Industrial Organization Program Meeting

February 10 and 11, 2012
Timothy F. Bresnahan, Stanford University, and Jakub Kastl, Stanford University and NBER, Organizers

Fernando Ferreira, University of Pennsylvania and NBER, and Amil Petrin and Joel Waldfogel, University of Minnesota and NBER

Trade and Welfare in Motion Pictures

Trade benefits consumers and producers. Importing consumers benefit from access to a wider variety of products, while exporting sellers experience higher profits by selling their products to a larger population of consumers. The effects of trade also can operate through product quality: larger markets can lead to larger investments in products and therefore to higher quality products, so that consumers in both the exporting and importing countries will experience a direct benefit from trade operating through quality. The movie industry is an auspicious context for exploring this phenomenon, because quality is produced exclusively with sunk costs, these sunk costs are high, international revenue is very important, and there are substantial obstacles to movie trade. Ferreira, Petrin, and Waldfogel develop a model of the global movie market, including consumers' movie demand (with country-by-movie-specific preferences), the relationship between movie budgets and quality, and the equilibrium yielding the movie budgets. This allows them to quantify the benefit of exporting for U.S. and foreign consumers, as well as the counterfactual impact on consumers and producers of various policies including elimination of European film subsidies and an expansion of the Chinese movie market.


Jennifer Brown, Northwestern University and NBER, and Dylan B. Minor, Northwestern University

Selecting the Best? Spillover and Shadows in Elimination Tournaments (NBER Working Paper No. 17639)

Brown and Minor consider how an elimination tournament's ability to select the most skilled competitor as the winner is shaped by past, current, and future competition. They present a two-stage model that yields the following main results: 1) a shadow effect - the weaker the expected future competitor, the greater the probability that the stronger player wins in the current stage; and 2) an effort spillover effect - previous effort reduces the probability that the stronger player wins in the current stage. They test their theory predictions using data from high-stakes tournaments and betting markets. The empirical results suggest that shadow and spillover effects influence match outcomes.


Brian Viard, Cheung Kong Graduate School of Business, and Shihe Fu, Xiamen University

The Effect of Beijing's Driving Restrictions on Pollution and Economic Activity

Viard and Fu evaluate the environmental and economic effects of Beijing's driving restrictions. According to daily data from multiple monitoring stations, air pollution falls 19 percent during every-other-day driving restrictions and 8 percent during one-day-per-week driving restrictions. According to data on hourly television viewership, the number of viewers during the driving restrictions increases 1.7 to 2.3 percent for workers with discretionary work time but is not affected for workers without that discretion. That is consistent with the restrictions' higher per-day commute costs reducing daily labor supply. The authors identify causal effectsfrom both time-series and spatial variation in air quality and intra-day variation in viewership. They provide possible reasons for the policy's success, including evidence of high compliance based on parking garage entrance records. Their results also provide new evidence on commute costs and labor supply.


Abe Dunn and Adam Hale Shapiro, Bureau of Economic Analysis

Physician Market Power and Medical-Care Expenditures

Dunn and Hale Shapiro empirically assess the degree to which physicians exploit their bargaining leverage over insurance carriers as a means to raise service prices. The researchers also examine the degree to which these potentially higher payments may translate into different levels of service utilization. They find that physicians are able to translate bargaining leverage into both higher fees and higher service utilization. Ceteris paribus, a cardiologist with high market power (concentration in the 90th percentile) will charge 26 percent higher prices and perform 24 percent more services than a cardiologist with low market power (concentration in the 10th percentile). The corresponding orthopedist will charge 29 percent higher prices and perform 6 percent more services. The authors provide evidence that the effect of bargaining leverage on service utilization may be explained by physicians responding to the negotiated service prices.

Jason Allen, Bank of Canada; Robert Clark, HEC Montreal; and Jean-Francois Houde, University of Wisconsin, Madison

Price Negotiation in Differentiated Product Markets: The Case of Insured Mortgages in Canada

Allen, Clark, and Houde measure market power in a decentralized market where contracts are determined through a search and negotiation process. The mortgage industry has many institutional features which suggest it should be competitive: homogeneous contracts, negotiable rates, and, for a given consumer, common lending costs across lenders. As a result, even with a small number of competing lenders, informed borrowers can gather multiple quotes offering interest rates that reflect the expected cost of lending. However, there is important heterogeneity in the ability of consumers to understand the subtleties of financial contracts, in their ability or willingness to negotiate and search for multiple quotes, and in their degree of loyalty to their main financial institutions. These authors propose and estimate a model to disentangle the different channels through which market power can arise for a given transaction in the mortgage environment. They find two main sources of market power. The first is search frictions; to quantify the, they develop a model that is sequential -- consumers are initially matched with a home bank to obtain a mortgage quote, and then can decide, based on their search costs, whether to gather additional quotes from banks in their neighborhood. The authors find that over the five-year period of the contract, the average search cost corresponds to an upfront sunk cost of between $1,047 and $1,590. The second main source of market power is switching costs. The authors estimate that consumers are willing to pay between $759 and $1,617 upfront to avoid having to switch banks.


Carlos E. Noton, University of Warwick

Revealing Bargaining Power through actual wholesale prices

Noton uses product-specific wholesale and retail prices to study bargaining power. He focuses on two outcomes of bargaining between coffee manufacturers and supermarkets in Chile: 1) the share of total profits that each player earns; and 2) the risk exposure to cost shocks that each player bears. He finds that Nestle, which accounts for almost 80 percent of the coffee market, obtains 70 percent of the total profits. Surprisingly, however, small manufacturers obtain between 30 and 50 percent of profits. These estimates suggest that a low degree of consumer substitutability can offset market size in terms of bargaining power. In terms of risk exposure, most cost shocks are absorbed by upstream manufacturers, and the small manufacturers bear more risk than the larger players. Supermarkets' pricing strategies also appear to play a role in the risk-sharing outcomes.


Liran Einav, Stanford University and NBER; Theresa Kuchler, Stanford University; Jonathan D. Levin, Stanford University and NBER; and Neel Sundaresan, eBay Research Labs

Learning from Seller Experiments in Online Markets (NBER Working Paper No. 17385)

The internet has dramatically reduced the cost of varying prices, displays, and information provided to consumers, facilitating both active and passive experimentation. Einav, Kuchler, Levin, and Sundaresan document the prevalence of targeted pricing and auction design variation on eBay, and they identify hundreds of thousands of experiments conducted by sellers across a wide array of retail products. They show how this type of data can be used to address questions about consumer behavior and market outcomes, and they provide illustrative results on price dispersion, the frequency of over-bidding, the choice of reserve prices, "buy now options", and other auction design parameters, and on consumer sensitivity to shipping fees. They argue that leveraging the experiments of market participants takes advantage of the scale and heterogeneity of online markets and can be a powerful approach for testing and measurement.