Criminal investigations: If, when, and to what extent does detective effort impact case outcomes?

Clearance rates have brought many to question the efficacy of criminal investigations. Unfortunately, existing research has only estimated what is in criminal investigations and do not seek to understand what should be. This knowledge gap has inhibited detectives and investigative administrators from making educated decisions about where and when to direct police resources. This endeavor seeks to overcome this shortcoming by estimating the time and activities associated with arrests and charges. Time-task logs from 184 detectives with reactive investigative units in the Houston Police Department were collected and paired with index offense data. Through a unique temporal sampling procedure, 602 index offenses were selected and detective effort among these cases was disaggregated into subsamples based on the case's suspect status at the time the case was assigned. The curvalinearity of these estimates was tested and recalculated into binary estimates when quadratic relationships were observed. Continuous and re-estimated quadratic variables were then evaluated for multivariate suitability and regressed to predict a cases likelihood of arrest and/or charge.;Results suggest that detectives have a great deal of discretion in if, when, and to what extent they work index crimes. Suspects who were known at the time the case was assigned were 6.4 times more likely to result in arrest and/or charge while controlling for detective effort. Time spent interviewing the suspect was the most consistent predictor of the likelihood of arrest and/or charge in each of the samples. The influence of all other individual investigative activities varied by the suspect status at the time the case was assigned. Detective effort mattered most among the most difficult to solve cases. These cases, however, received the least amount of investigative attention. Additionally, more time performing individual investigative tasks was not always positively associated with case outcomes. The policy implications for patrol officers, civilian employment in police departments, investigative administrators, detectives, and prosecutors are discussed. Finally, sampling omissions and measurement oversights that contributed to specification error and inhibited these analyses ability to observe between and within unit differences among a variety of traditional and non-traditional investigative outcomes are discussed as limitations and areas of future research.;KEY WORDS: Criminal investigations, detective effort, police.