Fifteen years ago this week, Congress passed the Government
and Results Act, a law whose sponsors hoped would help establish
a government guided by goals and monitored by results.
Senator William Roth of Delaware stood on the Senate floor and
For much too long, this focus on program performance and results
missing from the federal government. Agencies and managers are
expected to follow
proper procedures and spend their funds in an appropriate mannerto
all the ts and dot all the isbut rarely are their programs held
achieving measurable results toward any preestablished goals. Is it
then, that program performance suffers, and that public frustration
Fifteen years later, Roths successors can simply cut and paste his
into their own floor statements. Measuring government performance
has proven a lot harder that it sounds. Yet results really do
matteras the response
to Hurricane Katrina showed so clearly.
To become a high-performing enterprise, government needs to do the
Transform the culture of its organizations to work closely with
governments, nonprofits, and the private sector, domestically and
to achieve results.
Reassess missions and strategiesand the entire mix of policy tools
available to address its objectivespromoting practices that work
ending those that dont.
Have its leadership increase strategic planning,
address management challenges, use integrated approaches,
and enhance agencies results orientation
to give the public the demonstrable results
The July 2008 conference explored three broad
topics as part of the performance challenge track:
Demand for results drives new systems and behaviors.
Over the past decade or so, government at all
levels, in the executive and legislative branches,
has begun requiring short- and long-term plans,
including strategic goals, measurable objectives,
a system for assessing outcomes, and periodic reporting
on results. More recently, decision makers
have attempted to tie budget and other resource
decisions to agency performance. How are organizations
responding to this challenge?
Comparing performance with common standards. Casting
the net more broadly, how have governments begun
to make use of benchmarking, leading indicators,
and other learning organization methods to improve
and evaluate performance in a comparative context?
Fostering an organization-wide performance culture.
What new insights can be gleaned from federal,
state, and local efforts over the past decade to
evolve or transform the basic assumptions and
behaviors that lead to improved organization performance?
How are agencies setting standards,
holding organizations accountable, and considering
changes to human resources law, personnel
policies and systems, and other innovative ideas
(such as employment at will) in pursuit of a
Some speakers who contributed to this track offer
commentary on how we are addressing this challenge
and a glimpse at solutions, proven and experimental,
across the governmental landscape:
With respect to the emphasis on results, Stephanie
Shipman discusses both the challenges and progress at
the federal level, noting that interagency work groups
can move things along to achieve common goals.
Similarly, Jackie Werth shares some lessons learned
from measuring performance at San Diego Countys
Health and Human Services Agency.
In the area of comparing performance with common
standards, Jerry Ellig adds some lessons on
benchmarking from the Mercatus Centers performance
When it comes to fostering an organization-wide
performance culture, both Frank Bauer and Steve
Benowitz have some things to say about pay-forperformance
systems, and Benowitz adds a few
thoughts on building high-level support and bipartisan
trust for strategic human capital planning.
Demand for Results Drives New
Systems and Behaviors
by Stephanie Shipman
Federal agencies have made great progress in reporting
on results, but agencies still face performance
challenges reflecting the public sectors complex operating
Performance measurement challenges have been
both conceptual and practical. Many public-sector programs
have only limited control over their outcomes
because they (1) rely on actions by other parties (both
public and private), or (2) have ambitious goals influenced
by forces outside the programthe economy, climate,
or demographic trends. Reliance on third parties
tends to limit access to data on client-level effects or
data that can be readily combined across jurisdictions.
Managers are often torn between choosing ambitious
goals focused on health and safety benefits or using
lesser indicators of organizational change or intermediate
benefits for which they can feel responsible. Moreover,
while population statistics or satisfaction surveys
are often convenient and appear highly relevant, they
often do not measure program performance so much as
a mixture of program effects plus the effects of outside
Federal agencies have drawn on a variety of strategies
to address these challenges. To provide a balanced picture
of program performance, managers have selected portfolios
of measures to depict both concrete process improvements
as well as progress on benefits for program clients
and citizens. To improve access to client-level data, federal
agencies have collaborated with state and local partners
to select mutually useful performance measures and provided
assistance with reporting systems. To isolate program
contributions to population benefits, agencies have
invested in program evaluation. Program evaluations use
social science research methods to measure and compare
program and contextual influences to draw conclusions
about program contributions to observed results. Still,
measurement in some areassuch as basic research and
community developmentresists easy solutions. Interagency
work groups can speed progress in finding sound,
practical measurement solutions and encourage the leveraging
of programs to achieve common goals.
Stephanie Shipman, PhD, is assistant director of the Center for
Methods and Issues at the U.S. Government Accountability Office
(GAO), where she studies federal agency performance measurement and
evaluation activities and serves as coordinator for the federal
network. She can be reached at ShipmanS@gao.gov.
in San Diego
by Jackie Werth
San Diego Countys Health and Human Services
Agency (HHSA) is a complex organization with more
than three hundred discrete programs. Its organized
into six geographic service delivery regions, six operating
divisions, and several support divisions and provides
a full range of health, behavioral health, and social services.
Consequently, HHSA had no choice but to manage
strategically. Here are some lessons from its performance
Keep your approach simple and limit measures to a
critical few. HHSA settled on a one-page strategy
agenda with only seventeen result indicators. The
strategy agenda helps it communicate internally
and externally, in a very streamlined fashion, about
how it serves the community.
Do not be a slave to methods. Too often, organizations
overemphasize the balanced scorecard, logic
modeling, and other methods that are complex to
the average person. Adapting from these methods
while keeping it as simple and intuitive as
possibleto make incremental progress is best.
HHSA started by simply having the executives
identify the fifty most meaningful executive
performance measures from more than seven
hundred individual program measures. Only later
did it work to frame a strategy. Now, HHSA can
be much more sophisticated. Regions and programs
design shared strategies that align with larger
agency-wide goals and develop shared or aligned
performance measures for tracking progress.
Articulating solid measures is not enough. You also
need to connect your strategy to operations and
results. HHSA was fortunate to be able to build
upon the strengths of a General Management
System, a county-wide business model that aligns
strategy with budgeting, monitoring, rewarding
of employees, and leveraging efforts across departments.
Objectives and measures cascade through
all management systems.
Keep your progress reporting simple like your strategy.
This is why HHSA developed its own one-page
performance flash reports that offer a quick update
on key measuresperformance improvement
and risksincluding performance differences by
region and program. That way, the agency can
alert everyone, triggering conversations on what is
working and why and the areas where it needs to
mobilize to turn performance around.
Jackie Werth is a health services project coordinator for HHSA,
she facilitates a wide range of performance improvement projects.
she was a senior evaluator with GAO in San Francisco and
DC. She can be reached at Jackie.Werth@sdcounty.ca.gov.
Benchmarking Lessons from the
Mercatus Centers Performance
by Jerry Ellig
The Mercatus Center at George Mason University
produces an annual Performance Report Scorecard that
evaluates the quality of the annual performance reports
produced by the twenty-four federal agencies covered
by the Chief Financial Officers Act. The scorecard is
primarily a means of benchmarking best reporting practices.
As the quality of reports has improved, however,
we have also gleaned insight into helpful management
practices showcased in the reports.
Four of the most important are (1) know and measure
outcomes, (2) understand causality, (3) link results
with costs, and (4) acknowledge shortcomings and have
a plan to improve.
Know and Measure Outcomes
As George Harrison and others have said, If you
dont know where youre going, any road will take you
there. Agencies should know the outcomes their activities
are supposed to accomplish for citizens. They should
measure those outcomes so they know whether they are
Its not enough to have goals; agencies should also
know how their actions are supposed to affect outcomes
and assess how much of the change in outcome was
caused by the agencys actions.
Link Results with Costs
Linking results with costs, so decision makers know
how much it costs to produce a successful outcome, can
aid both agency managers and external budgeters.
Acknowledge Shortcomings and Have a Plan
The best performance reports candidly acknowledge
performance shortfalls and major management
challenges. Instead of trying to put the best spin on every
piece of information, they present plans and timelines
for resolving major problems.
All four of these practices make reports more transparent
and informative. More important, they are things
good managers ought to be doing anyway.
Jerry Ellig is a senior research fellow at the Mercatus Center at
Mason University and a coauthor of the Mercatus Performance Report
Fostering an Organization-Wide
by Frank Bauer
Pay increases in the public sector traditionally have
been based on length of service rather than performance
once a certain threshold of satisfactory performance has
been met. The federal General Schedule pay system is one
example of a system based largely on length of service.
Pay for Performance
Although most new employees (and most employees,
regardless of tenure) want to do a good job, they
quickly become accustomed to pay increases based on
length of service. Such pay increases become part of the
organizations culture: the values and norms shared by
people and groups in the organization. Recently, public
employers have attempted to change the culture by superimposing
a pay-for-performance system on the existing
culture. Employees and their union representatives
may resist this approach because it is inconsistent with
the values of their group.
New pay-for-performance systems that remain in
effect over time may, in fact, change the culture. However,
because public employees have access to elected
officials and the courts, their opposition has sometimes
resulted in a new system being abandoned (such as
MaxHR in the U.S. Department of Homeland Security)
or reduced in scope (such as the National Security
Personnel System in the Department of Defense) before
this cultural change can occur. An alternative approach
would be to develop or strengthen the performance
culture before implementing the new pay system.
Changing the Culture
Agencies could begin with an assessment of the
organizations readiness for a pay-for-performance
system. If necessary, management practices (other than
those dealing with pay) that communicate to employees
that performance is a critical factor in the organization
would then be introduced or enhanced. Research shows
that, in the private sector, high ratings on eighteen frequently
discussed management practices correlate with
successful financial performance. Such best practices
can be summarized into three categories: performance
monitoring, performance targets,
and human capital management.
Further research indicates that
noncorporate organizations (colleges,
universities, and state, city,
and county governments) score
lower than corporations on the
application of these management
Public agencies might consider
rigorous employment of
these best practices as a means of strengthening the performance
culture over time. A pay-for-performance system
would be introduced only after the organizations
culture recognized the critical importance of performance.
Readiness for the new system could be determined
through periodic assessment. This more customized
approach to strengthening the performance culture
would admittedly take longer to implement. Arguably, it
would also be more readily accepted by employees and
have lasting positive impact.
Building a Performance Culture
through Improved Human Resource
by Steve C. Benowitz
Executives and managers in public service are faced
with the task of building work cultures that support high
performance. The federal government and many state
and local governments have developed some form of
accountability systems to measure the outcomes of their
agencies. The public is demanding greater accountability
for government programs. Much is expected from public
agencies, and Americans would
rather do away with under-performing
programs. We need a
symbiotic relationship between
agency executives and managers
and the rank-and-file employees
upon whom we all depend for
Think of a performance
culture as the commitment of
the entire organization to work
together to meet organizational goals and commitments
on behalf of its many stakeholders, most important, the
public that the organization serves. With this in mind,
then, a performance culture can succeed only if the
relationships between employees at all levels are open,
transparent, and supportive. While executive leadership
is critical, employees at all levels have a substantive impact
on the ability of the organization to succeed, so
leadership is needed at all levels.
Recent studies of the federal workforce (such as
federal human capital surveys) indicate a high level of
employee engagement in their work. For example, 90
percent believe their work is important, 83 percent believe
they know how their work relates to agency goals
and priorities, and a similar number believe employees
cooperate to get work done. However, employees have
a much lower level of confidence in the ability of supervisors
and managers to ensure high performance. For
example, 78 percent see no link between performance
and pay, 61 percent say creativity and innovation are not
rewarded, and 71 percent say supervisors do not deal
with poor performers. How do we go about addressing
these challenges to create an effective, performancebased
Pay for Performance
Many believe the answer lies in pay for performance.
Often, comparisons are made with the private
sector, where pay-for-performance systems have been
used for many years. However, with the lack of trust
that employees have in their supervisors, no pay-forperformance
system by itself can provide the answer.
First, we have to model performance accountability systems
(measuring individual and group performance) in
a way that can be supported by managers and line employees.
Second, we have to fund these systems to ensure
that good performers are recognized with appropriate
levels of compensation. Finally, we have to ensure that
managers have the tools and support to address poor
At least two strategies should be employed. The first
is to ensure transparent communication from the top
levels of the organization and the willingness to listen to
employees as a way to engage them in the process. This
approach, what may be called inclusive leadership, will
require substantial changes in organizational culture in
many agencies. The second is to build trust between
managers and rank-and-file employees, which often will
take improved communication skills in both groups. This
will take a financial investment that is likely to become
more difficult to make under todays budget constraints.
Without these changes, improving organizational performance
or moving to accountability systems, such
as pay for performance, which many see as essential to
programmatic changewill be very difficult.
These strategies will help increase trust at an organizational
level, but the president must also identify
with and own this issue. Since enactment more than
thirty years ago of the Civil Service Reform Act, which
formalized the current structure of labor-management
relations in federal government, the two major parties
have not met at a high enough level to address current
issues. A summit of the parties, which also should
include appropriate members of Congress, can help in
identifying longer-term, systemic changes that can be
made to improve the trust and collaboration between
employees and executives/managers. Similar efforts at
state and local levels could help resolve their pressing
needs, especially the economic positions in which many
states now find themselves.