Many departments have put things into their pursuit policies that have the potential of exposing officers to greater risk, and make the job of policing that much harder. Departments adopt many of those policy measures in an attempt to reduce liability exposures, and many times it helps--but sometimes there are unintended consequences.
Consider the department that wants to limit the number of police units that can be involved in a pursuit. After all, the more units there are running around out there, the greater the likelihood of a crash, so it makes sense to limit the number of cars involved.
However, when you limit the number of units, you limit the number of officers. Most departments utilize one-person patrols, so "two units" means "two officers." If there are multiple suspects in a fleeing vehicle, that may not be enough, and the potential for an officer injury is increased.
What about a policy that stipulates that, no matter what, officers cannot exceed speed limits by more than (plug in your own arbitrary figure here) while in pursuit? Is that realistic on a clear, dry, wide open roadway at four in the morning?
Or what about a department that is concerned about a suspect running into a roadblock, so they stipulate in their policy that roadblocks will be considered to be deadly force, and will require supervisory approval before they can be implemented. Such a policy stance represents a far higher standard than the courts require, and is based on a flawed assumption that the likely result of a roadblock will be a fatality. If this were, in fact, the case, then such a standard would be appropriate. However, such outcomes are not the norm, so the fatal result is not "likely", and roadblocks do not meet the definition of deadly force.
Some departments cast intentional collisions in the same light, stipulating in writing that such techniques -- often referred to as "ramming" -- also represent deadly force, and should only be undertaken as a last resort. Again, this is a higher standard than the courts require, and involves an over-simplified view of contact techniques. The U.S. Supreme Court just made this clear in the recent case of Scott v. Harris.
No technique, in and of itself, represents any particular level of force, let alone deadly force. All techniques, whether they be related to managing the end of a pursuit, or related to any other use of force incident, must be viewed within their proper context, or as we more typically refer to it, within the totality of the circumstances.
Thus, a roadblock can be set up in such a way as to render it the equivalent of deadly force, and an intentional ramming of another vehicle can be undertaken in such a way as to constitute deadly force. But whether or not these things are "deadly force" will be based on the context within which they were used, not just the fact that they were used.
In fact, even though some departments undertake these policy measures in an attempt to reduce their liability exposure, there is a very real chance that they are actually making things worse. By defining some officers' actions as deadly force when they clearly may not be, departments may be discouraging officers from using perfectly viable techniques that could reduce the danger of a particular situation, thereby exposing officers and other citizens to an increased risk of harm.
And by defining some of these terms in a way that creates a higher standard for officer conduct than the courts require, departments may be playing right into the hands of plaintiff's attorneys
Departments make these types of policy mistakes because they are trying their best to do the right thing for themselves and their communities. This is not a problem unique to pursuit policies--it can be seen in many other types of department policy documents. The problem seems to be the most concentrated in policies related to higher risk activities, such as driving, use of force, and arrest practices, but there is evidence of it elsewhere.