This is the accessible text file for GAO report number GAO-13-757 
entitled 'FAA Facilities: Improved Condition Assessment Methods Could 
Better Inform Maintenance Decisions and Capital-Planning Efforts' 
which was released on September 10, 2013. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as 
part of a longer term project to improve GAO products' accessibility. 
Every attempt has been made to maintain the structural and data 
integrity of the original printed product. Accessibility features, 
such as text descriptions of tables, consecutively numbered footnotes 
placed at the end of the file, and the text of agency comment letters, 
are provided but may not exactly duplicate the presentation or format 
of the printed version. The portable document format (PDF) file is an 
exact electronic replica of the printed version. We welcome your 
feedback. Please E-mail your comments regarding the contents or 
accessibility features of this document to Webmaster@gao.gov. 

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. 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: 
GAO: 

Report to Congressional Committees: 

September 2013: 

FAA Facilities: 

Improved Condition Assessment Methods Could Better Inform Maintenance 
Decisions and Capital-Planning Efforts: 

GAO-13-757: 

GAO Highlights: 

Highlights of GAO-13-757, a report to Congressional Committees. 

Why GAO Did This Study: 

that FAA staff occupy, FAA is responsible for maintaining over 40 
percent of them—spending over $1 billion since 2008; most of the 
remaining facilities are leased and maintained by the lessors. Yet, 
according to FAA officials, air traffic control facilities-—the bulk 
of FAA-maintained facilities-—have a current maintenance backlog of 
about $259 million. Deferring maintenance on facilities could 
compromise safe airspace operations. 

The FAA Modernization and Reform Act of 2012 mandated that GAO study 
the condition of FAA-staffed facilities. This report addresses, among 
other things, (1) the condition of FAA-staffed facilities and the 
reliability of assessment methods; (2) the extent to which FAA has 
responded to identified safety deficiencies; and (3) the extent to 
which FAA’s actions to ensure that its facilities are in good 
condition follow leading practices. GAO analyzed data and documents 
from FAA, the Department of Labor, and GSA, and interviewed agency 
officials and others, including union representatives. 

What GAO Found: 

According to data provided by FAA and the General Services 
Administration (GSA), federally maintained facilities FAA occupies are 
generally in fair to good condition. FAA assesses the condition of 
staffed facilities in several ways, including direct inspections. 
However, the condition of approximately 75 percent of Air Traffic 
Organization’s (ATO) terminal facilities is based on estimates derived 
from the approximately 25 percent of facilities that have been 
inspected over the last 6 years. Our analysis of FAA’s statistical 
model for estimating the condition of uninspected terminal facilities 
found it to be imprecise; it uses one variable-—age of the facility-—
to estimate the facility’s condition. Adding other variables that are 
correlated with condition, such as facility replacement value and use, 
could potentially improve the estimate’s accuracy. In addition, 
facility condition data in FAA’s Real Estate Management System (REMS)—-
FAA’s database for tracking its inventory of real property assets, 
including the size, replacement value, and condition of each asset—are 
not derived from sound data collection practices and did not match 
facility condition data from ATO or other FAA organizational 
components responsible for maintaining these facilities. Inaccuracies 
in REMS data undermine its usefulness as a management tool. 

Table: Entities Responsible for Maintaining FAA’s Staffed Facilities, 
2013: 

Facility description: Operations (terminal & en-route facilities); 
FAA: 425; 
GSA: 0; 
Other (including non-federal): 137; 
Total: 562. 

Facility description: Administrative, technical, training; 
FAA: 80; 
GSA: 8; 
Other (including non-federal): 583; 
Total: 671. 

Facility description: Total; 
FAA: 505; 
GSA: 8; 
Other (including non-federal): 720; 
Total: 1,233. 

Source: GAO analysis of FAA and GSA data. 

[End of table] 

FAA has mechanisms to identify and mitigate safety deficiencies at FAA 
facilities. For example, FAA annually conducts a safety and health 
inspection, as required by regulation, at each of its staffed 
facilities under its Environmental and Occupational Safety and Health 
program. Identified deficiencies are tracked to ensure they are 
mitigated. FAA also established safety committees to maintain an open 
channel of communication between employees and managers concerning 
safety matters in the workplace. FAA employees may also report any 
hazards they identify to FAA managers at their facility or, as 
applicable, to their union representative or the Occupational Safety 
and Health Administration. 

FAA has taken actions to strengthen its capital planning process to 
help ensure its facilities are in good condition. In 2012, FAA created 
a Facilities Group Manager position within ATO to coordinate 
management of all its operations facilities, both staffed and 
unstaffed. ATO’s Facilities Group Manager is taking steps to develop a 
standard prioritization process focusing on safety, mission critical 
needs, and environmental requirements to modernize and sustain 
existing equipment, facilities, and services; this approach is to be 
used to develop FAA’s future budget requests. FAA is developing a 10-
year “Get Well” plan that, among other things, is to establish an 
approach to reduce the maintenance backlog for existing facilities. 
This plan, along with a facilities consolidation report the agency is 
developing in response to the FAA Modernization and Reform Act of 
2012, should help provide justification and cost projections for 
maintaining facilities in good condition. 

What GAO Recommends: 

To ensure more accurate and reliable data to help decision making on 
its facilities’ conditions, GAO is recommending that FAA (1) improve 
the precision of the methods used to estimate the conditions of 
uninspected terminal facilities and (2) implement a plan to improve 
REMS, consistent with sound data collection practices. DOT provided 
technical comments, which we incorporated as appropriate. 

View [hyperlink, http://www.gao.gov/products/GAO-13-757]. For more 
information, contact Gerald Dillingham, 202-512-2834, 
dillinghamg@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

FAA-Staffed Facilities Are Generally in "Good" to "Fair" Condition, 
but Opportunities Exist to Improve Assessment Methods and Condition 
Information: 

While the Number of Injuries and Illnesses Resulting in FAA's Workers' 
Compensation Claims Has Steadily Decreased from 2007 Through 2011, FAA 
Has Not Always Met Its Goals for Processing Claims: 

FAA Has Taken Steps to Identify and Abate Workplace Hazards: 

FAA Has Taken Steps to Strengthen Its Capital-Planning Process and 
Investment-Decision Practices That Are Generally Consistent with 
Leading Practices: 

Conclusions: 

Recommendations: 

Agency Comments: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Assessment of FAA's Statistical Model to Estimate the 
Condition of Terminal Facilities: 

Appendix III: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: Maintenance Responsibilities for FAA's Staffed Facilities, 
2013: 

Table 2: Aggregate FCI and Deferred Maintenance Values for En-route 
Air Traffic Control Centers and Inspected Terminal Facilities, 2010 
through 2012 (Dollars in Millions): 

Table 3: Comparison of Results from FAA and GAO Models for Estimating 
Deferred Maintenance at Selected Terminal Facilities (Dollars in 
Thousands): 

Table 4: Top Five Types of Injuries or Illnesses Claims Filed by FAA 
Employees and Accepted by OWCP, 2007 to 2012: 

Table 5: Ratio of FAA Employees to Number of Claims Filed by FAA 
Employees and Accepted by OWCP, 2007 through 2011: 

Table 6: FAA's Total Annual Workers' Compensation Claims Payments, 
2007 through 2012: 

Table 7: FAA CA-1 and CA-2 Filing Timeliness: Percentage within 14 
Days, Fiscal Year 2007 through 2012: 

Table 8: Leading Capital-Planning Practices for Using Information to 
Make Capital-Investment Decisions: 

Table 9: FAA's Actual and Projected Capital Improvement Budget Line 
Items, Fiscal Years 2008 through 2017 (Dollars in Millions): 

Table 10: FAA Facilities GAO Contacted: 

Table 11: Results of Replication and Predictive Validation of FAA 
Model: 

Table 12: Results of Replication and Predictive Validation of FAA 
Model: 

Figures: 

Figure 1: Eight FAA Air Traffic Control Towers Reported to Have the 
Same FCI and Deferred Maintenance Amount: 

Figure 2: Number of Workers' Compensation Claims of FAA Employees 
Accepted by OWCP, 2007 to 2012: 

Figure 3: Estimated versus Actual Deferred Maintenance in 2010: 

Abbreviations: 

AIRTRAF: Airport Traffic Control Tower Monitoring Inspection Program: 

ALO: Aviation Logistics Organization: 

ATO: Air Traffic Organization: 

BLI: budget line item: 

CIP: Capital Investment Plan: 

CMO: Certificate Management Office: 

EOSH: Environmental and Occupational Safety and Health: 

FAA: Federal Aviation Administration: 

FCI: Facility Condition Index: 

FRPP: Federal Real Property Profile: 

FSDO: Flight Standards District Office: 

GSA: General Services Administration: 

MIDO: Manufacturing Inspection District Office: 

NAS: National Airspace System: 

NextGen: Next Generation Air Transportation System: 

OMB: Office of Management and Budget: 

OSHECCOM: Occupational Safety, Health, and Environmental Compliance 
Committee: 

OSHA: Occupational Safety and Health Administration: 

OWCP: Office of Workers' Compensation Programs: 

PASS: Professional Aviation Safety Specialists: 

REMS: Real Estate Management System: 

RMSE: root mean squared error: 

TRACON: Terminal Radar Approach Control: 

T-SAP: Technical Operations Safety Action Program: 

WIT: Workplace Inspection Tool: 

[End of section] 

GAO:
United States Government Accountability Office: 
441 G St. N.W. 
Washington, DC 20548: 

September 10, 2013: 

The Honorable John D. Rockefeller IV: 
Chairman: 
The Honorable John Thune: 
Ranking Member: 
Committee on Commerce, Science, and Transportation: 
United States Senate: 

The Honorable Bill Shuster: 
Chairman: 
The Honorable Nick J. Rahall, II: 
Ranking Member: 
Committee on Transportation and Infrastructure: 
House of Representatives: 

For the Federal Aviation Administration (FAA) to safely operate the 
most complex airspace system in the world, its employees and systems 
need to be housed in facilities that are in good and safe condition. 
FAA staff occupy approximately 1,230 facilities that serve 
administrative and operational functions. FAA is responsible for 
maintaining almost half of these facilities--primarily those used for 
air traffic control operations. The General Services Administration 
(GSA) maintains the federally owned facilities that FAA occupies for 
administrative purposes. As for the remaining facilities, FAA occupies 
them under leases or other agreements whereby the owner provides 
facility maintenance. 

Since 2008, FAA has spent over $1 billion for repairs and improvements 
to the facilities it maintains. While some of the facilities that FAA 
maintains will become excess as a result of planned consolidation, 
most will likely remain in service to support FAA's mission. To 
sustain these facilities, FAA officials require accurate information 
about facility condition to plan and prioritize maintenance and 
repairs so that operational needs can be met in a safe working 
environment. FAA's efforts to address the needs of its aging 
facilities portfolio, meet facilities requirements related to 
implementation of the Next Generation Air Transportation System 
(NextGen),[Footnote 1] and move forward with facility consolidation 
plans will likely be affected by resource constraints as a result of 
expected lower future budgets. The challenges presented by these 
circumstances relate to long-standing problems that we have identified 
with federal real property, an area that we have designated as high-
risk.[Footnote 2] 

Congress mandated in the FAA Modernization and Reform Act of 2012 that 
we study the condition of FAA-staffed facilities, including 
administrative and operational facilities, employee occupational 
safety and health issues, and resources allocated to facility 
maintenance and renovation.[Footnote 3] This report follows up an 
interim report we issued on May 9, 2013, which discussed our 
preliminary findings.[Footnote 4] To meet the mandate, this report 
addresses the following questions: 

1. What is known about the conditions of FAA-staffed facilities and to 
what extent are the assessment methods and resulting data reliable? 

2. What have been the workplace injuries and illnesses reported by FAA 
employees in workers' compensation claims, and to what extent has FAA 
met its targets for processing claims? 

3. To what extent has FAA responded to safety deficiencies identified 
through its inspections and those conducted by the Occupational Safety 
and Health Administration (OSHA)? 

4. To what extent are actions FAA has taken to ensure that its 
facilities are in good condition, if any, consistent with leading 
practices for capital-investment decisions? 

To determine what is known about the conditions of FAA facilities and 
the extent to which the methods used to make those determinations and 
the resulting data are reliable, we collected condition data from 2010 
through 2012 from FAA and GSA. To assess the reliability of the FAA 
data we reviewed existing documentation, interviewed cognizant agency 
officials, and performed electronic testing for obvious errors in 
accuracy and completeness. To assess the reliability of the GSA data, 
we reviewed existing documentation and interviewed knowledgeable 
agency officials. We found that data in FAA's agency-wide Real Estate 
Management System (REMS) were not sufficiently reliable for describing 
facility condition, as discussed later in this report. However, we 
determined that data from other sources maintained by FAA 
organizational components with facilities maintenance responsibilities 
and from GSA were sufficiently reliable for this purpose. To determine 
whether the inspection methods used by FAA or GSA to obtain facility 
condition information were reliable for this purpose, we compared 
those methods with generally accepted industry standards and 
interviewed FAA and GSA facility officials and FAA's engineering 
consultants. We also interviewed FAA and union officials representing 
major aviation employee groups about facility conditions. To observe 
conditions and speak with knowledgeable staff to learn about any 
facility deficiencies and the projects for fixing them, we visited 
facilities supporting operations and administrative functions in each 
of FAA's three geographic service areas.[Footnote 5] To determine the 
most common types of workplace injuries and illnesses FAA employees 
have reported, if any, we reviewed FAA and Department of Labor (Labor) 
workers' compensation data from 2007 through 2012. To determine the 
extent to which FAA has complied with its targets for processing 
claims, we compared actual performance to the targets. To assess the 
reliability of the data, we reviewed existing documentation, 
interviewed knowledgeable agency officials, and performed electronic 
testing for obvious errors in accuracy and completeness. We found the 
data to be sufficiently reliable for the purposes of this report. 
Because of the variability of factors that may contribute to reported 
injuries and illnesses, our study does not attempt to link workers' 
compensation claims to facility conditions. To determine how FAA has 
responded to any safety deficiencies identified through inspections of 
its facilities, we reviewed FAA and OSHA inspection reports and 
documents pertaining to how FAA has addressed deficiencies. Regarding 
actions FAA has taken to ensure that its facilities are in good and 
safe condition, consistent with leading practices, we reviewed FAA's 
process for making capital decisions and compared it to our and the 
Office of Management and Budget's (OMB) leading practices.[Footnote 6] 
Appendix I has a more detailed discussion of our scope and methodology. 

We conducted this performance audit from July 2012 to September 2013 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence we obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

Background: 

Facility Maintenance Responsibility: 

Responsibility for maintaining the approximately 1,230 staffed FAA 
facilities--ranging from administrative buildings to operations 
buildings such as air traffic control towers--is distributed among 
FAA, GSA, and other entities, as shown in table 1.[Footnote 7] Within 
FAA, several components hold responsibility for facilities 
maintenance, including the Air Traffic Organization (ATO), Mike 
Monroney Aeronautical Center (Aeronautical Center), and William J. 
Hughes Technical Center (Technical Center). ATO maintains facilities 
supporting air traffic control operations at multiple locations across 
the country and overseas. These include en-route traffic control 
facilities that control high-altitude air traffic and terminal 
facilities that control low altitude air traffic. FAA's Aeronautical 
Center and Technical Center maintain facilities on their respective 
campuses in Oklahoma City and Atlantic City that provide research and 
training functions. GSA maintains some of FAA's administrative 
facilities including FAA's headquarters in Washington, D.C., and 
regional offices around the country. FAA also occupies several hundred 
facilities across the country through leases or other types of 
agreements under which the (non-federal) owner is responsible for 
maintenance. These include air traffic control towers and 
administrative facilities that serve a variety of functions.[Footnote 
8] 

Table 1: Maintenance Responsibilities for FAA's Staffed Facilities, 
2013: 

Facility Description: En-route Traffic Control Facilities: 
Responsible entity (number of facilities): 
FAA: 23; 
GSA: 0
Other (including non-federal): 0
Total: 23. 

Facility Description: Terminal facility sites, including Air Traffic 
Control Towers and Terminal Radar Approach Control (TRACON) facilities; 
Responsible entity (number of facilities): 
FAA: 402; 
GSA: 0; 
Other (including non-federal): 137; 
Total: 539[A]. 

Facility Description: FAA Mike Monroney Aeronautical Center[B]; 
Responsible entity (number of facilities): 
FAA: 37; 
GSA: 0; 
Other (including non-federal): 23; 
Total: 60. 

Facility Description: FAA William J. Hughes Technical Center; 
Responsible entity (number of facilities): 
FAA: 43; 
GSA: 0; 
Other (including non-federal): 0; 
Total: 43. 

Facility Description: Administrative/Other; 
Responsible entity (number of facilities): 
FAA: 0; 
GSA: 8; 
Other (including non-federal): 560[C]; 
Total: 568. 

Facility Description: Total; 
Responsible entity (number of facilities): 
FAA: 505; 
GSA: 8; 
Other (including non-federal): 720; 
Total: 1,223. 

Source: GAO analysis of FAA and GSA data. 

Note: FAA maintains some facilities that are not federally owned. For 
example, under the terms of some lease agreements, FAA maintains 
terminal facilities that are owned by non-federal entities such as 
airport authorities. 

[A] Some of the terminal facility sites contain more than one 
facility. According to FAA facility data, there are 543 facilities at 
these sites. 

[B] In 2013, Aeronautical Center staff occupy 60 facilities that 
include facilities that are either federally owned or leased from the 
Oklahoma City Airport Trust, which has primary responsibility for 
their maintenance. The Aeronautical Center also has an additional 68 
federally maintained and leased facilities that are considered 
unstaffed. 

[C] This includes 33 facilities that GSA leases on behalf of FAA for 
FAA staff; the property owners are responsible for their maintenance. 

[End of table] 

GSA's and FAA's internal organizational components responsible for 
facility maintenance collect information on the condition of their 
respective facility portfolios. This generally includes information on 
the Facility Condition Index (FCI)[Footnote 9] and deferred 
maintenance[Footnote 10]--two widely used benchmarks to determine the 
relative condition of public and private facilities. However, because 
of the variation in methods used to determine condition, the condition 
data for one type of facility may not be comparable to that for 
another type of facility. FAA also maintains an agency-wide Real 
Estate Management System (REMS) to collect information, including FCI 
data, on facilities in its inventory. FAA uses REMS for reporting 
required information to the Federal Real Property Profile (FRPP). 
[Footnote 11] 

Funding for facilities modernization, improvement, and replacement 
projects derives from FAA's Facilities and Equipment account.[Footnote 
12] The funding in this account is tied to the annually updated 
Capital Investment Plan, or CIP. This plan is to identify planned 
capital investments for the next 5 years consistent with the amount 
requested in FAA's annual budget submission. 

FAA Employees' Safety and Health: 

Two agencies within Labor administer programs related to safety and 
health issues in federal facilities: OSHA, which oversees compliance 
with occupational safety and health standards, and the Office of 
Workers' Compensation Programs (OWCP), which oversees certain programs 
to compensate employees for wages lost as a result of employment-
related injuries or illnesses. OSHA sets and enforces workplace 
standards affecting FAA facilities. As part of its enforcement 
efforts, OSHA conducts programmed and unprogrammed inspections of 
worksites to ensure compliance with its safety and health standards. 
[Footnote 13] Among other things, OWCP adjudicates federal employees' 
claims for compensation resulting from employment-related injuries and 
illnesses. FAA staff who sustain employment-related injuries and 
illnesses may file a workers' compensation claim with FAA for 
subsequent transmittal to OWCP.[Footnote 14] 

In addition, as required by law, FAA administers an environmental and 
occupational safety and health (EOSH) program that includes 
inspections aimed at preventing workplace injuries.[Footnote 15] Under 
its EOSH program, FAA annually inspects each of its staffed facilities 
and maintains inspection results in a central database. Also, under 
its EOSH program, FAA established Occupational Safety, Health, and 
Environmental Compliance Committees (OSHECCOM) to provide a method for 
employees to use their knowledge of workplace operations to assist 
agency management in improving policies, conditions, and practices 
concerning safety and health matters in agency workplaces. 

FAA-Staffed Facilities Are Generally in "Good" to "Fair" Condition, 
but Opportunities Exist to Improve Assessment Methods and Condition 
Information: 

Most Federally Maintained Facilities Occupied by FAA Employees Are in 
"Good" to "Fair" Condition: 

According to data available for the federally maintained facilities 
occupied by FAA staff, facilities are generally in "good" to "fair" 
condition based on their FCIs.[Footnote 16] This information results 
from the condition assessment processes and data sources used by each 
FAA organizational component that has facility maintenance 
responsibility as well as from GSA.[Footnote 17] Specifically: 

* ATO uses a technical-consulting firm to annually inspect a portion 
of en-route and terminal facilities. ATO's consultant annually 
produces "roll-up" reports that provide condition information for 
facilities that have been inspected during the current year as well as 
during the preceding 5-year period.[Footnote 18] For terminal 
facilities not inspected over this time period, ATO's consultant 
estimates their condition by using the condition data from inspected 
facilities; these estimates are contained in the annual terminal 
facilities roll-up report.[Footnote 19] 

* Aeronautical Center managers use a technical-consulting firm to 
inspect their federally maintained facilities on a biennial basis. 
Aeronautical Center managers use condition data obtained through these 
inspections and update them between inspection cycles to reflect 
repairs or improvements made to those facilities. 

* Technical Center managers used a technical contractor to inspect 
most of their federally maintained facilities in 2007 and 2008. 
[Footnote 20] Technical Center managers use condition data derived 
from these inspections to determine what improvements are needed. 

* GSA uses a standardized physical condition survey process 
administered by its on-site staff to biennially assess the condition 
of the facilities it maintains. GSA manages a database containing 
information derived from these surveys. 

Because of variations in the methods used to determine condition, the 
condition data for one type of facility may not be comparable to that 
for another type of facility. Our analysis of condition by facility 
type is presented below. 

En-route Air Traffic Control Centers and Terminal Facilities: 

According to ATO's most recent roll-up reports: 

* 23 en-route air traffic control centers are generally in "fair" 
condition having an aggregated FCI of 93.5 percent,[Footnote 21] with 
deferred maintenance totaling approximately $98 million and 
replacement values totaling about $1.5 billion. 

* 134 terminal facilities--the subset of terminal facilities that have 
been inspected--are at the transition from "good" to "fair" condition 
having an aggregated FCI of 94.9 percent, with deferred maintenance 
totaling about $82 million and replacement values totaling about $1.6 
billion.[Footnote 22] 

* 409 terminal facilities--the subset of terminal facilities that have 
not been inspected--have estimated FCI and deferred maintenance 
values. Because FAA's estimates for these uninspected facilities have 
substantial errors, we do not present condition data for them. We 
discuss in greater detail below and in appendix II the imprecision of 
FAA's estimating method and alternatives for improving the accuracy of 
its estimates. 

Based on our interviews with ATO and its technical consultant, we 
found the inspection and its related reporting process results are 
reliable indicators of the condition of inspected facilities based on 
generally accepted industry standards. However, as we explain below, 
limitations in ATO's process for estimating the condition of terminal 
facilities that have not been inspected may result in estimates that 
overstate or understate actual conditions. Moreover, based on our 
discussions with ATO and its technical consultant, we learned that 
condition data prior to 2010 were obtained using a different method 
than currently used. As a result, we determined that year-to-year 
comparison of condition data was appropriate only for data for 
inspected facilities from 2010 through 2012. Table 2 shows aggregate 
FCI and deferred maintenance values for en-route air traffic control 
centers and terminal facilities over this time period. 

Table 2: Aggregate FCI and Deferred Maintenance Values for En-route 
Air Traffic Control Centers and Inspected Terminal Facilities, 2010 
through 2012: 

Facility type: En-route centers; 
2010: 
FCI: 93.7%; 
Deferred maintenance: $91 million; 
2011: 
FCI: 93.7%; 
Deferred maintenance: $94 million; 
2012: 
FCI: 93.5%; 
Deferred maintenance: $98 million. 

Facility type: Inspected terminal facilities[A]; 
2010: 
FCI: 95.2%; 
Deferred maintenance: $61.8 million; 
2011: 
FCI: 95.5%; 
Deferred maintenance: $65.2 million; 
2012: 
FCI: 94.9%; 
Deferred maintenance: $81.9 million. 

Source: National Roll-Up Reports for En-route Facilities and Final 
MARS Roll-Up Reports for ATCT-TRACON Sites, FAA. 

Notes: Data for En-route Centers are based on the fiscal year while 
data for terminal facilities are based on the calendar year. 

The aggregate FCI is the summed deferred maintenance values divided by 
the summed replacement values, subtracted from 1. The aggregate 
deferred maintenance values represent the summed deferred maintenance 
amounts across the relevant subset of facilities. 

[A] The data shown are for only those terminal facilities that have 
been inspected and exclude deficiencies resulting from seismic, code, 
and standards assessments. For 2010, there were 114 inspected terminal 
facilities; for 2011, there were 121; and for 2012, there were 134. 

[End of table] 

Mike Monroney Aeronautical Center Facilities: 

According to fiscal year 2010 data, the aggregated FCI of 110 
Aeronautical Center facilities is 95.2 percent, or "good" condition, 
based on deferred maintenance totaling about $6.5 million and 
replacement values totaling about $135 million.[Footnote 23] Based on 
interviews with Aeronautical Center managers, we found the inspection 
method followed generally accepted industry standards and its related 
reporting process results and are reliable facility condition 
indicators. 

William J. Hughes Technical Center Facilities: 

The Technical Center's inspection process does not yield facility-
level condition data such as an FCI. Rather, the process assesses the 
conditions of the components--such as structural, electrical, and 
heating/cooling systems--that comprise each facility. Because 
available data for Technical Center facilities does not reflect 
facility-level conditions similar to what we reported for other 
facility types, we do not present them in this report. 

GSA-Maintained FAA Facilities: 

According to GSA data, the 8 staffed facilities it maintains for FAA 
have FCIs ranging from 95 percent ("good" condition) to 46 percent 
("poor" condition).[Footnote 24] GSA-maintained facilities where FAA 
is the principal occupant are the Hawthorne Federal Building (46.0 
percent FCI), Orville Wright Building (82.0 percent FCI), and Wilbur 
Wright Building (90.0 percent FCI).[Footnote 25] We found the 
facilities condition data for the FAA facilities that GSA maintains to 
be reliable indicators of these facilities.[Footnote 26] We also 
observed that GSA uses a different convention than FAA in categorizing 
facilities based on their FCI values. This results in GSA's having 
different FCI thresholds than FAA for determining if a facility is in 
"good," "fair," or "poor" condition.[Footnote 27] 

REMS Facility Condition Data Differ from Those Maintained by ATO and 
the Aeronautical Center: 

FAA collects information on all facilities in its portfolio (both 
federally owned and leased) and enters the data in REMS, FAA's 
repository for its entire real-property inventory. Information 
maintained in REMS includes the location, size, age, plant replacement 
value, operations and maintenance costs, repair needs, and FCI data 
for facilities. REMS is overseen by FAA's Aviation Logistics 
Organization (ALO), which provides guidance for users on updating REMS 
facilities data. FAA uses REMS for reporting required information to 
the FRPP.[Footnote 28] 

We initially assessed REMS data to obtain information on the condition 
of the agency's facilities but found it was not reliable for this 
purpose. For example, 16.5 percent (141) of the records for non-leased 
facilities in the initial REMS file FAA provided us had missing FCIs 
or an FCI of zero.[Footnote 29] We also found that the FCI data 
obtained from ATO and the Aeronautical Center did not consistently 
match FCI data in REMS. REMS uses an FCI that was modeled on DOD's 
facility-pricing index while ATO's and the Aeronautical Center's 
engineering consultants calculate FCIs from actual inspections. 
[Footnote 30] FAA officials recognized that ATO's and the Aeronautical 
Center's FCI data were more robust and indicated that we should use 
their data in lieu of REMS in our assessment of facilities that are 
maintained by FAA. 

Throughout the course of our review, we brought to ALO's attention 
deficiencies that we found with REMS, such as facilities with wrong 
addresses, incorrect use status, and erroneous lease costs. 
Furthermore, ALO independently recognized shortcomings with REMS and 
informed us that it has also been trying to correct errors with the 
data as well as address deficiencies we identified.[Footnote 31] For 
example, Aeronautical Center officials told us that whenever they 
tried to update REMS with their FCI data, the updated condition data 
would revert back to the original REMS data after a certain period of 
time. FAA officials and the consulting firm that manages REMS told us 
that a "business rule" in the programming for REMS caused the 
condition data updates entered by Aeronautical Center staff to be 
overwritten, causing the accurate up-to-date data to be replaced by 
the old data. ALO recently agreed to change this "business rule" to 
prevent updated facility condition data from being overwritten. In 
addition, ALO has taken steps to correct the errors with address, use 
status, and lease data. For example, ALO recently found 13 cases of an 
air traffic control tower noted as "unstaffed" and updated the 
database to show their use as staffed facilities. ALO officials also 
plan to replace the different identifiers currently used in REMS for 
denoting an air traffic control tower with just one primary 
identifier. ALO officials agreed that REMS data need to be better 
synchronized with other real property data in FAA's other 
organizational components. To be better stewards of FAA's real 
property, on June 24, 2013, FAA chartered a property accountability 
board--a cross-organizational entity that is to continue to undertake 
strategic, management, and operational initiatives to improve property 
management. 

We are encouraged with the steps FAA is taking to address the REMS 
deficiencies we identified to ensure a strong data validation and 
correction process, but the extent to which other deficiencies exist 
is not clear as FAA has not comprehensively assessed the quality of 
REMS data and raises questions as to whether the database is a useful 
tool for managing FAA's facilities portfolio. Federal internal control 
standards state that agencies have relevant, reliable, and timely 
information for decision-making and external-reporting purposes. 
[Footnote 32] This is enabled by sound and consistent data collection 
practices that ensure data are reasonably complete and accurate. OMB 
also has data quality guidelines for ensuring quality in information 
disseminated by federal agencies.[Footnote 33] Furthermore, leading 
practices for using information to make capital-investment decisions 
call for a needs assessment that makes use of an accurate and up-to-
date inventory of assets as well as current information on their 
condition.[Footnote 34] Comprehensively assessing the quality of the 
REMS data, determining if the data are appropriate for their intended 
use, and implementing corrective actions, as needed, to ensure that 
data are sufficiently complete and accurate with data maintained among 
other FAA organizational components would provide FAA with better 
assurance that its source for facilities information agencywide is 
reliable for making informed decisions. 

FAA's Method of Assessing the Condition of Uninspected Terminal 
Facilities Produces Error and Uncertainty: 

As indicated above, ATO uses a statistical method to estimate the 
condition of most terminal facilities; however, the limitations of 
this method produce estimates with substantial error and uncertainty. 
ATO has current inspection data for approximately one quarter of all 
terminal facilities (based on a 6-year rolling count of inspected 
facilities[Footnote 35]) and estimates the condition of the remaining 
facilities that have not been inspected. ATO's consulting firm 
annually develops a statistical model based on data obtained from the 
inspected facilities to estimate the condition of facilities that have 
not been inspected. In particular, the model uses facility age to 
estimate the FCI and deferred maintenance of each facility not 
inspected. In 2012, for example, ATO's consultant developed a 
statistical model based on inspection results of 134 inspected 
facilities to estimate the FCI and deferred maintenance at 409 
facilities not inspected. ATO's 2012 terminal facilities "roll-up" 
report presents condition information for these facilities. 

The following example derived from this roll-up report illustrates the 
limitations of using facility age as the sole determinant in 
estimating FCI and deferred maintenance. In the 2012 roll-up report, 
eight air traffic control towers, each 17 years old, were identified 
as having the same FCI and deferred maintenance amount, about 96.5 
percent and around $215,000 respectively, as estimated based on age 
(see figure 1). However, given differences in their locations, sizes, 
number of flight operations managed, and physical components, actual 
conditions at each are likely different. For example, two of these 
towers are located at two of the busiest airports in the country, 
Chicago, IL, and Los Angeles, CA. These towers are larger and 
experience more wear than other towers as a result of comparatively 
greater use by controllers. During our visit to the Los Angeles tower, 
officials showed us the single elevator that services the tower's cab 
and described its maintenance issues.[Footnote 36] These officials 
noted, at the time of our visit in January 2013, that this elevator 
had experienced 17 outages since May 2012 and would cost about 
$500,000 to repair.[Footnote 37] 

Figure 1: Eight FAA Air Traffic Control Towers Reported to Have the 
Same FCI and Deferred Maintenance Amount: 

[Refer to PDF for image: 8 photographs and associated data] 

ATO’s model predicts that these 8 air traffic control towers have 
exactly the same FCI (about 96.5 percent) and deferred maintenance 
(about $215,000) because they are all 17 years old, and age is the 
sole independent variable the model uses to predict FCI and deferred 
maintenance. However, given differences among tower locations, flight 
operations patterns, and facility structures, actual conditions at 
each tower are likely different. 

Bangor International (Bangor, Maine): 
Total operations: 25,995; 
Tower cab area (square feet): 350; 
Tower height (in feet): 111. 

Bellingham International (Bellingham, Washington): 
Total operations: 16,914; 
Tower cab area (square feet): 365; 
Tower height (in feet): 58. 

Los Angeles International (Los Angeles, California): 
Total operations: 603,381; 
Tower cab area (square feet): 665; 
Tower height (in feet): 252. 

Montgomery Regional (Montgomery, Alabama): 
Total operations: 28,937; 
Tower cab area (square feet): 380; 
Tower height (in feet): 109. 

Chicago O’Hare International (Chicago, Illinois): 
Total operations: 874,561; 
Tower cab area (square feet): 1,100; 
Tower height (in feet): 230. 

Worcester Regional (Worcester, Massachusetts): 
Total operations: 3,556; 
Tower cab area (square feet): 350; 
Tower height (in feet): 100. 

San Diego International (San Diego, California): 
Total operations: 183,985; 
Tower cab area (square feet): 477; 
Tower height (in feet): 104. 

Tallahassee Regional (Tallahassee, Florida): 
Total operations: 40,533; 
Tower cab area (square feet): 440; 
Tower height (in feet): 93. 

Sources: FAA (photographs) and GAO analysis of FAA data. 

[End of figure] 

To quantify the accuracy of the methods ATO's consultant used to 
estimate the FCIs and deferred maintenance of uninspected terminal 
facilities shown in the 2012 roll-up report, we compared FCI and 
deferred maintenance data derived from actual inspection of 134 
facilities[Footnote 38] to what their estimated FCI and deferred 
maintenance values would be using the statistical model. In reference 
to FAA's general categorization of facilities in "good", "fair", or 
"poor" condition based on their FCI values, we found that 41 percent 
of the inspected facilities would have been reclassified using 
estimated rather than actual FCI values. In particular, the 
statistical model classified 32 of the 134 facilities (24 percent) in 
a better condition category than actual inspection data showed, and it 
classified 23 of 134 facilities (17 percent) in a worse condition 
category than actual inspection data showed. For deferred maintenance, 
we found that, in 2012, 94 percent of facilities had estimated 
deferred maintenance values that differed by greater than 10 percent 
from actual values. 

As another approach to testing the accuracy of ATO's statistical model 
for estimating FCI and deferred maintenance, and assessing 
alternatives for improving the model, we used data presented in the 
2010 roll-up report as described below. Appendix II presents a more 
detailed description of our analysis. 

* Accuracy of FCI estimates. To assess the accuracy of FCI estimates, 
we compared actual FCI results (based on inspections) to the model's 
predictions for the same facilities. The goal was to calculate the 
uncertainty of the predicted FCIs, as calculated from the ATO model. 
We found that ATO's estimates were not highly precise. The model 
estimated that if ATO measured actual FCIs for a new sample of 
facilities, the actual FCIs would vary by an average of plus or minus 
3.4 percentage points in 95 percent of the new samples. That is a 
moderately large error, given that the middle 50 percent of actual 
FCIs ranged from 90.9 to 96.6 percentage points. In other words, the 
margin of error of the estimates was almost as large as the range of 
the actual FCIs. 

* Accuracy of deferred maintenance estimates. To assess the accuracy 
of deferred maintenance estimates, we compared actual deferred 
maintenance results (based on inspections) to the model's predictions 
for the same facilities. Similar to our findings on FCI estimates, we 
found that ATO's estimates for deferred maintenance were not highly 
precise. On average, the model's estimates varied from the actual 
deferred maintenance by $606,862. This is a moderately large error, 
given that the middle 50 percent of actual deferred maintenance values 
ranged from $231,420 to $663,944 in this time period. 

Our analysis shows that although FCI and deferred maintenance are 
correlated with age, the strength of the relationship is not strong 
enough to produce highly precise estimates of FCI and deferred 
maintenance for facilities that are not inspected. As a result, the 
true condition of the uninspected facilities is less precise and 
certain than FAA's roll-up reports suggest. Based on our analysis of 
FAA's data, we determined that increasing the number of variables in 
FAA's estimation method could potentially improve its accuracy. For 
example, we found that incorporating facility replacement value 
[Footnote 39] into the model for estimating deferred maintenance 
significantly reduced the error and uncertainty of the estimates. 
Table 3 illustrates the difference in accuracy between the ATO 
consultant's use of age as the only factor to predict deferred 
maintenance and our use of age and facility replacement value. On 
average, our model predicted deferred maintenance with 39 percent less 
error than ATO's current approach. 

Table 3: Comparison of Results from FAA and GAO Models for Estimating 
Deferred Maintenance at Selected Terminal Facilities: 

Facility: Dallas - Ft. Worth TRACON (Texas); 
Facility age (years): 36; 
Facility replacement value: $42,112,000; 
Deferred maintenance - actual[A]: $2,893,000; 
Deferred maintenance - FAA estimate[B]: $524,000; 
Deferred maintenance - GAO estimate[C]: $3,143,000. 

Facility: Greensboro Tower (North Carolina); 
Facility age (years): 36; 
Facility replacement value: $7,506,000; 
Deferred maintenance - actual[A]: $1,173,000; 
Deferred maintenance - FAA estimate[B]: $524,000; 
Deferred maintenance - GAO estimate[C]: $592,000. 

Facility: Teterboro Tower (New Jersey); 
Facility age (years): 36; 
Facility replacement value: $4,215,000; 
Deferred maintenance - actual[A]: $289,000; 
Deferred maintenance - FAA estimate[B]: $524,000; 
Deferred maintenance - GAO estimate[C]: $349,000. 

Facility: Woodring Tower (Oklahoma); 
Facility age (years): 36; 
Facility replacement value: $3,129,000; 
Deferred maintenance - actual[A]: $150,000; 
Deferred maintenance - FAA estimate[B]: $524,000; 
Deferred maintenance - GAO estimate[C]: $269,000. 

Facility: Yakima Tower (Washington); 
Facility age (years): 36; 
Facility replacement value: $3,487,000; 
Deferred maintenance - actual[A]: $295,000; 
Deferred maintenance - FAA estimate[B]: $524,000; 
Deferred maintenance - GAO estimate[C]: $296,000. 

Source: GAO analysis: 

[A] Actual deferred maintenance as shown in FAA's 2010 terminal 
facility roll-up report. 

[B] Estimate based on FAA consultant's model using facility age to 
predict deferred maintenance. 

[C] Estimate based on GAO model using facility age and facility 
replacement value to predict deferred maintenance. 

[End of table] 

For its part, ATO's consultant has documented in the roll-up reports 
the limitations of its model to estimate FCI and deferred maintenance 
and has recommended improvements, such as using facility size as a 
variable in the model. As indicated by our analysis, increasing the 
number of variables in the method for estimating FCI and deferred 
maintenance could potentially improve the FCI's accuracy, because the 
method would include other factors that are correlated with condition, 
such as usage and weather conditions. In addition, assessing and 
reporting the prediction error associated with any statistical 
estimate, regardless of the estimation method used, is useful in 
quantifying the degree to which the estimate varies from actual 
conditions. Also, inspecting the condition of a probability sample of 
terminal facilities, either as a supplement or alternative to FAA's 
current approach, may help ensure that FCI estimates accurately 
represent the population, particularly for reporting average and total 
condition across facilities. Having accurate and reliable facility 
condition data allows management to have a clear picture of repair 
needs that may affect future budget requests.[Footnote 40] 

While the Number of Injuries and Illnesses Resulting in FAA's Workers' 
Compensation Claims Has Steadily Decreased from 2007 Through 2011, FAA 
Has Not Always Met Its Goals for Processing Claims: 

The Most Common Types of Injuries and Illnesses Claimed Include 
Orthopedic and Mental, Emotional, or Nervous Conditions: 

Workers' compensation claims filed by FAA employees and accepted by 
OWCP from 2007 through 2012[Footnote 41] indicate that the five most 
common types of injuries and illnesses sustained by FAA employees 
ranged from mental, emotional, or nervous conditions to different 
types of orthopedic injuries (see table 4). According to FAA 
officials, orthopedic injuries represent the most frequent types of 
injuries filed by FAA employees and accepted by OWCP. These claims 
cite various causes of reported injuries and illnesses including 
falls, handling tools and materials, fume inhalation, and vehicle 
accidents. Workers' compensation claims data cannot be linked to 
facility conditions because of the variability of factors contributing 
to reported injuries and illnesses. 

Table 4: Top Five Types of Injuries or Illnesses Claims Filed by FAA 
Employees and Accepted by OWCP, 2007 to 2012: 

Nature of injury: Mental, emotional, or nervous condition; 
Number of Claims: 928. 

Nature of injury: Back sprain/strain or back pain[A]; 
Number of Claims: 826. 

Nature of injury: Sprain/strain of ligament, muscle, tendon (not 
back)[A]; 
Number of Claims: 700. 

Nature of injury: Pain, swelling, redness, stiffness (not in joint)[A]; 
Number of Claims: 615. 

Nature of injury: Joint pain/swelling/stiffness/redness[A]; 
Number of Claims: 472. 

Source: GAO analysis of Labor data. 

Note: Data are based on the date of injury indicated on certified 
workers' compensation claims filed by FAA employees and accepted by 
OWCP. Data for 2012 are from January through September. 

[A] FAA considers these injuries to be orthopedic injuries. 

[End of table] 

The Number of Workers' Compensation Claims Filed by FAA Employees and 
Accepted by OWCP Decreased from 2007 through 2011: 

The number of workers' compensation claims filed by FAA employees and 
accepted by OWCP has steadily decreased since 2007.[Footnote 42] 
According to OWCP data, in 2011, 907 injuries resulted in workers' 
compensation claims filed by FAA employees that were accepted by OWCP-
-a decrease of 32 percent since 2007 (see figure 2).[Footnote 43] Over 
the same time period, the number of permanent FAA employees increased 
from 44,423 in 2007 to 47,242 in 2011. The ratio of the number of FAA 
employees to the number of workers' compensation claims filed by an 
FAA employee and accepted by OWCP has increased from 33: 1 in 2007 to 
52: 1 in 2011 (see table 5). 

Figure 2: Number of Workers' Compensation Claims of FAA Employees 
Accepted by OWCP, 2007 to 2012: 

[Refer to PDF for image: vertical bar graph] 

Year: 2007; 
Number of claims: 1,336. 

Year: 2008; 
Number of claims: 1,118. 

Year: 2009; 
Number of claims: 1,045. 

Year: 2010; 
Number of claims: 949. 

Year: 2011; 
Number of claims: 907. 

Year: 2012; 
Number of claims: 615. 

Note: Data are based on the date of injury indicated on workers' 
compensation claims filed by FAA employees and accepted by OWCP. 

[A] Data for 2012 are from January through September. 

[End of figure] 

Table 5: Ratio of FAA Employees to Number of Claims Filed by FAA 
Employees and Accepted by OWCP, 2007 through 2011: 

Claims; 
2007: 1,336; 
2008: 1,118; 
2009: 1,045; 
2010: 949; 
2011: 907. 

Number of FAA employees; 
2007: 44,423; 
2008: 45,272; 
2009: 47,020; 
2010: 47,456; 
2011: 47,242. 

Employees per claim; 
2007: 33:1; 
2008: 40:1; 
2009: 45:1; 
2010: 50:1; 
2011: 52:1. 

Source: GAO analysis of FAA and DOL OWCP data. 

[End of table] 

FAA's total annual workers' compensation claim payments remained 
relatively constant during this period, ranging from $87 million to 
$91 million (see table 6). According to FAA, the annual payments have 
remained constant due to the large number of claims it must pay out 
that were accepted from previous years. For example, in fiscal year 
2012, FAA paid 2,885 workers' compensation claims at a total cost of 
$95 million. Of those claims, 766 were accepted from 1961 to 1983 and 
accounted for $46.7 million of the 2012 costs. An FAA official 
attributed the constant annual costs to the rising costs of medical 
care and the annual cost of living adjustment to the level of 
compensation being provided to FAA claimants. 

Table 6: FAA's Total Annual Workers' Compensation Claims Payments, 
2007 through 2012: 

Annual Payments: 
2007: $87.0 million; 
2008: $90.9 million; 
2009: $92.5 million; 
2010: $91.0 million; 
2011: $90.6 million; 
2012: $94.9 million. 

Source: FAA data. 

[End of table] 

FAA Uses Various Measures to Track the Performance of Its Workers' 
Compensation Program: 

FAA tracks its performance in processing employees' workers' 
compensation claims on an annual basis. According to FAA, it is 
required to process an employee's workers' compensation claim (form CA-
1 or CA-2) internally and send it to Labor's OWCP for evaluation 
within 10 business days or 14 calendar days. FAA typically sets an 
annual performance target relative to this requirement. From 2007 
through 2012, FAA has both exceeded and missed the target twice. FAA 
did not set a target for 2 of the 6 years. See table 7 for the annual 
targets and year end performance. 

Table 7: FAA CA-1 and CA-2 Filing Timeliness: Percentage within 14 
Days, Fiscal Year 2007 through 2012: 

Target: 
2007: N/A; 
2008: 65.3%; 
2009: 68.6%; 
2010: N/A; 
2011: 86.9%; 
2012: 89.5%. 

Year-end performance: 
2007: 77.1%; 
2008: 83.0%; 
2009: 84.4%; 
2010: 85.0%; 
2011: 85.0%; 
2012: 87.3%. 

Source: Labor data. 

[End of table] 

FAA also measures the performance of its workers' compensation program 
by comparing changes in the percentage of its workers' compensation 
costs against similar changes government-wide. From fiscal year 2007 
to 2012, FAA's annual workers' compensation costs increased 9 percent 
($87.0 million to $94.9 million) while the government-wide average 
increased 21 percent ($2.49 billion to $3.01 billion). 

Each year, FAA also estimates the amount of compensation payments and 
continuation of pay avoided when FAA either challenges a claim of 
questionable veracity and it is subsequently denied by OWCP, or an 
employee returns to duty following a disabling work injury as a result 
of FAA case management actions. From 2007 to 2012, FAA's annual 
compensation payments and continuation of pay avoided estimates for 
workers' compensation payments ranged from $8 million to $26 million. 

FAA Has Taken Steps to Identify and Abate Workplace Hazards: 

FAA's EOSH Program Includes Steps to Identify and Manage the 
Mitigation of Hazards: 

FAA has mechanisms to identify safety deficiencies at FAA facilities. 
For example, FAA annually conducts a safety and health inspection, as 
required by regulation,[Footnote 44] at each of its staffed facilities 
[Footnote 45] under its Environmental and Occupational Safety and 
Health (EOSH) program.[Footnote 46] The goals of FAA's EOSH program 
are to (1) ensure a safe working environment, (2) establish a uniform, 
national workplace inspections process to identify hazards with the 
potential to cause injury, illness, or death, and (3) monitor and 
evaluate FAA's compliance with workplace inspection and hazard 
abatement requirements, measured against applicable federal, state, 
and local regulations. According to FAA, trained FAA employees or 
contractors perform the annual EOSH inspections. Before inspecting a 
facility, inspectors must review the injury/illness records, previous 
inspection reports, open findings, reports of unsafe and unhealthful 
working conditions, and employee complaints. After inspecting a 
facility, according to FAA, inspectors brief the facility's managers 
and then enter the results in FAA's Workplace Inspection Tool (WIT), 
which is a central database that management uses to track and ensure 
that identified safety and health hazards are mitigated. 

As of March 15, 2013, FAA was tracking 123,368 separate hazards in 
WIT, with over 96 percent of those hazards abated and 4 percent open 
since WIT was implemented in October 2007. FAA officials stated that 
many of the open hazards are infrastructure-based improvements that 
will require significant funding to abate. For example, FAA reported 
that a $2 million infrastructure-based improvement is needed to abate 
extensive water intrusion and mold, which has caused the closure of 
three levels of the air traffic control tower at the San Juan, Puerto 
Rico, airport. The WIT was initially used by ATO but was recently 
expanded to the rest of FAA because of its effectiveness, according to 
FAA. 

FAA Contracted for an Evaluation of EOSH Programs: 

To better understand and improve its EOSH program, FAA commissioned a 
consultant's study, issued in December 2012, to evaluate the EOSH 
programs across the agency. The report concluded that ATO has 
established the most comprehensive program within the FAA. However, 
the report also noted that 8 out of 11 organizational components 
evaluated were either unaware of their responsibilities or had 
undertaken little or no action to implement a complaint program. 
According to FAA officials, ATO's EOSH program will absorb the EOSH 
programs of the other organizational components to mitigate the issues 
described in the report and ensure a uniform approach to workplace 
safety. 

FAA Established Safety Committees: 

In June 2008, FAA established an Occupational Safety, Health, and 
Environmental Compliance Committee (OSHECCOM) at the National, Region, 
Center, and field levels.[Footnote 47] The committees were created to 
provide a method for employees to use their knowledge of workplace 
operations to assist agency management in improving polices, 
conditions, and practices. The role of OSHECCOM is to assist FAA's 
organizational components in maintaining an open channel of 
communication between employees and management concerning safety and 
health matters in the workplace.[Footnote 48] 

According to FAA, the national OSHECCOM has improved the EOSH program. 
For example, the national OSHECCOM implemented a cardiopulmonary 
resuscitation and automated external defibrillator program that 
trained thousands of FAA employees to render emergency medical care to 
individuals in FAA facilities. According to FAA officials, FAA 
employees used the training at an airline training facility in Texas 
to save the life of a pilot who was in cardiac arrest. FAA's employee 
union organizations stated that the OSHECCOMs, on the national level, 
have improved the level of communication between FAA and employees and 
that they expect it to continue to improve as the EOSH programs from 
the rest of FAA are moved under ATO's EOSH program. In addition, 
according to union officials we met with at the Aeronautical Center, 
the center's OSHECCOM was able to identify and mitigate serious 
pedestrian safety issues on the center's campus. 

FAA Has Established Other Mechanisms to Report Safety Hazards at FAA 
Facilities: 

FAA employees may also report any hazards they identify to FAA 
managers at their facility or, as applicable, to their union 
representative or OSHA. According to FAA, facility managers then work 
to abate the hazard. In addition, FAA has agreements with employee 
union organizations to report safety concerns. Specifically, the Air 
Traffic Safety Action Program is an agreement between FAA and the 
National Air Traffic Controllers Association and the National 
Association of Government Employees that is designed to foster a 
voluntary, cooperative, nonpunitive environment for FAA air traffic 
employees to openly report safety concerns. Similarly, the Technical 
Operations Safety Action Program (T-SAP) is an agreement between FAA 
and the Professional Aviation Safety Specialists (PASS) that allows 
technicians represented by PASS and other non-bargaining unit 
Technical Operations employees the opportunity to report potential 
safety hazards voluntarily and confidentially. T-SAP is still a 
demonstration program, only available to employees in the Central 
Service Area, but FAA plans to expand this program. 

OSHA Inspections Provide Another Mechanism to Promote Workplace Safety: 

OSHA conducted inspections at 153 FAA facilities from 2007 to 2012. Of 
these inspections, 57 were unprogrammed, including those originating 
from employee complaints such as unsafe egress. The 96 other 
inspections were programmed inspections that targeted specific 
worksite hazards or high-hazard worksites. OSHA began one such 
targeted inspection program to focus on concerns related to egress and 
fire safety at federally owned air traffic control towers operated by 
FAA. Under OSHA's Airport Traffic Control Tower Monitoring Inspection 
Program (AIRTRAF), which was conducted from 2008 to 2010, OSHA found 
deficiencies at 70 of the 75 air traffic control towers it inspected. 
[Footnote 49] Specifically, 279 violations were cited at the 70 
towers. The most frequently cited violations at the towers were for 
the lack of smoke/fire stops throughout the towers, missing exit 
signage in emergency egress routes, storage of combustible materials, 
and failure to conduct stair pressurization-system testing. According 
to OSHA, as of February 2013, FAA has corrected these deficiencies, 
and all 75 towers now comply with the applicable egress and fire 
safety standards. Of the remaining towers, according to FAA, 77 are 
currently not compliant with OSHA standards for egress and fire 
safety, with an estimated cost of $57.5 million to repair them. 
According to FAA officials, it will take 4 years to upgrade the 
remaining towers to meet egress and fire safety standards. 

FAA Has Taken Steps to Strengthen Its Capital-Planning Process and 
Investment-Decision Practices That Are Generally Consistent with 
Leading Practices: 

FAA has taken steps to strengthen its capital-planning process to 
ensure its facilities are in good condition. We identified leading 
practices for using information to make capital-investment decisions 
primarily from GAO's Executive Guide and OMB's Capital Programming 
Guide (see table 8).[Footnote 50] We also drew from leading capital 
investment practices identified by the National Research Council. We 
compared FAA's actions related to capital-planning and investment-
decision making--focusing on operations facilities since they are the 
principal type of facility FAA maintains--to the criteria established 
in these guides. 

Table 8: Leading Capital-Planning Practices for Using Information to 
Make Capital-Investment Decisions: 

Leading practices: Needs assessment; 
Description: A comprehensive needs assessment identifies the resources 
needed to fulfill both immediate requirements and anticipated future 
needs based on the results-oriented goals and objectives that flow 
from the organization's mission. A comprehensive assessment of needs 
considers the capability of existing resources and makes use of an 
accurate and up-to-date inventory of capital assets and facilities, as 
well as current information on asset condition. Using this 
information, an organization can make decisions about where to invest 
in facilities. 

Leading practices: Review and approval of framework with established 
criteria for selecting capital investments; 
Description: Agencies should establish a formal process for senior 
management to review and approve proposed capital assets. The cost of 
a proposed asset, the level of risk involved in acquiring the asset, 
and its importance to achieving the agency mission should be 
considered when defining criteria for executive review. Leading 
organizations have processes that determine the level of review and 
analysis based on the size, complexity, and cost of a proposed 
investment or its organization wide impact. As a part of this 
framework, proposed capital investments should be compared to one 
another to create a portfolio of major assets ranked in priority order. 

Leading practices: Project prioritization; 
Description: Leading organizations have processes in which proposed 
capital investments should be compared to one another to create a 
portfolio of major assets ranked in priority order. 

Leading practices: Strategic linkage; 
Description: Capital planning is an integral part of an agency's 
strategic-planning process. It provides a long-range plan for the 
capital asset portfolio in order to meet the goals and objectives in 
the agency's strategic and annual performance plans. Agency strategic 
and annual performance plans should identify capital assets and define 
how they will help the agency achieve its goals and objectives. 
Leading organizations also view strategic planning as the vehicle that 
guides decision making for all spending. 

Leading practices: Long-term capital plan; 
Description: The long-term capital plan should be the final and 
principal product resulting from the agency's capital-planning 
process. The capital plan should cover 5 years or more and should 
reflect decision makers' priorities for the future. Leading 
organizations update long-term capital plans either annually or 
biennially. Agencies are encouraged to include certain elements in 
their capital plans, including a statement of the agency mission, 
strategic goals and objectives; a description of the agency's planning 
process; baseline assessments and identification of performance gaps; 
and a risk management plan. 

Source: GAO and OMB. 

[End of table] 

Needs Assessment: 

Leading practices suggest that to establish a baseline of condition 
and needs, organizations should maintain systems that track the use 
and performance of existing assets. FAA's organizational components 
with responsibilities for managing facilities--ATO, the Aeronautical 
Center, and the Technical Center--have utilized technical consultants 
to inspect and determine the relative condition of building structures 
and systems. The inspections have resulted in the development of 
maintenance and repair requirements to correct identified deficiencies 
as documented in summary reports. For example, ATO's terminal and en-
route and Aeronautical Center's facility inspection reports show 
maintenance and repair requirements for major components at each 
facility.[Footnote 51] Similarly, the Technical Center's master plan 
contains component-specific information for facilities that indicate 
operability or need for replacement. 

Leading practices also indicate that a comprehensive assessment of 
needs uses an accurate and up-to-date inventory of facilities as well 
as current information on their condition. While FAA's consultants' 
inspections provide current information on the condition of inspected 
facilities, the information on the condition of facilities that have 
not been inspected may be inaccurate. As previously discussed, ATO 
estimates the condition of most of the terminal facilities, and the 
method used to make the estimates produces results that may 
inaccurately reflect actual conditions and associated maintenance and 
repair needs. As a result, FAA cannot accurately determine any 
performance gap between current and needed conditions at its terminal 
facilities and is at risk of making suboptimal decisions about where 
to invest in maintenance and repairs. 

Review and Approval of Framework with Established Criteria for 
Selecting Capital Investments: 

Leading practices call for a comprehensive decision-making framework 
to review, rank, and select from among competing project proposals. 
Such a framework should include the appropriate levels of management 
review, and selections should be based on established criteria. In 
January 2012, FAA created a Facilities Group Manager position within 
ATO to coordinate management of all its operations facilities, both 
staffed and unstaffed. Prior to this organizational change, managers 
of the different facility types--terminal facilities, en-route 
centers, and unstaffed facilities--acted independently in identifying 
requirements and prioritizing facility modernization and replacement 
projects. According to the Facilities Group Manager, under this 
previous organizational structure program managers of the different 
facility types competed with each other for funding and did not ensure 
common practices in planning and implementing modernization and 
replacement projects. As a result, managers likely applied different 
criteria in determining requirements and prioritizing projects so that 
investment decisions might not have addressed the most important needs 
of the operations facilities (ATO) portfolio. Going forward, the 
Facilities Group Manager's goal is to ensure the use of common 
criteria to prioritize projects across facility types and make 
investment decisions that best align with mission needs. 

Project Prioritization: 

FAA organizational components responsible for managing facilities 
incorporate elements of leading practices in their approaches to 
prioritizing capital-investment projects. Leading capital-planning 
practices suggest that project prioritization processes use weighted 
criteria and consider long-term capital plans and project risk. FAA 
organizational components with facility management responsibility 
currently use weighted criteria that reflect risk considerations as 
illustrated by the following descriptions of their prioritization 
processes. 

* En-route centers: ATO's process for prioritizing projects at its en-
route centers weighs proposals (in decreasing order of importance) on 
their (1) direct operational requirement, (2) safety impact, (3) 
indirect operational requirement, and (4) FCI. ATO considers risk and 
places the most significant weight on projects to correct deficiencies 
that present direct risk to the safety of the National Airspace System 
(NAS). 

* Terminal facilities: In prioritizing projects for its terminal 
facilities, ATO categorizes requirements in eight categories (in 
decreasing order of importance): (1) waterproofing; (2) heating, 
ventilation, and air conditioning; (3) electrical; (4) elevators; (5) 
plumbing; (6) operations areas special needs; (7) exterior components; 
and (8) interior finishes. ATO then assigns a priority rating to 
projects giving highest priority to items that are degrading, where 
mitigation is difficult, and where adverse impacts regularly occur. 
Lower priority is assigned in cases where previously planned repairs 
and improvements are anticipated to address identified needs. 

* Technical Center: Technical Center bases its prioritization on the 
condition codes and importance factors assigned to individual 
components of its facilities. According to the Technical Center, each 
of the projects it has identified to date has been to repair or 
replace components that were in the poorest condition and of the 
greatest importance. 

* Aeronautical Center: The Aeronautical Center determines its project 
prioritization list based on three factors: (1) FCI of the facility, 
(2) mission criticality, and (3) a facility designation factor. The 
Aeronautical Center first identifies facilities with an FCI below 95 
percent and subsequently eliminates facilities that are planned for 
demolition or whose FCI is low based on deferred maintenance for non-
critical building components. The remaining facilities are then ranked 
using the following two factors, FAA mission criticality[Footnote 52] 
and FAA facility designation.[Footnote 53] Once a facility receives an 
overall ranking, the Center does a more in-depth study on the 
facility's actual condition and a business case analysis to determine 
detailed requirements. 

In reference to operations facilities, as indicated previously, ATO's 
Facilities Group Manager is taking steps to develop a standard process 
for prioritizing projects across all facility types (en-route centers, 
terminal facilities, and unstaffed facilities). This single process is 
to replace the separate prioritization processes for en-route centers 
and terminal facilities described above. The new process considers 
operations, risk, employee safety, mission-critical needs, and 
environmental requirements as prioritization criteria. According to 
FAA officials, the new project prioritization process will be fully 
implemented by the fiscal year 2015 budget cycle. 

Strategic Linkage and Long-term Capital Plan: 

Leading practices stress the importance of linking capital asset 
investments to an organization's overall mission and long-term 
strategic goals and the development of a long-term capital investment 
plan to guide the implementation of those organizational goals and 
objectives and to help decision makers establish priorities over time. 
FAA has capital asset investment-planning efforts under way related to 
the consolidation of operations facilities and the transition to 
NextGen, key strategic initiatives of the agency. These planning 
efforts include: 

* En-route center and TRACON consolidation: In November 2011, FAA 
approved an initial plan to consolidate en-route centers and TRACONs 
into six large, integrated facilities over the next two decades 
(through 2034). This long-term plan realigns and consolidates 
facilities based on operations, airspace responsibility, and 
geographic location. As we have previously concluded, reconfiguring 
facilities that handle air traffic control will be required to fully 
realize NextGen's capabilities, reduce operating costs in the long 
term, and address sustainability issues with FAA's current facility 
footprint.[Footnote 54] However, according to FAA officials, the 
implementation of this plan has been delayed due to significant 
funding reductions. 

* Terminal facility consolidation: FAA has been developing a national 
facilities consolidation report in response to the FAA Modernization 
and Reform Act of 2012. The act requires FAA to make recommendations 
for realigning and consolidating FAA services and facilities and 
provide justification, projected costs and savings, and proposed 
timing for implementing each recommendation.[Footnote 55] FAA 
officials told us that the agency chartered a collaborative workgroup 
with labor unions to develop a process to address the act's 
requirements. Their efforts will focus on TRACON realignments and 
consolidation. In July 2013, FAA officials noted that they had 
developed a process for identifying potential facilities for 
consolidation and realignment and were evaluating realignment options 
as required by the act, but had not yet identified which facilities 
would be consolidated or realigned or a time frame for developing such 
a list.[Footnote 56] 

* En-route and Terminal facility "Get Well" plan: ATO's Facilities 
Group Manager is leading development of a 10-year plan that, among 
other things, establishes an approach to reduce the maintenance 
backlog at its existing facilities. According to data provided by the 
Facilities Group Manager in May 2013, ATO's current overall deferred 
maintenance backlog for staffed operations facilities is $258.8 
million.[Footnote 57] The plan will be based on meeting a target FCI 
of 95 percent for all facilities and establish compliance with safety 
requirements as a primary focus. ATO expects to complete this plan by 
September 2013. 

These planning efforts are important to FAA's ability to balance 
facility sustainment needs with NextGen requirements to ensure proper 
prioritization of projects. FAA's Capital Investment Plan (CIP) 
provides insight into the agency's long-term strategies for meeting 
these requirements. Leading practices emphasize the importance of 
developing a long-term capital investment plan to guide the 
implementation of organizational goals and objectives and to help 
decision makers establish priorities over time. Among the information 
provided in FAA's CIP is projected funding for budget line items (BLI) 
for facility improvements and replacement as informed by the outcomes 
of the project prioritization processes discussed above. For example, 
table 9 illustrates BLIs specific to funding en-route-center and 
terminal-facility improvement and replacement projects. This shows 
that FAA's fiscal year 2014 through 2017 forecasts--with the exception 
of the terminal replacement budget line--are generally greater than 
actual past funding levels and significantly greater than levels in 
fiscal year 2013. ATO officials attributed increases in these budget 
lines in part to the better understanding of costs to upkeep 
facilities gained from its facility condition assessments. In 
reference to en-route centers, ATO officials noted another reason for 
expected increases in future budgets is a result of the aging of 
facilities that last received an influx of funds for modernization 
projects in the early 1990s. Overall, ATO officials expect these 
forecast funding levels to result in a decrease in deferred 
maintenance and help them achieve their 95 percent FCI target for 
facilities. 

Table 9: FAA's Actual and Projected Capital Improvement Budget Line 
Items, Fiscal Years 2008 through 2017: 

Budget line item: En-route center improvements; 
Actual: 
2008: $50.0 million; 
2009: $55.5 million; 
2010: $48.7 million; 
2011: $35.4 million; 
2012: $39 million; 
2013: $38.5 million; 
Projected: 
2014: $50.0 million; 
2015: $61.1 million; 
2016: $58.0 million; 
2017: $60.0 million. 

Budget line item: Terminal facility replacement; 
Actual: 
2008: $162.6 million; 
2009: $136.5 million; 
2010: $171.2 million; 
2011: $71.6 million; 
2012: $51.6 million; 
2013: $64.9 million; 
Projected: 
2014: $72.0 million; 
2015: $96.0 million; 
2016: $100.0 million; 
2017: $110.0 million. 

Budget line item: Terminal facility improvements; 
Actual: 
2008: $40.4 million; 
2009: $36.3 million; 
2010: $37.3 million; 
2011: $43.8 million; 
2012: $49.3 million; 
2013: $18.7 million; 
Projected: 
2014: $50.0 million; 
2015: $50.2 million; 
2016: $50.0 million; 
2017: $50.0 million. 

Source: FAA CIP data for fiscal years 2008 through 2017. 

[End of table] 

However, budget uncertainty may limit ATO's ability to fund facilities 
projects as forecasted as well as limit efforts to implement facility 
consolidation and NextGen transition plans. The Balanced Budget and 
Emergency Deficit Control Act of 1985, as amended,[Footnote 58] places 
limits on discretionary spending through 2021. ATO officials indicated 
that in preparing the update to the CIP, they recognize that future 
funding levels may be less than forecast in the current plan. As a 
result, they are developing a two-tier system to prioritize projects. 
For example, instead of maintaining all facilities at a 95 percent 
FCI, they would focus on maintaining only facilities at the 56 busiest 
airports at this level and facilities at all other airports at a 90 
percent FCI. 

Conclusions: 

FAA staff occupy an array of facility types including air traffic 
control towers, administrative offices, and research laboratories in 
executing the agency's mission to safely operate the NAS. The work 
environment defined by the condition of these facilities affects the 
safety and health of FAA employees. Inspecting facilities to gain 
information on their condition is important to FAA's ability to 
identify repair and modernization requirements and direct resources to 
address them. However, FAA's estimates of the condition of most 
terminal facilities--which represent a large segment of all FAA 
facilities--are inaccurate and lead to uncertainty in defining 
maintenance and repair needs. In addition, inaccuracies in FAA's 
agency-wide REMS data source and inconsistent data collection 
practices contribute to causing uncertainty about the composition and 
condition of its facility portfolio. Accurate and reliable information 
on facility condition is important not only to providing a safe and 
healthy working environment but also to prioritizing long-term needs 
for providing a physical infrastructure that enables FAA to 
effectively meet its mission while working to consolidate facilities 
and transition to NextGen. Having accurate and reliable information on 
facility conditions and associated maintenance and repair needs will 
help FAA optimize use of its resources, which are likely to be 
constrained in future years as a result of budget controls, and align 
its strategic and capital plans to reflect these constraints. 

Recommendations: 

To ensure that FAA has accurate and reliable information on the 
condition of its facilities and allow for more informed decision 
making on their maintenance and repairs and associated capital-
planning efforts, we recommend that the Secretary of Transportation 
direct the FAA Administrator to take the following two actions: 

* Improve the precision of the estimation methods used by ATO to 
determine conditions at terminal facilities that have not been 
inspected and assess and report the error associated with estimates of 
the terminal facilities' condition and deferred maintenance. 

* Develop and implement a plan to comprehensively assess and improve 
REMS, consistent with sound data collection practices, to ensure that 
the data are sufficiently complete, accurate, and synchronized with 
other real property data maintained by FAA organizational components. 

Agency Comments: 

We provided the Departments of Labor and Transportation and the 
General Services Administration with a draft of this report for review 
and comment. The Department of Transportation provided technical 
comments, which we incorporated as appropriate. Neither the Department 
of Labor nor the General Services Administration had comments on the 
draft report. 

We are sending copies of this report to relevant congressional 
committees, the Secretary of Transportation, and other interested 
parties. In addition, this report will also be available at no charge 
on GAO's website at [hyperlink, http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact me at (202) 512-2834 or dillinghamg@gao.gov. 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 key 
contributions to this report are listed in appendix III. 

Signed by: 

Gerald L. Dillingham, Ph.D. 
Director, Physical Infrastructure: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

Section 610 of the Federal Aviation Administration (FAA) Modernization 
and Reform Act of 2012 mandated GAO to study the condition of FAA-
staffed facilities.[Footnote 59] In response to this mandate, we 
addressed the following objectives: 

1. What is known about the conditions of FAA-staffed facilities and to 
what extent are the assessment methods and resulting data reliable? 

2. What have been the workplace injuries and illnesses reported by FAA 
employees in workers' compensation claims, and to what extent has FAA 
met its targets for processing claims? 

3. To what extent has FAA responded to safety deficiencies identified 
through its inspections and those conducted by the Occupational Safety 
and Health Administration (OSHA)? 

4. To what extent are actions FAA has taken to ensure that its 
facilities are in good condition, if any, consistent with leading 
practices for capital-investment decisions? 

To determine what is known about the conditions of FAA-staffed 
facilities and the extent to which the methods used to make those 
determinations and the resulting data are reliable, we focused on 
staffed facilities that FAA or the General Services Administration 
(GSA) maintains.[Footnote 60] We collected facility condition data 
from 2010 through 2012, including Facility Condition Indices (FCI) and 
deferred maintenance values for facilities, and compared the methods 
used by FAA and GSA in inspecting their facilities with generally 
accepted industry standards. Within FAA, we found two sources for 
facility condition data: the agency-wide Real Estate Management System 
(REMS) data and data from other FAA organizational components 
responsible for maintaining facilities. Early on, we found problems 
with REMS data and after discussions with REMS administrators and 
programmatic officials, we decided to use facility condition data from 
the FAA organizational components that are responsible for the 
facilities. Thus, we obtained facility data for en-route and terminal 
facilities from the Air Traffic Organization. We also obtained data 
from the Mike Monroney Aeronautical Center (Aeronautical Center) and 
the William J. Hughes Technical Center (Technical Center) for their 
facilities. For FAA's administrative facilities, we obtained condition 
data from the General Services Administration (GSA). The data we 
obtained represent different time frames (i.e., calendar year versus 
fiscal year). However, we used them because they were the only 
reliable data available. Further, the different sources of facility 
condition are based on differing methods. Therefore, in our analysis, 
we cannot compare the facility condition of one type of facility to 
other facility types. While we obtained the number of facilities 
leased by FAA and GSA, we did not obtain data on their condition 
because they did not exist or we determined that they were not 
reliable. 

To assess the reliability of the FAA and GSA data we reviewed existing 
documentation about the data and interviewed agency officials and 
technical-consulting firm officials knowledgeable about the data. We 
performed electronic testing of the FAA data for obvious errors in 
accuracy and completeness. We found the data for FAA's en-route 
facilities, inspected terminal facilities, Aeronautical Center 
facilities, and GSA-maintained FAA facilities to be sufficiently 
reliable for describing facility condition. FAA's Technical Center's 
inspection process does not yield facility-level condition data, such 
as FCI. Rather, the process assesses the conditions of the components-
-such as structural, electrical, and heating/cooling systems--that 
comprise each facility. Because available data for Technical Center 
facilities does not reflect facility-level conditions similar to what 
we reported for other facility types, we do not present them in this 
report. 

We did not find the condition data for FAA's uninspected terminal 
facilities to be reliable and conducted further analysis to articulate 
our concerns with the reliability and accuracy; our findings are 
presented in greater detail in appendix II. 

We also interviewed FAA and union officials representing major 
aviation employee groups about facility conditions. In addition, to 
observe conditions and to speak with knowledgeable staff to learn 
about any facility deficiencies and the projects for fixing them, we 
selected a non-probability sample of facilities supporting operations 
and administrative functions in each of FAA's three geographic service 
areas (see table 10). We also spoke with officials of the Technical 
Center by phone and obtained documents from them. 

Table 10: FAA Facilities GAO Contacted: 

Eastern Service Area: 
Eastern Service Area: FAA Headquarters; 
Eastern Service Area: William J. Hughes Technical Center; 
Eastern Service Area: Washington, D.C., En-route Center. 

Central Service Area: 
Mike Monroney Aeronautical Center; 
Albuquerque Flight Standards District Office (FSDO); 
Albuquerque En-route Center; 
Albuquerque Terminal Radar and Approach Control (TRACON) and Tower; 
Oklahoma City TRACON and Tower. 

Western Service Area: 
Long Beach FSDO; 
Western Pacific Region Office; 
Los Angeles En-route Center; 
Los Angeles Tower; 
Fullerton Tower (contract tower); 
Southern California TRACON. 

Source: GAO analysis. 

[End of table] 

To determine the most common types of workplace injuries and illnesses 
FAA employees have reported, if any, we reviewed FAA and Department of 
Labor (Labor) workers' compensation data from 2007 through 2012. 
[Footnote 61] To assess the reliability of the data we (1) reviewed 
existing documentation about the data, (2) interviewed agency 
officials knowledgeable about the data, and (3) performed electronic 
testing of the data for obvious errors in accuracy and completeness. 
We found the data to be sufficiently reliable for the purposes of this 
report. Because of the variability of factors contributing to causing 
reported injuries and illnesses, our study does not attempt to link 
workers compensation claims to facility conditions. 

To determine the extent to which FAA responded to safety deficiencies 
identified through its inspections and those conducted by the 
Occupational Safety and Health Administration (OSHA), we interviewed 
FAA and OSHA officials as well as reviewed FAA and OSHA documents 
regarding this issue. 

To determine the extent actions taken by FAA to ensure that its 
facilities are in good condition are consistent with leading practices 
for capital-investment decisions, we reviewed FAA's Capital Investment 
Plan's budget data (both actual and projected) for facility 
modernization and replacement, from 2008 through 2017. We also 
reviewed FAA's process for making capital-investment decisions and 
compared it to GAO's and Office of Management and Budget's leading 
practices. We also interviewed FAA officials on their reorganization 
efforts and reviewed documents on FAA's facility project 
prioritization schemes. 

We conducted this performance audit from July 2012 to September 2013 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

[End of section] 

Appendix II: Assessment of FAA's Statistical Model to Estimate the 
Condition of Terminal Facilities: 

FAA estimates the condition of many terminal facilities using a 
statistical model developed by a technical-consulting firm. The model 
estimates the condition of facilities that have not been physically 
inspected in the past 6 years, using data on the age and condition of 
facilities that have been inspected within that time period. The 
contractor has recommended that FAA revise its statistical models. In 
the 2012 roll-up report, the contractor notes that the baseline model 
is "preliminary and additional samples and adjustments to the formulas 
are needed for the results to be statistically valid." In this 
appendix, we describe our replication of the FAA model, using data 
from the agency's roll-up reports on facility condition, and assess 
the model's statistical precision. In addition, we discuss several 
alternative models we considered and their precision relative to FAA's 
baseline model. 

FAA Model Estimates Terminal Facility Condition with Substantial 
Imprecision: 

FAA uses a combination of direct inspection, assumptions about the 
decline of a facility's condition within 6 years of inspection, and 
statistical methods to estimate the condition of terminal facilities. 
For facilities that have been inspected in a given year, FAA uses the 
results of those inspections to estimate facility condition in that 
year. Over the next 5 years, FAA plans to update the condition of the 
facility from the baseline of its direct inspection, using a formula 
that incorporates assumptions about expected deterioration. After 6 
years have elapsed since the inspection, FAA uses statistical methods 
to estimate facility condition. 

The specific statistical method that FAA uses for these less recently 
inspected facilities is a regression model that predicts facility 
condition--either the Facility Condition Index (FCI)[Footnote 62] or 
deferred maintenance measured in dollars--as a function of facility 
age. Specifically, the model assumes that the unknown condition of a 
facility i is: 

[Refer to PDF for formula] 

with the intercept fixed at 100 in the model of FCI and 0 in the model 
of deferred maintenance. FAA's contractor fits the model to data on 
condition and facility age for facilities inspected within the past 6 
years. These inspection data estimate and, thus, the unknown condition 
of facilities that were not inspected recently. In the 2012 roll-up 
report, FAA reports that B = -0.203 for the model of FCI and B = 
$12,668 for the model of deferred maintenance. Since the model 
requires FCI and deferred maintenance to equal zero when age equals to 
zero, the model predicts a facility's condition using its age alone. 
For example, the 2012 model would predict that all 10 year-old 
facilities have a FCI of 98.0 and deferred maintenance of $126,680, 
regardless of any other characteristics that could vary across the 
facilities. 

All statistical estimates, including FAA's predictions of facility 
condition, have estimation error associated with them. In this 
application, we are specifically interested in the prediction error of 
FAA's model--that is, the amount by which FAA's condition estimates 
for uninspected facilities can be expected to vary from the true 
condition that would be measured in new physical inspections. We can 
model this prediction error for each facility as: 

[Refer to PDF for formula] 

which implies that: 

[Refer to PDF for formula] 

Statistical models of the type that FAA uses to estimate facility 
condition typically assume that the error of prediction, ei, varies 
randomly, such that E(ei) = 0 and Var(ei) = ó2. If one applies this 
assumption to FAA's model, the expected amount by which a new physical 
inspection of a facility with a fixed age would vary from FAA's 
estimate is given by: 

[Refer to PDF for formula] [Footnote 63] 

The expected error of a new inspection, given the FAA model's 
prediction, is equal to the variance of the predicted value, which 
depends on the size of the sample used to estimate B and facility age, 
plus a term o2 measuring the variance of y around its mean for the 
population of facilities at a fixed age.[Footnote 64] 

In order to assess the precision of the FAA model estimates, we 
assembled data on facility age and the results of physical inspections 
from the 2010 through 2012 roll-up reports. For the 2010 sample, we 
assumed that all inspections which FAA defined as "actual" provided 
new data for analysis, even though many facilities were not physically 
inspected in 2010 but, instead, had values that were updated from an 
inspection conducted within the prior 6 years. This assumption was 
reasonable for our analysis, because 2010 was the first year in which 
the inspections were available to analyze. For the 2011-2012 sample, 
we treated inspections as providing new data only if they were 
conducted in those years. FAA considered inspections within 6 years of 
2011 or 2012 to have provided "actual" data on condition for the 
purpose of model estimation, but in practice, these data were 
mathematically adjusted from prior years. From a statistical 
perspective, these data are not new observations, because FAA only 
measured condition once, in the year the inspection occurred.[Footnote 
65] 

We fit the FAA model to the 2010 sample of 114 facilities in order to 
replicate the estimates from the roll-up reports and assess their 
precision. The 2010 sample allowed us to assess the fit of the model 
on the same data used to estimate its parameters, also known as the 
"in-sample" fit. Using the model fitted to the 2010 sample, we 
predicted the condition of 33 facilities that were physically 
inspected in either the 2011 or 2012 samples. By comparing the 
condition predicted by the model to the actual condition measured in 
physical inspections, we estimated the model's "out-of-sample" 
predictive power for new, unknown data. Evaluating a model's ability 
to predict beyond the sample of data used to develop it helps avoid 
placing too much emphasis on random features of any one sample that 
can influence the model's estimates. Moreover, the purpose of FAA's 
model is to predict unknown facility conditions. 

Table 11 shows the results of our analysis, and figure 3 graphically 
depicts the 2010 data and model fit. We summarized the predictive 
accuracy of the FAA model using various statistics, in part due to the 
skewed distribution of FCI and deferred maintenance. Our estimated 
values of B-- -0.227 on the FCI scale and $14,565 on the deferred 
maintenance scale--were similar to the values FAA reported in its 2010 
roll-up report (-0.225 and $12,551). However, our replication 
confirmed the contractor's assessment that the FAA model does not 
predict facility condition with a high degree of precision. On the FCI 
scale, the root mean squared error (RMSE) of prediction--which is 
similar to the estimated variance of prediction error also reported in 
table 11---was 3.4 in the 2010 sample used to estimate the model and 
3.9 in the 2011-2012 validation sample. On the deferred maintenance 
scale, the RMSE was $606,862 in the estimation sample and $370,875 in 
the validation sample. The sample variance of FCI and deferred 
maintenance helps put the size of these results into context. In the 
estimation sample, the standard deviation was 5.0 and the middle 50 
percent of the FCI distribution ranged from 90.9 to 96.6. The middle 
50 percent of the deferred maintenance distribution ranged from 
$231,420 to $663,944, with a standard deviation of $565,421. These 
statistics are consistent with the large confidence intervals of 
prediction in figure 3. For example, the deferred-maintenance 
predictions for 10 and 40 year-old facilities have 95 percent 
confidence intervals ranging from 0 to $1,353,723 and from 0 to 
$1,797,599, respectively.[Footnote 66] Together, these results show 
that FAA's estimates of facility condition vary meaningfully from the 
results of physical inspections. 

Table 11: Results of Replication and Predictive Validation of FAA 
Model: 

Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 
2011-2012 Sample (prediction only): 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 
2011-2012 Sample (prediction only): 

Parameters: ß; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): -0.227; 
2011-2012 Sample (prediction only): NA; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 14,565; 
2011-2012 Sample (prediction only): NA. 

Parameters: ó; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 3.392; 
2011-2012 Sample (prediction only): NA; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 609,500; 
2011-2012 Sample (prediction only): NA. 

Fit statistics: 

Root mean squared error; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 3.377; 
2011-2012 Sample (prediction only): 3.869; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 606,862; 
2011-2012 Sample (prediction only): 370,875. 

Root median squared error; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 1.524; 
2011-2012 Sample (prediction only): 1.823; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 217,415; 
2011-2012 Sample (prediction only): 221,592. 

Mean absolute error; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 2.371; 
2011-2012 Sample (prediction only): 2.690; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 343,660; 
2011-2012 Sample (prediction only): 280,211. 

Median absolute error; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 1.524; 
2011-2012 Sample (prediction only): 1.823; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 217,400; 
2011-2012 Sample (prediction only): 221,592. 

Total error; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): NA; 
2011-2012 Sample (prediction only): NA; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 13,255,775; 
2011-2012 Sample (prediction only): -2,417,559. 

Total absolute error; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): NA; 
2011-2012 Sample (prediction only): NA; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 39,177,277; 
2011-2012 Sample (prediction only): 9,246,953. 

N; 
Facility Condition Index (0-100): 
2010 Sample (estimation and prediction): 114; 
2011-2012 Sample (prediction only): 33; 
Deferred Maintenance: 
2010 Sample (estimation and prediction): 114; 
2011-2012 Sample (prediction only): 33. 

Source: GAO analysis of FAA roll-up report data. 

Note: Entries in the "2010 sample" columns are estimates from a FAA 
statistical model of facility condition fit to data from the agency's 
2010 roll-up report on this topic. We used the estimated model to 
predict the condition of facilities that were physically inspected in 
2011 or 2012. Entries in the "2011-2012" sample columns are statistics 
summarizing the predictive accuracy of the 2010 model on these later 
data. 

[End of table] 

Figure 3: Estimated versus Actual Deferred Maintenance in 2010: 

[Refer to PDF for image: plotted point graph] 

Graph depicts the following: 

GAO predicted deferred maintenance (DM), using FY 2010 model fit; 
Actual DM, FY 2010; 
95% prediction interval; 
Predicted DM - age alone. 

Source: GAO analysis of FAA terminal facility data. 

[End of figure] 

Alternative Methods Can Improve the Precision of Some Facility 
Condition Estimates: 

Given the imprecision of FAA's model, we analyzed data available from 
the agency's roll-up reports to assess whether alternative methods 
could predict terminal facility condition more accurately. We measured 
the following variables for the same facilities and time periods 
described above: facility size (square footage); geographic location, 
including the facility's latitude and longitude; architectural design 
type; class of facility by function, such as whether a facility was a 
"tower with radar" or "combined TRACON and tower;" and whether FAA 
maintained and/or owned the facility (respectively). To measure 
facility value, we calculated the value implied by FAA's data on FCI 
and deferred maintenance. 

We used several approaches to systematically search for alternative 
prediction methods, collectively known as "data mining" techniques. As 
with our evaluation of the baseline FAA model, we used the 2010 sample 
for estimating and comparing models and the 2011-2012 sample for 
validating the predictions on new data. We would have preferred to 
divide the available data into three parts to estimate, refine, and 
test alternative models, as recommended by methodological guidance on 
data mining. The small sample size precluded this approach, however. 

Our first approach involved estimating a series of linear regression 
models that added various combinations of the variables above as 
predictors to the baseline FAA model. This "stepwise" search produced 
a model that minimized the Akaike Information Criterion (AIC), which 
is closely related to the residual sum of squares and R-squared 
statistics for the type of linear Normal models used by FAA. This 
procedure did not identify a more predictive model than the baseline 
FAA model, but it did identify a more predictive model of deferred 
maintenance: 

[Refer to PDF for formula] 

(GAO Model D1): 

Our second approach used recursive partitioning or "tree" methods to 
predict facility condition. This approach placed facilities into 
groups defined by the covariates, and then estimated condition as the 
mean within each group. Various classifications of facilities into 
covariate groups were searched to minimize the Akaike Information 
Criterion, which implicitly considered interactions and non-
linearities among the covariates. In this sense, a recursive 
partitioning search complemented our application of stepwise methods, 
which assumed that all variables had linear relationships and no 
interactions (except for latitude and longitude). The clustering, non-
linearity, and skewness in the distribution of facility condition (see 
figure 3) makes these assumptions somewhat unrealistic. Fitting a 
constant value for each group may be more appropriate, given these 
features of the data. Partitioning methods require "pruning," a 
process in which covariate groups of various sizes are searched to 
minimize the bias (average error) of the predictions while controlling 
their variance. We chose the pruning parameter that minimized the mean 
prediction error over 10 cross-validated samples of the 2010 
data.[Footnote 67] The final tree implied the following piecewise 
constant regression models for FCI and deferred maintenance: 

[Refer to PDF for formula] 

where I(.) is the indicator function and TRACONi equals 1 if the 
architectural design was a "Large TRACON" and 0 otherwise. Note that 
the intercept estimates the mean for large TRACONs. Fitting separate 
values according to whether the facility is a large TRACON is 
consistent with the data in figure 3, which shows that this class of 
facility often produces large outliers with respect to the mean value 
conditional on age and with respect to the marginal distribution of 
facility value.[Footnote 68] In addition, constant fits may better 
estimate the condition of very young facilities, which cluster toward 
a FCI of 100 and deferred maintenance of $0. 

In addition to searching broadly over the available data to improve 
prediction, we developed a model that simply adjusted the FAA baseline 
approach to predicting deferred maintenance by incorporating the 
facility's value (calculated as the cost of facility replacement): 

[Refer to PDF for formula] 

(GAO Model D3): 

This model makes the effects of age and value conditional on the other 
variable, so that deferred maintenance can vary across facilities of 
the same age. In contrast, the baseline FAA model assumes that all 
facilities of the same age have the same deferred maintenance, which 
ignores the fact that facilities of the same age vary widely in value 
and, thus, deferred maintenance. 

As shown in table 12, our alternative models predicted deferred 
maintenance more precisely than FAA's baseline model. On the deferred 
maintenance scale, our models improved the RMSE of deferred 
maintenance by at least $234,385 (39 percent) in the 2010 estimation 
sample and by $118,065 (32 percent) in the 2011-2012 validation 
sample. When assessed on other fit statistics, our models of deferred 
maintenance typically outperform FAA's baseline. The difference is 
largely due to the fact that our models incorporate facility value 
when predicting deferred maintenance, which should explain a large 
proportion of the variation across facilities of varying size. In 
contrast, we were unable to substantially improve FAA's model of FCI. 
The RMSE of our tree-selected model was 8.6 percent smaller than that 
of the FAA baseline in the 2010 sample, but the fit statistics did not 
consistently favor either model in the new 2011-2012 sample. FAA might 
prefer one model or another on these measures of fit, according to the 
goals of prediction. For example, if FAA sought to minimize to the 
error of the estimated total deferred maintenance, the total error or 
total absolute error would be appropriate measures of prediction 
accuracy. 

Table 12: Results of Replication and Predictive Validation of FAA 
Model: 

FCI (0-100): 

2010 Sample: 

FAA baseline; 
Root mean squared error: 3.377; 
Root median squared error: 1.524; 
Mean absolute error: 2.371; 
Median absolute error: 1.524; 
Total error: NA; 
Total absolute error: NA. 

GAO model F1; 
Root mean squared error: 3.088; 
Root median squared error: 1.519; 
Mean absolute error: 2.177; 
Median absolute error: 1.519; 
Total error: NA; 
Total absolute error: NA. 

2011-2012 Sample: 

FAA baseline; 
Root mean squared error: 3.869; 
Root median squared error: 1.823; 
Mean absolute error: 2.690; 
Median absolute error: 1.823; 
Total error: NA; 
Total absolute error: NA. 

GAO model F1; 
Root mean squared error: 4.221; 
Root median squared error: 1.663; 
Mean absolute error: 3.143; 
Median absolute error: 1.663; 
Total error: NA; 
Total absolute error: NA. 

Deferred Maintenance: 

2010 Sample: 

FAA model; 
Root mean squared error: $606,862; 
Root median squared error: $217,415; 
Mean absolute error: $343,660; 
Median absolute error: $217,400; 
Total error: $13,255,775; 
Total absolute error: $39,177,277. 

GAO model D1; 
Root mean squared error: $288,431; 
Root median squared error: $138,714; 
Mean absolute error: $195,861; 
Median absolute error: $138,713; 
Total error: $0; 
Total absolute error: $21,152,937. 

GAO model D2; 
Root mean squared error: $303,155; 
Root median squared error: $99,441; 
Mean absolute error: $175,479; 
Median absolute error: $99,438; 
Total error: $0; 
Total absolute error: $20,004,600. 

GAO model D3; 
Root mean squared error: $372,477; 
Root median squared error: $156,211; 
Mean absolute error: $238,010; 
Median absolute error: $156,205; 
Total error: -$1,106,601; 
Total absolute error: $25,705,067. 

2011-2012 Sample: 

FAA model; 
Root mean squared error: $370,875; 
Root median squared error: $221,592; 
Mean absolute error: $280,211; 
Median absolute error: $221,592; 
Total error: -$2,417,559; 
Total absolute error: $9,246,953. 

GAO model D1; 
Root mean squared error: $294,992; 
Root median squared error: $244,965; 
Mean absolute error: $232,250; 
Median absolute error: $244,965; 
Total error: -$1,802,472; 
Total absolute error: $6,270,737. 

GAO model D2; 
Root mean squared error: $263,292; 
Root median squared error: $126,139; 
Mean absolute error: $187,338; 
Median absolute error: $126,102; 
Total error: -$608,089; 
Total absolute error: $5,994,801. 

GAO model D3; 
Root mean squared error: $252,810; 
Root median squared error: $157,797; 
Mean absolute error: $187,201; 
Median absolute error: $157,797; 
Total error: -$810,940; 
Total absolute error: $5,054,426. 

Source: GAO analysis of FAA roll-up report data. 

Note: Entries are fit statistics of statistical models developed on a 
sample of 114 facilities from the 2010 FAA roll-up report. Results for 
the 2010 sample are the in-sample fit statistics, while results for 
the 2011-2012 sample are out-of-sample fit statistics. 

[End of table] 

Implications for Revising Prediction Models: 

Our analysis suggests that data available to FAA would allow the 
agency to substantially improve its model of deferred maintenance. 
Specifically, incorporating facility value along with age and perhaps 
facility class and architectural design type would better exploit the 
available data. Additional revisions also may be appropriate, 
depending on the available data. FAA could refine the mathematical 
form of the model over time, as new inspections are conducted to 
provide additional data for model validation. 

Although we were unable to improve upon FAA's model of FCI, the best 
available model still does not predict FCI with a high degree of 
precision. Additional data beyond the variables available to us in the 
roll-up reports may help identify a more predictive model. For 
example, data on typical weather conditions, usage, or building 
materials may improve predictions of FCI beyond what is possible using 
the variables we considered here and FAA's consultant has considered 
previously. If these data are available and are correlated with 
condition and if their collection and analysis costs less than the 
error caused by FAA's baseline model, incorporating additional 
variables could be a cost-effective solution to estimating FCI for 
facilities that have not been physically inspected in recent years. 

Our findings are consistent with those of FAA's consultant, which has 
recommended that FAA revise its statistical models. In its 2012 roll-
up report, the consultant notes that the baseline model is 
"preliminary and additional samples and adjustments to the formulas 
are needed for the results to be statistically valid." They 
recommended addressing outliers in the distribution of facilities and 
including additional predictors of facility condition.[Footnote 69] In 
a 2008 report, the contractor recommended adjusting the physical 
inspection process to ensure a representative sample is available for 
statistical analysis.[Footnote 70] The latter recommendation could be 
implemented as an inspection plan that selects facilities according to 
known probabilities, which would ensure that estimates of facility 
condition are unbiased and have known sampling variances. The sample 
could be designed in such a way to incorporate known maintenance 
problems or inspection needs, such as by selecting facilities for 
inspection with unequal probabilities from within strata defined by 
age, size, or other important variables. A probability sample could be 
particularly useful for reporting average or total condition across 
facilities. 

[End of section] 

Appendix III: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

Gerald L. Dillingham, Ph.D., (202) 512-2834, or dillinghamg@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, GAO staff who made key 
contributions to this report are Michael Armes, Assistant Director; 
Martha Chow; Colin Fallon; James Geibel; Kathleen Gilhooly; Daniel 
Hoy; Grant Mallie; SaraAnn Moessbauer; and Jeff Tessin. 

[End of section] 

Footnotes: 

[1] NextGen is a complex undertaking that requires acquiring new 
integrated air traffic control systems; developing new flight 
procedures, standards, and regulations; and creating and maintaining 
supporting infrastructure to create a more automated aircraft-
centered, satellite-based air transportation system. 

[2] GAO, High-Risk Series: Federal Real Property, [hyperlink, 
http://www.gao.gov/products/GAO-03-122] (Washington, D.C.: January 
2003) and GAO, High-Risk Series: An Update, [hyperlink, 
http://www.gao.gov/products/GAO-13-283] (Washington, D.C.: February 
2013). Our high-risk series identifies areas at high risk because of 
their greater vulnerabilities to waste, fraud, abuse, and 
mismanagement or major challenges associated with their economy, 
efficiency, or effectiveness. We identified federal real property as a 
high-risk area because of long-standing problems with excess and 
underutilized real property, deteriorating facilities, unreliable real 
property data, and costly space challenges shared by several agencies. 

[3] FAA Modernization and Reform Act of 2012, Pub. L. No. 112-95, § 
610, 126 Stat. 11, 117 (Feb. 14, 2012). 

[4] GAO, Preliminary Results of Work on FAA Facility Conditions and 
Workplace Safety, [hyperlink, http://www.gao.gov/products/GAO-13-509R] 
(Washington, D.C.: May 9, 2013). 

[5] FAA's service areas comprise the eastern, central, and western 
regions of the country. 

[6] GAO, Executive Guide: Leading Practices in Capital Decision-
Making. [hyperlink, http://www.gao.gov/products/GAO/AIMD-99-32] 
(Washington, D.C.: December 1998); OMB, Capital Programming Guide, 
Supplement to Office of Management and Budget Circular A-11: Planning, 
Budgeting, and Acquisition of Capital Assets (July 2012). 

[7] The FAA Unstaffed Infrastructure Sustainment program manages 
approximately 30,000 unstaffed facilities that house, protect, and 
support National Airspace System (NAS) communications, navigation, 
surveillance, weather, and other air traffic control equipment. This 
study is limited to FAA-staffed facilities maintained by FAA and GSA. 
For more information on unstaffed facilities, please see GAO, National 
Airspace System: Improved Budget Process Could Help FAA Better 
Determine Future Operations and Maintenance Priorities, [hyperlink, 
http://www.gao.gov/products/GAO-13-693] (Washington, D.C.: Aug. 22, 
2013). 

[8] Examples of FAA's administrative offices include its Certificate 
Management Office (CMO), Flight Standards District Office (FSDO), and 
Manufacturing Inspection District Office (MIDO). FAA's CMO is 
responsible for the certification, surveillance, and inspection of 
major air carriers and Flight Safety International's part 142 Training 
Centers; its FSDO responsibilities include airmen certification 
(licensing) for pilots, mechanics, repairmen, dispatchers, and 
parachute riggers; and the responsibilities of its MIDOs include 
airworthiness certification and the oversight of Manufacturing 
Designees. 

[9] As a formula, FCI is the value of the maintenance, repair, and 
replacement deficiencies of a facility divided by its current 
replacement value. The resulting fraction is then subtracted from 1 to 
express FCI as a percentage. Facilities can be grouped according to 
FCI values. ATO, for example, defines a facility with an FCI above 95 
percent as in "good" condition, 90 to 95 percent to be in "fair" 
condition, and below 90 percent to be in "poor" condition. 

[10] For its en-route and terminal facilities, FAA defines deferred 
maintenance as the cost of rebuilding or replacing components whose 
service life has exceeded their scheduled lifetime as of the forecast 
year. FAA does not include preventative maintenance and minor repairs 
in its deferred maintenance calculations. 

[11] FRPP is the inventory system for the federal real property 
portfolio. FRPP, which is overseen by OMB, includes data elements that 
agencies are mandated to report annually, including performance 
measures on asset utilization, condition, mission dependency, and 
operating cost. 

[12] Two budget accounts provide the principal source of funding for 
FAA facilities. The Facilities and Equipment account funds major 
modernization and improvement projects while the Operations account 
funds routine maintenance and repairs. 

[13] OSHA's programmed inspections are scheduled investigations that 
target specific worksite hazards or high-hazard industries and 
worksites. OSHA's unprogrammed inspections are conducted in response 
to imminent danger, fatalities or catastrophes, complaints, referrals, 
and as follow-up and monitoring of prior inspections. 

[14] For claims it approves, OWCP bases the benefits on an employee's 
wages at time of injury, the date the disability began or the date the 
disability recurred, his or her ability to work after the injury, and 
whether he or she has eligible dependents. OWCP charges agencies for 
the benefits provided to their injured employees. These agencies 
subsequently reimburse Labor's Employees' Compensation Fund from their 
next annual appropriation. Workers' compensation benefits paid to 
beneficiaries are adjusted for inflation and are not taxed or subject 
to age restrictions. 

[15] See 29 U.S.C. § 668, 29 C.F.R. pt. 1960, and Exec. Order No. 
12196, Occupational Safety and Health Programs for Federal Employees, 
45 Fed. Reg. 12769 (Feb. 27, 1980). 

[16] See footnote 9. 

[17] As discussed below, Technical Center managers use a process that 
assesses the conditions of a facility's components such as structural, 
electrical, and mechanical systems. This process does not yield an FCI 
for each facility but rather provides component-specific measures of 
condition. 

[18] FAA's consultant inspects each en-route facility approximately 
once every 4 to 5 years, and "roll-up" reports for these facilities 
contain actual inspection results for each of the 23 facilities. For 
terminal facilities, roll-up reports contain actual inspection results 
from facilities inspected over the most recent 6-year period. FAA's 
most recent terminal facility roll-up report, for example, contains 
information from 128 facility sites inspected from January 2007 
through December 2012. 

[19] FAA's consultant develops a statistical model based on data 
obtained from the inspected facilities to estimate the condition of 
facilities that have not been inspected. The model uses age to 
estimate the FCI and deferred maintenance of each facility not 
inspected. 

[20] The inspection covered 34 of the 43 federally maintained staffed 
facilities at the Technical Center. Technical Center managers 
indicated that they intend to award a new contract by early fiscal 
year 2014 to inspect the 9 facilities not previously inspected in 2014. 

[21] See footnote 9. 

[22] If seismic codes and standards deficiencies are included, the 
aggregated FCI for the 128 sites is 87.9 percent, or "poor" condition. 

[23] This includes both staffed and unstaffed facilities. According to 
Aeronautical Center officials, unstaffed facilities include guard 
sheds, shelters for NAS equipment, and other services such as Digital 
Remote Switching. Further, this deferred-maintenance estimate does not 
include costs for design services, demolition and restoration of 
architectural finishes, staging, moving services, and other expenses 
related to completion of the associated maintenance and repair work. 

[24] These facilities include the Orville Wright Building (Washington, 
D.C.), Wilbur Wright Building (Washington, D.C.), Hawthorne Federal 
Building (Hawthorne, CA), Senator Paul Simon Federal Building 
(Carbondale, IL), Prince J. Kuhio Federal Building and U.S. Courthouse 
(Honolulu, HI), and the Anchorage Federal Building and U.S. Courthouse 
& Annex (Anchorage, AK). FAA may be the sole tenant or part of a multi-
tenant occupancy of these facilities. 

[25] FCI information may not reflect current conditions. For example, 
in January 2013, we visited the Hawthorne Federal Building, which is 
located in FAA's Western-Pacific Region, and saw that the roof had 
been replaced, which should result in a higher FCI for the facility 
when it is next assessed. 

[26] Appendix I describes in further detail the steps taken to 
determine the reliability of the data. 

[27] GSA views an FCI of 90 percent and above as "good" condition, 
between 70 and 90 percent as "fair" condition, and less than 70 
percent as "poor" condition. 

[28] See footnote 11. 

[29] While FAA is required to maintain information on leased 
properties in REMS, FCIs are not one of the required data elements for 
these facilities. 

[30] According to FAA officials, the REMS model for estimating FCI 
incorporates the DOD Facilities Pricing Guide in part because FAA has 
comparable facilities to DOD, such as air traffic control towers. 

[31] In response to our questions about REMS data, FAA sent us revised 
REMS data sets on at least three different occasions. 

[32] GAO, Standards for Internal Control in the Federal Government, 
[hyperlink, http://www.gao.gov/products/GAO/AIMD-00-21.3.1] 
(Washington, D.C.: November 1999). 

[33] Office of Management and Budget, "Guidelines for Ensuring and 
Maximizing the Quality, Objectivity, Utility, and Integrity of 
Information Disseminated by Federal Agencies." 67 Fed. Reg. 8452 (Feb. 
22, 2002). Those guidelines state that, among other things, agencies 
"shall adopt a basic standard of quality (including objectivity, 
utility, and integrity) as a performance goal and should take 
appropriate steps to incorporate information quality criteria into 
agency information dissemination practices." 

[34] See footnote 6. 

[35] See footnote 18. 

[36] The cab is the elevated component of the air traffic control 
tower where controllers direct ground traffic, takeoffs, and landings. 

[37] According to officials at the tower, the elevator will undergo 
repairs in September 2013, partially funded by resources diverted from 
a repair project at another facility. 

[38] While there are 128 facility sites inspected, some of those sites 
have more than one facility or tower. Condition data were obtained at 
134 facilities within these 128 sites. 

[39] Facility replacement value is the total expenditure required to 
replace a facility, inclusive of construction, design, and project 
management and administrative costs. 

[40] GAO, Federal Buildings Fund: Improved Transparency and Long-term 
Plan Needed to Clarify Capital Funding Priorities, [hyperlink, 
http://www.gao.gov/products/GAO-12-646] (Washington, D.C.: July 12, 
2012). 

[41] OWCP data for 2012 are from January to September. 

[42] The Federal Employees' Compensation Act provides that a claim for 
compensation must be filed within 3 years of the date of injury. For a 
traumatic injury, the statutory time limitation begins to run from the 
date of injury. For a latent condition, it begins to run when an 
injured employee with a compensable disability becomes aware, or 
reasonably should have been aware, of a possible relationship between 
the medical condition and the employment. 

[43] For this report, we are using the date of injury to track the 
number of new workers' compensation claims filed by FAA employees and 
accepted by OWCP each year. These figures do not include workers' 
compensation claims that were rejected by OWCP or injuries for which 
an employee did not file a claim. 

[44] OSHA regulations require that all areas and operations of each 
workplace, including office operations, be inspected at least 
annually. 29 C.F.R. § 1960.25(c). 

[45] FAA defines a "facility" as a physical location where agency work 
or operations are routinely performed and include offices, buildings, 
structures, work areas, equipment, and outside grounds associated with 
the facility. 

[46] According to FAA, its occupational safety and health inspection 
program includes environmental aspects. 

[47] In March 1996, FAA originally established an OSHECCOM and 
reemphasized that effort in June 2008 with a revised charter. 

[48] Specifically, the OSHECCOM works to (1) establish a culture 
within FAA that facilitates an effective occupational safety, health, 
and environmental program; (2) improve intra-agency cooperation by 
establishing communication across the organizational components and 
promote a comprehensive occupational safety, health, and environmental 
program that can be implemented at all operational levels; (3) provide 
an open line of communication between management and nonmanagement 
employees regarding EOSH related matters; (4) monitor, assist, and 
support FAA's EOSH program; and (5) create a forum to address and 
discuss EOSH issues without fear of reprisal. 

[49] Because the unique design of air traffic control towers poses 
challenges to meeting standard OSHA egress and fire safety 
requirements, FAA developed an alternate standard specific to these 
facilities. Under its AIRTRAF inspection program, OSHA assessed air 
traffic control towers against this alternate standard. 

[50] See footnote 6. 

[51] Major building components include air handling units and 
ductwork, electrical distribution panels, lighting fixtures, lightning 
protection systems, uninterruptible power systems, and emergency 
generators, among others. 

[52] Mission critical assets are those that house high-cost, mission 
critical equipment and technological systems, or assets that support 
the continuity of operations at the Aeronautical Center. 

[53] FAA facility designation ranking is based on human occupancy 
levels and building size. The ranking with the highest score is a 
facility with over 151 employees and over 80,000 square feet of space. 
The next ranking is a facility with between 11 and 150 employees and 
2,500 to 80,000 square feet of space. The lowest ranking is for a 
facility with 10 or fewer employees and less than 2,500 square feet of 
space. 

[54] GAO, Department of Transportation: Key Issues and Management 
Challenges, 2013, [hyperlink, http://www.gao.gov/products/GAO-13-402T] 
(Washington, D.C.: Mar. 14, 2013). 

[55] FAA Modernization and Reform Act of 2012, Pub. L. No. 112-95, § 
804, 126 Stat.11,120 (Feb. 14, 2012). 

[56] Additionally, FAA's contract with the air traffic controllers' 
union requires that FAA notify the union as soon as possible but not 
less than one year in advance of the closure of a facility, facility 
consolidation, or inter-facility reorganization requiring reassignment 
of employees. This provision, along with the status of FAA's 
realignment and consolidation efforts, suggests that it will be no 
less than 2 years before realignments and consolidations occur. 

[57] This amount comprises about 6 percent of FAA's total deferred 
maintenance backlog for all staffed and unstaffed facilities. 
According to FAA officials, the backlog increases to $4.6 billion when 
factoring in the deferred maintenance for power systems, fuel storage 
tanks, and power cables as well as backlogs associated with 
environmental compliance and cleanup requirements. FAA expects this 
overall backlog is expected to grow to $7.5 billion in 2024. However, 
as we note earlier in this report, we have concerns regarding how FAA 
estimates the condition and deferred maintenance of those facilities 
that have not been inspected. 

[58] Pub. L. No. 99-177, title II, §§ 251-251A, 99 Stat. 1037 (Dec. 
12, 1985), as amended. 

[59] FAA Modernization and Reform Act of 2012, Pub. L. No. 112-95, § 
610, 126 Stat. 11, 117 (Feb. 14, 2012). 

[60] To differentiate responsibility for maintaining facilities, we 
use the phrase "FAA-maintained" or "GSA-maintained." 

[61] These data describe any claims submitted by employees for either 
(1) traumatic injury or (2) occupational disease or illness. In this 
report, we generally refer to these as claims for injuries and 
illnesses. 

[62] The FCI is equal to 100 - deferred maintenance/facility value. 
FAA has reported FCI as a percentage, but this scaling choice does not 
affect the analysis here. 

[63] See Jeffrey M. Wooldridge, Introductory Econometrics: a Modern 
Approach, 2ND ed. Mason, OH: Thomas South-Western, pp. 203-206. 

[64] This assumes that the population process generating the data used 
to estimate the model does not change between the time the model was 
estimated and the collection of new data. 

[65] More technically, these data could not have been new, random 
draws from a population data generation process, because they were 
calculated using a fixed mathematical formula. Our sample contrasts 
with the data FAA's contractor uses to estimate the parameters of its 
prediction model, which uses data from all inspections conducted in 
the last 6 years. 

[66] The confidence intervals are truncated at zero, because the 
deferred maintenance and FCI must be non-negative. 

[67] Trevor Hastie, Robert Tibshirani, and Jerome Friedman, The 
Elements of Statistical Learning (New York: Springer-Verlag, 2001) 
discusses cross-validation and pruning for recursive partitioning 
methods in more detail. 

[68] A similar version of the tree model of deferred maintenance is a 
three-way interaction of age, facility value, and an indicator for 
whether the facility was a large TRACON. The out-of-sample RMSE of the 
three-way interaction model was $272,169, compared to $263,292 for the 
tree model. Alternatively, one could estimate a model with the two-way 
interaction of age and value separately for large TRACONs and all 
other facilities. 

[69] Jacobs Engineering, "Organizational 2012 Final MARS Roll-Up 
Report for 128 ATCT-TRACON Sites (January 2007 - December 2012)," 
January 2013, pp. 24-26. 

[70] Jacobs Engineering, "Organizational Roll-Up Report for 109 ATCT - 
TRACON Sites," November 2008, Appendix C, pp. 2, 5. 

[End of section] 

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 website [hyperlink, http://www.gao.gov]. Each 
weekday afternoon, GAO posts on its website newly released reports, 
testimony, and correspondence. To have GAO e-mail you a list of newly 
posted products, go to [hyperlink, http://www.gao.gov] and select 
“E-mail Updates.” 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of 
production and distribution and depends on the number of pages in the 
publication and whether the publication is printed in color or black 
and white. Pricing and ordering information is posted on GAO’s 
website, [hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or 
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card, 
MasterCard, Visa, check, or money order. Call for additional 
information. 

Connect with GAO: 

Connect with GAO on facebook, flickr, twitter, and YouTube.
Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts.
Visit GAO on the web at [hyperlink, http://www.gao.gov]. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 
Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; 
E-mail: fraudnet@gao.gov; 
Automated answering system: (800) 424-5454 or (202) 512-7470. 

Congressional Relations: 

Katherine Siggerud, Managing Director, siggerudk@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, DC 20548. 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, DC 20548. 

[End of document]