Review of
An
Introduction to the Structural Econometrics of Auction
Data
By Harry J. Paarsch and Han Hong
with contributions by M. Ryan Haley
Cambridge, MA: The MIT Press, 2006.
By
Tong Li
Department of Economics
Vanderbilt University
This is the first book that gives a comprehensive introduction to the
structural econometrics of auction data (SEAD). This book is well structured,
carefully organized, and nicely written. What is now a vast literature
covering the structural analysis of auction data started almost fifteen years
ago with the seminal paper (Paarsch (1992)), written by one of the book's
authors. This subject is an exciting area of empirical industrial organization
in which researchers integrate economic theory, econometric techniques, and
data analysis to conduct rigorous economic analysis and to address policy
related issues. As someone who has been working in this area and advising
students on the subject, I have felt a strong need to have a textbook that
gives an effective coverage of the development of the field, offers insight on
the essence of the approach, and brings together the economic theory and
econometric tools to the analysis of auction data. This book without doubt
fills in the gap and will become an excellent source for researchers and
graduate students who are interested in this literature.
Analyzing auction data using the structural approach requires a special
integration of mathematics, economic theory, and econometric tools. The
authors have done a remarkable job of meeting the challenges that result from
needing such a broad toolset by introducing economic theory with examples,
presenting tools from recent developments in theoretical econometrics combined
with numerical methods, and providing a practical guide to the main concepts
and approaches in the field. The rapid and on-going development of the field
over the last fifteen years presents another challenge to the authors of this
book. I commend the authors' thoughtful organization of the book. While
focusing on the structural approach, that is to estimate an econometric model
closely derived from game-theoretic models, they restrict themselves to the
independent private value paradigm (IPVP), the simplest information
environment. This scope restriction allows the authors to offer more insight
on the approach, and to illustrate empirical methods in a relatively
straightforward way with a minimum of mathematics.
In the introduction the authors present an IPVP example to show how the
aggregate demand is connected to the distribution of the bidders' private
values, bringing up the central message that the essence of the SEAD is to
identify this distribution in order to estimate the demand function. This
objective also makes the analysis of auction data different from the standard
demand analysis because of the strategic behavior of bidders with incomplete
information. Then they depict how different auction formats generate
different information settings, and thus create different empirical problems.
They conclude the introduction by laying out the plan for the book and giving
several practical problems that are related to the basic probability theory,
statistical estimation and inference.
In the second chapter the authors provide an overview of auction theory with a
focus on single object auctions. The chapter covers the four most-studied
auction formats; the oral, ascending-price (English) auction; the first-price
sealed-bid auction; the oral, descending-price (Dutch) auction; and the
second-price, sealed-bid (Vickrey) auction. For each auction, the authors
derive the equilibrium-bidding function for risk-neutral or risk-averse
bidders. Then they discuss the revenue equivalence proposition, one of the
most intriguing results in auction theory, first demonstrated in the seminal
Vickrey (1961). Subsequently, the authors derive the optimal selling
mechanism based on the analysis of Riley and Samuelson (1981), which is
equivalent to deriving the optimal reserve price in an IPVP. While the book
focuses on the IPVP throughout, this chapter also covers two other important
information structures, the common value paradigm (CVP) and the affiliated
value paradigm (AVP). This chapter introduces asymmetry and risk aversion and
shows how these additional complications can change the standard results such
as the revenue equivalence proposition and optimal selling mechanism.
In Chapter 3 the authors discuss the SEAD in English auctions and Vickrey
auctions. Starting the introduction of the structural approach with these two
formats gives readers more insight and intuition on the essence of the
approach because they yield simple equilibrium bidding functions with trivial
relationships between observed bids and private values. Then they show how to
introduce covariates to control for observed auction heterogeneity. The
authors also present an application by Paarsch (1997) to illustrate how
structural analysis can be used to address policy related issues. They
discuss the work of Haile and Tamer (2003) that deals with incomplete data
arising from bid increment and jump bidding in English auctions. They
conclude by introducing asymmetric bidders and outlining a nonparametric
identification result established by Brendstrup and Paarsch (forthcoming).
In chapter 4 the authors focus on the structural approach to first-price
sealed-bid auctions and Dutch auctions, in which the equilibrium bidding
functions are no longer simple, but highly nonlinear functions of bidders'
valuations. That the equilibrium bidding functions are nonlinear in the
bidders' valuations raises numerous econometric issues. In this chapter the
authors discuss these issues intuitively and in great detail. They begin with
Paarsch (1989) and then present the nonparametric identification results by
Guerre, Perrigne and Vuong (2000) as well as the resulting nonparametric
estimation method. They also discuss how to extend the method to the case
where reserve prices are binding. Then they introduce various parametric
methods to estimate the structural model, including the simulated nonlinear
least squares of Laffont, Ossard and Vuong (1995), the maximum likelihood
estimator of Donald and Paarsch (1993, 1996), and the extreme-order,
generalized method of moments estimators of Donald and Paarsch (2002). Then
they consider various extensions of the standard IPV model to risk averse
bidders by Compo, Guerre, Perrigne, and Vuong (2000) and to bidders with
stochastic private values by Lu (2004). In great detail, they discuss
Krasnokutskaya (2004), which covers asymmetric bidding with unobserved auction
heterogeneity, and Li (2005), which covers auctions with endogenous entry.
Lastly, they present Brendstrup and Paarsch (forthcoming), which uses data
from fish auctions in Denmark to compare the expected revenues of the English
and Dutch auctions with asymmetric bidders.
Chapter 5 is devoted to the structural analysis of multi-unit auctions, in
which several units of the same object are sold to bidders. The theory of
multi-unit auctions is not as well developed as that of single-unit auctions,
and the structural analysis of multi-unit auctions is quite limited. The
authors have done a good job in highlighting the key theoretical results, and
presenting some important structural work in the area. After presenting the
Weber (1983) classification of multi-unit auctions into simultaneous-dependent
auctions, simultaneous-independent auctions, and sequential auctions; the
authors describe the uniform- and discriminatory-pricing rules. Then they
provide a detailed treatment of Ausubel and Cramton (2002), which contains a
result that belies the conventional wisdom. Ausubel and Cramton (2002) show
that bidders have an incentive to shade their bids, and that they do so
relatively more at low values than at high values. Ausubel and Cramton (2002)
also show that this behavior, refereed to as demand reduction, even holds
under the uniform-pricing rule. The authors also discuss the Ausubel and
Cramton (2002) efficiency results. Then they introduce the singleton demand
and multi-unit demand problems, and present Donald, Paarsch, and Robert
(forthcoming) that constructs a theoretic model of the multi-unit, sequential,
Milgrom-Weber clock auction with random demand, which was developed to analyze
the data of export permits for timber held in the Krasnoyarsk Region of
Russia. The chapter concludes with the presentation of the work by Brendstrup
and Paarsch (forthcoming) that considers a sequential clock auction with
nonrandom demand and establishes the nonparametric identification.
In chapter 6 the authors discuss some directions for further research. They
list three topics that are worth pursuing; simultaneous, dependent, and
sealed-bid auctions (e.g., Hortaçsu (2002a, b) and Chapman, McAdams, and
Paarsch (2005)); repeated games (e.g. Jofre-Bonet and Pesendorfer (2003)) and
multi-object auctions (e.g. Brendstrup and Paarsch (2005)). They also point
out that extending the methods described in the book to the CVP and the AVP
can be fruitful but more difficult than the other directions.
The Appendix covers basic statistical concepts and background material as well
as computational methods and tools that are useful in the SEAD. The problem
sets at the end of each chapter are thoughtfully designed and well balanced
between theory and empirical applications. Another nice feature of the book is
that it is accompanied by a CD containing sample data sets, providing readers
with an opportunity to conduct empirical analysis on their own.
While this book is suitable for a broad topic course in empirical industrial
organization, it can also be used to offer a graduate course in SEAD. SEAD is
a fast growing field. This book can serve as a standard textbook for the
course where extensions and new developments can be covered as well.
References
Ausubel, L.M., Cramton, P. (2002). Demand reduction and inefficiency in
multi-unit auctions. Mimeo, University of Maryland.
Brendstrup, B., Paarsch, H.J. (2005). Semiparametric estimation in models of
multi-object, sequential, English auctions. Mimeo, University of Iowa.
Brendstrup, B., Paarsch, H.J. Identification and estimation in sequential,
asymmetric, English auctions. Journal of Econometrics, forthcoming.
Campo, S., Guerre, E., Perrigne, I., Vuong, Q. (2000). Semiparametric
estimation of first-price auctions with risk averse bidders. Mimeo, University
of Southern California.
Chapmen, J.T.E., McAdams, D. Paarsch, H.J. (2005). Sequential, multi-unit,
sealed-bid, discriminatory-price auctions. Mimeo, University of Iowa.
Donald, S.G., Paarsch, H.J. (1993). Piecewise pseudo-maximum likelihood
estimation in empirical models of auctions. International Economic
Review 34:121-148.
Donald, S.G., Paarsch, H.J. (1996). Identification, estimation, and testing in
parametric empirical models of auctions within the independent private values
paradigm. Econometric Theory 12:517-567.
Donald, S.G., Paarsch, H.J. (2002). Superconsistent estimation and inference
in structural econometric models using extreme order statistics. Journal of
Econometrics 109:305-340.
Donald, S.G., Paarsch, H.J., Robert, J. An empirical model of the multi-unit,
sequential, clock auction. Journal of Applied Econometrics,
forthcoming.
Guerre, E., Perrigne, I., Vuong, Q. (2000). Optimal nonparametric estimation
of first price auctions. Econometrica 68:525-574.
Haile, P.A., Tamer, E. (2003). Inference with an incomplete model of English
auctions. Journal of Political Economy 111:1-51.
Hortaçsu, A. (2002a). Mechanism choice and strategic bidding in divisible good
auctions: an empirical analysis of the Turkish treasury auction market. Mimeo,
University of Chicago.
Hortaçsu, A. (2002b). Bidding behavior in divisible good auctions: theory and
evidence from the Turkish treasury auction market. Mimeo, University of
Chicago.
Jofre-Bonet, M., Pesendorfer, M. (2003). Estimation of a dynamic auction game.
Econometrica 71:1443-1489.
Krasnokutskaya, E. (2004). Identification and estimation in highway
procurement contracts under unobserved auction heterogeneity. Mimeo,
University of Pennsylvania.
Laffont, J.-J., Ossard, H., Vuong, Q. (1995). Econometrics of first-price
auctions. Econometrica 63:953-980.
Li, T. (2005). Econometrics of first-price auctions with entry and binding
reservation prices. Journal of Econometrics 126:173-200.
Lu, J. (2004). Stochastic private values in auctions: identification and
estimation. Mimeo, University of Southern California.
Paarsch, H. J. (1989). Empirical models of auctions and an application to
British Columbian timber sales. Mimeo, University of British Columbia.
Paarsch, H. J. (1992). Deciding between the common and private value paradigms
in empirical models of auctions. Journal of Econometrics 51:191-215.
Riley, J. G., Samuelson, W. F. (1981). Optimal auctions. American Economic
Review 71:381-392.
Vickrey, W. S. (1961). Counterspeculation, auctions, and competitive sealed
tenders. Journal of Finance 16:8-37.
Weber, R. J. (1983). Multiple-object auctions. In: Auctions, Bidding, and
Contracting: Uses and Theory, Engelbrecht-Wiggans, R., Shubik, M., Stark,
R.M., eds, New York: New York University Press.