3Q | 2015
Goal Setting: Meeting Stakeholder Expectations in an Increasingly Dynamic and Complex WorldBy Seymour Burchman and Mark Emanuel, Semler Brossy
Consulting Group
Incentive Compensation: Measures MatterBy Robert J. Greene, Ph.D., CCP, CBP, GRP, SPHR,
GPHR, CPHRC, Reward Systems Inc
Is the Safelite Auto Glass Case Study All It’s Cracked Up to Be?By Frank Giancola
Driving Compensation Strategy Alignment: Using Analytics to Benchmark Practices from European Normative DataBy Brian Levine, Ph.D., Min Park and Tom Jacob, Mercer
Cost-of-Living Adjustments By John Kilgour, Ph.D., California State University,
East Bay
Published Research in Total Rewards
06
18
28
39
46
56
Mission
WorldatWork Journal strives to:
❙ Advance the theory, knowledge and practice of total rewards management.
❙ Contribute to business-strategy development that leads to superior organizational performance.
❙ Provide an outlet for scholarly total rewards writing and research.
Editorial
Publisher Anne C. Ruddy, CCP, CPCU
Executive Editor Andrea Ozias
Managing Editor Jean Christofferson
Senior Editor Angelique Soenarie
Contributing Editor Jim Fickess
Review Coordinator/Permissions Editor Brealyn Nenes
Design
Art Director Jamie Hernandez
Senior Graphic Designers Kris Sotelo | Hanna Norris
WorldatWork Management Team
President and CEO Anne C. Ruddy, CCP, CPCU
Vice President and CFO Greg Nelson, CCP, CPA
Senior Vice President, Strategy and Organizational Effectiveness Betty Scharfman
Vice President, External Affairs and Practice Leadership Cara Welch, Esq.
Vice President, Human Resources Kip Kipley, CBP, SPHR
Circulation
Circulation Manager Barbara Krebaum
2015 WorldatWork Association Board
Lead Director David Smith, CCP, CBP, CECP
Secretary/Treasurer Jeff Chambers, WLCP
Director Bruce Clark, J.D.
Director Nathalie Parent, CCP, CBP, GRP, CECP, CSCP
Director Karen Ickes, CBP
Director Sara McAuley, CCP, WLCP
Director Alan Gardner
WorldatWork (www.worldatwork.org) is a global human resources asso-ciation focused on compensation, benefits, work-life and integrated
total rewards to attract, motivate and retain a talented work-force. Founded in 1955, WorldatWork provides a network of nearly 30,000 members in more than 100 countries with training, certification, research, conferences and community. It has offices in Scottsdale, Ariz., and Washington, D.C.
The WorldatWork group of registered marks includes: WorldatWork®, Wor ldatWork Society of Cer t i f ied Profess iona ls®, A l l iance for Work-Life Progress® or AWLP®, Cer tif ied Compensation Professional® or CCP®, Certified Benefits Professional® or CBP, Global Remuneration Professional or GRP®, Work-Life Certified Profess iona l™ or WLCP ®, Cer t i f ied Sa les Compensat ion Professional™ or CSCP™, Certif ied Executive Compensation Professional or CECP™, workspan®, WorldatWork® Journal and
Compensation Conundrum®.
This publication is a special benefit of membership.
Global Headquarters: In Canada: WorldatWork P.O. Box 4520 14040 N. Northsight Blvd. Postal Station A Scottsdale, AZ 85260 USA Toronto, ON M5W 4M4
Phone: 480-922-2020; Toll-free: 877-951-9191 Fax: 480-483-8352; Toll-free fax: 866-816-2962 Email: [email protected] Website: www.worldatwork.org
WorldatWork Journal (ISSN 1529-9457) is published quarterly by WorldatWork, 14040 N. Northsight Blvd., Scottsdale, AZ 85260, as a benefit to members, who receive an annual subscription with their membership. POSTMASTER: Send address changes to WorldatWork Journal, 14040 N. Northsight Blvd., Scottsdale, AZ 85260; 480-951-9191. Canada Post (CPC) publication #40823004.
WorldatWork neither endorses any of the products, services or companies ref er enced in this publication nor
does it attest to their quality. The views ex pressed in this pub li ca tion are those of the authors and should not be as cribed to the officers, mem bers or other spon sors of WorldatWork or its staff. Noth ing herein is to be construed as an at tempt to aid or hinder the adoption of any pending legislation, regulation or in ter pre tive rule, or as legal, ac count ing, actuarial or oth er such pro fes sion al ad vice.
Copyright © 2015. WorldatWork. All r ights reserved. WorldatWork: Registered Trademark ® Marca Registrada. Printed in U.S.A. No portion of this publication may be reproduced in any form without express written permission from WorldatWork.
Rejection rate: In the second half of 2015, the rejection rate for papers submitted to WorldatWork Journal was 40%.
Reprints: For bulk reprints contact: Gail Hallman at 800-352-2210, Ext. 8175, or [email protected].
Manuscripts: WorldatWork Journal welcomes manuscripts. See guidelines and review process at www.worldatwork.org, or contact any member of the editorial staff.
Letters: Readers are invited to submit letters for publi-cation. Letters are published as space permits and are subject to editing.
Email preferences: To change your email preferences and make sure you are receiving WorldatWork membership benefits via email:
❙ Log in to www.worldatwork.org.
❙ Click “My Profile.”
❙ Select “My email preferences and e-newsletter subscriptions.”
❙ Click “Modify.” Ensure WorldatWork email communications are delivered directly to your inbox and avoid company blocks and filters. Ask your technology department to allow WorldatWork communications to reach you. For more information call toll free, 877-951-9191.
2015 WorldatWork Society of Certified Professionals Board
Lead Director Tracy J O Kofski, CCP, CBP, GRP
Secretary Kevin Hallock, Ph.D.
Director Trevor Blackman
Director Carrolyn Bostick
Directo Robin Colman
Director Guillermo Villa
Director Karen Ickes, CBP
Director Kumar Kymal
Director Steve Pennacchio
Director Brit Wittman, CCP, CECP
Director J Ritchie, CCP
Director Robin Colman
Reviewers
WorldatWork Journal thanks the following individuals for reviewing manuscripts during the editorial cycle for the Third Quarter 2015 issue. Subject-matter experts, including members of WorldatWork’s advisory boards, review all manuscripts.
Wendy Criswell, CCP, CBP, WLCP, PHR I Alexandria Renew Enterprises
Andreas Spurlock, CCP I City of Las Vegas
Chris Ratajczyk, CCP, ACCP, SPHR
Melissa Shulman, CCP I Takeda Pharmaceuticals
Brad Worley, CCP, CBP I Animato
Brandon Conkle, CBP, CCP, GRP I Georgia Institute of Technology
Jan Rauk, CCP, CBP, GRP, SPHR I University of Idaho
Mark Day, CCP, GRP
Kelly Muechler
Wayne Camp, CCP, CBP, SPHR, CCCG
Susan Eichen, Mercer
John England, Pay Governance LLC
Daniel Purushotham, Ph.D., CCP, CBP I Central Connecticut State University
Mark Bussin, Ph.D., CCP, GRP I 21st Century Pay Solutions Group Pty
Stephanie Emhoff, CCP, CSCP I Cox Media Group
Dianne Auld, CCP, GRP, CSCP, WLCP I Auld Compensation Consulting
Dario Kosarac, CCP
Angela Keller, CCP, CSCP, SPHR
Jerry Colletti, CSCP
Fiona Heywood
Mike Ryan
Colin Jarvis
Executive SummariesThird Quarter 2015 | Volume 24 | Number 3
4 WorldatWork Journal
Goal Setting: Meeting Stakeholder Expectations in an Increasingly Dynamic and Complex WorldSeymour Burchman and Mark Emanuel | Semler Brossy Consulting Group
The challenge of setting meaningful, yet realistic, incentive plan goals has become
more difficult in an increasingly complex world. The authors provide three sets of
criteria to help guide companies in establishing meaningful goals that serve to meet
key stakeholder expectations.
Incentive Compensation: Measures MatterRobert J. Greene, Ph.D, CCP, CBP, GRP, SPHR, GPHR, CPHRC, Rewards Systems Inc
Getting measurement right is not easy but mandatory for incentive plans to accomplish
an organization’s goals. The right measures must be selected and the information
necessary to measure results accurately must be available. It may be necessary to
employ measures that are subjectively determined but all interested parties must accept
that they are appropriate. The article provides guidance in setting effective incentive-
plan measures under various business situations.
Is the Safelite Auto Glass Case Study All It’s Cracked Up to Be?By Frank Giancola
The Safelite Auto Glass case study is arguably the most prominent academic research
article that describes how financial incentives motivate employees to higher levels of
performance. Safelite is the largest provider of automotive replacement and repair
services in the United States. The article explains how the productivity of Safelite’s
auto windshield installers improved by 44%, after its pay plan was switched from an
hourly plan to a piece rate plan. This article examines the case, which was written by
a labor economist, from a HR management perspective and uses outside sources of
information to shed new light on what might have been behind the productivity gain.
As it turns out, evidence suggests that most employees improved their performance
primarily to save their jobs, as much as to increase their earnings by installing more
windshields under the new piece rate plan.
06
18
28
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
Executive SummariesThird Quarter 2015 | Volume 24 | Number 3
5 Third Quarter | 2015
Driving Compensation Strategy Alignment: Using Analytics to Benchmark Practices from European Normative DataBy Brian Levine, Ph.D., Min Park and Tom Jacob, Mercer
Rewards constitute a significant investment for all organizations. Thus, it is important
to strategically leverage this key investment most effectively. This includes formalizing
rewards programs to help drive employee attraction, retention and behavior — all of
which, in turn, drive business objectives. This article offers an objective approach to
both investigate differences in practice from norms and inform areas of focus. It uses a
unique application of predictive analytics to measure strategies for employee rewards
at a country level, taking advantage of incumbent-level data collected for pricing
jobs. Key differences are shared across a set of European countries examined, and
a blinded example illustrates the value of looking to company-specific differences.
The analysis presented in this article is based on extensive incumbent-level data for
seven European countries and encompasses approximately 640,000 employees across
1,750 organizations.
Cost-of-Living AdjustmentsBy John Kilgour, Ph.D., California State University, East Bay
This paper examines cost-of-living allowances or adjustments (COLA) for Social Secu-
rity, private-sector pension plans and various public-sector pension plans. While COLA
arrangements have largely been replaced by specified annual increases in private-
sector collective bargaining, the concept of indexing wages and salaries to inflation
has been adopted by Social Security and defined benefit pension plans in the public
sector. Inflation will cause income disparity between the public and private sectors and
create pressure to curtail public-employee pensions.
Published Research in Total Rewards
39
46
56
6 WorldatWork Journal
The perennial challenge of setting meaningful, yet
realistic, incentive-plan goals has become ever
more difficult in an increasingly complex and
rapidly changing business world.
Companies now must also accommodate the growing
importance of a range of stakeholders. Historically,
incentive plan goals have largely focused on meeting
the needs of shareholders and driving total share-
holder return (TSR). Yet, contemporary thinking has
companies taking a more holistic view, one that
explicitly recognizes the concerns of employees,
customers, suppliers, regulators, and the community
at large. (See The Growing Range of Stakeholders
on page 7.) Although shareholder expectations likely
remain primary, meeting the expectations of other
stakeholders is fundamental to generating long-term,
sustainable shareholder value creation.
The need to get it right by gauging performance
in ways that reflect stakeholder concerns has been
further heightened by the introduction of annual say
on pay, in which goal-setting rigor is subject to regular
and increasing scrutiny by institutional investors and
Mark Emanuel Semler Brossy Consulting Group
Seymour Burchman Semler Brossy Consulting Group
Goal Setting: Meeting Stakeholder Expectations in an Increasingly Dynamic and Complex World
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
7 Third Quarter | 2015
proxy advisers such as Institutional Shareholder Services (ISS) and Glass-
Lewis & Co. LLC.
With so much at stake, what is a company to do?
On the following pages, that question is answered in two ways. First, the article
describes three sets of criteria to help guide companies in establishing mean-
ingful goals that serve to meet key stakeholder expectations. Next, two cases are
provided to illustrate how these criteria can be used in varying situations.
CRITERIA TO GUIDE INCENTIVE GOAL SETTINGThe three criteria for effective goal setting in an increasingly dynamic and
complex world are:
1 | Adherence to stakeholder expectations
2 | Accommodation of organizational capabilities and the business environment —
the factors that affect the company’s ability to meet expectations
3 | Accordance with benchmarks and standards of success.
1. Adherence to
Stakeholder Expectations
In its simplest form, this first set of
criteria asks, “What should we do to
compete in this new world where
sustainable success requires that we
satisfy a broad range of stakeholders?”
The objective is to define what is
required to advance key shareholder
priorities, enable the company to
meet its various talent needs, as
well as meet the requirements of
customers, suppliers, regulators and
the community at large over the
short, intermediate and long terms.
To simplify the explanation of this
element of the framework, the focus
has been narrowed to two of the
company’s primary stakeholders —
shareholders and employees.
❚❚ Shareholders. Traditionally,
shareholder priorities have
been narrowly defined in terms
of creating shareholder value
(i.e., through stock price appre-
ciation and dividends). However,
The Growing Range of Stakeholders
Contemporary thinking highlights the need to consider a broad spectrum of stakeholders in a way that extends beyond the traditional view that shareholders are the singularly relevant stakeholder. Several factors have driven this shift in philosophy:
❚ The war for talent has gone global and innovation-fueled startups increasingly compete against established companies for the best and brightest.
❚ The pervasiveness of the Internet and mobile devices has made the acquisi-tion and retention of customers more difficult because they are increasingly knowledgeable of a company’s products, other customers’ level of satisfaction and competitors’ offerings.
❚ Supply chains have also become global and suppliers have become as funda-mental to product cost and quality as in-house operations.
❚ Meeting the ever-broader requirements of regulators has become a necessity, with fines or more onerous penalties a conse-quence of not doing so.
❚ The clout of the community at large has grown as interest groups and govern-ments demand more of corporations. Ignoring these outsiders heightens the potential for reputational risk.
8 WorldatWork Journal
issuers are encouraged to go both broader and deeper to focus on the key
drivers that influence sustainable TSR. This helps provide greater guidance to
the management team, focuses on aspects of performance that are more directly
controllable and enhances line of sight. Some of the drivers of TSR that might
be addressed are:
❚❚ Financial. The key financial measures that influence a company’s stock
price performance.
❚❚ Strategic. The strategic priorities that enable the company to build sustainable
competitive advantage and remain viable over the long term.
❚❚ Operational. The aspects of operational excellence (e.g., timeliness, reliability,
quality) that are fundamental to both financial results and the achievement of
strategic priorities.
❚❚ Employees. An organization’s talent goals should be defined comprehensively
to address the entire employee value proposition that enables the company to
attract, retain and motivate the desired talent (e.g., direct and indirect financial
rewards, affiliation, work content, career opportunities). The value proposition
should explicitly reflect the company’s unique talent needs. For example, a
struggling brick-and-mortar retailer making a foray into e-commerce might
appeal to expert-upon-hire e-commerce talent through incentive plan designs
that are sufficiently insulated from the results of the company’s brick-and-
mortar operations.
The focus in this article is on direct financial rewards. These need to be
sufficiently competitive to enable attraction and retention of key talent and be
appropriately structured to motivate and engage this talent to meet the expecta-
tions of the other stakeholders.
Balancing the shareholder and employee expectations with organizational capa-
bilities and the business environment requires that, over time, goals should result
in performance and payouts that:
❚❚ Are sufficient to generate returns above the company’s risk-adjusted cost of
capital and, in turn, drive shareholder value creation.
❚❚ Align relative pay and performance with peer companies.
❚❚ Represent a fair sharing of the value created between executives and shareholders.
❚❚ Are sufficient to keep executives retained, motivated and engaged.
2. Accommodation of organizational capabilities and the business environment
that affect the company’s ability to meet expectations
At its core, the second set of criteria asks, “What are we able to do?” The objective:
setting realistic goals given the organization’s strengths and weaknesses, as well
as the operating environment’s constraints and opportunities.
Organizational capabilities include the company’s talent, brand, infrastruc-
ture (e.g., systems and core processes, tangible assets, especially those that are
9 Third Quarter | 2015
proprietary), and intellectual
property. These include what is
available in-house and/or through
alliances and partnerships.
The assessment of the business
environment is meant to iden-
tify the various headwinds and
tailwinds that exist outside the
company. This assessment covers
a whole spectrum of factors,
including macroeconomic health
and trends, competitive influences,
technological advancements and
regulatory developments.
The more detailed the assessment
of the organizational capabili-
ties, the more valuable this set of
criteria can be in the goal-setting
process. What opportunities
exist to improve the company’s
support of the long-term strategy?
Over what timeframe? At what
cost and benefit?
The same is true for evaluating
the business environment. What
is a reasonable or likely range
of potential changes to environ-
mental factors over the established
timeframes and what is the associ-
ated impact on the business? This
process may begin as a qualitative
exercise, but should ultimately
become a quantitative one.
It is helpful to examine a range
of possible scenarios to better
understand and quantify the
impact on results for the different organizational capabilities and external head-
winds and tailwinds. (See Figure 1 on page 10.)
3. Accordance with benchmarks and standards of success
In its simplest form, this means “What does good performance look like on a
long-term, sustainable basis?” These benchmarks are the references against which
Proxy Advisers — the Shadow Stakeholder —
and Meeting Their Priorities
❚ Further complicating the challenge for issuers has been a heightened focus on the rigor of goal setting by another stakeholder in today’s say on pay-centric world — proxy advisers. ISS is the most influential of the proxy advisers, and in recent years, the frequency of ISS’ commentary on the rigor of goal setting has increased significantly.
❚ To date, ISS has largely taken an overly simplistic approach in its assessments of goal-setting rigor. Its assessments focus primarily on whether the goal (at target) increased year-over-year relative to the prior year target and/or actual results and whether payouts have been consistently above target. This approach often lacks in-depth analysis and consideration of the company’s specific circumstances, but frequently carries meaningful weight in ISS’ final vote recommendation. This element of ISS’ review is particularly troublesome for companies in turnaround situations, which might reasonably be expected to lower goals year-over-year to ensure continued engagement and motivation of employees.
❚ ISS has signaled that more is likely on its way with this topic. ISS’ 2014 policy survey indicated that 43% of investors believe that if performance goals are significantly reduced, target award levels should be modified commensurately (Institutional Shareholder Services 2014). In October 2014, ISS announced its acquisition of Incentive Labs. In its press release, ISS (2014) touted Incentive Labs’ proprietary analytical tools for assessing the rigor of performance targets and “measuring the efficacy of the link between pay and performance.”
❚ Given ISS’ present (and possibly increasing) focus on goal setting, the ability to describe the rationale and rigor underlying goals in the Compensation Discussion and Analysis (CD&A) is of the utmost importance, particularly for turn-around companies. To this end, the framework can serve as a foundation to structure more nuanced and thoughtful disclosures for share-holder and proxy adviser constituents.
10 WorldatWork Journal
the business measures itself and others measure it. They endure over the long
term and through economic cycles. Common references include:
❚❚ Historical performance of the company (e.g., average, best ever) and peers
(e.g., median, top-quartile)
❚❚ Best-in-class performance among the company’s own business units (where
there are comparables) and other relevant benchmarks for key parts of the
value chain, regardless of industry (e.g., firms that are exceptional in marketing
and sales, supply chain, manufacturing, managing inventories, serving
customers, etc.)
❚❚ Analyst expectations for the company and peers/industry
❚❚ Theoretical limitations (e.g., outcomes if existing processes are optimized).
These standards should be measured over a time horizon that is sufficient
to encompass multiple economic cycles and reflect the underlying timeframes
for key stakeholder expectations. When used for goal setting, these standards
should be considered with respect to the amount of value that is created by the
organization when achieved (i.e., the extent to which returns will exceed the
risk-adjusted cost of capital).
The most effective standards are established through rigorous benchmarking
in a multi-staged process. This process starts with the high-level financial
drivers of value creation (e.g., top-line revenues, bottom-line earnings, returns),
but then delves deeper into the second-order operational and strategic factors
that drive financial performance. Not only will this process inform goal setting
within a given incentive plan, but it will, by establishing a connection between
FIGURE 1 Illustration of Scenario Modeling
85
100
120
DownsideRisk
B4 B3 A4 A3 Baseline A1 A2 B1 B2 UpsideOppty.
Organizational Capabilities
A1 – Productivity improvement (+)A2 – Defer technology investment spending (+)A3 – New product risks (-)A4 – Accelerated hiring to fill key positions (-)
External Factors
B1 – Economic recovery improves sales trend (+)B2 – Decline in raw material costs (+)B3 – Increased regulatory costs (-)B4 – Competitors’ increased promotional activity (-)
11 Third Quarter | 2015
drivers, also help ensure consistency across the company’s various incen-
tive plans. In addition, it will help identify opportunities for performance
improvement and the actionable decisions that executives can make to drive
value creation.
CASE STUDIESUsing each of these three criteria, individually and in sum, will help provide a
structure and discipline to ensure that the goal-setting exercise meets the needs
of the company’s key constituents.
CASE STUDY 1: A CYCLICAL COMMODITY BUSINESS — TRUCKING Presenting challenge
Trucking and freight services are largely commoditized, and as a consequence,
business results are highly correlated with the macroeconomic cycle. Because
of the cyclical and unpredictable nature of the business, the actual operating
environment may vary significantly from the assumptions made during the
budgeting and planning process. Incentive payouts may thus vary widely inde-
pendent of management’s contributions to results. If gross domestic product
(GDP) growth is unexpectedly robust, achievement of the original goals may
warrant a smaller payout; if GDP growth is unexpectedly weak, achievement
of the original goals may warrant a larger payout. The trucking company’s
challenge in this case study was to design an incentive program that reflected
the uncertainty of the goals and ensured that payouts, on average, were fair
to shareholders.
Criteria 1: Stakeholder expectations
Given the cyclical nature of the business, key stakeholder expectations in trucking
are highly dependent upon the time horizon. Over the short term, shareholders
expect management to be flexible and responsive to a cycle-dependent operating
environment. Shareholders also expect that executives will manage the asset base
and costs at a level that is sustainable throughout the cycle.
Talent needs are equally cycle-dependent. Management needs to be retained
and engaged in a down-cycle when incentive plans are typically underwater.
Management also needs to be motivated to stretch for incremental opportuni-
ties in an up cycle when incentive plans are likely to be a layup. Thus, it is
critical to ensure that goals have an appropriate level of stretch and executives
have sufficient line of sight to the levers that can be pulled to drive results
in various market environments.
Over the long term, shareholders want management to optimize the asset base
and maximize operational efficiencies to ensure shareholder value is created
over the entire cycle, and multiple cycles, by generating returns that exceed the
company’s cost of capital.
12 WorldatWork Journal
Criteria 2: Organizational capabilities and external factors
The cyclical nature of the external operating environment is an important deter-
minant of top-line performance over both the near and longer term. Because it is
difficult to predict when the cycle might turn, the actual operating environment
may not reasonably be understood at the time the goals are set.
In addition to top-line growth, the ability to build value for shareholders is
largely a function of asset efficiency (e.g., fleet and supporting infrastructure) and
the development of effective systems (e.g., manpower planning, fleet management
and logistics) that enable the company to operate at the lowest possible cost while
meeting the needs of customers.
Criteria 3: Benchmarks and standards of success
Because top-line and bottom-line growth are both highly correlated with GDP
growth, the company focused on operating leverage as the basis for developing
a benchmark. Analysis of company and peer historical results and extensive
financial modeling indicated that, over the cycle, the operating leverage could
and should achieve a 3% premium in operating profit growth above GDP growth.
The Solution
Given the uncertainty associated with forecasting GDP growth, the appropriate-
ness of annual goals is more accurately judged at the end of the year when actual
GDP growth is known. Therefore, the company introduced a formulaic approach
for the annual plan whereby goals were adjusted after the fact to reflect how
the actual operating environment varied from the assumptions underlying the
budgeting process. (See Figure 2 on page 13.) Adjustment criteria were established
based on the sensitivity of results to assumed GDP growth.
This approach to goal setting recognized that exogenous factors are a key driver
of results and that management’s response to any given market environment is
an element that can be measured and rewarded. It also provided a mechanism
to provide relief (and improve retention) in a down cycle and require additional
stretch in an up cycle. Further, it avoided the need for blanket discretion and
reduced the incentive for management to sandbag the budget.
To complement the goal-setting solution, the company incorporated a number of
operational and non-financial goals in the annual incentive plan to encourage the
behaviors and decision making that would allow the company to generate returns
in excess of the cost of capital. Cost per ton shipped and equipment downtime
were included to determine improvements in fleet efficiency and management.
Measures of capacity utilization were used to assess whether managers were taking
the actions necessary to “right size” the business to the market environment. These
measures were seen as being controllable in the short term and across a variety
of market environments and were also required to achieve the desired long-term
3% premium in operating profit growth relative to GDP.
13 Third Quarter | 2015
For the long-term incentive plan, the company used the 3%-premium-to-GDP
standard as the basis for setting long-term operating profit goals (e.g., the multi-
year goal was set at 3% above forecasted GDP growth irrespective of positioning
in the cycle). For example, if GDP growth was forecast at 2.5% over the next three
years, the operating profit goal would be set at 5.5%. This approach to long-term
incentives rewarded and encouraged strategic investments and behaviors that
provided the foundation for long-term, sustainable value creation.
CASE STUDY 2: AN INDUSTRY-LEADING PERFORMER — RETAILThe study company, a retailer, had delivered industry-leading growth in prof-
itability and shareholder returns on a regular basis. Its strategy focused on
constantly improving its customers’ experience and making disciplined, long-
term investments in a seamlessly integrated omnichannel presence (i.e., brick
and mortar and e-commerce) to yield sustained growth and top-quartile TSR.
This strategy was supported by a results-oriented culture that sought constant
improvement without shying away from aggressive stretch goals.
FIGURE 2 After-the-Fact Adjustment Schedule for Case Study 1
Illustrative Adjustment SchedulePlan Payout as a Percentage of Target
GDP Growth
Actual Profit v. Budgeted Target
+0.6% to +1.5%
+1.6% to +2.4%
Expected: +2.5% to
+3.5%+3.6% to
+4.5%+4.6% to
+5.5%
170% 200% 200% 200% 200% 200%
160% 200% 200% 200% 200% 167%
150% 200% 200% 200% 200% 133%
140% 200% 200% 200% 167% 100%
130% 200% 200% 200% 133% 67%
120% 200% 200% 167% 100% 33%
110% 200% 200% 133% 67% 0%
100% 200% 167% 100% 33% 0%
90% 200% 133% 67% 0% 0%
80% 167% 100% 33% 0% 0%
70% 133% 67% 0% 0% 0%
60% 100% 33% 0% 0% 0%
50% 67% 0% 0% 0% 0%
40% 33% 0% 0% 0% 0%
30% 0% 0% 0% 0% 0%
14 WorldatWork Journal
The challenge for this company was twofold: to drive sustained improvement
in long-term returns and to ensure the alignment of pay with both absolute
and relative performance. That challenge was complicated by the company’s
successful track record, which created a high bar from which to grow. Coupled
with continuing pressure to set ever more difficult goals, the company could easily
find itself paying executives at the market median for performance materially
above competitive levels or paying little or nothing for average performance. In
either case, the outcome would likely be demotivating to executives and potentially
heighten retention risk.
Criteria 1: Stakeholder expectations
Shareholders were expecting the company to sustain its history of industry-leading
performance and continue to deliver long-term shareholder value creation. To do
this, the company needed to: 1) focus on a disciplined approach to investment in
its omnichannel strategy; 2) continue to deliver an outstanding customer experi-
ence; and 3) seek to retain and motivate the existing leadership team, which the
board and shareholders valued greatly.
From an employee perspective, pay and performance needed to be aligned so
that pay was commensurate with the results produced on both an absolute and
relative basis. To achieve this, goal setting and incentive plan design (e.g., target
pay opportunities, payout leverage) needed to be calibrated to ensure that
industry-leading performance was rewarded with industry-leading actual/realized
pay, while industry-matching performance needed to be rewarded with industry-
competitive actual/realized pay. Further, goals needed to be set to support the
high-performance culture.
Criteria 2: Organizational capabilities and business environment
Thanks to an established and revered brand, outstanding talent, great store loca-
tions and a well-integrated e-commerce platform, the company maintained a
position of strength. The company’s ability to continually raise the bar, however,
was becoming increasingly constrained because the low-hanging fruit for opera-
tional improvements had already been harvested and the pool of desirable store
locations had been shrinking. The company needed to identify new opportunities
for growth and profitability and preserve its high-performance talent and culture.
Although forecasts indicated that the challenging market environment would
make significant year-over-year improvements difficult in the subsequent year,
over the three-year planning horizon, the market environment was likely to
brighten considerably.
Criteria 3: Benchmarking
The company had a history of performing at or above the 75th percentile on
all high-level financial metrics, such as revenue, earnings growth and return on
15 Third Quarter | 2015
invested capital (ROIC). Also, it had consistently improved performance over prior
years and beaten analyst expectations, leading to top-quartile TSR growth. Setting
an “on-average, over-time” standard for revenue or earnings growth did not seem
appropriate since the continuous compounding of high-growth would, over time,
require unrealistic increases on a dollar value basis. Instead, the company focused
on ensuring continued discipline in its long-term investments and improvements
to operational efficiency by establishing a targeted ROIC of 5% to 7.5% above the
company’s weighted average cost of capital (WACC).
The company’s benchmarking exercise had showed that it consistently underper-
formed peers on inventory turns and payroll as a percent of sales, both of which
were attributable to its long-standing focus on delivering a superior customer
experience. Executives realized that new technologies offered solutions that could
reduce costs and inventory levels without sacrificing the customer experience. In
addition, the company could reap additional benefits from the continued rollout
and expansion of its omnichannel platform. Improvements in these areas would
allow the company to deliver returns in the targeted range.
THE SOLUTIONBoth the new annual and long-term incentives were designed to maintain the
company’s best-in-class performance while ensuring that pay was aligned with
performance. The annual plan also focused on key strategic changes that would
be needed to support this performance.
The key financial metric in the annual incentive plan was operating income
and goal setting was deliberately divorced from the budgeting process. The target
goal was set at what was considered to be median performance, and yielded a
target (i.e., 100%) payout. However, internally, all budgeting and planning were
focused on providing the progress needed to meet the on-average, over-time ROIC
goal, which was also generally consistent with 75th percentile earnings growth
(determined to be about 15% above median levels). Based on historical analysis
of peers, this level of earnings performance was consistent with a payout at 150%
of median. Thus, the payout curve was structured so it paid target for median
performance and 150% of target for stretch performance. (See Figure 3 on page
16.) Knowing that the macroeconomic environment presented headwinds for the
coming year, planned first-year growth was relatively modest. Subsequent growth
goals were set more aggressively to ensure that the annual goals delivered the
targeted multiyear performance levels.
Asymmetric payout curves were used to help emphasize the top-tier focus.
Thresholds were set at 90% of goal (median performance) and maximums were set
at 125% of goal, well above the stretch target of 115% of goal. Payouts were 67%
of target at threshold and 225% of target at maximum, providing significant upside.
To emphasize the changes needed to support both near-term and longer-term
performance, nonfinancial objectives were used to reinforce improvements in
16 WorldatWork Journal
inventory turns, employee productivity and the omnichannel rollout. Goals for
the first two metrics were initially targeted at peer levels and later exceeded
them. To ensure that the goals were met without diminishing the company’s very
positive customer experience, a modifier was added to reduce payouts if those
scores dropped.
The long-term incentive plan focused on ROIC. The ROIC goal was also set
at median, but again, internal planning was organized around delivering stretch
75th percentile performance (consistent with the targeted return of 5%–7.5%
above WACC). Although high threshold and maximum levels were used in the
long-term incentives (LTI), payouts were not as leveraged, reflecting that stock
price movements would provide the desired levels of realized pay. The trajectory
of annual operating income and inventory turn goals was established to ensure
that executives would achieve the long-term ROIC goal. To further reinforce
strong relative performance, a relative TSR modifier was also included.
FIGURE 3 Annual Incentive Leverage Curve
Step 1 — Company establishes a “stretch” budget that is representative of 75th percentile performance
Step 2 — A review of competitive pay data suggests that a 150% of target payout would typically deliver 75th percentile pay
Step 3 — Payout leverage curve is shifted to deliver a 150% of target payout for achieving the stretch budget
Step 4 — Company reaffirms that the sharing rates are appropriate above/below target and that the 100% payout is representative of what is typically considered to be 50th percentile performance
67%
$90
100%
$100
225% 225%
$125
150%
$115
Typical Payout Leverage Curve for 50th P Budget
Payout Leverage Curve for "Stretch"
(75th P) Budget
% o
f Tar
get P
ayou
t
Payout Leverage Curve
17 Third Quarter | 2015
CONCLUSIONThe complicated exercise of goal setting can be both simplified and made more
effective through the use of the three-part goal-setting framework. The overarching
goals of incentive compensation design — to advance key shareholder priorities
and meet the company’s talent needs — can be supported by an approach that
focuses on the expectations of key stakeholders and the company’s ability to meet
those expectations. It can also be supported by consideration of the benchmarks
that will be used to assess results. Not only can this framework be applied across
a spectrum of company circumstances, including companies in cyclical industries
and companies with track records of sustained high performance, but can also
help companies faced with compensation challenges operate effectively in today’s
increasingly uncertain and complex environment.
AUTHORS
Seymour Burchman ([email protected]) is managing director at Semler Brossy Consulting Group. Burchman has been an executive compensation consultant for over 20 years. His work focuses on reinforcing key strategies and leading to improved shareholder value through the identification of performance measures and goal-setting processes. He has served companies across a wide range of industries, including financial services, health care, high technology and publishing.
Mark Emanuel ([email protected]) is senior consultant at Semler Brossy Consulting Group. He joined SBCG in July 2007. He has broad experience consulting to senior management and boards of directors on a variety of compensation and governance related matters in both public and private settings. Emanuel serves clients across all industries but with a particular focus in financial services and retail apparel. Emanuel graduated from Harvey Mudd College with a bachelor of science degree in engineering and a minor in economics.
REFERENCES
Institutional Shareholder Services. 2014. “2014-15 Policy Survey Summary of Results.” Viewed: May 19, 2015. www.issgovernance.com/file/publications/ISS2014-2015PolicySurveyResultsReport.pdf.
“ISS to buy Incentive Lab.” Oct. 16, 2014. Institutional Shareholder Services. Viewed: May 19, 2015. www.issgovernance.com/iss-services-acquire-incentive-lab.
18 WorldatWork Journal
Substantial research supports the hypothesis that
aligning rewards with performance motivates
employees to put forth their best efforts and
focus on achieving the desired results. In order for that
motivation to materialize, employees must understand
what the objectives are, consider them attainable and
believe that the rewards for performing will be equi-
table, competitive, valuable and appropriate. Gallup
research indicates that “knowing what is expected”
is the most impactful factor on employee satisfaction
and effectiveness (Coffman and Buckingham 1999). So,
clarity up front about what the objectives are and how
results are measured is a prerequisite for motivation.
If a plan is to provide the incentive to perform well,
employees must know how the organization defines
performance and what their role is in helping the orga-
nization perform well.
Defining performance at the organizational, unit
and individual levels can provide clarity about criteria
and standards. The standards must be measurable, but
measurable does not necessarily mean quantitative.
Many measures are based on subjective judgments.
Robert J. Greene Ph.D., CCP, CBP, GRP, SPHR, GPHR, CPHRC
Incentive Compensation:Measures Matter
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
19 Third Quarter | 2015
For example, how well a customer believes he/she has been treated is that
person’s reality, and will influence future actions. In an incentive plan for
customer-service personnel based on customer evaluations of the service
provided, the metric for evaluating performance is subjective, but measur-
able. Often incentive-plan designers try to select only performance measures
that are countable. This may be wrongheaded because countable does not
always mean important. Although it may be more comforting to use hard
numbers when determining incentive awards and defending their fairness to
plan participants, it may result in underweighting criteria that by necessity
are based on subjective judgments.
USING MEASURES TO SET PERFORMANCE STANDARDSWhen performance measures are identified, a choice must be made as to how
the performance standards are set. There are three options:
1 | Compare current results to past results (look back).
2 | Compare current results to industry or competitor results (look around).
3 | Set standards at a level that is required in the future (look ahead).
No matter the approach used, the standards must be considered reasonable
by plan participants.
Based on Past Results
Basing standards on current results compared to past results is reasonable
if the present is similar to the past. But if a measure such as compounded
improvement in ROI is used in a troubled industry or struggling company,
basing standards on the past may be unreasonable. In the past, newspapers
in the United States often enjoyed ROIs of 20% or more. In today’s environ-
ment, returns of half that amount may be difficult to attain. So establishing
a threshold at which an incentive plan begins to be funded based on better
returns than those realized in the past would almost certainly result in no fund.
When setting plan thresholds and targets, it is advisable to examine long-
term trends to recognize what has happened in the past. Examples 1–4 present
different patterns of past results and each raises questions as to where the
threshold and target for a current year plan should be set.
Example 1: Improving Results
The trend evident in Example 1 shows that results have been improving
in the past 20 quarters. It might therefore be reasonable to assume that
establishing the plan threshold at the level of last year’s results is realistic.
Incentive plans are typically intended to reward improvement over what
would have probably occurred without them. This assumes management
does not believe that major changes have occurred that would significantly
affect current performance.
20 WorldatWork Journal
The fact that the results in
the past 10 quarters have
been relatively f lat might
cause management to set the
plan threshold even with the
past few quarters and require
improvement in order to
trigger an incentive fund.
Example 2: Growth,
Stability, Decline
This pattern in Example 2
poses considerable chal-
lenges. It appears results
showed major improve-
ment over the first seven
quarters, stabilized over the
next eight and then turned
sharply downward in the
most recent six quarters. If
the organization is oper-
ating in an industry subject
to major swings, consider-
able judgment is necessary
when establishing a plan
threshold and a target. This
pattern suggests variation
every two years. The most
recent three quarters suggest
a bottoming out and perhaps
an upturn, so management
may be comfortable setting
the current year threshold
and target at the level realized in the past three quarters. On the other hand, if
there is a pattern of industry fluctuations that cycle every six to eight quarters,
it would be reasonable to believe that results for the next four to six quarters
will improve substantially. If that is the expectation, the threshold and target
may be set at a level above the most recent results.
Example 3: Seasonality
Example 3 indicates there is seasonality in results. If the incentive plan is an
annual plan, it would be reasonable to average the four quarters in each of the
EXAMPLE 2 Growth, Stability, Decline
Last 20 Periods: Where to Set Baseline?
• •
• • • •
• •
• • •
• •
• • •
•
•
•
•
EXAMPLE 1 Improving Results
Last 20 Periods: Where to Set Baseline?
• • • •
• • • •
• • •
•
• • •
• •
•
•
•
21 Third Quarter | 2015
five prior years and set the
threshold and target based
on the trend of those aver-
ages. In this case, it would
indicate an upward trend,
arguing for a threshold and
target somewhat higher than
last year’s average perfor-
mance. On the other hand,
if the plan is tied to quarterly
results, as some sales plans
are, another pattern should
be taken into account. It
appears that the dispersion
of quarterly results is getting
larger, which may call for
consideration of moving to
an annual plan or manage-
ment may find the measure
being used is too erratic to
determine the size of the
incentive fund. In any event,
management should attempt
to determine the causes of
this pattern and consider
alternative strategies.
Example 4: Seasonality With
an Upward Trend
This pattern in Example 4
also suggests seasonality
in the results, as well as
an upward trend overall.
Because there is less dispersion than the pattern shown in Example 3, it
would be reasonable to conclude that results are on an upward trend and that
the threshold and target should be set at or above the level of results during
the prior year.
No matter what the pattern, if current results are to be compared to past
results, it is mandatory that a historical analysis is extended sufficiently into the
past to detect trends. It is also necessary to attempt to explain trends that are
apparent. For example, if an upward trend is attributable solely to significant
capital investment by the organization in technology, equipment or product
EXAMPLE 3 Seasonality
Last 20 Periods: Where to Set Baseline?
•
•
•
•
• • •
• • •
•
• • •
• •
• • • •
EXAMPLE 4 Seasonality With an Upward Trend
Last 20 Periods: Where to Set Baseline?
•
•
• •
• •
• • • • •
• •
• • •
• •
• •
22 WorldatWork Journal
improvement, management should set the threshold and target high enough to
ensure that the incentive fund rewards better performance rather than merely
the improvements attributable to the organization’s investments.
Based on Current Results of Industry/Competitors
When an organization is in an industry with identifiable competitors and the
results of those competitors can be determined, it may be advisable to use
industry/competitor results rather than past results to establish plan thresholds
and targets. This is particularly the case if the current environment differs from
the past and there is significant uncertainty about future economic conditions.
In some industries, associations gather performance data and share the aggre-
gate results with members. For example, some industry associations gather
operating and financial data in an annual survey that enables users to determine
how their profits, revenue growth, cost of sales and other measures of perfor-
mance compare to their peers.
One of the difficulties with the “looking around” approach is that competitors
often balk at sharing their results. Publicly traded organizations can use measures
such as total shareholder return as their basis for comparison to other publicly
traded organizations since some performance measures are readily available
through proxy statements and other required filings. However, if an organization
wishes to use performance criteria that are not reported publicly, it will have diffi-
culty using competitor results to establish performance standards for an incentive
plan. But even measures based on subjective data may be used if industry rankings
(i.e., J.D. Power and Associates; Consumer Reports) are accessible.
Based on Future Requirements
In some cases, standards must be set at minimum levels before incentive funds are
made available. Boards often mandate that a minimum level of shareholder return
or profits be produced before incentives can be paid. This minimum level may
be viewed as performance that justifies the cost of salaries and benefits and that
minimum level must be exceeded before any further compensation is made avail-
able. Profit-sharing plans often have a formula that prescribes an incentive fund
that equals a set percentage of profits after a predetermined return to shareholders.
Organizations facing economic difficulty may need to set the incentive plan
threshold at a survival level of performance. Turning around the fortunes of the
organization may require considerable effort on the part of competent people,
so getting, keeping and motivating the required effort may necessitate offering
incentive opportunity.
SELECTING MEASURESHistorically, measures used for organizational incentive plans were dominated
by past financial results. Kaplan and Norton (1992) were influential in providing
23 Third Quarter | 2015
a broader view and richer definition of organizational performance, including
customer and people measures to augment financial measures. In a later book,
Kaplan and Norton (2004) provided examples of how multiple measures can be
integrated. Financial and operating results can usually be based on quantitative
data in a relatively unarguable way.
Customer measures are more difficult to obtain due to the limited availability
of data and difficulty in getting at what the organization is trying to measure.
For example, market share for a product can generally be determined, but an
analysis of which customers are contributing to that share may be more diffi-
cult. It may also be challenging to measure customer loyalty (repeat purchases)
or share of wallet (what percentage of the customer’s expenditure goes to
the organization). Getting accurate measures can be especially important in
sales incentive plans, since different commission rates may be set for new
versus existing customers or purchases of new/more profitable products versus
recurring purchases of mature/less profitable products. One of the most chal-
lenging measures to use is customer satisfaction. Hotels often place customer
response cards in the room and airlines often send follow-up emails to gauge
customer experience. But these attempts tend to succeed in only hearing from
customers who are unhappy, limiting the responses to complaints rather than
valid indicators of satisfaction.
Finally, people measures pose significant challenges. HR functions are some-
times evaluated using measures such as average time to fill a vacancy, turnover
and implementing new systems on time and on budget. These quantitative
metrics can be misleading. Vacancies can be filled very quickly by lowering
selection standards. Turnover can be minimized by paying significantly above
market levels. And systems can be implemented quickly and on budget but not
deliver the quality desired. Some criteria require further definition to decide
how measures should be used and how standards are set. Whether human
resources should be charged with managing total turnover at a specified level
is debatable. Example 5 breaks down turnover into several types and in this
case what on the surface looked like a staggering level of turnover (24%)
turned out to be a much smaller number (13%) if only dysfunctional turnover
was used as the measure.
Example 5: Determining Dysfunctional Turnover
When incentive plan participants from human resources are measured on turnover,
the types of turnover should be factored in (see Example 5) and some consideration
given to their ability to affect results. Human resources may design sound recruit-
ment, onboarding and compensation programs but the impact of these programs
on retention will be determined by how well they are executed. The principle
of holding people accountable and rewarding them based only on things they
can influence should be followed. Therefore, a great deal of thought should go
24 WorldatWork Journal
into deciding what measures
drive incentive awards.
USING MEASURES TO EVALUATE INCENTIVE PLAN EFFECTIVENESSImplementing an incentive
plan to motivate employees
to produce improved perfor-
mance is a common practice.
Often it is difficult to deter-
mine if the plan had a positive
impact on the performance
measure the organization
wished to improve. Evaluating
results before and after plan
implementation can provide
information on the impact
of the plan. Examples 6–9
are illustrations of attempts
to determine whether the
desired results materialized.
Example 6: Improvement
Since Implementation
In Examples 6-9, X repre-
sents the plan and time
at which the plan was
implemented. In Example
6, it appears that results,
however measured, steadily
improved after implemen-
tation. Since historical data were not evaluated, it is unclear whether the
improvement would have occurred without the plan. This example illustrates
the importance of ensuring measures are tracked prior to taking an action so
that the impact of that action can be evaluated.
Example 7: Implementation During an Upward Trend
If the measure used in Example 6 had been tracked historically, a different
conclusion might be justified about the impact of the plan. It appears in
Example 7 that results were on an upward trend before plan implementation
and it is therefore difficult to give the plan much credit for improving results.
EXAMPLE 7 Implementation During an Upward Trend
Did “X” Make a Difference?
EXAMPLE 6 Improvement Since Implementation Did “X” Make a Difference?
º
º
º
º
º
º
º
º
º
X
º
º
º
º
º
º
º
º
º
•
•
•
• X
25 Third Quarter | 2015
It is possible that the plan
contributed to sustaining
improvement but t ha t
cannot be demonstrated by
tracking the measure.
Example 8: Improvement
After Plan Implementation
In Example 8, there is strong
support for concluding that
the implementation of the
plan changed the pattern
of results. The historical
data show results vaci l-
lat ing somewhat with a
sl ight downward t rend.
Immediately a f ter plan
implementation, the pattern
shifted abruptly, suggesting
an initial and ongoing posi-
tive impact on results. There
may have been other factors
that caused the pattern to
change. Those factors should
be explored to ensure that
at t r ibut ing improvement
solely to the incentive plan
accurately reflects reality.
Example 9: Increased
Improvement
After Implementation
Example 9 suggests the plan had an immediate positive impact on results
and that this was sustained, at least up to the present. Although there was
an upward trend in results prior to plan implementation, the slope was more
gradual than after implementation and there was a decided jump from just
before to just after implementation. The immediate jump might have been a
Hawthorne effect, caused by an increased focus on the results triggered by
communication during plan design and implementation. But it is reasonable
to conclude the plan had a positive impact on results that was sustained at
least over the periods measured.
EXAMPLE 8 Improvement After Plan Implementation
Did “X” Make a Difference?
EXAMPLE 9 Improving Results
Last 20 Periods: Where to Set Baseline?
º
º
º
º
º
º
º
• • º
• • • º
• •
X
º º
º º
º
º
º º
º
• • •
• •
• • •
•
X
26 WorldatWork Journal
CONCLUSIONIncentive plans can be a source of motivation, encouraging increased effort and
sharper focus on specific results the organization wishes to improve. Once the
criteria used to define performance are established, it is necessary to decide
how results relating to those criteria will be measured. Then the performance
standards must be determined so that the incentive plan threshold and target
can be established. Setting thresholds too low can result in an incentive fund
that is too large and does not provide a return that justifies the cost. Setting
thresholds too high can lessen the perception that the results required to
generate an award fund can be achieved, resulting in cynicism rather than an
increased motivation to perform.
The standards established for a plan can be based on what the organization
has achieved in the past, what competitors are achieving currently or what the
organization must achieve in the future. Historical results must be evaluated
if the threshold is set based on what has been done to increase the prob-
ability that it can be achieved and/or that it is set at a challenging level that
warrants an incentive award. If the threshold is set by comparing to industry/
competitor current performance, the organization must ensure the comparator
sample is reasonable and that accurate performance data can be obtained. If
the plan threshold is set at a future level, it is necessary to be able to create
reasonable forecasts.
Determining Dysfunctional Turnover
Is Dysfunctional Turnover Too High?
Avoidable Dysfunctional
7%
Dysfunctional9%
Involuntary3%
Voluntary13%
Functional4%
Unavoidable2%
External16%
Total Turnover24%
Internal 8%Functional 4%Disfunctional 4%
27 Third Quarter | 2015
A final use of measures is to determine if the adoption of an incentive plan made
a positive difference large enough to justify its cost. That determination should be
based on a comparison of pre- and post-plan implementation results.
Measures matter. The right measures must be selected and the information
necessary to measure results accurately must be available. It may be necessary
to employ measures that are subjectively determined but all interested parties
must accept that they are appropriate. The performance standards that are
attached to the measures must be set at a level that is both challenging and
attainable. Both the organization and incentive plan participants must believe
they are equitable and competitive. Getting measurement right is not easy but
mandatory for incentive plans to accomplish the organization’s goals.
AUTHOR
Robert J. Greene, Ph.D., CCP, CBP, GRP, GPHR, CPHRC ([email protected]) is the CEO of Reward Systems Inc. and a faculty member at DePaul University’s master of business administration and master of science in HR degree programs. He has more than 30 years of industry and consulting experience and has published more than 100 articles and book chapters, as well as the 2011 book “Rewarding Performance: Guiding Principles; Custom Strategies.” He was a principal designer of the CCP and GRP certifications and the first winner of the WorldatWork Keystone Award for achieving the highest level of excellence in the field.
REFERENCES
Coffman, Curt and Marcus Buckingham. 1999. First Break All the Rules. New York: Simon & Schuster.
Kaplan, Robert and David Norton. 2004. Strategy Maps: Converting Intangible Assets Into Tangible Outcomes. Boston: Harvard Business School Press.
Kaplan, Robert and David Norton. 1999. “The Balanced Scorecard — Measures That Drive Performance.” Harvard Business Review, January-February, 71-79.
28 WorldatWork Journal
Frank Giancola
Is the Safelite Auto Glass Case Study All It’s Cracked Up to Be?
When compensation professionals look for
support in academic research on the effec-
tiveness of financial incentives to motivate
employees, they are likely to find the case study of
Safelite Glass Co. (Lazear 2000). At the time of the study
in 1995, Safelite was the largest provider of automo-
tive glass replacement and repair services in the United
States. In an industry that was highly fragmented with
approximately 20,000 competitors, it had the largest
market share of about 13%.
The company was founded in 1947 and is based
in Columbus, Ohio. In the 1990s, it was owned by a
large private equity investment company and a new
management team was trying to inject new life into
the company. It employed about 4,000 employees and
had annual sales of about $400 million. Company
services were provided through 500 company-owned
and independently operated service centers and 1,000
mobile vans. It is now a subsidiary of Belron, a vehicle
glass repair and replacement group based in Belgium,
which operates in 35 countries and employs more than
24,000 people.
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
29 Third Quarter | 2015
In 1996, Safelite made the business news by replacing its hourly pay system for
windshield installers with a piece-rate system, as management believed produc-
tivity was below where it should have been. The new pay system paid employees
based on the number of windshields installed instead of on the number of hours
they worked. Following the change, the company experienced a remarkable 44%
increase in output per worker.
The article is a popular one since there are few case studies on the productivity
effects of establishing a new compensation system using financial incentives.
The article has been cited 1,797 times by other authors since it original publica-
tion in 1996, according to Google Scholar. By contrast, a journal article describing
a case study involving skill-based pay by a well-known and highly respected
compensation scholar, Gerald E. Ledford Jr., has been cited 28 times since its
publication in 1991.
In 2000, the Safelite story received the distinction of being made into a Harvard
Business School case study. Harvard believes that the case study method of instruc-
tion, which places students in the role of decision maker, is the best way to prepare
students for the challenges of leadership.
Given the visibility and recognition the study has experienced, compensation
professionals should have a clear understanding of what it says and does not say
about motivating employees with financial incentives. This author has looked
at the article closely and used sources not described in the article to provide a
clearer picture of the circumstances behind this change and its effects on the
employees and company.
Like all good case studies, a great deal of information is presented but the
reasons why the main characters acted as they did are not always spelled out.
Readers must sift through and piece together key facts from the case, investigate
outside sources of information and use their business acumen to get a better
picture of what happened at the company.
A GOOD PLACE FOR A PIECE-RATE PLANSafelite offered an excellent setting for installing a piece-rate pay plan due to the
nature of the work of its dominant classification, windshield installer.
The task of installing a new pay plan was also made easier by the availability of
a sophisticated IT system to facilitate plan administration.
❚❚ Work output was easy to measure and verify. It is not difficult to keep track of
the number of windshields each employee installed per day.
❚❚ Windshield installers worked independently and their output was easy to isolate.
The incentive pay plan did not affect teamwork because there was little team-
work involved in the work.
❚❚ If installers worked too fast to boost their pay and quality suffered, it was easy
to identify the responsible worker, who then had to install a new windshield
on his/her own time.
30 WorldatWork Journal
Professional and managerial jobs may not be as well suited to piecework, so the
study has limited applications.
The New Pay System
Initially, Safelite management approached its employees with the idea of
improving productivity by granting an across-the-board base wage increase and
requiring a higher level of minimum output from each employee. Employees
rejected the idea because it did not affect each of them in the same way.
The employees favored a piece-rate system that would allow some to earn more
if they desired and allow others to continue at their present pay if they, in the
words of the case author, “put forth less effort,” and earn a minimum guaranteed
rate. Under the new system, employees would earn either the guaranteed amount
or the piece-rate wage, whichever was higher.
New employees were not covered by the guarantee, and “those who were
not making it were fired or quit early.” From statements like these and outside
sources cited later in this paper, the company was operating with a de facto
minimum output standard of employee performance. The case did not state
what the performance standard was under the piece-rate plan and how many
employees were discharged for not meeting it.
It is also not clear from the article how many, and for how long, employees
could continue to receive the minimum guaranteed amount without experiencing
negative consequences. The case study does indicate that “Many employees
ended up in the guaranteed range.” Safelite, however, knew that this would
probably be a temporary situation, as the company had a high employee turnover
rate (40%). So if the existing turnover rate continued after the new plan went
into effect, many employees would soon be subject to the piece-rate plan, as
current employees covered by the guarantee would be replaced by new ones
who were not. Turnover increased slightly under the new plan.
The change in pay plans was introduced gradually in 1994 and 1995, which
allowed the researcher to see how an employee’s performance changed from the
old to the new plan. Under the old plan, and under the guarantee, employees
were paid $11.47 per hour. Under the new plan, they received $20 for each
windshield they installed. A source outside the case indicates that the top hourly
rate of $11.47 was earned after employees mastered their craft (Passell 1996).
The employees earned on average 10% more under the new plan. Ninety-two
percent of incumbents experienced a pay increase, with 25% receiving an
increase of at least 28%, according to the study author (no further breakdown
is provided). These numbers suggest that a minority of the employees were
working to increase their earnings, while a majority might have been working
just enough to keep their jobs.
31 Third Quarter | 2015
The Effect on Productivity
The case does not indicate what management’s productivity goal was for the
new pay plan. Productivity did increase substantially during the study — to 44%.
The increase was split between performance improvements of incumbent
employees and hiring more productive employees to replace less productive ones.
The average number of windshields installed per day by each employee
increased from 2.7 per day (13.5 per week) to 3.24 per day (16.2 per week).
The industry average was 3.0, and later statements and writings by company
officials stated the daily rate rose after the study to 3.9 at Safelite, an indica-
tion that the improvements attributable to the piece-rate plan lived beyond the
19-month period in the case study. According to the case study, the new piece-
rate plan was supported by a sophisticated information management system, but
the cost of such a system would probably not “overwhelm” the positive effects
on the bottom line of the productivity improvement.
It is not clear how the employees improved their productivity. Through brief
references in the case, it appears that worker productivity was a matter of effort
more than skill. However, separating one from other might be difficult. Although
certain low-skill jobs involving physical tasks might appear to depend heavily on
effort, they can favor people with good manual dexterity, strength and stamina.
More capable employees actually might have to expend less effort to succeed in
such a job than less able ones.
No mention was made of new equipment or methods being introduced during
this period that could account for the improvement or to training that might enable
employees to work more efficiently.
The case is noteworthy in one respect — the employees did not seem to
receive a commensurate share of the gains attributed to the productivity improve-
ment. Although productivity improved by 44%, wages increased by only 10%
on average. In addition, the number of windshields installed each day (3.9)
exceeded the industry average (3.0) by 30%, which seems like a reasonable basis
for greater sharing of the productivity gain.
Employee Turnover
The case report indicates that only about half of the 44% productivity improve-
ment was attributed to incumbent employees working at a higher level of
performance. The other half was due to employee turnover. New employees
were performing at higher levels than the ones who left the company. Turnover
was substantial during the 19-month period of the study. The turnover rate was
43%, which is slightly above, but not statistically different than, the 40% figure
that existed before the plan went into effect. The tenure effects were significant,
according to the study. One year of experience raised productivity by 10%.
The mean employee tenure was 8 months.
32 WorldatWork Journal
Conclusive evidence that poorer performing employees were leaving in higher
percentages than higher performing employees could be not be found.
The pattern did exist, but the difference between the two groups was not
statistically significant.
Interestingly, although high performers thrived under the new plan and their
separation rate declined by 17% under it, they were still leaving the company
in substantial numbers — 35% per year (Lazear 1999). Apparently, the prospect
of earning higher wages was not enough to keep some employees from leaving,
even though they were capable of thriving monetarily under the new plan. This
study demonstrated that there is more to retaining high-performing workers than
rewarding their superior performance. It is possible that some employees did
not believe that they were being adequately compensated for their contributions
under the piece-rate plan.
It is frustrating that the case study did not report the reasons why people
were leaving in large numbers, even before the new plan went into effect. It is
not known if the company was not keeping separation records, if the researcher
failed to consider the records, or if Safelite cared about the issue and had plans
to address it. An outside source said that prior to the pay change, the company
was not a demanding employer and had tried to address the turnover problem
with high levels of training and supervision (Passell 1996).
However, there is some evidence that management adopted a more hard-
line approach under the new pay arrangement. Early on under the piece-rate
plan, when a windshield installation was defective, the fix was assigned at
random to other employees in the center. As noted in the case, the company
hoped to create peer pressure from other employees to get the responsible
person to improve his/her performance or quit. Later, this aspect of the plan
was abandoned and the fix was assigned to the responsible installer. But here
again, it appears the company adopted another failed practice that had to be
abandoned. The responsible worker not only had to install a new windshield
on his/her own time, but had to reimburse the company for the cost of the new
windshield before paying jobs were assigned (Lazear 1996). The final practice
did not require reimbursement.
One advantage of having a high turnover rate may have come into play at Safelite.
If a company desires to reduce headcount, there is less of a need to lay people
off and incur severance costs when employees are quitting in sufficient numbers.
Outside sources of information prepared by the company indicate that reported
service center closures made this a plausible dynamic (Safelite 1997).
About 3,707 employees worked as installers during the study, which had an
average monthly headcount of 2,040 (Lazear 1996). Without providing demo-
graphic or employment details, the case study author indicates that the company
had a very different workforce at the beginning of the study than it did at the end
after the new plan had been fully installed (Lazear 1999).
33 Third Quarter | 2015
PERSONNEL ECONOMICSData in the Safelite case were analyzed and described by a Stanford University
labor economist, Edward Lazear, who was the chief economic adviser to President
George W. Bush. Lazear is considered to be the father of a branch of labor
economics, personnel economics, which involves the application of economic
theory and methods to a company’s HR systems.
One technique used is insider economics, which is the application of econo-
metric methods (advanced mathematics and statistics, such as predictive modeling)
to massive amounts of confidential company information to uncover patterns to
help explain the effects of HR practices on company performance. In these studies,
which include the Safelite study, researchers generally do not introduce changes
in HR practices. They only study how prior changes affect company performance
or compare the effects of different practices at several companies.
The approach sometimes requires the use of simplifying assumptions that results
in concrete, but limited applications (Gibbons and Woock 2007). Lazear alludes
to this outcome in his explanation of how personnel economics differs from the
traditional HR approach: “The typical human resources text is verbose and short
on general principle. Indeed, many books eschew generalization, arguing that
each situation is different. The economist’s approach is the opposite. Rather than
thinking of each human resources event as separate, the scientific method that
economists use places a premium on finding the underlying general principle and
on downplaying other factors” (Lazear 2000a).
He also has stated that, “The language of economics allows the personnel econo-
mist to remove complexity. Details may add to the richness of the description,
but the details also prevent the researcher from seeing what is essential.” Based
on these statements, labor economists’ work can become problematic when they
downplay key causal factors and important contextual factors that challenge their
general principle by misjudging the significance of matters relating to HR manage-
ment as well as economics.
MISSING INFORMATIONThe effects of the personnel economics approach on the Safelite study can be
seen in the following characteristics of the case study:
❚❚ Company and industry conditions. No information is provided on the condi-
tions in the company or industry that might help explain what is driving the
change in pay practices. Even basic information about the company, such as
number of employees, revenue and market share, is absent from the article.
(This author has provided that type of information in this paper.) The case study
provided one piece of potentially important information about the company
related to the lower cost per unit of installing windshields under the piece-rate
plan: “The firm’s earnings are up substantially since the switch to piece rates,
but this could reflect other factors.” The statement cannot be given much weight
34 WorldatWork Journal
as the case study did not describe what the other factors were and how much
the new plan contributed to the earnings improvement — information that was
available to the case study author.
❚❚ Workforce profile. No description is provided of the Safelite workforce that
might help readers understand why the workers behaved as they did following
the introduction of the new plan. Basic information, which HR professionals
have come to expect in case studies by management researchers, is missing. It
does not provide the gender, age, marital status, education, company tenure
and industry experience of the windshield installers. Union affiliation is impor-
tant, but not shown. An outside source indicates 330 of the 4,000 employees
in the company were unionized (Safelite 1997). A comparison of the profile of
low-productivity workers and high-productivity workers that might help readers
understand why the latter were more productive than the former is not there.
The reader does not know, for example, if the company had a core group of
long-service employees and how those workers performed.
❚❚ Hourly pay practices. A detailed description of how employees were paid
under the hourly pay system is needed to ascertain if employees were paid an
entry-level rate and then a fully qualified rate after demonstrating job profi-
ciency. If this was the case, it would be important to know what criteria were
used to determine if an employee was eligible for the higher rate. An outside
document refers to employees being paid the top rate after they “mastered the
craft” (Passell 1996). The case mentions that the rejected pay plan “required a
higher level of minimum output,” as if one were in place before the new plan
went into effect (Lazear 2000).
❚❚ Worker earnings. Actual annual earnings of employees were not provided,
which would have given readers a feel for how the new pay system was impacting
actual worker earnings. This is a basic item in any study of how a pay system
affects worker motivation.
❚❚ Rationale for change. Regarding the rationale for the change, readers
are told only that management believed “productivity was below where it
should have been.”
❚❚ Employee briefings. No information is provided on what the employees were
told about why the change in pay practices was being made, which could affect
their desire to stay and perform at a higher level or eek employment elsewhere.
❚❚ Turnover reasons. The reasons why the company had a very high employee
turnover rate were not described in the case.
❚❚ High performers. About 35% of the high-performing employees left
the company during the study, but the case study does not explain why
they left, even though they appeared to be thriving under the new plan.
The study is long on the statistics used to analyze the data, presumably to satisfy
economists, but short on key information about the employees, the company and
35 Third Quarter | 2015
the economic conditions in the industry to allow readers to apply the research
results. One might expect light treatment of the first item from an economist, but
the lack of information about the company’s financial condition, strategy and the
economics of the industry are somewhat unexpected because they relate directly
to the case and are economic factors more than HR factors.
Greater Information Gaps Revealed
There are additional serious gaps in information in the case study if one considers
outside sources of information. These sources show how management dealt with
underperforming employees and identified major operating conditions that existed
in the industry and company at the time the piece-rate plan was introduced:
Job Loss and Low Performance
The Safelite turnaround is described in a New York Times article (Passell 1996)
that shows that the threat of job loss was real for underperformers. It suggests
the real possibility that employees were motivated to perform at higher levels for
fear of job loss, as much as to earn additional income.
“Safelite had its own kind of glass ceiling,” Mr. Barlow, Safelite’s president and
chief operating officer, joked. “No more than three glass repairs per installer a
day.” His proposed solution: piecework with a safety net. Under the plan, installers
receive $19 to $23 for replacement jobs in the shop or at owners’ homes, and about
half that for multiple jobs in the lots of rental car companies. They can also make
commissions for selling auto parts to customers — from $1 for windshield wipers
to 10 percent of the invoice for an alarm system. Workers’ pay never falls below
a guaranteed floor. But in a typical month, said Bill Rapp, the Safelite treasurer,
two-thirds of the workers earn more than the minimum, and in the long run,
almost all do. Indeed, if an installer clings to the safety net for long, “it is treated
as a management problem,” Mr. Rapp said.
No information could be located in the research report on how many
employees were discharged for low productivity. Efforts to improve an employ-
ee’s ability to work successfully under the new plan, such as additional training,
were not mentioned.
Job Loss and Company Closings
Two outside documents provide information about another possible reason why
employees might have been concerned about losing their jobs that caused them
to perform at higher levels — the closure of company service centers. These
documents suggest that the new incentive plan had a goal of increasing produc-
tivity so that headcount could be reduced without degrading service, a possibility
not mentioned in the article. One factor mentioned in the case supports this idea.
If productivity increased as stated by 44%, and the business was not growing,
fewer employees would be needed to operate the service centers.
36 WorldatWork Journal
❚❚ In an article in an industry publication, Aftermarket Business, Safelite’s CEO
admitted that there were “too many glass replacement stores already dotting
the country” (Aylward 1993).
❚❚ A Securities and Exchange Commission (SEC) document filed in 1997 by the
company stated a new management team had undertaken a comprehensive
review of the company’s operations. In 1992, restructuring charges totaling
approximately $10.0 million were recorded, including $2.7 million in connec-
tion with the planned closing of 72 service centers. In 1993, the company
recorded $4.6 million in restructuring charges related to the planned closing
of about 70 service centers. In 1995, Safelite recorded $6.3 million in restruc-
turing charges. Of this amount, $5.6 million related to the planned closing
of 100 service centers (about 15 percent of the total of about 650) and
$0.7 million related to field management reorganization.
There is a pattern of closing service centers in the period shortly before
and during the case study. The readers do not know how the closures were
handled from a personnel standpoint and whether affected employees were
offered jobs at other centers to fill vacancies. In any event, employees were
aware of the contraction in jobs and that the new pay plan was reducing
the need for employees as productivity was increasing by 44%. In one sense,
employees were working harder to put themselves out of work. So it makes
sense to cover the closures in the case study because it is one important prob-
able reason for installing a pay plan that would raise productivity to allow for
a reduced headcount level.
The SEC document also showed that the company had much riding on the
success of the new pay plan. It was as a key component of its competitive
strength of being the low-cost provider: “The Company’s management esti-
mates that its performance incentive program has increased the productivity
of its installation associates from 2.5 installations per day in 1991 to 3.9 per
day in 1996 (56% gain), while the industry averages an estimated 3.0 instal-
lations per day” (Safelite 1997).
These documents raise important issues that the study does not address.
Were some employees motivated to perform at higher levels by the prospect of
earning more money or by the fear of losing their jobs? Was the incentive plan
installed to support reductions in headcount? One reading of the study might
lead the reader to believe that the case demonstrates that financial incentives
motivate employees to work harder to better themselves financially. But one
cannot be confident another significant reason for working harder, the potential
for job loss, was not at play. It appears as though it was.
37 Third Quarter | 2015
SUMMARYThe Safelite case lacks great value for HR practitioners and many business people
for three reasons.
1 | Limited focus. The case’s narrow focus on low-skilled workers in a high-
turnover industry who are paid a piece-rate limits its applicability beyond these
particular circumstances.
2 | Missing information. A substantial amount of relevant information, which
would help readers understand why the company and employees acted as
they did, is missing. For example, little is known about management’s reasons
for making the pay change, how much information employees were provided
about it, workforce demographics, employee turnover reasons, how employees
were paid under the hourly pay plan and why highly productive employees left
the company in large numbers after the change was put in place. In addition,
the business conditions that the company and industry were operating under
were not addressed.
This lack of basic, practical information and key unanswered questions
greatly limits the value of the case in the real world. The case study does
present an impressive array of numbers, charts, statistics and explanations of
how they were determined. This will primarily help people understand the
mathematical relationship between pay and productivity, not an organizational
or behavioral relationship.
Five years before the Safelite study was published, an article on labor
economics and psychology by Lazear recognized this issue. “Economics is
sometimes accused of being sterile, unrealistic and inhumane. We are often
charged with ignoring psychological and institutional issues that may have
most of the explanatory power. This stereotype has some truth, bit it is
becoming much less accurate. In labor economics and other areas, previous
non-economic issues are being systematically incorporated it to economic
analysis” (Lazear 1991). The Safelite case is arguably one instance were an
expanded analysis seems in order.
3 | Job security versus motivation. The case is said to demonstrate that
employees are sensitive to incentives and that paying on the basis of output
will induce them to supply more output. On one level it did this, but the primary
effect of the incentive plan on most employees was to establish a minimum
level of acceptable performance for continued employment. Do incentive plans
typically create fear of job loss in a vital segment of the workforce to motivate
performance? Do they produce massive turnover? Are they part of a plan to
reduce headcount? Do they vastly change the composition of a workforce?
Safelite’s predicament represents an extreme situation that most compensation
professionals will not face in their careers.
38 WorldatWork Journal
The study provides little knowledge compensation professionals can apply to
their jobs. If they encounter it in academic and in pursuit of advanced professional
certifications, hopefully they will use the information in this paper to provide new
issues to consider.
AUTHOR
Frank Giancola ([email protected]) has written more than 90 articles analyzing HR trends and practices. He has over 40 years of HR experience, 25 with Ford Motor Co, primarily in various compensation and benefits positions. He has taught HR and compensation courses at several colleges. He graduated from the University of Michigan with a bachelor’s degree in psychology-sociology and received a master’s degree in business administration and industrial relations from Wayne State University in Detroit. He is a regular contributor to WorldatWork publications and Online Community.
REFERENCES
Aylward, Larry. 1993. “Glass Roots.” Aftermarket Business, Oct. 1, 8.
Gibbons, John, and Christopher Woock. 2007. Evidence-Based Human Resources. A Primer and Summary of Current Literature. New York: The Conference Board Inc.
Lazear, Edward P. 2000. “Performance Pay and Productivity.” American Economic Review 90(5): 1346-1361.
Lazear, Edward P. 2000a. “The Future of Personnel Economics.” Economic Journal 110(467): 611–639.
Lazear, Edward P. 1999. “Personnel Economics: Past lessons and Future Directions. “ Working Paper 6957. National Bureau of Economic Research.
Lazear, Edward P. 1996. “Performance Pay and Productivity.” Working Paper 5672. National Bureau of Economic Research.
Lazear, Edward P. 1991. “Labor Economics and the Psychology of Organizations,” Journal of Economic Perspectives 5(2): 89-110.
Passell, Paul. 1996. “Earning it: Paid by the Widget, and Proud,” New York Times, June 16, F1.
Safelite Glass Corp. 1997. Form S-4. Securities and Exchange Commission. Feb. 18. Viewed on Feb 12, 2015. https://www.sec.gov/Archives/edgar/data/1033671/0000950135-97-000756.txt
39 Third Quarter | 2015
Driving Compensation Strategy Alignment: Using Analytics to Benchmark Practices from European Normative Data
Rewards constitute a significant investment for
all organizations. Thus, it is important to stra-
tegically leverage this key investment most
effectively. This includes formalizing rewards programs
to help drive employee attraction, retention and
behavior — all of which, in turn, drive business objec-
tives. However, such strategy most often is informed
by informal or qualitative evidence (e.g., relying on
surveys) to benchmark practices.
This article offers an objective approach to investigate
differences in practice from norms and inform areas of
focus. It uses a unique application of predictive analytics
to measure strategies for employee rewards at a country
level, taking advantage of incumbent-level data collected
for pricing jobs. Statistical tests can then assess how a
company’s strategy varies from the country norm.
Key differences are shared across a set of European
countries examined, and a blinded example illustrates
the value of looking to company-specific differences.
From the latter, important insights related to compensa-
tion imperatives are emphasized, including the:
❚❚ Extent to which organizations pay for performance
Tom Jacob Mercer
Min Park Mercer
Brian Levine Ph.D. Mercer
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
40 WorldatWork Journal
❚❚ Value of general versus company-specific experience
❚❚ Degree to which pay differs by gender after account for other relevant factors.
Steps companies can take to generate related insights from archival, employee-
level data also are highlighted.
METHODOLOGY: APPLYING PREDICTIVE ANALYTICSMercer collects extensive data to help organizations benchmark pay rates and
other aspects of the rewards package. Data on thousands of companies is main-
tained in the firm’s “Total Remuneration Survey.” To date, the use of this data
primarily has been for benchmarking: to understand typical pay outcomes in an
appropriately defined market or for a given role. This helps companies track what
happens outside their walls.
Mercer also has engaged with organizations to drive workforce strategy from
hard data, using predictive analytics to analyze administrative information stored
in their Human Resource Information Systems (HRIS) and related platforms.
The intent of this work is to understand which employee experiences or contribu-
tions the organization actually rewards through pay and progression. It is also
possible to measure how employees value different aspects of the employment
experience through parallel analysis of employee engagement and retention.
For more than 20 years, Mercer has proven predictive models to be highly effec-
tive in helping organizations understand how practices play out on the ground.
This type of analysis looks at what happens inside an organization, addressing
the effectiveness of an organization’s programs relative to its unique priorities.
The structure of the survey data allows for the marriage of these two approaches,
applying predictive analytics to assess compensation philosophies at an economy-
wide level, leveraging employee-level data across thousands of companies. Further,
it can be used to drill down on differences between economy-wide practices and
those of a specific company.
In summary, the analysis can compare typical practices outside the organization
with what occurs inside it, and from that, focus on what appear to be alignments
or misalignments with desired practice. For example, when looking at an organiza-
tion focused on pay for performance, a relatively weak association between pay
and above-average performance would recommend deeper-dive examination and,
potentially, programmatic changes.
The analysis presented in this article is based on extensive incumbent-level data
for seven European countries and encompasses approximately 640,000 employees
across 1,750 organizations in 2012. (Note: More recent data are used to support
the investigation of specific issued posed by Mercer clients.) The predictive algo-
rithm is a linear regression for each country to identify economy-wide drivers
of total cash compensation (TCC) (i.e., base pay rate plus, where applicable, the
most recent bonus).
41 Third Quarter | 2015
Furthermore, the analysis tested for organizational differences from econ-
omy-wide effects related to each explanatory factor in Mercer’s model by using
“interaction terms.” Interaction terms allow Mercer to estimate, for a given company,
a separate effect related to each variable in the model. Furthermore, interaction
terms provide the opportunity to test the statistical significance of each difference
from the economy-wide norm. The method provides for greater statistical power in
identifying differences when such interaction terms are considered one at a time,
especially in the case where a given company has a small number of employees.
All models account for basic items to address a critical set of issues:
❚❚ Age and tenure. In comparing these effects, the analysis can assess, on a
relative basis, whether the organization is more focused on general experience
(proxied by age, once tenure is taken into account) or firm-specific experience.
Essentially, the analysis assesses whether the organization is more focused on
paying for outside experience or building and retaining home-grown talent.
❚❚ Gender. The analysis can identify whether the organization is at risk with
regard to pay equity.
❚❚ Pay for performance. The analysis identifies employees who have received
an above-average bonus (which proxies for having achieved an above-average
performance level) in the most recent year. This explores the relative strength
of the organization’s pay-for-performance practice.
Models also include factors not detailed in this article, including:
❚❚ The class of the job based on Mercer’s International Position Evaluation (IPE)
methodology, to consistently account for differences in career levels between jobs
❚❚ The location in which the employee works, to account for differences in
pay across markets
❚❚ Organization-specific effects, to account for differences in pay-to-market philoso-
phies that generally transcend industry affiliation.
The goal is to continually tap the Mercer data to help organizations be more
strategic in setting compensation policies.
CAVEATS: SELECTION AND CAUSATIONTwo important caveats are an issue for any use of such data. First, as is typically
the case with survey data, the sample is not random; therefore, it is not fully
representative of an economy. Organizations select whether to provide data to
Mercer and, furthermore, whether to provide data for all of their employees in
a country. Figure 1 on page 42 provides information on the size of participating
organizations and their representation across countries and industries for the
reader’s consideration.
Second, the models presented in this article cannot account for a high level of
organizational specificity. For example, similar models run at the organizational
42 WorldatWork Journal
level to address gender pay equity would account for actual pay grades (which, of
course, vary greatly across organizations), performance rating histories, job histo-
ries and other differentiating information about an employee’s role. While the
analysis presented in this article is a useful starting point to raise areas of
concern and jump-start a strategy discussion, the authors recommend that it
be followed by a more detailed analysis that relies on a company’s own, more
detailed workforce data.
ECONOMY-WIDE RESULTSThe analysis focuses on seven countries: Finland, France, Germany, Italy, Sweden,
Switzerland and the United Kingdom. First, consider the results from the United
Kingdom in order to focus on the results of one economy before placing the
results in the context of other economies.
Relative Value of General Experience vs. Company-Specific
Experience in the Economy
As an economy, the United Kingdom generally appears to modestly value both
general and company-specific experience, though there is a tilt toward valuing
general experience more so than firm-specific knowledge. Every 10 years of
additional general experience, proxied by age, is associated with a 3.9% higher
TCC. Compare this to the same amount of firm-specific experience, proxied
by tenure, which is associated with only a 2.6% increase in TCC. This is all
FIGURE 1 Organization Demographics
This figure represents the distribution of industries within countries included in the dataset used for the analysis. This article focuses on the United Kingdom’s economy as an example.
Distribution
Industry
Finland, France, Germany, Italy, Sweden,
SwitzerlandUnited Kingdom
Durable 29.6% 26.2%
Consumer Goods 23.5% 24.0%
Nondurable (excluding Consumer Goods) 13.3% 9.7%
High-Tech 8.3% 7.0%
Services 6.4% 10.5%
Retail/Wholesale Trade 3.6% 4.9%
Energy 2.4% 4.3%
Insurance 0.8% 0.5%
Finance/Banking 0.7% 1.6%
Other 11.5% 11.3%
43 Third Quarter | 2015
after controlling for other relevant pay-related factors (e.g., job size, job family,
educational attainment, geographic location).
Extent to Which Women Are Falling Behind on Pay
There is potentially a significant gap. Women appear to be paid 9.3% less than
men who are in similar jobs and locations, with similar educational attainment
and experience. There may be additional legitimate factors within each organi-
zation that can explain some of this disparity. However, the size of the gap is
concerning (relative to past research using similar methodologies at the country
level) and suggests an economy-wide issue.
Extent to Which Performance Is Rewarded
As an economy, the United Kingdom tends to pay for performance. An employee
whose bonus exceeds his/her target (a proxy of a high performer) is paid better
on TCC by 7.6%, on average, after controlling for other relevant factors. Putting
these findings for the United Kingdom in the context of other nearby economies,
there are some important points of distinction. (See Figure 2.)
Specifically, the analysis reveals that the United Kingdom:
❚❚ Values firm-specific experience more highly relative to general experience than
other countries examined
❚❚ Has a larger, unexplained gender gap
❚❚ Pays significantly more for performance than the other countries examined.
FIGURE 2 Pay Drivers in the United Kingdom
The bars show the effects of various drivers on TCC. For example, in the United Kingdom, 10 years of additional general experience (proxied by age) is associated with a 3.9% increase in TCC, after controlling for factors such as years of service, Mercer IPE “position class” (a consistent, cross-company leveling variable) and educational attainment.
N=81,068, reflecting 325 companies
Years of Age: +10
Years of Service: +10
Female vs. Male
Typical Education: High School vs. University
Bonus Exceeds Target
IPE Position Class: +1
3.9%
2.6%
-9.3%
-1.3%
7.6%
6.6%
Total Cash Compensation
-10% 0% 10%
United KingdomPercent change in pay associated
with each driver comparison
44 WorldatWork Journal
Interestingly, similar analyses looking at IT employees in India show much
greater value than any of these countries linked to general experience, and
significant negativity associated with years of service. In India’s IT sector, which
has been rapidly developing, employees who sit still are penalized.
COMPANY SPECIFIC FINDINGS: IMPLICATIONS FOR STRATEGYConsider Company X, a multinational consumer goods organization with head-
quarters in the United States and more than 100,000 employees globally. Here,
the focus is on results for Company X’s less than 5,000 UK employees and
relevant strategy questions:
1 | Focus. General experience, proxied by age, is paid 66 percent more
at Company X (6.5% vs. 3.9% for every 10 years in the UK), making
it almost three times more valuable than firm-specific experience.
Strategy question: Should Company X be more focused on paying a premium
for outside experience than its local peers?
2 | Pay equity. Women are less underpaid at Company X than they are in
the average UK economy, but there remains a significant difference that is
concerning Strategy question: Can Company X improve the value proposi-
tion to its female talent and act to minimize compliance risk by engaging in
pay equity review?
3 | Pay for performance. Pay is rewarded highly, but is on par with other
companies in the United Kingdom.
FIGURE 3 Percent Change in TCC Associated with Each Driver Comparison
Driver comparison Finland France Germany Italy Sweden SwitzerlandUnited
Kingdom
Years of Age: +10 years(general experience)
5.7% 8.1% 5.4% 9.5% 4.4% 6.8% 3.9%
Years of Service: +10 years(firm-specific experience)
N/S 1.0% 2.6% -1.3% -0.4% 0.7% 2.6%
Female vs. Male -6.1% -4.1% -6.7% -3.4% -4.1% -5.8% -9.3%
Bonus exceeds target (performance) 5.9% 6.2% 5.7% 7.3% 7.0% 6.7% 7.6%
A comparison of the drivers of pay, TCC, across seven European economies. “N/S” implies an effect is not statistically significant at conventional levels.
N=637,844, reflecting 1,753 organizations.
45 Third Quarter | 2015
Strategy question: How strongly should pay vary with current-year performance?
How should pay-for-performance sensitivity vary across key workforce segments?
USING ANALYTICS TO GAIN A DEEPER VIEWOrganizations can gain great value from using predictive analytics to assess how
their rewards practices play out and, from that, in questioning whether those
practices are optimized. Data once solely used for benchmarking pay levels
can now be used, leveraging predictive analytics, to benchmark the efficacy of
practices. Furthermore, data that sits in the HRIS and supporting systems can be
harnessed to support more thorough diagnostics. Indeed, an increasing number
of organizations are tapping into the value of such data to support fact-based
compensation strategies.
AUTHORS
Brian Levine, Ph.D. ([email protected]) is partner and co-leader of Mercer’s Workforce Strategy & Analytics practice and fellow of Mercer’s research arm, the Workforce Sciences Institute. He is based in New York.
Min Park ([email protected]) is principal in Mercer’s Workforce Strategy & Analytics practice. She is based in Los Angeles.
Tom Jacob ([email protected]) is senior partner and global leader of Mercer’s Information Solutions Research and Insights Products. He is based in Philadelphia.
FIGURE 4 Company X vs. UK Norms
Pay drivers for Company X in comparison to the United Kingdom. Percentage increases are after controlling for factors such as years of service, IPE posi-tion class (a consistent, cross-company leveling variable) and educational attainment. The asterisk (*) next to a factor indi-cates that the effect is statistically significantly different from the norm.
Years of Age: +10*
Years of Service: +10
Female vs. Male
Typical Education: High School vs. University
Bonus Exceeds Target
IPE Position Class: +1*
6.5%
Company X
-20% 0% 40%30%20%10%-10%
UK economy-wide norm
Percent change in pay associated with each driver comparison
3.9%
2.6%2.6%
-7.0%-9.3%
-3.0%-1.3%
7.6%7.6%
12.5%6.6%
46 WorldatWork Journal
This paper examines cost-of-living allowances
or adjustments (COLA) for Social Security,
private-sector pension plans and various public-
sector pension plans.
The concept of the cost-of-living adjustments origi-
nated in collective bargaining in the auto industry in
1947. It quickly spread to other industries, due in part
to the “contract bar rule.” The National Labor Relations
Board would accept a petition for decertification of
a recognized union only between the 90th and 30th
day before the expiration of a labor agreement with a
maximum duration of three years (or after its expira-
tion). This greatly reduced the chances of the union
being “raided” by another union, a common occurrence
at the time. A “three-year book” also gave the employer
three years of “no-strike-clause” protection and unin-
terrupted production. These provisions quickly became
the standard and contributed significantly to labor-
management stability.
At that time and for years after, inflation was more of
a concern than it is today. If the union entered into a
multiyear agreement, it ran the risk of the real wages
John G. Kilgour, Ph.D., California State University,
East Bay
Cost-of-Living Adjustments
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
47 Third Quarter | 2015
(purchasing power) of its members being eroded by price inflation. One solution
was an annual “wage reopener,” with the possibility of a work stoppage. This
had little appeal to management. The other solution was a COLA provision that
adjusted wages to some measure of inflation.
While COLA arrangements have largely been replaced by specified annual
increases in private-sector collective bargaining, due in large part to the taming
of inflation, the concept of indexing wages and salaries to inflation has been
adopted by Social Security and to defined benefit (DB) pension plans in
the public sector.
ECONOMICS OF RETIREMENT INCOMEWhile COLAs were important
in multiyear labor agreements
during past periods of high
inflation, they are much more
important for participants in
DB pension plans. Retirees
and beneficiaries typically
live and collect retirement
benefits for many more
than three years.
Table 1 reports life expec-
tancy at birth and at age 65
for men and women. We are
accustomed to reading about
improved life expectancy
at birth. The improvements
have been impressive. Life
expectancy of a male born
in 1980 was 69.9 years. By
2013 it was 76.5, an increase
of 6.6 years (9.4%). Over the
same period, life expectancy
at birth for females went from
77.7 years to 81.3, an increase
of 3.8 years (4.9%).
For retirement income purposes, the more relevant measure is life expectancy
at age 65. For the period 1980–2013, male life expectancy at 65 went from 14.0
years to 18.0, an increase of 4.0 years (28.6%). For females, it grew from 18.4
years to 20.5 years, an increase of 2.1 years (11.4%). A male age 65 in 2013 can
TABLE 1 Life Expectancy at Birth and at Age 65,
Selected Years 1980–2013
LIFE EXPECTANCY
AT BIRTH AT AGE 65
Year Male Female Male Female
1980 69.9 77.5 14.0 18.4
1985 71.1 78.2 14.4 18.6
1990 71.8 78.9 15.1 19.1
1995 72.5 79.1 15.4 19.1
2000 74.0 79.4 15.9 19.0
2005 74.8 80.0 16.7 19.1
2010 76.1 80.9 17.6 20.2
2011 76.1 80.9 17.7 20.2
2012 76.4 81.1 17.9 20.4
2013 76.5 81.3 18.0 20.5
Additional Years*
6.6 3.8 4.0 2.1
Percent Increase*
9.4 4.9 28.6 11.4
* Calculated by author. Source: 2014 Social Security Trustees Annual Report. Table V.A3. Retrieved from www.ssa.gov October 12, 2014.
48 WorldatWork Journal
expect to live to 83.0 and a
female to 85.5. Women live
longer than men, but the
differential is shrinking.
Increased life expectancy
is usually considered good
news. We all want to live
longer and we want our
children and grandchildren
to live longer still. For the
Social Security program and
for DB pension plans, it is
not good news. The longer
participants and beneficia-
ries live, the longer they
collect benefits. In the
absence of a COLA provi-
sion, the longer they live, the
less the pension benefit is
worth in purchasing power.
Table 2 presents a picture
of the impact of inflation
on a basic (initial) pension
benefit of $12,000 per year
($1,000 per month) over
one through 30 years of retirement. While few people live to 95 (65 + 30),
lots of people retire earlier than age 65. At the current inflation rate of about
1.5%, a $12,000 basic benefit will be reduced to $10,317 (86.0%) in 10 years
and to $8,870 (73.9%) in 20 years. At a more likely long-run average inflation
rate of 3.0% per year, a $12,000 annual pension benefit will be reduced to
$8,849 (73.7%) in 10 years and to $6,526 (54.4%) in 20 years. Remember, life
expectancy at age 65 is 18.0 years for men and 20.5 years for women. Clearly,
a COLA provision is an important feature in any retirement-income plan.
SOCIAL SECURITYTo most people, the term Social Security means the Old Age, Survivors and
Disability Insurance (OASDI) program. It is funded by equal contributions
from the employer and employee (the self-employed pay both) and provides
retirement and disability benefits. The Social Security Act of 1935 has been
amended many times and now consists of a large number of benefit programs
including the Supplemental Security Income (SSI). SSI is a federal needs-
based welfare program that provides payments to low-income individuals and
TABLE 2 The Impact of Inflation on a $12,000
Per Year Basic (Initial) Pension Benefit.
Basic Benefit
Inflation Rate
Years Since Retirement
Future Value
Percent of Basic Benefit
12,000 1.5% 1 11,820 98.5%
12,000 1.5% 5 11,127 92.5%
12,000 1.5% 10 10,317 86.0%
12,000 1.5% 15 9,566 79.7%
12,000 1.5% 20 8,870 73.9%
12,000 1.5% 25 8,224 68.5%
12,000 1.5% 30 7,625 63.5%
12,000 3.0% 1 11,640 97.0%
12,000 3.0% 5 10,305 85.9%
12,000 3.0% 10 8,849 73.7%
12,000 3.0% 15 7,599 63.3%
12,000 3.0% 20 6,526 54.4%
12,000 3.0% 25 5,604 46.7%
12,000 3.0% 30 4,812 40.1%
Source: Calculated by author.
49 Third Quarter | 2015
families with limited assets. To many, SSI payments supplement retirement or
disability benefits under OASDI. SSI is financed from the general fund of the
U.S. Treasury.
The Social Security Amendments of 1972 established automatic cost-of-living
adjustments for OASDI and SSI benefits effective 1975. Before that, Congress
increased OASDI benefits by special legislation as deemed necessary.
OASDI and SSI benefits are indexed to the Consumer Price Index for Urban
Wage Earners and Clerical Workers (CPI-W), one of the several indexes published
by the Bureau of Labor Statistics. Specifically, the benefit adjustment is deter-
mined by comparing the price levels of a “market basket” of goods and services
for the third quarter ( July, August and September) of each year with the same
period in the previous year. Initially, benefits would increase automatically only
if the CPI-W increased by 3% or more. The “3% trigger” was eliminated in 1986
(Social Security Administration 2013).
During the 1970s and early 1980s, price
inflation was high and, thus, so were the
Social Security cost-of-living adjustments.
For example, in 1975 the COLA was 8.0%
and in 1980 14.3%, an all-time high. They
then declined and, as indicated in Table 3,
have been quite low since 2000. In 2010
and 2011, there was no increase at all due
to the Great Recession.
The COLA effective for January 2009
was 5.8% based on the CPI-W for the
third quarter of 2008 compared to the
same period in 2007. However, the Great
Recession hit in the fourth quarter of
2008 and continued into the first quarter
of 2009. A recession is defined as two
consecutive quarters in which gross
domestic product (GDP) declines. In 2009,
OASDI and SSI recipients received a 5.8%
benefit increase to compensate for price
inflation that had disappeared.
In 2010 and 2011 price inflation was
negative. When that happens (this was
the only case), benefits are not reduced.
They remain unchanged until the CPI-W
surpasses its former high, as it did in 2012.
Since then, the COLA has been low: 1.7%
in 2013; 1.5% in 2014; and 1.7% in 2015.
TABLE 3 Social Security COLAs,
2000–2015.
Year COLA
2000 2.5
2001 3.5
2002 2.6
2003 1.4
2004 2.1
2005 2.7
2006 4.1
2007 3.3
2008 2.3
2009 5.8
2010 0.0
2011 0.0
2012 3.6
2014 1.5
2015 1.7
Source: Social Security Administration (Official Website). History of Automatic Cost-of-Living Adjustments.
Retrieved from www.socialsecurity.gov Feb. 2, 2015.
50 WorldatWork Journal
The automatic CPI-W COLA is also used to adjust the Social Security wage
base, the cap on the amount of wages and self-employment income subject to
OASDI payroll taxes, and to adjust the retirement earnings test for those recipients
between 62 and the full retirement age (now 66), who continue to work. The U.S.
government uses the CPI-W COLA for its own pension plans.
PRIVATE-SECTOR PENSION PLANSThe most dramatic fact about private-sector pension plans is the massive shift
from DB plans to defined contribution (DC) plans, depicted in Table 4. DC plans
almost always provide retiring or terminating participants with a lump sum
distribution. There is no need for a COLA in a DC plan. In theory, the retiree
can protect the lump sum from inflation by prudent investment.
Table 4 tells only part of the story. Many nominal DB plans are hybrid cash
balance and pension equity plans that function more like DC plans and do
not need COLAs.
Private-sector automatic COLAs have received little attention in recent years.
A dated study found that the proportion of full-time workers participating in
“medium and large” pension plans (more than 100 participants) with any kind
of post-retirement benefit
adjustment declined dramat-
ically from 54% in 1983 to
10% in 1993. Moreover, the
proportion that had an
automatic COLA during that
period remained in the 3% to
5% range. In 1992, only 4%
of workers in both “medium
and large” establishments
and in “small” private-sector
establishments (less than 100
participants) had automatic
COLAs (Weinstein 1997).
Automatic COLAs were
never widespread in private-
sector pension plans. During
periods of high inf la-
tion, such as in the 1980s,
many sponsors voluntarily
improved retirement bene-
fits on an ad hoc basis. But
overall, they shied away
from automatic COLAs. One
TABLE 4 The Private Sector Shift from Single-
Employer Defined Benefit to Defined Contribution
Pension Plans
Basic Benefit
DEFINED BENEFIT PLANS
DEFINED CONTRIBUTION
PLANS
Year PlansActive
ParticipantsPlans
Active Participants
1980 145.8 23.7 340.8 18.4
1985 167.9 23.3 461.2 32.2
1990 111.3 21.2 598.2 33.9
1995 67.7 18.9 622.6 40.4
2000 47.0 17.3 685.6 48.9
2005 46.1 15.7 630.1 59.3
2010 45.1 12.8 653.1 70.0
2011 43.8 12.3 637.1 70.3
2012 42.2 11.8 637.1 71.9
Source: U.S. Department of Labor, Employee Benefit Security Administration. Private Pension Plan Bulletin Historical Tables.
Version 2. Released December 2014. Tables E1 and E8. Retrieved from www.dol.gov/ebsa Feb. 6, 2015.
51 Third Quarter | 2015
reason for this is that they did not want to add the inflation risk to the list of
other risks inherent in sponsoring a DB pension plan: investment, interest rate,
longevity and timing risk.
There may also have been complications associated with accounting and
funding due to the requirements of the Financial Accounting Standards Board
(FASB) and the Pension Benefit Guaranty Corp. (PBGC). The inflation risk does
not lend itself as well to risk pooling, duration spreading and probability theory
as do other pension risks. An automatic COLA would at least complicate the
calculation of the PBO (projected benefit obligation), net position (liabilities –
assets), funded ratio (assets ÷ liabilities) and, hence, the employer’s ARC (annual
required contribution). Of course, the sponsor could craft an automatic COLA to
control the “COLA risk” as is routinely done in the public sector.
PUBLIC-SECTOR PENSION PLANSThe public sector includes the U.S. government, 50 state governments and 89,004
local governments. Local governments include 19,522 municipal governments,
16,364 townships, 12,884 independent school districts and 37,203 special districts.
They sponsor 2,550 public employee retirement systems: 218 administered by
states and 2,332 by local governments (U.S. Census Bureau). These numbers do
not sum. Presumably, the missing 3,031 local governments are counties.
Public-sector pension plans are more generous than those in the private sector.
In contrast to the private sector, most public-sector employees are required to
contribute to their pension plans. Public employees typically remain with one
employer for a long time. When they do change employers, it often involves
staying in the same pension plan or transferring to one with reciprocity arrange-
ments. In addition, membership in public-employee unions has remained high
and those unions are active in political
fund-raising and election activities.
They often also exert influence through
membership on boards of trustees of
the public-pension funds.
Table 5 reports union membership
and the percent of employees repre-
sented in the private and public sectors
in 2014. Note how the union member-
ship and representation is especially
high at the local government level.
Public-employee union leaders routinely
attend county and city council meet-
ings and often have knowledge and
opinions that allow them to influence
outcomes on matters of importance to
TABLE 5 Union Membership and
Employees Represented in 2014
Percent Members
Percent Represented
Private Sector
6.6 7.4
Public Sector
35.7 39.2
Federal 27.5 31.6
State 29.8 32.8
Local 41.9 45.5
Source: U.S. Bureau of Labor Statistics, News Release dated Jan. 23, 2015. Union Membership 2014. Table 3.
52 WorldatWork Journal
their members, such as pensions. Traditional defined benefit pension plans with
COLAs have fared well in the public sector.
It is much harder to generalize about pensions in the public sector. Private-sector
DB pension plans are under one body of federal law: the Employee Retirement
Income Security Act of 1974 (ERISA) as amended. Public-sector pension plans are
largely governed by the legislation of the 50 states, many of which have multiple
laws that apply to different categories of employers or employees. Arrangements
vary significantly among jurisdictions. In addition, many state plans provide
pension services to county, city and other local governments. Some county plans
provide pension service to some of the local governments within the county.
FEDERAL GOVERNMENTThe United States government maintains a number of pension plans including
the Civil Service Retirement System (CSRS) and the Federal Employees Retirement
System (FERS). The Social Security Amendments of 1983 required that federal
employees hired after Jan. 1, 1984, be covered by Social Security and the newly
established FERS. Existing employees were allowed to remain in the CSRS or shift
to the FERS with Social Security. Given that more 30 years have elapsed, the vast
majority of federal employees are now in FERS.
Under CSRS and FERS, the COLA takes effect on Dec. 1 of each year and appears
on the January pension check. As is Social Security, it is based on the CPI-W
for the third quarter of the current year compared with the same period in the
previous year. Adjustments are not provided until age 62, other than for disability
and survivor benefits and for other special-provision retirees (public safety).
If the increase in the cost of living is 2% or less, the COLA is equal to the CPI-W
increase. If it is between 2% and 3%, the COLA is 2%. If it is more than 3%, it is
1% less than the CPI-W increase.
For 2013, 2014 and 2015, the COLA for both CSRS and FERS was the same at
1.7%, 1.5% and 1.7%. Before that, they differed with the CSRS rate being higher
than that of the FERS rate (federal retirement).
STATE AND LOCAL GOVERNMENTSThe 50 states administer 218 retirement systems. The 89,004 local governments
either operate their own pension plan or contract with the state or county for
pension coverage, thus creating a complex mosaic that makes generalization
difficult. In regard to COLAs, it is even more complex. In at least one state
(California), local governments contracting with a state plan may elect to pay for
more generous COLA provisions.
The Wisconsin Legislative Council has tracked approximately 85 major public
employee pension systems in even-numbered years since 1982. Table 6 on page
53 exhibits data on the types of COLA provisions in those plans for 1982, 1992,
2002 and 2012 (the most recent year for which data are available). Note how the
53 Third Quarter | 2015
two automatic COLA categories (Adjustments Limited to CPI and Automatic Percent
Increase) grew significantly from 1982 to 2002 and then contracted in 2012. The
contraction reflects the extensive amount of public-pension reform in recent years.
Rather than examine aggregates, the author will look at the COLA provi-
sions of the three large statewide public-pension plans in California with which
he is familiar.
CalPERSThe California Public Employees’ Retirement System (CalPERS) has 1.7 million
members, 587,959 of whom are annuitants (retirees and beneficiaries). CalPERS
has three categories of employers and members: “state” including the massive
California State University (CSU) system; “school” classified employees (secretaries,
custodians, etc.); and “public agencies” (local governments). As of February 2015,
there were 1,513 school districts with 442,088 members and 1,580 contracting
public agencies with 345,279 members under contract with CalPERS.
The COLA provision for state and CSU members (including faculty) is limited
to the actual rate of inflation up to a 2% maximum. There is a commonly held
misconception that the 2% maximum is simply applied to the gross pension each
year. It is a bit more complicated.
There is no COLA in the first year of retirement. It starts in the second year and
is then calculated annually and applied to the May 1 pension benefit thereafter.
CalPERS calculates the compounded 2% maximum each year as follows: (1.02 X
1.02) -1 = .04; (1.02 X 1.04) -1 = .061; (1.02 X 1.061) -1 = .082, etc. (rounded). The
result is then compared to the actual rate of inflation as measured by the Annual
Average of the All Urban Consumer Price Index (CPI-U) based on 1967 (1967 =
100). Contracting public agencies may elect to pay for a 3%, 4% or 5% maximum
COLA calculated in the same way.
TABLE 6 Types of COLA Provisions, Selected Years 1982–2012
1982 1992 2002 2012
Number Percent Number Percent Number Percent Number Percent
Adjustments Limited to CPI
34 42.5 43 50.6 39 45.9 30 34.5
Automatic Percentage Increase
10 12.5 13 15.3 22 25.9 24 27.6
Investment Surplus 3 3.8 4 4.7 3 3.5 5 5.7
Ad Hoc or No Provision 33 41.3 25 29.4 21 24.7 28 32.2
Total Plans in Survey 80 85 85 87
Source: Wisconsin Legislative Council. 1982, 1992, 2002 and 2012 Comparative Study of Major Public Employee Retirement Systems.
54 WorldatWork Journal
CalPERS provides another level of inflation protection called the Purchasing
Power Protection Allowance (PPPA). For state and school members, the PPPA
ensures that the purchasing power of the pension benefit will not fall below 75%
of its initial value. That would be when the hypothetical $12,000 per year pension
benefit used above would have been reduced to $9,000. At an inflation rate of
1.5% per year, the PPPA would kick in about 19 years. At a 3.0% rate of inflation,
it would take only about 9.5 years (author’s calculations).
For contracting public agencies, the guarantee is generally 80%. For a $12,000
annual pension, the trigger would be $9,600. At a 1.5% inflation rate, the PPPA
would take effect in 14.8 years and at 3.0% in 7.3 years. However, those public
agencies that have elected a more generous COLA percentage, it would kick in
later or perhaps never (CalPERS).
CalSTRSThe California State Teachers’ Retirement System (CalSTRS) covers certificated
teachers and community-college instructors (K–14). In 2014, it had a 603,702 active
and inactive members (420,887 active and 182,815 inactive) and 275,667 retirees
and beneficiaries.
The CalSTRS COLA provision is 2% of the basic allowance per year. Adjustments
are not tied to changes in the CPI and are reportedly not compounded. However,
if 2% is added to the base each year, there is some compounding ($12,000 X 1.02
= $12, 240; X 1.02 = $12,485, etc.).
CalSTRS also has a purchasing power protection arrangement that is currently
set at 85% of the initial benefit. This generous inflation protection is appropriate
because teachers in California (and 14 other states) are not covered by Social
Security with its 100% COLA (CalSTRS). Even if a teacher has a spouse who does
participate in OASDI and/or if he/she is entitled to a retirement benefit in his/her
own name, that benefit is greatly reduced by Social Security’s Windfall Elimination
Provision and the Government Pension Offset (Kilgour 2009).
UCRP The University of California Retirement Plan (UCRP) covers faculty and staff
employees at the 10 UC campuses, the Hastings College of Law and other UC
affiliates. In 2014, the UCRP had 218,921 total members: 120,407 active partici-
pants, 21,892 retirees and beneficiaries, and 35,027 separated (inactive) vested
participants (UCRP 2014).
The UCRP COLA provision is 100% of the CPI increase up to 2% and 75% of
any increase over 4% to a maximum of 6%. The CPI used is the average for the
Los Angeles and San Francisco metropolitan areas (University of California 2015).
The UCRP does not seem to have an automatic inflation protection feature, as
do CalPERS and CalSTRS. However, with its more generous COLA provision (up
to 6%), it may not need one.
55 Third Quarter | 2015
CONCLUSION In the absence of an automatic COLA (or timely ad hoc benefit increases), the
real value of a pension benefit depreciates over time due to price inflation. If
the retiree or beneficiary is blessed with great longevity, the benefit will be a
fraction of what it once was. In the absence of family or other support, the result
will be a lot of elder poverty. Fortunately, the OASDI and SSI are fully inflation
protected. Most public employees are largely protected, including the 30% of the
public-sector employees not covered by Social Security (mainly police, fire and,
in 15 states, teachers). The private sector is another matter.
COLAs are virtually nonexistent in private-sector pension plans. Single-employer
and multiemployer pension benefits, plus those paid by the Pension Benefit
Guaranty Corp. to participants of trusteed plans, atrophy with time. This further
dramatizes the retirement income disparity between the private and public sectors
and will add to the pressure to curtail public-employee pensions.
AUTHOR
John G. Kilgour, Ph.D. ( [email protected]) is a professor emeritus in the Department of Management at California State University, East Bay. He holds an MILR and Ph.D. in Industrial and Labor Relations from Cornell University. Kilgour has published numerous articles on compensation and benefits topics. He is a long-time member of WorldatWork and a frequent contributor to the WorldatWork Journal.
REFERENCES
CalPERS. California Public Employees’ Retirement System website. Viewed: Feb. 8, 2015. www.calpers.ca.gov.
CalSTRS. California Teachers’ Retirement System website. Viewed: Feb. 8, 2015. www.calstrs.com.
Federal Retirement. “Cost of Living Adjustments (COLA).” Viewed: Feb. 8, 2015. www.federalretirement.net/cola.htm.
Kilgour, John G. 2009. “Social Security in the Public Sector: The Windfall Elimination Provision and the Government Pension Offset.” Compensation and Benefits Review, September/October, 34–42.
Social Security Administration. 2013. “Cost of Living Adjustment.” Viewed: Feb. 2, 2105. www.socialsecurity.gov.
University of California 2015. “A Complete Guide to Your University of California Retirement Benefits.” p. 15. Viewed: Feb. 8, 2015. http://ucnet.universityofcalifornia.edu.
UCRP. University of California Retirement Plan (UCRP). 2014. “Summary of Plan Data for 2014.” Viewed: Feb. 9, 2015. http://ucnet.universityofcalifornia.edu.
U.S. Census Bureau. Viewed: Feb. 7, 2015. http://factfinders.census.gov.
Weinstein, Harriet. 1997. “Post-retirement Pension Increases.” Compensation and Working Conditions. Viewed: Feb. 4, 2015. www.bls.gov. (This study was based on “establishment data.” It excludes agriculture and the federal government and includes only full-time employees, not retirees.)
Published Research in Total RewardsA review of total rewards, compensation, benefits and HR-management research reports.
(Compiled by the editors from the WorldatWork Newsline column at www.worldatwork.org.)
56 WorldatWork Journal
Organizational Culture, Employer Brand Are Top Competitive Advantages in RecruitingCompanies that focus on offering higher salaries or promoting quickly alone may not be as competitive when recruiting talent, according to the results of an executive survey from Futurestep, a Korn Ferry company. 61% of respondents said organizational culture is the most important recruiting advantage for global organizations, followed by a leading employer brand at 26%.
“Focusing on culture and how that brand is represented in the marketplace has a critical impact on attracting and retaining the talent that will drive business success,” said Neil Griffiths, Futurestep global practice leader, talent communications and employer brand. “The survey results indicate that employers need to think more broadly about what attracts top talent to their organization.”
The survey also found that as the hunt for talent gets more difficult (61% of respondents said it’s harder to find qualified candidates than it was a year ago), there is a strong need for employers to closely evaluate and understand what attracts and motivates the ideal candidates for their company. While salary continues to be the “top negotiation sticking point” at 51%, “flexibility” comes in second at 33%, followed by “title” at 11% and “vacation” at 4%.
“The challenge is for organizations to listen to what employees want from their workplace, such as flexibility and, when possible, find a practical and effective way of delivering,” Griffiths said.
“In today’s digital, social and mobile world, it’s easier than ever to enable employees to work when, where and how they want to, as long as they remain productive.” There were more than 1,000 responses to the global survey, which was conducted in May.
Total Rewards and Employee Well-Being PracticesThis WorldatWork survey was conducted to identify traditional wellness plans and new trends in employee well-being. The objective was to gauge how many programs and initiatives organiza-tions offer. The survey also focused on how those offerings are expanding to include a more integrated well-being approach beyond one that is related solely to physical health.
Employers continue to depend on health and wellness initiatives to curb health-care costs and foster a successful and productive workforce. Ninety-six percent of organizations support well-being components and nearly three-quarters of organizations are increasing or considerably increasing their well-being offerings in the next two years. There were more than 400 responses to the survey, which was conducted in January.
Third Quarter 2015
© 2015 WorldatWork. All Rights Reserved. For information about reprints/re-use, email [email protected] | www.worldatwork.org | 877-951-9191
57 Third Quarter | 2015
Promotional GuidelinesWhile many aspects of promotional practices are largely unchanged from WorldatWork’s past two surveys on this topic in 2012 and 2010, the percentage of employees receiving a promotion has increased by two full points since 2010.
Respondents report that an average of 9% of their employees received a promotion during the past fiscal year.
This is a significant improvement from the 7% reported in 2010 and is further evidence that organizations are relaxing the purse strings in their pay budgets.
There has been little change however in the average amount of increase that employees are receiving. In determining promotional increase amounts, organizations continue to consider salary range position and the pay levels of similarly situated employees as primary determinants. Less than half (42%) budget separately for promotional activity. The prevalence of this practice appears to be falling as more organizations are either budgeting for promotions in their merit budget or simply funding with salary or vacancy savings. There were nearly 500 responses to the survey, which was conducted in December.
Trends in Employee RecognitionRecognition programs remain a mainstay in organizations, with a WorldatWork Survey finding 89% of organizations having programs in place. Sixty seven percent of organizations offer between three and six different programs, with 4.6 as the average number of programs. The average number of programs is back to the 2011 level after dipping to 3.9 in 2013. The prevalence of programs has experienced an interesting shift in the last few iterations of this survey, with a growing number of organizations relying on recognition programs that drive organizational results. However, the staple recognition program, length of service, still holds the top spot, with 87% of organizations offering this type of program. It has been at the top since the inception of this survey.
It is also interesting to note that in both 2011 and 2013 surveys, when asked to write in other programs that are offered in organizations, participants responded with “wellness programs.” With the increase in wellness programs and the continuing increase in programs to motivate specific behavior, WorldatWork speculated whether wellness programs and programs to moti-vate specific behavior were intermingled. Therefore, in this survey they were listed as separate programs. The results for “Programs that improve biometric indices through wellness initiatives” debuts as sixth on the list, with 32% of organizations offering these programs and the highest average number of employees recognized in the past 12 months (68%). There were nearly 500 responses to the survey, which was conducted in March.
Published Research in Total RewardsA review of total rewards, compensation, benefits and HR-management research reports.
(Compiled by the editors from the WorldatWork Newsline column at www.worldatwork.org.)