Sunk Costs and Quasi-Hyperbolic Discounting : Keeping Profitable Commitments by Considering Sunk Costs

Purpose: Managerial accounting education generally insists that managers should never consider sunk costs. This suggestion seems inconsistent with a common mode of thinking about future rewards: quasi-hyperbolic discounting. This paper aims to explore the conflict between sunk cost consideration and quasi-hyperbolic discounting and to illustrate when sunk cost consideration may be appropriate.Methodology: I conduct three numerical experiments, i.e. simulated experiments based on analytical models, to demonstrate how it can be beneficial to consider sunk costs in some circumstances. All three numerical experiments assume quasi-hyperbolic discounting. First, I test considering sunk costs with future rewards that are certain. Second, I test considering sunk costs with uncertain future rewards. Lastly, I test two different educational interventions to change decision makers' thought patterns.Findings: I find that considering sunk costs worsens decisions when there is bad news and improves them when there is good news. I find that an educational intervention that partially dissuades managers from considering sunk costs improves decisions when bad news arrives and worsens them when good news arrives. I also find that an educational intervention that reduces uncertainty improves decisions when bad news arrives and does not worsen these decisions when good news arrives.Originality/Value: I provide numerical examples of situations in which considering sunk costs is valuable. My findings on educational interventions provide information about the tradeoffs of teaching that sunk costs should never be considered