Friday, October 28, 2005

Wage and Salary - Key variables

Objective : To Understand variables that impact Basic salary & incentives design

From the viewpoint of the employee, the end product of any compensation program is a paycheck. The decision regarding the type of salary administration and/or structure system to be used do not, by themselves, deliver a paycheck to the employee. The wage determination must be personalized by making a further set of decisions.

The first compensation decision, the wage level, is an external organizational decision that determines the organization's competitive posture toward its human resources. The second major compensation decision is an internal organizational decision involving the structuring of the jobs within the organization. Putting these two decisions together in a wage structure provides the wage or range of wages that the organization perceives as equitable for each of its jobs. Although pay rates are determined for jobs, it is people who receive paychecks. So the next decision to be made is whether all people on a particular job are to receive the same pay or different pay; and if different, on what basis and how? These are not trivial questions.

The great majority of workers are paid through systems that provide for variable payment for the jobs. Such systems reflect the realization by management and employees that it is important to reward more than just minimal performance on the job. Thus management seeks to reward performance through merit-based and incentive pay systems, while employees and their unions ( if any) seek to have learning, proficiency and seniority rewarded.


The decision to participate assumes maintenance of an equilibrium between the inducements the organization offers and the contributions the person is asked to make. The organization must maintain, as a minimum, a balance of these two in the mind of the person, and, more realistically, a balance in the person's favor.

The ideas of J. G. March and H. A. Simon have been translated into equity theory. Pay system decisions can be regarded as focusing on individual equity. Equity theory states that a person compares his or her “inputs” or contributions with the “outcomes” from participation (I/O ratio). When this is hard to do directly, the person compares his or her I/O ratio with some other I/O ratio. Anything the person perceives as relevant goes into these input and output considerations.

Inputs and Outputs

Compensation decisions often focus upon the value of the job, both in the marketplace and within the organization. Although these are critical input factors, neither organizations nor individuals would be satisfied by making the employment exchange solely on this basis. To explain, compensation inputs can be classified into three general areas — job, performance and personal.2 Pay system decisions must incorporate the performance and personal factors into compensation, in order to provide a regular paycheck perceived as equitable to the employee.

Equity as a Cognitive Process

Experiencing equity is a cognitive process. People's perceptions determine whether their pay situation is equitable. Not all individuals within an organization are likely to perceive their pay situation the same, nor is the organization (through its management) likely to see the situation the same as the employees. This makes the creation of equity in the organization a difficult and recurring problem, not one that is determined once and for all.


Organizations are not powerless in this cognitive process. They can influence the perceptions of the person in a number of ways. First, they can define clearly the inputs required of the person. This allows the person to accept or decline the exchange in the way that a student stays or leaves a course after the professor hands out a syllabus. Second, organizations can affect (through communication and influence) the inputs and outcomes the person focuses on. Third, they can make certain responses to inequity more likely to occur than others. If an organization wishes to retain people, it may make quitting an unattractive way to solve feelings of inequity.


D. Katz claims that organizations seek three things from employees: (1) membership, (2) role behavior and (3) innovative and spontaneous behavior.3 Membership includes remaining with the organization and being present for work regularly. It provides consistency to the organization's labor force and reduces staffing and training costs.

Role behavior consists of doing the job as it is described and/or assigned. This is also needed for consistency and coordination of activities within the organization. To the extent that role behavior is explicitly spelled out and is seen as the basis for the person's input to the organization, this requirement is also covered under the decision to participate. However, not all required role behavior is easily spelled out in jobs, and all jobs have areas of discretion that allow the person freedom in accomplishing tasks.4

Innovative and spontaneous behavior addresses the organization's need for the person to adapt what he or she is doing, and how it is being done, to the constantly changing circumstances within the organization. Clearly this requirement is not covered in the decision to participate.

The decision to produce, then, moves the person beyond the minimum required just to maintain membership. It is what most managers call motivating their employees. A useful framework for this decision is provided by expectancy theory.5 This theory has three basic parts: (1) valence, (2) the performance-reward connection and (3) the performance-effort connection.


In expectancy theory valence means the strength of a reward. Does the person want the reward the organization is offering? Since our subject is pay, we can be confident that the answer is yes ... but not the same size yes for all people. People differ in how valuable money is to them compared with other things on and off the job. Content theories help us understand how people's need for money may be very different. The advantage of pay as reward, though, is that it is seen as a path to many different types of need satisfaction.

How much increase or difference in pay does it take to make the person respond? This is the difficult question of the proper size of a meaningful pay increase (SMPI).6 The organization must worry not only about whether pay is a motivator but also about whether it is offering enough to make it worthwhile for the person to produce beyond the minimum. As with the value of pay, the appropriate SMPI differs with a number of characteristics of the person, including current pay, age, experience and type of job.7

The Performance-Reward Connection

This may be the most important part of the decision to produce, since if the individual does not see the rewards he or she wants as being contingent on the behaviors or outcomes the organization wants, then the organization is not likely to obtain those outcomes. This connection would seem to be obvious, but in fact it is not. Managers find it difficult to always define the results and behaviors they desire. Also, it is difficult to measure and/or appraise whether these outcomes have occurred. In short, the definition of performance is difficult in and of itself.

The individual must understand what is requested and see its connection with the reward. This, like all understanding based on communication, is hard to realize perfectly. Most organizations claim they have a merit system of pay, but most employees do not perceive that merit is the primary basis on which pay adjustments are made. In some cases this perception is valid in that the organization says it uses merit but does not; in other cases the organization is rewarding merit but is not accurately communicating this fact to the employees.

The Performance-Effort Connection

People must feel that their efforts will affect their performance. This connection may seem obvious but it is not. There are many jobs in which variations in performance are impossible or inconsequential. To try to connect performance to reward in such jobs frustrates the incumbent. Also, individual effort is not a useful gauge in the many jobs whose tasks take two or more people to accomplish. Finally, the effort-performance connection highlights the fact that the person must perceive that he or she can adequately perform the task. All of these subjects should be taken into account in designing a pay system (and will be taken into account in some manner, even if by the default copying of some other organization's design and definitions).


The major way in which organizations allow for factors other than the job to enter into the determination of an individual's pay is to develop a range of pay for each job or grade of jobs. A rate range is a range of pay determined by the organization to be appropriate for anyone who occupies a particular job. A rate range consists of a minimum pay rate (the beginning hire rate), a midpoint (the market or job rate), and a maximum (the highest rate the organization is willing to pay for the job). The following sections cover single-rate wage systems, the rationale for rate ranges, two types of rate ranges, the manner in which a pay rate is set for individuals within a range, and the dimensions of range rates.

Single-Rate Wage Systems

Before discussing various aspects of rate ranges we should first consider situation in which there is no range. There a single rate is paid for the job and the individual receives just that rate. This pay rate is the market rate and may be paid to either a job or a pay grade. This is illustrated in figure 16-1, option a.

Figure 16-1. Alternative types of rate ranges

If a job rate is used, the wage line provides the job rate. The individual is paid in accordance with the number of points assigned the job by the job evaluation system, by the competitive value discovered in a review, a salary survey (see, or by the competitive value provided by a research analysis product such as ERI's Salary Assessor® software. Where the grade rate prevails, the individual is paid in accordance with the grade level assigned to the job.

This type of system is useful where performance variation and/or other personal characteristics are nonexistent or unimportant. Not all jobs allow for a significant difference in performance. Some assembly-line positions and lower-level service positions have very little discretion, so concern with differences in output or behavior are minimal. Other circumstances that lead to use of single-rate systems are (1) a strict technology that controls the output and (2) jobs for which the training time is short (a couple of hours or so) thereby making a learning curve inoperative. The individual in this type of system is paid for his or her time on the job and for completion of the job as directed.

Single-rate systems are simple to administer: once the pay rate of a person's job is identified, no further decisions need be made as to how much he or she is to be paid. The system can operate successfully if (1) there is little variation in output and (2) it is acceptable to the parties involved. Unions often like single rates because they eliminate judgment-based differences in pay.

Rationales for Rate Ranges

Any time individuals on the same job differ significantly in performance or personal characteristics that are perceived as relevant to either the organization or the person, differentiation by means of rate ranges may be in order. One study reported that the rationale for rate ranges in most large organizations was the need for performance differences, but in some cases industry practice was a major reason.8 Thus labor-market demands may also be a significant factor.

Rate ranges can serve other purposes for organizations. Retention is one of the most important of these. Experienced personnel can be made difficult to hire away by paying them above the market rate for the job. This is seen by the person as a significant reward for membership. Where there is a significant quality variation among people on the job, a rate range may represent an attempt by the organization to retain the best employees by paying them on the basis of quality.

Although performance is the reason most often given for rate ranges, this rationale should be scrutinized. Is movement in the range in fact related to performance? One major study challenged this assumption and found that performance was a very poor predictor of pay rate.9 There must be more than just an actual connection between pay rate and performance: there must also be a perception by the individual that this connection exists. The need for this perception makes communication very important in pay systems.

A further rationale for rate ranges is employee expectations. Few people are content to make the same wage and be dependent on changes in the total wage structure. for raises In particular, they may see that length of time on the job is an important input and expect a reward for it. But they may also see a number of factors other than performance as relevant to movement within the range. Personal factors having to do with the job are a good example. For instance, many employees who are going to school part time perceive that they should receive something for this. Employees may also perceive that they should receive more pay for a variety of non-work-related factors. Some of these factors, such as the birth of a new baby may be very important to the person but seen as irrelevant by the organization. Others, such as the person's sex, may be illegal to use as a differentiator of pay. It should also be noted that although some employees perceive the need for a rate range, they do not feel that performance should be the basis for this range.10

Another rationale for rate ranges may be collective bargaining. In contract negotiation the organization may agree to rate ranges or to an expansion of rate ranges as an alternative to a general increase. The union is likely to bargain for ranges in terms of movement within the range by seniority. The connection of performance and reward is not well served in this case.

Finally, the Internet has produced a wide array of sources by which employees can gain access to information regarding the competitive pay for their positions. While in the 1990's employees knew little about the competitive value of their jobs, the plethora of job/career information freely available on the Internet has changed this. If there were ever a reason for organizations to have a formalized method of administering salaries, it would be to forestall the number of hours wasted by management trying to disprove inflated salary averages reported on free Internet sites. More importantly, management must protect the organization’s bottom line by guarding against overpaying employees based upon the high rates reported by Internet sites focused on increasing their visitor hits to enhance their IPO values. The future challenge of compensation managers is clear for the next ten years ... employees walking into their offices with salary increase requests based upon free data from the Internet.

Types of Ranges

Having made the argument that rate ranges are useful and expected, we turn to how to develop rate ranges.

Step ranges

A common form of pay range consists of a series of steps, usually a specified distance apart, either in percentages or flat amounts.11 Step ranges may vary considerably in number of steps and the total range the steps cover. Clearly these two in combination will determine the size of each step. The point is that there are three variables present, and the determination of any two will decide the third.

Two basic types of step ranges are common. The first consists of a starting rate and a job rate (assumed to be the market rate), as in the single-rate system. New employees are brought in at the starting rate and then moved up to the job rate in a series of steps. If done properly, this movement corresponds with the learning curve of the job. The market rate is the maximum, since it is assumed that once the person has learned the job, performance differentials are minimal. This kind of system is illustrated in figure 16-1, option b. In this situation there would be a number of steps, most commonly three, between the starting rate and the job rate. This type of step system is most common in semiskilled blue-collar jobs.

The second type of step system places the market rate not at the top of the range but in the center of it. Other places, such as the one-third point or the two-thirds point, are also possible, but the middle is the most common. Employees are hired at the starting rate, as in the other step system, and progress to the midpoint over time is on the basis of learning job proficiency. Thus, a person at the midpoint of the range is assumed to be a satisfactory performer. Movement above the midpoint is assumed to be for performance, or other characteristics beyond the normal or average. This type of system is illustrated in figure 16-1, option c. It is used in a wide variety of office nonexempt jobs and lower-level exempt jobs where performance is important but not critical.

These two types of rate ranges are not mutually exclusive in an organization. Lower-level pay grades may have the type of range that ends at the midpoint, while higher grades have ranges extending beyond. The rationale for such a system is that the discretion in higher-level jobs in the organization allows for performance differences not permitted in lower-level jobs.

Movement within grades will be discussed later, but one point should be made here. A person who is moved from one step to the next usually retains the new step even when the overall wage structure is changed. In this way, adjusting the wage structure to meet labor-market changes automatically becomes a general increase for employees in a step system.

There is a further consequence of this type of system: all people tend to move to the top of the grade over time. Even if movement is by performance, a person can eventually reach the top and stay there regardless of future performance. This phenomenon in turn has a dramatic effect on the total wage bill. In a period of normal growth and turnover the average wage for the job classification will probably match the market rate as people start to climb the ladder while others leave. But in a low-turnover, no-growth situation the organization may soon be paying above market rate even if it sets the midpoint of the range at the market, because all the employees in the job are in the top steps.

Open ranges

In order to focus more clearly on performance and to avoid the problems of step ranges, more and more organizations are using an open-pay range. In this system the organization defines the midpoint, the maximum and the minimum of the range. Any one employee may be paid anywhere within this defined range. The function of the midpoint, as in the second type of step system, is that the average performer would be paid at this rate. Also as in the second step system, new employees would start at the bottom and move to the midpoint as they learned the job and became average performers. Payment above the midpoint can be reserved for above-average performance.

Unlike the second step system, the person's wage is not automatically adjusted when the wage structure is adjusted. At this point, the person's performance is reviewed and adjustment is made in relation to that performance.

Figure 16-1, options d and e, illustrate two types of open pay ranges. Option d has a series of steps up to the midpoint and an open range above the midpoint; option e has an open range from minimum to maximum.

With the increased emphasis on performance in organizations, open-range systems are becoming more popular. They provide more flexibility than a step system in granting pay increases and are more resistant to automatic increases. Finally, open ranges not only may make it easier to reward performance but are also useful when criteria other than performance are to be used.

Dimensions of Ranges

Any wage structure has a number of rate ranges and pay grades. This number can be a matter of the policy of the organization. Small organizations tend to have a small number of pay grades accompanied by wide pay ranges, broad definition of job titles, a great deal of movement within pay grades, little overlap between grades and limited promotion to higher grades. Some organizations have many grades, which tends to create an opposite set of characteristics.

When examining pay ranges we can determine the total wage structure with the help of three characteristics: the breadth of the rate range, the number of pay grades and the overlap (see figure 16-2). If one knows the bottom and top of the wage structure, the slope of the pay line, and any two of the three characteristics just cited, the third will be determined.

Range breadth

The breadth of the rate range is the distance from the top to the bottom of the range ... a to b in figure 16-2. It is the vertical dimension of the range. The breadth may be stated in dollar amounts or in percentages. The latter is more common and will be used here. The breadth of the range should vary with the criteria for movement within the range. Assuming that performance is the criterion, the breadth would represent the opportunity for performance differences in the job. Where ranges are narrow, the assumption is that performance differences are narrow and vice versa. In practice, hourly jobs have ranges of 10 to 20 percent, office jobs 15 to 35 percent, and managerial jobs 25 to 100 percent.

Figure 16-2. Parts of a wage structure

Factors other than potential performance differences may also affect range breadth. Organizations that promote intentionally fast encourage narrow ranges, since people do not stay within one grade very long. A wide range is encouraged if adjustments need to be large to be noticed by employees. Higher grade levels tend to have broader ranges for this reason. Broad ranges can accommodate a wide variety of jobs, as well as variable starting rates among jobs. These broad ranges indicate that the process of determining the market rate is not a precise one.

Establishing range maximums is particularly difficult. There is some logical maximum value for any job, regardless of how well it is performed. Ideally when this point is reached the person is promoted, either to a new job or by upgrading the tasks of the present job. Unfortunately, this may not be possible at the appropriate time. Realistically the person should be told that this is as high as he or she can go in the rate range and that any further salary adjustments will come from general increases.

Some organizations provide steps beyond the maximum of the range. There are usually two rationales for this ... seniority and recruiting. Long-term employees who will never be promoted and whose performance remains good are sometimes granted longevity increases beyond the maximum of the range. These usually take place after five or ten years at the top of the grade. Trouble in recruiting and retaining professional and managerial employees can be ameliorated by starting these people quite a ways up in the rate range; in order to retain them the organization must go beyond the maximum to provide any significant movement in grade.

Number of grades

The total number of pay grades in the wage structure can be a result of other calculations (mainly range breadth and overlap) or a conscious decision that forces the other two variables to adapt. The number of pay grades is reflected in the horizontal dimension of figure 16-2 (a to c). At one extreme, a structure with a single pay grade would have a minimum and maximum embracing the total wage structure and would include all jobs. At the other extreme, each job evaluation point on the horizontal axis would constitute a separate pay grade. In the latter circumstance two jobs would occupy the same pay grade only if they had identical job evaluation points ... a situation that would assume a very accurate job evaluation plan.

A large number of pay grades often coincides with a narrow range, permitting a large number of promotions and multiple classifications in job families in the organization. A small number of pay grades allows for flexibility, in that it assigns people to a wide range of jobs without changing their pay grade. Not surprisingly, number of pay grades is associated with size and number of levels in the organization. It also seems reasonable that organizations with a fluid, organic structure would have a minimum of pay grades whereas more structured and bureaucratic ones would have more.

Clearly there is no optimum number of pay grades for a particular job structure. In practice, the number of pay grades varies from as few as 4 to as many as 60. But 10 to 16 seems to be most common. With few grades there are many jobs in each grade and the increments from one grade to another are quite large. The presence of many grades has the opposite characteristics.

A number of considerations help to determine the appropriate number of grades. One is organization size: the larger the organization, the more pay grades. A second is the comprehensiveness of the job structure. A structure that covers the whole organization will tend to have more pay grades than one that deals only with one job cluster. Third, the type of jobs in a structure makes a difference. Production jobs whose pay policy line is relatively flat will tend to have fewer pay grades than a managerial structure that has a steep slope. The last determinant is the pay-increase and promotion policy of the organization. A large number of pay grades allows for many promotions but entails narrow ranges and a narrow classification of jobs. A small number of pay grades, accompanied by wide ranges was traditionally thought of as unreasonable in that cost control of salary administration would be lost. In the late 1980's, this reasoning was badly shaken.


The final pay range determinant is the degree of overlap between any one pay grade and the adjacent grade (c to d in figure 16-2). Overlap allows people in a lower pay grade to be paid the same as or more than those at a higher grade. The rationale for such a phenomenon is that a person at a lower pay grade whose performance is very good is worth more to the organization than a new person at the higher pay grade who is not yet performing effectively. This reasoning seems to work: seldom are there complaints about overlap.12

As with the number of grades, overlap can be either a determining variable or the determined variable. Overlap will work well where there are many wide pay grades. A conscious decision to keep overlap to some maximum (such as 50 percent) will reduce one of the other two variables.

Some overlap is desirable, but there are problems. The main one comes about in promotions. A person high up in a rate range who is promoted may start in the new rate range higher than the job rate of the new grade. But not to give the promoted person a pay raise is hardly to have promoted him or her. Organizations generally set some policy that any promotion be accompanied by some specified minimum increase, such as one step in the new rate range or a specified percentage. The designers of career paths in some organizations reduce this problem by placing the next job in the sequence more than one pay grade above the present one.


Rate ranges make possible different pay rates for individuals in the same job and/or grade level. Operating such ranges calls for some method that differentiates between employees. Such a method must provide a decision framework for positioning each person within the range.

Open rate ranges facilitate a pay-for-performance approach to individual pay determination. The present section will focus on movement within grades in a step system. It should be noted, though, that an open range system can also accommodate the methods of progression discussed.

Step Rates

Most government and some private organizations divide their entire rate range into a number of steps. (One should always be aware of the influence of government systems in compensation. For example, with half the paychecks in Canada being written by governmental agencies, one cannot overlook these step approaches.) This number is a function of the breadth of the rate range, the time required to achieve proficiency in the job, whether there are to be steps beyond the market rate, and a determination of the size of a meaningful pay increase. At least three steps are almost always used. A general step system is illustrated in option c in Figure 16-1.

Step rates facilitate the granting of pay increases by determining the amount that any increase will take. Of course, it may be possible to move a person two steps, but this is always done in predetermined amounts. Such increases can be considered a disadvantage as well as an advantage. Many organizations prefer to be able to grant a wide variety of increases to better relate pay to their pay-increase policy.

Methods of Progression

All methods of progression specify how a person moves from the bottom of the range to the top of the range. The major difference among them is the criteria for movement. The major methods are automatic progression, a combination of merit and automatic progression and merit progression. An organization does not have to restrict itself to only one method; it may use different methods for different jobs or even different methods for a single job at different parts of the rate range.

Automatic progression

This type of progression (sometimes referred to as scheduled increases) consists of wage increases based automatically on length of service. In some situations, such as basic industries, there are a small number of increases often in rapid succession (every three months) to the maximum rate for the job. These are jobs in which proficiency can be gained in a short time. On the other hand, some governmental organizations may have many steps (five or more) and grant increases once a year. In these situations longevity on the job leads to higher proficiency, and the organization wishes to reward continuity of employment.

A major source of variation in automatic plans is the nature of the maximum rate ... whether it is the market rate or an above-market rate. Organizations that move only to the market rate tend to have rate ranges with a small number of steps and a short time frame for progression. They are interested not so much in rewarding longevity as in encouraging learning the job. Organizations that move beyond the market rate are specifically rewarding longevity on the job; they tend to spread out the progression to the top of the grade over a long period.

Automatic progression does not have to be totally automatic. A fully automatic progression plan is actually a variation of the single-rate or flat-rate system. If all employees can expect to reach the maximum of the rate range after a given period on the job, the assumption is that the maximum is the real rate for the job.

Variation can be introduced in two ways. First the time period may vary from step to step. For instance, some systems move people rapidly to the midpoint and then much more slowly; the extended steps beyond the midpoint are clearly tied to longevity. The second variation introduces a little merit into the system by either denying movement to the next step for poor performance, giving good performers a double-step jump, or shortening the time period between step increases.

Merit considerations in automatic plans should not be overemphasized. The system is designed to be automatic, and variations are seen as exceptions, not the rule. In most systems that allow either movement ahead or denial of increases, these alternatives are rarely used: the problems they pose for administration of the workplace are not perceived by supervisors to be worth the advantages they offer. Unions commonly accept rate ranges but insist on automatic progression and encourage maximum rates that are above the market rate.13

Organizations make much more use of automatic progression than might be assumed. Studies indicate that in most areas of the country and in most industries, automatic progression is the norm and not the exception.14 But this may be changing. The emphasis on productivity in the United States is translating itself into a search for ways to make employees more productive. Focusing on performance instead of longevity is part of this trend.15

Combinations of automatic and merit progression

We have just seen that some introduction of merit is possible even in automatic progressions that focus on longevity. It is possible also to design progressions that try to balance merit and longevity. These progressings usually let employees focus on different criteria at different places in the pay range.

Probably the usual combination is automatic progression to the midpoint ... the market rate ... and progression beyond the midpoint on the basis of merit. The rationale for this method of progression is that all employees can be expected to reach average proficiency within a certain time on the job; this period matches the automatic movement to the midpoint. However, not all employees exceed average performance on the job, and movement from the midpoint on should be based on performance that is above average. If the organization does a good job of matching time taken to reach the midpoint with time taken to reach proficiency in the job, then labor costs are equalized; if these are out of balance, then labor costs are higher or lower than is optimum.

The rate range can take one of two forms in this case. The first looks like option c in figure 16-1, with a series of steps from bottom to top and the market rate as the middle step. The distinguishing feature of this form is how movement is determined after the midpoint has been reached. In the second form there is a series of steps up to the midpoint but an open range from that point on with movement of any degree possible and decided by merit. This form is illustrated in figure 16-1, option d.

Another method is to combine longevity and merit at all points in the range. Under this arrangement all employees receive an automatic adjustment, but those with above-average performance receive more, such as a two-step jump. It is also possible to hold back those who are not performing well. The latter action is rare but can be effective in probationary situations.

The areas of prevalence of these different methods are hard to determine. It appears that automatic methods are most typical of factory jobs and combination methods most typical in office situations.16

Automatic-progression methods are simple to administer since they are purely mechanical adjustments made by time in grade. Introducing merit complicates the pay decision by adding a judgment about how well the person is doing the job. Then a way must be developed to incorporate this judgment into a wage increase. This makes administration more complex and, if the judgments are perceived as arbitrary, raises concerns about the equity of the system. The advantage is that a connection is made between performance and reward, and this may be worth the trouble.

Merit progression

A pure merit progression employs an open rate range with only the minimum, maximum, and midpoint defined, as in option e in figure 16-1. Movement within the range is based strictly on performance, and there are no adjustments for general increases. This pay-for-performance system requires an integration of performance appraisal with pay determination. What we cover here is movement between steps of a pay grade, as in figure 16-1, option c, on the basis of merit. The rationale for merit progressions is that the movement to proficiency is actually an improvement in performance and should be treated as such; people differ in their rate of improvement to proficiency, and this should be taken into account; it is performance that the organization wants and should pay for.

In practice, a merit progression is usually a combination of merit and longevity. The initial decision to move a person from say, step 3 to step 4 is based on performance, but from that time on the person retains step 4 when adjustments to the wage structure are made, thereby remaining at the same relative position in the range. If step 4 is one step above the midpoint, the assumption is that this person is always above average in performance, but actually the person needs only to maintain a level of performance that will not result in termination. Further, unless the performance-appraisal system is tied consistently to the merit pay adjustments, either the system tends to be seen as arbitrary or supervisors tend to grant the same increase to all employees and thus destroy the performance-reward connection.

In a bad economy. In all step systems most employees eventually get to the top of the pay range. In a merit progression method the good performer should get there faster than the average or poor performer. This phenomenon of getting to the top of the range tends to be hidden when the organization is growing and times are good. But when growth stops, then promotions slow up, employees stay on their current job, movement to the top of the range is accelerated, and the organization finds that all employees are at the top of the range. Labor costs thus become very high at exactly the time the organization can least afford them. From the employee's perspective, the only pay increases received are those that occur through wage structure adjustments, and these are likely to decrease in these circumstances. This lack of wage increases makes the potential for feelings of inequity increase considerably.

Actual practice. Most organizations and their management claim that they use a merit progression system. But studies show that up to 80 percent of employees are at the top of their rate range.17 The problem is compounded when management mixes up general pay increases with merit pay. Granting all employees the same pay increase and announcing it as a merit increase destroys the concept of merit. Lower-level supervisors, in particular, find it uncomfortable to deal with merit pay, which requires him or her to make competitive distinctions between employees. For these supervisors it is often cooperation and not competition that is important. Because of the inflation of the late 1970s, annual pay increases are almost institutionalized in organizations today. This makes merit progression something of a misnomer, especially where organizations simply call all pay adjustments merit increases.

Union acceptance. Unions generally do not support merit-progression systems (as was the case in the Year 2000 when the NEA voted against merit pay for teachers). They question the objectivity of performance criteria and see the supervisor rewarding things other than getting the job done. Further, they are interested in getting their members to the top of the rate range as fast as possible. Unions can complicate the merit system through grievances. Some unions will automatically file a grievance if all members do not receive an adjustment or if they do not receive the maximum adjustment. This not only increases administrative costs but considerably burdens the performance-appraisal system.18

Nonunionized people in the organization look at what happens to union members, and management knows this. Therefore, management tends to give those not in the union what union members received and maybe a little more. Organizations do try to deal with their nonunion sectors more on a merit basis than on a longevity basis, and to the degree that above-average employees receive more, the merit principle does work.19

Rate Ranges and Recruitment

To this point we have assumed that the organization has been hiring people who are just qualified and moving them up in the range as they learn the job. But what if it hires a person who can do the job from the beginning? Clearly this person should be hired at the market rate (the midpoint). In actuality, then, people are likely to be brought into the organization anywhere up to the midpoint of the range, based upon their qualifications. Thus a system that ends at the market rate has a flat rate for hiring fully qualified employees.

The labor market may complicate the rate range when there is a shortage of applicants. When it is hard to recruit, one way organizations adjust is to raise the starting pay to wherever in the range it must go in order to obtain people. This may result in hiring rates at the top of the rate range or above. This extreme situation makes any upward movement within the grade difficult or impossible for the person. A person who is then expected to stay in the grade for three or more years before promotion can only look forward to general increases.

Correcting Out-of-Line Rates

The rate range defines the minimum and maximum that a person may be paid for a given job. For a number of reasons an individual's pay may be more or less than the prescribed range. The organization needs policies for dealing with these out-of-line rates.20

Terms of the trade

Many a new compensation analyst has been tested by management with the question, "Do you know what a green circle is?" This question separates the college student from the practitioner. It refers to the case in which a person is paid less than the minimum of a grade. This occurs, for example, when a person is promoted into a position in a higher pay grade, but not given a pay increase (because all increases may have been frozen by top corporate management).

Underpaid employees. As stated, a person paid below the minimum of the rate range for his or her job is said to carry a green-circle rate. This situation usually occurs when the wage structure is changed upward and the individual was at the bottom of the rate range. Little question exists regarding the appropriate response: the underpaid employee should have his or her pay raised to the minimum of the range, immediately if possible or in a couple of steps. If the person is performing adequately, the difference between his or her rate and the minimum of the range should be made up by the employer.

Of course it is possible, for a number of reasons, that the employee is not worth the minimum of the range. Even so, there are usually adjustments that can be made. For instance, if the labor market is very tight and marginal workers must be hired and retained, a lower classification involving job redesign to accommodate the person's skills would be in order. This same reasoning could apply to older and handicapped employees who cannot fully carry out their jobs. On the other hand, redesign may be unnecessary where there is already a lower-level job to which the person can be assigned. Or a trainee rate may be appropriate if the employee is still learning the job.

Usually there will be a few underpaid employees, and a policy of bringing their rates into line immediately protects the integrity of the pay system. But if many employees are underpaid, a careful review is required: not only may the costs of adjustments be high but also equity between the newly raised employees and other employees on the job may require a phasing in of increases. Also, all underpay situations should be examined for racial or gender discrimination.

Overpaid employees. A person paid above the maximum of the range for his or her job is said to receive a red-circle rate. Other names for this situation are ringed, flagged, or personal rates, red allowances, overrates, and personal out-of-line differentials. The variety of terminology suggests that this is a common problem in organizations, that it stems from a number of sources, and that it is more difficult to deal with than the problem of underpaid employees.

Solutions to overpay vary from doing nothing to reducing the pay to the top of the range. Both approaches can cause equity problems, both in others and in the person affected.21 The most common solutions are the following:

  1. Freeze the pay until general increases catch up with the current pay.

  2. Transfer or promote the person to a job in an appropriate pay grade.

  3. Freeze the pay for a limited period, such as six months. Then attempt either of the previous strategies. If this is unsuccessful, reduce the pay at the end of the period.

  4. Red-circle the job and not the person.

  5. Eliminate the differential after a period such as a year or gradually over time.

A number of less common arrangements also exist. One, the adder, is a payment to the employee in quarterly installments of the difference between his or her rate and the maximum of the range. The employee is given 100 percent of the differential the first year, 75 percent the next year, and so on until there is no differential. The advantage of the adder is that the top rate for the job is made clear and both the person and the organization are aware of the exceptional and temporary character of the differential.

Another possible solution is a lump sum payment. For example, the employee may be paid the difference times 2080 hours and have his or her pay rate brought immediately into line. Any solution to overpay involves questions of equity. Overpayment is usually not the fault of employees, and any reduction in pay will be seen as unfair by them. On the other hand, there is also the perception of equity by other employees, so some action is always called for. All the actions just described try to balance these two perceptions in arriving at an equitable solution. Failure to correct red-circle rates means that range maximums are meaningless; people may be paid more than their job and performance are worth to the organization; and organizational resources are being diverted into paying these rates rather than rewarding others' good performance.


The pay rate of an individual reflects a number of considerations, of which performance is only one. Other variables found to influence pay are the person's performance appraisal, pay history, present position in the range, and experience; the time since the last pay increase; the amount of that increase; pay relationship within the work area and other parts of the organization; labor-market conditions; the financial condition of the company; and of course the previous decisions regarding wage level and structure. The interaction of these forces determines whether a person receives an increase, and if so the amount of that increase.

Linking Pay to Performance

Judging from this list of variables, it is clear that an organization that claims it uses a merit system is likely to be exaggerating somewhat. Although almost all companies would claim that performance is the primary variable in their determination of individual pay, not many have a system that directly links pay to performance. One study of a large organization showed that "the careers that people make for themselves in large-scale organizations attenuate the role of pure performance in pay."22 However, it is becoming increasingly important that organizations do connect performance with pay.

Linking pay and performance is difficult at certain times. During the late 1970s and 1980s, when inflation was running rampant, an organization had to offer very large increases to be seen as rewarding merit and not just keeping its employees up with inflation. Even if an organization is committed to pay for performance, its employees are the ones who have to perceive the relationship.


One particularly sticky problem is that of wage compression. This occurs when new people are brought into a pay grade at the same, a higher, or even a somewhat lower rate than people currently in it. This is most obvious in the case of new hires who are brought in at pay rates almost the same as those of employees who have been there a year. Rates for new hires are determined by the external labor market. Unless one pays that amount, the new employee will not accept the job offer. Current employees have their wages set by the internal labor market, which is an administrative decision. As we have noted, the particular pay rate for an individual is determined by a number of factors, of which the market is just one. The result is that new hires make too much in relation to those already working. This was especially true for new college hires in 1999 and 2000, when they were paid historically high rates.

Compression is also likely to occur with first-line supervisors of nonexempt employees who are paid overtime; with sales managers, whose sales staff can make more selling on commission than the manager; and with middle management, who are squeezed between top management and the increases given to lower-level employees. The last is very evident in government jurisdictions. All three examples differ somewhat from the case of new hires, in that they involve a hierarchy, and the perception of unfairness is related to an inadequate distance between organizational levels.

Solutions to compression depend upon what type it is and how serious it appears to management. One obvious solution is to ignore it. This is possible if people are moving rapidly and the problem is mostly one of timing. The person feeling the inequity can be told that it will disappear shortly. A second possible solution is to adjust the internal structure more completely to the external realities. This may be an expensive alternative but is necessary if the organization is experiencing turnover and employee discontent. In the set of three examples just cited, the most likely solution would be a policy statement that a particular distance ... say 15 percent ... be maintained between levels. Rather than change the rate range for the supervisory jobs, however, organizations often pay this differential as a bonus based upon the wages of the subordinates.23

Integration of The Wage Structure and Individual Pay Determination

Changing the wage structure results in more money being spent on wages by the organization. This usually translates into pay adjustments for employees. But there needs to be some way for these two disparate events to come together. The vehicle for this is the budgeting process. On the wage structure side, what is required is an indication of how much of a change is to be made in the structure and how much money that will take. That money in turn becomes the organizational input data for individual pay determination. The question now is how to allocate the money provided by the wage structure adjustment.

Budget Allocation

The design decisions discussed provide a framework for deciding how the budgeted money is to be spent. Where a single-rate system is used, the pay adjustment is a general increase. The basic question then becomes one of timing. When should the general increase be granted? If an increase is given at the first of the budget period, then the percentage of the wage structure movement is the same as the general increase. But if the general increase is held off, the percentage can be larger and still fall within the budget. Remember, however, that this larger percentage is built into the next year's budget.

In organizations using an automatic-increase system, a change in the wage structure changes all steps in all grades. But there is an additional cost ... the movement of people from one grade to the next. So the total increase in the wage bill will be more than the increase in the wage structure. The exact difference depends on the timing of the step increases and on estimates of turnover. If all step increases are granted at one time then the impact is even, but if they are staggered by some criterion such as anniversary date then the organization needs to prorate the increases depending on when in the year they are granted. For instance, a 5 percent step increase given a person on July 1 is a 2.5 percent change for the year. But again, these adjustments increase the total wage bill beyond the cost of the wage structure adjustment. On the other hand, turnover tends to reduce the total wage bill since replacements are ordinarily hired at steps lower than those occupied by the people who left.

Merit-progression systems add another layer of complexity to the problem. In automatic systems the increases can be planned because the variables are known. Performance ... the merit system variable ... is less predictable. Organizations deal with this by developing a budget for a sector that shows how much it can spend to increase the wage bill in that sector. The same kind of considerations now go into the planning of each of the other sectors. The major complication is that the increase amounts will vary considerably among people. Last, a decision needs to be made as to whether all increase funds will be allocated on the basis of merit or whether there will be a general increase and a merit pool.

Decision Makers

In a merit-progression program the supervisor becomes a key person in the pay decision, for it is he or she who decides upon the performance of the individual. Thus the pool of money available for wage adjustments is ordinarily controlled by the supervisor, to be dispensed within the guidelines provided by the compensation specialist. This supervisor must really believe in the value of a merit program for it to work. There are considerable pressures upon him or her not to allocate this money on the basis of merit. In brief, validating this decision in the minds of employees is difficult and may lead to feelings of inequity. In addition, supervisors are often much more concerned with cooperation than they are with outstanding performance.

An advantage of simpler individual pay determination is that the decision making can be more centralized and does not involve as much judgment. In this way consistency of treatment is maintained, which leads to feelings of equity. In an automatic-progression system the compensation specialist can make all the appropriate decisions and implement the program without having to coordinate their efforts with line management at all. This is convenient, but the program then becomes that of the compensation department. Line management feels divorced from the compensation program, perceiving that they have little ability to motivate their employees.

Even if line management has a say in this determination of the exact amount people are to be paid, there is a series of other decisions framing this decision and limiting its impact. These decisions start with the wage-level determination, the form and shape of the wage structure, and the design of individual pay determination.


A nontraditional method of compensation that is gaining popularity is that of paying for the job knowledge of the employee rather than the job.24 The focus of these plans is not the specific job the person is currently performing but the range of jobs the person can perform. This form of compensation has been made popular by some of the experiments in quality of work life, such as that undertaken by the Gaines pet food plant in Topeka, Kansas.25 In this type of system, employees are hired at a base rate that is determined by the labor market. Movement in pay then occurs as the employee learns new tasks used in creating the product. The top rate is for those employees who can do all the jobs in the work unit.

There are essentially two types of knowledge-pay systems. The first is a multi-skill plan. Here pay is linked to the number of different skills the employee learns and can perform. The second is an increased-knowledge plan, wherein the employee's pay is related to the increased knowledge required by a particular job category. The latter is a more appropriate plan where the jobs have a progression of difficulty and the employee can learn them over time. Top rates in this type of plan are for those employees who can act as troubleshooters and trainers of others. The increased-knowledge program is similar to the type of compensation system often developed for professionals: skill-based pay plans.

Administrative Issues

A number of issues are important in establishing a pay-for-knowledge compensation plan. Let's briefly consider some of them.

Progression or promotion

When a person learns a new or improved skill, is this to result in a promotion or an in-grade increase? If the latter, then there would have to be very few grade levels and very wide rate ranges to accommodate the number of skills to be learned. If the former, then there would have to be a large number of grade levels since the person would move up a grade with each new skill. Each grade level could be a single rate.


In a pay-for-knowledge system performance can be considered the same as learning. But what about those who learn better than others? If the progression is by grade and not by promotion, then a range can be included in each grade that can be used for variation in performance, which is defined as doing the task better. Furthermore, a person's skills may deteriorate over time if he or she does not use them. Some organizations provide for retesting or refresher sessions to keep skill levels up.

Whom and what

Deciding employees and jobs to include in the system is always a problem in pay-for-knowledge systems. The ideal situation for such a plan is a clearly defined production unit that produces a discrete product through tasks that all employees can learn. But even here there are positions around the edge, including that of the supervisor, on which inclusion decisions need to be made.26 Even if the whole production unit is included, there may be some tasks that all employees are not able to learn. The tasks that constitute a pay increase must also be delineated.

Maintaining stability

Pay-for-knowledge plans require job rotation. This can unsettle the production process if there are more employees learning new roles than employees well trained in their roles. Also, bottlenecks in rotation can prevent employees from moving up when they desire. In these situation, special rates can be used to hold the person until an opening occurs.

Maxing out

As in all pay systems, there is the problem of the person who reaches the top of the range, or in this case who has learned all the jobs. The only increases available to this person are those granted across the board. There is no real answer to this problem except for those employees who move up from their current roles.

Advantages and disadvantages

Pay-for-knowledge plans can have significant benefits, but they are not suited for all situations. Here are some of the advantages and disadvantages.

Flexibility. It is clear that organizations can obtain increased employee flexibility from this type of plan. The employee looks forward to learning and changing. Changes in demand can be adapted to more readily since employees can be moved to whatever task needs to be done at the moment. Staffing can be leaner since employees can fill in for one another. It is further argued that this flexibility leads to a higher quality of output since the employees know all of the tasks and can therefore focus on the overall product rather than on the specialized task.

Employee satisfaction and commitment. Lawler's studies seem to indicate that employees in a pay-for-knowledge setting have higher levels of job satisfaction, particularly satisfaction with their pay.27 This would seem to reinforce the point that employees emphasize their personal contributions more than the organization does. This higher satisfaction is claimed to lead to lower absenteeism and turnover.

Costs. The disadvantages are basically that the system is more expensive to operate. First, hourly labor costs are higher since the person is being paid for skills not currently being employed. Second, there are training costs, both in the design of the training and in the fact that trainees are performing the tasks. Finally, there are the administrative costs of keeping track of where these employees are and what pay rate they should have, given the training they have received.

Organizational integration. Pay for knowledge has often been associated with a broader program of quality of work life in the organization. These programs are usually considered experiments and as such are separated from the rest of the organization. The success of this type of plan, as with all new ideas, depends in large part on how top management views it. If it is supported, the probabilities of success are increased. But these types of plans are often perceived as threatening by top-level staff, and this can lead to the demise of the program. A major consideration is the equity question with other employee groups, the supervisors, and the labor market. Higher wages of pay-for-knowledge employees may not accord with the equal pay for equal work doctrine. They may also create compression with supervision, and cause companies to pay above-market rates for labor.


Organizations wish to pay for more than just the job that the employee does. Employees contribute both in terms of membership (staying on the job) and being productive while on the job. Both of these sets of contributions need to be rewarded by the organization. Wage structures deal with rewarding these sets of contributions by establishing rate ranges for jobs. This allows for variable pay rates for employees on the same job and/or in the same pay grade.

The breadth of the rate range (distance from top to bottom) is a matter of judgment for the designer of the wage structure. Further, the decision is interrelated with other factors in the wage structure, namely the distance from top to bottom of the entire wage structure, the number of pay grades, and the amount of overlap between grades.

The design of rate ranges may vary from a structured set of steps a given percentage apart to an open range in which only the minimum, midpoint and maximum are defined. Picking the type of range depends largely on the factors that the organization wishes to reward. Step systems do a good job of rewarding membership and seniority. Open ranges allow the organization to more clearly recognize variable performance. There is an aspect of rewarding both in either case, so the choice is one of emphasis and not of kind.

In administering the movement of employees within rate ranges, compensation specialists face a number of problems. Recruitment in the labor market may require the organization to hire new employees at advanced positions on the range. This in turn can lead to compression, as current employees are paid less than new employees. Keeping employees within the rate range is a constant problem. One of the most pervasive problems is keeping the focus of increases on performance; supervisors and employees alike are more comfortable with seniority increases. Last, while other aspects of compensation administration are often centralized in the hands of compensation staff, the determination of pay increases within grade must involve all supervisors in the organization.

We have also examined a radically different type of pay system, that of skill-based pay. In this system employees are paid for the range of skills that they bring to the job that are useful in performing the job. As employees learn more skills they are paid more. These types of plans can provide the organization with a well-trained work force, flexible as to work assignments and interested in the work. It can also be more costly, require too many people in training, and be difficult to integrate with the traditional wage structure of the organization.

1 J. G. March and H. A. Simon, Organizations (New York: John Wiley, 1963).
2 D. W. Belcher and T. J. Atchison, "Compensation for Work," in Handbook of Work, Organization, and Society, ed. R. Dubin (Chicago: Rand McNally, 1976).
3 D. Katz, "The Motivational Basis of Organizational Behavior," Behavioral Science, April 1964, pp. 131-33.
4 E. Jaques, Equitable Payment (New York: John Wiley, 1961).
5 See also V. Vroom, Work and Motivation (New York: John Wiley, 1964).
6 L. A. Krefting and T. A. Mahoney, "Determining the Size of a Meaningful Pay Increase," Industrial Relations, 1977, pp. 83-93.
7 F. Krzystofiak, F. Newman, and L. Krefting, "Pay Meaning, Satisfaction, and Size of a Meaningful Pay Increase," Psychological Reports, 1982b, pp. 660-62.
8 U.S. Department of Labor, Salary Structure Characteristics in Large Firm, Bulletin no. 1417 (1963).
9 M. Haire, E. E. Ghiselli, and M. E. Gordon, A Psychological Study of Pay, Journal of Applied Psychology Monograph no. 636 (1967).
10 L. D. Dyer, D. P. Schwab, and J. A. Fossum, "Impacts of Pay on Employee Behaviors and Attitudes: An Update," Personnel Administrator, January 1978.
11 Much of this discussion of types of rate ranges uses information from R. J. Greene, "Which Pay Delivery System is Best for Your Organization?" Personnel, May-June 1981, pp. 51-57.
12 For a discussion of the proper overlap see B. R. Ellig, "Pay Inequities: How Many Exist in Your Organization," Compensation Review, third quarter 1980.
13 S. H. Slichter, J. J. Healy, and E. R. Livernash, The Impact of Collective Bargaining on Management (Washington, D.C.: Brookings Institution, 1960).
14 U.S. Department of Labor, Bulletin no. 1625-90.
15 E. E. Lawler, Pay and Organizational Development (Reading, Mass.: Addison-Wesley, 1981).
16 U.S. Department of Labor, Bulletin no. 1417.
17 B. K. Scanlon, "Is Money Still the Motivator?" Personnel Administrator, July-August 1970, pp. 8-12.
18 Slichter, Healy, and Livernash, Impact, p. 604.
19 Ibid.
20 See Ellig, "Pay Inequities."
21 G. W. Torrence, "Correcting Out-of-Line Rates of Pay," Management Record, September 1960, pp. 10-13.
22 T. H. Patten, "Merit Increases and the Facts of Organizational Life," Management of Personnel Quarterly, Summer 1968, pp. 33-38.
23z For some problem statements and solutions see R. Kemp, "Salary Compression Workshop," in Regional Conference Proceedings, 1978 (Scottsdale, Ariz.: American Compensation Association, 1978).
24 G.D. Jenkins and N. Gupta, "The Payoffs of Paying for Knowledge," in Labor-Management Cooperation Brief, U.S. Department of Labor, Bureau of Labor-Management Relations and Cooperative Programs, 1985.
25 R. E. Walton, "How to Counter Alienation in the Plant," Harvard Business Review, November-December 1972, pp. 70-81.
26 E. E. Lawler and G. E. Ledford, “Skill Based Pay,” Working Paper no. 84-18, Center for Effective Organizations, University of Southern California (Los Angeles, 1984).
27 Lawler, Pay and Organizational Development.