Home > Economic Efficiency and the Merger of Superior Propane with ICG Propane

Economic Efficiency and the Merger of Superior Propane with ICG Propane

An Economic Justification for a Price Standard in Merger Policy:

The Merger of Superior Propane and ICG Propane

Richard O. Zerbe Jr.[1]

Sunny Knott


1.0. Introduction

The Federal Trade Commission (FTC) and the Department of Justice, the United States government offices responsible for antitrust regulation, apply a price standard to the evaluation of mergers. These agencies seek to prevent mergers that increase prices even if the efficiencies produced from the merger offset the loss of consumer surplus. A price increase as low as 1% might be opposed if the market and, therefore, the absolute consumer loss were large enough regardless of the efficiencies from the merger.[2] Such a policy violates the criteria of traditional welfare economics.

Professor Robert Lande and others (Lande 1982; Fisher, Lande, and Ross 2000) have argued powerfully for a price standard based on the legislative intent underlying the development of major antitrust laws. The legislators, Lande contends, were concerned with protecting consumers from harm caused by increases in monopoly power; namely monopolistic price increases. Lande’s “Wealth Transfer Thesis” outlines the history behind creation of the major antitrust regulations (1982). Much of the antitrust law was inspired by the significant increases in industrial concentration during the late 1800's and early 1900’s. The Sherman Act of 1890 marked the beginning of modern antitrust regulation and created a foundation for future legislation including the Federal Trade Commission Act, the Clayton Act, and the Celler-Kefauver Act. Legislation developed after the Sherman Act kept the basic theory of the original and “represented either extensions of the same ideas to different economic arenas, or attempts to better implement the same fundamental principles” (1982). One of the main reasons for enacting antitrust legislation, Lande argues, was to protect consumers from the effects of monopoly power and he provides many examples of Congress explicitly articulating the need for antitrust law in order to protect consumers and encourage economic efficiency. Congress believed that monopolies “unfairly extract(ed) wealth from consumers,” consumers “were entitled to purchase competitively priced products,” and some members believed unfair prices essentially equaled extortion (Lande 1982). “Congress was willing to risk some immediate efficiency losses in order to benefit consumers ultimately” and “artificially high prices were condemned not for causing allocative inefficiency but for “unfairly” transforming consumers’ wealth into monopoly profits” (Lande 1982). Congress thought equity and unfairness important, and efficiency, although also an important issue, was not the only motive for antitrust law. The goals of antitrust legislation include moral and ethical concerns about protecting consumers and small businesses from the redistributive effects of monopolies. Lande summarizes by saying “each antitrust law grew in part out of a desire to define and protect consumers’ property rights, an antipathy toward corporate aggregations of economic, social, and political power, and a concern for small entrepreneurs” (1982).

Others have offered arguments countering Lande’s theory and instead saying that equity is not the proper goal of antitrust institutions. For example, Bork (1978), Posner (1976), and Areeda and Turner (1980) argue that economic efficiency should be the goal of antitrust policy. Certainly there has been no economic justification provided for using a price standard for mergers.� While the issue has never been directly addressed by any case[3], the 1997 revisions to the Federal Trade Commission and the Department of Justice Horizontal Merger Guidelines appear to side with Lande.� The 1997 revisions greatly expanded the Efficiencies section of the Guidelines (Section 4) and now include extended treatment of how the agencies (called "the Agency") would evaluate a merger with both anticompetitive and procompetitive effects.� According to the new Guidelines, the central issue is whether such a merger would harm consumers in the relevant market.[4]� In other words, if a merger was likely, absent efficiencies, to harm consumers in the relevant market, the agencies would not challenge the merger if it would produce efficiencies that would forestall the harm.� The revisions even illustrate consumer harm with a price increase in the relevant market.� By neglecting to use any of the tools of traditional welfare analysis, including producers surplus, consumers surplus, or deadweight loss, the Guidelines appear to have fashioned a per se price standard.

2.0. The Merger

Superior Propane and ICG Propane provide retail sales and distribution services to residential and commercial propane customers. They have historically competed against each other in a number of overlapping geographic and product markets, and they are the only two companies that supply propane to end-users throughout Canada (Commissioner of Competition v. Superior 2000b). In 1997, Superior had 130 and ICG had 110 branches and satellite locations across Canada (ibid.). On December 7, 1998, Superior acquired all outstanding shares of ICG (the merger) but the implementation of the merger was delayed pending the Competition Tribunal's initial decision.[5]

Historically, Canadian antitrust policy has differed from U.S. antitrust policy.[6]� The merger would very likely be disallowed under the United States’ price standard because of the expected 8% average increase in the price for propane. But, the Canadians have used traditional economic welfare criteria in which a merger is allowed if the economic welfare gains exceed the economic welfare losses not including moral sentiments.� The recent Canadian propane merger between Superior Propane and ICG, however, challenges this position because a Canadian federal court required the Competition Tribunal to consider equity effects during the merger evaluation process.

The Commissioner of the Competition Bureau in Canada contested the merger between Superior Propane and ICG Propane before the Tribunal because the merger was expected to create significant monopoly power and price increases for propane. The Tribunal, persuaded by the large efficiencies created by the merger, ruled in favor of the companies, but the Commissioner appealed the decision to the Canadian Federal Court of Appeals on grounds that certain effects, in particular equity effects, were not given proper consideration as required by the Competition Act of 1986.� The Federal Court found for the Commissioner and remanded the matter back to the Tribunal, which later ruled in favor of the companies. In June of 2002 the Commissioner has again appealed. [7]

The case presents us with three interrelated challenges: (1) how does one make a decision when economic efficiency conflicts with equity; (2) is there an economic justification that would prevent the propane merger and that would support Professor Lande's finding of sentiments underlying U.S. antitrust law as well as the practice of the Federal Trade Commission not approving mergers that raise prices, and; (3) can these first two issues be reconciled through an expanded definition of economic efficiency?

3.0. Canadian Antitrust Institutions[8]

In Canada antitrust policy and enforcement reside with the Competition Bureau, which enforces the Competition Act.� The structure is shown in the following chart.

Chart 1: Canadian Antitrust Institutional Structure

The Competition Bureau resides under an umbrella ministry, Industry Canada, which works to improve the functioning of the Canadian economy through programs aimed at increasing technological innovation, and promoting investment, trade, and a competitive and fair economy (Industry Canada 2002). A cabinet level Minister heads Industry Canada and an appointed public official called the Commissioner heads the Bureau of Competition. Six Deputy Commissioners control different branches within the Bureau with one branch responsible for mergers. The Competition Bureau does not have independent power to prevent mergers and when seeking to enforce a decision about a merger the Commissioner must apply for a hearing before the quasi-judicial Competition Tribunal, which then rules on the matter. The purpose of the Tribunal is to allow for expedited hearings on antitrust matters. The Tribunal operates independently of any government agency and consists of up to four judicial members, and up to eight lay members. Cases are heard by assigned (usually three) Tribunal members; those reviewing the Superior case consisted of one judicial member and two laypersons (an economist and a business representative). Parties to antitrust cases may appeal Tribunal decisions to a Federal Appellate Court, from which decisions may be appealed to the Supreme Court of Canada.

The Bureau only reviews mergers that have significant competition issues and qualify as notifiable transactions. Notifiable transactions account for a small portion of total transactions. Companies that are involved in a merger that qualifies as a notifiable transaction must notify the Commissioner of their intentions to merge. In order for a merger to qualify as a notifiable transaction, the parties to the merger must have assets in Canada or gross revenues from sales exceeding $400 million (about $251 million U.S.) and generally the merger must increase one party's gross revenues from sales by over $35 million (Industry Canada 2000).

In 1989 the reviewed mergers accounted for approximately 10% of total mergers, but in 2000 they accounted for about 25% of all mergers (Goldman 1989; Crosbie & Co. Inc. 2000). The number of reviewed mergers is most likely increasing because of an increase in the size and/or holdings of the merging companies. During a three-year period (1986 through 1989) 39 out of 335 reviewed mergers required further attention but almost none proceeded to the Tribunal (Goldman 1989). The Bureau managed the 39 mergers in various ways, which included requiring monitoring (19 of the mergers), restructuring the merger to alleviate anticompetitive issues (9), voluntary abandonment by the firms (7), and Tribunal consideration as solely a final alternative (4) (ibid.).

When the Bureau reviews a merger it assesses the impact of the merger on future competition. This includes evaluating the size of the market share and market concentration, and evaluating impediments to entry. Additionally, the Bureau must review the merger in light of certain qualitative factors outlined in Section 93 of the Competition Act. Historically, however, even if the Bureau predicts that the merger will have a negative impact on competition the merger participants can offset this by showing that the merger will result in cost savings of greater magnitude than the loss (The Competition Act Section 92.1).

4.0. Canadian Propane Consumers [9]

������������� Superior and ICG Propane tend to serve businesses and individuals that reside in rural areas.� The businesses are more likely small and often use propane for essential purposes like cooking and heating. Residential customers are older on average, nearly a third are retired, and they generally have less education. Households use propane for a variety of purposes including heating, cooking, and running a hot water heater. A smaller fraction of customers use propane solely for unessential or less essential amenities such as heating a swimming pool or operating a fireplace. Residential consumers of propane also tend to earn lower levels of income with approximately 63% earning less than the Canadian median income (Canadian Council on Social Development 1996). The relatively high cost of switching to different fuels makes it difficult to change to new fuel sources. Substituting natural gas could be a cheaper option for those with access, but often this fuel is not available. Electricity often costs more per BTU than propane and switching to electricity would involve buying new appliances and some rewiring of the house (Commissioner of Competition v. Superior 1999c). Switching from propane to heating oil costs over $4,000 and the use of heating oil has other costs including bad odor and higher levels of indoor and outdoor environmental pollution (ibid.).

5.0. The Tribunal’s Findings and First Decision

The Tribunal found that the propane merger would substantially lessen competition in 89% (66 of 74) of the local markets for the supply of propane (Commissioner of Competition v. Superior 2001).� In 16 out of the 66 markets the merged firm would have a pure monopoly or near monopoly, with a market share ranging from 97% to 100%. In another 16 markets, where substantial market concentration already existed, the merger would remove healthy competition, and competition would be reduced in the remaining 34 markets (ibid.). The Tribunal found that the merger would lessen competition substantially in the coordination services offered to national account customers, leaving the merged entity as the only firm in Canada serving this market (ibid.).� National account customers have multiple business locations throughout the country, but use only one supplier for propane services.

������������� Despite these adverse effects on competition, the majority of the Tribunal found that the merger was saved from divestiture by application of the efficiencies defense found in the Competition Act.[10] The efficiencies defense rests on the cost savings from a merger.� Under Section 96, the Tribunal shall not prohibit a merger, even if it reduces competition, if it finds that the merger “ . . . is likely to bring about gains in efficiency that will be greater than, and will offset, the effects of any prevention or lessening of competition that will result or is likely to result from the merger . . .” (The Competition Act Section 96.1). The Tribunal’s interpretation of the clause “effects of any prevention or lessening of competition that will result” considers economic efficiency as the only “effect” of the merger.� The Tribunal defines economic efficiency in traditional terms, which they label the “Total Surplus Standard.”

6.0. Economic Efficiency

The Total Surplus Standard is simply the normal (Kaldor-Hicks or potential Pareto) criteria for welfare typically used in economic analysis, which is the sum of consumer and producer surpluses.� A positive sum, such as the one found in the propane merger, would create additional societal wealth.

The� Tribunal found that the proposed merger would raise prices, but would lower costs.� The merger would then be economically efficient if the efficiencies created by the merger were greater than the loss of consumer surplus so as to offset the negative effect of the merger’s price increase. The merger’s estimated effects are shown in Figure 1 below. We assume for simplicity that the long run supply curve is flat and is equal to marginal and average costs (a constant cost industry).� We assume that, before the merger, competition existed and monopoly profits equaled zero. The assumptions of a flat supply curve and of competition imply that producer surplus did not exist before the merger.� Price before the merger is determined by the intersection of demand and supply at P1. Gains in efficiency from the merger will reduce costs from AC1 to AC2 and after the merger the price will be determined by the monopolist and will be above costs at, say, P2.� The merger will create producer surplus both because of the reduction in costs and because of the gain in monopoly pricing power.� Since price is expected to increase, as shown in Figure 1, consumer surplus will fall.� The loss of consumer surplus will consist of the deadweight loss, which is lost to everyone, and the income transfer (welfare neutral transfer) which was initially consumer surplus and now becomes producer surplus. That is, both the deadweight loss (DWL) and the income transfer were consumer surplus before the merger.� After the merger, the producer surplus will be the efficiency gains plus the income transfer.� These will also represent economic profits.� Since the wealth transfer is a gain to producers and a loss to consumers, traditional economic analysis regards this as an offsetting transfer with no net welfare effect. The deadweight loss represents surplus lost to society because the monopolist can raise price above the supply curve (above average and marginal costs).� The economic welfare effects of the merger will then be found by comparing the deadweight loss with the efficiency gain and this is shown as follows:

W = (CS +PS)

Since CS = - (DWL + Wealth Transfer) and

PS = + (Efficiency Gain + Wealth Transfer), then

W = - (DWL + Wealth Transfer) + (Efficiency Gain + Wealth Transfer), or

W = - DWL + Efficiency Gain

Figure 1 Effects of Monopoly Power

6.1. The Economic Efficiency Standard Applied to the Superior Merger

The court estimated Superior Propane’s efficiency gains over a ten year period at $29.2 million per year, the deadweight loss at less than $3.0 million per year, and the wealth transfer from consumers to producers at approximately $40.5 million per year (Commissioner of Competition v. Superior 2000b).[11] The results of applying the Total Surplus Standard are shown in Table 1.

Table 1: The Total Surplus Standard

Change in Welfare

Millions of dollars per year

Deadweight Loss (-CS)


Wealth Transfer (-CS, +PS)


Efficiency Gain (+PS)




The deadweight loss and wealth transfer were calculated based on a method developed by Michael Ward, an expert witness for the Competition Bureau. The actual deadweight loss and transfer depend on the size of the price increase, the elasticity of demand, and the shape of the demand curve. Ward assumes a linear demand curve, looks at three elasticities, and calculates the DWL and transfer for various price increases based on his assumptions about the demand curve and the elasticity of propane (Commissioner of Competition v. Superior 1999b). According to Ward, the average expected price increase for the merger is 8% and the demand for propane in the Canadian market is likely inelastic.� The fact is, however, that from the beginning of 1999 (Superior acquired ICG in December of 1998) to the end of 2001 the real price has increased by 9.44% on average. The real price increases for previous ICG customers have been harsher with increases of over 24% during the same time period.[12]

The Tribunal concluded that economic efficiency as the sum of producer and consumer surplus was the correct measure and that the deadweight loss is the only anticompetitive effect that requires balancing against the efficiency gains in operation. Since the efficiency gains exceed the deadweight loss, the Tribunal deemed the merger efficient and beneficial to society. The wealth transfer from consumers to producers of $40.5 million per year for the Superior Propane case is irrelevant under the Tribunal’s interpretation of “effects.”

7.0. The Commissioner’s Appeal�������������

The Commissioner appealed the Tribunal’s decision to the Federal Court of Appeals. The Commissioner disagreed with the Tribunal’s interpretation of the Competition Act in three areas: the interpretation of the purpose clause (Section 1.1), the definition of “effects” under subsection 96.1, and the adoption of the Total Surplus Standard under subsection 96.1. Generally, the Commissioner’s appeal concerns the scope of the “efficiency defense” and in particular the meaning of "the effects of any... lessening of competition." The Commissioner seeks to expand the meaning of effects beyond the interpretation of the Tribunal.

The purpose clause requires that the Commissioner protect small and mid-sized companies from unequal competition and protect consumers from unfair prices and limited product choices in addition to ensuring economic efficiency.� The Commissioner interprets the purpose clause as an indication that the Tribunal should recognize other effects in addition to economic efficiency when considering a merger. The Commissioner argues that the purpose clause is inseparable from the rest of the law and any interpretation must coincide with the scheme of the statute and the intent of Parliament.

The Commissioner argues further that the effects are never clearly defined within the law and should be interpreted with respect to the Purpose Clause. Economic efficiency is not the only effect that prevents or decreases competition, and the commissioner argues that other effects require consideration. According to the Commissioner, the Tribunal in considering economic efficiency as the sole standard did not protect consumers and smaller businesses from the anticompetitive effects of the monopoly as required by the Purpose Clause. The wealth transfer from consumers to producers results in an increased concentration of wealth in business and decreases a consumer’s equal footing in the marketplace. The monopoly increases price, which limits consumers’ entitlement through law to competitive prices and diverse product choices.

The key issue raised by the Commissioner’s appeal is whether the legal effects of an anticompetitive merger are limited to the loss of resources to the economy as a whole, that is to traditional economic efficiency, or whether they include a wider range of effects, including effects on equity or other social values such as protecting consumers from monopolistic exploitation. The issue for economists is whether or not things people care about other than traditional goods should be included in efficiency. In short should moral sentiments be considered in determining efficiency?

8.0. Townley’s Economic Defense

The Commissioner’s expert economist, Dr. Peter G.C. Townley, says that in a situation where a merger causes the price of the good in question to rise, consumers lose and merging firms gain.� The distinction between winners and losers is clear, and one would wish to assess relative income levels and impacts explicitly. Townley states that before the Tribunal accepted the Total Surplus Standard it should have assessed the nature of the good in order for proper consideration of equity effects.

Townley points out that a given price increase applied to a good that is more inelastic in demand will create smaller deadweight losses but greater inequity as compared with a more elastic demand (Commissioner of Competition v. Superior 1999a).� This is because the wealth transfer will be larger with the more inelastic demand. Goods with inelastic demand are generally necessities or goods that lack close substitutes.� Townley notes "if the good in question is regarded as a necessity, its demand will be price inelastic and a price-increasing merger will lead to a relatively small deadweight loss.� However, it is precisely in this case that equity concerns are greatest" (ibid.).[13] He notes that to disregard these effects does not accord with society’s tax treatment of necessities. Governments “forgo an 'efficient' tax (and) presumably they do so because such a tax would be perceived as inequitable. That is, relatively poor households spend proportionately more on necessities than relatively wealthy households.� A tax on such goods would fall disproportionately on the poor, and thus be considered regressive” (ibid.). Propane, in the Canadian market context, exhibits qualities of a necessity and a good without close substitutes.

8.1. The Effects of Elasticity

To see the effects of elasticity consider Figure 2 where the demand curve D1 is more inelastic than D2. Suppose that initially the price is P1 determined by a flat supply curve at this price. A monopolist that maximizes profits would reduce production by one-half, given a horizontal cost curve, regardless of whether he faces the more elastic or more inelastic curve. The effects on price, however, differ greatly for the two demand curves. A monopolist faced with the more elastic curve would price at P2. If the monopolist faced instead the more inelastic curve, it would price at P3.� At these prices the sizes of the deadweight loss and the wealth transfer are proportional to the sizes of the price increase so that both the wealth transfer and deadweight loss are greater in the case of the more inelastic curve. A monopolist is unable to increase price significantly with an elastic demand curve, because quantity demanded decreases in a greater proportion to the price increase.� The monopolist prefers to face inelastic demand, as the resulting economic profits are greater. Townley’s statement that the deadweight loss will be less for a more inelastic demand is true only for a price increase of a given magnitude.� The price increase will be greater as will the deadweight loss and the wealth transfer for the more inelastic demand.

Economic efficiency traditionally suggests that a merger is more likely considered desirable when the companies face a more elastic demand curve and as a result produce less deadweight loss. Equity effects reach the same conclusion, as the wealth transfer effect is less for the more elastic demand curve, as long as one gives more weight to consumer than to producer surplus.

������������� ������������� ������������� Figure 2 Graphical Depiction of Elastic and Inelastic Demand

8.2. The Four Standards

Townley considers four standards for evaluating the effects of monopoly power: (1) the Total Surplus Standard (defined earlier as the efficiency gains minus the deadweight loss), (2) the Price Standard, (3) the Consumer Surplus Standard, and (4) the Balanced Weight Standard. The latter three are outlined below and a table follows which evaluates the propane merger based on each standard.

The merger would only pass the Price Standard if it improves or maintains the welfare of consumers.� A merger may accomplish this if the downward pressure on price through decreased marginal costs outweighs the upward pressure on price from the increase in market power.� Any merger that causes an increase in price will fail the Price Standard.

������������� The Consumer Surplus Standard compares the loss of consumer surplus with the efficiency gains to the producer. The standard assigns a weight of zero to those producer gains that arise from monopoly power.� Thus the wealth transfer from a price increase would count as a loss to consumers but not as a gain to producers. This method makes the value judgment that exploitation of monopoly power is bad, but resource savings are good.� If the loss of consumer surplus is greater than the efficiency gains, then the merger will fail the Consumer Surplus Standard. The Consumer Surplus Standard is measured by comparing the loss of consumer surplus (LCS) equal to the DWL plus the wealth transfer with the efficiency Gains (E)

if LCS > E then the merger fails the Consumer Surplus Standard

if LCS < E then the merger passes the Consumer Surplus Standard

������������� The final method Townley introduces involves measuring a weight (w) that just balances consumer losses and producer gains (the Balanced Weight Standard).� The decision-maker would then decide whether the balanced weight is higher or lower than what is desired. In order to solve the equation, the producer’s gains are assigned a weight of 1 and the total surplus equation is set equal to zero (see below). Solve for w and w gives the additional weight consumers would need to balance their losses with the producer's gains. The decision-maker(s) would make a value judgment concerning whether consumer dollars should be weighed w% more than producer dollars.

������������� In order to assess the effects of a merger on society, Townley recommends the following steps (Commissioner of Competition v. Superior 1999a):

    Subject the proposed merger to the Price Standard. If the merger satisfied this test, merging firms would be better off and consumers would at least be no worse off than before the merger. If a merger failed to satisfy the Price Standard, a second step would be required.Assess the merger with the Total Surplus Standard.� If the proposed merger failed this test the deadweight loss would outweigh the efficiency gains and there would be little reason to approve the merger.Use the Consumer Surplus Standard to evaluate the merger. The Consumer Surplus Standard gives a zero weight to any gains to the producer because of a price increase due to the increased monopoly power. Compare the efficiency gains with the loss of consumer surplus (wealth transfer + deadweight loss).Calculate a balanced weight that just balances gains and losses. One can assign a weight of 1 to producer’s gains and a weight of w to consumers' losses. Here, the value of w is the one that just causes the weighted change in total surplus to be equal to zero. This is the solution to:

Then policy makers must determine if the weight lies within an acceptable range.

8.3. Measuring the Superior Merger According to the Four Standards.

Table 2 below suggests that the merger should be rejected using two of the standards, the Price Standard and the Consumer Surplus Standard.� It would be accepted by the Total Surplus Standard as discussed previously.� The balanced weight approach rejects the merger if the weight for consumer losses is 60% or greater than that given to producer gains.

Table 2: The Standards Evaluated




Total Surplus Standard


The efficiency gains of $29.2 million outweigh the deadweight loss of $3.0 million so the merger is accepted by this standard.

Price Standard


Superior Propane is expected to raise prices by 8% on average because of their increase in monopoly power so the merger is rejected by the price standard.

Consumer Surplus Standard


Compare the loss of consumer surplus (the wealth transfer plus the deadweight loss) with the efficiency gains. The loss of consumer surplus is equal to the wealth transfer of $40.5 million plus the deadweight loss of $3 million for a total of $43.5 million. The merger would be rejected by the consumer surplus standard because the efficiency gains equal only $29.2 million.

Balanced Weights Standard

Depends whether decision-makers are willing to weight consumers losses 60% more than producers gains

The balanced weight equation is set equal to zero and reads as follows.

TS = 0 = (1)(40.5million) + (1)(29.2million) – ((w)(40.5million + 3million))

so w = 1.602

Decision-makers would then decide whether the weight attributed to consumers is greater than the weight attributed to producers by a factor greater than 60.2%.

9.0. The Federal Court's Decision�������������

The Federal Court of Appeal found that the majority of the Tribunal had erred by limiting the considered effects to changes in producer and consumer surplus, and by failing to ensure that all of the objectives of the Competition Act, and the particular circumstances of each merger, were given consideration in the balancing exercise mandated by the Act (Section 96.1).

The Federal Court of Appeal remanded to the Tribunal and required consideration of other anticompetitive effects, especially those included in the objectives of the Purpose Clause. The Court of Appeal expected the Tribunal to exercise its judgment when determining the exact effects to be considered.

������������� The Tribunal was required to use the Balanced Weights Standard (as advocated by Townley and the Commissioner) or a similar method, which incorporates the additional effects of the merger, instead of the efficiencies defense� (Commissioner of Competition v. Superior 2000c). Using this methodology, the Tribunal would determine the anticompetitive effects of the merger by taking into account a range of factors, but would not assign to each a fixed a priori weight.� The factors include: the deadweight loss, the wealth transfer from consumers to producers resulting from the increase in prices through the exercise of market power, the loss of product choices and services currently associated with the product, the impact of the merger on small to medium sized businesses, and the prevention of competition and the creation of a monopoly or near monopoly in some or all of the relevant markets (ibid.).

10.0. The Tribunal's Response

The Tribunal issued its response (Reasons and Order) to the Court on April 4, 2002.� The Tribunal left its approval of the merger unchanged. It cites a statement from the Appeal Judgment in a previous case, Air Canada, as showing its concept of the nature and scope of its response.

The Tribunal need only identify and assess, “the effects of the prevention or lessening of competition” for the purpose of section 96 and decide whether the efficiencies that the Tribunal has already found to have been proved by the respondents are likely to be greater than, and to offset, those effects (par 7).

The Tribunal members based their response to the Court remand on the arguments that they are 1) constrained by the language and history of the Competition Act to give primary importance to (traditional) economic efficiency,[14] and 2) any adverse "effects" including those on smaller businesses and on equity were both insufficiently clear and substantial to justify a conclusion contrary to economic efficiency. The Tribunal found that the legislative history clearly indicates that the merger provisions were not driven by consumer interest.� The Tribunal concludes that adopting an approach that prevents efficiency-enhancing mergers in all but rare circumstances must be wrong in law (par 83).

They differentiate the Canadian Competition Act and policy, which they find emphasizes economic efficiency, from United States merger law and policy, which they find emphasizes protection of small business and especially protection of the consumer from price increases due to market power (par 131).� The Tribunal notes that the makers of Canadian law and policy have historically understood that economic efficiency is particularly important in Canada as it allows businesses to achieve economies of scale that allow them to compete effectively with the United States and the international market.[15]

The Tribunal finds equity considerations insufficient to change its previous conclusion based on traditional economic efficiency.� They give the following reasons: efficiency is more important than equity as a matter of law; equity effects are too difficult to measure; their proper weight is uncertain in both law and economics; and, regardless, in this case, such effects appear sufficiently small and are unlikely to overturn the efficiency effects under any reasonable weighting scheme.

The Tribunal notes that they do not know how to compare equity with efficiency and the Federal Court did not offer any guidance.[16] They conclude, nevertheless, that the Commissioner’s expert economist, Peter Townley, who gives some weight to equity effects, suggests the best approach.� They find two difficulties with implementing his approach, however, the determination of the equity effects and the weight to use if they can be determined.� They reject Townley's suggestion that equal weight be given to all consumers, so that the entire transfer would receive some weight, on the ground that there is no reason for this in considerations of welfare economics or law (par 327 and par 371). The Tribunal's discussion appears to suggest that consumers would need to be sorted by income class, with different weights assigned according to these classes and then compared with the income class of the beneficiaries of the merger, primarily the stockholders of the merging companies.� The Tribunal finds this task virtually impossible to accomplish.[17]� That is they find the costs of compensation including all its relevant transactions costs to be too high.

Nevertheless, the Tribunal finds it necessary to attempt a consideration of equity as much as its present data will allow.[18] They note that among household purchasers (as opposed to business purchases most of which end up in households) 102,465 households fall into the lowest income quintile and consume bottled propane. The Tribunal finds that this group will pay about $2.6 million more under the merger and find that this constitutes the adverse redistributive effect, assuming they are using the propane for essential purposes (par 368).[19] The Tribunal then finds the maximum loss, including adverse equity effects, to be the sum of the maximum deadweight loss of $6 million (the mean being $3) plus the $2.6 million in adverse equity transfer for a total of approximately $8.6 million (par 370).� They go on to note that there is no statutory basis under the Act or in U.S antitrust law for such equal weighting;

perhaps the adverse redistributive effects should weigh twice as heavily as efficiency losses, in which case the . . . effects would not exceed $11.2 million. Alternatively, since efficiency concerns are paramount in merger review, perhaps adverse redistributive effects should be weighted half as much as deadweight losses.� In the instant case, it is clear that the adverse redistributive effects are, on the evidence, quite small.� Accordingly, the Tribunal is of the view that under any reasonable weighting scheme, the gains in efficiency of $29.2 million are greater than and offset all of the effects of lessening and prevention of competition attributable to the merger under review (par 371).

As shown above, it is possible to quarrel with these rough calculations. The Tribunal fails to give equity standing to those in the in the fourth income quintile class, does not give any standing to foreigners and gives primary standing to efficiency effects. For example, the Tribunal only considers people in the lowest income bracket among those negatively affected by the merger. This seems an arbitrary cutoff and people that make less than the median income may be the objects of altruism as well. According to a study completed by Statistics Canada, 11.1% of consumers or 241,991 individuals in the second lowest income bracket are “fueled by propane.” Given the same assumptions as the Tribunal this amounts to about $4.8 million, which would give a total equity loss of $13.4 million for consumers who may be legitimate objects of altruistic sentiments.

Our quarrel is however more fundamental.

11.0� Harberger’s Approach to the Trade Off of Equity and Efficiency

Economists have attempted to treat equity issues not withstanding their historical taboo.� Their approach has been to assign weights to the changes in income levels of different individuals. Weights have been suggested that depend on the preferences of administrators (Feldstein 1974, McGuire and Garn 1969), on the requirements of justice (Rawls 1971), and on the willingness to pay (WTP) of others (Harberger 1978).� Harberger's approach is grounded in an expanded version of traditional economic criteria based on traditional considerations of economic value.

Harberger’s basic notion is that one should not give more weight to losses or gains of any group than the costs of compensating or transferring money to them by the most efficient alternative method. So he would put an upper bound on the value of equity equal to the costs of redistribution. Harberger suggests that moral sentiments about transfers probably only arise with the appearance of gains or losses by the poor, and that the costs of transferring money, i.e. the waste, is likely to be no more than 20% of the amount of compensation made. Thus the limit to credit for favorable transfers to the poor or unfavorable transfers from the poor should be up to 20% of the amount transferred.[20] Harberger's argument is that a weight of more than 20% is inefficient as equity can be achieved more cheaply by using the efficient alternative method.

Applying the Harberger criterion to the propane merger does not justify preventing it. The deadweight loss is $3 million per year and the wealth transfer is $40 million per year for a total consumer loss of $43 million per year. Efficiency gains equal about $29 million. The administrative costs of achieving a transfer to poorer citizens by the most efficient available method would be 20% of $43 million or $8.6 million.� This is the sum that must be added to the loss of $43 million. The benefit-cost calculation is then the $43 million consumer loss plus the $8.6 million WTP for equity for a total of $51.6 million.� The benefits consist of the $40 million gain to the producers from the wealth transfer plus their efficiency gains of $29 million per year for a total of $69 million per year. Thus benefits exceed costs by $17.4 million per year. This can be summarized as follows.

Table 3: Accounting for Equity

Change in Welfare

Millions of dollars per year

Change in Consumer Surplus


Change in Producer Surplus


Transfer Compensation Cost




The merger produces efficiency gains sufficient so that it remains justified by this modification or extension of traditional economic criteria (Kaldor-Hicks). To this point we have found no economic justification in equity considerations to disagree with the Tribunal's decision nor have we discovered a basis in the consideration of this Canadian case to explain U.S. antitrust merger policy. The questions we now ask of Harberger’s approach are (1) does it include the relevant moral sentiments and (2) is 20% on the dollar the upper bound on the costs of the transfer?� We answer no to both questions and the implication is that the social loss from inequitable results can be very high.

To consider these questions we consider an expansion of Harberger's approach. Harberger's approach is in the spirit of traditional efficiency in that it rests on the WTP of individuals to purchase a good.� It is the good itself, equity, that is non-traditional.

12.0.� An Economic Justification for A Price Standard

Recently, an efficiency approach (KHZ) has been suggested (Zerbe 2001) that incorporates moral satisfaction including equity effects. KHZ expands on the Harberger criteria, inter alia, by exploring more generally the value of moral sentiments.[21] A relevant assumption of this approach is that all goods for which there is a WTP (such as moral sentiments) are economic goods.[22] The KHZ criterion is met when the sum of the WTP's for a change, including moral assumptions, exceed the sum of the absolute value of the WTA's.

By moral sentiments we mean those involving concern for the well being of objects or beings other than oneself. These sentiments include immoral sentiments as when one wishes harm to others.� One may care about others as a result of kinship, empathy, envy or hatred, or as a matter of justice. Charity is an expression of moral sentiment.� One may care about others from their perspective (one cares about their utility function) and this is called non-paternalistic altruism.� Or in the case of paternalistic altruism, one may care about others from one's own perspective as when a parent requires a child to eat spinach when the child would rather not.�

The implicit assumption in the Harberger analysis is that transfers to poorer citizens are fungible; it is as desirable to transfer money to one poorer person as to another equally poor. Moral sentiments may involve general considerations of income distribution, so that it is a matter of giving weight to a more equitable income distribution rather than the compensation of specific persons as a matter of justice. Then the Harberger approach seems appropriate (Zerbe 1998a and 1998b).�

This will not, however, in general be the case. The relevant moral satisfaction will depend on the specificity of the requirements for felt justice. When the matter of equity or justice concerns a specific population, the value of equity can only be considered in terms of the cost of the particular compensation required and the social loss from imposing the inequity or allowing it to continue. Where justice is a matter of compensating particular people, rather than general income distribution, the cost is not limited to the waste inherent in transfers since no such transfer mechanism will generally be in place. Instead the costs will include the cost of creating such a mechanism where it is not in place.

That is to say, when justice requires the compensation of particular people, the Harberger approach is no longer the correct one. The correct approach will allow for the possibility that the costs of compensation and the WTP for justice is greater than 20%, and in some cases much greater. Given there is no mechanism to compensate those threatened unjustly, the cost of transfers may be quite large so that the project imposes substantial costs on society. That is, the demand curve for moral satisfaction in a particular case may not be limited by the costs of redistribution in general.

12.1. Costs of Compensation

The costs of compensation will depend on (1) determining the injured, (2) determining the extent of their injury, and (3) carrying out the actual compensation.

12.1.1. Determining the Injured

Determination of the consumers injured by monopolistic price increases can prove difficult because these increases often harm consumers secondarily. Many industries use propane as an input in the production of their goods and consequently, a price increase for propane will affect the price of those goods that require propane as an input. A considerable proportion of propane is sold to companies for industrial type uses (89.3%) and a significantly smaller proportion (10.7%) is sold for residential end use applications (Commissioner of Competition v. Superior 2000a). So, the probability is high that consumers will be hurt through secondary price increases. To determine who these customers are (and the extent of their injury) would be costly so, as a prudent assumption, actual compensation is best limited to those directly harmed.

12.1.2. Determining the Injury

The direct injury (ignoring secondary effects) will be the difference between the price paid and the price (call it the base price) that would have existed were it not for the merger, times the amount purchased.[23] The most difficult determination will be that of the base price.� Since the injury will be recurring it is necessary to determine the base price over some time period. If compensation is based on the amount purchased, as is logical, this will itself affect the demand for propane and the monopoly price that could be charged, which is an undesirable effect of attempted compensation. An alternative method is price regulation with its apparent costs. Where the base price is easy to determine, such regulation would be less costly as would the costs of determining injury. We have no estimates of the costs of determining injury in a case such as this but clearly it could be large.

Price could be regulated directly based on the difference between a base price and the actual price.[24] Suppose there is a benchmark available to use for the competitive price such as the price in the United States or in other competitive market areas.� In this case the firms will be required to return an amount to each customer representing the difference between the new price and the base price times the quantity bought by the consumer. In this case the company has no incentive except to charge the competitive price.� In most cases, however, the base price may be a source of considerable controversy. The greater the difficulty in determining the base price, the closer the process approaches traditional public utility regulation which generally is costly.

12.1.3. Costs of Carrying out Compensation

Discussions with representatives of firms who contract to carry out compensation resulting, for example, from a consumer fraud case brought by the FTC, indicate that the direct costs of compensation vary greatly and can be high.� The costs of compensating consumers vary depending on the number of consumers, the costs of determining injury for each consumer, the availability of addresses for consumers, whether or not these addresses are available in electronic or paper form, and the accuracy of the addresses.� Where the accuracy of addresses is poor, costs can reach 90 cents on the dollar. In addition, it is not uncommon that around 20% of consumers fail to open or cash compensation checks usually because they don't realize a check is enclosed.[25]

The potential costs of determining the injured, the amount of injury and carrying out the compensation can be high. The high costs of actual compensation have implications for the value of inequitable results.

12.2.� The Conditions Under Which A Per Se Price Standard Would be Efficient

Suppose for a project there are users, consumers, and altruists who care about the consumers. A project will be KHZ efficient when the sum of net value for users and consumers plus the value arising from the moral sentiments of the altruists, minus costs to the altruists,� is greater than zero.� We assume for simplicity the project's costs to the altruists are zero.� The value for altruists will be their WTP for the project for its direct benefits to them and their WTP for its moral value to them. The moral value will be the smaller of the least cost of purchasing the same moral satisfaction elsewhere or the gross value to the altruists of the moral satisfaction from the project. The maximum moral satisfaction can never be greater than average moral satisfaction per altruist times the number of altruists times the net gain to the objects of altruism.

Suppose initially that all consumers are equal objects of altruism and that their net gain (which may be negative) from the merger is NGc.� Let  represents the aggregate level of altruism per unit of the altruistic good, usually called the warm glow parameter.� Thus  is the willingess to pay per dollar for moral satisfaction. (If� i is the average warm glow per altruist, then�  = Ni� where N is the number of altruists.) Then the maximum moral value cannot be greater than NGc.

The moral value may, however, be less or even zero even though�  > 0. For a particular equity good, say good i, there will be a WTP based on the availability of and price for substitutes and complementary goods.� If perfect substitutes are available at the same or lower price, the consumer surplus from good i will be zero or negative.� If the price of perfect substitutes is higher than the price of good i, or if the substitutes are imperfect, there will be a consumer surplus from purchase of good i. The purchase of the good will generate consumer surplus as long as its demand is above the price of substitutes,.[26] The consumer surplus from purchase of the good can not be greater than the cost of purchasing the same satisfaction by other means. Suppose the relevant alternative means of purchasing moral satisfaction is the direct transfer of cash.

In transferring cash to these consumers there will be a loss per dollar transferred due to the costs of determining the injured, the extent of injury and the costs of carrying out the actual transfer.� We will call this cost .� Assume initially that this parameter is less than $1.00. If, for example, 40 cents is lost from each dollar transferred, then  = .4. The amount of money that would need to be transferred to give the consumers one dollar will then be 1/(1- ), or $1.67. The moral value for some project, say project A, will be the smaller of the cost of purchasing equivalent moral satisfaction elsewhere and the moral value of the project itself. The amount that must be transferred X is found from X (1-) = NGc or X = NGc/(1-). Thus the moral value (MV) from purchase of good i will be the smaller of� NGc/(1-) - NGc� and the value determined by the warm glow parameter , so that moral value will equal the smaller of :

MV = NGc[/(1-)]������������� �������������  < 1.������������� ������������� ������������� (1)


MV = NGc ������������� ������������� ������������� ������������� ������������� ������������� (2)

������������� ������������� ������������� ������������� ������������� �������������

The value of  will be quite large, even if the average level of altruism is small, as long as there are many altruists. For example if there are 100,000 altruists and the warm glow parameter for each is 1 cent per dollar (i =.01),� the average  would be 1000. Thus in the case where compensation is not possible, modest altruistic values, a modest number of altruists and a modest number of people who are the objects of altruism are sufficient to render a project KHZ inefficient that would otherwise be efficient because moral sentiments are not considered.

Table 4 shows that for any  equal to or greater than 10 and any cost of transfers equal to or greater than 40%, the multiplier will be greater than 1.67. When this is applied to all consumer losses, the Tribunal's welfare conclusion is reversed.

We have regarded  as simply the loss per dollar transferred. We could more broadly and accurately call  the cost of obtaining equivalent moral satisfaction. In some cases the transfer may not be the cheapest or even a viable substitute. Just as apples and oranges are imperfect substitutes but are both foods, so there will be different categories of moral satisfaction. There are goods for which cash is a poor substitute. Some cheaper substitute than direct cash transfers may be available so that the relevant  may be less than the cost of money transfers. For example, the efficient remedy for an unjust insult may be an apology and not a cash transfer. On the other hand, direct transfers of cash may be a poor substitute and the costs of the cheapest substitute greater than the cost of transferring money. Thus even if the cost of money transfers is less than 1,  may be greater than 1. Where  is greater than 1, there is no possibility of gaining value from a substitute good and the moral value can be very large indeed.

The total value of the project will be the net gains to the users (i.e. the merging parties) and consumers plus the net existence value.� This will be the smaller of

W= NGc/(1-)������������� + NGu������������� ������������� ������������� ������������� ������������� ������������� (3)

W =NGc(1+) + NGu������������� ������������� ������������� ������������� ������������� ������������� (4)

From these equations we can determine the multiplier that is to be applied by the loss to consumers to obtain the social loss including altruism for selected values of  and . This is shown in Table 4 below.

Table 4: First Multiplier

alpha 






phi 









































































The total gain attached to the net gain of consumers will be the smaller of NGc(1-) and NGc(1+).� These expressions determine the multipliers calculated in Tables 4 and 5.� For example, when phi is zero, the costs of purchasing equivalent moral satisfaction is zero and equation 1 shows that MV = 0.� The multiplier is then 1 as shown in the row when phi is equal to 0.� This means that the results are not changed by the inclusion of moral values.� When phi is equal to say 0.9, the total value attached to consumer gains is 10(NGc)������������������������������������������������������������������������������������������������������������������������������ so that the multiplier is 10.� When phi is greater than 1, then the warm glow parameter  governs, as shown in the last row of Table 4.� Since  will be an aggregate figure over all altruists, it will be large for policies which affect the moral sentiments of many people.[27] Thus only the last two columns of Table 4, or those not shown but with even higher 's, are likely to be relevant.� Table 4 then shows that where the waste parameter is greater than 1, the welfare loss from a policy with a component broadly viewed as unjust, will be welfare (KHZ) inefficient.� When phi is greater than one there is no substitute available for the purchase of moral satisfaction.� When the project is widely viewed as unjust, alpha will be large so that the multiplier to be given to the net gain to consumers will be large, e.g. 1,000, as shown in Table 4.�

The Tribunal points out quite correctly that not all of the consumers are relatively poor and thus they will not all be the object of altruistic sentiments.� Nevertheless as long as the costs of compensation are prohibitive, it remains true that relatively low levels of sentiment and relatively low numbers of altruists are sufficient to conclude the merger should be prohibited.�� Let  be the fraction of consumers who are the object of altruism and let their loss be proportional to the fraction they represent.� Let M be the multiplier needed to reverse the Tribunal’s decision.� The gross consumer loss is estimated to be $43 million per year. The gross gain is $69 million per year. The welfare conclusion of the Tribunal will be reversed as long as $43M > $69 so that M > 1.6/.� Table 5 below shows the necessary values of M for different values of . Thus under reasonable and indeed quite conservative assumptions it is possible to obtain a multiplier which if applied to the consumer loss is sufficient to change the Tribunal's welfare conclusion.

Table 5: Second Multiplier














Table 5 shows that even if the portion of customers is small who are the objects of altruism, the multiplier is likely to render the project KHZ inefficient whose waste parameter is say 95% or greater and which engages the sympathies of several hundred people or more.� For example, suppose the net loss to that portion of consumers who engage moral sentiments is just one-tenth the total consumer loss of $43 million, or $4.3 million.� If phi is .95 the multiplier will be 20 assuming that the warm glow parameter is at least this large.� Thus, the total effect attached to consumer losses will be 20 times the loss of those consumers who are the object of moral sentiments.� This will be $86 million, enough to offset the efficiency gains.

13.0 Compensation

A project with compensation under KHZ will be different from one without (Zerbe 1998a, 1998b,2001, 2002a).� The extra value of a project with compensation is NGc while the cost of compensation will be NGc/(1-).� The difference will be

NGc[- -1]/(1-).

This will be positive as long as


If, for the propane case, we assume that = 40%, a merger with compensation would be preferable to one without for any  > .67.� Even if  = .9, a merger with compensation is superior to one without as long as  > 10. So for likely values of  and , a merger with compensation is preferable to one without the merger as long as compensation, even expensive compensation, is possible.�� Possibly the explanation for absence of compensation in many cases is that  >1.� If all consumers are objects of altruism, then for any value of phi equal to or greater than 40% and any alpha equal or greater than 1.67, the Tribunals conclusion is reversed.� If, as seems more likely, 60% of consumers are objects of altruism, then the Tribunals conclusion is reversed for any phi greater than 63% and an alpha greater than 2.7.� These are values that would seem to be easily met or exceeded.� The costs of compensation appear likely to be higher than 63 cents per dollar transferred and an alpha sufficiently large is found if there are a mere 270 people who are each willing to pay one cent to prevent the negative effect on consumers.

14.0 Objections to the Purchase of Moral Satisfaction

The conclusions in this paper rest on the willingness to consider moral satisfaction in welfare analysis.� As far as I am aware six arguments have been advanced against such conclusion. It is said that their inclusion results in� (1) the acceptance of projects that fail to pass a potential compensation test (PCT) [Winter 1969, Milgrom 1993], (2) the assignment of different weights to different individuals [Quiggin 1997], (3) the inclusion of purely redistributive policies [Quiggin 1997], (4) the inclusion of undesirable sectarian sentiments [Quiggin 1997], (5) that such inclusion results in double counting [McConnell 1997, Diamond and Hausman 1993, Diamond and Hausman 1994], and (6) such sentiments may be different to measure. I have considered these at length elsewhere (Zerbe, 2002a, 2002b) and find them either incorrect or inclusive.� I will thus only mention objects to these criticism briefly� here. First, KHZ does not require a potential compensation test.� Second,� including moral sentiments can result in a failure to pass this test only when  <1.� This will not be the case where issues of moral concern are� important. Third, KHZ does not assign different weights to different individuals as the value of a dollar is the same for all who receive.� The apparent weight arises instead from the fact that some people care about others. Fourth to label some sentiments as unworthy of consideration is not the job of the analyst and clearly involves a slippery slope.� The inclusion of moral sentiments does not result in double counting and arises from an assumption that the cost of direct transfers, , implicit assumption is zero, which, as we have seen is not likely and if it occurs results not in double counting because the existence value, i.e. the value to altruists is zero.�� Finally, moral sentiments may be difficult to measure in many cases but the analysis here shows that their consideration can nevertheless inform public policy.

15.0 Conclusion

Thus there is a case that can be made on economic welfare grounds to support the Commissioner's stance that the merger should not be allowed, or the stance that, if it is allowed, compensation should be required.� This case requires, contrary to the traditional welfare definition, that moral sentiments be considered an economic good.� This assumption is, however, consistent with KHZ, a recently developed expanded definition of traditional Kaldor-Hicks efficiency.� It is true that we have not proven the case for disallowing the merger.� We do not know whether or not, nor how soon, entry might occur reducing monopoly power.� We do not know the effects of a decision preventing the merger on the competitive balance between Canadian firms and other, primarily United States firms. In short this is not a full general equilibrium treatment.� Nevertheless we have shown that within the constraints of the data available a robust welfare economics analysis suggests the likelihood that the merger is undesirable.

In addition, it is clear that the analysis here provides even greater support for United States merger policy and for Professor Lande's advocacy of a price standard because the issue of economies of scale is less important in the U.S. than in Canada.� Moreover, Professor Lande's argument about the intent of the antitrust laws can itself have an economic efficiency justification. Only three assumptions are needed: (1) it is costly to compensate those harmed, (2) people care about equity so that inequity is costly, (3) the costs of mistakes from treating cases on a per se price basis, rather than on a case by case analysis of economic efficiency, minus the administrative savings under a per se rule, are less than the smaller of the first two categories of cost.[28]�� These assumptions do not appear unreasonable.


Areeda, Phillip, and Donald F. Turner (1980). Antitrust Law: An Analysis of Antitrust Principals and their Application. Boston: Little, Brown.

Bork, Robert (1978). The Antitrust Paradox: A Policy at War with Itself. New York: Basic Books.

Canada Market Research Ltd. (1997). “Defining Pricing Strategy in the Context of Customer Value Analysis (Commercial and Residential Markets): Research Report.”

Canadian Council on Social Development (1996). “Average and Median Total Family Income by Province, 1996.” Found at http://www.ccsd.ca/factsheets/fsavin96.htm

Commissioner of Competition, The v. Superior Propane Inc. et al. [1999a] “Affidavit of Peter G.C. Townley.” Competition Tribunal (C.T.).

Commissioner of Competition, The v. Superior Propane Inc. et al. [1999b] “Affidavit of Michael R. Ward.” Competition Tribunal (C.T.).

Commissioner of Competition, The v. Superior Propane Inc. et al. [1999c] “Expert Affidavit of R. Schwindt and S. Globerman.” Competition Tribunal (C.T).

Commissioner of Competition, The v. Superior Propane Inc. and ICG Propane Inc. [2000a]. Final Ruling. Competition Tribunal (C.T.).

Commissioner of Competition, The v. Superior Propane and ICG Propane Inc. [2000b]. “Memorandum of Fact and Law of the Appellant.” Federal Court of Appeal.

Commissioner of Competition, The v. Superior Propane Inc. and ICG Propane Inc. [2000c]. Final Ruling. Federal Court of Appeal.

Commissioner of Competition, The v. Superior Propane Inc. and ICG Propane Inc. [2001]. “Memorandum of the Commissioner of Competition on the Redetermination Proceedings.” Competition Tribunal.

Commissioner of Competition, The v. Superior Propane Inc. and ICG Propane Inc. [2002]. “Reasons and Order Following the Reasons for Judgment of the Federal Court of Appeal Dated April 4, 2001.” Competition Tribunal.

Competition Act, The, R.S., 1985, c. C-34, s. 1; R.S., 1985, c. 19 (2nd Supp.), s. 19.

______, ______. (2000). “Mergers and Acquisition in Canada.” Crosbie & Co. Inc. From a reference in, “A Comparison of Basic Merger Notification Requirements in Canada, the United States and the European Community,” authored by Michael A. Sullivan and Mario Brillant.

Diamond P. A. and J. Hausman, (1993). "On contingent Valuation Measurement of Nonuse Values" in Hausman, Jerry A. (ed.) Contingent valuation: A critical assessment. Contributions to Economic Analysis, Amsterdam; London and Tokyo; North-Holland; distributed in the U.S. and Canada by Elsevier Science, New York, 220: 417-35.

Director of Investigation and Research v. Air Canada et al. 51 C.P.R. (3rd) 131, 1993 C.C.T,D. No. 19.

Feldstein, Martin (1974). “Distributional Preferences in Public Expenditure Analysis.” Chaps. 6 and 14, in Redistribution Through Public Choices, Harold M. Hochman and George E. Peterson, eds. New York: Columbia University Press, pp. 136-161.

Fisher, Alan A., Robert H. Lande, and Stephen F. Ross (2000). “Counterpoint: Legalizing Merger to Monopoly and Higher Prices: The Canadian Competition Tribunal Gets it Wrong.” Antitrust 15(1):71-80.

F. T. C. v. Butterworth Health Corporation, 946 F.Supp. 1285 (D.Mich. 1996).

F.T.C. v. University Health Inc., 938 F.2d 1206 (11th Cir. 1991).

Goldman, Calvin S. (1989). The Competition Act: Our Track Record To Date, Notes for an Address. Found at http://strategis.ic.gc.ca/SSG/ct01508e.html. Calgary, March 29, 1989.

Harberger, Arnold (1978), “On the Use of Distributional Weights in Social Cost-Benefit Analysis,” Journal of Political Economy 86:87.

Industry Canada, Competition Bureau (2000). “Procedures Guide: Notifiable Transactions and Advance Ruling Certificates under the Competition Act.” Found at http://strategis.ic.gc.ca/pics/ct/ct01499e.pdf.

Industry Canada (2002). “The Department.” Found at http://www.ic.gc.ca/cmb/Welcomeic.nsf/ICPages/Department.

Lande, Robert H. (1982), “Wealth Transfers as the Original and Primary Concern of Antitrust: The Efficiency Interpretation Challenged.” Hastings Law Journal 34: pages 873-887.

McConnell, K. E. (1997).� "Does Altruism Undermine Existence Value? Journal of Environmental Economics and Management 32: 22-37.

McGuire, M.C., and H.A. Garn (1969). “The Integration of Equity and Efficiency Criteria in Public Project Selection.” The Economic Journal, 79(316), pages 882-893.

Posner, Robert (1976). Antitrust Law: An Economic Perspective. Chicago: University of Chicago Press.

Rawls, John (1971). A Theory of Justice. Cambridge: Belknap Press.

Schwartz, Lawrence, September 1992. " Comment and Analysis: The 'Price Standard' or the 'Efficiency Standard':? Eomment on the Hillsdown Decisioin" Canadian Competition Policy Record,� pp 42-47.

U.S. Department of Justice and Federal Trade Commission. 1992 Horizontal Merger Guidelines. Available from: http://www.ftc.gov/bc/docs/horizmer.htm (visited October 1, 2001).

Zerbe, Richard O. Jr. (1998a). “An Integration of Equity and Efficiency.” Washington Law Review. 73: 349-361.

Zerbe, Richard O. Jr. (1998b). “Is Benefit Cost Analysis Legal? Three Rules”, Journal of Policy Analysis and Management. 17 (3): 419-456.

Zerbe, Richard O. Jr. (2001). Economic Efficiency in Law and Economics. Northampton: Edward Elgar.

Zerbe, Richard O. Jr. (2002a). “Can Law and Economics Stand the Purchase of Moral Satisfaction?” Research in Law and Economics 20.

Zerbe, Richard O. Jr. (2002b). “The Purchase of Moral Satisfaction.” Working Paper.


[1]  This work was supported in part by the Center for the Study and Improvement of Regulation at Carnegie Mellon University and the University of Washington and in part by the Earthquake Engineering Research Centers Program of the National Science Foundation under Award Number EEC-9701568.�� This article builds on Zerbe (2001).

We would like to thank Keith Anderson, Ronald Bond, Aaron Finkle, Richard Higgins, John B. Kirkwood, Robert Lande, Steven Lewis, John Peterman, Peter Townley, Toby Savell, and members of the faculty colloquim at the University of Washington School of Law for useful information or comments.

[2] We thank Richard Higgins for clarification on FTC and Department of Justice policy regarding price increases.

[3]University Health, however, states that significant economies from an acquisition cannot be recognized unless they ultimately would benefit competition and, hence, consumers. See F.T.C. v. University Health Inc., 938 F.2d 1206, 1223 (11th Cir. 1991).� This is cited with approval by F. T. C. v. Butterworth Health Corp., 946 F. Supp. 1285 (D.Mich. 1996).

[4] The relevant language states:� "The agency will not a challenge a merger if cognizable efficiencies are of a character and magnitude such that the merger is not likely to be anticompetitive in any relevant market.� To make the requisite determination, the Agency considers whether cognizable efficiencies likely would be sufficient to reverse the merger's potential to harm consumers in the relevant market, e.g. by preventing price increases in that market."� FTC and Department of Justice Horizontal Merger Guidelines, Section 4, 1997 (footnote omitted).

[5] Currently, Superior and ICG have nearly completed implementation of the merger. They have integrated their corporate offices and market operations completely, and will complete the optimization of customer delivery routes by the end of 2002. See Superior’s 2001 annual report at http://www.superiorpropane.com/html/income.html.

[6] For a discussion of the euqity- efficency tradeoff in the Canadian context as compared with the American see

Schwartz (1992)

[7] The Commissioner appealed the Tribunal’s decision (June 2002) and has an initial hearing date set in October 2002 in the Federal Court (see http://www.fct-cf.gc.ca/business/hearings/appeal_e.shtml).

[8] Helpful and informative conversations with Douglas Tentland, Competition Law Officer at the Competition Bureau, improved this section considerably.

[9] All consumer data comes from a 1997 survey completed by Canada Market Research Ltd. for Superior Propane entitled, “Defining Pricing Strategy in the Context of Customer Value Analysis (Commercial and Residential Markets): Research Report.”

[10] One of the Tribunal members offered a dissent. Christine Lloyd, a lay member of the Tribunal, proposed a different approach to evaluating the merger, which called for an evaluation of the quantitative and qualitative losses due to the merger. She pointed to the purpose clause in the Competition Act as exemplifying the importance placed on other effects. The Purpose Clause reads “the purpose of this Act is to maintain and encourage competition in Canada in order to promote the efficiency and adaptability of the Canadian economy, in order to expand opportunities for Canadian participation in world markets while at the same time recognizing the role of foreign competition in Canada, in order to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy and in order to provide consumers with competitive prices and product choices” (The Competition Act Section 1.1).� Lloyd’s approach to evaluating the merger included additional qualitative effects like the reduction of consumer choice because of the loss of a competitor, the complete absence of choice in 16 markets, the company’s ability to exercise market power over national account customers, and the wealth transfer (Commissioner of Competition v. Superior 2000a).

[11] The loss of consumer surplus could be less than the stated amount because of improvements to delivery services by the merged propane company, or increases in quality and improved innovation. In stockholder literature, Superior claimed that services would improve following the merger, but did not use this argument in court and did not give specific examples of improvement (see http://www.superiorpropane.com/html/income.html). We thank Aaron Finkle and Leigh Anderson for this observation.

[12] Price increases adjusted for inflation by obtaining monthly inflation rates from a Canadian government website called Canadian Economy Online http://canadianeconomy.gc.ca/english/economy/#bp.

[13] Figure 2 below shows that this is not strictly true. Under the assumptions used there, more inelastic demand will result in a greater price increase and a greater DWL.� Where costs fall, I am unable to sign the relative change in the wealth transfer as between more and less elastic demand curves.

[14] The Tribunal concludes: “It is clear to the Tribunal that the Parliamentary Committee endorsed the view that efficiency was the paramount objective of the merger provisions of the Act” (par 80). The Tribunal believes that when section 96 is placed in the context of history of the Competition Act they are to focus on competition and resource allocation not considering much redistributional effects. "It was a clear concern of Bill C-256 that redistributional effects of anti-competitive mergers saved by efficiency gains not harm consumers beyond a reasonable time period. This concern was successively de-emphasized in subsequent bills"� (par 49).

[15] The Tribunal notes that the primary reason to amend the Combines Investigation Act in 1986 was "the need to strengthen Canadian business and provide an incentive for productivity in the face of aggressive international competition to which the government was committed and which would ultimately benefit consumers" (par 81).

[16] “ . . . (I)t is not clear how adverse redistributive effects, which are not losses of real savings, could be netted against real savings (of resources)."� Since the Tribunal interprets 'savings of resources' as economic efficiency, the Tribunal is stating that they do no know how to balance efficiency with equity (par 57).

The Court instructed the Tribunal to consider redistributive effects but it did not prescribe the method by which the Tribunal would perform its task. “The Tribunal must follow this instruction in light of the clear legislative history that indicated that the merger provisions were not driven by consumer interest. The Tribunal concludes that adopting an approach that prevents efficiency-enhancing mergers in all but rare circumstances must be wrong in law” (par 83). (emphasis added).

[17] They note that “ . . .the redistributive effects are generally difficult to identify correctly, and will involve multiple social decisions.� Given the informational requirements of such assessments, the assumption of neutrality could be appropriate in many circumstances” (par 329).

[18]One reason the equity effects are small is that they fall in part on foreigners (which the Tribunal claims it can not consider as a matter of law).� In so far as the effects on foreigners are not considered then the Tribunal's approach departs from the traditional theoretical welfare standard. In traditional economic welfare theory all effects and persons are to be counted in the welfare analysis. In practice this is necessarily not the case both because full measurement is never possible and for reasons of policy, which may also be traced ultimately back to measurement issues. The question of whose values are to count has become known as the issue of economic standing. Zerbe (1998, 2001, 2002) has considered these issues elsewhere.

[19] They use a demand elasticity of -1.5.

[20] Zerbe (1998a, 1998b) has elaborated on the Harberger criterion.

[21] KHZ (Kaldor-Hicks-Zerbe) builds on Kaldor-Hicks to expand an approach to equity suggested by Harberger.� KHZ assumes simply that (1) all values count, or more precisely all goods and sentiments for which there is either a WTP or a WTA are economic goods; (2) gains and losses are to be measured by the WTP and WTA respectively and from a psychological reference point that is determined largely by legal rights, and (3) transactions costs of operating within a set of rules are included in determining efficiency.� The rationale for these assumptions, along with some other ones, may be found in Zerbe (2001).

[22] Zerbe argues for a modification of the traditional approach that includes moral sentiments (Zerbe 1998a, 1998b, 2001, 2002a, 2002b). In Zerbe 2002a and 2002b particularly he considers the arguments that have been made against such inclusion and shows that they are either incorrect or not compelling.

[23] In addition there is the welfare loss triangle which we ignore here.

[24]A court has found that nonprofit hospitals will not likely exploit monopoly power.� See F. T. C. v. Butterworth Health Corp., 946 F. Supp. 1285 (D.Mich. 1996).

[25] In arriving at these conclusions I spoke with Toby Savell of the FTC, a representative of the Garden City Group, Tom Litala, representatives of Analytics Inc., Nancy Johnson and Richard Simmons, and representatives of Gilardi & Co., Ron Heard and Tom Sinden.� We also talked with Richard Higgins, John Peterman, and Ronald Bond of the Law and Economics Consulting Group.

[26] This will occur only if the other sources of moral satisfaction are imperfect substitutes or if the cost of achieving it in the particular case is below the price elsewhere.

[27] These sentiments could, however, be offsetting, in which case  may be small.

[28] Elsewhere Zerbe suggests other requirements (Zerbe 2001, 2002a, 2002b).� One equally important assumption for a viable expanded definition of economic efficiency is to count the transaction costs of carrying out a project as part of the economic costs.

Recent Documents:

Set Home | Add to Favorites

All Rights Reserved Powered by Free Document Search and Download

Copyright © 2011
This site does not host pdf,doc,ppt,xls,rtf,txt files all document are the property of their respective owners. complaint#downhi.com