Issue Position: Law Endorsement Policy

Issue Position

There are areas of law enforcement where we have enough evidence to guide policy. High-speed chase policy is a good example and is discussed below. However, law enforcement, like education, is a field that has many studies that are so poorly conducted that they do not provide practical guidance. This is reprehensible because there are many practical studies of law enforcement that could be conducted efficiently and inexpensively. Police officers are already on the job and required to collect the necessary data. It is only necessary have the cooperation of a competent statistician before initiating policy changes.

A few good studies of crime prevention strategies may be found at http://evidencebasedprograms.org/wordpress/?page_id=43. Evaluations of substance abuse prevention and treatment at may be found at http://evidencebasedprograms.org/wordpress/?page_id=46. However, good studies of street-level policing are scarce.

For example, street-level drug enforcement policies can be divided into four groups.

Traditional policing consisting of routine preventive patrols and arrests in response to calls for service.
Hotspots policing that is focused on small geographic regions with high crime rates, repeat offenders, or repeat victims.
Problem-oriented policing that is geographically targeted at drug hotspots and uses a variety of interventions involving other agencies to correct fire, safety, and other code violations.

Community-wide policing activities that involve partnerships with non-police agencies and use a wide array of interventions targeted at relatively large areas. These approaches have been compared in published studies. However, the quality of the work is usually so poor as to make the studies useless and contradictory. The problem is similar to the problem in forensics in which the reliability of DNA evidence has been validated, but the same cannot be said even for such familiar procedures as fingerprint analysis.

Even small, local jurisdictions could usefully track 911 calls. They could, for example, determine which officers are most likely to settle domestic disputes so that there are no subsequent requests for service from the same source. These officers could then be used to mentor others.

One area in which useful progress has been made is in studying high-speed pursuits. A discussion of this topic follows. (Most of the evidence cited below is from Evidence-Based Decisions on Police Pursuits by David P. Schultz, Ed Hudak, and Geoffrey P. Alpert in the March 2010 F.B.I. Law Enforcement Bulletin.)

Step 1 -- State the Evidence

The majority of all police pursuits result from a stop for a minor traffic violation. The violators typically flee because they do not have a valid driver's license, have alcohol in the car or are driving under the influence, are driving a stolen car, or some other offense that does not justify the risks described below.

Roughly one-third of chases end with a crash.

One percent of high-speed pursuits result in a fatality. One person per day is killed in the United States. Twenty percent result in injury. Of all innocent third-parties who are inadvertently killed or injured by police, 42% are hurt in high-speed chases. One percent of all police officers who die in the line of duty are killed during high-speed pursuits.

When police terminate a chase by stopping, driving away from the fleeing person, and turning off their flashing lights, drivers being chased typically resume normal driving within a short period of time. In one survey of chases that involved aerial surveillance, normal driving was typically resumed within 90 seconds. Interview reports from police officers and offenders suggest that offenders typically slow down within two blocks in an urban setting, or two miles on a freeway.

The performance and judgment of both pursuing police and the offender are impaired by their hormonal response to stress (the adrenaline rush).

If police refrain from chasing all offenders or terminate their pursuits, there is no significant increase in the number of offenders who flee.

Police policies concerning starting or terminating a chase usually require an officer to make a judgment that balances the need to apprehend against the risk of injury, death, and property damage.

Step 2 - Evaluate the Evidence

As is often the case, the evidence seems to be making a case. It is trivial to note that the incidence of mayhem in high-speed chases is sufficiently high so that most people would want to see it reduced. The issue, of course, is how do changes that reduce mayhem affect our ability to enforce the law?

Steps 3 and 4 -- List the Options and Prioritize

In this discussion, options and priorities are not separate issues because department policies frequently blend the two. Officers are frequently required to balance the need to catch an offender with the risk.

One piece of data that is clearly missing is the frequency with which the offender in a terminated chase is later caught. The absence of such easily acquired data should be an embarrassment to government officials at all levels that are involved in writing laws or policies that govern police conduct. For example, if we found that nearly all offenders are later arrested, it would be obvious that almost all pursuits are a needless risk.

I will not review the various policies adopted by different police agencies. However, the continuing mayhem associated with police chases suggests that policy responses to that mayhem have been inadequate. Given that eliminating high-speed chases does not increase the frequency with which offenders flee, it seems that policies should be generally more restrictive.

Existing examples of more restrictive policies commonly require a supervisor to authorize continued pursuit. That is undoubtedly a good idea. However, supervisors must rely, in large part, on information provided by the officer who initiates the chase and may, himself, be affected by the adrenaline rush and the natural desire of police officers to catch criminals.

General policies may have supplemental rules that police are required to follow. The City of Cincinnati, for example, prohibits officers from exceeding the speed limit by more than 20 miles per hour. However, Cincinnati news reports indicate that, in the heat of the chase, officers frequently fail to follow the rules. (http://news.cincinnati.com/article/20120223/NEWS/302240029/City-police-routinely-flout-high-speed-chase-policies)

It may easily be observed that many existing policies reflect good ideas that are contradicted by the evidence. For example, it is often assumed that failure to pursue fleeing offenders will result in an increase in the number of offenders who flee. That is a good idea that turns out to be untrue. Similarly, it may be believed that offenders who flee are often guilty of some violent crime. Also untrue.

Street-level police officers must be included in the successful writing and implementation of policy. However, the officers must be educated with regard to the available evidence. All parties, including officers, should be aware of 1) the bias of officers toward making arrests, 2) the effect of the adrenalin rush on judgment, and 3) the possibility that police work attracts risk-taking personalities (unstudied to my knowledge).

Everyone involved in developing policy should be aware that violence in our entertainment media may desensitize people to the effects of violence, to make it seem exciting rather than evil. It follows that everyone involved in setting policy should examine their personal experience with cop shows and train themselves to focus on evidence rather than their emotional response to issues.

Everyone involved in writing policy should have knowledge of the policies of other districts, and how those policies are implemented. They should have knowledge, not only of the general policies, but of specific rules of conduct that may have a greater influence on officer behavior. Telling an officer that she must consider public safety is not the same as telling her she cannot pursue through a tunnel or over a bridge. Special attention should be focused on the policies of districts that can provide even fragmentary data on outcomes.

Any organization writing policy should include consideration of the ease with which they might acquire their own useful data. However, it is folly to collect data and then try to decide what it means. Any large university is almost certain to have a qualified statistician and a criminologist who are willing to work with street-level officers to develop programs that answer important questions without overly-burdening officers with paper work.

Finally, even small agencies should be aware of their ability to provide themselves with useful data. In general, law enforcement data is so easily acquired that small agencies should consider working together to develop policy and acquire larger data sets.

Source