Primed for Crime? Investigating Dispatcher-Officer Inconsistency in Classification of Behavioral Crisis Calls to Police
Files
Publication or External Link
External Link to Data Files
Date
Authors
Advisor
Citation
DRUM DOI
Abstract
Emergency call-takers and dispatchers play a critical but understudied role in shaping responses to behavioral health crises. Before responders arrive on scene, ambiguous, caller reported information must be interpreted, classified, and translated into a singular call type that structures police response. Call classifications assigned by dispatchers may prime officers’ expectations of an incident, which may have important implications for whether individuals are routed toward treatment or further into the criminal legal system. However, research on dispatch decision-making is limited, and little is known about the prevalence, predictors, or consequences of classification error, particularly as it relates to behavioral crisis calls.
This dissertation addresses these gaps by examining (1) the prevalence and patterns of crisis call classifications assigned by dispatchers and responding officers; (2) individual, temporal, and environmental predictors of co-identification (i.e., calls classified as behavioral crises by both dispatchers and responding officers or, calls where the officer was “primed” for a crisis-related interaction); and (3) the downstream benefits of co-notification (i.e., crisis priming) for police response. Using administrative data from a large police department, this study analyzes behavioral crisis-related calls for service and employs multivariate regression models to estimate associations between call characteristics, classification outcomes, and incident response.
Findings indicate that, within discrepantly classified calls, officer-identified crises (i.e., calls initially dispatched as a non-crisis incident but re-classified by responding officers) are more prevalent than dispatcher-identified crises (i.e., calls initially dispatched as a crisis incident but re-classified by responding officers).Additionally, non-crisis priming is associated with factors linked to perceptions of threat and ambiguity, including age, gender, time of day, and historical call patterns. Moreover, non-crisis priming is associated with more restrictive and resource-intensive responses, including increased likelihood of emergency petition, longer time on scene, and greater odds of co-occurring use of force incidents. Together, these findings suggest that initial dispatcher call classifications may function as important primes associated with subsequent police responses to crisis encounters. Implications and future directions are discussed.