Understanding market demand for agricultural products through consumer research : the coffee example
A theoretical model of coffee consumption in the U.K. is proposed, which is estimated and used to examine the influence of habit formation and advertising in the period 1957-80. This work challenges both the assumption of symmetrical consumer response and the statistical source for measuring coffee consumption. The model allows for asymmetrical consumer reactions. Explanatory variables are: price of coffee and of tea, income, advertising and the strength of the coffee drinking habit. This work is original in terms of interpreting and quantifying product field advertising and habit formation; and for allowing a minimal threshold level of predictors. Mistakes, repeated printing errors and unpublished changes in definitions were found in the statistics of domestic coffee supplies 0 Household coffee purchases estimated by the National Food Survey (N.F.S.) are consistently over-reported. Causes investigated provide grounds for correcting estimates by pooling N.F.S. with Family Expenditure Survey; the result is consistent with adjusted supplies. Advertising effect on demand is separated into two aspects. The first action increases sales by attracting new buyers, while protecting consumers from competitors' propaganda. The second action increases sales to habitual customers, while manufacturers are competing through advertising for a larger brand share. The transmission medium is a factor in both effects. The strength of the habit shifts market demand function. A routine way of thinking prevails under stationary conditions; yet shifts in the function occur in a non-stationary situation which initiates the problem-solving way of thinking. In the model, addiction can be either absolute or relative to changes in other factors. All the evidence supports the general model proposed, which shows that a non-symmetrical functional effect prevails and demonstrates the existence of an adjustment period. Irrefutably, coffee consumption depends on former consumption levels, coffee price, price-ratio tea to coffee, income and advertising.