The Michigan State study reviewed existing methods of determining manpower requirements. This study notes that "... no method is free from arbitrary, subjective judgments, some of the methods that have been and are being used by police departments to justify manpower are completely inadequate except as psychological levels which can be used to move administrators, naive taxpayers and councilmen." (p.3)

Specifically, the study points out that the most prevalent approaches used are (1) comparison of police per population ratios, and (2) comparison of crime rates. Each method will be discussed below.

Comparison of Police per Population Ratios

Utilization of this method involves a comparison between communities of approximately similar population and number of police officers per 1,000 population. It is usually expressed as a ratio. The current national average is 2.0 officers per 1,000 computed on an annual basis. The MSU study cited earlier points out that this method is based on false assumptions as follows:

First, it ignores community differences such as social and
economic conditions, street construction and design, popula-
tion characteristics, culture and customs of the people, and
finally, the characteristics of the courts and other agencies
involved in the processes of criminal justice.

This method also assumes equal competency on the part of police officers. Because of the wide variation in recruitment, selection and training standards, we know that this assumption is totally invalid. Finally, this method assumes that what is sufficient for one department is sufficient for another. By way of example, the National Commission on Productivity analyzed data on police per 1,000 population and found that:

wide dispersion exists among cities of like size, much larger than the dispersion as a function of population size. For example, among cities of 60,000 to 70,000 population, there is an almost four-fold variation (0.7 to 2.7 employees per thousand), with a mean of 1.6 employees per thousand.

and a half times variation (1.4 to 3.6 employees per 1,000).
Although, larger cities have a larger number of employees per
capita on the average, there are many small cities that have
more police per capita than larger cities.8

As this discussion indicates, the use of comparative ratios of police per 1,000 population among like-sized cities leaves much to be desired. The attractiveness of this measure rests in its ease of computation; however, it is clear that it lacks any objective validity.

Comparison of Crime Rates

This is another popular method of manpower justification used

by many U.S. police agencies. This method starts with a base period and assumes that available manpower at that time was sufficient to deal with the crime problem. Fluctuations in subsequent crime rates are analyzed and are used to justify requests for additional manpower. The earlier referenced Michigan State Study points out the following deficiencies of

this method:

...it does not take into consideration other aspects of the
police duties which may in some instances be more important
than crime. It is not at all unusual to find a police depart-
ment spending eighty to ninety percent of its time performing
services for the public; therefore, they could spend a dis-
proportionate amount of time on crime and cut the rate, yet
still need more manpower.9

Additionally, crime statistics can be manipulated even without the user being dishonest or in violation of procedures.

With respect to this latter point, enormous rate changes can occur when improved reporting procedures are instituted. Recent research clearly indicates that the actual, as opposed to the reported, crime rate is at least two to five times higher.

The Home Office Police Research and Development Branch conducted an extensive study in 1967 and 1968 entitled Assessment of Police Establishment by Formula. The state purposes of this research were to:

Analyze the apparent differences in standards between one
force and the other.

Examine the factors which have a bearing on police establishments and to select those which are reliable and measurable.

Create a common standard of manning based on acceptable
factors.

It is first worth quoting several of the major assumptions of this study before discussing its findings. With respect to "adequate" police protection, the authors of this study stated:

Although in their 1962 report the "Royal Commission on Police"
did not attempt to define precisely what is meant by an "ade-
quate" standard of policing, it accepted the view that in most
cases the existing establishment represented an up to date as-
sessment of the total number of police officers required to
provide adequate police protection. (p. 3)

. . In the absence of a more objective criterion we have no alternative but to take such generally accepted standards of policing as a basis on which to build a formula. (p. 3)

In short, this group found that there is at present "no absolute standard of policing, a formula is not likely to provide a single number of men required by a force at a particular time."

The study goes on to point out the reasons for this situation. As they note, some towns have always enjoyed a high standard of policing because they have strong community spirit, adequate resources, a pleasant environment, and so forth. On the other hand, other communities have always been short of policement because of the poverty of the city, even balances between political parties, (which inhibit both parties from supporting actions that lead to tax increases), or an inability to attract a sufficient number of able police recruits because of the unattractive characteristics of the town combined with high wage rates in local industry. They conclude that:

own standard of policing and "adequate" policing now means
different standards of protection to citizens in different
parts of the country. (p. 16)

The study group developed a series of "formulae" for computing manpower requirements based primarily on population, crime rate, road miles, and total acreage. Three types of formulae are presented for large urban, suburban, and rural areas. After reviewing these methods in detail, it is concluded that they are based on false assumptions and are invalid for our purposes. This is not to say that they are wrong, only that they represent the state of the art at the present time.

To develop a "truly" objective manpower requirement for law enforcement, it would be necessary to locate (in economic terms) the point of diminishing manpower returns. This "point" is where the amount of social and monetary costs incurred by adding an additional police officer is greater than the benefits derived from his or her presence. As John Angell has noted in commenting on this issue:

The trouble with attempting to apply this is the impossibility
of isolating all the social and economic costs involved. Since
this is not presently possible, it is impossible to determine
the point of diminishing returns

In the absence of such a measure, many theorists recommend basing manpower requirements on weighted variables such as:

8.

9.

10.

11.

12.

13.

Number of poolrooms, bowling alleys, saloons and clubs that need unusual police protection

Amount of heavy industry

Social and cultural conditions

Nationality and composition of population

Age and sex of population

Proportion of males to females and married to unmarried.

The trouble with this method is that these are subjective fac

tors thought to have sche relationship to the demand for police services. The few studies available that have attempted to correlate police service demand with social and demographic variables are generally ambiguous. For example, the Rand Corporation in studying the relationship of demand for service to socioeconomic data in 30 large cities found:

the most disturbing and surprising results occur for the correlation of Index Crimes per capita with the demographic variables (percent non-white, percent poor under $3,000, percent age 15-29, percent moving within past two years). No significant pairwise correlations exist at all.

The Urban Institute Report on police productivity studied this question and produced different results. They found that:

In no case did the socioeconomic variations explain more than 11 a small part of the effectiveness and productivity variations.

Again, based on these results, aside from the fact that no comparable studies are available for recreation areas, we must conclude that the use of this method is simply not applicable.

The most common method used by sophisticated U.S. police agencies to estimate needed manpower requirements is the evaluation of past workloads. For example, if in year "X", the police agency investigated 2,000 traffic accidents, and each accident required 2 man-hours to fully investigate, then we know that 4,000 man-hours are absolutely necessary to process this type of event. In justifying its budget request the police agency then projects the expected number of accidents for the coming year, applies the workload factor and computes necessary manpower needs. However, such a method also has some major conceptual difficulties. The Michigan State study mentioned earlier states that this method is based on several assumptions, which, if wrong, render the estimates inaccurate: The department is administratively and organizationally sound (e.g., administrative overhead is at a minimum; the activities being used by the department are most effective in achieving the goals; and organizational relationships are proper).