Matthew T. Holt and Andrew M. McKenzie, "Quasi-Rational and Ex Ante Price
Expectations in Commodity Supply Models: An Empirical Analysis of the U.S.
Broiler Market", Journal of Applied Econometrics, Vol. 18, No. 4, 2003,
pp. 407-426.

The bi-monthly time series data that were used in this study are contained
in the file broiled3.dat, an ASCII file in DOS format that is zipped in the
file hmquasi.zip.  Also included in the hmquasi.zip file is the file
broild3.for, also in DOS format, that contains the basic FORTRAN code used,
in conjunction with GQOPT, to estimate the parameters of the preferred
model.

The data contained in broild3.dat consist of 177 bi-monthly observations
starting with the third bi-monthly period in 1966 and ending with the fifth
bi-monthly period in 1995.  There are 21 variables.  With the exception of
dummy variables and the year/bi-month indicator, all variables represent
natural logarithms, multiplied by 100.  In estimation the first ten and the
last two observations are omitted to initialize data lags and leads.  The
data set broild3.dat is organized according to the list below.

Column		Series

1.              year/bimonth (i.e., 6702 is the first bi-month of 1967, etc.)
2.              cattle futures price
3.              cattle spot price
4.              hog futures price
5.              hog spot price
6.              corn futures price
7.              corn spot price
8.              soybean meal futures price
9.              soybean meal spot price
10.             iced broiler futures price
11.             twelve-city broiler wholesale price
12.             Georgia dock wholesale broiler price
13.             dummy variable for bi-month one
14.             dummy variable for bi-month two
15.             dummy variable for bi-month three
16.             dummy variable for bi-month four
17.             dummy variable for bi-month five
18.             hatch of broiler type chicks
19.             young chicken, ready to cook production
20.             turkey spot price
21.             linear time trend

The feed cost variable used in the model is constructed according to the
formula fc=ln(70*cs/56)+(30.*sms/20))*100, where cs denotes the corn spot
price and sms the soybean meal spot price.  Additional details regarding the
data may be found in section 3 of the article.


Matthew T. Holt
Department of Economics
Box 8110
North Carolina State University
Raleigh, NC  27695-8110

e-mail:	Matt_Holt@ncsu.edu