CHAPTER 3: Human Resource Planning: Goals, Trends, and ...



whereas CHAPTER 3:  Human Resource Planning:  Goals, Trends, and Forecasting

Organizations need people just as they need raw materials, equipment, and other materials in order to function successfully.  In fact, it is not uncommon to hear managers acknowledge: "Our people are our most important asset."' Organizations undertake human resource planning to enable them to meet their future "people" needs in the same way in which they plan for their nonhuman resources.

In light of our perspective in Chapter 1, we will define human resource planning as an information-decision-making process designed to ensure that enough competent people with appropriate skills are available to perform jobs where and when they will be needed As such, "it entails defining the organization's human needs for particular positions and assessing the available pool of people to determine the best fit."' The relationship of human resource planning to the other personnel management processes is depicted in Figure 3. 1.

In this chapter we will first make some introductory comments concerning the goals and trends in human resource planning.  We will then discuss the basic information necessary for human resource planning-job analysis.  We have included the subject of job analysis here because firms must first know the nature of jobs to be filled before they can engage in accurately forecasting either the supply or demand of labor for their organizations.  As can be noted in Figure 3.2, before any human resource planning can begin, jobs must first be analyzed and job descriptions and job

           Figure 3.1   Personnel decision making: human resource          planning.

specifications developed.  Finally, we will discuss forecasting the internal demand for personnel. In Chapter 4 we will discuss the specifics for determining both the internal and external supplies of personnel available to the organization.  In addition, we will consider specific human resource decisions themselves: filling job vacancies and reducing the number of employees when necessary.  For example, the 1980-1982 recession necessitated that firms undertake reductions not only of rank-and-file workers, but of middle managers as well.  A model of this human resource planning process is shown in Figure 3.2.

Goals and Objectives of Human Resource  Planning

Effective human resource planning can help managers meet organizational subgoals as well as wider objectives such as profitability and the needs of employees in the organiza-

tion.' For example, effective planning may

1.   Serve to stabilize employment levels when demand for a firm's product is variable, thus reducing the firm's unemployment      compensation liability costs due to layoffs, providing more      job security to the firm's employees, and minimizing the costs of overtime during periods of peak demand.

Figure 3.2   Human resource planning model for any particular job at any particular time.

2.   Prevent young college recruits from leaving the firm after expensive training programs because they lack opportunities for promotion.

3.   Reduce the problems of managerial succession by permitting plans for replacements to be drawn up in advance in case key executives resign or die.

4.   Make it possible to allocate financial resources so that departments will have the necessary people to produce the firm's desired output.'

One example of the way in which human resource planning can provide for employment stabilization involves the use of         temporary personnel ("temporaries") by organizations.  Although such temporaries have been used primarily to fill in for employees when they are on vacation or ill, many businesses are now finding other creative uses for temporaries which can provide cost savings to firms.  Guillet has suggested, for example, that planned staffing may result in using temporary sales personnel during the peak Christmas holidays, thereby minimizing costs associated with        overtime pay for permanent employees.'

In addition, temporaries may be used "on a long-term basis     when employee turnover is high or production low, especially in tedious or routine jobs."' Using temporaries may provide savings not only in overtime costs, but costs associated with absenteeism, employee turnover, and recruitment problems, and savings on        other benefits as well.'

Relationship of Planning to Other Personnel  Processes

From a systems view, human resource planning is mutually interrelated with many of the organization's other endeavors        in personnel management. The strongest relationship exists between human resource planning and selection.  In fact, all selection efforts really are an integral part of the whole human resource planning process.  Organizations that have either stable or increasing human resource needs must go into the external labor market and hire employees even though they generally follow a promotion-from within policy, as we will see later.

In addition, human resource planning is related to both performance appraisal and training and development.  Performance appraisals can pinpoint the skills that will be required for employees to move into higher-level positions via promotion, while training and development efforts may then be designed to provide these skills.

To meet organizational goals, human resource planning seeks to ensure that the organization's demand for individuals at any particular time will be just met by available human resources.  This view assumes that "stockpiling" employees at levels greater than needed and being understaffed are both undesirable.  This assumption represents a major difference between planning for human resources and planning for nonhuman resources.  Although it is generally unacceptable to stockpile or build inventories of human resources, organizations may find it necessary or desirable to build up raw materials or finished-goods inventories.

It is unacceptable to hold human resource inventories for three reasons.  First, human resources are costly and it may be difficult to justify the expense of excess personnel.  As the previous example of the employment of temporaries indicated, there are sounder and more cost-effective options available to personnel planners in business firms.  Second, excess people are not engaged in productive work, and are likely to be bored and frustrated by the lack of anything constructive to do.  Such boredom and frustration can create problems because excess people may make unnecessary work for productive people and may even inhibit the firm's total productive efforts.  Third, since human resources (particularly skilled and professional people) may be in short supply, taking productive workers out of the economy's labor pool may be considered socially unacceptable.'

It is equally undesirable for an organization to operate with too few employees.  As with "stockpiled" employees, individuals may feel frustrated, but in this case because of overwork rather than a lack of productive activity.  This situation may also be dysfunctional to an organization's goals.  Consider, for example, a department store during the holiday season with a shortage of sales personnel.  In addition to the frustrations experienced by employees, such understaffing may also result in loss of employee efficiency.  Customers may respond to long lines and excessive waiting by taking their business elsewhere, with resultant loss of sales by the organization.  Having too many or too few employees may create numerous problems for organizations-problems that can be reduced or eliminated through effective human resource planning.

Trends in Human Resource Management

Unfortunately, the human resource planning efforts of organizations have often been inadequate by failing to emphasize the truly systematized approach geared toward meeting overall objectives.  As Lopez and others have noted:

     Some organizations have perceived manpower planning primarily   in terms of budgeting to control labor costs; others have   viewed it as a management development technique; still others see it as a table of back-ups and replacements for current   employees; and finally, others have viewed it as a means of   establishing a human resources informational system and a       personnel inventory.  Since each of these approaches is   necessarily limited in scope, it is a small wonder that the    state of the art in human resource planning has limped along      quite slowly.9

Toward More Sophisticated Human Resource Planning

In recent years, both personnel practitioners and researchers have emphasized some of the basic facets of personnel decision making that we stressed in Chapter 1: (1) taking systems and contingency approaches, and (2) developing more sophisticated human resource forecasting and planning models.  For example, the growth of equal employment opportunity regulations in recent years has increased the awareness of human resource planners of the effects of external changes on personnel systems.

Two observations are in order regarding these more sophisticated approaches.  First, more complex planning systems have generally been used in larger firms.  Large organizations generally must undertake complex human resource planning and can afford the higher costs of such approaches.  Second, although a wide range of human resource models have been developed, some of these models have ignored so many "real life" personnel variables that they have had virtually no practical application.  On the positive side, there have been numerous quantitative models that have been very useful to organizations.  This chapter will focus on the basic concepts involved in some of these more useful models.

There are a number of reasons for the recent increase in the use of more sophisticated human resource planning models.  For example:

           Organizations simply have been growing larger and more     complex, requiring more sophisticated approaches.  This has   been especially true in those organizations in which    interdependencies have increased."'

The invention and development of the computer has made possible the analysis of complex human resource problems that would previously have been so time-consuming as to be cost prohibitive or virtually impossible to deal with by manual computations.

"The manpower mix in organizations had gradually come to focus around highly skilled managerial and technical talent."" Such personnel have at times been in short supply, and more of a lead time has been required for their training and development.

Once an integrated, well-thought-out human resource planning program has been initiated, managers tend to appreciate its benefits and work together with the firm's human resource specialists in developing viable programs-"they are more willing to plan in this area, if only they are shown how to begin.""

Problems Inherent in More Sophisticated Human Resource Planning

Despite these reasons for the growth of more sophisticated human resource planning, such approaches have faced a number of problems.  First and most obviously, there is an inherent mathematical complexity associated with efforts to model human resource systems.  In addition, however, there are two less obvious difficulties: (1) a lack of certainty surrounding human resource needs in the future, coupled with (2) the existence of an acquisition lead time for meeting those needs.  Even if an organization's human resource planning experts were completely uncertain about the number of operations researchers that would be needed on July 28, 1998, the organization would face no problems if it could at that future time instantaneously obtain any number of such personnel to meet its objectives."

In actuality, however, lead times are needed to recruit and train new personnel, and to train and promote existing employees for new positions or assignments.  Acquisition lead times have become more of a problem in recent years because of the needs for more highly skilled managerial and professional personnel.  Since this trend is expected to continue in future years, the problem of acquisition lead times creates forecasting difficulties for most organizations."

Finally, human resource plans must be updated more frequently in firms (or in any of their subsystems) in which greater uncertainty exists.  As one observer has noted:

     Increasing instability and the greater uncertainties associated      with certain job requirements (e.g., research and development   or marketing) indicate a requirement for more up-to-date      information on emerging needs.  This manpower data is    increasingly subject to change, and organizational needs   dictate timely information with appropriate systems support's

    

Job Analysis

Before engaging in human resource planning, management must first define what work is to be performed and determine how tasks can be divided into jobs.  The assignment of tasks to jobs is commonly known as job design.  Even after jobs have been defined, however, current information about their content must be maintained.  This information-gathering process, called job analysis, serves several useful functions for the organization.

First, since the tasks comprising most jobs will change over time owing to technological innovations or other reasons, job analysis is an ongoing process within the organization.  Up-to-date job analysis may provide a firm with an indication of when jobs need redesigning because of task content changes.  For example, a firm may purchase newer machinery which reduces the time required to perform the tasks comprising the job by 50 percent.  When such situations arise (as they do), job occupants may be assigned new additional responsibilities in order to keep them busy in performing a full day's work.

Second, identifying the work to be performed provides the basis for effective human resource planning.  Such planning cannot be done properly by management unless it knows fairly precisely what a "sales," "engineering," "receptionist," or any other type of job consists of.  For example, a firm may have hired its present receptionist many years ago when the job consisted of operating a telephone switchboard and greeting visitors.  With technological innovations such as automated switchboards and display writers, the tasks performed by that receptionist have changed over the years.  Ongoing job analysis would identify those new duties, providing a better understanding of the current responsibilities of the receptionist.  Thus, should the present receptionist leave, the organization can seek someone truly capable of filling the job.

Relationship of Job Analysis to Other Personnel Functions

Accurately describing what each job entails is essential for the recruitment and selection process.  Job analysis results in the development of not only job descriptions but job specifications as well-that is, what characteristics the individual needs to possess to be qualified for any given job.  For example, in the case of the receptionist, the degree of typing skills needed by an applicant (for example, 30 words per minute or 80 words per minute) would be included in the job specification.  This kind of specification enables the firm to engage in sounder recruitment and selection procedures.

Job analysis also provides valuable data for performance appraisal.  This is so because supervisors can compare an individual's actual job performance with the tasks required to be carried out as specified in the design of the jobs.  In addition, job analysis data provide a basis for training and development efforts in the organization.  In order to train individuals to perform particular jobs effectively, we need to define for them (and for the trainers) what tasks employees are expected to carry out.

Job analysis data are also extremely valuable in wage and salary administration.  Through job evaluation (discussed in Chapter 12), jobs in an organization can be compared with each other and differential levels of pay established for jobs calling for greater or lesser skills.  Basic to evaluating and comparing jobs for wage and salary purposes is a clear statement of what each job entails.

Finally, job analysis has become increasingly important as the federal government has sought to compel firms to end discriminatory hiring, promotional, wage and salary, and other personnel practices.  For example, accurate job descriptions may represent a prime defense against charges of discrimination by proving that a woman performing one job and being paid less than a man performing another is really working at a different skill level.  As a consequence, organizations may be able to defend themselves against charges that the woman is being discriminated against on the basis of her sex as prohibited by the Equal Pay Act of 1963.

Undertaking Job Analysis

There are four basic techniques generally used for gathering information about different jobs: (1) interviews, (2) observation, (3) questionnaires, and (4) the use of diaries or logs.  Each of these four different techniques possesses certain advantages and disadvantages for organizations seeking to analyze jobs.

Interviews In some cases, job analysts may interview the employee who is performing the job, the supervisor, or both.  This technique may be time-consuming, and the danger exists that employees may exaggerate the importance of their jobs.  In addition, the persons interviewed may forget certain duties or responsibilities which are an important part of the job, but which are undertaken on an infrequent basis.  For example, an assembly line worker may also be required to inventory supplies on a monthly basis.  Although this responsibility may be extremely important, it may be overlooked during the interview since it occurs on an infrequent basis.

Observation A second approach to information gathering in job analysis is having the analyst actually observe the individuals performing a job and record observations while doing so.  In some instances this method is very useful, but under certain condi-

tions it becomes more difficult if not impossible.  This is especially true with jobs in which some elements of the complete job cycle occur at intervals that are infrequent or unpredictable.  In the example of the monthly inventory duties cited above, simply observing the worker may not reveal that this important monthly responsibility is an integral part of the job.

Questionnaires Probably the least costly method of collecting job analysis data is by using questionnaires.  Well-designed questionnaires have been claimed to be "the most efficient way to collect a wide array of job data and information in a short time.  However, there is the danger that a respondee will either not complete the questionnaire, complete it inaccurately, or take an excessively long time to return it."" Although the questionnaire is less expensive generally than interviewing, questionnaires contain many of the same problems: information may not be complete, or it may be inaccurate in describing the actual tasks performed.

Diaries or Logs A fourth approach is to utilize diaries or logs in which employees record their daily activities and tasks.  In addition, those activities performed at infrequent intervals (for example, the monthly inventory mentioned earlier) must also be noted.  Because of the difficulties connected with this procedure, portable tape recorders may be utilized to assist in maintaining a record of such activities.

Although it is difficult (and dangerous) to generalize, one authority has suggested that "any combination of the preceding four methods may provide better results than one used by itself."" Whatever technique or combination of techniques is chosen, it is important to determine just what objectives among those mentioned are to be met by the analysis.

The Position Analysis Questionnaire  (PAQ)                          

The four techniques just discussed are general procedures used for undertaking job analysis.  In recent years there has been increased concern with quantification of the process of job analysis.  One such technique utilizing a structured questionnaire is the Position Analysis Questionnaire (PAQ).

The PAQ was developed in the early 1970s through the efforts of McCormick and others.  It consists of 194 job elements of a "worker-oriented" nature which are divided into six major categories.  Table 3.1 depicts these six major categories, along with illustrative job elements.  The job elements are normally rated by the analyst on a scale of 0 to 5. However, as the developers themselves note, the use of the PAQ is far from simple."

Table 3.1 PAQ MAJOR DIVISIONS AND  ILLUSTRATIVE  JOB  ELEMENTS

      Source: Ernest J. McCormick, Angelo S. DeNisi, and James B. Shaw, "Use of the Position

Analysis Questionnaire for Establishing the Job Component Validity of Tests," Journal of Applied Psychology 64 (February 1979):53.  Copyright 1979 by the American Psychological Association.

Major drawbacks of the use of the PAQ involve its length and the high-level vocabulary and complicated scoring requirements.  This technique has been well researched, however, with positive results found regarding both its reliability and validity." In addition, McCormick and his fellow researchers found that the "results indicate that aptitude requirements and rates of pay can be predicted reasonably well from such quantified job analysis data.

Job Descriptions and Job Specifications

Job analysis provides two types of information-job descriptions and job specifications.  We have shown the relationship of job analysis to job descriptions and specifications in Figure 3.3.

Job Descriptions  A job description describes the characteristics of the job to be filled.  In larger organizations, job descriptions are often written by personnel specialists, while in smaller ones, supervisors usually develop descriptions for the jobs in their departments -if descriptions are developed at all! One source that has proven useful to many organizations in developing job descriptions is the Dictionary of Occupational Titles (DOT)." Compiled by the Department of Labor, DOT contains actual descriptions of more than 20,000 occupations covering virtually all jobs in the United States.  We have provided one example of a job description from DOT in the box on page 47 in order to assist readers in better understanding what it is that their university professors do!

In general, job descriptions tend to be less specific at higher levels of the organization.  The reason for this is fairly obvious: Lower-level positions, such as ones on a highly repetitive assembly line, are preprogrammed to a much greater extent than, for example, that of vice president for marketing.  The vice president's job may involve everything from entertaining visiting dignitaries to supervising the development of marketing projections of a particular product over the next three to five years.

It is also important to recognize that most jobs are more dynamic than their descriptions suggest.  First, activities in most organizations will gradually change over time as a result of technological innovations or other reasons (for example, an employee may be given an added duty as a member of the company-wide safety committee).  As

Figure 3.3  The relationship between job analysis, job descriptions, and job specifications.

PORTION OF DOT JOB DESCRIPTION FOR UNIVERSITY PROFESSOR

FACULTY MEMBER, COLLEGE OR UNIVERSITY: Conducts college or university courses for undergraduate or graduate students.  Teaches one or more subjects such as economics, chemistry, or medicine within a prescribed curriculum.  Prepares and delivers lectures to students.  Compiles bibliographies of specialized materials for outside reading assignments.  Stimulates classroom discussions.  Compiles, administers, and grades examinations, or assigns this work to others.  Directs research of other teachers or graduate students working for advanced academic degrees.  Conducts research in particular field of knowledge and publishes findings in professional journals.  Performs related duties, such as advising students on academic and vocational curricula and acting as adviser to student organizations. . . . May be designated according to faculty rank in traditional hierarchy as determined by institution's estimate of scholarly maturity as ASSOCIATE PROFESSOR (education); PROFESSOR (education); or according to rank distinguished by duties assigned or amount of time devoted to academic work as RESEARCH ASSISTANT (education); VISITING PROFESSOR (education).

     Source.- Dictionary of Occupational Titles, 4th ed. (Washington, D.C.: U.S. Government Printing Office, 1977), p. 67.

a consequence, job descriptions need periodic updating.  Second, regardless of the description of a job (except possibly highly repetitious production jobs), the personalities of individuals currently filling the position will moderate the job description.  Unfortunately, this fact is often not taken into account in organizations.

These two points may be seen in Figure 3.4, which is a description of a secretarial job in a large eastern organization.  The incumbent has held this position since 1980, and the job description was updated in 1985 to account for expanded responsibilities.  For example, with a larger department, the individual holding the secretary A position directs the work of three other full-time secretaries, while in 1979 it required supervising only two others.  With the firm having expanded its efforts abroad in 1985, the responsibilities for Item 10-making travel arrangements-have been complicated by much more foreign travel among the managerial and professional members of the department.

The fact that human personalities may have an important bearing on how jobs are actually performed may also be illustrated by a situation that took place for over two years in the department in which secretary A actually worked.  Secretary A was an excellent typist and could take shorthand.  Unfortunately, she often had difficulty reading her own writing! On the other hand, Alex (one of the three secretaries supervised by secretary A) could read secretary A's shorthand quite easily.  As a result of this unique situation, many managerial and professional personnel dictated a letter to secretary A which was later typed by Alex.  This added responsibility of Alex was obviously not included in the job description, but on several occasions this arrangement helped get the typing done on time.

Figure 3.4  The Jones Corporation, clerical job description.

Job Specifications As can be seen in Figure 3.3, job analysis also provides the basis for developing job specifications.  Job specifications are a statement of the human qualifications required to perform the job.  Among the categories of human qualifications which are often included in job specifications are the following:"

     Educational standards which may spell out, for example, degrees      attained, such as a B.S. in engineering.

     Experience requirements, such as two years of general clerical experience for the position depicted in Figure 3.4. This job    requires directing the work of from two to ten clerical     employees, and therefore would normally require some prior      experience in actually performing clerical duties.

     Skills requirements, such as typing 80 words per minute, or     being able to take dictation at 100 words per minute.

     Physical requirements, which might include such factors as      lifting 100 pounds in the performance of the job.

     Certification or licensing requirements, such as a practical    nursing license in order to work in a nursing home, or a     license as a registered representative in order to work for a      securities firm as a salesperson.

Although job specifications are extremely important for human resource planning, recruitment, selection, and so forth, there are two primary problems with their development.  First, some human requirements are fairly easy to translate into specific, well-defined skills (for example, typing and dictation).  Other human requirements (especially those desired in managerial personnel), such as "personality...... the ability to get along with others," or "leadership skills," are much more difficult to translate into precise specifications.

Second, accurate job-related specifications have become increasingly important since the 1971 Griggs v. Duke Power Company decision by the United States Supreme Court.  In addition, in the past two decades personnel decision makers have become increasingly constrained by other laws and regulations, as we will see in Chapter 6.

Human Resource Forecasting

To make effective decisions, managers need to be able to forecast both what their human resource needs will be in the future and from where these resources will be obtained.  As Tichy and others have noted, "the selection, promotion, and placement process includes all those activities related to the internal movement of people across positions and to the external hiring into the organization.  The essential process is one of matching available human resources to jobs in the organization."" Thus, there is both a "demand" and a "supply" side of human resource forecasting.

From the demand side, the organization will need to predict the number of managers, technicians, and so forth that it will need at particular times in the future.  It will also need to forecast what its internal supply of human resources will be.  This process involves an evaluation of current organizational members and their potential for promotion or transfer to other organizational positions as they become vacant.

Organizations must also be concerned with forecasting external human resource demand and supply.  Such forecasting entails an analysis of both the demand of other firms for various categories of human resources, as well as the supply of these resources in the external labor market.  For example, an accounting firm seeking to hire new accountants would need to consider both the external supply of new accountants graduated by business schools five years hence and other firms' demand for those human resources.

In the following sections we will examine a number of forecasting techniques that have been developed to deal with organizations' human resource demand, their own internal supply, and external demand and supply.  Some of the same forecasting techniques are applicable for each of these different types of situations.

 Forecasting the Demand for Human Resources

Length of Forecasting Time Human resource plans must be updated more frequently in firms operating in more uncertain environments.  In addition, the farther into the future that any firm makes its human resource forecasts, the greater the uncertainty will be with respect to their accuracy.  Some observers have indicated, for example, that long-term forecasts are generally not too accurate, even in firms using reasonably sophisticated human resource planning techniques.

Scarborough and Zimmerer have provided one explanation for this problem:

     By its very nature long-range forecasting is less exact and is more complex due to the extended time frame in which the     dynamic variables involved have to fluctuate.  This is especially true in the area of forecasting human resources.       For example, long-term manpower projections are complicated by the difficulties in foreseeing union and governmental    restrictions and in foreseeing specific organizational policy changes.26

Thus, even though it is feasible to develop relatively accurate short-term forecasts for one or two years in the future, intermediate and long-range forecasts may require adjustments owing to the rapid decline in accuracy after longer periods.

Regardless of organization type, length of the planning period, or degree of certainty faced, organizations must focus their human resource demand forecasting on three related factors linked by two conversions.  First, an organization must examine both internal and external variables that will enable it to make a reasonably accurate projection of its output.  Externally, for example, the nation's birth rate has been declining, with the birth rate per thousand declining from 22.7 in 1960 to 18.4 in 1970 to 15.9 in 1979 .17 This decline has resulted in a smaller potential market for baby food, and in the longer run, declining enrollments in colleges and universities (assuming that a constant percentage of the population continues to go to college).  Internally, the baby food manufacturer may have already decided to diversify into other product lines, and may have already begun construction of a plant for producing a new line of products.

Knowledge of such factors will be of value in forecasting product or service demand.  For example, the baby food manufacturer must convert birth rate statistics into the number of jars of baby food of each type and flavor it will be able to sell each year.  Colleges and universities must convert these birth statistics (along with numerous other variables such as the increased demand for adult education) into projections for future student demand in each department for each academic year.  In progressive, personnel-proactive firms, the organization's top personnel officer may participate in decision making regarding the number and types of products produced.  When this is the case, the personnel manager may provide valuable data on the available supply of human resources (both internal and external) to meet projected sales levels.

Forecasts for Specific Jobs                    

When sales forecasts have been developed,they must be converted into specific human resource needs for specified jobs throughout the firm.  In some cases, such a conversion may be relatively simple, while in other instances it may be quite difficult.  For example, it may be fairly straightforward to predict the number of direct employee hours to build a tire, cure it, inspect it, and so on.  Based on previous production data, the firm can simply take the time required in each stage of producing one tire, multiply these figures by sales forecasts, and arrive at a prescribed number of work hours required for production.  Compiling such data for the manufacturing of all tires produced will enable the firm to calculate the number of direct-labor jobs required at each stage of the manufacturing process.

It would be more difficult, however, for the firm to establish the indirect-to-direct labor hour ratios.  For example, the number of setup people required on a production line might increase if the firm's output level increases by 10 percent, but not necessarily by this same percentage.  It is even more difficult to project the numbers of clerical workers, managers, technicians, and professionals necessary to accommodate any changes in sales.  If, for example, total tire sales increased, the number of first-level supervisors might not increase at all.  All supervisors might simply have their span of control increased instead.  Even if the number of supervisors increased, their superiors might be able to handle this increase in the same fashion.

Two additional points concerning output-human resource-demand conversions deserve attention.  First, many companies rely heavily on historical data in making these projections.  Such data may involve a relatively constant percentage increase in annual sales, fairly stable employee absenteeism or turnover rates, and so forth.  However, these data are only valuable to the extent that the relevant variables remain the same in the future as in the past.  In some situations, historical data may be good predictors of the future, while in others they may not.  For example, utilizing historical data regarding the growth in baby food demand following World War II would not be a good predictor of growth of such demand in the 1980s.

Second, because of uncertainties about the future, some organizations will develop a range of output and human resource forecasts contingent upon the overall economic outlook for the firm's output.  For example, such projections may include estimates which are based on sales remaining constant, increasing by 10 percent, or decreasing by 10 percent during the coming year.

Several different forecasting techniques have been used to cope with the problems involved in human resource demand forecasting.  We will focus our attention on four of these: (1) statistical projections, (2) logarithmic learning curve analysis, (3) one general-purpose computer model, MANPLAN, and (4) the Delphi technique.

Traditional Statistical Projections

Traditional statistical projections such as correlation and regression analysis have been found useful by many firms in converting output forecasts to specific human resource needs.  Frantzreb, looking at ways of forecasting human resource demand, has described the value of such tools:

     In most firms sales are directly related to the numbers of      sales personnel.  For other types of employees, other      indicators may be more appropriate: production levels, number     of active accounts, loan volume (for a bank), telephones in service (for a phone company), population (for a governmental      entity), etc.  Usually, the relationship between staff levels   and any of these indicators is explored by correlation     analysis, most often with the help of linear or multiple      regression.28

Rowland and Sovereign have described one firm's experience with the actual use of statistical projections.  The firm based its forecast on a correlation of personnel with the number of major product units shipped.  Initially, the forecast was based on a seven-year trend line and firm estimates of anticipated future orders over the next time period.  The human resource demand forecasts were then broken up into divisional requirements, and the firm utilized this method of forecasting with considerable success." The company recognized the limitations inherent in the use of historical data, and made periodic adjustments in its final forecasts to reflect changing conditions in the economy and the firm.  This firm was deemed typical of many companies, particularly in rapid growth situations, where demand for human resources would be increasing rather than decreasing.

One final caveat regarding the use of statistical techniques cannot be overemphasized: "Any purely statistical approach to forecasting has an inherent bias-a bias in favor of the past.  ""l The assumption underlying such analyses is that past patterns will continue into the future.  However, many contingencies could occur which would alter that assumption.  For example, product lines could be eliminated, contracts could be lost, or changes in government regulations could effectively eliminate (or create) markets for the firm." For these and other reasons, any statistical forecasting technique must be periodically updated in order to accommodate such changes in the organization's environment.

Logarithmic Learning Curve Analysis

In 1936 T. P. Wright published an article specifying how direct labor hours would decrease in the aircraft industry as additional units of aircraft were produced." This decrease in labor hours reflects organizational learning rather than individual learning.

Wright's model has been referred to as describing "manufacturing process curves," .1 experience curves," "learning curves," or more precisely as "logarithmic learning curves," since the model is based on logarithmic calculations.

Wright's model indicated that as the number of units of production doubles, the number of direct labor hours it takes to produce a unit decreases by a constant percentage.  Thus, if it takes 10,000 hours to produce the first unit, and the learning percentage is 80 percent, it would take 8000 hours to produce unit two, 6400 hours to produce unit four, and so forth.  This relationship for both 80 percent and 70 percent curves is illustrated in Figure 3.5.

Regardless of the number of direct man-hours required to produce the first unit, all 80 percent curves have the same shape, as do all 70 percent curves, and so forth.  Thus we have a "family" of learning curves.  Figure 3.5 illustrates both the 80 percent and 70 percent curves, and it is provided because most logarithmic learning curves that exist fall within this range.

As Figure 3.5 and Table 3.2 show, once the learning percentage is established it is relatively simple to determine the number of direct labor hours needed for the second,

Figure 3.5 Log learning curve graph.

fourth, eighth, or sixteenth units produced, and so on.  For the third, fifth through seventh, ninth through fifteenth, and so on, however, the direct labor hours required cannot be precisely determined through simple arithmetic.  For these numbers a mathematical model is needed.

One additional aspect of the logarithmic learning curve model is of critical importance.  In real life it would be rare for all points to fall exactly on any logarithmic learning curve.  This problem may be overcome by using the statistical approach of logarithmic least squares regression analysis to fit a line to the data points.  Although discussion of this approach is beyond the scope of this book, many other sources describe this statistical technique."

The logarithmic learning curve model helps meet two important human resource planning objectives:

1.   Effecting more efficient production scheduling since approximate improvement in work performance is predictable.

2.   Making possible the more efficient control of hiring and laying off human resources over the life of a contract.

In addition, the model may be used to meet other organizational objectives such as establishing production goals and evaluating performance in meeting these goals, and as a negotiation tool in purchasing."

Although the learning curve was originally geared toward aircraft production, other industries have also found it applicable.  For example, in one survey of nonaircraft industries, 61 percent of the respondents indicated that they utilized it in manufacturing

Table 3.2  LOG LEARNING CURVE DATA

emphasizes the decreasing times needed to produce additional (doubled) units, but not the fact that cumulative direct labor hours increase as more units are produced.  For the 80 percent curve, for example, 1000 plus 800 or 1800 hours would be required to produce the first two units; another 702.1 hours (not calculated here) would be required for the third unit; and 640 for the fourth unit; therefore, cumulative labor hours for the first four units would be 3142. 1.

Some industries that have utilized learning curve applications are metalworking, textile manufacturing, candy making," automobile assembly, apparel manufacturing, and the production of large musical instruments."

The learning curve has not been utilized without criticism.  Young, for example, has indicated that aerospace firms' learning curves have sometimes been poorly formulated initially.  In such instances, using curves once actual production has begun poses a real danger of obscuring problem areas.  In these cases, "Being 'on the curve' implies that all is as it should be, when in fact it may be far from it. " 31 More recently, however, it has been concluded that the relevance of the learning curve in labor-intensive manufacture extends beyond the aerospace and related defence industries ... the learning curve phenomenon is not an artifact of defence-contracting procedures or a self-fulfilling prophecy, since two of the firms studied had never attempted learning curve analySiS.39

Two final observations are in order.  First, from a contingency point of view, logarithmic learning curve analysis is most applicable in labor-intensive industries.  One would not find the model applicable in most continuous-process operations, where, for

example, the rate of automated chemical manufacturing may be relatively constant.  Second, since this tool can help firms deal not only with human resource planning problems but with other problems as well, research and refinement of the application of this technique continues.

Computer Analysis

A number of other quantitative models have been developed to forecast human resource needs.  In this section we will describe one such successful computer model: MANPLAN.  Such general-purpose models are important, because when they can be utilized by firms, the cost of development is much less.

MANPLAN was developed by General Electric" to overcome human resource modeling problems (such as the overwhelming mathematical complexity that can be brought into such planning efforts).  General Electric decided to develop a general purpose simulator (MANPLAN) to be made available to users of the GE computer time sharing service.  This computer program interrogates its human user for the information it needs for its forecast, asking such questions as:

1.   How many different product lines do you manufacture?

2.   How many months does your forecast cover?

3.   How many units of refrigerators, washers, dryers, air      conditioners, and fans do you forecast for the next 10 months?"

Once these questions are answered and fed into the computer, the computer can produce a forecast estimating average human resource levels required to meet product demand.  MANPLAN also provides for ranges of possible human resource needs for any period.  The personnel manager can use the model to test the impact of various human resource decision-making strategies.  For example:

     Possible changes in current manning levels of various employee      classifications by means of promotion or downgrading could be   tested to locate minimum lay-off strategies.... Any number of experiments with various numbers of upgrades or downgrades are      permissible until you have finally reached an optimum schedule     .43

     One final merit of MANPLAN is that running the computer model is relatively inexpensive.

The Delphi Technique

The Delphi technique for forecasting is different from the human resource models we have discussed thus far, because it is basically subjective in nature.  For this reason, it is often viewed with suspicion by members of the quantitative school.  The objective of the Delphi technique is to predict future situations by integrating the independent opinions of experts.

A major goal of the Delphi technique is to avoid direct confrontation of experts, since some individuals may be unduly influenced by others because of status differences,

resulting in compromise of good ideas.  This lack of direct confrontation is accomplished by having an intermediary provide the experts with a sequential series of questionnaires concerning the forecasting along with the controlled written feedback to each expert.  During each round of written interrogation, each expert making forecasts independently specifies the assumptions concerning the forecasting problem, identifies source materials that would be helpful in revising the forecast estimates, and is given the same kinds of information developed by each of the other experts whose names are not associated with these data.

The intermediary "gathers the data requests of the experts and summarizes them along with the experts' answers to the primary question. . . . The developers of the Delphi argue that the procedures are more conducive to independent thought and allow more gradual formulation to a considered opinion."" Successive revisions of these procedures are continued until a composite forecast is obtained.

These rounds of information and decision making provide each expert with an iterative (or step-by-step) feedback loop in which the expert receives successive rounds of reactions of others which may be helpful in forecasting.  Such successive rounds usually result in the opinions of the experts converging and thereby providing a viable composite forecast.

The Delphi technique, which originally obtained its name from the widely celebrated oracle of Delphi who lived in the fourth century A.D., dates back to at least the late 1940s at the famed "think tank"-the RAND Corporation in Santa Monica, California.  The technique has become more and more popular, and numerous companies have initiated Delphi forecasting projects to gain insights as to how future events might affect industry and the world in the future." Among the firms that have utilized the Delphi for various types of projections are McDonnell Douglas, Weyerhaeuser, Smith Kline and French, and TRW, Inc        .41

Of special interest is a one-year human resource forecast utilizing Delphi that was successfully carried out by a retailing firm concerned with the number of buyers needed " one year from now."" The firm also utilized linear regression equations to forecast its personnel needs, and found t ' hat the composite Delphi forecast came closer to ascertaining the actual number of buyers needed in the following year than did the regression analysis.  The researchers who conducted this study, however, pointed out that there are certain limitations to the use of the Delphi technique.  For example, even though the experts are instructed not to discuss their Delphi problem, they may actually do so.

Problems have occurred in other organizational uses of Delphi, for example, in integrating the forecasts of the experts (who are able to make no decisions) with the regular decision-making processes of top management.  Further, some authorities indicate that one-year human resource demand forecasts are often easy to make, but that longer-range projections become increasingly difficult.  Thus, one could legitimately ask how successful the use of the Delphi would be in developing, for example, a five- or ten-year human resource forecast.

In spite of such limitations, the popularity of the Delphi appears to be increasing.  We have no reason to expect that the trend toward greater usage will reverse itself.  "At the minimum, the Delphi appears to be highly useful in generating preliminary insights into highly unstructured or undeveloped subject areas such as manpower planning.""

Summary

In this chapter we have focused attention on several aspects of human resource forecasting and planning.  The process of human resource planning is closely related to other personnel functions, such as selection, training and development, and performance evaluation.  Before planning can begin, however, jobs must first be analyzed, and job descriptions and job specifications must be developed for each position.

In order to make effective planning decisions, managers must first be able to forecast what their human resource needs will be.  We have discussed a number of techniques that firms can utilize in forecasting their demand for human resources.  Correlation and regression analysis, logarithmic learning curves, and computer analysis (such as MANPLAN) have all been found to be useful mathematical techniques for forecasting personnel needs.  In addition, the nonquantitative Delphi technique has been used by business firms for forecasting purposes.

Organizations must also determine what sources can be utilized to supply needed personnel, and then develop techniques used to forecast the internal supply of personnel.  In addition, once such forecasts are developed, firms must then create human resource plans for specific situations.  These topics will be discussed in Chapter 4.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download