An Index Approach to Measuring Product Differentiation: A Hedonic Analysis of Airfares

THE airline industry has always captured the imagination, but never more so than during the current deregulatory period. To understand the sources of economic rents that are the driving forces for change in this industry, we employ an interdisciplinary analysis drawn from the economics and strategy literature. This analysis allows us to identify sources of rent that are geographically based (local monopoly sources) and others that stem from economies of scale-based advantages in providing nationwide service (oligopoly sources). We examine the economic and strategic characteristics of these diverse sources of rent and suggest some policy concerns about the emerging airline market structure. An interdisciplinary approach is taken because economic models normally assume that firms have identical factor costs and similar production opportunities. Yet much of our ability to frame the forces for change in the airline industry begins with the recognition that firms came into the deregulatory period with differing factor costs and differing locational endowments and, hence, differing production opportunities. Building on Porter'sl concept of strategic groups, we offer evidence that air carriers

Since Vives (1984), Cheng (1985) and Okuguchi (1987), the equilibria in Bertrand and Cournot oligopolies with product differentiation have been known to differ. Okuguchi (2005) has shown that Bertrand priceadjusting oligopoly with product differentiation and symmetric firms is quasi-competitive but not perfectly competitive in the limit state of infinite number of firms. This paper formulates and analyzes two types of Cournot output-adjusting oligopoly with product differentiation and symmetric firms, in one of which symmetric firms producing the same goods are assumed to collude, and in the other, collusion is ruled out. The limit states are shown to be different in two oligopolies but they are both quasi-competitive. It is well known, since the seminal paper of Ruffin (1971), that the Cournot oligopoly without product differentiation and symmetric firms is quasi-competitive in the sense that the equilibrium industry output increases with an increase in the number of firms, and that the equilibrium product price converges to the competitive one under a set of reasonable conditions on the inverse market demand function and the firm’s cost function. Okuguchi (1973) has extended Ruffin’s result taking into account asymmetric firms and adapting the method of the proof of Debreu and Scarf (1963) of the equivalence of the core and the perfectly competitive equilibrium in the limit state of a pure exchange economy. Two kinds of equilibria are conceivable for Cournot oligopoly without product differentiation and with symmetric firms with identical cost functions. The first equilibrium emerges when all firms behave with a priori knowledge that their equilibrium outputs are the same. The second equilibrium occurs when all symmetric firms do not know in advance that their equilibrium outputs are identical. This is the case of non-collusion and was first systematically analyzed by Ruffin (1971). In section 2 we will analyze the differences between the two equilibria in Cournot oligopoly without product differentiation and with symmetric firms. The difference of the equilibria in Cournot oligopoly without product differentiation and with symmetric firms and with or without collusion was first suggested by Kemp and Shimomura (1995). We will find that the Cournot oligopoly equilibria are quasi-competitive but do not converge to a perfectly competitive equilibrium when the number of firms becomes infinite and when collusion exists. If oligopoly consists of firm producing differentiated products, the firm may choose price or output as a strategic variable. If its strategy is price, the oligopoly is called Bertrand oligopoly with product differentiation, and if it is output, we have Cournot oligopoly with product differentiation. In Bertrand oligopoly with product differentiation, the firm chooses its profit-maximizing price assuming that its rivals’ prices are not affected by its own price choice. On the other

PurposeThe purpose of this paper is to review previous research and to propose a model for the impact of barriers to entry on the market strategy of an entrant firm, where product/market scope and product differentiation are central strategy components. The paper asks, what is the impact of barriers on market strategies of entrants? Are early and late entrants affected in different ways?Design/methodology/approachA model and propositions are developed‐based on a review of previous research. The model applies the contingency perspective and company cases exemplify the model.FindingsIt is proposed that a firm that enters a market late and faces extensive barriers would choose a broader product/market scope and differentiate its products to a larger extent than an early entrant. It is also proposed that incumbents' market strategies indirectly affect the market strategy of an entrant firm as incumbents' market strategies interact with barriers, and the effects are due to entry timing.Research limitations/implicationsThe study contributes theoretically as it extends current knowledge of the impact of barriers to entry on strategy. Management of entrant firms are advised to strive for a fit between barriers and market strategy and consider the propositions.Originality/valueThe model and the propositions concern barrier effects on two key components of the market strategy of an entrant firm: product/market scope and product differentiation. Another important value is that the model accounts for interactions between incumbent strategies and barriers to entry, and effects on the market strategy of an entrant firm.

In competition law the concept of “significant existing market power” is often considered as a prerequisite to examining whether a business conduct gives rise to liability and, traditionally, the value of the Lerner index (the proportional price–cost margin) is used to measure the size of market power. In this paper we discuss in detail the role of the size of existing market power as a predictor of the size of the reduction in welfare generated by anticompetitive actions/conducts. We concentrate on monopolization or abuse of dominance conducts in which an exclusionary action by the dominant firm eliminates one of the rival firms. The main point which emerges from our analysis is that the source of market power is very important in understanding how changes in the size of extant market power affect the size of the reduction in welfare. We consider vertical and horizontal product differentiation and market structure as alternative sources of market power. We show that in contrast to Kaplow (Goals of competition law, Edward Elgar, pp 3–26, 2012), a significant extant market power requirement can be justified if either a Total Welfare Standard (TWS) or a Consumer Surplus Standard (CSS) is used and that this will be the case if the market power is the result of horizontal product differentiation or the result of a smaller initial number of competing firms. Further, we show, again in contrast to Kaplow (Goals of competition law, Edward Elgar, pp 3–26, 2012), that such a requirement may not be justified under either a CSS or a TWS—as when the market power is the result of vertical product differentiation. We also examine how extant market power and market share vary with the degree of product differentiation and market structure. We show that market shares vary inversely with the Lerner index as horizontal differentiation increases and directly as vertical differentiation increases and as the number of firms decreases, thus proving the irrelevance in many cases of market share as a predictor of market power.

This Article defines the notion of market power and how in conjunction with trademark rights give rise to elements that are deemed anticompetitive in a free market society. This Article uses legal arguments to consider how important developments in antitrust economics, particularly product differentiation and monopolistic competition, have contributed to the notion that trademarks are a source of market power. The Article uses a number of cases in the field of trademarks to underscore the key points that trademarks are a source of market power. These case developments contribute to the monopolistic tendencies of trademarks and describe how such tendencies are associated with the theory on market power and product differentiation. Empirically, the Article examines beer products from a single large corporation and the various trademarks/brands to determine whether such brands are a source of market power, effectively giving that manufacturer a monopoly on the beer market. A discussion of product hopping in pharmaceuticals is used to supplement the theories and evidence from the beer market. The Article also develops a theory of branded monopoly and suggests that, as a result of single ownership of trademarks and brands that are abundant from a single owner trademark’s market power, questions relating to antitrust foreclosure are often raised, despite the fact that market power is not anticompetitive per se. If it is recognized that trademarks are a source of market power, and hence, a core concern for antitrust law and policy, then the legal foundations of the current trademark system would need a radical redesign. If, on the other hand, it is recognized that trademarks are a source of market power, but do not conflict with antitrust law, and antitrust enforcers are to ignore conducts such as market foreclosure and other barriers to entry as a result of excessive trademarks and brands, then both antitrust and trademark law can continue to co-exist in the current system.

Preferential trade agreements (PTAs) usually result in increased trade amongst member countries and lower prices within the PTA. Thus the markets are assumed to be “more competitive” because lower prices are taken to imply “decreased market power” due to a reduction in price-cost margins. In this paper we empirically examine the relationship between changes in market power and product differentiation within the context of a duopoly framework with restricted entry. Using industry level trade data from the EEC we show that the formation of a PTA may increase the market power of PTA-exporters and lower the market power of non-PTA exporters. In addition, we show that these market power effects are more pronounced for less differentiated products.

Estimating the residual demand curve facing a single firm

A Nash-Cournot model, based on generating units' marginal cost function, is set up to study oligopolistic generators' market power and pricing strategy in the unilateral open electricity market. Factors that affect generators' market power are analyzed. We find that in unilateral open electricity markets, oligopolistic generators always tend to raise its bidding price by making full use of its market power, which raise the overall electricity price in the whole market. In order to reduce the generators' market power, the whole market total installation capacities must be increased and there must be no one generator whose capacity is too much relative to others. Further, measures must be taken to effectively monitor the operation of the market and to motivate the generators to undertake cost-cutting strategies through technical innovation. Reasonable competitive bidding system must also be set up to prevent collusion.

Competition and Entry in Small Airline Markets

This chapter discusses the economics of competition law, with particular emphasis on the role of market and market power. It begins with an analysis of competition from an economic perspective, taking into account the meaning of competition, the importance of the concept of market power, and competition in relation to the nature of strategic interaction. It then considers the conceptual foundations of welfare economics, the factors of production and commodification, the distinction between firms and consumers, monopoly vs. perfect competition, and the social costs of market power. It also reviews the history and many dimensions of market definition, the concept and function of market definition in EU and UK competition law, substitutability and the Hypothetical Monopolist Test, the ‘Cellophane Fallacy’ and ‘Reverse-Cellophane Fallacy’, aftermarkets, and brands and product differentiation. Finally, the chapter provides an overview of barriers to entry and/or expansion and the future of market definition.

The Collins et al. (2011) study (hereafter, CRC) examines how the competitive strategies of airlines impact the persistence of their operating earnings. The study then investigates if the persistence of earnings components, such as unit prices, costs, and productivity changes, also varies by competitive strategy. The study uses 590 observations of quarterly financial performance from 14 airline companies, over the period 1996 to 2008, to examine the impact of different competitive strategies on the persistence of profit margins and asset turnover. Two airlines' competitive strategies are identified: network carriers and low-cost providers. The paper documents that network airlines (differentiators) have more persistent operating profit margins than low-cost carriers. Little difference is noted for total asset turnover across business models. The study also demonstrates that partitioning the change in profit margin into growth, price-recovery, and productivity components more fully explains future profit margins than current profit margin alone.The paper is a nice blend of the literatures in industrial economics and both financial and management accounting. There are two strings in the paper that are interesting to management and management accountants. One is how a firm's competitive strategy impacts the persistence of profit margins and asset turnover ratios, which are the two key components of return on assets using the DuPont formula. The other interesting issue is the examination of how volume, price, and productivity components of profit margin changes impact earnings persistence differently across competitive strategies. As this latter part of the paper uses traditional variance analysis and productivity measures, it is particularly interesting to management accountants. Below, I discuss these two inter-related issues.One of the primary contributions of the paper is linking a firm's competitive strategy to earnings persistence in an industry setting. Earnings persistence is defined as the degree of first-order auto-correlation in a time-series of earnings numbers. Prior studies in economics and accounting have documented a number of factors that are related to a firm's earnings persistence. Early on, Stigler (1963) related the degree of an industry's concentration to the stability of accounting returns, and found that the industry pattern of returns was extremely stable through time in concentrated industries. In fact, the persistence differences in returns between firms in highly concentrated industries versus those in industries with low concentration became more noticeable two or three years out. Gort and Singamsetti (1976) confirmed Stigler's (1963) finding by examining return on assets metrics for more than 500 manufacturers. They found that although firms in concentrated industries did not have either higher levels of returns, or more variable returns, than firms in industries with little concentration, their accounting returns were more persistent than those in less-concentrated industries. Lustgarten and Thomadakis (1980), early on, documented that stock prices responded more strongly to the earnings of firms in concentrated industries, suggesting that the information contained in earning changes is positively related to persistence.Lev (1983) was one of the first in the accounting literature to examine factors driving annual earnings persistence. Using annual data on 385 firms from the period 1947–1973, Lev (1983) examined the link between four economic variables and the auto-correlation in a firm's annual earnings, sales, and return on equity. Lev (1983) used barriers to entry in place of industry concentration as a determinant of persistence, and found that firms in industries with high barriers to entry exhibited more persistent earnings than firms with low entry barriers. He also documented that firms in industries selling durable goods had less persistent earnings than firms in industries selling nondurable goods or services. Capital intensity, measured as depreciation expense divided by sales, was negatively related to earnings persistence, suggesting that firms with high operating leverage have less persistent earnings than firms with little operating leverage. The fourth variable, sales, was not related to persistence, even though larger firms have less volatile earnings. Using a variety of annual earnings persistence measures, Baginski et al. (1999) found that the higher-order ARIMA models document a positive association between barriers to entry, measured at the firm level as the sum of advertising and research and development over sales, and a negative relation with durable goods and capital intensity. Like Lev (1983), they also found that sales was not related to earnings persistence. However, Baginski et al. (1993) found a positive relationship between sales and earnings persistence using quarterly earnings. Finally, in a manner suggestive of Lustgarten and Thomadakis (1980), Kormendi and Lipe (1987) linked persistence to the strength of the association between accounting earnings and returns.One noticeable shift in the research is the close link between market concentration and barriers to entry. While market concentration and barriers to entry are closely related, Bain (1956) documented that barriers to entry was the primary explanatory of high returns. Mann (1966) confirmed Bain's (1956) findings by examining return on equity measures for 30 industries a decade latter. By restricting entry, firms are able to retain abnormally high profits. The early industrial economics literature discusses four sources of barriers to entry: economies of scale, product differentiation, absolute cost advantages, and capital costs.The competitive strategy literature suggests that it is a firm's business strategy, in addition to industry characteristics, that determines the persistence of profits. That is, while many of these barriers to entry are at the industry level, firms' strategies also play an important role. For instance, the competitive strategy literature (Porter 1980; Grant 2010) argues that pursing a product differentiation strategy helps firms retain a long-run competitive advantage. A differentiation strategy erects competitive barriers, as it tends to result in a price premium as the power of buyers is reduced due to the lack of product substitutability (Miller 1988). With this backdrop, CRC provide an interesting test of the competitive strategy literature using the airline industry. By limiting their study to a single industry, the study is able to focus directly the link between competitive strategy and earnings persistence. The paper provides two reasons why network carriers are expected to have more persistent earnings than the low-cost carriers. The first argument is that these airlines have a larger investment in infrastructure and higher fixed costs than the low-cost carriers. Because replicating this large investment would be extremely costly for other airlines, the investment by the network carriers results in a barrier to entry. In addition, the authors argue that entry is also made difficult because the network carriers have created a comparative advantage in running a complex hub system by developing special expertise to manage it effectively.One of the cornerstones of the paper is the delineation of competitive strategies. Rather than developing an empirical measure of strategy through survey or other empirical methods, such as in Banker et al. (2006), the study relies on the Bureau of Transportation Statistics (BTS) to define the two dominant competitive strategies in the airline industry and to identify the airlines in each group. The BTS (2005) defines a network airline as a carrier that operates "a significant portion of their flights using at least one hub where connections are made for flights on a spoke system." Alternatively, J.D. Power (2011) defines network carriers as "airlines that operate multi-cabin aircraft and use multiple airport hubs." With respect to low-cost carriers, BTS defines them as "those that the industry generally recognizes as operating under a low-cost business model." J.D. Power defines them as "airlines that operate single-cabin aircraft with typically lower fares." It would be useful to be able to determine the distinct attributes of the two operating strategies, as this would help readers understand the specific characteristics of each strategy. A more refined definition should also help facilitate generalizing the results to other industries where firms use different activities to compete. For example, descriptive statistics are provided for the two groups of airlines in Table 3 of the CRC paper. It is clear that the low-cost airlines have lower operating expenses per available seat mile and that the network airlines have higher prices per revenue passenger mile. However, there is little information on other dimensions of the two strategies, such as variety measures on customer and aircraft types, labor specialization, or hub concentration. Rather, the primary distinguishing factors seem to be that the network carriers are much larger in terms of assets, revenues, market share, and seat miles than the low-cost airlines.The theoretical and empirical work discussed above suggests that a differentiation strategy can result in barriers to entry that then lead to persistence earnings. While it is easy to argue that the empirical models used in the current study suffer from specification errors and low power, the results generally document that the profit margins of the network carriers are more persistent than those of the low-cost carriers. Turnover differences are not noticeable in the basic models. The current paper argues that network carriers adopt a business model based on a full service or differentiation strategy. There are two aspects of the paper's argument that are unclear. One deals with the nature of the network carriers and whether their business model is one of differentiation. The other is whether the margin persistence observed in the paper can be ascribed to entry barriers.Exhibit 1 presents a summary of some of the key drivers of entry barriers and persistence as discussed in the literature. Many of the factors listed in the exhibit were identified above. Other factors are identified in Waring (1996). Waring (1996) examined the persistence of firms' annual earnings using industry metrics, and found that skill level of employees, unionization, and switching costs, in addition to more traditional measures of barriers to entry, were positively related to persistence. There are aspects of the network carriers that suggest a differentiation strategy. These include the multiple classes of airlines and customers, amenities, and flying to small cities, as well as internationally. But many of these airlines are legacy carriers with high union representation and gate dominance. It is possible that rather than erecting barriers via differentiation, these network carriers create barriers by controlling scarce resources such as landing sites or concourses, wielding political clout, or predatory pricing tactics. Thus, it is not clear that a differentiation strategy is responsible for creating entry barriers, to the extent they exist.The other issue is that other factors, such as firm size, operating leverage, unionization, buyer switching costs, or some other factor, could be driving the earnings persistence, rather than entry barriers. Focusing within one industry certainly helps control for many factors, but not sufficiently to ascribe competitive strategy to the persistence differences. Thus, it is not clear if, in fact, it is the competitive strategy, through the erection of entry barriers, that is driving the observed differences in earnings persistence. For instance, conference participants were concerned about how size differences, as well as the international activities of the network airlines, could have been influencing the results. In summary, because it is very difficult to ascribe the documented differences in earnings persistence to strategy differences, future research is warranted.A very interesting aspect of the paper is the examination of how the growth, price, and productivity components of profit margin changes (and revenue growth and revenue price components of asset turnover) impact persistence differently across competitive strategies. By combining variance and productivity analysis, Banker et al. (1989) provide a framework for isolating the sources of changes in operating income. Their approach partitions operating income changes into changes in sales growth, productivity, and price recovery. Horngren et al. (2006) link these components to competitive strategy by stating that productivity gains would be more prevalent for firms following a cost leadership strategy, whereas product differentiators would be focused in price recovery. CRC use these results and construct operating income changes for the airlines in their study as follows:In the model, operating income (OI) changes are due to revenue (REV) changes and expense (EXP):In this formulation, RPM is revenue passenger miles, ASM is available seat miles, and LF is the load factor for the period. An empirical problem with their measures is that their growth measures are very closely related. In the CRC formulation, all costs are treated as varying with seat miles. Thus, growth in revenues and growth in costs are virtually identical, with a correlation coefficient reported in Table 4 of the CRC paper of 0.992. The portion of the operating income change due to growth, plus, can also be measured as:In this formulation, income growth equals the growth in revenue passenger miles times the prior period's operating income. This reformulation would allow for a more stable estimation of the relationship between overall growth and operating earnings margin persistence.Other than the measurement issue addressed above, the approach CRC use is interesting. Because there is little previous research using this partition to measure the persistence of these components, there is little theory to guide their investigation. The authors argue that because network airlines cater to more business travelers than low-cost carriers, network carriers face a more inelastic demand. This results in more persistent revenue components for the network carriers. Moreover, CRC argue that low-cost carriers are more nimble at adjusting for changes in demand and, therefore, they are able to adjust support resources more quickly than the network carriers as conditions warrant. The authors find that while the growth persistence measures are generally not different across airline type, the operating margin changes unit price and cost changes, as well as productivity gains, of network airlines are more persistent than those of the low-cost carriers. There is some difference in the turnover revenue components for the quarterly model, but no annual differences are noted. Overall, this is a very interesting approach that can be replicated in other industries.The Collins et al. (2011) study examines how the competitive strategies of airlines impact the persistence of operating earnings. It finds that that network airlines (differentiators) have more persistent operating profit margins than low-cost carriers. The study also demonstrates that partitioning the change in profit margin into growth, price recovery, and productivity more fully explains future profit margins than current profit margin alone. A very useful feature of the paper is that by focusing on one industry, the paper is able to control for many variables that have previously been associated with earnings persistence. Clearly, the paper extends our knowledge of how competitive strategy impacts the persistence of earnings.The study raises some interesting questions for future research. One avenue is to continue intra-industry research on the links between strategy and earnings persistence. A benefit of this intra-industry approach is that it controls for many of the other factors related to persistence. Another avenue, and one that is particularly relevant for management accountants, is to continue CRC's work on the links between earnings persistence, strategy, and the components of earnings changes. In this manner, we can help identify which strategic operating initiatives, such as those related to pricing or productivity, result in lasting changes in profitability.

This paper applies a model of market power measurement under product differentiation to the case of the gasoline market in California, using data for the period 1983–1989. Our results show that there is a considerable degree of product differentiation among major brands. This allows firms to exercise local market power over their own specific products, but there are also signals of an important degree of global market power. However, none of the four pure market structures analyzed (price taking, monopolistic competition, Cournot oligopoly and collusion) seems able to explain by itself the behavior of the whole market.

In the recent years, a vast number of economic studies focused on the art market and on its development, as well as on the artists as agents of a labour market with features that are particular of art market. However, really few studies analyzed the functioning of the art market as a whole with a theoretical approach, often considering the artist as nothing more than a simple producer of artworks. This dissertation aims to enrich the economics literature with the above-mentioned missing parts. In Chapter 2 we build a model for studying the price formation in the private art market in a context of complete information and full rationality, using a game theory approach. In this first part, the key variables in this market are the market powers of the agents in the market, that define what is the channel through which a certain artwork reaches the public market. Chapter 3 firstly better defines what are the economic and cultural values and how their formation works for artworks, and then it introduces in the just-created framework the effects of information asymmetry among the agents operating in the market, and behavioural aspects in the choices of some of them: besides market powers, also information has a key role in the art market,and it mainly depends on artists’ characteristics, as fame and talent. Chapter 4 empirically investigates these artists’ characteristics, creating a new measure for them, through the use of hedonic regression and quantile hedonic regression, finding that this measure presents a bimodal distribution in all the specifications we implemented, coherently with the superstar theory in the art market, and that there exist various effects than influence the ranking created using our measure. Chapter 2 Chapter 2 analyses the pricing of artworks just created by an artist and sold for the first time in the art market, investigating the price-formation mechanism at work in the private art market. In particular, a game theory approach is used to consider the possible channels (paths) that a new artwork can take to reach a collector or an auction house, assuming price-maximizing agents with full information on market powers and reserve prices. The study is aimed to identify the relationships between the artwork prices and the market power of agents operating in each channel of the market. What we find is that the market power of each of the agents is key to identify the market channel that will be preferred by the artist and, then, her incentive in creating a new artwork. At the same time, the importance of art market intermediaries as the galleries in the formation of art prices is attested and their incentives to enter and remain in the market are confirmed. Chapter 3 Chapter 3 aims to explore the role of cultural and economic value, and of the information on private art market price formation; in this Chapter, we develop a bargaining game model in which we explicitly consider the effects of these issues. Furthermore, we introduce artists’ fame and talent as determinants of an artwork’s cultural and economic value. Assuming artists, galleries, and collectors have different levels of information on the quality of the artworks and on the characteristics of the artists, we study the behaviour of these agents and the potential emergence of disappointment for the sophisticated collectors and undertreatment for the unsophisticated ones. Artworks will be treated as credence goods, experience goods, or search goods, depending on the level of information each agent has at his disposal; this influences the way the price is formed and how the surplus of the trade is shared among the agents in certain particular trade channels. Chapter 4 In Chapter 4, we aim to empirically analyze those characteristics of the artists that influence cultural and economic values, in order to understand how the art market considers them in the formation of prices. To do so, using a unique hand-collected dataset from the Artist Re-sale Rights (ARR) archives of the Societ`a Italiana degli Autori ed Editori (SIAE), we build a novel index of artists’ talent and fame in the Italian art market. The distribution of this measure and the ranking of the artists’ index is analyzed in several model specifications, and the dynamics of the index is also qualitatively studied. Despite of the existence of a price effect, a time effect, and an art genre effect, coherently with the superstars theory, the bimodality of the distribution is confirmed in all our empirical findings.

intra-industry trade and in the neotechnology theory of international trade; (ii) to indicate the limitations of Hufbauer's measure of product differentiation; and (iii) to propose an alternative measure of product differentiation. The mist surrounding points (i) and (ii) has no doubt been cleared. However, the alternative suggested by the authors is by no means perfect. It is with this last aspect that this note is concerned. The authors have proposed the construction of "hedonic" price indexes using regression techniques to capture the product differentiation effect in international trade in place of Hufbauer's coefficient of variation based on unit values. This suggestion is indeed an improvement over Hufbauer's measure. However, the problem of measurement of product differentiation in international trade is not solved. This is because a researcher will be faced with as many export prices1 as there are export markets; the differential being due to differences in factors and magnitude of factors operating in different export markets. The authors themselves have acknowledged price variation in different export markets as a problem in the measurement of product differentiation in international 1 The authors have not clarified whether unit values or export prices are to be used for constructing "hedonic" price indexes for measuring product differentiation in international trade. However, the evidence, namely, (i) the limitations of unit values which constitute

Assuming symmetry across firms and constant unit costs Perloff and Salop (1985) show: If product differentiation increases, prices rise in a symmetric equilibrium. This raises the question of whether, in general, more product differentiation leads to higher market prices. Giving up the symmetry and the constant unit costs assumptions we present examples in which at least one firm lowers its equilibrium price when product differentiation increases. We formulate a model of product differentiation and state and discuss, within the theory of supermodular games, conditions ensuring that all firms raise their prices in a Nash equilibrium if product differentiation increases.