Coast Guard: Non-Homeland Security Performance Measures Are	 
Generally Sound, but Opportunities for Improvement Exist	 
(16-AUG-06, GAO-06-816).					 
                                                                 
Using performance measures, the Coast Guard explains how well its
programs are performing. To do so, it reports one "primary"	 
measure for each program (such as percent of mariners rescued)	 
and maintains data on other, "secondary" measures (such as	 
percent of property saved). Concerns have been raised about	 
whether measures for non-homeland security programs accurately	 
reflect performance, that is, they did not rise or fall as	 
resources were added or reduced. For the six non-homeland	 
security programs, GAO used established criteria to assess the	 
soundness of the primary measures--that is, whether measures	 
cover key activities; are clearly stated; and are objective,	 
measurable, and quantifiable--and the reliability of data used to
calculate them. GAO also used these criteria to assess the	 
soundness of 23 selected secondary measures. Finally, through	 
interviews and report review, GAO assessed challenges in using	 
measures to link resources to results.				 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-816 					        
    ACCNO:   A58739						        
  TITLE:     Coast Guard: Non-Homeland Security Performance Measures  
Are Generally Sound, but Opportunities for Improvement Exist	 
     DATE:   08/16/2006 
  SUBJECT:   Evaluation criteria				 
	     Evaluation methods 				 
	     Internal controls					 
	     Performance measures				 
	     Program evaluation 				 
	     Agency missions					 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-816

     

     * Results in Brief
     * Background
     * Non-Homeland Security Primary Performance Measures Are Gener
          * Although the Six Primary Measures Are Generally Sound, Three
          * Existing Procedures Help Ensure Reliable Internal Data, but
     * More than a Third of the Secondary Performance Measures Asse
     * Challenges Exist in Using Measures to Link Resources to Resu
          * Primary Performance Measures Cover a Key Activity, but Not E
          * Performance Results Can Be Affected by Factors Other than Re
          * Coast Guard Has Developed a Range of Initiatives to Forge Be
     * Conclusions
     * Recommendations for Executive Action
     * Agency Comments
     * GAO Contact
     * Acknowledgments
     * GAO's Mission
     * Obtaining Copies of GAO Reports and Testimony
          * Order by Mail or Phone
     * To Report Fraud, Waste, and Abuse in Federal Programs
     * Congressional Relations
     * Public Affairs

Report to the Subcommittee on Fisheries and Coast Guard, Committee on
Commerce, Science, and Transportation, U.S. Senate

United States Government Accountability Office

GAO

August 2006

COAST GUARD

Non-Homeland Security Performance Measures Are Generally Sound, but
Opportunities for Improvement Exist

GAO-06-816

Contents

Letter 1

Results in Brief 5
Background 8
Non-Homeland Security Primary Performance Measures Are Generally Sound and
Data Are Generally Reliable, but Weaknesses Exist 13
More than a Third of the Secondary Performance Measures Assessed Are
Generally Sound, and the Remainder Have Weaknesses 19
Challenges Exist in Using Measures to Link Resources to Results, but the
Coast Guard Is Working on Ways to Address Them 21
Conclusions 28
Recommendations for Executive Action 30
Agency Comments 31
Appendix I Objectives, Scope, and Methodology 32
Appendix II Secondary Performance Measures 35
Appendix III Ongoing Coast Guard Initiatives to Link Resources Used to
Results Achieved 38
Appendix IV Comments from the Department of Homeland Security 40
Appendix V GAO Contact and Staff Acknowledgments 41
Related GAO Products 42

Tables

Table 1: Coast Guard's Non-Homeland Security Programs 2
Table 2: Soundness of Primary Measures and Reliability of Data Used to
Calculate the Primary Measures for the Coast Guard's Non-Homeland Security
Programs 13
Table 3: Source of Data Used to Calculate Non-Homeland Security Primary
Performance Measures 17
Table 4: Soundness of Selected Non-Homeland Security Secondary Performance
Measures 19
Table 5: Soundness of Secondary Measures for Coast Guard's Non-Homeland
Security Programs 36
Table 6: Coast Guard Non-Homeland Security Secondary Performance Measures
Not Assessed 37
Table 7: Ongoing Coast Guard Initiatives to Link Resources Used to Results
Achieved 38

Abbreviations

Corps United States Army Corps of Engineers BARD Boating Accident
Reporting Database DHS Department of Homeland Security GPRA Government
Performance and Results Act of 1993 OMB Office of Management and Budget
PART Program Assessment Rating Tool

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office

Washington, DC 20548

August 16, 2006 August 16, 2006

The Honorable Olympia J. Snowe Chair The Honorable Maria Cantwell Ranking
Minority Member Subcommittee on Fisheries and Coast Guard Committee on
Commerce, Science, and Transportation United States Senate The Honorable
Olympia J. Snowe Chair The Honorable Maria Cantwell Ranking Minority
Member Subcommittee on Fisheries and Coast Guard Committee on Commerce,
Science, and Transportation United States Senate

Recent years have seen a marked shift in the Coast Guard's
responsibilities. The events of September 11, 2001, shifted the Coast
Guard's priorities and focus toward homeland security responsibilities,
such as protecting the nation's network of ports and waterways. At the
same time, however, the agency's traditional non-homeland security
programs, such as rescuing people at sea and directing oil spill cleanup
efforts, remain an integral part of its operations. In all, the Coast
Guard has six non-homeland security programs (see table 1), and
collectively, the effort that goes into them constitutes 50 percent of the
Coast Guard's fiscal year 2006 enacted budget.1 Recent years have seen a
marked shift in the Coast Guard's responsibilities. The events of
September 11, 2001, shifted the Coast Guard's priorities and focus toward
homeland security responsibilities, such as protecting the nation's
network of ports and waterways. At the same time, however, the agency's
traditional non-homeland security programs, such as rescuing people at sea
and directing oil spill cleanup efforts, remain an integral part of its
operations. In all, the Coast Guard has six non-homeland security programs
(see table 1), and collectively, the effort that goes into them
constitutes 50 percent of the Coast Guard's fiscal year 2006 enacted
budget.1

1The Coast Guard's six non-homeland security programs account for about
$4.2 billion of the Coast Guard's $8.4 billion fiscal year 2006 enacted
budget. The remaining $4.2 billion is for its five homeland security
programs-ports, waterways, and coastal security; illegal drug
interdiction; defense readiness; undocumented migrant interdiction; and
other law enforcement activities, including U.S. Exclusive Economic Zone
enforcement.

Table 1: Coast Guard's Non-Homeland Security Programs

Source: GAO analysis of Coast Guard documents.

aStarting with the fiscal year 2007 budget, OMB has designated the Coast
Guard's drug interdiction and other law enforcement programs as
non-homeland security missions for budgetary purposes. However, at the
time of our review, Coast Guard officials told us that, in terms of
measuring performance, the agency still categorized these programs as
homeland security missions as delineated under section 888 of the Homeland
Security Act of 2002, 6 U.S.C. S: 468.

bThe U.S. Exclusive Economic Zone is defined as an area within 200 miles
of U.S shores in which U.S. citizens have primary harvesting rights to
fish stocks.

Since the changes that increased the Coast Guard's homeland security
responsibilities, Congress has paid renewed attention to the Coast Guard's
ability to carry out its non-homeland security programs. To help gauge its
performance in these areas, the Coast Guard collects data on 45
performance measures, such as the percentage of mariners successfully
rescued from imminent danger and the number of oil spills and chemical
discharges. When reporting its performance, the Coast Guard follows the
instructions of its parent agency, the Department of Homeland Security
(DHS), and reports one measure for each program. For example, for the ice
operations program, the Coast Guard reports on the annual number of days
certain waterways are closed because of ice, and for the aids to
navigation program, the Coast Guard reports on the number of collisions,
allisions, and groundings.2 These performance measures, which we call
"primary measures" in this report, are intended to communicate Coast Guard
performance and provide information for the budgeting process to Congress,
other policymakers, and taxpayers. Beyond the six primary performance
measures, the Coast Guard also uses a variety of other performance
measures to manage its programs, called "secondary measures" in this
report. There are three key publications that DHS and the Coast Guard use
to report the Coast Guard's non-homeland security primary performance
measures-the DHS Performance and Accountability Report, the DHS fiscal
year budget request, and the Coast Guard's fiscal year Budget-in-Brief.

2The Coast Guard defines an "allision" as a collision between a vessel and
a fixed object.

Our recent analyses have raised concerns about whether the primary
measures accurately reflect what the Coast Guard is accomplishing with the
resources it expends. In April 2004, we testified that despite substantial
changes in the distribution of resources among programs, performance
results appeared largely unaffected, and the Coast Guard had limited data
and no systematic approach to explain the lack of a clear relationship
between resources expended and performance results achieved.3 You asked us
to consider whether shortcomings in the primary measures might explain why
there was no apparent connection between resources expended and results
achieved for the non-homeland security programs. In response, we evaluated
the primary measures for the Coast Guard's six non-homeland security
programs with regard to two key characteristics: (1) their soundness-that
is, whether the measures cover the key activities of the program, are
clearly stated and described, and are objective, measurable, and
quantifiable-including having annual targets-and (2) the reliability of
the data used to calculate the measures-that is, whether controls are in
place to ensure the timeliness, completeness, accuracy, and consistency of
the data.4 You also asked us to provide information on some of the
secondary measures that are used in the Coast Guard's six non-homeland
security programs. Our report addresses three questions:

           o  Are the primary performance measures for the Coast Guard's six
           non-homeland security programs sound, and are the data used to
           calculate them reliable?

           o  Are selected secondary performance measures for four of the
           Coast Guard's non-homeland security programs sound?

           o  What challenges, if any, are present in trying to use the
           primary measures to link resources expended and results achieved?

3GAO, Coast Guard: Key Management and Budget Challenges for Fiscal Year
2005 and Beyond, GAO-04-636T (Washington, D.C.: Apr. 7, 2004); and Coast
Guard: Relationship between Resources Used and Results Achieved Needs to
Be Clearer, GAO-04-432 (Washington, D.C.: March 2004).

4The criteria for assessing soundness are not equal, and failure to meet a
particular criterion does not necessarily preclude that measure from being
useful; rather, it may indicate an opportunity for further refinement.

To conduct our analysis of the soundness of the primary performance
measures, we relied primarily on a set of criteria that we had previously
developed.5 These criteria were developed based on the Government
Performance and Results Act of 1993 (GPRA) and Office of Management and
Budget (OMB) guidelines for agency performance measures.6 We used our
judgment to assess whether these measures met our criteria. We also
reviewed the fiscal years 2005 and 2006 DHS Performance and Accountability
Report, the fiscal years 2006 and 2007 DHS budget requests, and the Coast
Guard's fiscal years 2006 and 2007 Budget-in-Brief.  To conduct our
reliability analysis, we relied primarily on comparisons of Coast Guard
data collection methods and internal control processes with GPRA and the
Reports Consolidation Act of 2000 requirements, as well as commonly
accepted standards and practices.7 Our reliability analysis assessed only
the specific data fields used to collect and report data for the six
non-homeland security primary performance measures, and not the relevant
databases as a whole. We reviewed and analyzed information collected and
assembled at Coast Guard headquarters as well as at four Coast Guard field
locations.8 To the extent possible, we also reviewed secondary measures
for four of the six non-homeland security programs.9 To identify and
assess the challenges in trying to use the primary measures to link
resources expended and results achieved we interviewed Coast Guard
officials at agency headquarters to discuss how measures are used in
resource and budget allocation decisions and reviewed previous GAO reports
on performance measures, performance reporting, and the link between the
Coast Guard's resources used and results achieved. We conducted our work
from July 2005 to August 2006 in accordance with generally accepted
government auditing standards. More details about the scope and
methodology of our work are presented in appendix I.

5GAO, The Results Act: An Evaluator's Guide to Assessing Agency
Performance Plans, GAO/GGD-10-1 .20 (Washington, D.C.: April 1998).

6GPRA, Pub. L. No. 103-62, 107 Stat. 285 (1993).

7The Reports Consolidation Act of 2000, Pub. L. No. 106-531, 114 Stat.
2537.

8The field locations we selected were District 1 (Boston, Massachusetts);
District 7 (Miami, Florida); District 9 (Cleveland, Ohio); and District 13
(Seattle, Washington). We selected these field locations because of the
number and types of non-homeland security programs that are performed at
these locations. We reviewed activities at multiple offices or units at
each location.

                                Results in Brief

Although some opportunities for improvement exist, the Coast Guard's
primary performance measures for its six non-homeland security programs
are generally sound, and the data used to calculate them are generally
reliable. All six measures are generally sound in that they cover key
program activities and are objective, measurable, and quantifiable, but
three are not completely clear, that is, they do not consistently provide
clear and specific descriptions of the data, events, or geographic areas
they include. For example, the primary performance measure for ice
operations, "domestic icebreaking-annual number of waterway closure days,"
actually only reflects closures for certain waterways within the Great
Lakes region. Although these waterways are the main location for domestic
icebreaking, icebreaking also takes place on the East Coast. While this
caveat is included in some accompanying text, the description is
inconsistent across department and agency publications. For instance, the
DHS fiscal year 2005 Performance and Accountability Report notes that the
measure is made up of nine critical waterways, but the DHS fiscal year
2007 budget request reports that the measure consists of seven critical
waterways, while the Coast Guard's fiscal year 2007 Budget-in-Brief does
not mention the number of waterways included in the measure. With regard
to data reliability, although the processes the Coast Guard uses to enter
and review its own internal data are likely to produce reliable data for
the performance measures we reviewed, we identified weaknesses with
processes used to review the reliability of data gathered from external
sources. Specifically, we found that neither DHS nor the Coast Guard has
policies requiring review or consistent verification processes for these
data. Instead, the processes vary for different data sources. For example,
the Coast Guard tests the reliability of state-provided data used for its
marine safety program's primary measure, but does not test the reliability
of Army Corps of Engineers (Corps) data or review the Corps' data
reliability procedures for data used for its marine environmental
protection program's primary measure. While, according to a Corps
official, the Corps does have some controls in place, without, at a
minimum, familiarity with the internal controls used by the Corps to
ensure the reliability of these data, the Coast Guard cannot provide
assurance that the data are reliable.

9The four programs we selected were aids to navigation, living marine
resources, marine environmental protection, and search and rescue. We
selected these programs because they had the largest budget increases of
the six non-homeland security programs (as reflected in the fiscal year
2005 budget and the Coast Guard's fiscal year 2006 budget request) and
because they are programs of particular interest because of events
surrounding Hurricane Katrina. Further, we selected only those measures
that Coast Guard officials said were high-level, strategic measures used
in performance budgeting, budget projections, and management decisions. In
addition, we did not assess any of the secondary measures that were in
development at the time of our report. The 23 secondary measures we
assessed for these four programs represent more than half of the 39
high-level, strategic secondary measures used to manage the six
non-homeland security programs.

For the four non-homeland security programs we assessed, more than a third
of the secondary performance measures are generally sound (9 of the 23),
while opportunities for improvement exist for the remainder (14 of the
23). More specifically, for the 14 secondary measures, we found (1) the
Coast Guard does not have measurable targets to assess whether program and
agency goals and objectives are being achieved for 12 measures, (2) the
Coast Guard does not have agencywide criteria or guidance to accurately
reflect program results and ensure objectivity for 1 measure, and (3) the
Coast Guard does not clearly state or describe the data or events included
in 1 measure. For example, a secondary measure for the search and rescue
program, "percent of lives saved after Coast Guard notification," does not
clearly state that it excludes incidents in which 11 or more lives were
saved or lost in a single case. While including such large incidents in
performance measures would skew annual performance results, it is
important for the Coast Guard to identify these exclusions, either through
a footnote or accompanying text, to ensure that events such the rescues of
Hurricane Katrina-when the agency rescued more than 33,500 people within a
few weeks-are recognized; otherwise, performance results could be
misinterpreted or misleading to users.

Although the primary performance measures are generally sound and data
used to calculate them are generally reliable, even sound performance
measures have limits to how much they can explain about the relationship
between resources expended and results achieved. Specifically, we
identified two challenges that stand in the way of establishing a clear
link between resources and results. One challenge involves the difficulty
of capturing an entire program such as ice operations or marine
environmental protection in a single performance measure. The Coast Guard
follows DHS guidance in reporting a single measure per program, which is
consistent with our prior work on agencies that were successful in
measuring performance and implementing GPRA.10 However, reporting some
secondary measures or additional data in venues, such as the Coast Guard's
annual Budget-in-Brief or program-specific publications, could provide
additional context and help to more clearly articulate to stakeholders and
decision makers the relationship between resources expended and results
achieved. For instance, reporting data on the annual number of search and
rescue cases in the search and rescue program, in addition to its primary
measure, "the percent of mariners in imminent danger who are rescued," can
provide greater context for the program's activity level. This is
important because while the percentage of mariners saved may remain
consistent from year-to-year, the number of cases, number of lives saved,
and the resources used to achieve this result can vary. The second
challenge involves the Coast Guard's ability to account for factors other
than resources that can affect program results. Some of these factors are
external to the agency-and perhaps outside of its ability to influence.
Because of the potentially large number of external factors, and their
sometimes unpredictable or often unknown effect on performance, it may be
difficult to account for how they-and not the resources expended on the
program-affect results. For example, a change in fishery regulations
reduced the number of search and rescue cases in Alaska because it
provided greater flexibility for fishermen to choose when they would fish
for certain fish stocks-this flexibility allowed them to choose different
timeframes and therefore safer weather conditions for their fishing
activities. Developing a system or model that could realistically take all
such factors into account may not be achievable, but, the challenge is to
develop enough sophistication about each program's context so the Coast
Guard can more systematically consider these factors, and then explain
their influence on resource decisions and performance results. Recognizing
these limitations, and responding to recommendations we have made in past
reports, the Coast Guard has developed a range of initiatives that agency
officials believe will help explain the effects of these factors and
decide where resources are best spent.11 Some of these initiatives have
been ongoing for several years, and according to agency officials, the
extent and complexity of the effort, together with challenges presented in
integrating them into a data-driven and comprehensive strategy, requires
additional time to complete. Currently, the Coast Guard does not expect to
fully implement many of the initiatives until 2010, and thus it is not
possible to assess their likely impact in linking resources and results
until they are further developed and operational.

10GAO, Executive Guide: Effectively Implementing the Government
Performance and Results Act, GAO/GGD-96-118 (Washington, D.C.: June 1996);
and Tax Administration: IRS Needs to Further Refine Its Tax Filing Season
Performance Measures, GAO-03-143 (Washington, D.C.: November. 2002).

To improve the quality of program performance reporting and to more
efficiently and effectively assess progress toward achieving the goals or
objectives stated in agency plans, we are recommending that the Secretary
of the Department of Homeland Security direct the Commandant of the Coast
Guard to take steps to further improve the soundness of the 3 primary
measures and 14 secondary measures we found to have weaknesses, develop
and implement a policy to review the reliability of all external data that
is used in calculating performance measures, and report additional
information-besides the one primary performance measure-in appropriate
venues to better inform stakeholders and decision makers about the
relationship between resources expended and results achieved. In
commenting on this draft, DHS and Coast Guard officials generally agreed
with our findings and recommendations, and provided technical comments
that we incorporated.

                                   Background

The Coast Guard has responsibilities divided into 11 programs that fall
under two broad missions-homeland security and non-homeland security-which
are recognized in the Homeland Security Act. To accomplish its wide range
of responsibilities, the Coast Guard is organized into two major commands
that are responsible for overall mission execution-one in the Pacific area
and the other in the Atlantic area. These commands are divided into nine
districts, which in turn are organized into 35 sectors that unify command
and control of field units and resources, such as multimission stations
and patrol boats. In fiscal year 2005, the Coast Guard had over 46,000
full-time positions-about 39,000 military and 7,000 civilians. In
addition, the agency had about 8,100 reservists who support the national
military strategy or provide additional operational support and surge
capacity during times of emergency, such as natural disasters.
Furthermore, the Coast Guard also had about 31,000 volunteer auxiliary
personnel help with a wide array of activities, ranging from search and
rescue to boating safety education.

11In 2004, we recommended that the Coast Guard identify the intervening
factors that may affect performance and systematically assess the
relationship among these factors, resources used, and results achieved.
GAO-04-432 .

For each of its six non-homeland security programs, the Coast Guard has
developed a primary performance measure to communicate agency performance
and provide information for the budgeting process to Congress, other
policymakers, and taxpayers. The Coast Guard has also developed 39
secondary measures that it uses to manage these six programs. The Coast
Guard selected and developed the six primary measures based on a number of
criteria, including GPRA, DHS, and OMB guidance as well as legislative,
department, and agency priorities. When viewed as a suite of measures, the
primary and secondary measures combined are intended to provide Coast
Guard officials with a more comprehensive view of program performance than
just the program's primary measure. Some of these secondary measures are
closely related to the primary measures; for example, a secondary measure
for the marine environmental protection program, "annual number of oil
spills greater than 100 gallons and chemical discharges per 100 million
tons shipped," is closely related to the program's primary measure,
"5-year average annual number of oil spills greater than 100 gallons and
chemical discharges per 100 million tons shipped." However, other
secondary measures reflect activities and priorities that are not
reflected in the primary performance measures. For example, a secondary
measure in the search and rescue program, "percent of property saved,"
reflects activities not captured in the program's primary measure,
"percent of mariners in imminent danger who are rescued."

In 2004, we compared trends in performance results, as reported by the
Coast Guard's primary performance measures, with the agency's use of
resources and found that the relationship between results achieved and
resources used was not always what might be expected-that is, resources
expended and performance results achieved did not have consistent
direction of movement and sometimes bore an opposite relationship.12 We
reported that disconnects between resources expended and performance
results achieved have important implications for resource management and
accountability, especially given the Coast Guard's limited ability to
explain them. In particular, these disconnects prompted a question as to
why, despite substantial changes in a number of programs' resource hours
used over the period we examined, the corresponding performance results
for these programs were not necessarily affected in a similar manner-that
is, they did not rise or fall along with changes in resources.13 At that
time, the Coast Guard could not say with any assurance why this occurred.
For example, while resource hours for the search and rescue program
dropped by 22 percent in fiscal year 2003 when compared to the program's
pre-September 11, 2001 baseline, the performance results for the program
remained stable for the same period. These results suggest that
performance was likely affected by factors other than resource hours. One
set of factors cited by the Coast Guard as helping to keep performance
steady despite resource decreases involved strategies such as the use of
new technology, better operational tactics, improved intelligence, and
stronger partnering efforts. Coast Guard officials also pointed to another
set of factors, largely beyond the agency's control (such as severe
weather conditions), to explain performance results that did not improve
despite resource increases. At the time of our 2004 report, the Coast
Guard did not have a systematic approach to effectively link resources to
results. However, the Coast Guard had begun some initiatives to better
track resource usage and manage program results, but many of these
initiatives were still in early stages of development and some did not
have a time frame for completion.

12 GAO-04-432 .

Like other federal agencies, DHS is subject to the performance-reporting
requirements of GPRA. GPRA requires agencies to publish a performance
report that includes performance measures and results. These reports are
intended to provide important information to agency managers,
policymakers, and the public on what each agency accomplished with the
resources it was given. The three key annual publications that DHS and the
Coast Guard use to report the Coast Guard's non-homeland security primary
performance measures are the DHS Performance and Accountability Report,
the DHS fiscal year budget request, and the Coast Guard's fiscal year
Budget-in-Brief. The DHS Performance and Accountability Report provides
financial and performance information to the President, Congress and the
public for assessing the effectiveness of the department's mission
performance and stewardship of resources. The DHS annual budget request to
Congress identifies the resources needed for meeting the department's
missions. The Coast Guard's annual Budget-in-Brief reports performance
information to assess the effectiveness of the agency's performance as
well as a summary of the agency's most recent budget request. These
documents report the primary performance measures for each of the Coast
Guard's non-homeland security programs, as well as descriptions of the
measures and explanations of performance results. While these documents
report performance results from some secondary measures, DHS and the Coast
Guard do not report most of the Coast Guard's secondary measures in these
documents.

13The Coast Guard maintains information on how assets, such as cutters,
patrol boats, and aircraft are used. Each hour that these resources are
used is called a resource hour. Resource hours do not include such things
as the time that the asset stands idle or the time that is spent
maintaining it.

GPRA also requires agencies to establish goals and targets to define the
level of performance to be achieved by a program and express such goals in
an objective, quantifiable, and measurable form. In passing GPRA, Congress
emphasized that the usefulness of agency performance information depends
to a large degree on the reliability of performance data. To be useful in
reporting to Congress on the fulfillment of GPRA requirements and in
improving program results, the data must be reliable-that is, they must be
seen by potential users to be of sufficient quality to be trustworthy.
While no data are perfect, agencies need to have sufficiently reliable
performance data to provide transparency of government operations so that
Congress, program managers, and other decision makers can use the
information. In establishing a system to set goals for federal program
performance and to measure results, GPRA requires that agencies describe
the means to be used to validate and verify measured values to improve
congressional decision making by providing objective, complete, accurate
and consistent information on achieving statutory objectives, and on the
relative effectiveness and efficiency of federal programs and spending.14
In addition, to improve the quality of agency performance management
information, the Reports Consolidation Act of 2000 requires an assessment
of the reliability of performance data used in the agency's program
performance report.15

OMB's Program Assessment Rating Tool (PART) is designed to strengthen and
reinforce performance measurement under GPRA by encouraging careful
development of outcome-oriented performance measures.16 Between 2002 and
2005, OMB reviewed each of the Coast Guard's six non-homeland security
programs.17 OMB found that four programs-ice operations, living marine
resources, marine environmental protection, and marine safety-were
performing adequately or better, and two programs-aids to navigation and
search and rescue-did not demonstrate results. OMB recommended that for
the aids to navigation program, the Coast Guard develop and implement a
better primary performance measure that allows program managers to
understand how their actions produce results. Specifically, OMB
recommended using an outcome-based measure, the number of collisions,
allisions, and groundings, as a measure for the program, instead of the
measure that was being used-aid availability. For the search and rescue
program, OMB recommended that the Coast Guard develop achievable long-term
goals for the program. Since these reviews, the Coast Guard has
implemented a new primary performance measure for the aids to navigation
program, "5-year average annual number of distinct collisions, allisions,
and groundings," and developed new long-term goals for the search and
rescue program's primary performance measure, that is rescuing between 85
and 88 percent of mariners in imminent danger each year from fiscal year
2002 through 2010.

14GPRA, Pub. L. No. 103-62, 107 Stat. 285 (1993).

15The Reports Consolidation Act of 2000, Pub. L. No. 106-531, S: 5, 114
Stat. 2537, 2539-40.

16OMB's PART review is a systematic method of assessing the performance of
program activities across the federal government used by OMB to review
federal agency programs. The PART review is a series of questions that
assess different aspects of program performance in which agencies under
review must answer; responses must be evidenced- based. Agencies must
clearly explain their answers and include relevant supporting evidence
such as agency performance information, independent evaluations, and
financial information. PART reviews provide an overall rating for each
program that includes effective (the program is well managed), moderately
effective (the program is well managed but needs improvements), adequate
(the program needs to improve accountability), ineffective (the program is
unable to achieve results), results not demonstrated (the program does not
have acceptable performance goals or targets).

17OMB reviewed the aids to navigation and search and rescue programs in
2002, the living marine resources and marine environmental protection
programs in 2003, the ice operations program in 2004, and the marine
safety program in 2005.

Non-Homeland Security Primary Performance Measures Are Generally Sound and Data
                  Are Generally Reliable, but Weaknesses Exist

While the six non-homeland security primary performance measures are
generally sound, and the data used to calculate these measures are
generally reliable, we found weaknesses with the soundness of three
measures and the reliability of the data used in one measure (see table
2). All six measures cover key program activities and are objective,
measurable, and quantifiable, but three are not completely clear, that is,
they do not consistently provide clear and specific descriptions of the
data, events, or geographic areas they include. The Coast Guard's
processes for entering and reviewing its own internal data are likely to
produce reliable data. However, processes for reviewing or verifying data
gathered from external sources vary from source to source, and for the
marine environmental protection measure, the processes are insufficient.

Table 2: Soundness of Primary Measures and Reliability of Data Used to
Calculate the Primary Measures for the Coast Guard's Non-Homeland Security
Programs

Legend: Yes Weaknesses identified

Source: GAO analysis of Coast Guard primary performance measures.

Although the Six Primary Measures Are Generally Sound, Three Have Weaknesses

While the six primary performance measures are generally sound-in that the
measures cover key activities of the program, and are objective,
measurable, and quantifiable-three of the measures are not completely
clear. The primary performance measures for the ice operations, living
marine resources, and search and rescue programs do not consistently
provide clear and specific descriptions of the data, events, or geographic
areas they include. It is possible these weaknesses could lead to
decisions or judgments based on inaccurate, incomplete, or misreported
data. The three programs with primary measures that are not completely
clear are as follows:

           o  Ice operations. Further clarity and consistency in reporting
           the geographic areas included in the ice operations primary
           performance measure, "domestic ice breaking-annual number of
           waterway closure days," would provide users additional context to
           discern the full scope of the measure. Despite its broad title,
           the measure does not reflect the annual number of closure days for
           all waterways across the United States, but rather reflects only
           the annual number of closure days in the Great Lakes region,
           although the Coast Guard breaks ice in many East Coast ports and
           waterways. According to Coast Guard officials, the measure focuses
           on the Great Lakes region because it is a large commerce hub where
           the icebreaking season tends to be longer and where ice has a
           greater impact on maritime transportation. While this limitation
           is included in accompanying text in some documents, the
           description of the limitation is inconsistent across department
           and agency publications. The DHS fiscal year 2005 Performance and
           Accountability Report notes that the measure is made up of nine
           critical waterways within the region, but the DHS fiscal year 2007
           budget request reports that it consists of seven critical
           waterways, while the Coast Guard's fiscal year 2007
           Budget-in-Brief does not mention the number of waterways included
           in the measure. In addition, Coast Guard program officials said
           that the measure only reflects closures in one critical
           waterway-the St. Mary's River. Coast Guard program officials at
           agency headquarters said that they are in the early stages of
           developing a new primary performance measure that will incorporate
           domestic icebreaking activities in areas beyond the Great Lakes.
           However, until a better measure is developed, the description of
           the current measure can confuse users and might cause them to
           think performance was better or worse than it actually was.

           o  Search and rescue. While the primary performance measure for
           the search and rescue program, "percent of mariners in imminent
           danger who are rescued," reflects the program's priority of saving
           lives, it excludes those incidents in which 11 or more lives were
           saved or lost. According to Coast Guard officials, an agency
           analysis in fiscal year 2005 showed that 98 percent of search and
           rescue cases involved 10 or fewer people that were saved or lost.
           Coast Guard officials added that large cases involving 11 or more
           people are data anomalies and by excluding these cases the agency
           is better able to assess the program's performance on a
           year-to-year basis. While we understand the Coast Guard's desire
           to assess program performance on a year-to-year basis, and to not
           skew the data, in some instances this type of exclusion may
           represent a significant level of activity that is not factored
           into the measure. For example, during Hurricane Katrina, the Coast
           Guard rescued more than 33,500 people. While including such large
           incidents in the performance measure would skew annual performance
           results, it is important for the Coast Guard to recognize these
           incidents, either through a footnote or accompanying text in
           department and agency publications. Not clearly defining the
           measure and recognizing such incidents may cause internal managers
           and external stakeholders to think performance was better or worse
           than it actually was.

           o  Living marine resources. Similar to the ice operations primary
           measure, the living marine resources primary performance measure,
           "percent of fishermen in compliance with regulations," is not
           consistently and clearly defined in all department and agency
           publications. The Coast Guard enforces federal regulations,
           similar to agencies across law enforcement, not by checking
           fishing vessels at random, but instead by targeting those entities
           that are most likely to be in violation of fishery regulations,
           such as vessels operating in areas that are closed to fishing.
           Because the Coast Guard targets vessels, the primary measure does
           not reflect the compliance rate of all fishermen in those areas
           patrolled by the Coast Guard, as could be inferred by the
           description, but rather is an observed compliance rate, that is,
           the compliance rate of only those fishing vessels boarded by Coast
           Guard personnel. The description of this performance measure is
           inconsistent across department and agency publications. For
           example, in the DHS fiscal year 2005 Performance and
           Accountability Report and the Coast Guard's Budget-in-Brief, this
           measure is described as an observed compliance rate, but the DHS
           fiscal year 2007 budget request does not clarify that this measure
           represents an observed compliance rate rather than the compliance
           rate of all fishermen in those areas patrolled by the Coast Guard.
           A measure that is not consistently and clearly stated may affect
           the validity of managers' and stakeholders' assessments of program
           performance, possibly leading to a misinterpretation of results.

Existing Procedures Help Ensure Reliable Internal Data, but Procedures Do Not
Exist to Check Reliability of All External Data

While the Coast Guard has controls in place to ensure the timeliness,
completeness, accuracy, and consistency of internal data it creates-that
is, original data that Coast Guard personnel collect and enter into its
data systems-the agency does not have controls in place to verify or
review the completeness and accuracy of data obtained from all external
sources that it uses in calculating some of the primary performance
measures. The internal data used to calculate the six primary performance
measures are generally reliable-in that the Coast Guard has processes in
place to ensure the data's timeliness, completeness, accuracy, and
consistency. These controls include data fields, such as pick lists and
drop-down lists, that allow for standardized data entry, mandatory data
fields to ensure all required data are entered, access controls that allow
only authorized users to enter and edit data, requirements for entering
data in a timely manner, and multiple levels of review across the agency.
To ensure data consistency across the Coast Guard, each of the six
non-homeland security programs has published definitions or criteria to
define the data used for the primary measures. However, the Coast Guard
acknowledges that in some instances these criteria may be open to
subjective interpretation, such as with the search and rescue program. For
example, when entering data to record the events of a search and rescue
incident, rescuers must identify the outcome of the event by listing
whether lives were "lost," "saved," or "assisted." While program criteria
define a life that is lost, saved, or assisted, there is potential for
subjective interpretation in some incidents.18 Through reviews at the
sector, district, and headquarters levels the Coast Guard attempts to
remedy any inconsistencies from interpretations of these criteria.

While the Coast Guard uses internal data for all six of its non-homeland
security primary performance measures, it also uses external data to
calculate the primary performance measures for two programs-marine safety
and marine environmental protection (see table 3). The Coast Guard's
procedures for reviewing external data are inconsistent across these two
programs. For example, while the Coast Guard has developed better
processes and controls for external data used in the marine safety
program's primary performance measure-such as using a news clipping
service that gathers media articles on recreational boating accidents and
fatalities and using a database that gathers recreational boating injury
data from hospitals-the agency does not have processes to test the
reliability of external data used in the marine environmental protection
program's primary performance measure. The extent to which controls are
used to verify external data for the marine safety and marine
environmental protection primary measures is described below.

18The U.S. Coast Guard Addendum to the United States National Search and
Rescue Supplement defines lives lost, saved, and assisted. A life saved is
defined as a life that would have been lost had the rescue action not been
taken, including actually pulling a person from a position of distress or
removing them from a situation that would likely have resulted in their
death had the action not been taken. A life assisted is defined as those
persons who are provided assistance that did not meet the criteria for
lives saved but did receive some assistance, however, persons merely
onboard a vessel that is provided assistance directed at the vessel (such
as providing repairs or fuel) are not necessarily assisted. To count a
life as lost there must be a body recovered; otherwise it is considered a
life-unaccounted-for. Lives lost before notification are those lives lost,
which to the best of the reporting unit's knowledge, occurred before
notification of the incident was made to the Coast Guard and lives lost
after notification are those lives lost that occurred after notification
was made to the Coast Guard.

Table 3: Source of Data Used to Calculate Non-Homeland Security Primary
Performance Measures

Source: GAO analysis of Coast Guard data.

aThe Marine Information Safety and Law Enforcement database is a Web-based
database used to track marine safety and law enforcement activities
involving commercial and recreational vessels. The system provides query,
reporting, and file-downloading capabilities to the Coast Guard marine
safety and law enforcement operating programs.

bCoast Guard District 9 (headquartered in Cleveland, Ohio) develops weekly
icebreaking reports by compiling information from icebreaking cutters
operating within the district. Information in these reports includes data
on the number of vessels beset in ice that were assisted, the number of
waterways closed because of ice, the duration of any waterway closures,
and the number of vessel transits through critical waterways. These
reports are sent directly from the cutters to the district office and
compiled into an annual report that is sent to Coast Guard headquarters.

cTo calculate the marine environmental protection primary performance
measure, the Coast Guard uses data from the Army Corps of Engineers on the
amount of oil and chemicals shipped in the United States.

dTo calculate the marine safety primary performance measure, the Coast
Guard uses state data on recreational boating deaths and injuries.

           o  Marine safety. To calculate the marine safety program's primary
           performance measure, "5-year average annual number of deaths and
           injuries of recreational boaters, mariners, and passengers," the
           Coast Guard uses internal data on deaths and injuries for mariners
           and passengers, as well as external data on recreational boating
           deaths and injuries from the Boating Accident Reporting Database
           (BARD)-a Coast Guard managed database-that relies on data
           collected and entered by the states. In 2000, the Department of
           Transportation Office of Inspector General reported that
           recreational boating fatality data collected from the states
           consistently understated the number of fatalities, in part because
           a precise definition of a recreational boating fatality did not
           exist.19 To improve the reliability and consistency of the data,
           the Coast Guard created a more precise definition and clarified
           reporting criteria by providing each state with a data dictionary
           that describes the definitions for all required data fields. In
           addition, to improve the timeliness of incident reporting, the
           Coast Guard created a Web-based version of BARD for electronic
           submission of recreational boating accident data. According to
           Coast Guard officials, this system allows Coast Guard staff to
           verify, validate, and corroborate data with each state for
           accuracy and completeness prior to inclusion in the measure.

           According to Coast Guard officials, a recent Coast Guard analysis
           showed that these efforts have improved the error rate from an
           average of about 6 percent to about 1 percent annually. However,
           despite these improvements, the Coast Guard acknowledges that some
           incidents may still never be reported, some incidents may be
           inaccurately reported, and some duplicate incidents may be
           included. Coast Guard officials told us that the agency continues
           to work to reduce these errors by developing additional steps to
           validate data. These recent steps include using a news clipping
           service that gathers all media articles concerning recreational
           boating accidents and fatalities and using a database that gathers
           recreational boating injury data from hospitals.

           o  Marine environmental protection. In contrast, the Coast Guard
           does not have processes to validate the reliability of external
           data used in the marine environmental protection program's primary
           performance measure, "5-year average annual number of oil spills
           greater than 100 gallons and chemical discharges per 100 million
           tons shipped." Each year the Coast Guard uses internal data on oil
           spills and chemical discharges, as well as external data from the
           Corps on the amount of oil and chemicals shipped annually in the
           United States to calculate this measure. However, the Coast Guard
           does not review the Corps' data for completeness or accuracy, nor
           does it review the data reliability procedures the Corps uses to
           test the data for completeness or accuracy. Coast Guard officials
           said that they did not take these steps because they had thought
           the Corps performed its own internal assessments, but they were
           also unaware of what these assessments were or whether the Corps
           actually performed them. While, according to a Corps official, the
           Corps does have some controls in place, an official at the Coast
           Guard agreed that the Coast Guard would benefit from having, at a
           minimum, some familiarity with the internal controls used by the
           Corps.

19Department of Transportation Office of Inspector General, Audit of the
Performance Measure for the Recreational Boating System, MA-2000-084
(Washington, D.C.: April 2000).

 More than a Third of the Secondary Performance Measures Assessed Are Generally
                    Sound, and the Remainder Have Weaknesses

More than a third (9 of the 23) of the secondary performance measures
assessed are generally sound-that is, they are clearly stated and
described; cover key activities of the program; and are objective,
measurable, and quantifiable (see table 4). However, as described below,
weaknesses exist for the other 14 of these 23 measures. More specifically,
for the 14 secondary measures, we found (1) the Coast Guard does not have
measurable targets to assess whether program and agency goals and
objectives are being achieved for 12 measures, (2) the Coast Guard does
not have agencywide criteria or guidance to accurately reflect program
results and ensure objectivity for 1 measure, and (3) the Coast Guard does
not clearly state or describe the data or events included in 1 measure.
These weaknesses do not allow the Coast Guard to provide assurance that
these performance measures do not lead to decisions or judgments based on
inaccurate, incomplete, or misreported information. More detail on all of
the secondary measures we assessed is in appendix II.

Table 4: Soundness of Selected Non-Homeland Security Secondary Performance
Measures

Source: GAO analysis of Coast Guard secondary performance measures.

           o  Measures without measurable targets. Twelve secondary
           measures-11 living marine resources measures and 1 marine
           environmental protection measure-do not have annual targets to
           assess whether program and agency goals and objectives are being
           achieved.20 According to Coast Guard officials, these measures do
           not have targets because the focus of the program is on the
           primary performance measures, and not the inputs and outputs
           reflected in these secondary measures. However, without any
           quantifiable, numeric targets, it is difficult for the Coast Guard
           to know the extent to which program and agency goals and
           objectives are being achieved.

           o  Measure without criteria or guidance to accurately reflect
           program results and ensure objectivity. One of the search and
           rescue program's secondary performance measures that we analyzed,
           "percent of property saved," does not have criteria or guidance
           for agency personnel to objectively and consistently determine the
           value of saved property. Despite this lack of criteria on how to
           consistently and objectively determine property values, data from
           this measure are reported in both the Coast Guard's annual
           Budget-in-Brief and the DHS fiscal year Performance and
           Accountability Report. Coast Guard officials said it would be
           difficult to develop such criteria because of the large number of
           boats and vessels and their varying values. Officials added that
           Coast Guard personnel generally do not have access to, and do not
           follow up to obtain, insurance or damage estimates for saved
           property. In addition, we found that Coast Guard units do not
           consistently record property values across the agency. For
           example, some units do not record property values at all, other
           units record property values only when the actual value can be
           determined, and other units estimate property values using a
           $1,000-per-foot-of-vessel-length rule of thumb. Without any
           criteria or guidance to determine property values, the Coast Guard
           cannot provide assurance that agency personnel are consistently
           and objectively making these determinations across the agency, and
           whether the measure accurately reflects program results.

           o  Measure not completely clear. Similar to the primary
           performance measure for the search and rescue program, one of the
           search and rescue program's secondary measures we analyzed,
           "percent of lives saved after Coast Guard notification," reflects
           the program's priority of saving lives, but excludes those
           incidents in which 11 or more lives were saved or lost in a single
           case. As with the primary measure, including such large incidents
           in performance measures would skew annual performance results, and
           thus it may be appropriate to exclude them. However, it is
           important for the Coast Guard to recognize, either through a
           footnote or accompanying text, the exclusion of these
           incidents-such as during Hurricane Katrina, in which the agency
           rescued more than 33,500 people-because otherwise, performance
           results could be misinterpreted or misleading to users.

20The 11 living marine resources performance measures without measurable
targets are the (1) number of domestic fisheries enforcement resource
hours, (2) number of active commercial fishing vessels by major fishery,
(3) number of domestic boardings by major fishery, (4) boardings per
active commercial fishing vessels by major fishery, (5) number of
significant violations by major fishery, (6) number of significant
violations per domestic resource hour, (7) status of fish stocks, (8)
number of Coast Guard members trained at Regional Fishing Training
Centers, (9) cost per Coast Guard member trained at Regional Fishing
Training Centers, (10) number of Marine Affairs graduates on active duty
and, (11) percent of Marine Affairs graduates in Marine Affairs-coded
billets. The one marine environmental protection performance measure is
the Tokyo and Paris memorandum of understanding port state control reports
measure.

 Challenges Exist in Using Measures to Link Resources to Results, but the Coast
                    Guard Is Working on Ways to Address Them

While the primary measures for the Coast Guard's six non-homeland security
programs are generally sound and use reliable data, challenges exist with
using the primary measures to assess the link between resources expended
and results achieved. Ideally, a performance measure not only tells
decision makers what a program is accomplishing, but it also gives them a
way to affect these results through the decisions they make about
resources-for example, by providing additional resources with a degree of
confidence that doing so will translate into better results. Even sound
performance measures, however, may have limits to how much they can
explain about the relationship between resources expended and results
achieved. For the Coast Guard, these limits involve (1) the difficulty of
fully reflecting an entire program such as ice operations or marine
environmental protection in a single performance measure and (2) the
ability to account for the many factors, other than resources, that can
affect program results. Recognizing these limitations, and responding to
recommendations we have made in past reports, Coast Guard officials have
been working on a wide range of initiatives they believe will help in
understanding the effects of these other factors and deciding where
resources can best be spent. According to Coast Guard officials, although
the agency has been working on some of these initiatives for several
years, the extent and complexity of the effort, together with the
challenges presented in integrating a multitude of initiatives into a
data-driven and comprehensive strategy, requires additional time to
complete. At this time, the Coast Guard does not expect many of the
initiatives to be implemented until 2010. Until these initiatives are
developed and operational, it is not possible to fully assess the overall
success the agency is likely to have in establishing clear explanations
for how its resources and results are linked.

Primary Performance Measures Cover a Key Activity, but Not Every Activity
Conducted under a Program

Performance measures are one important tool to communicate what a program
has accomplished and provide information for budget decisions. It is
desirable for these measures to be as effective as possible in helping to
explain the relationship between resources expended and results achieved,
because agencies that understand this linkage are better positioned to
allocate and manage their resources effectively. The Coast Guard follows
DHS guidance in reporting a single measure per program, and doing so is
consistent with our prior work on agencies that were successful in
measuring performance and implementing GPRA.21 Previously, we found that
agencies successful in measuring performance and meeting GPRA's
goal-setting and performance measurement requirements limited their
measures to covering core program activities essential for producing data
for decision making and not all program activities. Each of the Coast
Guard's primary measures for its six non-homeland security programs meets
our criteria of covering a key activity. None of them, however, is
comprehensive enough to capture all of the activities performed within the
program that could affect results. For example, the primary performance
measure for the marine environmental protection program relates to
preventing oil and chemical spills. This is a key program activity, but
under this program the Coast Guard also takes steps to prevent other
marine debris and pollutants (such as plastics and garbage), protect
against the introduction of invasive aquatic nuisance species, and respond
to and mitigate oil and chemical spills that actually do occur. As such,
resources applied to these other activities would not be reflected in the
program's primary measure, and thus, a clear and direct relationship
between total program resources and program results is blurred.

In some cases, it may be possible to identify or develop a performance
measure that fully encapsulates all the activities within a program, but
in many cases the range of activities is too broad, resulting in a measure
that would be too nebulous to be of real use. Coast Guard officials told
us that developing primary measures that incorporate all of the diverse
activities within some programs, as well as reflect the total resources
used within the program would be difficult, and that such a measure would
likely be too broad to provide any value for assessing overall program
performance. As such, officials added that performance measures provide a
better assessment of program performance and resource use when all of a
program's measures-both primary and secondary-are viewed in conjunction as
a suite of measures.

21 GAO/GGD-96-118 ; and GAO-03-143 .

Performance Results Can Be Affected by Factors Other than Resources

A second challenge in establishing a clearer relationship between
resources expended and results achieved is that many other factors can
affect performance and blur such a relationship. Some of these factors can
be external to an agency-and perhaps outside an agency's ability to
influence. At the time of our 2004 report, Coast Guard officials also
pointed to these external factors outside of the agency's control to
explain performance results that did not improve despite resource
increases. Because of the potentially large number of external factors,
and their sometimes unpredictable or often unknown effect on performance,
it may be difficult to account for how they-and not the resources expended
on the program-affect performance results.

Such factors are prevalent in the Coast Guard's non-homeland security
programs, according to Coast Guard officials. They cited such examples as
the following:

           o  Changes in fishing policies off the coast of Alaska had an
           effect on performance results in the search and rescue program.
           For many years, commercial sablefish and halibut fishermen were
           allowed to fish only during a 2-week period each year. Given the
           limited window of opportunity that this system provided, these
           fishermen had a strong incentive to go out to sea regardless of
           weather conditions, thereby affecting the number of the Coast
           Guard's search and rescue cases that occurred. In 1994, these
           regulations were changed; in place of a 2-week fishing season with
           no limits on the amount of fish any permitted fisherman could
           harvest, the regulations set a longer season with quotas. This
           change allowed fishermen more flexibility and more opportunity to
           exercise caution about when they should fish rather than driving
           them to go out in adverse weather conditions. Following the change
           in regulations, Coast Guard statistics show that search and rescue
           cases decreased in halibut and sablefish fisheries by more than 50
           percent, from 33 in 1994 to 15 in 1995. However, Coast Guard
           officials said that because of the large number of search and
           rescue cases in the district during these two years-more than
           1,000 annually-this policy change only had a minimal impact on the
           amount of resources the district used for search and rescue cases.

           o  Vagaries of weather can also affect a number of non-homeland
           security missions. Unusually severe weather, such as Hurricane
           Katrina, for example, can affect the success rates for search and
           rescue or cause navigational aids to be out of service. Even good
           weather on a holiday weekend, can increase the need for search and
           rescue operations-and consequently affect performance
           results-because such weather tends to encourage large numbers of
           recreational boaters to be out on the water. Harsh winter weather
           can also affect performance results for the ice operations
           program.

           o  Results for the marine environmental protection primary
           performance measure, "the 5-year average annual number of oil
           spills greater than 100 gallons and chemical discharges per 100
           million tons shipped" can be affected by policies and activities
           that are not part of the marine environmental protection program.
           For example, according to Coast Guard officials, a foreign
           country's decision to institute a more aggressive vessel
           inspection program could reduce spills caused by accidents in U.S.
           waters, if the inspections uncovered mechanical problems that were
           corrected before those vessels arrived in the United States. While
           not captured in the primary performance measure, the Coast Guard
           tracks such information through a secondary measure, "the Tokyo
           and Paris memorandums of understanding port state control
           reports."22

This small set of examples demonstrates that, in some situations, other
factors beyond resources expended may influence performance results.
Developing a system or model that could realistically take all of these
other factors into account is perhaps impossible, and it would be a
mistake to view this second challenge as a need to do so. Rather, the
challenge is to develop enough sophistication about each program's context
so that the Coast Guard can more systematically consider such factors, and
then explain the influence of these factors on resource decisions and
performance results.

22The Tokyo and Paris memorandums of understanding are agreements between
the U.S. and other countries to promote maritime safety and environmental
protection, and eliminate sub-standard shipping through port controls that
include enforcing applicable treaties. These treaties include various
construction, design, equipment, operating, and training requirements
related to maritime safety, environmental protection, and security. The
Tokyo memorandum of understanding includes 19 countries and the Paris
memorandum of understanding includes 22 countries.

Coast Guard Has Developed a Range of Initiatives to Forge Better Links between
Resources and Results

The Coast Guard is actively seeking to address such challenges, as those
discussed above, through efforts, some of which have been under way for
several years. In 2004, we reported that several initiatives had already
begun, and we recommended that the Coast Guard ensure that its strategic
planning process and associated documents include a strategy for
identifying intervening factors that may affect performance and
systematically assess the relationship among these factors, resources
expended, and results achieved. Shortly thereafter the Coast Guard
chartered a working group to investigate its then more than 50 ongoing
initiatives to make recommendations on their value, contribution, and
practicality, and to influence agency decisions on the integration,
investment, and institutionalization of these initiatives. The working
group's product was a "road map" that clearly defined executable segments,
sequencing, and priorities. These results were then documented in a
January 2005 Coast Guard internal report that summarized these
priorities.23 Agency documents indicate that the Coast Guard later reduced
these 50 original initiatives to the 25 initiatives considered to be the
most critical and immediate by evaluating and categorizing all 50
initiatives based on their ability to contribute to the agency's missions.
These 25 initiatives, listed along with their status in appendix III,
involve a broad range of activities that fall into seven main areas, as
follows:

           o  Measurement. Five initiatives are intended to improve the
           agency's data collection, including efforts to quantify input,
           output, and performance to enhance analysis and fact-based
           decision making.

           o  Analysis. Eight initiatives are intended to transform data into
           information and knowledge to answer questions and enhance decision
           making on issues such as performance, program management,
           cause-and-effect relationships, and costs.

           o  Knowledge management. Three ongoing initiatives are intended to
           capture, evaluate, and share employee knowledge, experiences,
           ideas, and skills.

           o  Alignment. Three initiatives are intended to improve the
           consistency and alignment of agency planning, resource decisions,
           and analysis across all Coast Guard programs.

           o  Access. Two initiatives relate to making data, information, and
           knowledge transparent and available to employees.

           o  Policy and doctrine. Three initiatives are intended to develop
           new and maintain current Coast Guard management policies.

           o  Communication and outreach. One initiative is intended to
           assist and guide program managers and staff to understand and
           align all aspects of the Coast Guard's overall management
           strategy.

23U.S. Coast Guard, Institutional Research Road Map (Washington, D.C.:
January 2005).

We found that one of the initiatives that the working group deemed
important and included among the most critical and immediate initiatives,
relates, in part, to the first challenge we discussed-that is, developing
new measures and improving the breadth of old measures to better manage
Coast Guard programs and achieve agency goals. Coast Guard efforts have
been ongoing in this regard, and our current work has identified several
performance measures that were recently improved, and others that are
currently under development. For example, to provide a more comprehensive
measure of search and rescue program performance, the Coast Guard is
improving its ability to track lives-unaccounted-for-that is, those
persons who at the end of a search and rescue response remain missing.
According to Coast Guard officials, the agency is working on and
anticipates being able to eventually include data on lives-unaccounted-for
in the primary performance measure. Also, the Coast Guard began including
data on the number of recreational boating injuries, along with the data
on mariner and passenger deaths and injuries and recreational boater
deaths, which can help provide a more comprehensive primary measure for
the marine safety program. In addition, recently, OMB guidance began
requiring efficiency measures as part of performance management, and in
response, the Coast Guard has started developing such efficiency measures.
The Coast Guard is also developing a variety of performance measures to
capture agency performance related to other activities, such as the
prevention of invasive aquatic nuisance species (marine environmental
protection), maritime mobility (aids to navigation), and domestic and
polar icebreaking (ice operations).

Many of the Coast Guard's other ongoing initiatives are aimed at the
second challenge-that is, developing a better understanding of the various
factors that affect the relationship between resources and results. This
is a substantial undertaking, and in 2005, upon the recommendation of the
working group, the Coast Guard created an office to conduct and coordinate
these efforts.24 This office has taken the lead in developing, aligning,
implementing, and managing all of the initiatives. Together, the
activities cover such steps as (1) improving measurement, with
comprehensive data on activities, resources, and performance; (2)
improving agency analysis and understanding of cause-and-effect
relationships, such as the relationship between external factors and
agency performance; and (3) providing better planning and decision making
across the agency. Coast Guard officials expect that once these
initiatives are completed, the Coast Guard will have a more systematic
approach to link resources to results.

The Coast Guard has already been at this effort for several years but does
not anticipate implementation of many of these initiatives until at least
fiscal year 2010. The amount of time that has already elapsed since our
2004 report may raise some concerns about whether progress is being made.
However, as described in the examples below, many of these are complex
data-driven initiatives that make up a larger comprehensive strategy to
better link resources to results, and as such, we think the lengthy time
frame reflects the complexity of the task. According to Coast Guard
officials, the agency is proceeding carefully and is still learning about
how these initiatives can best be developed and implemented. Three key
efforts help show the extent of, and interrelationships among, the various
components of the effort:

           o  Standardized reporting. The Coast Guard is currently developing
           an activities dictionary to standardize the names and definitions
           for all Coast Guard activities across the agency. According to
           Coast Guard officials, this activities dictionary is a critical
           step in continuing to develop, implement, and integrate these
           initiatives. Officials added that standardizing the names and
           definitions of all Coast Guard activities will create more
           consistent data collection throughout the agency, which is
           important because these data will be used to support many other
           initiatives.

           o  Measurement of readiness. Another initiative, the Readiness
           Management System, is a tool being developed and implemented to
           track the agency's readiness capabilities by providing up-to-date
           information on resource levels at each Coast Guard unit as well as
           the certification and skills of all Coast Guard uniformed
           personnel. This information can directly affect outcomes and
           performance measures by providing unit commanders with information
           to reconfigure resources for a broad range of missions. Tracking
           this information, for example, should allow the unit's commanding
           officer to determine what resources and personnel skills are
           needed to help ensure the unit has the skills and resources
           necessary to accomplish its key activities, or for new programs or
           activities. Coast Guard officials told us that the Readiness
           Management System is in the early stages of being implemented
           across the agency.

           o  Framework for analyzing risk, readiness, and performance.
           According to Coast Guard officials, the information from the
           Readiness Management System will be integrated with another
           initiative currently under development, the Uniform Performance
           Logic Model. This initiative is intended to illustrate the causal
           relationships among risk, readiness management, and agency
           performance. Coast Guard officials said that by accounting for
           these many factors, the model will help decision makers understand
           why events and outcomes occur, and how these events and outcomes
           are related to resources. For example, the model will provide the
           Coast Guard with an analysis tool to assist management with
           decisions regarding the allocation of resources.

24The Office of Performance Management and Decision Support was
established by the Coast Guard Chief of Staff on August 11, 2005.

The Coast Guard currently anticipates that many of the 25 initiatives will
initially be implemented by fiscal year 2010 and expects further
refinements to extend beyond this time frame. While the Coast Guard
appears to be moving in the right direction and has neared completion of
some initiatives, until all of the agency's efforts are complete, it
remains too soon to determine how effective it will be at clearly linking
resources to performance results.

                                  Conclusions

It is important for the Coast Guard to have sound performance measures
that are clearly stated and described; cover key program activities; are
objective, measurable, and quantifiable-including having annual targets;
and using reliable data. This type of information would help Coast Guard
management and stakeholders, such as Congress, make decisions about how to
fund and improve program performance. We found that the Coast Guard's
non-homeland security performance measures satisfy many of the criteria
and use data that are generally reliable. The weaknesses and limitations
we did find do not mean that the measures are not useful but rather
represent opportunities for improvement. However, if these weaknesses are
not addressed-that is, if measures are not clearly stated and
well-defined, do not have measurable performance targets, or do not have
criteria to objectively and consistently report data, or processes in
place to ensure external data are reliable-the information reported
through these measures could be misinterpreted, misleading, or inaccurate.
For example, without either processes in place to review the reliability
of external data used in performance measures, or a familiarity with the
controls used by external parties to verify and validate these data, the
Coast Guard cannot ensure the completeness or accuracy of all of its
performance results.

While the Coast Guard's measures are generally sound, even sound
performance measures have limits as to how much they can explain about the
relationship between resources expended and results achieved. The Coast
Guard continues to work to overcome these limitations by developing a
number of different initiatives, including but not limited to developing
and refining the agency's performance measures. Although the agency
appears to be moving in the right direction, until all of the Coast
Guard's efforts are complete, we will be unable to determine how effective
these initiatives are at linking resources to results. In the interim, an
additional step the Coast Guard can take to further demonstrate the
relationship between resources and results is to provide additional
information or measures in some of its annual publications-aside from the
one primary measure used in department publications-where doing so would
help provide context or provide additional perspective. For example, this
could be done in other venues-such as the Coast Guard's annual
Budget-in-Brief, or any program-specific publications-where reporting some
secondary measures or additional data could provide more context or
perspective on programs, and could help to more fully articulate to
stakeholders and decision makers the relationship between resources
expended and results achieved. Reporting supplemental information on such
things as the percentage of aids to navigation available and in need of
maintenance, the annual number of search and rescue cases, and icebreaking
activities beyond the Great Lakes region would provide additional
information on the annual levels of activity that constitute the aids to
navigation, search and rescue, and ice operations programs; information
that external decision makers, in particular, might find helpful.
Reporting these measures would be useful to provide additional information
to Congress on activities being conducted that may require more or less
funding while the Coast Guard continues its work on the many initiatives
it has ongoing aimed at better linking its performance results with
resources expended.

                      Recommendations for Executive Action

To improve the quality of program performance reporting and to more
efficiently and effectively assess progress toward achieving the goals or
objectives stated in agency plans, we recommend that the Secretary of
Homeland Security direct the Commandant of the Coast Guard to:

           o  Refine certain Coast Guard primary and secondary performance
           measures by:

                        o  further clarifying the ice operations primary
                        measure by clearly and consistently describing the
                        geographic area and number of waterways included in
                        the measure; the living marine resources primary
                        measure by clearly and consistently reporting the
                        scope of the measure; and the search and rescue
                        primary measure and the search and rescue "percent of
                        lives saved after Coast Guard notification" secondary
                        measure by reporting those incidents or data that are
                        not included in the measures;

                        o  developing measurable performance targets to
                        facilitate assessments of whether program and agency
                        goals and objectives are being achieved for the 11
                        living marine resources secondary measures and the 1
                        marine environmental protection secondary measure,
                        "Tokyo and Paris memorandums of understanding port
                        state control reports," that lack annual targets; and

                        o  establishing agencywide criteria or guidance to
                        help ensure the objectivity and consistency of the
                        search and rescue program's "percent of property
                        saved" secondary performance measure.

           o  Develop and implement a policy to review external data provided
           by third parties that is used in calculating performance measures
           to, at a minimum, be familiar with the internal controls external
           parties use to determine the reliability of their data.

           o  Report additional information-besides the one primary
           measure-in appropriate agency publications or documents where
           doing so would help provide greater context or perspective on the
           relationship between resources expended and program results
           achieved.

                                Agency Comments

We provided a draft of this report to the Department of Homeland Security,
including the Coast Guard, for their review and comment. The Department of
Homeland Security and the Coast Guard generally agreed with the findings
and recommendations of the draft and provided technical comments, which we
incorporated to ensure the accuracy of our report. The Department of
Homeland Security's written comments are reprinted in appendix IV.

As agreed with your office, unless you publicly announce the contents of
this report earlier, we plan no further distribution of it until 30 days
from the date of this letter. We will then send copies to the Secretary of
Homeland Security; the Commandant of the Coast Guard; the Director, Office
of Management and Budget; and make copies available to other interested
parties who request them. In addition, the report will be available at no
charge on GAO's Web site at http://www.gao.gov.

If you have any questions about this report, please contact me at
[email protected] or (202) 512-9610. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. GAO staff who made major contributions to this report are
listed in appendix IV.

Stephen L. Caldwell Acting Director, Homeland Security and Justice Issues

Appendix I: Objectives, Scope, and Methodology

For our first objective-whether the primary performance measure for the
Coast Guard's six non-homeland security programs are sound, and the data
used to calculate them are reliable-we used previously established GAO
criteria to determine the soundness of the primary performance measures.1
Specifically, we used our judgment to assess whether the measures are (1)
clearly stated and described; (2) cover a key program activity and
represent mission goals and priorities; (3) objective, that is whether
they are open to bias or subjective interpretation; (4) measurable, that
is, represent observable events; and (5) quantifiable, that is, are
countable events or outcomes. A measure should be clearly stated and
described so that it is consistent with the methodology used to calculate
it and can be understood by stakeholders both internally and externally.
Measures should also cover key program activities and represent program
and agency goals and priorities to help identify those activities that
contribute to the goals and priorities. To the greatest extent possible,
measures should be objective, that is, reasonably free of bias or
manipulation that would distort an accurate assessment of performance.
When appropriate, measures should be measurable and quantifiable,
including having annual targets, to facilitate future assessments of
whether goals or objectives were achieved, because comparisons can be
easily made between projected performance and actual results.

In addition, to further assess the soundness of the primary performance
measures, we interviewed program officials from each non-homeland security
program and reviewed planning and performance documentation from each
program office at the headquarters, district, and sector levels. Program
officials we spoke with included headquarters officials responsible for
developing and implementing performance measures in each program, as well
as officials at the district and sector levels responsible for collecting
and entering performance data. We reviewed documentation on Coast Guard
policies and manuals for performance measures, Coast Guard annual
performance plans and reports, commandant instructions, prior GAO reports,
Office of Management and Budget Program Assessment Rating Tool reviews for
each program, and Department of Homeland Security annual reports.

To determine the reliability of data used in the primary measures, we
assessed whether processes and controls were in place to ensure that the
data used in the measures are timely, complete, accurate, and consistent,
and appear reasonable. We reviewed legislative requirements for data
reliability in both the Government Performance and Accountability Act of
1993 and the Reports Consolidation Act of 2000 and reviewed Coast Guard
standards and procedures for collecting performance data and calculating
results. In addition, we interviewed agency officials at Coast Guard
headquarters, as well as at the district and sector levels, regarding
standardized agencywide data collection, entry, verification, and
reporting policies, and inquired as to if and how these procedures
differed across programs and at each level of the organization. We
observed data entry for the Marine Information Safety and Law Enforcement
database at Coast Guard district and sector offices in Boston,
Massachusetts; Miami, Florida; and Seattle, Washington; a district office
in Cleveland, Ohio; as well as at an air station in Miami, Florida; and a
marine safety office in Cleveland, Ohio, to check for inconsistencies and
discrepancies in how data are collected and maintained throughout the
agency. We selected these field locations because of the number and types
of non-homeland security programs that are performed at these locations.
We also spoke with information technology officials responsible for
maintaining the Marine Information Safety and Law Enforcement database.

1 GAO/GGD-10-1 .20.

For our second objective-whether selected secondary performance measures
for four of the Coast Guard's non-homeland security programs are sound-we
selected measures in addition to the primary performance measures for the
aids to navigation, living marine resources, marine environmental
protection, and search and rescue programs. We selected these programs
because they had the largest budget increases between the fiscal year 2005
budget and the Coast Guard's fiscal year 2006 budget request, and are
programs of particular interest because of events surrounding Hurricane
Katrina. In addition, we did not assess any of the secondary measures that
were in development at the time of our report. For these four programs, we
assessed the soundness of only those other performance measures that Coast
Guard officials said were high level, strategic measures used for
performance budgeting, budget projections, management decisions, and
external reporting. The 23 secondary measures we assessed for these four
programs represent more than half of the 39 high-level, strategic
secondary measures used to manage the six non-homeland security programs.
To assess the soundness of the selected 23 secondary measures, we used the
same GAO criteria and followed the same steps that we used to determine
the soundness of the primary performance measures.

For our third objective-the challenges, if any, that are present in trying
to use these measures to link resources expended to results achieved-we
interviewed Coast Guard budget officials at agency headquarters to discuss
how performance measures are used in resource and budget allocation
decision making processes. We reviewed previous GAO reports on performance
measures, performance reporting, and the link between the Coast Guard's
resources expended and results achieved. We also interviewed program
officials at Coast Guard headquarters about ongoing initiatives the agency
is developing and implementing to link resources expended to results
achieved.

We conducted our work from July 2005 to August 2006 in accordance with
generally accepted government auditing standards.

Appendix II: Secondary Performance Measures

Appendix II provides our findings for the soundness of the high-level,
strategic secondary measures we assessed (see table 5), as well as a list
of those high-level, strategic secondary measures we did not assess (see
table 6). Because of the large number of secondary measures for the Coast
Guard's six non-homeland security programs, we assessed the soundness of
secondary measures for the aids to navigation, living marine resources,
marine environmental protection, and search and rescue programs, and we
did not assess the soundness of secondary measures for the ice operations
and marine safety programs.

Table 5: Soundness of Secondary Measures for Coast Guard's Non-Homeland
Security Programs

Legend Yes Weaknesses identified

Source: GAO analysis of Coast Guard secondary performance measures.

aFour secondary measures-(1) port state annual detention ratio; (2) port
state 3-year average detention ratio; (3) Tokyo and Paris memorandums of
understanding port state control reports; and (4) annual number of
distinct collision, allision, and grounding events-are each used by the
aids to navigation, marine environmental protection, and marine safety
programs.

bThe marine environmental protection program secondary measure, 5-year
average annual number of distinct collision, allision, and grounding
events, is also the primary performance measure for the aids to navigation
program.

Table 6: Coast Guard Non-Homeland Security Secondary Performance Measures
Not Assessed

Source: Coast Guard.

Appendix III: Ongoing Coast Guard Initiatives to Link Resources Used to
Results Achieved

Appendix III provides a list of the Coast Guard's ongoing initiatives to
improve the agency's planning, resource management, and decision support
systems to more closely align performance with resources. (See table 7.)

Table 7: Ongoing Coast Guard Initiatives to Link Resources Used to Results
Achieved

Source: Coast Guard.

Appendix IV: Comments from the Department of Homeland Security

Appendix V: GAO Contact and Staff Acknowledgments

GAO Contact

Stephen L. Caldwell, Acting Director, Homeland Security and Justice
Issues, (202) 512-9610, or [email protected]

Acknowledgments

In addition to the individual named above, Billy Commons, Christine Davis,
Michele Fejfar, Dawn Hoff, Allen Lomax, Josh Margraf, Dominic Nadarski,
Jason Schwartz, and Stan Stenersen made key contributions to this report.

Related GA Related GAO Products

Coast Guard: Station Readiness Improving, but Resource Challenges and
Management Concerns Remain. GAO-05-161 . Washington, D.C.: January 31,
2005.

Coast Guard: Relationship between Resources Used and Results Achieved
Needs to Be Clearer. GAO-04-432. Washington, D.C.: March 22, 2004.

Coast Guard: Comprehensive Blueprint Needed to Balance and Monitor
Resource Use and Measure Performance for All Missions. GAO-03-544T.
Washington, D.C.: March 12, 2003.

Performance Reporting: Few Agencies Reported on the Completeness and
Reliability of Performance Data. GAO-02-372 . Washington, D.C.: April 26,
2002.

Coast Guard: Budget and Management Challenges for 2003 and Beyond.
GAO-02-538TU . Washington, D.C.: March 19, 2002.

Coast Guard: Update on Marine Information for Safety and Law Enforcement
System. GAO-02-11 . Washington, D.C.: October 17, 2001.

Tax Administration: IRS Needs to Further Refine Its Tax Filing Performance
Measures. GAO-03-143 . Washington, D.C.: November 22, 2002.

The Results Act: An Evaluator's Guide to Assessing Agency Performance
Plans. GAO/GGD-10-1 .20. Washington, D.C.: April 1998.

Agencies' Annual Performance Plans under the Results Act: An Assessment
Guide to Facilitate Congressional Decision Making. GAO-GGD/AIMD-10.1.18.
Washington, D.C.: February 1998.

Executive Guide: Effectively Implementing the Government Performance and
Results Act. GAO/GGD-96-118 . Washington, D.C.: June 1996.

(440432)

GAO's Mission

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( www.gao.gov ). Each weekday, GAO posts newly
released reports, testimony, and correspondence on its Web site. To have
GAO e-mail you a list of newly posted products every afternoon, go to
www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: [email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional Relations

Gloria Jarmon, Managing Director, [email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Washington,
D.C. 20548

Public Affairs

Paul Anderson, Managing Director, [email protected] (202) 512-4800 U.S.
Government Accountability Office, 441 G Street NW, Room 7149 Washington,
D.C. 20548

www.gao.gov/cgi-bin/getrpt? GAO-06-816 .

To view the full product, including the scope

and methodology, click on the link above.

For more information, contact Stephen L. Caldwell at (202) 512-9610 or
[email protected].

Highlights of GAO-06-816 , a report to the Subcommittee on Fisheries and
Coast Guard, Committee on Commerce, Science, and Transportation, U.S.
Senate

August 2006

COAST GUARD

Non-Homeland Security Performance Measures Are Generally Sound, but
Opportunities for Improvement Exist

Using performance measures, the Coast Guard explains how well its programs
are performing. To do so, it reports one "primary" measure for each
program (such as percent of mariners rescued) and maintains data on other,
"secondary" measures (such as percent of property saved). Concerns have
been raised about whether measures for non-homeland security programs
accurately reflect performance, that is, they did not rise or fall as
resources were added or reduced. For the six non-homeland security
programs, GAO used established criteria to assess the soundness of the
primary measures-that is, whether measures cover key activities; are
clearly stated; and are objective, measurable, and quantifiable-and the
reliability of data used to calculate them. GAO also used these criteria
to assess the soundness of 23 selected secondary measures. Finally,
through interviews and report review, GAO assessed challenges in using
measures to link resources to results.

What GAO Recommends

GAO made recommendations to clarify, develop targets, establish criteria,
and review external data for certain performance measures and improve the
Coast Guard's overall reporting of results. DHS and the Coast Guard
generally agreed with the recommendations in this report.

While some opportunities for improvement exist, the primary measures for
the Coast Guard's six non-homeland security programs are generally sound,
and the data used to calculate them are generally reliable. All six
measures cover key program activities and are objective, measurable, and
quantifiable, but three are not completely clear-that is, they do not
consistently provide clear and specific descriptions of the data, events,
or geographic areas they include. Also, the processes used to enter and
review the Coast Guard's own internal data are likely to produce reliable
data; however, neither the Department of Homeland Security (DHS) nor the
Coast Guard have policies or procedures for reviewing or verifying data
from external sources, such as other federal agencies. Currently, the
review processes vary from source to source, and for the primary measure
covering marine environmental protection (which concerns oil and chemical
spills), the processes are insufficient.

Of the 23 secondary performance measures GAO assessed, 9 are generally
sound, with weaknesses existing in the remaining 14. These weaknesses
include (1) a lack of measurable performance targets, (2) a lack of
agencywide criteria or guidance to ensure objectivity, and (3) unclear
descriptions of the measures.

Two main challenges exist with using primary measures to link resources to
results. In one case, the challenge is comprehensiveness-that is, although
each primary measure captures a major segment of program activity, no one
measure captures all program activities and thereby accounts for all
program resources. The other challenge involves external factors, some of
which are outside the Coast Guard's control, that affect performance. For
example, weather conditions can affect the amount of ice that must be
cleared, the number of aids to navigation that need repair, or mariners
that must be rescued. As a result, linking resources and results is
difficult, and although the Coast Guard has a range of ongoing initiatives
to do so, it is still too early to assess the agency's ability to
successfully provide this link.

Soundness of Primary Measures and Reliability of Data Used to Calculate
the Primary Measure for the Coast Guard's Non-Homeland Security Programs

Legend: Yes Weaknesses identified

Source: GAO analysis of Coast Guard primary performance measures.
*** End of document. ***