Consumption of lowfat and skim milk has increased substantially over the past decade. This study investigates whether the change is due to price and expenditure effects or to a more fundamental preference change in milk demand. Parametric and nonparametric analytical approaches provide a comprehensive analysis of structural change in milk consumption in New York State. A nonparametric approach first finds evidence of structural change. A parametric likelihood-ratio test then confirms the existence of structural change using a Kalman filter specification. The value of this technical analysis of milk preferences is its implication for labeling initiatives. Milk fat labels have allowed consumers to act on a new set of preferences, thereby improving consumer welfare.

“Commodity promotion” consists of many activities, each designed to contribute to a consumer's product knowledge or influence tastes. However, both knowledge and tastes are unobservable, or latent, variables influencing demand. This paper specifies a dynamic structural model of fresh fruit demand that treats promotion and other socioeconomic variables as “causal” variables influencing these latent variables. Estimating this state-space model using a Kalman filter approach provides estimates of both the system parameters and a latent variable series. The results show that these latent effects contribute positively to apple and other fruit consumption, while reducing banana consumption.

Several recent papers have used annual changes and monthly data to estimate demand systems. Such use of overlapping data introduces a moving average error term. This paper shows how to obtain consistent and asymptotically efficient estimates of a demand system using seasonally differenced data. Monte Carlo simulations and an empirical application to the estimation of the U.S. meat demand are used to compare the proposed estimator with alternative estimators. Once the correct estimator is used, there is no advantage to using overlapping data in estimating a demand system.

We propose a regime-switching model that allows demand to respond asymmetrically to upward and downward advertising changes. With the introduction of a smooth transition function, the model features smooth rather than abrupt parameter changes between regimes. We apply the model to nonalcoholic beverage data in the United States for 1974 through 2005 to investigate asymmetric advertising response. Results indicate that a decrease in milk advertising had a more profound impact on milk demand than an increase did. An increase in milk advertising had no impact on milk demand, but a decrease could have an own-advertising elasticity up to 0.049.

Using U.K. data that are consistent with utility maximizing behavior by consumers, we construct aggregates for both consumption goods (nondurables and services) and monetary assets that are consistent with economic aggregation theory. Using these aggregates and the stock of durable goods, we estimate the elasticities of substitution between various consumption goods and monetary assets. These estimates are compared to the corresponding results from conventional monetary and consumption aggregates. The results give important information for monetary policy and the monetary transmission mechanism. In particular, these substitution estimates provide insight into the recent changes in expenditure on durable goods. We also show that the use of conventional U.K. consumption and monetary aggregates often give incorrect estimates of substitution and can provide misleading policy insights.