Competition, Product Proliferation and Welfare: A Study of the U.S. Smartphone Market
We develop and estimate a structural model of the U.S. smartphone market. Based on the estimates, we study (1) whether there are too few or too many products in the market from a welfare point of view; and (2) how competition affects product offerings in the market. We find that there are too few products and a reduction in competition further decreases the product offerings. These results suggest that merger policies should be stricter when we take into account the effect of merger on firms' product choices in addition to its effect on prices.
This project studies product proliferation in the U.S. smartphone market. The smartphone industry has been one of the fastest growing industries in the world with billions of dollars at stake. The worldwide smartphone sales grew from 122 million units in 2007 to 1.4 billion units in 2015. Global revenue was more than 272 billion dollars in 2015. Moreover, product proliferation is prominent in the smartphone industry. For example, on average, Samsung offered 13 smartphones in the U.S.\ market in a given month during our sample period (January 2009 to March 2013). This average is 11 for HTC and 8 for Motorola. There is substantial within-firm dispersion on price and product quality, implying that firms offer a wide selection of smartphones across the quality spectrum and charge different prices for them.
In this project, we study two questions about product proliferation in the U.S. smartphone industry. First, are there too few or too many products form a welfare point of view? On the one hand, there may be excessive product proliferation because a profit-maximizing firm will have a product in the market as long as it is profitable to do so. Some of the profit may come from business stealing. Since firms do not take this negative externality into account, there can be too many products from a welfare point of view. On the other hand, different from a social planner, firms do not internalize consumer surplus either. Assuming that consumer surplus is increasing in the set of products in the market, there may also be too few products. In sum, whether there are too few or too many products in the market is an empirical question. Second, how does competition affect product offerings? On the one hand, there may be excessive product proliferation because a profit-maximizing firm will have a product in the market as long as it is profitable to do so. Some of the profit may come from business stealing. Since firms do not take this negative externality into account, there can be too many products from a welfare point of view. On the other hand, different from a social planner, firms do not internalize consumer surplus either. Assuming that consumer surplus is increasing in the set of products in the market, there may also be too few products. In sum, whether there are too few or too many products in the market is an empirical question.
Combining these two research questions, this paper sheds light on how to adjust the leniency of competition policies when product offerings are endogenous. If competition leads to too many products, and merger reduces product offerings, then the merger policy should be more lenient because the consumer welfare loss due to potentially higher prices may be offset by the savings of fixed costs in producer surplus. Conversely, the policy should be stricter if merger reduces product offerings when there are already too few products in the market.
To study the two research questions of interest, we develop a structural model describing consumer demand and firms’ product and price decisions. We estimate the model using data from the Investment Technology Group (ITG) Market Research. Using these data, we estimate demand and marginal cost. We find that, on average, consumers prefer smartphones with a longer battery talk time, a higher camera resolution, a later generation of the chipset and a larger screen size. We also find sizable consumer heterogeneity in their willingness-to-pay for quality. On the supply side, not surprisingly, we find that marginal cost is increasing in quality and decreasing over time. We also obtain bound on fixed costs of having a product in the market.
Based on the estimated model, we conduct a set of counterfactual simulations to address the research questions of the project. We find that dropping the lowest-quality product in a market leads to a welfare decrease, even considering the maximum possible savings in fixed costs. Adding a discontinued product, however, increases welfare as long as the fixed cost of the product is not too large. These results suggest that there are inefficiently too few products in the market, and that the effect of firms not including consumer surplus in their objectives dominates the effect of firms not internalizing the business-stealing externalities. Through a merger simulation of Samsung and LG, we find that a reduction in competition further decreases the product offerings, implying that the direct effect of merged firm internalizing the business-stealing externality dominates the indirect effect of a reduction in the pricing competition. These findings together suggest that in markets where product offerings are important, a merger policy should be stricter when the merger effect on product offerings is taken into account.
Intellectual Merit of Project
The intellectual merit of this project lies in developing and estimating a demand and supply model of the smartphone industry, and using this model to study research questions relevant to this industry and of general interest. Moreover, in the merger simulations, we have to solve a computational issue. For example, the Samsung-LG has 27 potential products. Its product portfolio can be any subset (of any size) of these 27 potential products. There are more than 134 million such product portfolios. To deal with this computational challenge, we propose a heuristic algorithm to solve for a firm’s best response and embedded this algorithm in a best-response iteration to solve for the post-merger equilibrium.
Broader Impact of Project
We believe this research is important for three main reasons. First, the smartphone industry is economically and socially important. The smart phone industry has been one of the fastest growing and most innovative industries in the world with billions of dollars at stake. In 2013, the price of a mainstream product, such as an iPhone 5, Galaxy S4/S5 or HTC One/8x, ranges from $300 to over $650. Samsung shipped out over 80 million smart phones in the third quarter of 2013 worldwide, and Apple more than 30 million. The industry has affected millions of consumers’ daily life. Second, the economic issues studied in the project are not special to this industry. Product proliferation is common in many industries. Prominent examples include the PC market, the printer market and the CPU market. Therefore, the answers to the research questions studied in this project and their policy implication is of general interest. Third, the computational challenge faced by this project is also common in economic researches. Our algorithm therefore may be useful for other researches when the decision variable is discrete and the action space is prohibitively large.