Laura J. Kornish
Assistant Professor, Marketing Division

Leeds School of Business

The University of Colorado at Boulder

Publications

"Repeated Commit-or-Defer Decisions with a Deadline:  The Influenza Vaccine Composition"  with Ralph L. Keeney, Operations Research, May/June 2008.

Abstract
Seasonal products have an effective inventory deadline, a time by which the inventory must be ready to distribute. The deadline creates an incentive to start early with production. However, opportunities to gather information that might change production decisions provide an incentive to defer the start of production. We study the resultant dynamic decision problem with alternatives that commit to one of several courses of action now and an alternative to defer the commitment in order to gather more information about the possible consequences of each alternative. The deadline increases the effective cost of gathering information, as that cost includes the value sacrificed by reducing the time available to produce inventory. We frame our model using the annual influenza vaccine composition decision: deciding between strains of the virus to include, which must happen in the spring to allow time for vaccine production before the fall flu season begins. Our analysis describes the optimal decision strategies for this commit-or-defer decision. Many insights are drawn from this model that could contribute to more informed flu vaccine composition decisions. We comment on the relevance of this commit-or-defer decision model to a firm’s production decisions for other seasonal products with an inventory deadline such as fashion goods.

"Technology Choice and Timing with Positive Network Effects" European Journal of Operational Research, August 2006.

Abstract
When two competing and incompatible products coexist in a market, potential users face a choice between the two products and the alternative of deferring the decision. This paper examines the choice between the two substitutes, where each one is subject to a positive network effect. That is, a user of one of the products experiences an increase in the value for each additional person using the same product. We examine this buy or wait problem, either for an individual or a manager making the investment on behalf of a firm, by formulating and analyzing a decision-theoretic model. To model the stochastic evolution of market share, we build on the generalized Polya urn of Arthur, Ermoliev, and Kaniovski (1987), allowing for composition of the market to affect not just the relative market shares but also the absolute growth rate of the market. We show that the optimal strategy is defined by a pair of market penetration thresholds that depend on the market composition. Looking at the effect of the network effects on the optimal strategy, we find that more pronounced network effects can either raise or lower the penetration thresholds. 

"Discipline with Common Agency: The Case of Audit and Non-Audit Services" with Carolyn B. Levine, The Accounting Review, January 2004.

Abstract
Using a common agency model, we investigate the strategies of self-interested auditors (the agent) hired by both managers (for non-audit services) and shareholders (for conducting an audit) of the same firm.  In a single period model, managerial discretion over consulting fees can influence auditors to issue reports that are more favorable than warranted.  Shareholders, represented by an audit committee, cannot recover truth-telling.  Removing the current restriction on contingent audit fees allows audit committees to offset the incentives provided by management and instead provide the auditor incentives to report truthfully.  Extending the model to a multiperiod framework, the audit committee can motivate truth-telling by making retention decisions which are contingent on outcome.  Auditors will consider the impact of overreporting on their ability to generate future audit fees from the same client.

"Smart Agents: When Lower Search Costs for Quality Information Increase Price Sensitivity" with Kristin Diehl and John G. Lynch, Jr., Journal of Consumer Research, June 2003.

Abstract
Recent consumer research suggests that lowering search costs for quality information reduces consumer price sensitivity by creating greater perceived differentiation among brands (e.g., Kaul and Wittink 1995; Lynch and Ariely 2000). We argue that lowering quality search costs by smart agents can have the opposite effect on differentiation and price sensitivity. Smart agents screen through a universe of alternatives, recommending only a handful well-matched to the customer’s quality preferences. In this research, we ask and answer the following questions: In markets in which price and quality are uncorrelated, will the use of screening agents increase or decrease prices paid compared to searching from an unordered list of options? Will increasing the size of the store’s underlying assortment increase or decrease prices paid when options have been screened on quality?  In markets where higher priced goods have higher quality, will the use of screening agents increase or decrease prices paid and quality selected? Experiments 1 and 2 test the effect of quality screening when price and quality are uncorrelated. We then present an analytic model for markets in which price and quality are correlated. We deduce that ordering can cause price and quality to increase or decrease depending on the slope of the price quality relationship in comparison with the relative importance of price in the utility function. We find support for this model in Experiment 3. 

"Pricing for a durable-goods monopolist under rapid sequential innovation" Management Science, November 2001.

Abstract
A durable-goods monopolist who will be introducing new and improved versions of his product must decide how to price his products, keeping in mind the relative attractiveness of the current and future products. Dhebar (1994) has shown that if technology is changing too quickly and the producer cannot credibly commit to future prices and quality, then no equilibrium strategy exists. That is, there is no credible strategy for the future product that the producer can commit to in the first period.  We show that an equilibrium pricing strategy exists if the monopolist does not offer upgrade pricing, that is, special pricing to consumers who have bought an earlier version. We show the possible purchase patterns in equilibrium and derive the optimal pricing strategy.

"On Optimal Replacement Thresholds with Technological Expectations" Journal of Economic Theory, December 1999. 

Abstract
Balcer and Lippman (1984) develop a model to analyze the "buy or wait" problem under technological change. They show that this dynamic problem has a threshold solution: if the difference between the best available technology and the one currently held exceeds a certain threshold, then buy. They also claim that the threshold is increasing in the "discovery potential." That is, the faster the technology is changing, the higher the threshold. In this note, we point out an error in the proof of that theorem and further, we provide a numerical counterexample to the claim.

Working Papers

“Optimal Referral Bonuses with Asymmetric Information: Firm-Offered and Interpersonal Incentives” with Qiuping Li (July 2008)

Abstract
Referral bonuses, in which an existing customer gets an in-kind or cash reward for referring a new customer, are a popular way to stimulate word-of-mouth. In this paper, we examine key firm decisions regarding such bonuses. Biyalogorksy, Gerstner, and Libai (2001) also study referral bonus programs; a key difference is that we study the role of recommendations not just in spreading awareness (as they do), but also in providing assessments. We start with the idea that people have a variety of reasons for making product recommendations, including placing a value on a friend’s outcome with a product they recommend. We apply that idea in a context of asymmetric information: a customer combines his knowledge about the product and his familiarity with friends’ tastes, making him more informed than the friends. Thus, the recommendation is a signal about the value of the product to the friend.

We show that these two ideas, “concern for others’ outcomes” and “recommendation as signal,” drive the following main results. On setting bonuses: the more concern customers have for their friends’ outcomes, up to a point, the higher the optimal bonus should be. But, with enough concern, bonuses should not be used at all. On setting prices: greater concern for friends can drive the price up or down, depending on the population characteristics. But, with enough concern, the optimal price drops to a low level. On firm profit: when it is optimal to use a bonus program, there is generally a negative association between the bonus level and profit. This implies that bonus programs that require only small bonuses will be the most profitable. On consumer welfare: not surprisingly, we often see higher optimal bonuses associated with lower consumer surplus. However, if customers are quite risk-averse in their recommendations, not only can bonuses be an effective tool to expand recommendations, but they can also increase consumer surplus.

“New Product Launch Date Decisions: Promotion and Production” with Moren Levesque and Lisa Maillart (May 2007)

Abstract

Many new products, including hybrid cars, Harry Potter books, video game consoles, and holiday toys, are released to the market with great fanfare and, interestingly, sometimes experience shortages at launch. The newsvendor model reflects the fact that such shortages are an inherent possibility when demand is uncertain. We extend the newsvendor model to account for the time the firm spends building inventory prior to launch. We analyze the tradeoffs between spending money on promotional activities to increase demand, spending time to build inventory, and getting to market quickly. We show that, when considered in isolation, more promotional activity (production) typically increases (decreases) the chance and expected magnitude of shortage. However, when the promotion and production decisions are coordinated, their effects on shortage are more subtle; neither investments aimed at increasing the production rate nor a reduction in promotional activities necessarily alleviate the firm’s risk of experiencing a shortage upon launch.

"Before the Fork in the Road: Investments in Research on Competing Alternatives" (May 2006)

Abstract
In many decisions, firms face binary choices: one alternative or another can be pursued, but not both. This paper examines the logic behind initially exploring multiple approaches when constraints or the structure of returns will ultimately force such a choice. Many reasons have been offered as rationales for diversification of investment, such as risk aversion and economies of scope. In this paper, I describe a new rationale for diversification, based on an uncertainty-resolution or “value of information” explanation, and derive the conditions under which it holds. If a single alternative will eventually be picked, investment in research in multiple alternatives has the greatest incremental value over investment in research on a single alternative in the following scenarios: 1) when, with joint information, a good outcome from the investigation of an alternative might nonetheless result in another alternative’s being picked because the other alternative has an even better outcome, and, likewise, 2) when a bad outcome might result in a alternative’s being picked because the other alternatives are even worse. Even when the outcomes for different alternatives are probabilistically independent, decisions about research cannot be made separately for each alternative; the ultimate comparison of the alternatives links the research decisions. With probabilistic dependence, if each alternative has a non-zero value of information individually, the incremental value of joint information over individual information is highest when the alternatives are mildly positively dependent.

"Search and the Introduction of Improved Technologies" with Steve Lippman and John Mamer (October 2004)

Abstract
Modelling R&D as standard sequential search, we consider a monopolist who can implement a sequence of technological discoveries during the technology search process: he earns revenue on his installed technology while he engages in R&D in order to find improved technology. What is not standard is that he has several opportunities to introduce improved technology.  We show that his optimal policy is characterized by thresholds: introduce the newly found technology if and only if it exceeds the threshold as a function of the state of the currently installed technology and the number of remaining introductions allowed. We also analyze a non-stationary learning-by-doing model in which the monopolist's experience in implementing new technologies imparts  increased capability in generating new technologies. Because this non-stationarity model is not in the class of monotone stopping problems,a number of surprising results hold and several seemingly obvious properties of the stationary model no longer hold. 

Laura Kornish | Leeds School of Business | University of Colorado at Boulder

Under construction 7/2008