Do Managers Overreact When in Backlog? Evidence of Scope Neglect from a Supply Chain Experiment

Abstract

Problem definition: We empirically examine a complementary behavioral source of the bullwhip effect that has been previously overlooked in the literature: that individuals order more aggressively (i.e., overreact) when they face shortages than when they hold inventory. Methodology/Results: We conduct a behavioral experiment using the beer distribution game. We estimate decision rules using multilevel modeling approaches that overcome several drawbacks of the estimation methods used in the earlier literature. We find robust evidence that, contrary to the overreaction when in backlog hypothesis and reports from popular press, decision makers order less aggressively and become insensitive to the scope of the problem when in backlog—a scope neglect phenomenon. Managerial implications: We propose a dual-process theoretical account predicated on affective reactions to explain this scope neglect. Our results suggest that affective reactions under novel operating conditions or dramatic events in supply chains, such as the COVID-19 pandemic, can overwhelm cognitive processing of managers and make them fail to recognize the full scope of the problems faced and update decision models accordingly. Understanding the cognitive-affective drivers of ordering behaviors that generate supply chain instability is important in designing interventions to mitigate their negative effects.

The impact of information visibility on ordering dynamics in a supply chain: a behavioral perspective

“…Oliva et al . (2021) find positive but not statistically significantly at 10 percent estimating aggregate supply‐line coefficients for all positions. Importantly, Sterman and Dogan (2015) capture two feedback processes going through the supply line of orders with the supplier (the channel‐order backlog): one balancing and one reinforcing (Figure 2).…”

“…The estimated coefficient for own inventory in the baseline condition is −0.12, a value that has the expected sign and a magnitude lower than some studies (e.g. Sterman, 1989; Croson and Donohue, 2002, 2003, 2006; Croson et al ., 2014; Sterman and Dogan, 2015) and equal to others (Oliva and Gonçalves, 2007, and Oliva et al ., 2021). The estimated coefficient for supply line in the baseline condition is +0.05, a value that has an unexpected sign but of similar magnitude in comparison with most other studies.…”

“…While a positive and significant effect of supply line is counterintuitive, it has been reported in previous published studies particularly for estimations of individual nonfactory players (e.g. Oliva and Gonçalves, 2007; Oliva et al ., 2021). In particular, Oliva and Gonçalves (2007) find positive ( β SL = +0.02) and statistically significantly (1 percent) supply‐line coefficients estimating aggregate nonfactory positions.…”

“…A few studies focus on the supply chain (e.g the beer game) as the unit of analyses (e.g. Oliva and Gonçalves, 2007, Oliva, Abdulla & Goncalves 2021, and this study), obtaining coefficients estimates for different information cues directly. The dependent variable across all studies focus on either the orders placed or the order variability.…”

“…Behavioral research finds that the Bullwhip Effect persists, even when researchers control for operational causes in their experiments. Behavioral causes of the Bullwhip Effect include underestimating the supply line (Sterman, 1989), reaction to backlog (Oliva and Gonçalves, 2007), phantom ordering (Sterman and Dogan, 2015), coordination risk (Croson et al ., 2014), and scope neglect (Oliva et al ., 2021). Different attributes can influence those behavioral causes such as the cognitive profile of people (Narayanan and Moritz, 2015), retailer shortages in a competitive environment (Villa et al ., 2015), and access to POS or inventory information (Croson and Donohue, 2003, 2005, 2006).…”