Multiproduct firms' entry and exit decisions: The case of the United States sport utility vehicle market

Empirical analysis of multiproduct firms' decisions to offer new products or withdraw existing ones is relatively scarce. Theoretical models are rooted in Hotelling's theory of spatial competition and product differentiation, which focus on firms' decisions on where to locate their product(s) in quality space. Multiproduct firms face conflicting pressures when deciding where to locate new products. They may locate products close together to take advantage of local economies of scope, but such a strategy could cannibalize sales from existing products. Firms' product location choice can also be a strategic decision to deter entry. Product proliferation deters entry by segmenting the market while spreading products throughout the product space can preempt entry by filling empty market niches. Theoretical models Produce mixed results about the outcome of a multiproduct firms' entry decision depending on various assumptions in the models.; This research applies techniques from Stavins (1995) on entry and exit of multiproduct firms to the SUV segment of the U.S. auto industry. Rapid growth of the number of sport utility vehicles offered in the 1990s makes it a natural case study of multiproduct firms' entry and exit decisions. A hedonic regression is used to reduce the multidimensional characteristics of SUVs to a single quality measure. The empirical analysis reveals both U.S. and Japanese firms disperse their models throughout quality space. SUV manufacturers' product location decisions reveal the Japanese firms are concerned about cannibalizing sales from existing products while U.S. firms seek to take advantage of local economies of scope. Moreover, SUV models in areas with consumer valuation beyond that attributed to quality are less likely to exit while models in crowded areas are more likely to exit.; The entry of new firms and products in the SUV market may have changed the degree of competition in the market. Attempts were made to apply New Empirical Industrial Organization techniques to test whether entry resulted in a more competitive environment, but the data were not well suited for the modeling technique chosen. The problems encountered are likely to be found elsewhere, however, highlighting the strong assumptions of the model for applied econometrics.