Premium
Quarterly Demands for Meat in Canada with Alternative Seasonality Hypotheses
Author(s) -
Hassan Zuhair A.,
Johnson S. R.
Publication year - 1983
Publication title -
canadian journal of agricultural economics/revue canadienne d'agroeconomie
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.505
H-Index - 37
eISSN - 1744-7976
pISSN - 0008-3976
DOI - 10.1111/j.1744-7976.1983.tb01052.x
Subject(s) - multicollinearity , seasonality , econometrics , sample (material) , seasonal adjustment , ordinary least squares , statistics , population , economics , mathematics , linear regression , variable (mathematics) , demography , sociology , mathematical analysis , chemistry , chromatography
It is common practice to use ordinary least squares with seasonal dummies in linear regression analysis of seasonally unadjusted time series data. The approach is based on a hypothesis of fixed and constant shifts in the intercept for each of the quarters (Jorgenson). This method is expensive in terms of potential multicollinearity problems and degrees of freedom (Green and Doll; Ladd; Madalla 1977). Also, the hypothesis that the quarterly effects are fixed may be overly restrictive, especially if the dummy variables are included simply to support the estimation of a structural relationship from the pooled quarterly sample partitions. This paper presents results on the structure of consumer demand for meats in Canada. These results, while of interest for their implications on the structure of consumer demand, are intended primarily to illustrate the consequences of alternative seasonality hypotheses. The sample is for the twelve year period, 1965 through 1976. Linear demand relationships are estimated using a number of alternative methods for accommodating seasonal effects. Results for the demand models with the alternative seasonality hypotheses are quite different, implying that researchers should exercise care in selecting a representation for seasonality that is consistent with the process generating the data, the theory, and the institutional setting governing the population from which the sample is drawn.