Food Safety, Habits, and Rational Expectations in U.S. Meat Demand

Abstract

The objective of this dissertation is to explore the theoretical and empirical implications of a meat demand model with rational habits. To introduce consumption dynamics, habit persistence is used to motivate intertemporally related preferences. The impact of food safety information on meat consumption is systematically analyzed. Important differences between myopic habits and rational habits are underscored. Theoretical predictions are tested using U.S. consumption data and food safety indices compiled based on articles from four major U.S. newspapers during the 1980(3)-2005(4) period.

Assuming rational expectations, Hansen's (1982) generalized method of moments (GMM) is implemented to investigate the Euler equations implied by a demand model with habits. Empirical evidence suggests that, at quarterly frequencies, habit persistence dominates inventory behavior in beef consumption during the post-1998 sample period, while pork and poultry demands exhibit mild degrees of inventory adjustment overall for this period. A plausible explanation for the dominance of habits in beef demand since 1998 is that the low carb-high protein fad may have helped to increase the degree of habits for beef. This is consistent with the theoretical prediction that an increase in the perceived benefits of long-term consumption of a good is likely to strengthen the degree of habit persistence of that good. With GMM estimates of preference parameters at hand, demand elasticities to price and food safety shocks that are expected to be transitory or permanent can be computed. The standard procedure is to linearize the Euler equations and derive analytical results for elasticities. Nevertheless, linearization is not attempted here. Instead, a numerical procedure is invoked to compute approximate solutions to the Euler equation under various shock senarios. The simulated elasticities are sensible and consistent with theoretical predictions. Specifically, using 2004(4) prices and food safety levels as benchmark values, the short-run own-price elasticity of beef demand for a permanent price increase is more than twice of its counterpart for a price increase that is expected to be transitory. The long-run ownprice elasticity of beef for a permanent price change is about 30% higher than the short-run elasticity. The food safety elasticities are generally small in magnitude with the long-run own-effect elasticity on beef of about -0.015 being the largest.

A model for food safety policy analysis should be closely tied to the objectives of consumers and to the decision rules of policy makers, and should explicitly embed expectations that are most relevant for consumers when consumption is temporally dependent. Although some reduced-form dynamic models of meat demand have been developed to capture dynamic interactions among demands over time, these are unlikely to be adequate in modeling expectations. Food safety shocks that were largely transitory in the past may be perceived by consumers as being more persistent in nature if happen again in the future, and vice versa. This is the essense of the classic Lucas (1976) critique applied to food safety analysis.

The distinct own-price elasticities of beef demand for price changes of different durations provide corroborating evidence that the Lucas critique can be quantitatively important in modeling meat demand.