ESTIMATING THE IMPACT OF CONSUMER EXPECTATIONS OF COUPONS ON PURCHASEBEHAVIOR - A DYNAMIC STRUCTURAL MODEL

We examine the basic premise that consumers may anticipate future prom otions and adjust their purchase behavior accordingly. We develop a st ructural model of households who make purchase decisions to minimize t heir expenditure over a finite period. The model allows for future exp ectations of promotions to enter the purchase decision. Households wit h adequate inventory of the product may face a trade-off of buying in the current period with a coupon or defer the purchase until next peri od, given their expectations of future promotions. Thus, we provide a framework for examining the impact of consumer expectations on choice behavior. The target audiences for our paper are (a) empirical researc hers who intend to make structural models part of their applied resear ch agenda; and (b) managers who value and seek to understand the impac t of consumers' coupon expectations on current purchase behavior. Our research objective is to provide an empirical framework to examine whe ther and to what extent consumers anticipate future coupon promotions and adjust purchase behavior. The central premise of our approach is t hat a rational consumer minimizes the present discounted value of the cost of a purchase where cost in a single period consists of purchase price, inventory holding cost, gains from coupons, and potential stock out cost. We aim to test whether our hypotheses regarding the various elements of the cost structure are supported and that whether consumer s take into account future discounted cost when making current purchas e decisions. The research methodology we adopt is relatively new in ec onometrics and known as the estimable stochastic structural dynamic pr ogramming method. The methodology amounts to incorporating a maximum l ikelihood routine embedded in a dynamic programming problem. The dynam ic programming problem is solved several times within a maximum likeli hood iteration for each value of the state space elements and for each value of the parameters in the parameter set. The state space in our model consists of purchase and nonpurchase alternatives in each time p eriod, coupon availability and no coupon availability in each time per iod, level of inventory in each time period for each household, and co nsumption rate of each household. We use scanner panel data on purchas es in the disposable diaper product category and promotions. We estima te the inventory holding and stockout costs, brand-specific value of c oupons, and consumers' expectations of future coupons. The key insight s and lessons learned can be summarized as follows: (1) Our results ar e consistent with the notion that consumers hold beliefs about future coupons, and that such beliefs affect the purchase decision. We find t hat the dynamic optimization model performs significantly better than a single-period optimization model and a naive benchmark model. (2) We find a high and significant stockout cost, consistent with the essent ial nature of the product category. Our estimate of the holding cost y ields a reasonable annualized percentage value when converted to the c ost of capital. We find that consumer valuation of coupons differ mark edly across brands. (3) Our empirical evidence supports the notion tha t consumers hold beliefs about future coupon availability. We also fin d that the expectations about future coupons, estimated endogenously, differ depending upon whether or not a coupon was available in the cur rent period. Thus, the proposed model structure yields rich managerial insights and facilitates several ''what if'' scenarios. A possible li mitation of our model, and estimable structural models in general, is the computational cost. While it is possible to conceptually extend th e state space to accommodate variations across households and add a ri cher parameter structure, each addition multiplies the size of the sta te space and the computation time. For this reason, we have kept the s tate space as tight as possible and refrained from additions that woul d otherwise enable us to incorporate heterogeneity in consumer decisio ns. For example, we assumed that consumers are similar other than refl ected by their purchase behavior. We built a category purchase inciden ce model rather than a brand choice model. We refrained from including unobserved heterogeneity in the parameters. We chose to opt out of mo deling autocorrelation and other time-dependent error term patterns in the likelihood function. Thus, we have made an effort to build a stru ctural model that reasonably reflects consumer purchase behavior witho ut requiring expensive computation. Currently, there are developments in econometrics to approximate the computation of the valuation functi ons without sacrificing much accuracy. When these methods are well dev eloped we expect that structural models will become more commonplace i n marketing.