« Back to Results

Big Data in the Modern Economy

Paper Session

Saturday, Jan. 5, 2019 8:00 AM - 10:00 AM

Atlanta Marriott Marquis, International 5
Hosted By: American Economic Association
  • Chair: Christopher Tonetti, Stanford University

The Impact of Big Data on Firm Performance: An Empirical Investigation

Patrick Bajari
,
University of Washington
Victor Chernozhukov
,
Massachusetts Institute of Technology
Ali Hortaçsu
,
University of Chicago
Junichi Suzuki
,
Amazon

Abstract

In academic and policy circles, there has been considerable interest in the impact of “big data” on firm performance. We examine the question of how the amount of data impacts the accuracy of Machine Learned models of weekly retail product forecasts using a proprietary data set obtained from Amazon. We examine the accuracy of forecasts in two relevant dimensions: the number of products (N), and the number of time periods for which a product is available for sale (T). Theory suggests diminishing returns to larger N and T, with relative forecast errors diminishing at rate 1/√N+1/√T. Empirical results indicate gains in forecast improvement in the T dimension; as more and more data is available for a particular product, demand forecasts for that product improve over time, though with diminishing returns to scale. In contrast, we find an essentially flat N effect across the various lines of merchandise: with a few exceptions, expansion in the number of retail products within a category does not appear associated with increases in forecast performance. We do find that the firm’s overall forecast performance, controlling for N and T effects across product lines, has improved over time, suggesting gradual improvements in forecasting from the introduction of new models and improved technology.

Big Data and Firm Dynamics

Maryam Farboodi
,
Princeton University
Roxana Mihet
,
New York University
Thomas Philippon
,
New York University
Laura Veldkamp
,
Columbia University

Abstract

Big data is transforming the modern economy. Data has become a valuable asset, because it allows a firm to learn about its customers and produce more valuable goods. Data has changed firm dynamics as well: More production creates more data, which makes production more valuable. We construct an aggregate model of competition among growing firms that accumulate data, and explore how data affects firm valuation, growth and competition.

Nonrivalry and the Economics of Data

Chad Jones
,
Stanford University
Christopher Tonetti
,
Stanford University

Abstract

Data is nonrival: a person's location history, medical records, and driving data can be used by any number of firms simultaneously without being depleted. Nonrivalry leads to increasing returns and implies an important role for market structure and property rights. Who should own data? What restrictions should apply to the use of data? We show that in equilibrium, firms may not adequately respect the privacy of consumers. But nonrivalry leads to other consequences that are less obvious. Because of nonrivalry, there may be large social gains to sharing data across firms, even in the presence of privacy considerations. Fearing creative destruction, firms may choose to hoard data they own, leading to the inefficient use of nonrival data. Instead, giving the data property rights to consumers can generate allocations that are close to optimal. Consumers appropriately balance their concerns for privacy against the economic gains that come from selling data to all interested parties.

A/B Testing

Eduardo Azevedo
,
University of Pennsylvania
Alex Deng
,
Microsoft
José Luis Montiel Olea
,
Columbia University
Justin Rao
,
Microsoft
Glen Weyl
,
Yale University

Abstract

Large and thus statistically powerful “A/B tests” are increasingly popular in business and policy to evaluate the efficacy of exogenous interventions. Yet the ability of such precise tests to detect small improvements may be of limited value if most gains accrue from rare and unpredictable large successes that can be detected using tests with smaller samples. We show that if the tails of the (prior) distribution of true effect sizes is not too fat, the standard approach of trying a few high-powered experiments is quite sensible. When this distribution is very fat tailed however, a “lean experimentation” strategy of trying more but smaller interventions is preferred. We measure this tail parameter using experiments from Microsoft Bing’s EXP platform and find extremely fat tails. Our theoretical results and empirical analysis suggest that even simple changes to business practices within Bing could dramatically increase revenue.
JEL Classifications
  • O4 - Economic Growth and Aggregate Productivity
  • L1 - Market Structure, Firm Strategy, and Market Performance