Competing Industrial Standards and the Impact of Trade Liberalization

Abstract

The main purpose of this study is to illustrate, with simple trade theory, the relationship between competing industrial standards and trade liberalization. We assume that there are two competing industrial standards in an international context, each of which applies to a group of differentiated products. A product can be used only in combination with other products based on the same industrial standard. We examine the impact of trade liberalization (i.e., a decline in trade costs) on consumers’ choice of a standard. It will be shown that the degree of indirect network effects, captured with substitution between differentiated products, plays an important role as a determinant of the impact of trade liberalization.

Keywords:

• Competing industrial standards
• trade liberalization

JEL CLASSIFICATION :

• F12

Acknowledgements

We are greatly indebted to Jota Ishikawa, the anonymous referees and the Editor of this journal for detailed and insightful comments.

Notes

¹Addressing this point, the OECD (2006: Ch. 2) reports that between 1996 and 2004 the annual increase in OECD countries’ exports of ‘software goods’ was 5%, while imports increased by an average of 6.5% per year.

²In this study we will use the term ‘standard,’ not in the sense of government regulation, but in the universal sense of the set of technical specifications that enable compatibility among products.

³The seminal contributions on the indirect network effect are by Chou and Shy (1990) and Church and Gandal (1992). See Gandal (2001, 2002) and Farrell and Klemperer (2007) for surveys of the relevant literature. In the international context, see Iwasa and Kikuchi (2008) and Kikuchi (2005, 2007) for analyses of trade liberalization in the presence of network effects.

⁴Motorola’s share of the world market dropped from 40% in 1994 to 32% in 1995, as use of the European GSM standard grew. Still, it is larger than Nokia’s share in the world market (22%). See, also, Lembke (2002).

⁵Burton and Saelens (1987, p. 291) note that, while sales by Japanese firms accounted for 43.5% of all sales in the US market in 1981, Japanese firms held only a 15.2% share of the Western European market in 1983. (Note that the Japanese televisions that are exported to the Western Europe are based on PAL system.) Rohlfs (2001) discusses this in terms of network effects.

⁶See Matsuyama (1992) for elaboration on this point. This assumption implies that, for one country’s producers, the cost of converting to the other standard is extremely high. Although this assumption is restrictive, it is often argued that the existence of (high) conversion costs affects foreign firms’ behavior (Gandal & Shy, 2001).

⁷As a first step to incorporate competition between industrial standards, we will concentrate on the case of identical technology between standards. In order to analyze the interaction of technology gap and standard competition, an extension for asymmetric technologies needs further consideration.

⁸For discussion on the role of market entry in the presence of fixed costs, see Melitz (2003).

⁹From equation (11) we obtain

which gives the value of ρ* corresponding to the interior equilibrium. Note that , , ∂ψ(ϵ, t)/∂ϵ>0, and ∂ψ(ϵ, t)/∂t<0.

¹⁰Figure 4 indicates that, the smaller (resp. larger) t becomes, the higher the possibility of (∂W/∂t)<0 (resp. (∂W/∂t)>0) becomes.

¹¹For this point, see, Kemp and Shimomura (2001) and Fujiwara (2005).

¹²In case (2c), since there are only Foreign standard products, trade liberalization does not change the nature of trade flows.

¹³See Lawrence (1987).

¹⁴It is important to note that point I is unstable. One interpretation of this situation is as follows: suppose that initially σ is sufficiently close to (but larger than) 2 for the point I to be stable. Then, some parameter change increases σ, which changes the slope of the ϕ curve. In this situation, at least initially, the economy might stay at point I. Thus, the results in case (2b) need to be interpreted with great caution.

¹⁵Another possible equilibrium is point I′. However, since potential Foreign firms enter due to improved access to the Home market and more consumers switch to Foreign products, it seems to be natural that point N will be selected as the new equilibrium.

¹⁶We use ‘^∼’ to denote Foreign market values.

¹⁷This approach relates to that of Melitz (2003). In an alternative approach, one can assume that the fixed costs are paid only once. Then, the equilibrium number of products is obtained by adding both markets’ demand conditions. Although that approach is desirable, deriving explicit solutions from such a system of simultaneous equations might not be feasible. An extension for this setting needs further consideration. We would like to thank one referee for pointing this out.

¹⁸See, for example, Helpman and Krugman (1985).

¹⁹Sufficient condition for (∂W/∂t)>0 can be obtained similarly.

²⁰Clearly, h satisfies

²¹One can verify that