« Back to Results

Firm Responses to Incentives and Regulation

Paper Session

Saturday, Jan. 6, 2018 8:00 AM - 10:00 AM

Marriott Philadelphia Downtown, Meeting Room 405
Hosted By: Econometric Society
  • Chair: Mitsuru Igami, Yale University

Measuring the Incentive to Collude: The Vitamin Cartels, 1990-1999

Mitsuru Igami
Yale University
Takuo Sugaya
Stanford University


Why do some cartels survive for a decade but others collapse within a few years? Models of collusion and repeated games are usually difficult to identify from observational data, but the vitamin cartels, one of the most prominent cases in recent history, provide direct evidence on their internal organization in the form of American court documents and European antitrust enforcement. Our estimates suggest the cartel leader's incentive to collude diminished significantly at the time of the vitamin C cartel's actual collapse in 1995, mainly because of deteriorating demand conditions and unexpected increases of fringe supply, whereas the markets for beta carotene, vitamin A, and vitamin E remained stable until prosecution in 1999. We also find that an earlier consummation of the 2001 BASF-Takeda merger would have saved the vitamin C cartel.

Certification, Reputation and Entry: An Empirical Analysis

Xiang Hui
Massachusetts Institute of Technology
Maryam Saeedi
Carnegie Mellon University
Giancarlo Spagnolo
Stockholm School of Economics
Steven Tadelis
University of California-Berkeley


More precise information about past performance of incumbent firms
has implications for entry decisions and affects the quality distribution
of firms in the market. However, this relationship has not been well
studied empirically, because of limited data. We exploit a policy
change on eBay as a quasi-experiment, in which the "eBay Top Rated
Seller" badge substituted the previous "Powerseller" badge,
by making the requirements more stringent and adding new performance measures for qualification. This change results in some incumbent
sellers losing their badge, implying that the quality
of both certified and uncertified sellers increases. This change can
generate countervailing incentives for entrants of different quality
levels. On one hand, it may induce more entry by high-quality firms by increasing
their future payoff conditional on receiving the badge. On the other hand, it
may induce more entry by low-quality firms, because the average
quality of uncertified sellers with whom they may pool increases. We exploit the differential impact of the policy change on different subcategories of the eBay
marketplace for identification. First, we document a negative correlation
between the number of entrants and the share of badged sellers across
categories. We find that after the policy change, categories that
experience a higher reduction in the share of badged sellers experience a larger increase in the number of entrants. However, this difference tends to disappear
once the market adjusts to the new equilibrium after about six months.
Second, we find that the distribution of quality provided by entrants
changes and has fatter tails. In other words, a larger share of
entrants provides qualities from the first and last two deciles of
the quality distribution after the policy change. This finding is
consistent with the prediction that entrants from the extremes of
the quality distribution have stronger incentives to enter, although
it could be due to similar entrants changing their behavior. Third,
we find almost no change in the quality provided by incumbents. This
last finding suggests that the increase in the quality provided by
entrants at the tails of the distribution is more likely to
come from improved selection rather than from a change in entrants'
behavior. The results have relevant implications for the design of
reputation and certification mechanisms in electronic and other markets,
and shed some light on the concerns expressed by the U.S. senate and
the EU that allowing public buyers to use past performance information
when selecting contractors may hinder entry by novel or foreign firms
in public procurement markets.

Understanding Disparities in Punishment: Regulator Preferences and Expertise

Karam Kang
Carnegie Mellon University
Bernardo Silveira
Washington University-St. Louis


This paper provides an empirical framework to evaluate regulatory discretion and apply it to California water quality regulation. We identify and estimate an adverse selection model of the regulator-discharger interaction to measure the extent to which the regulator's environmental preferences and administrative or political costs affect penalties and compliance. We find that, because of the heterogeneity in compliance costs, the disparities in penalties would decrease by at most 15 percent, even if the regulator were homogeneous across dischargers. We also find that introducing a one-size-fits-all policy would lead to an increase in violations by large facilities and those located in densely populated areas.

A Dynamic Model of Vehicle Ownership, Type Choice, and Usage

Kenneth Gillingham
Yale University
Fedor Iskhakov
Australian National University
Anders Munk-Nielsen
University of Copenhagen
John P. Rust
University of Maryland
Bertel Schjerning
University of Copenhagen


This paper develops an estimable structural microeconometric model of car choice and usage that features endogenous equilibrium prices on the used-car market. Households buy and sell cars in the market and car owners choose how much to drive their car in a finite-horizon model. Moreover, we explicitly model the choice between scrapping the car or selling it on the used-car market. We estimate the model using full-population Danish register data on car ownership, driving and demographics for the period 1996--2009, covering all Danish households and cars. Simulations show that the equilibrium prices are essential for producing realistic simulations of the car age distribution and scrappage patterns over the macro cycle. We illustrate the usefulness of the model for policy analysis with a counterfactual simulation that reduces new car prices but raises fuel taxes. The simulations show how equilibrium prices imply that the boom in new car sales come at the cost of accelerated scrappage of older cars. Furthermore, the model gives predictions on tax revenue, fuel use, emissions, the lifetime of vehicles as well as the composition of types and ages of cars in the future.
Matthew Carl Weinberg
Drexel University
Brian McManus
University of North Carolina-Chapel Hill
Yao Luo
University of Toronto
Marc Rysman
Boston University
JEL Classifications
  • L2 - Firm Objectives, Organization, and Behavior
  • L5 - Regulation and Industrial Policy