863521

research-article2019
CBRXXX10.1177/0886368719863521Compensation & Benefits ReviewThomas and McKenzie

Article
Compensation & Benefits Review

Ten Years After: The

2019, Vol. 51(1) 27­–43
© 2019 SAGE Publications
Article reuse guidelines:
Development of a University sagepub.com/journals-permissions
DOI: 10.1177/0886368719863521
https://doi.org/10.1177/0886368719863521

Staff Pay System—Reflections journals.sagepub.com/home/cbr

and the Lessons Learned

Steven L. Thomas , Lyn M. McKenzie,

Missouri State University

Abstract
This article documents the development and implementation of a new staff pay system for a large,
comprehensive, public university. It discusses decisions that were made, alternatives chosen,
important process issues and outcomes, as a guide to administrators and human resource staff
into what can be expected as new job structures, pay and performance management systems
are developed. The authors review program successes and remaining challenges from the
perspective of 10 years after system implementation.

Keywords
compensation administration, job evaluation, performance management, wage and salary surveys,
pay for performance

By 2005, it was widely recognized that the inadequate, either for recognizing and reward-
current 25-year-old staff pay system at a major ing performance or for providing adequate
Midwestern university (referred to as performance feedback.
“Metropolitan University”) was no longer This article reviews the process of recog-
adequate to serve the needs of either employ- nizing the need for change in the compensa-
ees or the university. The job evaluation sys- tion system and describes the decisions made
tem no longer adequately captured tasks and and implementation activities undertaken to
work requirements for many jobs, particularly develop a new pay structure for Metropolitan
for information technology and other profes- University staff. We recognize that although
sional specialty jobs. The pay system’s com- many human resource professionals do work
pensable factors described some tasks and with compensation systems, relatively few
processes no longer performed, while simul- have had the opportunity to actively partici-
taneously omitting other important job activi- pate in the comprehensive planning, develop-
ties or requirements. Pay ranges did not ment and implementation of a new pay system
adequately reflect market pay levels, and absent substantial involvement of external
compensation policies regarding hiring rates consultants. The authors have discovered that,
and movement through pay ranges had
become barriers to staffing and retention. Corresponding Author:
Steven L. Thomas, College of Business Administration,
Many job descriptions were outdated and not Missouri State University, 901 S. National, Springfield,
reflective of actual work performed. The per- MO 65804, USA.
formance review process was perceived to be Email: steventhomas@missouristate.edu
28 Compensation & Benefits Review 51(1)

although challenging, it is possible to design, the need to revise the staff pay system. The
develop and implement a comprehensive grade and step system in place since the
compensation system by tapping into internal 1970s was no longer labor marker sensitive.
resources and expertise. Our experience sug- New classified staff started out at uncompeti-
gests that human resource administrators and tive early step salaries at the bottom of the
analysts will benefit from a review of our pay range, regardless of their skill or experi-
decisions, processes and challenges in the ence levels, making staffing difficult for
management of their own systems. We pro- many jobs. Pay ranges by the early 1990s
vide a rather detailed description of the forces were relatively broad, and pay increases for
leading to the design and development of new classified staff occurred through a series of
job evaluation systems, pay structures and ten to eleven semi-automatic pay steps. The
performance management tools for university role of performance or merit in pay determi-
staff. We describe system implementation, nation was negligible. Very few employees
and we explore some problems and issues we failed to receive a pay step increase based on
encountered as we developed a new pay struc- adverse performance reviews, and most
ture, all from the perspective of 10 years of employees moved from the bottom to the top
after implementation. of the pay range one-step per year over about
10 years. Conversely, senior employees at
Recognizing the Need for a the top of the pay range were rarely the ben-
eficiaries of pay increases that kept them
New Pay System whole with either cost-of-living or growth in
The development and implementation of a market pay. It was clear that the need for a
compensation system for a major comprehen- new approach to compensating employees
sive university is a complex and time-con- was great.
suming process. The complexity stems from
having to make strategic decisions about job Early Committee Work on
evaluation processes, job structures and crite-
rion measures. The time commitment is due to
Pay and Performance
(1) inertia and bureaucracy in recognizing the Several Metropolitan University committees
need for change, (2) administrative approvals and task forces had addressed pay and perfor-
and budgeting for both strategic and opera- mance management issues since the early
tional decisions and (3) the myriad activities 1990s. In 1994, a Merit Pay Task Force con-
that must be undertaken to implement the stra- ducted a comprehensive attitude survey of all
tegic and operational decisions made. This staff using the Job Satisfaction Survey (JSS).
includes the necessity to conduct job analysis This instrument produced scores on several
for large numbers of jobs that have inadequate dimensions of overall job and pay satisfaction
job descriptions, in addition to the actual job with normative data that provided useful diag-
evaluation design and implementation pro- nostic information.1 The JSS, along with sev-
cesses. Another major implementation activ- eral additional questions created by the internal
ity contributing to the time demands at consultant, revealed mixed and sometimes
Metropolitan University was the commitment conflicting perceptions about compensation
to involve large numbers of staff and adminis- and other organizational issues. On the posi-
trative employees in planning, implementa- tive side, staff were generally satisfied with the
tion and evaluation activities. quality of immediate supervision, their rela-
This article describes activities undertaken tions with coworkers, the work itself and the
from 1994 to 2005 that established the impetus mission of the organization. The surveys
for change, and the roughly 3-year period from revealed a relatively committed group of
August 2005 to March 2008, to develop a new employees who were proud of their work, who
pay system from Metropolitan University staff. felt they had good working relationships with
There had been a long-standing recognition of both peers and their immediate supervisors.
Thomas and McKenzie 29

However, employees expressed consider- information technology, were frustrated with

able dissatisfaction with relationships and their lack of discretion in hiring. They argued
communications with administrators, with that they were losing the more highly skilled
pay structure and administration in general candidates because of their inability to offer a
(including the way pay increases were deter- salary above the pay range minimum. To deal
mined) and, in particular, with pay level. One with this problem, the task force suggested
of the most seemingly inconsistent results was several important objectives for a new pay
attitudes toward performance-based pay. system:
Although employees embraced the concepts
of performance-based pay, and the notion that 1. A pay-level policy needed to be
higher pay raises should go to those who per- explicitly articulated for each different
formed their jobs well, they were generally job family, and pay ranges should be
reluctant to embrace performance-based pay redesigned to accommodate the stated
systems. An in-depth analysis of the survey pay-level policy.
results, and focus group interviews, suggested 2. The hiring rate should not be lower
deep-seated mistrust in the university admin- than the 30th percentile relative to sur-
istration’s ability or willingness to develop vey data.
adequate performance management systems 3. The midpoint of each range should
or to fund systems that were in place. reflect the intended pay-level policy.
Survey results and comments from staff 4. Hiring managers should have more
pointed to the realization that Metropolitan discretion to hire particularly well-
University lacked a performance-based cul- qualified applicants at levels up to the
ture. Survey responses pointed to a significant pay range midpoint.
perception among rank-and-file staff that 5. The range maximums should reflect
administration did not care about employees the maximum value of each job in the
and was not fair and honest in its dealings with grade and should be somewhat above
employees and that its communication with market medians as determined by
employees was poor. Survey comments wage surveys.
included one particularly salient example from 6. All pay ranges should be adjusted
the 1980s that generated hard feelings: A new annually for overall market growth in
pay Step Zero, created as a cost-cutting mea- wages and salaries.
sure, forced new employees to start at a lower
salary. Based on the results of this survey, and Additionally, to address pay equity issues,
the deliberations of the Merit Pay Task Force, the committee recommended that where
a set of recommendations evolved that restructuring grades and ranges revealed that
included three components: (1) the need to some individuals were paid above the pay
develop a new pay structure, (2) the necessity range for their grade, their pay should be fro-
of creating a more effective performance man- zen until range adjustments could catch up.
agement system and (3) an initiative to com- Individuals paid below the new pay range for
municate more effectively with employees. their jobs should be adjusted into the range as
It became abundantly clear that the pay appropriate for their performance and tenure.
structure had far too many pay steps within The Merit Pay Task Force recommended
each pay range and an unjustifiably long time specific changes in the pay step system.
to progress through the pay range. In some Instead of moving from the bottom to the top
instances, those at the top of the pay range of the pay range over 10 years at the rate of
were paid substantially above the relevant one step per year, the Task Force originally
labor market means, while those at the bottom made a recommendation that step progression
of the pay range were frequently somewhat should correspond to the learning curve for
below market. In addition, many managers, the job. Movement through the automatic
particularly in areas such as skilled crafts and steps to the range midpoint should occur in no
30 Compensation & Benefits Review 51(1)

more than 2 years based on satisfactory per- was recommended that groups of incumbents
formance reviews at each designated step, and and supervisors, with staff support, should be
movement beyond the midpoint would be on involved in deciding what constituted good
the basis of merit. job performance by identifying all the impor-
Ultimately, the final recommendation was tant dimensions of performance for each job
that step increases should be completely elim- group and arriving at performance standards
inated and that employee movement through for each dimension. Subsequently, another
the pay range should be based on a Merit Grid internal consultant with expertise in perfor-
approach in which raises (as constrained by mance management (whose work was not
the pay increase budget) were determined by part of the Compensation Analyst Team),
incorporating both an employee’s perfor- worked with staff to develop performance
mance review and position in the pay range. dimensions, criteria and standards, which
Better performers, and those lower in the pay were translated into a new, more valid and
range, would receive higher percentage substantially less contaminated, appraisal
increases. Employees at the top of their pay instrument. Eventually, all supervisors
range would receive only an across-the-board received training in the use of the new
increase based on indexed growth in wages appraisal instrument.
and salaries. The Task Force suggested that An important part of the appraisal system
pay range adjustments should be based on involved careful training of supervisors in
overall wage and salary indexes, not on how to observe, record and document perfor-
Consumer Price Index as the former is a less mance. Just as important was training in how
volatile and more appropriate indicator of to conduct appraisal feedback sessions. Top
market salary growth. administrators were encouraged to create a
As noted above, the JSS survey revealed a climate that required all managers and super-
low level of employee satisfaction with com- visors to take employee appraisal and feed-
munications from administration, and that back seriously and to avoid the politics of
was particularly true with regard to under- appraisal that could undermine program suc-
standing how pay and benefit decisions were cess. Administrators were tasked with provid-
made. The Task Force addressed this problem ing support for the development of sound
by recommending that a custom report be pro- appraisal systems and were encouraged to
vided annually to each employee that pro- reward accurate and timely appraisals from
vided each with information on his or her total department heads and supervisors.
compensation, including what the university The Merit Pay Task Force argued that
paid for in legally required benefits and supervisors and administrators, themselves,
fringes. Labor market information should also should be appraised on the accuracy and time-
be included so each employee could compare liness of employee appraisal and feedback. By
his or her total compensation package to the June 2003, a Staff Salary Task Force chaired
relevant labor market. Perhaps more impor- by the Human Resources Director supported
tant was support for a strategic imperative that many of the recommendations described in
the process of developing a new pay system the 1994-1995 Merit Pay Task Force. It con-
should involve considerable employee partici- cluded that adopting a system that involved
pation, from all levels and job families, eliminating the current grade and step struc-
through focus groups and committees. The ture, establishing new salary ranges and
need to involve key employees in making adjusting employee salaries into new pay
decisions and communicating processes and ranges would be an initial step toward achiev-
outcomes to others was paramount. ing one of the key performance measures for
The Task Force recognized the necessity of Metropolitan University: achieving and main-
developing a new and more valid performance taining classified staff salaries competitive
management system. The flaws in the unstruc- with equivalent positions in the metropolitan
tured system in place were recognized, and it labor market.
Thomas and McKenzie 31

Table 1.  Board of Governor’s Recommendations and Implementation Steps.

Recommendations Implementation steps

I. The existing step-and-grade system for 1. E stablish the purpose of the new pay plan; decide the
classified staff would be abandoned in favor method and level of involvement.
of a system with open pay ranges. 2. R eview job descriptions, conduct job analysis and
II. The distinction between classified and rewrite job descriptions.
unclassified staff would be abandoned in 3. D evelop a new job evaluation instrument with which
favor of a new system, where all staff fall to evaluate most staff jobs.
under the same system for salary purposes. 4. E valuate all jobs using the new job evaluation
III. Future pay increases would be determined instrument to establish a job hierarchy based on job
through a Compensation Matrix process, evaluation results.
where pay increases would be based on 5. C reate a salary survey database.
the overall pay increase budget, individual 6. D evelop pay grades and ranges by combining the job
performance appraisal results and the hierarchy with the salary survey data.
individual’s position in the pay range. 7. Administration and communication activities including
process of notifying employees and handling appeals.

The University occurring mostly during March 2006. As a

Compensation Committee result of feedback received from the commu-
nications during these town hall meetings, and
On his appointment to Metropolitan other concerns expressed to the University
University in 2005, one of the first initiatives Compensation Committee during the spring
undertaken by the new University President of 2006, the Committee’s final report was
was the establishment of a 12-member completed and presented to the president in
University Compensation Committee, com- April 2006. The final report was subsequently
posed of administrators, faculty and staff rep- submitted to, and approved by, the University’s
resenting a wide range of areas and levels, and Board of Governors in May 2006. The
charged with making recommendations University Compensation Committee reaf-
regarding a number of important compensa- firmed many of the findings of 1995-1995 and
tion-related matters affecting both faculty and 2003 task forces in recommending that the
staff. Under the leadership of the Associate current staff grade-and-step system be aban-
Provost, the University Compensation doned and that a new pay system be devel-
Committee met weekly (either as the entire oped for the majority of university staff. The
committee or as subcommittees of faculty and new pay system would have market-driven
staff) during the fall of 2005 into January pay grades and ranges in which both equity
2006, to address the complex issues relating and performance would be applied via a merit
to performance-based pay and job and pay grid approach called the “Compensation
structure modifications. Matrix” to establish pay increases for all staff.
The University Compensation Committee We offer a more complete description of these
worked throughout the winter intersession recommendations below (see Table 1).
period of late 2005 and early 2006 to deliver
an initial set of compensation-related recom- Developing the New Pay
mendations to the University President. The
initial draft proposal, completed in late
System
January of 2006, went through a series of revi- The development of the new pay system
sions based on feedback from the University began in 2006 with consideration of the major
President and others. The seventh draft of the decisions involved in developing a pay struc-
revised University Compensation Committee ture, including establishing the purpose of the
recommendations was presented to faculty plan, deciding among different plans and
and staff during a series of town hall meetings methods and obtaining the involvement of
32 Compensation & Benefits Review 51(1)

relative stakeholders.2 For Metropolitan families. This perspective was carefully con-
University, the purpose of the plan was to sidered. In traditional job-based methods such
establish a pay structure that integrated each as the point-factor system ultimately adopted,
job’s pay with its value, to communicate how pay is based on characteristics of the job, such
job hierarchies and pay levels were developed as the level of education, experience, respon-
and to create a system that could be used to set sibility or other factors required by the work.
pay for both new and changing jobs over a Pay is linked to the job, and everyone per-
relatively long time period. By establishing a forming the job is paid within a given pay
clear set of criteria to develop a job and pay range for that job.
structure, and maintaining transparency in the A move toward person-based pay would
process, bias and inequity could be limited represent a substantial departure from tradi-
and acceptance of the final system ensured. tional pay systems. In person-based systems,
There was concern with the strategic aspect pay is based on skills, knowledge sets or com-
of the new system. It was important to demon- petencies possessed by the individual. Person-
strate strategic congruency by ensuring that based approaches allow people to increase
Metropolitan University’s mission was sup- pay by taking on additional training and/or
ported by the job evaluation system. This developing new job knowledge or skills by
issue proved to be somewhat difficult to going through a certification process to estab-
address because a comprehensive university lish proficiency. The person-based approach
has multiple constituencies, and mission state- was advanced by some university groups as
ments that reflect issues not immediately supporting the university’s continuous learn-
related to job value. It is much easier, for ing mission by placing emphasis on undertak-
example, to demonstrate how job evaluation ing training and developing job skills. In these
criteria support the value of continuous learn- systems, pay is no longer attached to the job
ing than it is to develop criteria demonstrating but to the person. Person-based approaches
a commitment to inclusion. As the job evalua- allow for more staffing flexibility as pay does
tion system was developed, however, there not change because changing work tasks does
was an awareness of how job evaluation crite- not involve changing jobs.
ria rewarded aspects of work that supported Despite these advantages, the move toward
organizational success, and we sought to limit a person-based system was rejected for two
evaluation criteria to those that reflected job main reasons. First, there was a concern that a
attributes consistent with strategic objectives person-based system would put increasing
of the university, or, at the very least, not pressure on training resources since new train-
inconsistent with the Metropolitan University ing programs and certification processes
mission. would have to be developed and implemented.
A much greater fear, however, was the recog-
nition that as more individuals moved to the
Person-Based Pay
top of skill or knowledge blocks over time,
The second major decision addressed was to actual work performed would not require the
select from different approaches to establish- skill or knowledge levels achieved by those
ing a pay structure. There are only two con- employees, and for which the university was
ceptually different ways to establish pay paying. The university would be paying for
structures: job-based and person-based struc- skill sets and knowledge possessed by the
tures.3 One of the most strategically important employee but not required to perform the
decisions at this point was the decision to tasks. An additional concern by the compen-
reject person-based pay systems in favor of sation team was that it would be exceedingly
traditional point-factor job evaluation. There difficult to price knowledge or skill sets from
was significant interest from some quarters to available salary survey data. Ultimately, the
move toward person-based pay systems by decision was to approach job evaluation using
using knowledge-based pay within some job traditional point-factor analysis.
Thomas and McKenzie 33

Another critical decision was whether to or revised for each of four job families.
use a single plan or multiple plans. A single Committees were developed to review job
plan involves having a single set of job evalu- evaluation plans and to conduct the actual
ation criteria and applying them to all jobs. A job evaluation process. Job evaluation was
multiple plan involves describing different conducted from November 2006 to October
sets of related jobs (called “job families”) and 2007. Salary survey data were collected,
developing different rating criteria for each databases were developed and pay ranges
different job family. Metropolitan University were developed and analyzed during the lat-
had a large, very heterogeneous set of jobs ter half of 2007. The proposed new pay sys-
with multiple incumbents, so the decision was tem was presented to Administrative
to designate four job families and four sepa- Council over three sessions in late 2007.
rate job evaluation plans as described in more Staff were officially notified of the new pay
detail below. system in early 2008, and after a series of
communication meetings and appeals pro-
cesses, the new pay system went into effect
Process Design and on July 1, 2008.
Implementation
Our last decision to make was whom to The University
involve in design and implementation. Compensation Committee’s
Because of the problems identified above Recommendations and
through attitude surveys and focus groups
over the quality of communication and trans-
Implementation Steps
parency, the decision was to address the per- The University Compensation Committee’s
ceptions of unfairness and information flows report contained important recommendations
through incorporating the widest level of par- for both faculty and staff. For faculty, the rec-
ticipation in plan design and implementation ommendations included the establishment of
possible. The decision was to design and a new merit-based pay system in which future
develop the program internally. Plan design- pay increases were based on a combination of
ers included an internal consultant, an experi- equity and performance (see Thomas &
enced faculty member in the College of Richardson,4 for a detailed description of its
Business and a compensation analyst from operation). The major policies adopted by the
human resources. Implementation involved Board of Governors based on the University
more than 30 staff involved in Compensation Compensation Committee’s recommenda-
Committees and the Task Force, with more tions are described in Table 1, along with a
than 100 staff working in 12 job analysis and number of distinct implementation steps
job evaluation committees. needed to accomplish the desired outcomes.
Pay system development required con- The implementation of the Board
current work on a number of activities. The Governor’s Recommendation III required all
entire process was communicated to six steps outlined in Table 1 to establish each
University staff from late summer through individual’s position in the pay range. In addi-
the fall of 2006. Traditional job analysis tion, the determination of individual employee
methods including a task inventory and job performance would have to be accomplished
analysis interviews were employed from through the development of a new perfor-
mid-2006 to mid-2007 to provide informa- mance management system, training manag-
tion for restructuring administrative support ers, supervisors and employees in the
and clerical jobs. Job descriptions for other performance management process, resolving
jobs were reviewed and updated over the administration and communication issues and
same time period. Simultaneously, job eval- implementing the system. By early 2008, all
uation plans were developed, updated and/ these activities had been completed.
34 Compensation & Benefits Review 51(1)

The Staff Performance that this appraisal instrument would be appro-

Management System priate for merit purposes with minor modifica-
tions and extensive training. Under the
The University Compensation Committee’s direction of the Human Resources Training
recommendations that stimulated the develop- and Development Manager and the internal
ment of a new pay system were grounded on consultant, the modifications and training
the work of previous committees that recog- were completed during by the end of 2006.
nized the inadequacy of the old pay system, The University Compensation Committee
and staff members on the Committee were recommendations recognized that the aban-
generally aware of these problems. There was donment of the step system would create pay
also concern that pay increase policies for fac- inequities because individual employees
ulty, unclassified staff and classified staff be would be at different steps within the old sys-
designed as nearly similar as possible to avoid tem. Therefore, different individuals in multi-
the “us-versus-them” mentality and to incumbent jobs would likely be at different
enhance feelings of equitable treatment across pay rates. Before the performance manage-
different groups of jobs. An additional charge ment system could be implemented, a mecha-
presented to the University Compensation nism for recognizing this inequity had to be
Committee that made this committee different applied to the pay system. An initial step in
from its predecessors, and one that had a addressing equity occurred in 2006 in which
major impact in shaping the Committee’s rec- 0.75% of the year’s pay increase was devoted
ommendations, was the imperative that ele- to equity. A more permanent mechanism, the
ments of performance-based pay must be Compensation Matrix, was designed to pro-
integral to the design of the pay system. Thus, vide slightly larger percentage pay increases
the University Compensation Committee’s to equally performing individuals who were
recommendations for the staff pay system lower in their pay ranges once the pay ranges
involved moving to a pay increase system that were established.
placed more emphasis on individual perfor-
mance, eliminating the distinction between The Process of Establishing
classified and unclassified staff and abandon- a Job Hierarchy Using Job
ing the system of automatic step increases.
The University Compensation Committee
Evaluation
issued a final report that embraced the funda- The traditional approach to establishing a job
mental principle that valid and fair measure- hierarchy involves conducting job evaluation
ment of performance is essential for the by developing a point-factor plan and evaluat-
implementation of any performance-based pay ing each job using that plan.5 This is a long
plan. While the Committee recognized the and involved process requiring substantial
developmental purpose of performance mea- expertise. Therefore, it is common to hire a
surement, it recognized that the focus of the consulting firm to develop and implement job
Committee’s work was on using performance evaluation systems. This approach, however,
evaluations in compensation decisions. The is expensive with costs well into six figures
Committee reviewed various performance (estimates at the time were in the range of
appraisal models and eventually recommended $250,000). The University’s solution was to
a performance evaluation tool developed in- rely on internal expertise by reassigning a
house a few years previously but never fully compensation analyst from the College of
implemented. Beginning in the late 1990s, a Business, to the role of internal consultant, to
staff task force charged with looking at perfor- work with the Human Resources
mance management issues developed and pilot Compensation Analyst to develop and imple-
tested a performance appraisal instrument with ment a new job evaluation system. This
a focus on performance feedback and Compensation Analyst Team began work on
employee development. The Committee felt the new job evaluation system in the summer
Thomas and McKenzie 35

Table 2.  Steps Involved in the Point-Factor Job Evaluation Process.

1. A set of compensable factors was defined. These included factors such as knowledge, skill,
responsibility, effort, experience or other attributes required by the job that reflect what is valued by
the organization and the labor market.
2. Each compensable factor was scaled with three to nine different degree levels that would distinguish
jobs from one another (e.g., Level 1 of an Experience factor might be a job requiring no experience,
while Level 4 might be a job requiring 5 years of experience).
3. The factors were differentially weighted by small and medium-size enterprises, and based on the
weights, each factor level was assigned a specific point value.
4. Each job was evaluated by a committee to assign a level for each factor to that job. That job received
the points associated with that level.
5. Each job’s point score was calculated by simply adding the points the job received for each factor, and
that point score was to establish a job hierarchy.

Table 3.  Job Families and Descriptions of Jobs in Each.

Job Family 1 (JF1) Administrative support and clerical jobs, some library employees, many student
service jobs and other related jobs
Job Family 2 (JF2) Skilled crafts and trades, custodial and maintenance jobs, safety and security
jobs and other related jobs
Job Family 3 (JF3) Information and computer technology, technicians and related jobs
Job Family 4 (JF4) Professional, administrative and managerial jobs

of 2006 with graduate assistant support and inadequate to capture the unique attributes of
clerical, information technology and financial some job families. Since Metropolitan
support from relevant university functions, but University had an incredibly diverse group of
required no substantial incremental budgetary jobs, the Compensation Analyst Team decided
commitment. A point-factor job evaluation to use a multiple plan. Table 3 describes the
plan was developed to establish a job hierarchy four job families identified for which separate
based on the point values of each job as deter- job evaluation plan was developed.
mined by a set of scaled and weighted compen- The advantage of using a multiple plan
sable factors. That point system was developed with four different job families was that each
through the steps described in Table 2. job evaluation instrument developed for a job
As previously noted, one of the initial deci- family could include factors specific to that
sions the Compensation Analyst Team had to job family. That made it somewhat easier to
make was whether to use a single or multiple evaluate jobs and to produce more consistent,
job evaluation plan. Establishing a single or intuitive and market-sensitive outcomes. The
multiple plan involved deciding whether the disadvantage was that the value of jobs in dif-
organization would use one set of compensa- ferent job families could not be directly com-
ble factors for all jobs or if it was desirable to pared because factors, scales and weights
separate jobs into different groups (job fami- were different in each plan. A job evaluated at
lies) and develop a different set of factors and/ 500 points in Job Family 1 (JF1), for example,
or scales for each family. Using a single plan would not necessarily be worth the same as a
is possible when jobs are similar. It allows 500 point job in JF4, because the compensa-
direct comparisons of all jobs (because all are ble factors or factor weights would be differ-
evaluated on the same scale), and it is useful ent, and the points would correspond to
for comparable worth analysis. However, different dollar values. Although not a consid-
when jobs differ widely in job requirements eration in the development of this pay system,
and tasks performed, a single plan is often the inability to compare point levels across
36 Compensation & Benefits Review 51(1)

Table 4.  Job Evaluation Compensable Factors by Job Family (JF).

JF1 JF2 JF3 JF4

Professional knowledge, skill and technical mastery x
Educational and experience equivalencies chart x x  
Educational requirements of the job x  
Complexity and technical mastery x x  
Skill requirements: Crafts and trades skills x  
Supervisory responsibility x x
Responsibility for the work of others x  
Job controls, guidelines, scope and impact x x
Guidelines x x  
Contacts x x  
Interactions with others x
Relationship complexity x  
Physical effort and work environment x  
Work environment x x  
Physical demands x x  
Responsibility for facilities and resources x  
Budgetary control x  
Work impact and effect x  
Managerial responsibility x

job families can actually be an attractive fea- deleted, modified and added to fit each of the
ture, because it creates a mechanism that four job families. The job evaluation systems
allows an organization to establish different for each job family were not developed simul-
pay-level policies for different job families taneously. Instead, they were developed as job
that have different strategic importance or analysis and new job descriptions were com-
market sensitivity. pleted, and that process proved to be easier in
Writing compensable factor definitions is a some job families relative to others. Each job
difficult and time-consuming task. A pilot evaluation instrument was reviewed by a
instrument developed in 2000 by the internal committee that included job incumbents and
consultant was revived and used as a starting supervisors, and numerous changes were
point for the development of a new job evalu- made based on feedback received. The revi-
ation manual. The internal consultant incorpo- sions continued even as the instruments were
rated language found in a variety of being utilized, as problems were recognized
well-known job evaluation plans and modi- in the application of the job evaluation plans.
fied the language for use with Metropolitan A complete description of the final job evalu-
University. Much of the language came from ation plan and the compensable factors are
U.S. Bureau of Labor Statistics resources not included here, but they are available in
including the National Compensation Survey the Job Evaluation Manual available from
(2003)6 and various versions of the U.S. Civil the authors or online on request. However, an
Service Commission’s Factor Evaluation overview of the compensable factor structure
System (1997).7 The language from these by job family is shown in Table 4. As one can
resources, along with that from other widely see, some measures such as knowledge, skill
available job evaluation resources, was or educational requirements of the job or
adopted and modified to fit the needs of responsibility requirements are common
Metropolitan University’s job structure. across job families although the factors may
A number of modifications were required be defined or weighted differently in different
to the original instrument as factors were job families. Other factors such as managerial
Thomas and McKenzie 37

responsibility are used in only one job family one Job Evaluation Committee of about 10
(in this case, JF4). members for each job family. For JF4, the
largest of the job families, five different com-
Job Analysis and Job mittees representing different University
functions were established to conduct job
Descriptions evaluation. The sole function of each commit-
Before the job evaluation process could begin, tee was to take every job in each job family,
the Compensation Analyst Team had to ensure compare the job with each compensable fac-
that accurate job descriptions existed so that tor and assign a factor level from each factor
decisions about job content made by job eval- to each job. When each committee’s work was
uation committees could be made with some completed, each job had been assigned a level
degree of certainty. The Compensation for each factor, but the work of creating a
Analyst Team recognized early on that a com- complete job structure required the weighting
plete job analysis of all staff jobs, though and scaling of the factor structure to establish
highly desirable, would take too long and cost a point total for each job.
too much. Pilot testing of JF1 during the To develop a point scale, the internal con-
spring of 2006 revealed a considerable num- sultant applied multiple regression techniques
ber of problems with many administrative to the data set to establish factor weights and
support positions to the extent that a fairly create factor point scales for each job evalua-
large-scale restructuring of many administra- tion plan. The approach required regressing
tive support and clerical positions was neces- current employee pay levels (the midpoint of
sary. The Compensation Analyst Team current pay ranges) on internal job evalua-
decided to perform a complete job analysis of tion results (compensable factor levels) so
most clerk, secretary, administrative secretary that the new pay system would not be severely
and executive assistant jobs using a task out-of-step with current pay levels. The
inventory and individual job interviews. The objective was to incorporate a “policy captur-
lengthy process occurred from August 2006 to ing” approach in weighting and scaling the
April 2007. The result was that many JF1 jobs compensable factors. This approach was as
were reclassified into academic administra- much political and practical as it was objec-
tive assistant, administrative assistant, admin- tive and analytical. The constraint facing the
istrative specialist and executive assistant job Compensation Analyst Team was that the
series. This provided a more consistent place- cost of implementing a new compensation
ment of individuals in jobs and an overall program could not be prohibitive. Therefore,
reduction in the number of jobs. This process the final pay outcomes were constrained by
was complete by summer of 2007. The the fact that they could not stray too far from
remainder of the jobs in JF1 and the jobs in current pay levels. In other words, the new
the other three job families did not receive a pay system had to “capture” or otherwise
complete analysis. Instead, unit managers and generally support or defend current pay lev-
supervisors were asked to conduct a complete els and differentials. The committee was
review of job descriptions in order to make aware that a policy capturing approach could
sure that job specifications were accurate and be open to criticism for perpetuating past dis-
ensure that task statements reflected the work crimination and inequities in the current pay
as it was actually performed. system, and there was some concern about
conflict between the internal job structure
and market pay levels.
Job Evaluation
Regression coefficients were used to weight
About 80 individuals with broad campus rep- factors and then to develop a scale of point val-
resentation working in four Job Evaluation ues for each factor level. The resulting scale
Committees conducted job evaluation in the was applied to the job evaluation ratings to
fall of 2007. For JF1, JF2 and JF3, there was yield a total point score for each job. Initially,
38 Compensation & Benefits Review 51(1)

the fit between point values and pay was did not directly establish the pay levels for
uneven with R2 values of .70 to .85, indicating jobs. It was necessary to complete pricing by
only moderate fit. Some evaluation errors and collecting market pay data for jobs to anchor
analytical issues were addressed, and the final pay levels within each pay grade. The
R2 values from about .80 to about .92 indicated Compensation Analyst Team developed a
adequate fit. Several months later, regression detailed salary survey database used to price
analysis using market salary survey data in benchmark jobs. Not all University jobs had
place of current employee pay showed sub- market comparisons, in fact market pay data
stantially equivalent fit. Both internal pay rates were available for only 35% to 50% of all
and market pay generated a very similar set of staff jobs. While a few unique jobs such as
factor rates and job structures. “Program Dance Musician” had no market
The Compensation Analyst Team estab- data readily available, neither did some rela-
lished pay grades by sorting jobs so that jobs tively common occupations such as lock-
within a narrow range of point values were smith. Therefore, the team identified a number
grouped into a pay grade. This was accom- of benchmark jobs (jobs common in the labor
plished via a “cluster” approach in which the market and for which salary data were avail-
team looked for natural breaks in the range of able) within each job family, and these were
point values. Grade breaks were chosen in the jobs used to establish market pay. Pay lev-
such a way as to minimize problems with els for jobs that were unable to be bench-
reporting relationships and existing hierarchi- marked were established based on the salary
cal structures. Once grades were defined, all data for other jobs within the same pay grade
jobs in the same grade were treated as the under the assumption that all jobs in a grade
same for pay purposes. Each job family are equally valued. In other words, if there
received new pay grade numbers such that the were 12 jobs in a grade, we might be able to
first digit indicated the job family and the sec- get pay data for five to seven of them, includ-
ond digit the grade. Within a job family, each ing some jobs that might have data from mul-
pay grade number designated a pay range (the tiple surveys. It would then be possible to
higher the grade number, the higher the pay collect 15 to 30 market medians or weighted
level). However, there was substantial pay means and 10 to 20 minimum, maximum or
range overlap across job families (e.g., the percentile values for that grade. All jobs in
JF1 Grade 19 might have a higher pay range that grade would then be paid based on the
than JF2 Grade 22). Final job hierarchy and overall benchmark job data.
grade structure were subsequently communi- A different salary survey database was
cated to each job evaluation committee and to established for each job family by carefully
job family leadership. Review of the job comparing job descriptions for University
structure by the Compensation Analyst Team, jobs with job descriptions for survey jobs.
and the resulting feedback from job family When matches were obtained, descriptive
leadership, resulted in modification of ratings data, including salary means, medians, mini-
for a small number of jobs. Changes in job mum and maximum values and 25th and 75th
evaluation outcomes were typically the result percentiles were incorporated into our data-
of job descriptions that had not been properly base. As many as 30 to 40 different salary sur-
revised, postevaluation changes in job content veys were used to establish market pay. The
or simple evaluation error. major surveys used included the College and
University Personnel Association database,
Developing Pay Grades and the proprietary Salary.com database, the
Chamber of Commerce database and the AIM
Ranges database (an industrial group) and association
The job evaluation process created an internal data provided to us by various managers and
job hierarchy and allowed us to establish pay administrators. Many sources of pay data were
grades. However, the job evaluation process costless based on membership in professional
Thomas and McKenzie 39

organizations, some were in the public was close to market means and/or medians
domain, while others (such as Salary.com) and (2) pay ranges would be constructed in a
cost several thousand dollars to purchase. manner that produced the least amount of con-
However, we spent less than $10,000 in total flict with existing pay levels. In general, pay
on data collection. Geographical pay data ranges were 40% to 60% and midpoint differen-
were selected for comparison purposes corre- tials were 12% to 15%. However, the two deci-
sponding to what was considered to be the sion rules produced some unevenness in the
relevant labor markets for each job group. For grade structures including pay ranges that were
example, we used local data for positions such broader than desirable and/or midpoint differ-
as skilled crafts and trades, but we used entials that were too flat. The Compensation
regional data for professional jobs such as a Analyst Team’s position was that unevenness
development director. The various databases in the structure was necessary to reduce the
allowed us considerable choice in selecting conflict with current pay levels and the need
relevant labor market data. The Compensation for excessive pay adjustments that were not
Analysis Team did have to resist pressure market justified.
from some internal sources to use data that As a result of these parameters, the original
were favorable to specific jobs but probably grades were modified and adjusted by the
not representative of the actual market. Compensation Analyst Team during the range
Pay ranges were established for each job development process. This involved adding or
grade within each job family by aggregating deleting a grade or changing cut scores
pay data for all surveys and benchmark jobs between pay grades. Further changes to jobs
and by carefully analyzing survey metrics. were made after input from job family lead-
The data used to determine pay ranges ers. Based on a limited number of inconsisten-
included the median, mean, percentiles and cies in the job and pay structures, a few jobs
range values within and across survey data were reevaluated and a small number of indi-
sets. In many cases, pay data were based on viduals were moved into different grades based
hundreds or thousands of incumbents. In other on those results. One of the most significant
cases, grades with a small number of jobs changes was the removal of about 30 jobs from
might have limited data. While, the Salary. the JF4 structure. Higher level jobs, including
com database was the only source that was the Provost, Associate Provost, Vice President,
purchased, it provided a set of unique analyti- jobs with faculty appointments and physicians
cal tools that allowed a lot of discretion in and pharmacists were removed because they
how to “slice” the data, and it also allowed were not appropriate for the job evaluation sys-
estimation of relevant local market pay data, tem or were of poor fit. Pay grade and range
where sample sizes were small by using a final results were presented to Administrative
forecasting tool and regional pay data. This Council for approval in December 2007.
“virtual” local labor market data proved to be
extremely useful. Pay range midpoints for Outcomes of the New Pay
each grade were established by creating a pay System, Communications
policy line using linear regression analysis
and regressing market pay medians on the job
and Appeals
evaluation points of each benchmark job. The A new pay system was developed for more
resulting R2 values for pay policy lines than 1,200 employees in about 670 jobs. The
exceeded .90 for all job families, suggesting a new market-based pay ranges produced esti-
good fit between market pay and our job eval- mates that 149 employees would be below the
uation system. bottom of the new pay ranges. Those employ-
Once pay survey data were collected, the ees were brought into pay ranges at a cost of
determination of pay ranges was based on about $256,000 (about 0.6% of payroll) at its
these principles: (1) pay ranges would be con- implementation on July 1, 2008. At least two
structed so that the midpoint of each pay range dozen employees were projected to be above
40 Compensation & Benefits Review 51(1)

pay range maximum values. In general, no Analyst Team remained concerned about
action was taken on these employees because some unresolved issues that could affect the
most were only slightly over pay range maxi- ultimate success of the pay initiative. These
mums. The overall projected compa ratio (a concerns and recommendations are pre-
ratio of actual pay to range midpoint) of .947 sented in Table 5.
suggested that Metropolitan University staff
were, on average, paid slightly more than 5%
Observations From the
below the median of the pay ranges (in our
case, slightly below the average of the labor Compensation Analyst
market pay from the data we had collected). Team: 10 Years After
Employees were notified of new pay grade By 2018, the job evaluation tool, the pay struc-
and ranges and job titles in late January 2008. ture and the performance evaluation process
Following the notification process, a series of had been in place for 10 years. Has the com-
town hall meetings were scheduled for late pensation system worked as it was designed?
January and/or early February to explain the What have we learned? What problems have
pay system and its processes to employees, to we encountered? Has the performance man-
address questions or concerns and to explain agement tool worked? What would we do dif-
the appeal process for employees who felt that ferently? The answers to these questions are
their jobs had been incorrectly evaluated. equivocal and complicated by contextual
Employee appeals were heard by an Appeals forces outside of the systems themselves. The
Committee consisting of a subset of members job evaluation system has generally worked
from the original Job Evaluation Committees. well. It remains in place with some minor
There were 37 appeals involving 41 employ- modifications to factor structures.
ees; 17 appeals were approved, affecting 24 One important question that remained was
individuals. Appeal issues included misevalu- how did staff react to the new pay system. As
ation of the job description, incorrect job described above, a JSS of University staff pre-
assignment (multi-incumbent JF1 positions) dating the job evaluation initiative provided
and incorrect job family assignment. the baseline information on work attitudes.
The measure (the JSS from Spector9) was
Some Concerns and Caveats chosen because a body of research supported
at the Time of Initial the unique dimensional job satisfaction tax-
onomy it provided and because it included a
Implementation set of normative scores for each dimension.
The development of a new staff pay system As noted previously, the results did show that
required 20 months, a substantial commitment staff scores were above survey norms for
of University resources, the cooperation of quality of immediate supervision, relation-
hundreds of University staff, and the direct ships with coworkers and the nature of work.
involvement of scores of individuals. The However, the JSS revealed scores lower than
result was a new pay system in which jobs had norms for pay, benefits, contingent rewards,
been slotted into pay grades based on compen- operating procedures and communications.
sable factors reflecting knowledge, skills, The ad hoc items suggested strong support of
responsibilities and work efforts valued by the and identification with the organization’s mis-
University, and into associated pay ranges sion but considerable dissatisfaction for the
reflecting the value of those factors in the mar- step-and-grade system in place at the time and
ketplace. The hope was that the result would significant resistance to a performance-based
be a pay system that was fair to all in that it pay initiative.
was both internally equitable and externally Attitudinal measures taken after the imple-
competitive. On its completion, the compensa- mentation of the new pay system did not
tion initiative was considered to have been include a comprehensive JSS as recom-
very successful. However, the Compensation mended by the Analyst team. Instead, general
Thomas and McKenzie 41

Table 5.  Pay System Compensation Analyst Team Concerns and Recommendations.

1. On pay system implementation on July 1, 2008, the salary survey data were from 6 to 18 months old
depending on the database. Since the data had not been aged, University salaries relative to market
could fall as much as an additional 3%.
2. Pay ranges should be adjusted annually by an amount equivalent to the wage and salary index for
public sector employees (the BLS Employment Cost Index for state and local government workers
wages and salaries). This would maintain pay grades at market and allow the compa ratio to serve
as a consistent metric for evaluating pay policy. Since this index is typically 2% to 3% annually, the
University would have to be able to generate a salary increase pool of at least that size to avoid falling
further behind market salaries.
3. The pay ranges be audited every 3 years using actual salary survey data to make certain that the pay
ranges remain accurate.
4. There was some concern that staffing levels across the University were uneven and that job creation
was not uniformly controlled. The Team urged investigating opportunities to improve pay for current
employees through modifying work process, eliminating unnecessary work tasks, and restructuring
process and jobs.
5. Human Resources should develop a detailed guide to writing job descriptions using the lessons
learned in job evaluation committee meetings. Chiefly, the job description should have contents that
are consistent and improved task statements and should allow the application of the job analysis and
job description issues.
6. Human Resources should audit all job descriptions at least once every 3 years, noting that Human
Resources will likely need to add at least one 0.5 FTE (full-time equivalent) position to adequately deal
with job analysis and job description issues.
7. Human Resources should measure employee job satisfaction periodically using the Job Satisfaction
Survey,8 or other tool, to assess the success of the compensation initiative by tracking employee
attitudes.
8. Concerns remained that budgetary issues might impede the operation of the Compensation Matrix
and make it difficult for employees to move through the pay range.

Note. BLS = Bureau of Labor Statistics; Team = Compensation Analyst Team; University = Metropolitan University.

conclusions about employee attitudes were ranges annually and review market data as
based on limited survey information, input the Compensation Analyst Team recom-
from focus groups and various committees mended because of cost concerns. State bud-
and feedback from individuals. The consen- get cuts interfered with the operation of the
sus was that satisfaction with both communi- merit grid system and contributed to increased
cations and operating procedures was dissatisfaction with both pay level and con-
significantly improved. tingent rewards.
However, satisfaction with the pay level The job evaluation initiative produced
and contingent rewards failed to produce a enhanced communication with staff regard-
long-term change. In the 2 years following ing how their work is valued and how pay
the implementation of the pay system, satis- decisions were made. We feel that one of the
faction with both did initially rise. This was most important aspects of the original pro-
likely the result of increases in pay range cess was involving large numbers of staff in
breadth that benefited many previously plan design and implementation. The
“topped-out” employees, several one-time Compensation Analyst Team remains con-
pay adjustments for green-circled employees, vinced that, although it contributed to the
and relatively robust pay increase budgets in investment in time, including respected
these 2 years. However, within 3 years of the administrative and operative employees in
new pay system implementation, it appeared the committees that reviewed plan design and
to deteriorate to preimplementation levels. did the actual job evaluation allowed the job
The university was reluctant to adjust pay evaluation process to be widely understood
42 Compensation & Benefits Review 51(1)

and accepted. We also feel that using numer- expectations, but no more than 15% could be
ous small group meetings to communicate the rated as exceptional (no unit was forced to rate
process and its outcomes was critical to its any employee as unsatisfactory). These distri-
success. Ten years later, the job evaluation butions proved to be problematic for staff
system has proven reliable for determining because there were numerous situations in
the appropriate pay grade for new jobs and which a supervisor was responsible for only
reclassifications. The job hierarchy continues one or two direct reports, and that did not allow
to be accepted 10 years later, and employees him or her to meet the distribution guidelines.
continue to recognize that there is a consis- This constraint was unwelcome and was heav-
tent process in place to make pay decisions. ily resisted. University administration provided
One important element that we did not ini- guidance to senior management in each of the
tially realize was the need for salary adminis- cost centers to review the distribution of per-
tration rules applicable to the new formance scores annually within their cost cen-
compensation system for implementing ter and work with managers to reassess
changes to salaries resulting from promotions, performance ratings when patterns of leniency
transfers and reclassifications. Our old salary were noted. These policies were not evenly
administration guidelines were based on the applied across cost centers and units.
old grade and step system. We researched sal- Faculty, although not in the job evaluation
ary administration policies at peer institutions system, used the same Compensation Matrix
and developed new salary administration rules structure to decide pay increases. In the faculty
based on best practices at those institutions. system, pay increases were a function of equity
While we are satisfied with the job evalua- and merit as well as salary increase budgets, with
tion plan and pay structure outcomes, we have the same distributional limits on ratings as for
faced many challenges, particularly with the staff. The system for faculty collapsed because
performance management tools. The original many department heads ignored the performance
grade-and-step system had been replaced by distribution rules, some deans ignored the equity
the Compensation Matrix approach for estab- distribution requirements and the Office of the
lishing pay increases and movement through Provost did not provide enforcement.
the pay range. In the Matrix, pay increases are The internal consultant and the Compensation
dependent on three variables: one’s perfor- Analyst Team had previously cautioned the uni-
mance appraisal, one’s current position in the versity president and other high-level adminis-
pay range and the salary increase budget. This trators about relying on a pay for performance
tool has been a victim of practical problems, mechanism. Metropolitan University lacked a
administrative failures and lack of funding. strong culture of performance management. It
One of the original concerns regarding the had no successful experience with performance-
performance appraisal tool was that there based pay, and its experience with performance
would be severe leniency in performance rat- evaluation tools was a history in which few
ings. This concern was simply an outgrowth of individuals, either faculty or staff, received poor
the fact that the university had never really performance ratings. Raises had been either
developed a performance system that was valid automatic step increases or across-the-board
and had never really required managers to pay increases. Increasing political pressure
make anything other than routine performance from the state to demonstrate some sort of per-
decisions. To address this, the decision was formance-based mechanism for pay increase
made to use a modified forced distribution sys- forced the university’s hand. However, budget
tem in which unit administrators were given cuts in the years following the implementation
limits on the portions of each unit’s employees of the new pay system resulted in several years
who could be rated at the highest performance without pay increases. This rendered the perfor-
level. Recognizing that most employees per- mance management systems developed from
formed adequately, managers were allowed to 2005 to 2008 to be meaningless from a motiva-
rate about 85% of its employees as meeting tional perspective, as pay increases became
Thomas and McKenzie 43

nonexistent or across-the-board. Because of Funding

years with no budget pay increases, administra- The authors received no financial support for the
tive rules did not allow for a merit component to research, authorship and/or publication of this
be included in pay increase decisions, and article.
movement through pay ranges based on across-
the-board pay increases became the norm. ORCID iD
In sum, pay system development and its Steven L. Thomas https://orcid.org/0000-0001
associated outcomes have been generally suc- -5774-2708
cessful. Because of processes designed to have
the broadest possible participation and com- Notes
munication, and the efforts to collect extensive
data and involve a large number of employees 1. Spector, P. E. (1985). Measurement of human
service staff satisfaction: Development
in analysis and decision making, the pay struc-
of the Job Satisfaction Survey. American
ture has enjoyed substantial support, and inter- Journal of Community Psychology, 13,
nal equity is generally accepted. However, the 673-713.
system has not been without problems. Pay- 2. Newman, J., Gerhart, B., & Milkovich, G. T.
level issues do exist and are largely the result (2017). Compensation (12th ed.). New York,
of external events, notably budgetary short- NY: McGraw-Hill Education.
falls. Satisfaction with pay level and contin- 3. See Note 2.
gent rewards remains lower than hoped for 4. Thomas, S. L., & Richardson, P. (2013).
because of the inability to adequately fund the Using an equity/performance matrix to
merit grid and advance individuals through the address salary compression/inversion and per-
pay range. The system is “merit” in name only formance pay issues, with Peter Richardson.
Administrative Issues Journal: Education,
as budgets allowed for only across-the-board
Practice, and Research, 3(1), 20-33.
pay increases of 1% to 2%, and even those 5. See Note 2.
were irregular. Though the direct pay system 6. U.S. Bureau of Labor Statistics. (2003).
remains viable, the performance management National Compensation Survey: Guide
system remains only as a developmental tool for evaluating your firm’s jobs and pay.
for staff, and it has little practical administra- Retrieved from https://www.bls.gov/ncs/ocs/
tive purpose. For faculty, it has been relegated sp/ncbr0004.pdf
to the position of a relatively pointless bureau- 7. U.S. Civil Service Commission. (1977).
cratic exercise. Unfortunately, that is the situa- Instructions for the Factor Evaluation System.
tion at many public universities. While we Washington D.C.: Government Printing
have been able to demonstrate that where Office.
8. See Note 1.
internal expertise is available, effective direct
9. See Note 1.
pay systems can be designed and implemented
successfully, and without incurring great
expense. However, a total compensation sys- Author Biographies
tem depends on successful implementation Steven L. Thomas is a professor of Management
that is dependent on external forces (most at Missouri State University, and has published
notably funding) that marginalize even the numerious articles on compensation, staffing, and
best efforts of the institution. employee rights and responsibilities. He has a PhD
in human resource management from the University
Declaration of Conflicting Interests of Kansas.
The authors declared no potential conflicts of inter- Lyn M. McKenzie is the Director of Compensation
est with respect to the research, authorship and/or at Missouri State University, and has many years of
publication of this article. experience as a compensation analyst.