicitly take into account demandconditions when setting their prices or quantities both under Cournot and under thecartel Prices in competitive oligopolistic markets may be less stable than underperfect competition, all else equal

uranium 308 The reader may wish to speculate when the period of the cartel was.

7.1.2.2 Before and After

The “before-and-after” methodology uses the historical time series of the prices ofthe cartelized goods as the main source of information It looks at the prices beforeand after the cartel and compares them with the prices that prevailed during thecartel The damages are then calculated as the difference between the cartel pricesand the prices under competition multiplied by the amount of sales during the cartel:

The before-and-after method just links with a straight line the price levels ring before and after the cartel In cases where there is an underlying trend in thedata one can take into account the trend to determine the hypothetical prices underperfect competition In the example of the uranium cartel presented, there seems to

occur-be a declining trend in the price of uranium 308 right occur-before the cartel (measured inconstant 2005 dollars) When competition is re-established, prices settle at a levelslightly lower in real terms than that which predates the cartel In this case, a simplebefore-and-after calculation of the damages in real terms could resemble the areaabove a line drawn between a competitive price of say $21 per pound in 1974 and

a competitive price of say $18 in 1989 It is important to note, however, that there

is a very important caveat to this calculation: namely that the cartel is alleged tohave lasted between 1972 and 1975 although the high prices clearly lasted for farlonger Thus an important question is whether those higher prices persisted becausecoordination arrangements had been settled during a period of explicit collusion and

The “before -and- after” methodology uses the historical time series of the prices ofthe cartelized goods as the main source of information It looks at the prices beforeand