Economics of Digitization

March 6, 2015
Shane Greenstein of Northwestern University, Josh Lerner of Harvard Business School, and Scott Stern of MIT, Organizers

Samuel Paul Fraiberger and Arun Sundararajan, New York University

Peer-to-Peer Rental Markets in the Sharing Economy

Peer-to-Peer Rental Markets in the Sharing Economy Samuel Paul Fraiberger, New York University Arun Sundararajan, New York University

Fraiberger and Sundararajan develop a new dynamic model of peer-to-peer, Internet-enabled rental markets for durable goods in which consumers may also trade their durable assets in (traditional) secondary markets, transaction costs and depreciation rates may vary with usage intensity, and consumers are heterogeneous in their price sensitivity and asset utilization rates. The researchers characterize the stationary equilibrium of the model. They analyze the welfare and distributional effects of introducing these rental markets by calibrating their model with U.S. automobile industry data and two years of transaction-level data they obtained from Getaround, a large peer-to-peer car rental marketplace. The authors’ counterfactual analyses vary marketplace access levels and matching frictions, showing that peer-to-peer rental markets change the allocation of goods significantly, substituting rental for ownership and lowering used-good prices while increasing consumer surplus. Consumption shifts are significantly more pronounced for below-median income users, who also provide a majority of rental supply. Their results also suggest that these below-median income consumers will enjoy a disproportionate fraction of eventual welfare gains from this kind of 'sharing economy' through broader inclusion, higher quality rental-based consumption, and new ownership facilitated by rental supply revenues.


Kevin Williams, Yale University, and Thomas W. Quan, University of Minnesota

Product Variety, Across-Market Demand Heterogeneity, and the Value of Online Retail

This paper quantifies the effect of increased product variety in online markets on consumer welfare and firm profitability. Williams and Quan show the gains may be small if consumer tastes vary geographically and brick-and-mortar stores cater to the local demand. The researchers use an original data set from a large online retailer containing millions of transactions. However, the large choice set leads to many products having zero local market shares. The authors propose a modification to Berry (1994) and Berry, Levinsohn, Pakes (1995), where both national and local market shares are used to recover geographically varying mean utilities. The researchers' two step approach is easy to implement and fits their data well. Their results indicate that products face substantial heterogeneity in demand across markets, with more niche products facing greater heterogeneity. Failing to account for across-market demand heterogeneity grossly overstates the consumer welfare gain of increased online product variety, and on the supply side they find traditional retail chains can generate a substantial increase in revenue by localizing assortments.


Weijia Dai, University of Maryland; Ginger Zhe Jin, University of Maryland and NBER; Jungmin Lee, Sogang University; and Michael Luca, Harvard University

Optimal Aggregation of Consumer Ratings: An Application to Yelp.com (NBER Working Paper 18567)

Consumer review websites leverage the wisdom of the crowd, with each product being reviewed many times (some with more than 1,000 reviews). Because of this, the way in which information is aggregated is a central decision faced by consumer review websites. Given a set of reviews, what is the optimal way to construct an average rating? Dai, Jin, Lee, and Luca offer a structural approach to answering this question, allowing for (1) reviewers to vary in stringency and accuracy, (2) reviewers to be influenced by existing reviews, and (3) product quality to change over time. Applying this approach to restaurant reviews from Yelp.com, the researchers construct optimal ratings for all restaurants and compare them to the arithmetic averages displayed by Yelp. Depending on the interpretation of the downward trend of reviews within a restaurant, the researchers find 19.1-41.38% of the simple average ratings are more than 0.15 stars away from optimal ratings, and 5.33-19.1% are more than 0.25 stars away at the end of their sample period. Moreover, the deviation grows significantly as a restaurant accumulates reviews over time. This suggests that large gains could be made by implementing optimal ratings, especially as Yelp grows. Their algorithm can be flexibly applied to many different review settings.

Erik Brynjolfsson, MIT and NBER, and Kristina McElheran, University of Toronto and MIT

Data in Action: Data-Driven Decision Making in U.S. Manufacturing

Brynjolfsson and McElheran investigate the adoption and performance effects of data-driven decision-making (DDD) in U.S. firms. Their study relies on a novel survey of managerial practices conducted by the U.S. Census Bureau combined with other detailed data on information technology (IT) investment and firm performance in the manufacturing sector. Using a differences-in-differences approach, the researchers find that plants adopting DDD exhibit an average increase of 3% in value-added between 2005 and 2010. For the average plant in the sample, this increase is on par with the productivity benefit from investing an additional $5 million in IT capital, or $60 thousand per employee over the five-year period. Because IT use is managerially and statistically distinct from IT investment, the authors take this as evidence that it can be just as important for firm performance. Moreover, DDD displays important complementarities with other organizational choices such as employee education and the allocation of decision rights. Furthermore, the variation in the timing and distribution of DDD effects suggests there are complex and costly organizational adjustments as firms adapt to new technological possibilities.


Michela Giorcelli, Stanford University, and Petra Moser, Stanford University and NBER

Copyrights and Creativity: Evidence from Italian Operas

This paper exploits variation in the adoption of copyright laws within Italy – as a result of variation in the timing of Napoleon's military victories – to examine the effects of copyrights on creativity. To measure variation creative output, Giorcelli and Moser use new data on 2,598 operas that premiered across eight states within Italy between 1770 and 1900. These data indicate that the adoption of copyrights led to a significant increase in the number of new operas premiered per state and year. The researchers find that the number of high-quality operas also increased – measured both by their contemporary popularity and by the longevity of operas. By comparison, evidence for a significant effect of copyright extensions is limited. Their analysis of alternative mechanisms for this increase reveals a substantial shift in composer migration in response to copyrights. Consistent with agglomeration externalities, the researchers also find that cities with a better pre-existing infrastructure of performance spaces benefited more copyright laws.


Hong Luo, Harvard University, and Julie H. Mortimer, Boston College and NBER

Copyright Enforcement in Stock Photography

Digital technologies facilitate online piracy but also make it easier to detect unauthorized use of copyrighted goods. Detection of infringement provides copyright owners the opportunity to monetize their products ex-post through settlement. This paper studies the effectiveness of different enforcement methods in the context of infringing use of digital images by businesses. Luo and Mortimer use a novel, proprietary dataset from a leading agency in the stock photography industry. Two field experiments exogenously vary (1) the requested settlement amount and (2) the wording of the request letter. The researchers find that, on average, a substantial reduction in the requested amount generates only a small increase in the settlement probability. In contrast, given the same lower requested amount, a message that informs infringers of the price reduction and acknowledges the possibility of unintentional infringement has a large positive effect on settlement. The authors also find that adding a deadline, after which the forgiven amount is added, has a positive and significant effect on the pre-deadline outcome. Their findings highlight the importance of ex-post licensing in contexts with unintentional infringement, and suggest that reciprocity may play a role in settlement communications with small businesses.