Forecasting HR demand and Supply 

Virtually all HRP begins by determining future human resource requirements. It is done by estimating (as closely as possible). Estimation determines how many employees the organization will need in each job category by the end of planning period. The purpose of HR forecasting is to estimate staff requirements at some future time period. Such factors are of two types: 

• The external and internal demand for HR 
• The external and internal demand of HR 

Following are the factors to be remembered: 

1. Internal supply forecasts related to conditions inside the organization. For example, the age distribution of the workforce, terminations, retirements, etc. 
2. Both internal and external demand forecasts, on the other hand, depend mainly on the behavior of some businesses factor. They can be projected sales, product volume to which personnel need can be related. 
3. Unlike internal and external supply forecasts, external and internal demand forecasts are subject to many uncertainties. They can be in the areas of domestic or global economic conditions, in technology, and in consumer behavior, etc. 

1. Forecasting Human Resource Demand 

It is derived from the information generated in the first step of HRP, i.e., "setting the organizational goals" (as explained earlier). Human resource demand forecast is a determination of need for people. It tries to determine the appropriate types and skills for given time periods in the future such as one, three and five years hence. 

Step one of HRP process should provide the HR planners with data on such factors: 

1. Projected/estimated monetary volume of sales, 
2. Units/quantity to be produced 
3. Number of clients to be served, 
4. New facilities to be constructed, 
5. New functional departments to be created and so on. 

Problems/Issues in Demand Forecasting 

In contrast to supply forecasting, demand forecasting is contingent (dependent) upon multiple uncertainties. Some of them are: 

• Change in technology, 
• Consumer attitudes and pattern of buying behavior, 
• Local, national and internal economics, 
• Number, size and types of contracts won or lost,
• Government regulations that might open new markets and close off old ones. They are just to name a few.

As a result, forecast of human resource demand are often more subjective than objective (quantitative) . Therefore, in practice, a combination of the two is typically used (they are given in below figure ). Generally, HR planners must use a variety of techniques to project future HR demand. The techniques available can range from judgmental (qualitative) to sophisticated quantitative models. It may be stated that managers generally follow more than one technique to minimize the errors. Some of the important and popular techniques/methods of demand forecasting are: 

A. Judgmental Techniques( Forecasting HR Demand)

These techniques are easy to describe and understand but applying them may be complex because they require a variety of data and experience.

a. Managerial Judgement/ experience: This category consist estimates made by managers who are very familiar withy the products, processes and jobs in the business. It is suitable only for short-term forecast (e.g. up to 1 or 2 years). Managers of different departments make estimates of future HR demand judgmentally. They convert information on short-term future business activity into numbers and types of employee needed. Role of thumb and decision rules are also subset of this category. There are three ways to administer the managerial judgment method.

• The top-down approach (top managers rely on their knowledge to best guesses about what future employee demand will be ).
• The bottom-up approach (lower-level managers make initial estimates for their units and they are refined by higher level of managers).
• Mixed approach (that is a proper combination of above two approaches).

b. The Delphi technique : It is another but more refined technique of subjective/judgmental approach for demand forecasting. It is a systematic way of obtaining and refining judgements by a group of experts. It was originally developed by the Rand Corporation USA (in the late 1940s). It was developed to forecast technological developments. Nowadays it is also used in HR demand forecasting. The procedures followed in Delphi Technique are as below:

• A panel of supervisors/ managers very familiar with the demand of HR at hand is formed.
• Those experts work privately and not in an assembled group. This is done to avoid errors and conformity that can  be produced by face-to-face peer group social pressure.
• Each expert is asked to make predictions of future events.
• A moderator (HRP expert) collects the predictions, summarizes them, and distributes them to the panel members for another round of forecasts.
• Above round is repeated until the experts' opinions begin to agree.
• At last, convergence/ uniformity of opinion takes place and an acceptable forecast is obtained.

c. Nominal grouping techniques (NGT) : A next technique under judgmental approach is NGT. In this technique several experts (HR mangers) sit around a conference table. They independently list their ideas on a sheet of paper (or in computer). After 10 to 20 minutes, they take start expressing their ideas to the group. As these ideas are presented, they are recorded on larger sheet of paper (or projector screen). It is done so that everyone can see all the ideas . They refer those ideas in next parts of the session. In this technique:

• Group members (HR experts) sit together, but they are not allowed to discuss face-to-face with each other.
• Members secretly write a list of general problem-areas and potential solution to a problem.
• The grouping is only for name sake, that's why it is called NGT. (Note that in the Delphi technique, all members need not be physically present at one place).

Delphi vs NGT : Which one and where? Although the two techniques are similar in process, the Delphi technique is more frequently used to generate predictions about HR but NGT is used more for identifying current organizational problems and solutions to those problems. 

b. Objective/Statistical Techniques

These techniques are based on past data. All of those judgmental techniques are less complex and rely on less data than those based on the statistical methods discussed next. Objective or statistical forecast tend to dominate in practice. Some of popular techniques are:

a. Ratio-trend analysis: It is more scientific than judgmental techniques. It is also the quickest forecasting technique. The technique involves studying past ratios. For example, between the number of workers and sales in an enterprise and forecasting future ratios, making some allowance for changes in the organization or its methods. The following table show how an analysis of actual and forecast ratios, between the numbers of routine proposals to be processed by an insurance company's underwriting department and the number of underwriters employed could be used to forecast future requirement. 

It is a popular mathematical technique to forecast HR demand. It means making estimates based on the ratio between (a) some casual factor (like sales volume) and (b) number of employees required (e.g. number of salesman). It can also be used to help forecast other employee requirements. For example, one can calculate salesman-secretary ratio and thereby determine how many new secretaries will be needed to support the extra sales staff. Ratio analysis assumes that productivity remains about the same. If sales productivity were to increase or decrease, then the ratio between sales to sales person would also change. 

b. Simple linear regression analysis: In it a projection of future demand is based on a past relationship. The relationship can be between the organization's employment level and a variable related to employment, such as sales If a relationship  can be established between the sales and the level of employment, predictions of future sales can be used to make future predictions of future employment. This technique is not perfect in itself because other factors also affect the demand for manpower. For example, the level of sales may double but the level of employment necessary to meet this increase may be less than double. Therefore, there is the need for multiple regression analysis. 

c. Multiple linear regression analysis: It is an extension of simple liner programming regression analysis. Instead of relating employment to one variable, several variables are used  in this technique. For example, instead of using only sales to predict employment demand, productivity data and equipment use data may also be used. Multiple regression analysis produces more accurate demand forecast than simple linear regression analysis. The reason is that it incorporate several variables relate to employment. Due to its complexity, only large organizations use multiple regression analysis.

d. Work standard data technique: Work standard data techniques are more useful for direct production tasks. It can also be used for maintenance and clerical activities. The projected units of employees by applying the established time standards. Given below is a highly simplified example of this technique.

1. Planned output for next year  =20,000 units

2. Standard hours per unit =5

3. Planned hours for the year = 1,00,000

4. Production hours per man/ year = 2,000

5. Number of direct workers required (4/5)=50

Again, this technique (for direct workers) can be combined with ratio-trend analysis to forecast for indirect workers, by establishing the ratio between the two categories.

Limitations: This technique generally can not be used for determining professional, administrative and executive needs. So, Judgmental techniques are most appropriate for this positions.

e. Project/ venture analysis technique: New venture analysis will be useful when new ventures demand human resource planning. This technique requires planners to estimate HR requirements in line with companies that perform similar operations or projects. For example, a hydropower company that plans to open a power generation plant can estimate its future manpower needs by determining employment levels of other power plants/projects.

f. Computerized forecast techniques: These techniques are getting popularity even to all the above techniques for the purpose of simplicity and accuracy. Ready made software is also available to forecast employee requirements. With such a system, a HR specialist (working with line managers) complies the information needed to develop a computerized forecast of staff requirements. Typical data needed: It include direct labour hours to produce one units of product and three levels of sales projections (i.e. minimum, maximum and probable) for the product line in question. Based on such data, a typical program generates figures on " average staff levels required meeting product demands" . Computer also provides separate forecasts for : 

• Direct labour (such as assembly workers),
• Indirect staff (such as secretaries), and 
• Exempt staff (such as executives).

Forecasting HR supply 

Though the available supply of human talent seems to be easier to determine than projected needs, but there a number of complexities in this decision as well. HR managers need to consider current inventory, productivity level and turnover rates among others. Current supply of labor is the best starting point for forecasting a firm's future supply of labor. It is also called the forecasting the supply of inside candidates. 

Before determining how many outside candidates to hire, HR managers need to forecast how many candidates for their projected job openings will come from within the organization from the existing ranks. Like in demand forecasting, there are basically two techniques to help forecast internal HR supply viz,. judgmental and statistical. 

1. Judgmental Techniques: - Replacement planning and succession planning 

2. Statistical Techniques: :Transition/Markov matrix and gain and loss analysis 

1. Judgmental Techniques 

Two popular judgmental techniques used by organizations (to make internal supply forecast) are replacement planning and succession planning. 

a. Replacement planning: It uses replacement charts. They are developed to show the names of the current occupants. It gives information about the candidate's position in the organization and the names of likely replacements. Replacement charts make it readily visible where potential vacancies are. They also show what types of positions most urgently need to be filled. Potential vacancies can be estimated by the present performance levels of employees currently in jobs. The sample replacement chart is shown in above figure. It shows potential vacancies that may possibly occur in those jobs. It also shows the incumbents whoa re not outstanding performers. The incumbents (candidates) are listed directly under the incumbent. Such a listing can provide the organization with a good estimate of what jobs are likely to become vacant. It also shows the candidate who will be ready to fill the vacancy. 

b. Succession Planning: A succession plans is the process of anticipating future staff needs. It also covers preparing plans for meeting these manpower needs internally (a HRIS can help in preparing succession plans) . With the help of succession plans, the positions likely to vacant in future can be estimated before-hand. Succession planning also helps to take timely actions which can be taken to prepare lower level managers for succession. Vacancies in organizations occur due to retirements, resignations, promotions, transfers, deaths. Some of these vacancies are anticipated while others are unexpected. In the absence of succession plans, the management may find it difficult to immediately search for appropriate talent within the organization. The process of developing succession plan includes different activities. They are setting a planning horizon, identifying replacement candidates for each position, assessing current performance and readiness for promotion, etc. Moreover, it also includes activities identifying career development needs, and integrating the career goals of individuals with organization's goals. 

To remember: Succession planning is very similar to replacement planning. The only difference is that succession planning tends to be longer and more developmental and to offer greater flexibility. Other differences of them  are given below: 

Differences between Replacement and Succession Planning

Major differences between replacement and succession planning are given below

Variable	Replacement Planning	Succession Training
Time frame	0-12 months	12-36 months
Readiness	Best candidate available.	Candidate with best development potential
Commitment level	Designated preferred replacement candidate.	Merely possibilities until vacancies occur.
Focus of planning	Vertical lines of succession within units or functions	Development of a pool of talent candidates with capability to take any of several assignments
Development action planning	Usually informal, merely a status report	Usually extensive specific plans and goals set for each person.
Flexibility	Limited by the structure of the plans, but in practice, decision reflect a great deal of flexibility.	Plans are conceived as flexible intended to promote development and thinking about alternatives
Experience base applied	Each manager’s best judgment based on personal observation and experience	Plans are the result of inputs and discussion from multiple managers.
How candidates are evaluated.	Observation of performance on the job over time, demonstrated competence; progress through the function.	Multiple evaluations by different managers of the candidates on varied job assignments; testing and broadening early in careers

 

2. Statistical Techniques

These techniques require extensive analyses of past patterns of employee flows. Statistical techniques are used to project future flows. Most popular techniques are:

a. Transition of Markov Matrix: A key component of most statistical techniques is a transition or Markov matrix. It is used to model movements of employees within or across organizational units. A simple matrix is shown in table given below. The probabilities in part A of that figure represent average rates of historical movements. They show the movements between and out of the three job categories during a particular period of time (for example, last year or the mean of the last three years). To learn how to read the matrix, consider the middle managers. Read the table across the row from left to right. The data shows that historically an average of 10 percent of those managers has been promoted to top management each year. Further more, while 80 percent of them have stayed in middle management, 5 percent have been demoted to lower management. Similarly, 5 percent have left the unit of interest. similar analysis can be done concerning the flows of top-and lower-level managers.

b. Gain and loss analysis: HR managers can identify future flows and internal availability of human resources through gain and loss analysis. The steps used in this technique are: 

• Determine number of losses (i.e. transfers, quits, promotions/demotions, discharges, etc.)
• Determine number of gains (i.e. transfers, promotions, etc. )
• Determine anticipated net internal supply (beginning inventory-projected loss+ projected gain = net supply).