In fact, and quite surprisingly, one of the key findings of the Rand Report was that 50% of the nation's detectives could be eliminated without having a significant effect on clearance rates in the country (O'Connor). This conclusion flies in the face of conventional wisdom on police work, yet was fully supported by the exhaustive study. The study was funded by the National Institute of Justice and was designed to monitor the effectiveness of detectives in clearing cases in a variety of situations. Surveys were designed and then sent to local and county police departments throughout the country with at least 150 fulltime personnel or a municipal jurisdiction of more than 100,000 people. In total, three hundred agencies were solicited, of which 153 responded with answers to the survey. While all of the data from those 153 agencies was used in the compilation of the study, twenty-five of the respondents were selected for a more detailed analysis. This analysis included visits by researchers, direct interviews, and observations of detective work in order to better inform the conclusions of the study (Purpura 167).
The findings of the study were quite comprehensive and far-reaching. Specifically, though, the researchers determined that difference in detective organization, differences in training, differences in staffing, differences in workload, or differences in procedures had no significant effect on the crime rates, clearance rates, or arrest rates. In fact, it became apparent from the research that detective provided only superficial treatment of more than 50% of the cases that would pass across their desks in the course of their jobs (Purpura 167).. With such a low rate of interaction with their caseload, it is little wonder that detectives and police departments have had such a low rate of crime clearance. This last fact might even suggest that more detectives are the answer to the problem, as these would easily be able to handle some of the cases that went unnoticed. But the Rand study did not concur with this assumption, finding instead that more detectives will not affect clearance rates.
Part of the problem, the study found, lay in the nature of detective work. Investigators spent most of their time on the job reviewing reports, filing paperwork, and attempting to locate and interview victims in cases that would ultimately go unsolved (Purpura 168). Again, one might guess that the problem is manpower -- more detectives would mean more hands to finish paperwork. But the researchers found that the single largest determinant of whether or not a crime would be cleared is whether or not the victim provided useful information to onsite responders, especially information regarding the perpetrator of the crime. Other factors, such as the presence of forensic evidence, were much less important (Purpura 168).
In the intervening years since the original Rand study that challenged our conventional wisdom on detective work, duplications of the study and similar research has confirmed the original conclusions. A University of Maryland study that considered homicide clearance rates in four large cities in the country in 1994 and 1995 found that the rates were affected by two categories of factors: the nature of the crime and police practice and procedure (Wellford and Cronin 3). The former category of factors obviously is outside the realm of police control, but the study suggests that police departments and detectives can control the latter. Interestingly, the researchers conclude that clearance rates will increase most significantly when changes are made to response times, faster notifications of detectives, and securing witnesses at the scene (Wellford and Cronin 6). This conclusion runs eerily parallel to the conclusions of the Rand study. No suggestion was made that more detectives or changes in detective training could improve clearance rates, but rather that the initial response time to a crime -- that makes it more likely key evidence and witnesses will be found -- were the most significant determinants of increased clearance rates.
Other duplications of the Rand study reveal similar conclusions. When the offender was identified at the scene of the crime, clearance rates were much higher than when he or she was not. Research conducted by the Police Executive Research Forum found that arrests were much more likely on a case based on how quickly the crime was reported and detectives were able to get to the scene. The likelihood of an arrest stood at 33% if the crime was reported while in progress, but dropped to only 10% if it was reported only a minute later. If the crime was reported more than fifteen minutes after the fact, the likelihood of an arrest dropped to only 5% (Purpura 168). Time and availability to credible witnesses, then, seemed to be the most important factors in determining whether or not an arrest could be made and clearance achieved. The conclusion we can draw from these studies is obvious: changes in detective work are relatively meaningless in affecting clearance rates. Instead, factors that are generally outside of the purview of police departments seemed to be most crucial such as witness statements obvious evidence.
In most cases, then, we see that crimes are solved based on the information that is gathered by the police not because there are more detectives in the field or because they are better trained (Livingston). In the case of murder cases, clearance rates have very little to do with the availability of detectives or in their access to specialized training. Instead, we find that clearance rates are most affected by the choices that killers make and the types of crimes that are committed. Murder cases that involve friends or family -- such as crime of passion -- have much higher clearance rates because witnesses generally know who is involved. When murders occur between strangers -- such as robberies that lead to murder or when a killer takes minimal steps to avoid detection -- capture is much less likely and clearance rates are significantly lower. For example, in New York City, where crimes between strangers are much more common than in other parts of the country, clearance rates for homicides are closer to 50%, down significantly from the national average that pushes 70% (Livingston).
In the case of nonviolent crimes, it is even harder to clear cases. A separate study by the Rand Corporation found that when a thief can escape the scene of his crime before police arrive, the clearance rates are between 2% and 3% (Livingston). It is only because police are able to arrive at many of these types of crimes while they are in progress that the clearance rate is as high as it is, just below 20%. Nonetheless, this extremely low rate of case clearance means that the most important factor in the case of nonviolent crimes is police response time, not detective availability. Detectives are rarely the first responders at a crime scene, and certainly not the officers who are tasked with responding to crimes in progress. Thus, it seems quite unlikely that changes in the number of detectives or procedures related to them will have a significant impact on clearance rates.
The conclusions we can draw from this data is clear: putting more police detectives on the job or changing some police procedures related to detectives will not have an appreciable effect on clearance rates or crime rates. We have already established that crime rates and clearance rates are not positively associated, despite commonsense notions to the contrary. Clearance rates, too, cannot be improved with changes to the number of type of detectives. Instead, numerous studies -- including the pivotal Rand study of the 1970s -- conclude that the most important factors related to case clearance are response time and the availability of useful information from witnesses on the scene. Since neither of these factors can be directly affected by changes in the number or kind of detectives, we can conclude that such changes will not be crucial in improving case clearance rates or reducing crime rates overall.