Special Session Summary Does Holding on to a Product Result in Increased Consumption Rates?

Dilip Soman, University of Colorado at Boulder
[ to cite ]:
Dilip Soman (1997) ,"Special Session Summary Does Holding on to a Product Result in Increased Consumption Rates?", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 33-35.

Advances in Consumer Research Volume 24, 1997    Pages 33-35



Dilip Soman, University of Colorado at Boulder

Consumer behavior research has focused largely on the act of choosing between products or on events that are antecedent to the purchase occasion. The actual product usage occasion has received relatively little attention (but see Folkes, Martin and Gupta 1993, Wansink 1996). One fundamental question relates to how consumers evaluate products at the time of consumption and consequently how willing they are to consume them. An examination of this evaluation process is important at two levels. At the theoretical level, it provides an understanding of how consumers evaluate transactions and consumption experiences in general, and the role of time on these evaluations in particular. At a practical level, this translates into an understanding of how delayed consumption and bulk buying might increase consumption rates, category sales and profits. The three papers in this special session focused on the effect of the temporal delay between purchase and consumption and the supply quantity on consumption rates. The papers address different aspects, use different kinds of data and methods of analysis and use diverse paradigms in studying these issues.

Consider a consumer making a routine cash and carry purchase where she gives up a quantity of money in exchange for goods or services. Research on Prospect Theory and Mental Accounting (Thaler 1985) and the principles of hedonic editing (Thaler and Johnson 1990) would suggest that consumers will attempt to integrate the costs and benefits of the transaction and hence will not experience a separate "loss" at the time of payment and a "gain" on receipt of the product benefit.

However, as the costs and benefits associated with a transaction become separated in time, Gourville and Soman (1996) argue that the hedonic impact of the upstream cost diminishes. Consequently, the payment is "depreciated" and it results in the reduced valuation of the downstream benefits. Since the impact of the cost decreases over time, the benefits take on the nature of "pure" or "windfall" gains. If this happens, it would result in increased consumption rates and greater willingness to forego the benefits. Papers presented in this session examined different aspects of this process.



Eldar Shafir, Princeton University

Richard Thaler, University of Chicago

In this paper presented by Richard Thaler, the authors explored situations in which consumers had to buy objects that could only be consumed properly at a later date (e.g. wine, concert tickets). Their presentation attempted to investigate the mental accounting processes that occur in such circumstances and asked the following specific questions: First, when the consumer buys the good in advance, is this act considered "spending" or "investing"? Then, when the consumption takes place, does the consumer feel as if she is spending or simply drawing down some asset? Does the accounting in the latter case depend on the market value of the good in question, or whether the item is consumed or simply lost?

Thaler presented results of the following questions posed to the subscribers of a wine auction newsletter:

1. Suppose you bought a case of a good 1982 Bordeaux in the futures market for $20 a bottle. The wine now sells at auction for about $75 a bottle. You have decided to give one bottle of this wine to a friend as a gift. Which of the following best captures your feeling of the cost to you of giving away this bottle? (Check one).

a) Giving away the bottle does not feel like it costs me anything. I paid for the bottle already, many years ago, and probably don’t remember exactly what I paid for it anyway. _____

b) Giving away the bottle fees like it costs me $20, the amount I roughly remember paying for it. _____

c) Giving away the bottle feels like it costs me $20, the amount I originally paid for it, plus whatever the interest would have been on the money I paid. _____

d) Giving away the bottle feels like it costs me $75, the amount it would take to replace it. _____

e) Giving away the bottle feels like I am saving $55, because I am able to give a $75 gift for which I only paid $20. _____

2. Which answer would you choose if instead of giving the bottle away, you dropped the bottle and broke it? How much would it feel like you had lost in this case?

(a) $0 _____

(b) $20 _____

(c) $20 plus interest _____

(d) $75 _____

(e) a $55 saving (relative to a bottle bought recently) _____

Responses to these questions revealed that consumers, even sophisticated wine collectors, adopt the convenient mental accounting fiction that they are investing when they make their initial purchase but that they are not spending money when they eventually drink the bottle. Few respondents, other than economists, paid much attention to the market value of the good in question.



John Gourville, Harvard University

Dilip Soman, University of Colorado at Boulder

In the second presentation John Gourville and Dilip Soman presented their theory of payment depreciation and transaction decoupling. Under the label "Transaction Depreciation", they proposed that as the time between cost incursion and benefit consumption increases for any given transaction, the downstream benefit will increasingly take on the look and feel of a "windfall" gain. As a result, they predicted that individuals will be willing to consume such a product more readily, lend such a product more freely, or forego its benefits more willingly.

This presentation built on the sunk cost research of Thaler (1985) and others (Arkes & Blumer 1985). To highlight the sunk cost effect, Thaler argues that a family will be more willing to drive thorough a snowstorm to attend a basketball game had they purchased tickets to that game than had they been given (free) tickets. He attributes this difference in behavior to the desire to psychologically integrate the costs and benefits of a consumer transaction, resulting in the unused tickets being perceived as a "pure loss" when paid for but as a "foregone gain" when free. Gourville and Soman proposed that this sunk cost effect is moderated by time, with upstream costs becoming less salient and increasingly "depreciated" at the consumption occasion as the time between costs and benefits increases. As a result, the downstream benefits increasingly take on the characteristics of a pure or "windfall" gain. Gourville presented the results of several of their studies to support this argument.

In one study, they showed that individuals were significantly more likely to use their car under adverse conditions when they had fully paid for that car three years prior than when they had financed that car over the past three years with payments only recently completed. In a second study, they found theater-goers to be significantly more likely to miss a show due to illness if tickets for that show had been purchased six-months in advance as opposed to one-day in advance. Further, they found that the likelihood of missing the theater event due to illness was the highest when the tickets were free, and that the passage of time did not significantly affect the likelihood in the free ticket scenarios.

Consumers may also undervalue the benefit if they are not able to correctly assign a cost to each unit of consumption. This is referred to as "Transaction Decoupling", and can happen through packaging (large multiple unit pack sizes vs. small unit pack sizes), physical format of the transaction (2 liter bottles of soda vs. six packs, season ticket in the form of a pass vs. a booklet of tickets) or bundled pricing. They showed that ski vacationers are more likely to forego a day of skiing (and demanded lesser compensation if the weather was bad) if their season ticket was in the form of a pass as compared to when it is in the form of a booklet of tickets. In the case of bulk purchasing, both "Payment Depreciation" and "Trasaction Decoupling" would act to reduce valuation of the product leading to the prediction that it would be consumed more readily.



Kusum Ailawadi, Dartmouth College

Scott Neslin, Dartmouth College

Ailawadi and Neslin’s presentation used scanner data to provide empirical evidence for the increased consumption rates that arise from promotion-induced stockpiling of certain product categories. This paper identified two mechanisms-fewer stockouts and faster usage-by which promotion can increase category demand. Both stem from the power of promotions to induce consumers to buy more of the product category either by increasing the number of purchase occasions or by encouraging larger purchase quantities per purchase occasion. The use-faster mechanism is examined by modeling weekly consumption as a function of inventory levels and incorporating this into a purchase incidence model. The usage rate model requires only one parameter to be estimated. This parameter measures the degree to which consumers are willing to change their usage rate depending on their inventory levels. Ailawadi discussed two product categories, yogurt and ketchup, across which the authors expected the consumption effect of promotion to differ substantially.

The model was estimated via maximum likelihood procedures. An iterative grid search was used to obtain the usage rate parameter that maximizes the log-likelihood of the purchase incidence model. Results showed that the fit of the Ailawadi and Neslin model was significantly better than a status quo model using a constant usage rate. Further, the estimates of the usage rate parameter for very different for the two product categories, thus providing strong discriminant validity for the model. As hypothesized, usage rate for yogurt was much more flexible than that for ketchup.

A Monte Carlo simulation of the model was used to quantify the total consumption effect. The simulation provided both face and discriminant validity for the results. In the yogurt category, the simulation illustrated how promotion can increase consumption by loading consumers up with product that they consume quickly. In the ketchup category, a smaller increase in consumption was shown to be distributed over a much longer period of time. In addition, the simulation showed how the model can be used to decompose the long-run effects of promotion into brand switching versus increased consumption.




Richard Thaler, University of Chicago

This session dealt with the general issue of the evaluation of the product at the time of consumption and the resultant "flexibility" in the consumption rate. The first two papers raised the notion that the consumption rate of a product may be related to the "price tag" (or value) that the consumer attaches to a unit of consumption at the time of consumption. This value is a function of how salient the cost is at the time of consumption. The third paper used real world data to show that usage rates vary with inventory holding, at least for some product categories, therefore promotion can expand category demand by inducing consumers to stockpile the product. The three papers in this session examine different aspects of this underlying phenomenon.

Richard Thaler, the synthesizer for the session, stressed the importance of consumption research in marketing. He pointed out the similarities among the three papers and called for further research that provided more "real world" evidence of increased consumption rate. Thaler spoke about the sales of ski passes and how it affected consumption rates. Consistent with some results presented earlier in the session, he prposed that if ski passes were sold in advance as a package, consumers were likely to purchase more tickets than they would use and eventually not feel bad about the unused (and hence wasted) tickets. He also gave the example of vacation homes as a purchase where some of the phenomena raised in the session were particularly applicable. Finally he engaged the audience in a discussion with the authors. Issues that came up in the audience discussion included comments about the importance of consumption research, specific comments on the research reported int he three papers and other areas of application.


Arkes, Hal and Catherine Blumer (1985), "The Psychology of Sunk Cost, " Organizational Behavior & Human Decision Processes, 35(1), 124-140.

Folkes, Valerie S., Ingrid M. Martin and Kamal Gupta (1993), "When to Say When: Effects of Supply on Usage," Journal of Consumer Research, 20 (3), 467-477.

Gourville, John and Dilip Soman (1996), "Payment Depreciation: The Effects of Temporally Separating Payments from Consumption," Working Paper, University of Chicago, Graduate School of Business.

Thaler, Richard H. (1985), "Mental Accounting and Consumer Choice," Marketing Science, 4, 199-214.

Thaler, Richard H. and Eric J. Johnson (1990), "Gambling with the House Money and Trying to Break Even: The Effect of Prior Outcomes on Risky Choice," Management Science, 35, 643-660.

Wansink, Brian (1996), "Can Package Size Accelerate Usage Volume?," Journal of Marketing, 60(3), 1-14.