MEASURING THE IMPACT OF ADVERTISING AND PROMOTION: SINGLE-OR MULTI-EQUATION SYSTEMS? A CASE STUDY OF THE WASHINGTON APPLE INDUSTRY

This paper reports the results of a case study analyzing the impacts of promotion and advertising conducted by the Washington Apple Commission (WAC) using both single- and multi-equation systems. The analyses are compared in terms of the comparability and coherency of results and in terms of the types of information that can be generated by the multi-equation model but not by the single-equation model. Due to the availability of detailed data from the WAC, the multi-equation system incorporates measures of actual lines of print media within an explicit model of print media as well as a measure of the multiplier effect associated with WAC promotion expenditures relative to the overall (retail-matched) level of print media. This information cannot be effectively included in a single-equation analysis. The multi-equation system also includes supply-response and price linkages between F.O.B. and retail levels. Monthly data are used for all systems. The results of this study suggest that in terms of predicting the quantity effects of promotion expenditures, if one conditions on F.O.B. price levels the effects of non-trade activities from the single equation are not statistically significant and seasonality is not a major factor in determining monthly demand. Conversely, within the multi-equation approach, results of this study suggest that non-trade activities did have a measurable impact on the demand for Washington apples and that seasonality is a major factor in determining monthly demand. In terms of predicting the price effects of promotion expenditures, if one conditions on supply elasticities, the approaches provide similar predictions of price impacts. However, the simultaneous supply feedback is significant in the apple industry case, so, by taking the interaction of supply and demand into account, the more complex multiple-equation approach predicts notable supply response-induced mitigation in price increases originally induced by promotion efforts relative to when impacts of promotions are measured under scenarios of a fixed supply function. Actual cost estimates by type of analysis are also provided for each approach. The estimates are based on actual costs incurred in conducting the analyses. The results of the case study provide a basis for assessing the necessity, as well as the benefits and costs, of conducting a detailed and complex multi-equation promotion evaluation relative to a simpler and more focused single-equation analysis of promotion effectiveness.