Files
Abstract
There are a number of estimation methods available
to the econometrician. They range in complexity
from least squares, which takes only a few minutes
with pencil and paper for a simple problem, to full
information maximum likelihood, which requires
a large-capacity electronic computer for effective
application. A list of the common methods, given
in order of increasing complexity, includes least
squares, two-stage least squares, instrumental variables,
limited information, and full information.
Recursive systems are fitted by least squares which
makes estimation relatively simple, and they have
other desirable statistical properties. Following a
brief discusssion of the special characteristics of recursive
systems, this paper presents an application of
the recursive method to an analysis of the hog industry.
The author is indebted to Anthony S. Rojko
and Hyman Weingarten of the Economic Research
Service for their helpful suggestions during the
preparation of this paper.