4.2  Applying the efficiency defence in cases where price changes

If a price increase is likely to result from the merger, or a dominant firm is likely able to raise prices above competitive levels, then, in the spirit of Williamson (1968), it is necessary to compare the magnitude of potential efficiencies to the allocative efficiency costs of the price increase (ie, the deadweight loss triangle in Figure 1). Price rises in domestic markets could be the result of mergers or practices relating to exporting and non-exporting firms. Indeed, in a small economy the domestic detriments that were otherwise quite significant may or may not be outweighed by export benefits of some mergers and practices. Figure 5 shows how an increase in producers’ surplus due to exports (resulting from increased efficiencies through merging for example) may outweigh a reduction in consumers’ surplus due to a price rise in the domestic market. If a merger increases efficiencies and thereby reduces average cost, producer surplus increases with area C and D. If this merger creates market power resulting in a price rise on the domestic market, area B is transferred from (domestic) consumers to producers and the little triangle next to it becomes deadweight loss (DWL). Clearly, the loss in consumers’ surplus is more than made up for by the increase in producers’ surplus. Even if area C and E would be considered a loss of ‘potential’ consumers’ surplus, the export effect may still outweigh the welfare loss of a higher domestic price (compare areas B + DWL and D).

Figure 5 – Export benefits may outweigh loss in domestic consumers’ surplus

Under US merger law, any price increase is sufficient to block a proposed merger. However, if one adopts a total welfare standard, it can be shown that there can be a misalignment between private incentives to enter an industry and the entry levels that maximize total economic welfare.

It is not always the case that more firms imply improved efficiency. In this case we have the following two results: First, Von Weizsacker (1980) shows that under the assumption of Cournot oligopoly behaviour, the free-entry outcome involves excessive entry. Perry (1984) examines this result when the conjectural variation term is allowed to vary. Brander and Spencer (1984) find similar results under the assumption that firms engage in tacit collusion in an environment of free entry. Simply put, it may be that free entry will achieve the optimal number of firms. Also, Winter (1993) shows that, under the assumption of differentiated Bertrand oligopoly behaviour, the free entry equilibrium involves an excessive number of differentiated product offerings (see also Spence 1976 and Dixit and Stiglitz 1977).

The problem identified in both these strands of the literature reflects the following two characteristics of the entry process. The last entrant reduces price and thus increases total welfare by an amount equal to the deadweight loss. That is, the last entrant generates a positive externality on the economy ie, there is a real resource cost of setting up an additional firm. This cost is manifested through an increase in all firms’ average total cost—each entrant reduces the quantity sold by each, and thus fixed costs must be averaged over a smaller amount of output. That is, the last entrant imposes a negative cost externality on existing competitors.

These results are reviewed in Mankiw and Whinston (1986) who identify the fundamental forces that give rise to each of them, recognising that the opposite result had been found in earlier literature (eg, Spence 1976 and Dixit and Stiglitz 1977). Amir (2001) provides further insights on this issue for a more general cost function. Amir points out that marginal cost changing with output may quantitatively strengthen the conclusion of other authors, and goes on to emphasise that merger analysis should explicitly take scale into account.

In sum, as long as one retains a total welfare standard, it remains that an effective efficiency defence is important even if prices may be increased above competitive levels. The analysis may be more complex in such cases because one must compare the deadweight loss triangle to the magnitude of efficiency gains. The excess-entry literature implies that the efficiency gains are larger in a fairly wide set of circumstances. However, whether excessive entry will occur in a small open economy, particularly one that is open to imports, is an open question the answer to which is likely to depend on the relevant good or service characteristics. It may well be less a potential issue in small relative to large domestic markets.

Sustainability arises when there is a zero-profit equilibrium.[43] Non-sustainable cases arise when demand intersects with a firm’s U-shaped average cost curve beyond the point of minimum average cost, but not far enough out to admit two or more homogeneous firms producing at minimum average cost. In this circumstance industry equilibria generally will not exist if strategies are purely in prices, certainly not one with zero economic profits, because when the incumbent prices at average cost—let alone marginal cost—it will be profitable for an entrant with the same cost curve to enter. No equilibrium exists because ex ante no firm would rationally be the first to enter because it knows it will be subject to cream skimming entry. The “no-equilibrium” result is sensitive to the simple strategies of the game: if long-term contracts with consumers are credible they can restrict cream skimming and thereby enable an equilibrium, or if there is competition in the quality dimension then an equilibrium may well exist. The sustainability issue therefore arises most often in the context of the delivery of a homogeneous good. Ultimately the possibility of sustainability is an empirical issue. But, because non-sustainability requires a single firm’s average cost curve to rise over the output range of the relevant market and yet the market does not support two firms with the same cost structure at minimum average cost, it is unlikely to be a major issue in the markets of very small economies. It is not an issue for tradable goods and services since aggregate demand in this case is very large, and where a single firm can supply (almost) all the market at minimum average cost it is likely that the average costs will not increase over the relevant range of market demand of a small economy.[44] From the point of view of competition law, the issue on non-sustainability is that there may not exist an equilibrium if zero (economic) profits are insisted upon; although we reiterate that the issue does not arise in situations where there is product differentiation and dynamic competition that entails long term contracting.

We return to examine competition law in small economies after consideration of dynamic efficiency in modern economies and linkages with trade policy.