Are frequency and intensity of participation decision-bearing aspects of consumption? An analysis of drinking behaviour

Analysing quantity consumed per-period implies a restriction on decisions about how frequently and intensely to consume. This 'quantity-hypothesis' is rejected based on estimated equations for the frequency and intensity of spirit consumption. Consequently, a simultaneous equation system for frequency and intensity is estimated as they enter the budget constraint as a multiplicative term. Income and education levels are found to be negatively related to intensity but not frequency. This may reflect differences in the shadow prices of frequency and intensity for different socioeconomic groups or heterogeneous preferences for intoxication and health. Distinguishing between these aspects of drinking patterns is important for the evaluation of the price-responsiveness of harmful drinking, restrictions on availability and the causes of inequalities in health.