10 Contribution-Based Compensation

Air Force Research Laboratory

When talking about pay and performance, the late Dr. George Abrahamson liked to characterize managers and employees as performance co-conspirators. Dr. Abrahamson had been around the block a few times: Director of the Poulter Laboratory and Senior Vice President of the Sciences Group at Stanford Research Institute (SRI), chief scientist for the U.S. Air Force, and a return at the end of his career to SRI as senior technical advisor. But people at the Air Force Research Laboratory (AFRL) remember George best for inspiring AFRL’s contribution-based compensation (CCS) system.

Performance is traditionally defined as a measure of how well an employee accomplishes a set of tasks. An employee who performs assigned tasks well is considered a good performer; an employee who performs assigned tasks poorly is considered a poor performer. Managers strive to get the most work they possibly can out of their group of employees. A good manager plays on employees’ strengths by designing jobs that allow each employee to succeed. If a football player is best at blocking, you make him an offensive lineman; if he is good at dodging and juking, you make him a running back. But as George so often pointed out, the result is that employees always end up in jobs they can perform well. Hence managers and employees are co-conspirators—conspiring so that performance, defined in this traditional way, is always good.

The problem arises when pay is tied to this traditional measure of performance. If everyone succeeds in performing assigned job tasks, how does the measure of that performance provide any insight into how much an employee should be paid? Once the conspiracy concept is understood, as George so often pointed out, it becomes clear that the proper basis for determining pay is not traditional performance—how well an employee does what he or she is told to do—but contribution—how that employee contributes to the organization and its mission.

In the 15 years that AFRL has been using CCS, the federal government has tried, unsuccessfully, to implement pay-for-performance systems at the Department of Defense (DoD) and the Department of Homeland Security. Even in the business world, doubts have emerged regarding pay-for-performance, as studies have shown that it demotivates employees and interferes with meaningful feedback on performance and development. Given the current sentiment on pay-for-performance, AFRL’s accomplishments stand out even more prominently. For 15 years, AFRL has been carefully measuring the contributions of its employees and paying them more when they contribute more. For these same 15 years, leaders, managers, and employees have all reported extremely positive results and high levels of satisfaction with the critical elements of CCS.

How Does CCS Work?

The Defense Authorization Act of 1995 authorized DoD to establish personnel demonstration projects in research laboratories. This permitted those laboratories to experiment with HR management practices and policies that would not otherwise have been allowed under the statutes governing personnel management in the federal government (Title 5 U.S.C.). Using this demonstration project authority, AFRL designed its CCS and implemented it in March 1997.

AFRL’s CCS design reflects the contribution-based approach George Abrahamson first applied during his tenure at SRI. The fundamental design of the system is relatively simple, although the specific tools and processes used can seem complex.

Every year, individual employees are evaluated, much like they would be under a traditional performance appraisal system, except that the standards used for the evaluation are designed to measure the employee’s contribution to the mission and goals of AFRL rather than success in performing the specific objectives of a given job. As a result of this evaluation, each employee receives a contribution score that places his or her contribution along a broad continuum representing the full range of possible contribution levels at AFRL. The employee’s contribution level is then compared with his/her compensation level. If the employee’s contribution level exceeds his compensation level, his pay is increased so that his new pay properly reflects his current contribution level. If the employee’s contribution level is below his compensation level, then he receives no pay increase until his contribution catches up to his pay. That’s the basic design, as depicted in AFRL’s operational manual (Figure 10-1).

FIGURE 10-1 CCS Standard Pay Line Example

The y-axis represents the range of possible pay levels for the work covered by AFRL’s system. In this case, the range includes all possible pay levels for scientists and engineers; it spans the pay range for GS-7 to the pay range for GS-15. The x-axis represents the range of possible contribution scores AFRL’s scientists and engineers can receive, and it spans a score of 1 to a score of 5. The diagonal line in the middle of the chart, running upward from left to right, is known as the standard pay line (SPL). That line establishes the normal expected relationship between an employee’s contribution score and his or her compensation. For example, an employee with a contribution score of 3 should be paid $60,000, which can be determined by following the vertical line upward from the 3 score on the x-axis to the point where it intersects the SPL, then moving to the left to locate the compensation level on the y-coordinate corresponding to that point on the SPL. In this way, the expected compensation level for all employees in the system can be determined based on their contribution scores.

The two diagonal lines running in parallel to the SPL are known as the upper rail and the lower rail, and the space between them is known as the equitably compensated area. The purpose of these lines is to soften the implied precision of the SPL. What the upper and lower rails communicate, using the same example, is that an employee with a contribution score of 3 would be equitably compensated if paid anywhere between approximately $55,000 (the lower rail) and $65,000 (the upper rail), rather than suggesting that the employee must be paid exactly $60,000 (the SPL).

The chart also includes three dots, labeled “A,” “B,” and “C.” These dots represent three hypothetical employees who have been plotted on the chart to reflect their current compensation levels and their contribution scores. Employee A, for example, is paid approximately $40,000 and has received a contribution score of approximately 1.8. This places Employee A in the equitably compensated area, suggesting that Employee A’s pay is consistent with her level of contribution. Employee B, however, falls below the lower rail into an area known as the undercompensated area. This suggests that Employee B is not paid enough for what she is contributing and should be paid more. Employee C falls above the upper rail into an area known as the overcompensated area. This suggests that Employee C is paid too much for what she is contributing and should either step up her contribution or be paid less. The dashed lines point to the target pay range on the pay line based on contribution.

By plotting individual employees according to their contribution scores and compensation levels, AFRL managers can clearly identify where they need to focus their available compensation dollars to bring pay into alignment with contribution. AFRL has established policy on exactly how to make compensation decisions within this framework. The basic principle is that those in the equitably compensated range receive the general pay increase granted to all federal employees, and the funds that would have gone into step increases and promotions are now pooled and used to move those who fall in the undercompensated area into the equitably compensated area. Those who fall in the overcompensated area receive no pay increase. This is an oversimplification of the exact mechanics of CCS, but it reflects the general principles that guide AFRL’s policy.

Using CCS, AFRL plots the entire workforce according to their contribution and compensation (Figure 10-2). The outliers become clearly identifiable, helping AFRL understand where to focus its compensation resources and highlighting performance issues it needs to address.

Employees are constantly moving through the middle area between the rails. Those employees below the lower rail are most likely employees who have stepped up their contribution, moving their score to the right, and the chart indicates that their compensation needs to move up to reflect this increased contribution. For those above the upper rail, the message is usually received loud and clear that they either need to increase their contribution to justify their pay or they need to move to another position. Attrition rates for those in the overcompensated area run as high as 50 percent per year. While the overall pattern may remain similar from year to year, high performers are constantly moving up and low performers are constantly moving out, which is exactly what keeps those dots clustered in the equitably compensated area.

FIGURE 10-2 Overall AFRL Plot (2011)

Measuring Contribution

The hard part of CCS is how to measure contribution. First, this measurement is very different from the traditional concept of performance. Traditional performance measures quality, quantity, and timeliness. Traditional performance may also measure success in achieving agreed-upon results, such as producing a report or making a sale. What traditional performance almost never measures is the value of the employee’s work to the organization. Two different people may fully meet their performance goals; traditional performance doesn’t seem to care whether those two successes are of equal or wildly different value to the organization.

The concept of contribution is to begin where performance ends. While performance focuses on measuring how well you did what you did and assumes that its value is accounted for elsewhere, contribution ensures that you did your work well and then focuses on measuring the value of those work accomplishments. It’s all about impact of an employee’s work on the organization’s mission and goals.

Contribution tends to measure the same kinds of things that a traditional HR system measures through job evaluation systems for purposes of employee progression through defined career levels. For example, the traditional federal position classification system measures the level of knowledge a job requires, the degree of supervision required, and the scope and impact of the work. These sorts of factors are naturally aligned with the degree of impact the work has on the agency’s mission results. A job that requires a high level of knowledge, operates independently of direct supervision, and has broad scope is likely to have a greater impact on the agency’s mission.

The following specific factors that AFRL uses to measure employee contribution align closely with traditional job evaluation factors:

• Factor 1: Problem solving. Measures project scope and level of impact, technical complexity and creativity, and recognition of work.

• Factor 2: Communications. Measures contributions to scientific papers and reports, breadth of responsibility, level and diversity of audiences for scientific work, and level of oversight required.

• Factor 3: Technology management. Measures contract and technology management responsibilities.

• Factor 4: Teamwork and leadership. Measures role on project teams, breadth of influence, supervision exercised, and guidance received.

These factors are not measured along a qualitative scale with markers such as “acceptable” or “outstanding,” but rather along a value scale with four defined levels. The value scale spans the full range of contribution that might be expected for the broad career field of science and engineering as applied at the lab, including the contributions usually associated with new entry-level employees as well the contributions of experienced scientists and scientific leaders. By contrast, traditional performance management systems define a performance range that might be expected for a single job title at a particular grade level. Unlike a narrow performance scale, the contribution scale is a tool for total management of an employee’s career development and progression as well as for determining annual pay adjustments and bonuses.

At AFRL, this contribution scale is applied by teams of managers evaluating groups of employees within a major organizational unit. This approach ensures a high degree of consistency in how the contribution factors are applied. The contribution assessment begins with a self-assessment by the employee, with inputs from the employee’s team leader to ensure that all contributions have been captured and well articulated. The employee’s immediate supervisor then assesses the employee’s contribution against the contribution factors in the AFRL scale and assigns a preliminary contribution level for each factor. Following the immediate supervisor’s assessment, the branch head and all the section heads within the branch meet to review and calibrate the contribution levels of all employees in the branch. This group review by managers is repeated again at the division level, with the division head and all the branch heads participating.

Through this collaborative process, a contribution score is developed for each employee in each factor, along a scale of 1–5 and with up to one decimal place. For example, an employee might receive a score of 3.2 on one factor, 3.9 on another, and 2.4 on another. All the scores are averaged to determine the employee’s final contribution score.

Another value-added feature of the contribution-based approach is that it allows the organization to rank-order its entire workforce. All employees in a division, for example, can be listed in rank order by their contribution score as a way of intuitively validating the results of the contribution assessment and reaching agreement that the relative contributions of different employees have been accurately recognized in their respective scores.

Once the contribution score has been determined, the pay-pool manager (a senior military leader or senior executive designated to manage the CCS process for the pay pool) convenes a meeting with the division heads to review each employee’s compensation and contribution in relation to the standard pay line. Pay adjustments are determined, with the goal of moving all employees closer to the standard pay line.

AFRL Results

Since implementing the CCS system in 1997, AFRL has conducted detailed annual evaluations to assess its impact and effectiveness and to determine its potential value for other parts of the government. Fourteen annual evaluation reports have now been prepared.

The highlights of the impact the CCS system has had over the years include the following:

• Employee grievances gradually dropped by approximately 50 percent and formal complaints (equal employment opportunities and unfair labor practices) disappeared entirely for a seven-year stretch, briefly reappearing when CCS was expanded to include non-scientific and nonengineering employees.

• Comparisons of AFRL to control-group populations consistently show significantly higher levels of management satisfaction with the CCS approach (50 percent higher or more), particularly in areas such as authority to set pay and authority to advance employees.

• Despite added management flexibilities, salaries at AFRL have remained in line with the salaries for government scientists and engineers at comparable organizations such as NASA and the Department of Energy.

• Mean contribution scores at AFRL have risen over the life of the system. The dispersion of contribution scores has also increased, suggesting that the contribution factors have become more effective over time in differentiating among employees, even as mean scores increased.

Michelle Williams has been Director of the AFRL demonstration project office since 2004. She comments:

Advantages and Limitations of CCS

For most of us, if we think back on our careers, we will recognize that the progress we have made was driven primarily by our professional growth and maturity, not by the particulars we accomplished one year versus the other. Over time we increased our professional value because we demonstrated we could take on larger roles and contribute more to our organization. This professional growth was probably recognized by a series of promotions; any significant jumps we may have experienced in our compensation were the direct result of those promotions.

Contribution-based compensation is a systematic way of assessing, managing, and recognizing this long-term pattern of professional growth. That is the system’s true strength. In the past, similar approaches were called “rank-in-person” systems. Their focus was on recognizing the increased value individual employees demonstrated and letting that increased value drive the employee’s rank, rather than having rank be driven exclusively by the job an employee was assigned. CCS does exactly the same thing but within a defined framework that enables managers to determine and communicate contribution in a consistent and systematic way. It’s a rank-in-person system that can be easily understood and consistently administered. Exhibit 10-A offers a quick comparison of CCS with traditional pay-for-performance approaches.

If CCS is rightfully placed in the context of career progression, it quickly becomes immune from many of the criticisms traditionally leveled against pay-for-performance. After all, the critics of pay-for-performance are not critical of career progression and the importance of paying more for work of a higher level or value. Their criticism is focused on the more narrow practice of trying to pay more or less for the same level of work, through a measure of how well that work was done.

Daniel Pink’s recent book, Drive: The Surprising Truth About What Motivates Us (Riverhead Books, 2009), offers one of the better arguments against pay-for-performance. Pink contends that extrinsic incentives are not nearly as effective as intrinsic incentives. Extrinsic incentives, he says, can be counterproductive. Some of the scientific literature he cites likens the phenomenon to stage fright. People who can sing beautifully in the privacy of their home may choke when put before an audience. Likewise, employees may find that powerful financial incentives cause them to choke on the job and perform less effectively than they would in the absence of such incentives.

Pink argues that organizations would be much more effective if they focused on inspiring their employees by nurturing intrinsic incentives, rather than controlling them through intricate “carrots and sticks” performance systems that reward or punish specific actions or behaviors. He goes on to explain what kinds of intrinsic incentives are most effective, breaking them down into the following:

• Autonomy, the satisfaction of having some control over what one does and how one works

• Mastery, the satisfaction of getting better and better at doing something

• Purpose, the satisfaction of doing something that matters.

The way in which CCS differs from traditional pay-for-performance systems aligns with Pink’s notion of intrinsic incentives. Compare CCS to the three elements of intrinsic motivation that Pink identifies:

• Autonomy. A critical element of CCS is that employees have a high degree of freedom to shape their career growth and development. They are not placed in narrowly defined jobs with narrow performance standards. The general direction they are given is to further the research goals of the laboratory. CCS contribution factors simply provide a framework to help them understand how to contribute, over the course of their entire career, to the specific goals that are important to the lab. But employees remain free to define many of the details of their career path.

• Mastery. In CCS, increased contribution equals increased mastery. The contribution factors serve as a framework for defining and measuring mastery in the context of AFRL. Employees are driven by the intrinsic motivation of achieving mastery; CCS simply recognizes, after the fact, when that mastery has occurred.

• Purpose. Higher contribution also equals higher purpose. The progressive contribution levels within AFRL’s CCS serve as descriptors of higher and higher purpose—higher and higher impact not only on AFRL’s goals but on DoD’s goals and on the safety and security of our country. For those who are motivated by having an impact, CCS empowers them to find ways to have that impact, recognizes when they have done so, and pays them more for it.

Neither Pink nor any of the other critics of pay-for-performance are critics of paying some people more and other people less. They acknowledge, for example, that market factors are a legitimate basis for determining pay. They also agree that career and professional growth are legitimate differentiators of pay. The brand-new employee fresh out of school should be paid less than the experienced veteran who has demonstrated what he or she can do. This is exactly what CCS does. Another way to understand CCS is as a well-defined internal labor marketplace. It provides definition of the kind of work that is more or less valuable in the AFRL marketplace and empowers employees to work within that market to pursue their own goals and interests— knowing how the market will reward them for the things they choose.

CCS has worked exceptionally well at AFRL and is clearly well-suited to the laboratory environment. Many other DoD laboratories have adopted CCS or versions of it, including the Naval Research Laboratory, the U.S. Army Tank Automotive Research Development and Engineering Center, and the Naval Sea Systems Command. In addition, the acquisition workforce in selected organizations throughout DoD is compensated under an approach similar to CCS.

However, CCS is not for every organization. The system would likely not work so well in a highly structured production environment. It is well suited to knowledge workers and to organizations that perform creative, design, or strategic functions. Any dynamic business environment that requires employees to continually develop, grow, take on new tasks, adapt to change, and respond to new customer requirements is an excellent candidate for a contribution-based model. Contribution-based models pull employees out of their position boxes and place them in much larger mission and business boxes. The message is not: Do what is in your position description, do it well, and we will reward you. Instead, the message is: Take initiative to bring about results that impact the mission and we will pay you what that is worth.

Exhibit 10-A

Key Features of CCS versus Traditional Pay-for-Performance

In pay-for-performance systems, the performance appraisal is always conducted in the context of the job. It asks, “Did you do your job well?” Contribution-based pay systems look beyond the job. The question becomes, “What is the value of what you did in terms of accomplishing the mission?” If you transformed your job into one of higher value, it recognizes that. If you diminished your job into one of lesser value, it recognizes that. If you took on an entirely new job, it recognizes that as well.

Most performance appraisal measures focus on quality, quantity, and timeliness: “Did you do a good job?” “Were you productive?” “Were you timely?” Contribution assessments look at the level of work done and its value. In this sense, contribution assessment is similar to the old position classification evaluations designed to determine the appropriate level of a position. In fact, many of the contribution factors used in the DoD laboratories look strikingly similar to factors in position classification standards.

Since performance appraisal systems do not look at the level or value of work, they presume that the level-of-work dimension has been accounted for through a position classification. Position classification determines that a given job is at a certain level and is thus worth a certain range of pay. Performance appraisal then only needs to focus on helping determine where in that range the employee’s pay should fall, assuming higher performers are worth more in that range than lower performers.

Since contribution-based pay systems look explicitly at the level or value of work, they do not presume a position classification system that determines grade level. In fact, they replace position classification. Under contribution-based systems, there is no need for traditional position classification because an employee’s contribution score determines what band level his or her position is at, not vice versa. For this reason, contribution-based systems are ideally suited for broad pay-banding constructs. In broad pay-banding, important distinctions among different levels of work often get lost because pay ranges are so broad. Contribution-based systems provide a methodology for reintroducing level-of-work considerations into pay-setting.

Because performance is defined relative to the job, the factors and standards used to measure performance must also be job-specific. For example, an office assistant may have a performance factor that measures typing errors while a management analyst may have a performance factor that measures the depth and thoroughness of studies conducted. Because each is being measured in the context of the job being performed, it would be impossible to develop a common set of performance factors that apply to both the office assistant and the management analyst.

Under contribution-based systems, the opposite is true. A single set of factors and standards must be applicable to very broad sets of jobs in order to compare the value of work done and place it along a complete continuum. Both the office assistant and the management analyst would be measured against the same factors (e.g., leadership) to determine their relative value to the organization.

Again, because performance appraisal does not measure the value of work, it cannot be used as a definitive indicator of pay. Rather, it indicates only if someone should be paid higher or lower in the pay range. Contribution assessments, on the other hand, impact pay across the organization’s entire pay range. An employee’s contribution score answers the question, “How much should this employee be paid?” If the employee’s pay is higher than the pay range for the position, the employee should get promoted. The contribution score drives pay within the range for a band, as well as movement to the next band.

Under performance-based pay systems, employees who perform at the same level year after year will get the same performance rating year after year and thus the same accompanying pay increases. Under contribution-based pay systems, pay increases are the result of the delta between contribution and pay. So the employee whose contributions are at a given level year after year will continue to be paid at that level until the level of contribution increases.

Because performance pay involves adjustments only within a range, it is best understood as a symbolic reward; it tends to be relatively small. The big pay jumps occur by getting promoted to a job with a higher pay range. Contribution-based pay increases, on the other hand, reflect the results of a fundamental compensation review. Is this employee paid appropriately given the employee’s impact on mission and business goals? Consequently, contribution-based pay increases can be very large and are often the only way pay levels get adjusted. Under many contribution-based systems, getting promoted has no direct impact on pay. All pay increases must be earned by demonstrating one’s value to the mission. Promotions are simply an after-the-fact recognition that an employee’s value and therefore pay level have moved to the next band.

In a performance-based system, compensation is based primarily on the level of the position, with small increases possible based on performance. When employees get promoted to a higher level position, their compensation is increased, often substantially, based on a projection that, in the new position, the employee will do work that is substantially more valuable to the organization. In contribution-based systems, employees receive compensation increases only after they have demonstrated that they have actually worked successfully at a higher level and can contribute in ways that are of greater value to the organization.

Performance-based systems build off positions. The position defines the factors and standards for measuring performance. Positions define the pay range. The only way employees can escape the position is to get reassigned or promoted. Under contribution-based systems, the position becomes practically irrelevant. Employees can contribute very high on the scale or very low on the scale, regardless of their position. Contribution-based systems offer a structured way of moving to a person-based concept of personnel management.