Unintended responses to IT‐enabled monitoring: The case of the electronic logging device mandate

Abstract

Research shows that electronic monitoring technologies can reduce illicit agent (e.g., frontline worker) behavior along monitored dimensions such that principals (e.g., firms) benefit. Drawing on theory from criminology regarding the concept of offense displacement, we explain why the benefits from increased monitoring may be mitigated if agents subsequently increase other types of illicit acts not subject to greater monitoring. We investigate this possibility by examining how a mandate requiring millions of truck drivers to adopt electronic logging devices to record their working hours affected different safety behaviors and outcomes. Using data from millions of driver inspections and crash counts before and after the mandate and a differences‐in‐differences identification strategy, we find greater monitoring due to the mandate achieved its first‐order effect of reducing the frequency that drivers were in non‐compliance with work hour rules (the monitored dimension). However, drivers for small firms that were most affected by the mandate also appear to have increased their frequency of unsafe driving (e.g., speeding). We also find that crash counts for small firms did not fall relative to large firms, and may have increased. These results call into question whether increased electronic monitoring has improved safety in this industry.

The consequences of self‐reporting biases: Evidence from the crash preventability program

“…Safety has been linked to financial performance (Miller & Saldanha, 2016), driver pay (Faulkiner & Belzer, 2019), driver turnover (Miller, Fugate, & Golicic, 2017), technology adoption (Cantor et al, 2016), and market prices (Scott & Nyaga, 2019). Scott et al (2020) analyzed how a recent federal mandate affected safety compliance, unsafe driving, and accident rates. We are the first to study the Crash Preventability Program, which is a novel policy change that affects the safety incentives of all transportation firms.…”

The hidden side of sustainable operations and supply chain management: unanticipated outcomes, trade-offs and tensions

“…Interestingly, nonsustainability OSCM research seems to have a more visible focus on such outcomes than sustainable OSCM studies. Since at least the examination of the bullwhip effect (Lee et al, 1997), OSCM scholars have found unintended consequences in various contexts, such as purchase price variance (Emiliani et al, 2005), the modeling of equilibrium disclosure and pricing strategies (Guan and Chen, 2015), revenue sharing in the mobile value chain (Choi, 2018), price pressure on suppliers (Schleper et al, 2017;Carnovale et al, 2019) or the monitoring of agents through IT (Scott et al, 2020). Moreover, Jonsson and Holmstr€ om (2016) investigate (un)intended consequences from a broader conceptual perspective in their supply chain planning study and emphasize the need to present both evidence of intended outcomes (when things work as designed) and unintended ones (when they do not).…”

To err is human: Exploratory multilevel analysis of supply chain delivery delays

“…Technology has played an important role in developing processes resilient to human error (Adelman, 2019). Important technology-related operational improvements have been achieved across such diverse contexts as retail (Asare et al, 2016), construction (Nnaji et al, 2019), healthcare (Douglas & Larrabee, 2003;Harrington et al, 2011;Knoedler, 2003), retail pharmacies (Elliott et al, 2014), outsourcing to developing country suppliers (Thomas & Ray, 2019), nuclear power generation (Stock & McFadden, 2017), municipal government (Sanford & Bhattacherjee, 2007), logistics (Cantor et al, 2016;Hickman et al, 2015;Scott et al, 2021), aviation (DeFlorio, 2016;Peysakhovich et al, 2018;Rusu et al, 2012), pharmaceutical manufacturing (Markarian, 2019), and automotive manufacturing (Ponnaluri, 2019;Subit et al, 2017). Error reduction technologies include computerized physician order entry (Cowan, 2003;Harrington et al, 2011), vendor-managed inventory (Asare et al, 2016), automated driving technologies (Ponnaluri, 2019;Subit et al, 2017), bar-coded medication administration (Douglas & Larrabee, 2003;Harrington et al, 2011), collaborative decision support systems (Rusu et al, 2012), smart construction helmets (Nnaji et al, 2019), point-of-sale technologies (Asare et al, 2016), eye tracking technology (Peysakhovich et al, 2018), and many other innovative technologies, with a goal of automating human variability out of processes (Granot, 1998).…”

“…Explanations include organizational culture (Averett, 2001;Sanford & Bhattacherjee, 2007), unintended consequences that actually cause an increase in errors, such as reduced nursephysician interaction when computerized physician order entry systems are used (Cowan, 2003), overreliance on technology (Ngwenyama & Nielsen, 2014;Peysakhovich et al, 2018), late employee involvement (Averett, 2001), weak influence skills (Ngwenyama & Nielsen, 2014;O'Conor & Smallman, 1995), inadequate communication (Averett, 2001;Cowan, 2003), lack of perceived benefit (Ngwenyama & Nielsen, 2014), employee resistance (Averett, 2001;Markarian, 2019), technology alert fatigue (Cowan, 2003;Yusuf & Sahroni, 2018), and failure to consider the relationship between a technology, the firm in which it is implemented, and its environment (Gangwa et al, 2014). Scott et al (2021) describe the case of a monitoring technology that achieved its primary goal of preventing truck drivers from exceeding their maximum daily hours of service, but led to the unintended consequence of causing drivers to speed or drive unsafely. Clearly, social and managerial factors are important to a technology's success, in addition to its unique features (Averett, 2001;Ngwenyama & Nielsen, 2014).…”

“…This suggests the importance of a socio-technical systems approach, based on two principles (Bednar & Welch, 2020;Leitch & Warren, 2010;Winter et al, 2014): (a) the interaction between human and technical factors creates the conditions for successful outcomes, and (b) optimization of technology or human factors, alone, is suboptimal and will lead to less successful outcomes. The human element introduces uncertainty, the potential for error, and unintended consequences (Harrington et al, 2011;Scott et al, 2021;Taylor & Robinson, 2015). For example, decision support systems may introduce new opportunities for human errors (Beiger & Kichak, 2004), including over-trusting a technology (Tsai et al, 2003), overriding information not perceived as beneficial ( Van der Sijs et al, 2009), and reduced vigilance (Harrington et al, 2011).…”