CONTENTS

Coupon Redemption and Cheese Purchase Timing

Manager's Viewpoint

Why Economic Analysis in the Evaluation Strategy?

Editor's Notes

Director’s Corner

Selected Readings

Next Meeting

 

NEC-63 Spring 1996 May 26-27, 1996 Cancun, Mexico Agricultural Commodity Promotion Policies and Programs in the Global Agrifood System

Coupon Redemption and Cheese Purchase Timing

by Brian W. Gould

Household-level analyses of the impacts of commodity promotion programs typically hypothesize effects on both quantity and timing of product purchases. With commodity promotion there may be a direct positive impact on amount purchased and a concurrent reduction in the length of time between purchases. This reduction in interpurchase time is often referred to as “purchase acceleration.” In addition to the direct effects, commodity promotion efforts can also generate an indirect effect.

With households holding a given inventory, the reduction in interpurchase time implies smaller purchases per occasion. The possible conflict of the direct and indirect quantity impacts implies that unless shorter interpurchase time is recognized, comparing quantity consumed with and without product promotion may overestimate the true amount by which a particular promotion campaign increases demand. As a first step in investigating the relationships between purchase timing, quantity purchased, and commodity promotion, we focused on the dynamics of the consumer purchase process. We examined the effect of the use of one type of commodity promotion, coupon-based price deals, on interpurchase times for cheese, a frequently purchased food commodity.

We conducted this analysis using a data set which consists of a household panel observed over 170 consecutive weeks from April 1991 to June 1994. In addition to quantity and price data, we also used information on coupon usage and household demographic characteristics in the analysis. Only cheese purchased for at-home consumption was included in this analysis. On each purchase occasion a household in the panel recorded purchase date, UPC code, expenditures, and quantity purchased.

We used event history analysis to analyze purchases of four cheese types: all (excluding cottage), processed, natural cheddar, and cottage cheese. We used a variety of duration models in our analyses under alternative assumptions concerning the shape of the frequency distribution of the amount of time between purchase occasions (interpurchase time), the role of household characteristics in the determination of such distributions, and the presence of unobserved heterogeneity (variation) across households in these distributions.

Results

In our duration models, we hypothesized that the distribution of cheese-specific interpurchase times was affected by the use of coupon-based price deals as well as household demographic characteristics. In addition to the two variables used to capture the effect of coupon redemption on the timing of cheese purchases, we also included several variables as distribution “shifters”--household size, age composition of household members, ethnicity of main meal planner, ratio of household income to poverty threshold income, lagged purchases, whether or not there was a shelf-price change since last purchase, dummy variables identifying the Thanksgiving/Christmas period and summer months, and urbanization of residence area.

For all cheese, we found that over 18 percent of the purchase weeks occurred with the use of some type of cents-off coupon. This is similar to what we found for processed cheese. Only 3 percent of cottage cheese purchase occasions involved the use of some type of coupon. An examination of average interpurchase time indicated that there was some coupon-driven acceleration of purchases. For processed cheese there was a 1.9 week difference in mean interpurchase time for weeks associated with coupon use versus non-coupon use. This compared with .6 weeks for cheddar and 1.1 weeks for cottage cheese.

Using the results from the duration models applied to the four cheeses, we conducted likelihood ratio tests on the null hypotheses that the distribution of interpurchase time was not affected by household demographic characteristics. This hypothesis was rejected for all four cheese types. We also tested and rejected the null hypothesis that coupon redemption does not affect the length of time between purchases.

Using our model results, we simulated “survival” probability profiles for a variety of household types. These profiles represent the probability of a household not purchasing a particular cheese for alternative interpurchase times (weeks). Figure 1 presents an example of these profiles for processed cheese. Profiles are presented for coupon-using black, hispanic, “nonminority,” and single person households as well as non-coupon-using households. For processed cheese, non-coupon-using households have a 5-week survival probability of 25 percent. This compares to 34 percent and 32 percent for black and single-person coupon-using households, and 19 percent and 15 percent for “nonminority” and hispanic coupon-using households. For cheddar cheese, non-minority coupon-using and non-coupon-using households have very similar survival probability profiles, with black coupon-using households exhibiting the lowest survival probability profiles.

Summary and Areas for Future Research

The use of coupon-based incentive programs continues to be an important marketing tool. The present analysis investigated one facet of coupon usage, its impact on the timing of purchases of cheese, a frequently purchased, non-durable commodity. We estimated a series of duration models for four cheese classifications: all, processed, natural cheddar, and cottage cheese. Within these duration models, we incorporated household demographic characteristics to allow the distribution of interpurchase time to vary across households. The results of likelihood ratio tests indicated that the household characteristics included as distribution shifters were statistically significant explanatory factors.

Two variables were used to account for the impact of coupon redemption on interpurchase times. Likelihood ratio tests clearly rejected the null hypothesis that coupon use has no impact on the timing of cheese purchases. As hypothesized, the use of coupons resulted in reduced interpurchase times for all cheeses. This impact, however varied across cheese type, especially when considering the type of household doing the purchasing.

As previous analyses have shown, coupon promotion has direct impacts on purchase timing and quantity purchased,as well as indirect impacts on quantity purchased that may counteract the direct impacts. One fruitful area for future research would be the development a model which takes into account the simultaneous decisions of coupon use, interpurchase time, and quantity purchased. Previous analyses have not recognized the simultaneous nature of these consumer decisions. Our work provides a foundation for future research to analyze the net effect of coupon promotion on commodity demand and whether or not the costs of such promotion are justified.

Brian W. Gould is a Senior Scientist with the Wisconsin Center for Dairy Research and the Department of Agricultural Economics at the University of Wisconsin-Madison.

[ top ]