is the application of statistical methodology to economics. Used
in virtually every area of economics, econometrics is often a second
field for people doing empirical research. The Triangle area
includes large number of econometricians covering a wide range of
theoretical and applied interests in the field. The Econometrics
Workshop is run jointly by the University of North Carolina - Chapel
Hill, Duke University, North Carolina State University, and the SAS
Institute. The group also holds the annual Harold Hotelling
Triangle Economics Conference.
is Henry A. Latane Distinguished Professor of Economics and Adjunct
Professor of Statistics at the University of North Carolina at Chapel
Hill and Research Professor of Economics at Duke University. Before
joining the UNC faculty, he was, successively, Assistant, Associate,
Full, and Drexel Professor of Statistics and Economics at North Carolina
State University. Gallant has held visiting positions at the University
of Chicago, Duke University, and Northwestern University. He received
his A.B. in mathematics from San Diego State University, his M.B.A.
in marketing from the University of California at Los Angeles, and
his Ph.D. in statistics from Iowa State University. He is a Fellow
of both the Econometrics Society and the American Statistical Association.
He serves on the Board of Directors of the National Bureau of Economic
Research and has served on the Board of Directors of the American
Statistical Association and on the Board of Trustees of the National
Institute of Statistical Sciences. He is co-editor of the Journal
of Econometrics and past editor of The Journal of Business
and Economic Statistics.
interested in fitting models from the sciences to data for the purpose
of statistical inference. Typically these models will involve a nonlinear
parametric component that describes features of the model where the
underlying scientific theory is explicit and a nonparametric component
that accounts for features where the scientific theory is vague. Appropriate
statistical methods for these problems are usually computationally
intensive. Methodological interests are in developing statistical
methods and numerical algorithms for fitting these models. Theoretical
interests are in deriving the statistical properties of proposed methods,
particularly the asymptotic properties of estimators of functionals
of the nonparametric component. Applied interests are primarily within
economics and finance.
Professor and Chair of Economics, is an econometrician interested
in statistical problems associated with large survey data sets. His
current work involves the development of estimation methods to predict
outcomes such as contraceptive choice and recent fertility. His primary
measurement focus is on correctly estimating the effects of external
factors such as resource inputs on ultimate outcomes. He is working
Mroz and Gustavo Angeles to extend economic structural models
for use in the study of the determinants of fertility when government
program placement is targeted. They are applying their methods to
data sets from Peru and Tanzania and presented a paper on this topic
at the 1998 PAA meeting. Guilkey is currently on the Carolina Population
Center Advisory Council and is one of the CPC statistical advisors.
(Yale (1980, Economics) and University of Washington (1973, Mathematics))
is currently working on his third contribution to econometrics.
"What is Fractional Integration," Review of Economics
and Statistics (forthcoming, 1999), opens a discussion of the
relation between fractional integration, which is a generalization
of the notion of a unit root, and structural change. The former
has to this point been mainly a technical econometric concept and
the latter has been primarily a matter of substantial practical importance.
Parke introduces a new construction, known as an error duration model,
that is an alternative to ARMA models for a certain class of time
series processes. A fractionally integrated process has a surprisingly
elegant representation as an error duration model. Parke
shows that fractional integration and structural change are essentially
two views of one process. This paper is available in pdf format from
the Department's working
Algorithm for FIML and 3SLS Estimation of Large Nonlinear Models,"
Econometrica 50 (1982), 81-95, provides a solution to the longstanding
numerical challenge of FIML estimation of large simultaneous equations
models. Related papers (Journal of Econometrics (1980),
Annals of Statistics
(1986), International Economic Review (1986)) demonstrate the
algorithm and investigation theoretical issues related to techniques
for estimating subvectors of a large parameter vector. Parke's
FIML algorithm and other econometrics for large simultaneous equations
models are in the Fair-Parke Program, which you can find at fairmodel.econ.yale.edu/fp/fp.htm.
Likelihood Based-Prediction Functions" (with Thomas F. Cooley)
Econometrica, 58 (1990), 1215-1234, and related papers
(Journal of Econometrics (1987, 1989), propose a theory of predictive likelihood that complements the
notion of likelihood in estimation theory. Existing forecasting
methodologies relied either on linear models or simulation techniques
to contruct predictive confidence intervals. The predictive
likelihood approach constructs a predictive likelihood function for
general nonlinear models.
topics have included empirical work on stock price volatility (American
Economic Review, 1992) and the pre-WWI gold standard era (Journal
of Money, Credit, and Banking, 1995). Several of these
papers are available at www.jstor.org.
Parke is also the architect of the EconModel project at www.econmodel.com.
is interested in the estimation of continuous time models for financial
data. A working paper discussing numerical techniques for implementing
a simulation-based maximum likelihood estimator for scalar, fully-observed
stochastic differential equations is available at www.unc.edu/~gdurham.
Future work will extend these results to multivariate and partially
is developing extensions of Parke's error duration ideas. His
results help to show the relation between fractional integration and
common notions of structural change.
Bollerslev, Hall, Dickey, ...
Introduction to Econometric Theory
theory, expectation, conditional expectation, modes of convergence,
limit and interchange theorems, and the asymptotics of maximum
likelihood, generalized method of moments, and efficient method
of moments. Fall. Gallant, Parke.
Economics 271 or equivalent. One semester coverage of basic econometrics.
Topics include: regression under ideal and nonideal conditions;
special models, including simultaneous equations models; and applications
and econometric computer programs. Spring. Guilkey, Mroz, Parke.
Economics 271 and Mathematics 147. Economics 273 constitutes a
one-semester treatment of the fundamental theory of econometrics.
Topics covered include asymptotic distribution theory, linear
and nonlinear models, specification testing techniques, and simultaneous
equations models. Fall. Guilkey, Parke.
Time Series Econometrics
Economics 273. Covers stationary univariate and multivariate time
series models, spectral analysis methods, nonstationary models
with time trends, unit roots and cointegration, and special topics
such as conditional volatility, the Kalman filter and changes
of regime. Spring. Gallant, Parke.
Nonlinear Econometric Methods
Economics 273. Density estimation, nonparametric regression, neural
nets, nonlinear regression, generalized method of moments, seminonparametric
time series, estimating stochastic differential equations and
nonlinear latent variables. Fall or Spring. Gallant.
Cross Sectional Econometrics
Economics 273. Maximum likelihood methods for limited dependent
variables. Longitudinal data models and methods. Hazard models.
Multivariate models with limited dependent variables. Fall or
Spring. Guilkey, Mroz.
Research in Econometrics
Economics 273 and Economics 274, 275, or 276. Seminar on special
topics in econometrics. Spring. Gallant, Guilkey, Parke.