Railroad Safety: Continued Emphasis Needed for an Effective Track Safety
Inspection Program (Chapter Report, 04/22/94, GAO/RCED-94-56).

The Federal Railroad Administration (FRA) has improved its track
inspection program and has a sound strategy for correcting weaknesses
that GAO flagged in earlier reports.  To further strengthen rail safety,
FRA needs to incorporate site-specific data on passenger and hazardous
materials traffic in its inspection plan and improve the reliability of
accident and injury data.  Because of gaps and inaccuracies in the
plan's data, FRA officials do not believe the plan will help them
develop annual strategies for targeting inspections of high-risk track.
Also, the usefulness of the staffing and evaluation components is
limited because they rely on the plan's questionable data.  FRA has
enhanced its daily oversight of track safety efforts.  FRA has not,
however, always enforced its policy that inspectors examine track
inspection records to review a railroad's compliance history before
physically inspecting track.  FRA faces a difficult challenge in
revising the safety standards for excepted track.  FRA intended that
exceptions apply to little-used lines that, for economic reasons, could
not be brought up to minimal safety standards.  Yet the number of
reported accidents and cited defects on excepted track has increased,
and FRA is concerned that railroads have abused the excepted track
provision.  Regulations prohibit FRA inspectors from writing violations
for excepted track and do not require railroads to fix cited defects.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-94-56
     TITLE:  Railroad Safety: Continued Emphasis Needed for an Effective 
             Track Safety Inspection Program
      DATE:  04/22/94
   SUBJECT:  Railroad accidents
             Reporting requirements
             Railroad safety
             Railroad regulation
             Railroad transportation operations
             Data integrity
             Inspection
             Safety regulation
             Internal controls
             Transportation safety
IDENTIFIER:  FRA Regional Inspection Points Program
             FRA National Inspection Plan
             FRA Automated Track Inspection Program
             FRA Quality Improvement Program
             Florida
             Georgia
             Kentucky
             FRA Railroad Inspection Reporting System
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Committees

April 1994

RAILROAD SAFETY - CONTINUED
EMPHASIS NEEDED FOR AN EFFECTIVE
TRACK SAFETY INSPECTION PROGRAM

GAO/RCED-94-56

Improving Track Safety


Abbreviations
=============================================================== ABBREV

  AAR - Association of American Railroads
  ANPRM - Advance Notice of Proposed Rulemaking
  ATIP - Automated Track Inspection Program
  CFR - Code of Federal Regulations
  DOT - Department of Transportation
  FRA - Federal Railroad Administration
  GAO - General Accounting Office
  ICC - Interstate Commerce Commission
  NIP - National Inspection Plan
  NTSB - National Transportation Safety Board
  QIP - Quality Improvement Program
  RIP - Regional Inspection Points Program

Letter
=============================================================== LETTER


B-240287

April 22, 1994

The Honorable Ernest F.  Hollings
Chairman
The Honorable John C.  Danforth
Ranking Minority Member
Committee on Commerce, Science
and Transportation
United States Senate

The Honorable John D.  Dingell
Chairman
The Honorable Carlos J.  Moorhead
Ranking Minority Member
Committee on Energy and Commerce
House of Representatives

The Rail Safety Enforcement and Review Act of 1992, Public Law
102-365, enacted September 3, 1992, requires GAO to report to the
Congress on the effectiveness of the Department of Transportation's
(DOT) enforcement of track safety standards, with particular
attention to recent relevant accident experience and data.  Within
DOT, the Federal Railroad Administration (FRA) establishes and
enforces regulations for railroad safety.  This report acknowledges
the improvement that FRA has made in its track safety program and
recommends actions to further strengthen track safety. 

We are sending copies of the report to the appropriate congressional
committees, the Secretary of Transportation, and the FRA
Administrator.  We will make copies available to others on request. 

This work was performed under the direction of Kenneth M.  Mead,
Director, Transportation Issues, who can be reached on (202)
512-2834.  Major contributors to this report are listed in appendix
II. 

Keith O.  Fultz
Assistant Comptroller General


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

The Rail Safety Enforcement and Review Act of 1992 requires GAO to
report to the Congress on the effectiveness of the Department of
Transportation's (DOT) enforcement of track safety standards.  Within
DOT, the Federal Railroad Administration (FRA) sets standards and
develops regulations for the safety of the nation's 635 railroads and
over 200,000 miles of track. 

Since 1987, GAO has issued many reports identifying weaknesses in
FRA's rail safety inspection and enforcement programs and recommended
improvements in the data and methods used to oversee the rail
industry.  This report addresses (1) the improvement FRA has made in
its track inspection program by correcting problems noted in past GAO
reports and (2) the implementation problems that still limit the
effectiveness of track inspections.  In addition, this report
addresses an emerging rail safety issue requiring FRA's
attention--federal oversight of track not currently subject to track
safety standards (excepted track). 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

FRA's rail safety program has three basic components:  planning,
staffing, and evaluation.  The planning component, FRA's new National
Inspection Plan, is a computer model that incorporates risk factor
data--derived from past inspections, accident and injury reports, and
records of passenger, hazardous materials, and total freight
traffic--to produce hourly inspection goals for each railroad, by
state.  The staffing component will allow FRA to implement these
inspection goals by indicating where inspectors are needed most to
ensure the timely inspection of high-risk track:  track on which many
accidents have occurred, track located near population centers, or
track used to carry passengers or hazardous materials.  The
evaluation component, which collects data on how much track is
inspected each year, enables FRA to determine whether it is meeting
the inspection goals.  Wherever a gap exists between the planning and
evaluation components, FRA is to develop strategies, including plans
for reallocating its inspectors, to ensure that the goals are met. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

FRA has improved its track inspection program, and its strategy for
correcting the weaknesses that GAO identified in previous reports is
sound.  To further strengthen rail safety, FRA needs to incorporate
site-specific data on passenger and hazardous materials traffic in
its inspection plan and improve the reliability of accident and
injury data.  These data are critical in setting inspection goals and
targeting inspection time to high-risk track.  Currently, because of
gaps and inaccuracies in the plan's data, FRA regional officials do
not view the plan's goals as providing them with a systematic basis
for developing annual strategies for targeting inspections to
high-risk track.  In addition, the usefulness of the staffing and
evaluation components is limited, since they rely on the plan's
questionable data. 

FRA has enhanced its daily oversight of track safety activities. 
Communication with participating states improved after FRA agreed
with the states on ways to avoid duplicative inspections.  In
addition, inspectors have applied track safety regulations and
reported track defects more consistently since GAO last reviewed this
issue.  However, FRA has not always enforced its policy that
inspectors examine track inspection records to review a railroad's
compliance history before physically inspecting track. 

FRA faces a difficult challenge in revising the safety standards for
excepted track.  FRA intended designations of excepted track to apply
to little-used lines that, for economic reasons, could not be brought
up to minimal safety standards.  However, the number of reported
accidents and cited defects on excepted track has increased, and FRA
is concerned that railroads have abused the excepted track provision. 
The track safety regulations do not allow FRA inspectors to write
violations for excepted track and do not require railroads to fix
cited defects. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4


      FRA'S TRACK INSPECTION
      STRATEGY CONTAINS WEAKNESSES
      AND OPPORTUNITIES
-------------------------------------------------------- Chapter 0:4.1

The National Inspection Plan, the cornerstone of FRA's new inspection
strategy, was revised in 1992 to better target FRA's routine
inspections of high-risk track.  However, an important element of the
plan, a Regional Inspection Points program, was not completed.  This
program was intended to incorporate data on the volume of rail,
passenger, and/or hazardous materials traffic carried on each route. 
In response to budget constraints, FRA stopped collecting these data
and substituted data on the miles of track inspected over 3 years in
its inspection plan for 1993--a substitution that did not pinpoint
which track carried passengers or hazardous materials. 

In implementing the inspection plan, FRA recognized that changes in
the goals would be needed to compensate for limitations in the risk
factor data.  Hence, FRA established a process through which the
regions could recommend adjustments to headquarters in the hourly
goals for each railroad in a state.  However, the regions have not
reported some adjustments to headquarters but have relied instead on
their inspectors to determine what track to inspect--a practice GAO
criticized in the past as unresponsive to the congressional mandate
that FRA develop a systematic approach to targeting limited
inspection resources to high-risk track.  Furthermore, FRA sets its
inspection goals in accordance with the location of its inspectors
rather than of high-risk track. 


      FRA HAS ENHANCED DAILY
      OVERSIGHT OF TRACK SAFETY
      ACTIVITIES
-------------------------------------------------------- Chapter 0:4.2

GAO recommended in 1990 that FRA inspectors improve communication
with state rail inspectors by coordinating their inspection
territories and sharing the results of their inspections.  During
1993, GAO found that FRA had defined track inspection territories in
12 of the 15 states reviewed.  In addition, FRA had revamped its
safety training programs and issued new enforcement manuals to reduce
variation in inspectors' application of safety standards throughout
the rail industry.  But because of budget constraints, FRA did not
fund its track safety training program in 1993.  GAO's analysis of
track inspection results from 1989 to 1992 indicates that track
inspectors have applied the safety standards more uniformly. 

Despite these improvements, FRA and state inspectors examined
railroads' track inspection records at only 60 percent of the
railroads they visited between 1989 and 1992.  FRA policy requires
inspectors to prepare for an inspection by reviewing a railroad's
track inspection records in order to gain an understanding of the
railroad's compliance history.  However, FRA inspectors have not
always implemented this policy because they have not had access to
original inspection records--especially since many railroads began to
centralize or automate these records--and they have considered
reviews of photocopied or electronic records unacceptable to FRA. 


      SAFETY OF EXCEPTED TRACK HAS
      DECLINED
-------------------------------------------------------- Chapter 0:4.3

Federal track safety standards do not apply to about 12,000 miles of
track designated by the industry as excepted; travel on such track is
limited to 10 miles per hour, no passenger service is allowed, and no
train may carry more than five cars containing hazardous materials. 
The safety of this track has declined over the past few years.  GAO
found that the number of track-caused accidents on excepted track
increased from 22 in 1988 to 65 in 1992--a 195-percent increase.  The
number of defects cited on excepted track also increased from 3,229
in 1988 to 6,057 in 1992.  With few exceptions, FRA cannot compel
railroads to correct these defects. 

According to FRA, the railroads have applied the excepted track
provision far more extensively than envisioned (i.e., on certain yard
track and little-used branch lines away from public roads and
populous areas).  One railroad classified about 80 percent of its 400
miles of track as excepted.  Other railroads transported hazardous
materials on excepted track through residential areas or
intentionally designated track as excepted rather than comply with
minimum safety standards.  According to FRA, these designations often
occurred after the agency advised the railroads that their track did
not meet safety standards. 

In November 1992, FRA announced its intention to review the excepted
track provision.  In response, the Association of American Railroads
expressed its disapproval of manipulating the excepted track rule to
avoid compliance but maintained that FRA already had the necessary
enforcement tools.  FRA, however, views the existing regulations as
inadequate because its inspectors cannot write violations for
excepted track and railroads are not required to correct defects. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

To overcome the problems identified in FRA's rail safety inspection
program, GAO recommends that the Secretary of Transportation direct
the FRA Administrator to gather the data needed to successfully
implement the National Inspection Plan and continue to improve FRA
inspectors' daily oversight of track safety activities.  GAO further
recommends that FRA strengthen its enforcement authority to allow
inspectors to require minimum safety standards on excepted track that
poses an imminent hazard.  GAO's detailed recommendations appear in
chapter 5. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:6

In commenting on a draft of this report, DOT agreed with most of
GAO's findings and concurred wholly or partially with all of GAO's
recommendations.  However, DOT maintained that the costs of gathering
site-specific data on passenger and hazardous materials traffic
prevented FRA from including these data in its inspection plan. 
While continuing to believe that these data are essential to target
inspections to high-risk track, GAO modified its recommendation to
endorse a pilot program in one FRA region, through which FRA could
collect site-specific data and assess the cost-effectiveness of
extending this effort nationwide.  Additional comments and responses
appear in chapter 5 and appendix I. 


INTRODUCTION
============================================================ Chapter 1

Under the Federal Railroad Safety Act of 1970, as amended, the
Federal Railroad Administration (FRA) is responsible for regulating
all aspects of railroad safety for the nation's 635 railroads and
more than 200,000 miles of track.\1 FRA's safety mission includes (1)
establishing federal rail safety rules and standards; (2) inspecting
railroads' track, signals, equipment, and operating practices; and
(3) enforcing federal safety rules and standards.  The railroads are
primarily responsible for conducting safety inspections of their own
equipment and facilities to ensure compliance with federal
regulations, while FRA monitors the railroads' actions.  The National
Transportation Safety Board (NTSB) also plays a role in ensuring
railroad safety.  Established by the Congress in 1966, NTSB
investigates transportation accidents, determines their probable
causes, and issues safety recommendations. 


--------------------
\1 The act directed the Secretary of Transportation to prescribe
regulations for all areas of rail safety.  The Secretary delegated
this responsibility to the FRA Administrator. 


   FRA'S APPROACH TO RAILROAD
   SAFETY
---------------------------------------------------------- Chapter 1:1

To carry out its safety mission, FRA has established eight regional
offices under the direction of an Associate Administrator for Safety. 
Inspectors specialize in one of five inspection disciplines:  track,
locomotive power and equipment, operating practices, signal and train
control, and hazardous materials.  The primary responsibility of the
inspectors is to conduct routine inspections of railroads.  When an
inspection reveals noncompliance with a federal safety regulation,
the condition is listed as a defect on an inspection report.  When an
inspector identifies a defect that poses an immediate safety hazard
or when noncompliance persists, a violation report is prepared and
submitted to FRA's Office of Chief Counsel to be used to assess a
civil penalty against the railroad.\2

This report focuses on FRA's track inspection discipline.  In 1993,
FRA had a total of 56 track inspectors operating out of its regional
offices.  These track inspectors report to a supervisory track
specialist in each region.  FRA also had 20 chief inspectors who
inspect small railroads in all disciplines, including track.  In
addition, 29 states have their own track safety inspectors; these
inspectors--54 in total--participate with FRA under cooperative
agreements to monitor railroads' compliance with the safety
regulations.  The state inspectors forward their inspection reports
to supervisory track specialists in the appropriate FRA regional
office.  FRA and state track inspectors monitor the performance of
the rail industry's track maintenance workforce of about 25,000,
including 5,000 track inspectors.  In 1992, the FRA and state track
inspectors conducted 17,000 inspections of 348,326 miles of track and
identified more than 100,000 track defects. 

FRA's track safety standards are contained in part 213, title 49, of
the Code of Federal Regulations (49 C.F.R.  213).  Last revised in
1982, the standards prescribe minimum safety requirements for
railroad track that is part of the general railroad system of
transportation.  This includes virtually all track over which
commercial railroads operate.  The standards define (1) track
inspection requirements that railroads must follow; (2) the condition
of the track structure--such as the roadbed, crossties, rails, and
switches--that must be met to operate trains safely at given speeds;
and (3) the geometry of curved track, e.g., the gage (distance
between rails), alinement, and elevation of outer rails.  These
standards vary, depending on the track class.  The stricter the
requirements that must be met, the higher the maximum allowable
operating speed.  As table 1.1 shows, FRA has defined six classes of
track and designated maximum train speeds for each. 



                          Table 1.1
           
           FRA Track Classes and Associated Maximum
                       Operating Speed

                  (Speed in miles per hour)


Track class                            Freight     Passenger
--------------------------------  ------------  ------------
1                                           10            15
2                                           25            30
3                                           40            60
4                                           60            80
5                                           80            90
6                                          110           110
------------------------------------------------------------
Source:  49 C.F.R.  part 213.9(a). 

The Rail Safety Enforcement and Review Act of 1992 (which amended the
Federal Railroad Safety Act of 1970) required FRA to review and
revise its track safety standards.  As part of this review, FRA held
four workshops in which it discussed revisions to the standards with
representatives of the railroad industry, railroad employee unions,
and other interested parties.  FRA is required to complete its
efforts by September 1994. 


--------------------
\2 Both defects and violations are instances of regulatory
noncompliance; violations are considered more serious.  Penalties are
not assessed for defects, although railroads are expected to correct
the defective conditions. 


   FRA'S RESPONSE TO PRIOR GENERAL
   ACCOUNTING OFFICE REPORTS
---------------------------------------------------------- Chapter 1:2

Since 1987, we have issued many reports identifying weaknesses in
FRA's rail safety inspection and enforcement programs.  For example,
in July 1990, we reported on FRA's progress in meeting the
requirements of the Federal Railroad Safety Authorization Act of 1980
that FRA submit to the Congress a system safety plan to carry out
railroad safety laws.  As part of the plan, FRA was directed to
develop an inspection methodology that considered carriers' safety
records, population centers, and the volume and usage of track.  The
statute required the Secretary to give appropriate priority to
inspections of track and equipment involving passenger and hazardous
materials routes.  The House report accompanying this provision
stated that FRA should be able, through its national plan, to target
safety inspections to high-risk track--track with a high incidence of
accidents and injuries, located in populous urban areas, carrying
passengers, or transporting hazardous material. 

In our 1990 report, we found that the National Inspection Plan (NIP)
that FRA had developed did not include volume data on passenger and
hazardous materials routes--two important risk factors.\3 In an April
1989 report, we also noted problems with another NIP risk
factor--accidents and injuries.\4 We found that the railroads had
substantially underreported and inaccurately reported the number of
accidents and injuries and their associated costs.  As a result, FRA
could not integrate inspection, accident, and injury data in its
inspection plan to target high-risk locations and thereby achieve
maximum effectiveness with its limited inspection resources. 

In response to our 1990 report, the FRA Administrator stated that FRA
could greatly improve its safety inspection program by

     targeting [its] resources more effectively on the basis of the
     wealth of statistics at [its] command.  Given the size of [the]
     inspector force in relation to the size of the railroad
     industry, [FRA could] maximize the effect of [its] resources
     only by deploying them as scientifically and strategically as
     possible. 

We also reported in 1990 that FRA inspectors were not filing safety
violations uniformly.  We noted that inconsistencies existed because
FRA did not provide adequate training for its inspectors and did not
provide clear guidance on issuing violations.  In addition, we noted
communication problems between federal and state inspectors that
resulted in overlapping inspection territories and led state and FRA
inspectors to inspect the same track within days of one another.  We
recommended that FRA routinely exchange inspection plans and
periodically meet with state inspectors to discuss their inspection
activities. 


--------------------
\3 Railroad Safety:  New Approach Needed for Effective FRA Safety
Inspection Program (GAO/RCED-90-194, July 31, 1990). 

\4 Railroad Safety:  FRA Needs to Correct Deficiencies in Reporting
Injuries and Accidents (GAO/RCED-89-109, Apr.  5, 1989). 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:3

We conducted our review in response to a requirement of the Rail
Safety Enforcement and Review Act of 1992 (P.L.  102-365) that we
report to the Congress on the effectiveness of FRA's enforcement of
track safety standards.  Following discussions with the office of the
Senate authorizing Committee,\5 we agreed to assess FRA's progress in
implementing recommendations made in our past reports concerning
track safety.  Specifically, we sought to determine

  what progress FRA has made in improving its track inspection
     program by correcting problems noted in our past reports and

  what implementation problems, if any, continue to limit the
     effectiveness of track inspections. 

In addition, over the course of our review we found an emerging issue
related to rail safety that FRA must address in the near term.  As a
result, this report also provides information on FRA's efforts to
revise track standards for track not currently subject to federal
oversight--excepted track. 

To assess the actions FRA has taken to strengthen its track
inspection program, we (1) reviewed internal FRA reports and
memoranda describing these actions and (2) analyzed the inspection
plans and supporting data for calendar years 1992 and 1993.  We
interviewed officials from three headquarters offices--Safety
Enforcement, Safety Analysis, and Office of Policy--about their roles
in developing and implementing FRA's new track inspection program
strategy.  We also reviewed track inspection records and related
program documents and interviewed officials from four of FRA's eight
regional offices--Atlanta, Georgia; Chicago, Illinois; Kansas City,
Missouri; and Philadelphia, Pennsylvania.  We selected these regions
to provide a geographically balanced view of FRA's inspection and
enforcement efforts.  In addition, we interviewed an NTSB rail safety
official about FRA's track safety program and reviewed NTSB reports
on track-caused train accidents. 

To assess the effectiveness of the actions FRA has taken to
coordinate its efforts with those of the states, conduct uniform
inspections, and resolve other problems identified in our past
reports, we interviewed Office of Safety officials, including FRA's
state coordination program manager and a recently appointed training
official.  We reviewed records of training programs and minutes of
meetings between FRA and state rail safety officials.  We analyzed
the results of track inspections for the period from 1989 to 1992. 
We interviewed officials in FRA's Atlanta, Chicago, Kansas City, and
Philadelphia regions to learn how FRA track inspectors planned and
carried out their inspections and coordinated their activities with
track inspectors in 15 states located in those regions.  We also
discussed coordination with FRA and related issues with rail safety
officials in six states--Illinois, Iowa, Maryland, Missouri,
Tennessee, and Virginia--within the four regions, and with officials
of the National Association of Regulatory Utility Commissioners. 

In addressing the question of excepted track, we interviewed
officials in FRA's Office of Safety Enforcement and in the Chicago
and Philadelphia regions.  We attended FRA workshops in Denver,
Colorado; Newark, New Jersey; and Washington, D.C., where FRA
discussed proposed revisions to its track safety standards, including
revisions to excepted track rules.  We reviewed the proceedings of
these workshops and obtained and analyzed position papers on excepted
track.  We also analyzed FRA data on accidents reported and defects
cited on excepted track and the results of an FRA special inspection
of a railroad with considerable excepted trackage. 

Finally, we discussed track quality and related safety issues with
representatives of the Brotherhood of Maintenance of Way Employees,
the Association of American Railroads, and the American Short Line
Railroad Association.  FRA used information from several automated
systems to plan its inspection activities and to measure the results
of those activities.  We analyzed data from FRA's Railroad Inspection
Reporting System, accident and injury reporting systems, and other
systems supporting the NIP. 

We conducted our review from January to November 1993 in accordance
with generally accepted government auditing standards. 

In this report, we present our findings in chapters 2, 3, and 4 and
our conclusions and recommendations in chapter 5.  We provided both
the Department of Transportation (DOT) and FRA with a draft of our
proposed report.  DOT responded on behalf of both agencies.  Its
principal observations and our general responses to these
observations are summarized at the end of chapter 5.  DOT's written
comments appear in full in appendix I, together with our detailed
responses to specific comments. 


--------------------
\5 Senate Committee on Commerce, Science, and Transportation. 


OPPORTUNITIES EXIST TO IMPROVE
FRA'S NEW TRACK INSPECTION PROGRAM
============================================================ Chapter 2

FRA has made progress in developing a comprehensive strategy for
inspecting the nation's rail tracks as required by the Federal
Railroad Safety Authorization Act of 1980\1 and thereby in meeting
the law to develop a system safety plan.  Its safety program now has
three basic components:  a revised NIP, a staffing allocation model,
and a Quality Improvement Program (QIP) (now being implemented). 
When fully implemented, the track portion of the inspection program
will enable FRA to identify and target limited resources for
inspections to high-risk track and evaluate whether its efforts are
improving the overall safety of the nation's track. 

However, FRA still faces challenges similar to those we cited in our
previous reports.  First, it has not obtained and incorporated into
the NIP site-specific data on two critical risk factors--the volume
of passenger traffic and the volume of hazardous materials
transported.  Second, it has not improved the reliability of another
critical risk factor--the reporting of accidents and injuries
nationwide.  FRA was developing a program to provide this
information, but the program was not completed in time for the 1993
or the 1994 NIP.  Because of these inadequacies in the NIP, FRA
regional offices generally do not view the inspection goals produced
by the NIP as providing a systematic basis upon which they can
develop annual strategies for targeting inspections to high-risk
routes.  In addition, the usefulness of the staffing allocation model
and the QIP is limited; the proposed staffing model relies on NIP
data, and the QIP is intended as a check on the extent to which
inspectors are meeting the NIP goals. 


--------------------
\1 Public Law 96-423; 94 Stat.  1811. 


   FRA'S TRACK INSPECTION STRATEGY
   HAS THREE COMPONENTS
---------------------------------------------------------- Chapter 2:1

FRA's rail safety program has three basic components:  planning,
staffing, and evaluation.  The planning component--a revised NIP--is
a computer model that evaluates data on risk factors, including past
inspection results, accident and injury information, and the volume
of passenger, hazardous materials, and total freight traffic.  The
NIP produces inspection goals--expressed in hours of inspection
time--for each railroad, by state and by discipline.  For the track
discipline, the NIP is designed to identify which track segments pose
the greatest risk of accident or harm to the public and which should
therefore receive the most inspection resources. 

For the track discipline, seven risk factors are evaluated and given
equal weight in the NIP model:  (1) regional inspection points, (2)
track-caused accidents, (3) track-caused casualties (fatalities and
injuries), (4) passenger traffic, (5) freight tonnage, (6) hazardous
materials tonnage, and (7) track defects.  The Regional Inspection
Points (RIP) program is a computer-based program intended to identify
and periodically update the universe of inspection points in each
discipline to provide a baseline in the NIP for allocating resources. 
For the track discipline, the RIP program is designed to contain a
national inventory of all inspection points (track segments) with
data on the physical characteristics of all track in each inspector's
territory. 

FRA proposes to use the NIP data in a staffing allocation model that
will enable the agency to determine how many inspectors are needed in
each discipline and how they can best be allocated.  The staffing
model will also incorporate data from a computer-based program (the
QIP) that gathers and develops data on inspectors' workloads.  The
QIP, which allows FRA to measure progress in meeting NIP goals,
consists of daily activity reports that inspectors fill out to show
the time they spend inspecting and doing other activities.  Track
inspectors' reports provide information on the number of miles of
track inspected each year and the number of defects cited.  Where a
gap exists between the planning and evaluation components, FRA is to
develop strategies, including plans for reallocating its inspectors,
to ensure that its goals are met. 


   CHALLENGES REMAIN IN ENSURING
   THAT THE NIP IS FULLY
   IMPLEMENTED
---------------------------------------------------------- Chapter 2:2

FRA revised its NIP for 1992 in an attempt to comply with the Federal
Railroad Safety Authorization Act of 1980, which required FRA to
develop an inspection methodology that considered the volume of
passenger and hazardous materials traffic in specific locations and
railroads' safety records.  FRA has not yet collected information on
passenger and hazardous materials traffic and is still working to
ensure the reliability of the data submitted by industry on accidents
and injuries. 


      RIP PROGRAM DOES NOT INCLUDE
      SITE-SPECIFIC DATA ON
      PASSENGER AND HAZARDOUS
      MATERIALS TRAFFIC
-------------------------------------------------------- Chapter 2:2.1

In the 1980 act, the Congress required the Secretary to give
appropriate priority to inspections of track and equipment involving
passenger and hazardous materials routes.  FRA developed the RIP data
collection program in response to the act and our 1990 report, which
concluded that FRA's safety inspection program should incorporate
data on the location of track (its proximity to population centers),
frequency of track usage, and volume of passenger and/or hazardous
materials traffic.  FRA began collecting RIP data in all disciplines
in early 1991 but had not completed the effort when we conducted our
study in 1993.  While track inspectors had prepared an inventory of
inspection points for each train route, the inventory included only
data on physical characteristics, such as miles of track, type of
track, type of traffic, and location.  Data on the volume of
passenger and hazardous materials traffic and on the frequency of
track use were not collected.\2

Because the RIP data were unavailable, FRA entered data on the miles
of track inspected by FRA inspectors over a recent 3-year period into
the 1993 NIP.  According to FRA officials, this information was a
poor substitute for the RIP data because it could not pinpoint which
track carried passenger traffic or hazardous materials.  The 1993 NIP
noted that FRA intended to collect information on the volume of
passenger and hazardous materials traffic, since this information was
the cornerstone of the plan.  However, FRA safety officials noted
that, because of budget constraints, they have indefinitely suspended
their efforts to collect the required information.  In addition, an
FRA official stated that FRA may have to impose a new reporting
requirement on the industry to obtain information on hazardous
materials; FRA is, however, reluctant to impose such a requirement. 


--------------------
\2 The volume of passenger traffic is measured in passenger miles,
while freight traffic is measured in train miles.  A passenger mile
is defined as one passenger traveling 1 mile; a train mile is defined
as one train traveling 1 mile.  Hence, a passenger train that carries
300 passengers a distance of 1 mile equals 300 passenger miles, while
a freight train that travels a distance of 1 mile travels 1 train
mile. 


      RAILROADS DO NOT ACCURATELY
      REPORT ACCIDENT/INJURY DATA
-------------------------------------------------------- Chapter 2:2.2

Railroads must report monthly to FRA all accidents involving damages
exceeding a biennially adjusted dollar threshold ($6,300 in 1993), a
death, or an injury on railroad property that requires medical
treatment (49 C.F.R.  225).\3 The NIP includes accident and injury
data, since track with a higher preponderance of accidents poses a
higher risk to rail safety.  In 1989, we reported that some railroads
substantially underreported and inaccurately reported accidents and
injuries, and we recommended improvements in the internal controls
for reporting this information. 

FRA agreed with our April 1989 recommendations and increased the
number of inspections it conducted for compliance with
accident/injury reporting requirements by 50 percent over 1989
levels.  In addition, the agency revised its accident/incident
reporting guide to clarify the reporting codes and will publish
revised rules on the accident/incident reporting requirements.  FRA
has also audited the large railroads' reporting procedures and a
sampling of the smaller railroads' procedures to help determine
whether regulatory changes in internal control procedures are needed. 

A January 1993 FRA report found that railroads had improved their
reporting of more serious injuries.  However, this report also found
that (1) in many cases, carriers' records were still inaccurate and
unreliable, and many reportable employee injuries were not reported;
(2) initial accident reports were often not updated to reflect more
accurate information; and (3) most nonreporting or inaccurately
reporting railroads had internal communication and/or control
problems.  Other FRA internal reports have identified the
underreporting of accidents as a continuing industry problem. 

Our review of the accuracy of the accident/injury data base confirmed
FRA's internal findings.  We assessed whether FRA's data base
included information on 39 track-caused accidents that NTSB
investigated during 1990 and 1991.  FRA's data base included
information on all 39 accidents that NTSB investigated.\4 However,
for five accidents, the reports of casualties differed.  At our
request, FRA followed up on one accident and determined that the
railroad had not reported one death and two serious injuries that
resulted from the accident.  In addition, while NTSB reported $35.3
million in damages resulting from the 39 accidents, FRA's data base
reported only $26.2 million in damages.\5 The damages reported by
NTSB were over 25 percent higher in 22 of these cases.  This finding
supports FRA's conclusion that accident reports are not updated as
more reliable information on damages becomes available. 

FRA safety officials recognize that they have not implemented many
proposed corrective actions to improve railroads' reporting of
accidents and injuries.  In March 1990, FRA issued an Advance Notice
of Proposed Rulemaking (ANPRM) on methods to improve rail carriers'
reporting.  In comments on a draft of this report, FRA stated that it
has been developing and will publish a Notice of Proposed Rulemaking
to revise its current accident/incident reporting requirements. 


--------------------
\3 Railroads are required to report to FRA all deaths and all
injuries that occur on or adjacent to railroad property except for
injuries requiring one-time first-aid treatment.  Railroads must also
report train accidents (which FRA defines as collisions, derailments,
and other occurrences) for which damage to railroad equipment and
track exceed a current dollar threshold.  In addition, railroads are
required to report all rail-highway crossing accidents regardless of
damage to equipment and track. 

\4 NTSB investigates only about 100 of the most serious rail
accidents that occur each year; about 3,000 accidents are reported
annually to FRA. 

\5 To ensure that we were evaluating comparable data, we excluded
cargo losses and other damages to nonrailroad property from the NTSB
calculation.  FRA did not include cargo losses in its report. 


      OTHER RISK FACTOR DATA ARE
      NOT COMPARABLE OR ACCURATE
-------------------------------------------------------- Chapter 2:2.3

Because risk factor data are not collected in a comparable format,
their usefulness in the NIP is diminished.  While the RIP data
identifying track segments to be inspected were intended to be
route-specific, the NIP's fourth, fifth, and sixth risk
factors--passenger traffic, freight tonnage, and hazardous materials
tonnage--were intended to provide state-specific information.  We
found that some railroads did not separate freight from passenger
traffic.  In addition, the information in the data base
underrepresented the hazardous materials tonnage and total freight
tonnage transported on smaller railroads.  Errors in the NIP goals
have resulted. 

For the passenger traffic risk factor, FRA collected data from Amtrak
and from commuter and excursion railroads to enter into the NIP. 
According to the 1993 NIP, these data should represent train miles
operated in passenger service.  Because FRA had access to Amtrak's
schedules, it was able to calculate train miles operated in passenger
service for all Amtrak routes.  But because schedules for local
commuter and excursion lines were not as readily available, FRA
relied on data these railroads had already reported to FRA.  However,
we found that the reported data did not separate passenger and
freight train miles.  As a result, the method FRA used to collect
passenger information is likely to have given disproportionate weight
(too many inspection hours) to railroads that operate both passenger
and freight trains. 

In addition, we found errors in the number of train miles operated in
passenger service as reported in the NIP.  For example, the Wisconsin
Central Railroad operated no regular passenger service in the state
of Wisconsin.  Nevertheless, the risk factor data showed 2.6 million
train miles in passenger service for this railroad for a recent
12-month period.  Officials in FRA's Office of Safety Analysis were
not able to explain this error.  They also could not explain data
showing 568,000 passenger train miles for the Iowa Interstate
Railroad--a railroad that carried no passengers in 1992. 

The method that FRA used to collect statewide information on
hazardous materials and overall freight tonnage may also produce
errors in the final NIP goals.  FRA used sample data reported to the
Interstate Commerce Commission (ICC).  ICC annually collects a
1-percent sample from the waybills (records of total goods shipped)
of Class I railroads.\6 The records used for the sample cover the
entire transportation history of a commodity's movement, which
includes movement across non-Class I railroads.  According to FRA,
the sample would not include shipments that occurred exclusively on
non-Class I railroads.  As a result, FRA officials said, non-Class I
railroads do not have the same chance of being represented in the
sample as the Class I railroads.  FRA stated that because of the
limitations in the waybill sample, they did not use hazardous
materials tonnage data to distribute inspection hours between large
and small railroads.  According to the FRA official responsible for
the NIP's operation, the ICC information may contain other errors,
since FRA did not determine the overall reliability of the data base. 


--------------------
\6 The 15 Class I railroads are the nation's largest; each has annual
gross operating revenue in excess of $250 million.  In calendar year
1992, these railroads accounted for 87 percent of the nation's total
train miles and 78 percent of the reported train accidents. 


   DATA GAPS LIMIT THE NIP'S
   EFFECTIVENESS
---------------------------------------------------------- Chapter 2:3

Because of the NIP's inadequacies, FRA regional officials generally
do not view the NIP output as a systematic basis upon which they can
develop annual strategies for targeting inspections to high-risk
routes.  In addition, the usefulness of the staffing allocation model
and of the QIP is limited because the proposed staffing model relies
on NIP data and the QIP is intended as a check on the extent to which
inspectors meet the NIP goals. 


      INSPECTORS SEE LIMITED USE
      FOR THE CURRENT NIP GOALS
-------------------------------------------------------- Chapter 2:3.1

In implementing the NIP, FRA recognized that changes in hourly
inspection goals would be needed for various reasons to compensate
for limitations in the risk factor data.  To control the accuracy and
reliability of the NIP goals, FRA established a process through which
regional officials could recommend adjustments to the hourly goals
for each railroad in the state.  Regional officials we spoke with
have responded to the adjustment process by either (1) adjusting the
NIP goals and reporting the adjustments to headquarters or (2)
adjusting the NIP goals and not reporting the adjustments to
headquarters.  In both cases, the regions have relied primarily on
their inspectors to determine where and when inspections are
needed--a practice we criticized in our July 1990 report as not being
responsive to the congressional mandate for FRA to develop a
systematic approach to targeting limited inspection resources to
high-risk track. 

In part, the regions have relied more on their inspectors than on the
NIP goal-setting process because they have questioned the overall
usefulness of the NIP goals.  For example, the Northeast Illinois
Regional Commuter Rail Corporation operated commuter service mainly
on track owned and maintained by four other railroads.  In 1992 and
1993, the NIP allocated 306 and 205 hours, respectively, for track
inspections of this commuter railroad.  The FRA Region 4 Director
stated that the NIP guidelines required him to transfer the
inspection hours from the commuter railroad to the railroads that
owned and maintained the track on which the commuter railroad
operated.  However, he did not make the transfer because he did not
think it served a useful purpose. 

The regions have also relied on their own inspectors because they
have perceived that headquarters staff have not made adjustments in
the NIP when the regions have proposed them.  For example, in May
1992, after the Toledo, Peoria and Western Railway Corporation
acquired more than 250 miles of track from the Atchison, Topeka and
Santa Fe Railway Company in Illinois and Indiana, FRA Region 4
requested that FRA headquarters adjust its inspection hours to
reflect the change in ownership.  However, the Office of Safety
Analysis did not adjust the NIP for either 1992 or 1993.  We also
found that headquarters staff did not adjust the 1993 NIP to reflect
a similar change in ownership, as requested by Region 2. 

The regions further questioned the usefulness of the NIP goals after
headquarters found errors in the NIP goals for 1993.  In April 1993,
FRA's Office of Safety Analysis discovered a computer programming
error while extracting data on track-caused accidents from the
accident/injury data base.  Although the office corrected the
programming error and recalculated the 1993 NIP goals, it did not
issue a revised NIP to the regions.  According to an Office of Safety
Analysis official, the impact of the error was too minor to issue a
revised NIP.  However, we found examples in FRA's eight regions in
which revisions in the NIP goals could have resulted in a 25-percent
increase or decrease in the number of inspection hours for certain
railroads. 


      USEFULNESS OF FRA'S STAFFING
      MODEL AND QUALITY
      IMPROVEMENT PROGRAM IS
      LIMITED
-------------------------------------------------------- Chapter 2:3.2

In July 1990, we reported that FRA needed to target its inspection
resources to the areas posing the greatest risk.  Since then, FRA has
been developing a staffing model that will incorporate NIP risk
factor data and inspection coverage standards to estimate how many
inspectors FRA will need and how it can best allocate its inspectors
among the inspection disciplines and regional offices.  The NIP data
would identify high-risk railroads and indicate the number and
location of the routes requiring inspection coverage.  Because of the
problems cited in the NIP data, we believe the staffing model will
not be effective in determining the number and location of the
inspectors needed in each discipline. 

At the time of our review, the staffing allocation model was still
under development, and FRA was allocating its inspection goals
according to the current location of its inspectors rather than
according to the location of the railroads or routes that pose the
greatest risk to safety.  However, even if the limitations in the NIP
are addressed, FRA will not be able to move inspectors to the sites
that the allocation model determines are most critical.  FRA
officials stated that it would be too costly to move inspectors on
the basis of the allocation model's results.  Eventually, through
attrition among inspectors and the assignment of new inspectors to
the disciplines and regions with the greatest need, FRA expects to
implement the goals of the staffing allocation model.  However, FRA
stated that inspection goals may still be assigned to regions that
cannot accomplish their mission with the available staff. 

The third component of FRA's rail safety program, the QIP, measures
inspectors' productivity and allows FRA to determine the extent to
which hourly inspection goals are being achieved.  FRA and state
inspectors are producing daily activity reports and reporting the
information to the QIP data base.  Although the QIP information will
become important in determining whether the NIP is meeting its
intended goals and improving the overall quality of the track
inspection program, its usefulness will not be fully realized until
the NIP and staffing allocation model are improved. 


   USE OF OTHER TOOLS COULD
   IMPROVE FRA'S TRACK INSPECTION
   PROGRAM
---------------------------------------------------------- Chapter 2:4

FRA uses an Automated Track Inspection Program (ATIP) vehicle to
monitor railroads' compliance with track geometry standards.  If
targeted properly, the ATIP vehicle could supplement coverage by
inspectors in regions where the staffing allocation model showed a
shortage of track inspectors to meet the NIP goals.  FRA also does
not include in its staffing allocation model the 54 state inspectors
who assist FRA track inspectors in monitoring railroads' compliance
with track safety standards.  These state inspectors are not
currently included in the NIP. 


      FRA HAS NOT CONSIDERED ATIP
      VEHICLE IN PLANNING AND
      IMPLEMENTING NIP GOALS
-------------------------------------------------------- Chapter 2:4.1

FRA's ATIP vehicle is a self-propelled railcar that contains
electronic sensing and data-processing equipment.  While traveling up
to 80 miles per hour, the vehicle can measure track conditions and
geometric defects.\7 The vehicle produces detailed exception reports
listing the specific type and number of geometric defects in the
track.  In 1991, the ATIP vehicle inspected 21,023 miles of track and
identified 15,708 geometric defects--892 alinement, 4,903 gage, and
9,913 cross-level defects.\8 In comparison, FRA's inspectors covered
333,114 miles of track during 1991 routine inspections. 

The relationship of the ATIP vehicle to other elements of FRA's track
inspection program has been unclear since an ATIP vehicle was first
used in 1974.  According to FRA's track enforcement manual, track
inspectors should use the ATIP vehicle's results for planning and
prioritizing their routine inspection visits and should follow up
within 60 days of the vehicle's operation to ensure that all defects
have been located and corrected.  However, the 1993 NIP does not
address whether track inspected by the ATIP vehicle should receive
less routine coverage.  The Office of Safety generally schedules the
vehicle so that it covers the northern states during the summer
(traveling west to east) and southern states during the winter
(traveling east to west).  This scheduling process is independent of
any NIP results that would indicate where inspector resources are
insufficient to meet the hourly inspection goals.  NTSB noted in its
1985 review of the ATIP program that such vehicles, which are also
operated by railroads in their track inspections, must be used in a
well-planned track survey program to supplement the activities of
track inspectors. 


--------------------
\7 FRA's ATIP vehicle is operated under contract by Ensco, Inc.  The
contractor received $500,000 in fiscal year 1993 to operate the ATIP
for FRA. 

\8 At the time of our review, ATIP results for 1992 were not
available. 


      FRA HAS NOT INCLUDED STATE
      INSPECTORS IN ITS INSPECTION
      PLANS
-------------------------------------------------------- Chapter 2:4.2

In our 1990 report, we recommended that FRA include the resources
represented by state inspectors in its staffing allocation model. 
According to the 1993 NIP, state inspectors augment but do not reduce
the number of FRA inspectors assigned to a specific state.  As a
result, according to FRA officials, more track inspections occur in
states that have state inspectors.  As stated in the 1993 NIP, state
resources were not included in federal plans because of funding
uncertainties or changes in state inspection priorities.  For
example, Minnesota state inspectors generally reviewed track
rehabilitation efforts rather than monitored railroads' compliance
with FRA track safety standards. 

However, FRA's Region 3 considered existing state inspector resources
in its decision to reallocate federal inspections to understaffed
states.  The region reallocated inspection time for Florida, which
had three state track inspectors, to Georgia and Kentucky, which had
no state track inspectors.  Thus, although the Office of Safety does
not plan to include state inspectors in the NIP planning process, a
regional FRA director shifted inspectors to ensure uniform coverage
among the various states in the region. 


FRA HAS ENHANCED DAILY OVERSIGHT
OF TRACK SAFETY ACTIVITIES
============================================================ Chapter 3

FRA's NIP and track enforcement manual state that, to effectively
monitor railroads' compliance with track safety standards on a daily
basis, the agency must (1) coordinate its work with that of state
rail inspectors, (2) promote compliance by uniformly applying the
track safety standards, and (3) review railroads' inspection records. 
Since we conducted our previous reviews in 1990, we found that FRA
had improved coordination with state rail inspectors by clearly
defining inspection territories and had ensured more uniform
application of track safety standards by redesigning its training
program.  However, inspectors were reviewing railroads' inspection
records at only 60 percent of the railroads they visited, despite the
FRA requirement that inspectors review all records before physically
inspecting track.  In addition, many railroads are consolidating
their records at one location, making inspections of records more
difficult for FRA and state inspectors. 


   FRA HAS IMPROVED COORDINATION
   WITH STATE RAIL INSPECTORS
---------------------------------------------------------- Chapter 3:1

Twenty-nine states have their own track safety inspectors.  Under
cooperative agreements with FRA, these inspectors monitor the
railroads' compliance with track safety standards.  FRA believes that
state inspectors are an integral part of overall inspection
activities and provides guidance to regional managers on how to
ensure that FRA and state inspectors work effectively together.  We
reported in 1990 that poor communication with the states decreased
the safety program's efficiency.  We recommended that FRA improve
communication with the states by coordinating inspection territories
and sharing the results of inspections.  During 1993, we found that
FRA had coordinated its inspection territories with 12 of the 15
states we reviewed.\1 However, FRA provided the results of its safety
inspections to only 5 of the 15 states we contacted.  Accordingly, in
10 states, the state inspectors did not have complete information on
a railroad's responsiveness to prior safety reports before conducting
their inspections. 


--------------------
\1 The four FRA regions in our study encompass 15 of the 29 states
that participate in FRA's track safety inspection program. 


      INSPECTORS GENERALLY
      COORDINATE INSPECTION
      TERRITORIES
-------------------------------------------------------- Chapter 3:1.1

In 12 of the 15 states we reviewed, FRA and state inspectors had
effectively coordinated their itineraries.  In seven states, FRA and
the states had divided the state into mutually exclusive inspection
territories.  Defining territories prevents overlapping inspections,
since each inspector is assigned specific railroads or routes to
inspect.  For example, Nebraska has three track inspectors--two from
FRA and one from the state.  The state is divided into three
territories, and one inspector is assigned to each territory.  FRA
regional officials stated that, with defined inspection territories,
FRA and state inspectors will not burden the railroads with
duplicative inspections. 

Five other states coordinated their inspection activities without
dividing the state into exclusive territories.  In three of the five
states, the state inspectors routinely sent their monthly inspection
schedules to FRA.  In the remaining two states, the FRA and state
inspectors communicated informally with one another periodically to
avoid duplicating inspection coverage. 

FRA and three states did not effectively coordinate their
inspections.  One state in FRA's Region 6 did not routinely inform
FRA of its inspection plans.  The region's track specialist stated
that he was not able to coordinate with the state.  He noted that FRA
inspectors working in the state usually receive information from the
railroads after a state inspection has occurred.  Similarly, FRA
Region 3 inspectors found it difficult to coordinate inspection plans
with two states.  An official from one of the states acknowledged
this lack of coordination and said that state and federal inspections
often overlapped. 


      FRA REGIONS DO NOT ALWAYS
      SHARE INSPECTION RESULTS
      WITH STATES
-------------------------------------------------------- Chapter 3:1.2

The FRA regions in our review did not consistently share information
about their inspection activities with state track inspectors.  Only
5 of the 15 states received the results of FRA's track inspections. 
In addition, while three of the four FRA regions invited state
inspectors to attend civil penalty settlement conferences, state
inspectors did not always attend because of travel limitations. 

FRA requires all FRA and state inspectors to review prior FRA track
inspection reports in order to understand a railroad's compliance
history (responsiveness to the prior inspection reports) before
conducting an inspection.  While all four FRA regions kept prior
track inspection reports on file, only 5 of the 15 states in our
review received the results of FRA's track inspections.  In some
instances, regional track specialists believed that defining
territories precluded the need for sharing the results of track
inspections with state inspectors.  However, FRA and state inspectors
with exclusive territories may inspect different sections of track
owned by the same railroad.  For example, FRA and state inspectors in
Illinois had exclusive territories.  However, in 1992, FRA and state
inspectors both inspected track owned by the Burlington Northern,
Conrail, CSX, and Santa Fe railroads in Illinois.  Information from
all of these inspections would provide the inspectors with a more
complete understanding of the railroads' compliance history. 

Inspectors can also obtain a more complete understanding of a
railroad's compliance history by attending settlement conferences. 
When an inspection results in a regulatory violation, FRA meets with
the railroad to settle the civil penalty assessment.  In three of the
four regions we reviewed, FRA invited state inspectors to these
settlement conferences, particularly when the state inspector had
identified the violations that precipitated the civil penalty.  FRA's
Region 2 did not invite state inspectors to settlement conferences
because in the past the state inspectors had declined when invited. 
According to FRA officials, by attending these meetings, FRA and
state track inspectors can better understand the quality of their
violation reports and obtain information on violations that other FRA
inspectors have assessed against the railroad. 

Because of the costs involved, state inspectors may not always attend
settlement conferences.  The Manager for Engineering and Safety in
the Rail and Water Division of the Iowa Department of Transportation
stated that the state's costs for attending are not reimbursed.  The
FRA Region 6 track specialist said that Iowa state inspectors do not
attend settlement conferences.  Illinois track inspection officials
can attend only settlement conferences held in the state. 


   FRA HAS APPLIED TRACK SAFETY
   STANDARDS MORE UNIFORMLY
---------------------------------------------------------- Chapter 3:2

In July 1990, we reported that FRA and state inspectors had not
applied safety rules and standards uniformly throughout the railroad
industry.  We recommended improved guidance and training to address
these problems.  In response, FRA revamped its safety training
program and issued new enforcement manuals to ensure more uniform
inspections.  As of February 1993, 73 percent of FRA and state track
inspectors had attended the revised classes.  Partly as a result of
these efforts, the range of defects cited per inspection has declined
from 3 to 15 defects per inspection in 1989 to 5 to 8 defects per
inspection in 1993. 


      FRA HAS REVISED ITS TRAINING
      PROGRAM FOR FEDERAL AND
      STATE TRACK INSPECTORS
-------------------------------------------------------- Chapter 3:2.1

In August 1991, FRA hired a Director of Safety Training for the
Office of Safety to develop a training program to improve the
consistency with which inspectors apply the FRA safety standards. 
The trainee orientation program was established first.  This program
trains newly hired inspectors without previous railroad experience in
all five inspection disciplines.  After 3 years, FRA assigns the
trainee to one of the inspection disciplines.  As of February 1993,
25 trainees were participating in this program; 9 trainees were
expected to complete the program by the end of 1993.  According to
FRA officials, since these inspectors will not work independently
until they have completed their training, the full effect of the
trainee orientation program on consistency may not be known for a few
years. 

FRA also revised its training program for track inspectors.  A basic
track course was designed for new hires with previous railroad
experience.  An advanced track course was designed for experienced
FRA and state track inspectors.  In the advanced track course,
participants learn, for example, how to analyze ATIP-generated data
and how to identify defects unique to the structure or geometry of
the track.  Table 3.1 shows that 73 percent of the track inspectors
active as of February 1993 completed one or more of these core
courses during 1991 and 1992. 



                          Table 3.1
           
                 Training Program Attendance



            Basic  Advance                No         Percent
                         d    Both    record        attendin
            track    track  course        of  Tota         g
             only     only       s  training     l     class
----------  -----  -------  ------  --------  ----  --------
FRA             6       27      11      12\a    56        79
 inspectors
State           9       19       8        18    54        67
 inspectors
============================================================
Total          15       46      19        30   110        73
------------------------------------------------------------
\a This number includes two inspectors that FRA hired in 1993. 

Source:  GAO's presentation of data from FRA. 

At the time of our review, FRA had not yet evaluated whether the
revised training programs had improved the consistency with which
inspectors applied the safety standards.  In addition, FRA was not
able to fund the basic and advanced track courses in 1993.  However,
in commenting on a draft of this report, FRA stated that it had
scheduled a basic track course and an advanced track course for
fiscal year 1994.  Approximately one-half of the federal and state
track inspectors will attend these courses.  In addition, FRA has
scheduled regional meetings for federal and state inspectors that
include training in regulatory compliance and enforcement procedures. 


      FRA HAS REDUCED VARIATION IN
      THE APPLICATION OF TRACK
      SAFETY STANDARDS
-------------------------------------------------------- Chapter 3:2.2

We analyzed track inspection results from 1989 to 1992 to determine
whether any change had occurred in inspectors' application of track
safety standards.  We found that the range of defects per inspection
cited by the regions had decreased since 1989, indicating that track
inspectors may have interpreted the track safety standards more
consistently.  We also found less variation in the number of
violation reports filed--a further indication that inspectors may
have applied the standards more consistently. 

To compare the range of defects per inspection cited by FRA's regions
from 1989 to 1992, we divided the total number of track defects each
region found by the total number of track inspections each region
conducted.  Figure 3.1 shows that, from 1989 through 1992, the range
of defects per inspection cited each year narrowed.  In 1989, the
regions cited from 3 to 15 defects per inspection; in 1992, they
cited 5 to 8 defects per inspection. 

   Figure 3.1:  Defects per
   Inspection, 1989-92

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from FRA. 

We also identified the number of violation reports that FRA and state
track inspectors filed from 1989 to 1992 and compared these results
with the results of a similar analysis that we performed for our 1990
report to determine whether any change had occurred in the frequency
with which track inspectors recommended violations as a result of an
inspection.  Figure 3.2 compares the number of violation reports
filed by FRA and state inspectors from 1986 to 1988, as we reported
in 1990, with the number of violation reports filed from 1989 to
1992. 

   Figure 3.2:  Track Inspectors'
   Violation Reports

   (See figure in printed
   edition.)

Source:  GAO's analysis of FRA's violation reports. 

The results of the analysis show that the modal class of track
inspectors--the violation category with the greatest number of track
inspectors--changed from 1 to 3 violation reports to 4 to 10
violation reports.  In other words, the frequency with which FRA and
state track inspectors filed violation reports moved toward a more
normal distribution.  The shift depicted in figure 3.2, from the
irregular distribution for 1986-88 to the bell-shaped (normal)
distribution for 1989-92, may indicate that track inspectors applied
the standards more uniformly during the later period. 


   RAILROADS' TRACK INSPECTION
   RECORDS ARE NOT REVIEWED
---------------------------------------------------------- Chapter 3:3

FRA policy advises FRA and state track inspectors to prepare for an
inspection by reviewing a railroad's track inspection records.  From
1989 through 1992, FRA and state inspectors examined records for 60
percent of the railroads they visited.  As more railroads consolidate
their records at one location, FRA and state inspectors may have more
difficulty reviewing records. 


      INSPECTORS DO NOT ALWAYS
      EXAMINE RAILROADS'
      INSPECTION RECORDS
-------------------------------------------------------- Chapter 3:3.1

Federal regulations require that railroads maintain and inspect their
track in accordance with track safety standards and maintain records
of each inspection.  FRA and state track inspectors monitor
railroads' compliance with track safety requirements.  A railroad's
track inspection records are a good source of information about the
extent to which a railroad has met the regulatory requirements and
about the type and location of the track problems.  Accordingly,
FRA's track enforcement manual recommends that federal inspectors
review these records to prepare effectively for a physical track
inspection.  We analyzed track inspection data for 1989 through 1992
to determine the frequency with which inspectors reviewed records at
the railroads they visited.  FRA inspectors reviewed the records of
1,202, or 60 percent, of the 1,987 railroads they inspected during
the 4-year period.  Table 3.2 shows that although FRA inspected about
31 percent more railroads in 1992 than in 1989, the percentage of
railroads whose records inspectors reviewed declined by 6 percent. 



                          Table 3.2
           
               Inspection of Railroads' Records

                                 Number of
                           railroads whose     Percentage of
            Total number           records   railroads whose
            of railroads        inspectors      records were
Year           inspected          reviewed         inspected
----------  ------------  ----------------  ----------------
1989                 409               270                66
1990                 488               300                61
1991                 553               308                56
1992                 537               324                60
============================================================
Total              1,987             1,202                60
------------------------------------------------------------
Source:  GAO's presentation of data from FRA's Railroad Inspection
Reporting System. 

Regional track specialists stated that inspectors first physically
inspect track and may review the railroad's records afterwards.  The
FRA Region 4 track specialist said that most inspectors probably
review records only when track conditions indicate that the railroad
may not have met standards for frequency of inspection and repair. 
FRA Region 2 officials said that inspecting records is good winter
work when weather conditions are difficult for inspecting track.  The
track training official said that inspectors are more comfortable
working outdoors than in an office searching automated and manual
inspection records. 


      RAILROADS ARE CONSOLIDATING
      AND AUTOMATING THEIR
      INSPECTION RECORDS
-------------------------------------------------------- Chapter 3:3.2

In recent years, many railroads have moved their original track
inspection records to a single location, thereby making it difficult
for FRA and state inspectors to review these records.  In some cases,
the states do not allow their inspectors to travel outside the state
to conduct their inspections.  In addition, both FRA and state
inspectors are further constrained by their understanding that FRA
requires inspectors to review original inspection reports rather than
photocopies. 

FRA and state officials stated that obtaining access to railroads'
original inspection records is a continuing problem because so many
of these records are centrally located.  Some railroads maintain
photocopies of records at local offices where inspectors may review
them.  However, according to FRA officials, inspectors can base
defect and violation reports only on reviews of original documents. 
Therefore, even though photocopies may be available locally,
inspectors need to travel to review original documents before they
can write defect or violation reports, or they need to rely on other
FRA or state inspectors to review original records on their behalf. 

Gaining access to original records is particularly problematic for
state inspectors, who typically cannot travel out of state.  For
example, three times in the last 10 years, Missouri track inspectors
have asked FRA to review the Kansas City Southern's track inspection
records located in Shreveport, Louisiana.  In 1992, FRA responded by
mailing photocopies of all the pertinent records to the Missouri
inspector.  However, as previously stated, the state inspectors
understood that they could not use photocopies as a basis for
recommending defects or violations. 

Recognizing the trend toward consolidating track inspection records
at a few locations, FRA is considering allowing track inspectors to
inspect automated records, thereby allowing access from remote
locations.\2 However, the Director of FRA's Office of Safety stated
that FRA would first need to address certain legal and human resource
issues, such as how FRA would ensure that railroads have established
adequate safeguards against improper access to the system.  In
addition, the Director stated that FRA must determine the amount and
type of training inspectors would require to access and review
records in the system.  Finally, a rail union official said that FRA
would need to ensure that the railroads implemented their automated
record-keeping systems consistently. 

In its written comments on a draft of this report, DOT said that
inspectors may review photocopies of inspection reports and that
these copies are equal in value to the originals.  According to DOT,
FRA will, by technical bulletin, reemphasize the importance of
including record reviews as an integral part of the inspection
process. 


--------------------
\2 If FRA suspected that a railroad had falsified an inspection
report, FRA would need to review the original report. 


SAFETY STANDARDS ON EXCEPTED TRACK
NEED TO BE IMPROVED
============================================================ Chapter 4

Federal track safety standards do not apply to about 12,000 miles of
track designated by the industry as excepted.  In an Advance Notice
of Proposed Rulemaking (ANPRM) issued in November 1992, FRA noted
that it intended to review the excepted track provision and determine
whether the provision should be revised.\1

According to industry officials, the railroads will abandon excepted
track lines if they are required to meet minimum federal safety
standards.  However, we found that the number of reported accidents
on excepted track almost tripled between 1988 and 1992.  During this
same period, the percentage of defects doubled for excepted track
while remaining constant or declining for all other classes of track. 
In addition, FRA accident data showed that industry had not always
complied with regulations restricting the transport of hazardous
materials on excepted track. 


--------------------
\1 In commenting on a draft of this report, FRA stated that by
December 1994 it would issue a Notice of Proposed Rulemaking revising
the safety standards for excepted track. 


   EXCEPTED TRACK IS NOT SUBJECT
   TO TRACK SAFETY STANDARDS
---------------------------------------------------------- Chapter 4:1

Since 1982, FRA has allowed railroads to designate segments of track
as "excepted," or exempt from many track safety standards.  The
designation of a track as excepted is left to the discretion of the
industry.  FRA estimated that the industry had designated about
12,000 miles of track as excepted.  When FRA first adopted the
regulations, it intended that the provision would apply to certain
yard and little-used branch lines on comparatively level terrain. 
FRA also intended that excepted track would not pass through
populated or residential areas where a derailment would endanger
persons along the right-of-way.  FRA also placed restrictions on the
operation of trains over excepted track.  Specifically, (1) trains
may not operate on the track at speeds greater than 10 miles an hour,
(2) passenger trains may not use the track, (3) trains with more than
five hazardous materials cars may not use the track, and (4) trains
carrying hazardous materials may not use the track if it is located
along a public road or highway. 

The requirements for excepted track differ considerably from those
for track in classes 1 through 6.  For example, FRA need not be
notified when a track owner designates track as excepted.  In
contrast, FRA must be notified when a track owner changes the
operating speed for a segment of track, because a change in the
operating speed effectively changes the track's classification--for
example, from a Class 2 track to a Class 1 track.  Moreover, for
excepted track, the railroads do not have to meet many of the track
safety standards that apply to other classes of track.  For example,
they do not have to maintain structures, such as roadbeds, crossties,
rails, and switches, to a prescribed level, and they do not have to
comply with requirements for the geometry of curved track, such as
the distance between the rails (gage), the alinement, and the
elevation of the outer rails. 

The means to correct deficiencies also differ for excepted track and
other classes of track.  FRA and state inspectors are required to
inspect excepted track and note any serious deficiencies on a track
inspection report.  However, because the excepted track is exempt
from the track safety standards, the track owners cannot be compelled
to correct the deficiencies noted.  In addition, FRA inspectors
cannot propose violations for specific track defects; they can
propose violations for operational deficiencies, such as traveling
more than 10 miles per hour and failing to identify excepted track. 
Among the approximately 25,000 defects on excepted track reported
between 1989 and 1992, only 175 (fewer than 1 percent) were
operational deficiencies upon which FRA could propose a violation. 
The remaining 99 percent were track-related defects that railroads
were not required to correct and whose correction FRA could not
require by filing a potential violation. 

FRA's track enforcement manual states that if the condition of the
track continues to constitute a hazard to life and limb and the track
owner fails to alleviate the hazard, the inspector is to notify his
or her supervisor of the problem.  The enforcement manual does not
indicate what further actions the supervisor should take to ensure
the prompt resolution of the critical problems cited by the
inspector. 


   INDUSTRY CITES ADDITIONAL COSTS
   TO UPGRADE EXCEPTED TRACK
---------------------------------------------------------- Chapter 4:2

The Association of American Railroads (AAR) and the American Short
Line Railroad Association stated that it would not be economical for
railroads to upgrade their excepted track to meet minimum federal
safety standards.  The industry designated track as excepted when the
low volume of traffic did not generate sufficient revenues to cover
the cost of rehabilitating the track. 

In a 1993 survey by the American Short Line Railroad Association, 110
railroads estimated that it would cost about $182 million to upgrade
their nearly 4,000 miles of excepted track.  The survey indicated
that rail service would be abandoned on 66 segments of the excepted
track before the railroads would make this investment.  In addition,
the survey noted that the loss of rail service would affect 637
shippers.  According to the Association, railroads will rehabilitate
track to meet minimum safety standards only when sufficient revenues
will be generated to cover the cost of maintaining the track.  In a
similar study in 1993, AAR estimated that it would cost $100 million
to upgrade excepted track to Class 1. 

AAR noted in comments on FRA's ANPRM that it did not endorse the
industry's manipulation of the excepted track rule or any other
section of the regulations to avoid compliance and enforcement.  It
stated that FRA had the necessary enforcement tools to remedy blatant
disregard for the excepted track regulations. 


   THE NUMBER OF ACCIDENTS AND
   DEFECTS ON EXCEPTED TRACK IS
   INCREASING
---------------------------------------------------------- Chapter 4:3

The number of accidents and the number of defects cited by inspectors
is increasing on excepted track.  FRA's accident- reporting system
showed that the number of reported track-caused accidents on excepted
track increased from 22 in 1988 to 65 in 1992--a 195-percent
increase.  Reported accidents on excepted track from all causes
increased from 57 in 1989 to 106 in 1992. 

The number of derailments on excepted track may be understated
because railroads are required to report only accidents that reach a
damage threshold of $6,300.  An FRA Regional Director stated that,
because trains are required to move at slow speeds on excepted track,
many derailments on excepted track will not meet the reporting
criteria.  Even if derailments chronically occur at a location, they
may not show up in the reported data.  Frequent derailments raise the
risk of injury to persons on the train or along the track. 

The number of defects cited on excepted track has also increased. 
Our analysis of data from FRA's Railroad Inspection Reporting System
found that the number of defects cited on excepted track had
increased from 3,229 in 1988 to 6,057 in 1992 (see fig.  4.1). 
Meanwhile, from 1988 to 1992, the number of defects on Class 1 track
decreased from 58,694 to 50,787, and on Class 2 track from 24,839 to
18,960.  During this period, the proportion of defects on excepted
track doubled--from 2.6 to 5.8 percent of the defects on all
track--while the proportion of defects cited on track classes 1
through 6 either remained constant or declined slightly.  On Class 1
track it remained about 50 percent of the defects on all track, and
on Class 2 track it declined from about 20 percent to about 18
percent. 

   Figure 4.1:  Number of Defects
   Cited on Excepted Track

   (See figure in printed
   edition.)

Source:  GAO's presentation of data from FRA's Railroad Inspection
Reporting System. 

However, the number of defects on excepted track may be understated,
since inspectors are required to report only those defects they
consider to be serious.  For example, in September 1992, a track
inspector in Region 4 cited eight defects on a 43-mile section of
excepted track owned by the Indiana Hi-Rail Corporation.  On the
inspection report, the inspector noted that the defects cited were
typical of many other conditions that he had observed on the line
segment but had not specifically cited as defects.  The inspector
also noted that it was difficult to inspect the line because about 50
percent of the line segment was covered by vegetation. 


   FRA CITES ABUSES OF EXCEPTED
   TRACK PROVISION
---------------------------------------------------------- Chapter 4:4

In its ANPRM, FRA stated that railroads had applied the excepted
track regulation far more extensively than the agency had envisioned
in 1982.  FRA cited instances in which railroads were repeatedly
using excepted track to transport hazardous materials through
residential areas or were intentionally designating track as excepted
rather than repairing it to comply with minimum safety standards. 
According to FRA, these designations often occurred after FRA advised
the railroads that they were not in compliance with track safety
standards, even though the railroad should have taken remedial action
before the FRA inspection. 

During workshops on revisions to the track safety standards, FRA
provided additional evidence to support its view that railroads had
abused the excepted track provision.  Workshop participants described
a series of problems with a railroad operating over 400 miles of
track; about 80 percent of the track was classified as excepted. 
Hazardous materials were regularly moved over much of the track.  FRA
had frequently cited the railroad for serious track defects, such as
broken rails, damaged ties, and excessive vegetation on the roadway
on all routes, including those used to move hazardous materials cars. 
Because the railroad was not required to correct the defects on the
excepted track, it often did not do so. 

After a June 1992 derailment that involved hazardous materials cars
and caused two injuries, FRA began a comprehensive safety inspection
of the railroad.  In its December 1992 report, FRA found more than 30
potential violations of safety regulations.  On one excepted track
segment, FRA inspectors identified 22 track defects that posed an
imminent hazard that could lead to a derailment.  FRA did not issue
violations for the track defects because the track was excepted. 
However, FRA proposed eight violations on the basis of 55
accident/injury reporting defects.  For example, one proposed
violation involved the nonreporting of a train derailment because the
derailed cars were not repaired and the railroad did not estimate the
damages.  FRA also found that the railroad had falsified inspection
records and listed clerical staff as engineers. 

FRA safety officials deferred sending the proposed violations to the
Office of Chief Counsel for the assessment of civil penalties,
pending the railroad's response to the safety problems found.  During
a follow-up inspection in June 1993, FRA found some improvements in
the railroad's operating procedures but no improvements in track
conditions.  Some of the continuing track problems were serious.  For
example, FRA found track crosslevel\2

variances of more than 6 inches.  The maximum crosslevel variance
allowed for Class 1 track is only 3 inches.  In 1992, the railroad's
employees cited crosslevel problems as the cause of 11 railcars
overturning during a rerailing operation.  In addition, inspectors
found broken rails, misaligned track and other geometric defects,
excessive vegetation covering the tracks, and substandard tie
conditions.  Despite these problems, FRA could not issue violations
because the track defects occurred on excepted track.  FRA decided in
November 1993 to allow the railroad until the spring of 1994 to
correct the serious safety problems.  At that time, FRA will decide
whether to process the accumulated violations. 


--------------------
\2 A crosslevel variance is a difference in elevation between one
rail and another.  A railroad may, for example, raise the outer rail
on a curve to tip the rail cars inward in order to offset the effect
of centrifugal force. 


CONCLUSIONS, RECOMMENDATIONS, AND
AGENCY COMMENTS AND OUR EVALUATION
============================================================ Chapter 5


   CONCLUSIONS
---------------------------------------------------------- Chapter 5:1

Since we last reported on the weaknesses in FRA's safety inspection
program, the agency has committed significant time and resources to
developing a complex strategy to ensure that its resources are
targeted to the areas posing the highest risk.  As currently
developed, FRA's strategy is a major step in the right direction. 
When fully implemented, the strategy will enable FRA to better target
limited inspector resources and further improve the quality of the
inspection programs.  However, the cornerstone of FRA's new
inspection strategy--the National Inspection Plan--is still an
incomplete safety plan.  The plan's Regional Inspection Points (RIP)
program component does not collect site-specific data for two
critical risk factors--volume of passenger traffic and volume of
hazardous materials traffic.  These gaps in the RIP data weaken FRA's
overall inspection strategy, since the RIP program feeds into the
NIP, the staffing allocation model, and the Quality Improvement
Program.  Similarly, errors in the accident/injury data base weaken
the NIP and, in turn, the other components of the safety enforcement
program.  Accordingly, FRA needs to focus its resources on completing
the RIP program and on improving the reliability of the data on which
the overall effectiveness of the NIP model depends.  These steps
would also reduce the need to artificially adjust NIP goals on the
basis of information from FRA's regional offices. 

The ATIP vehicle offers the Office of Safety a tool to improve its
existing track safety enforcement program.  When the overall goal of
the safety enforcement program is to develop a systematic approach to
targeting inspector resources to high-risk track, the unsystematic
scheduling of the ATIP vehicle is inefficient.  When the planning,
staffing, and evaluation components of the safety enforcement program
are fully operational and effective, the ATIP vehicle could help FRA
supplement inspection time in regions that have inadequate numbers of
federal and state inspectors. 

Although FRA has not yet fully implemented its overall track
inspection strategy, it has enhanced its daily oversight of track
safety.  Communication with participating states has improved
considerably, resulting in the more efficient and effective use of
track inspection resources.  FRA regions have reached agreements with
most states we reviewed either on defining inspection territories or
on otherwise coordinating inspection visits to avoid duplicative
inspections.  Our analysis of track inspections also showed that from
1989 to 1992 FRA had reduced the variation in its inspectors'
application of track safety standards.  FRA's redesigned track safety
training program and other measures may have promoted more uniform
enforcement of track safety standards.  Continued support for the
track safety training program is warranted, even though classes were
not funded in 1993. 

However, some aspects of FRA's daily oversight still need attention. 
FRA and state track inspectors have not implemented FRA policy and
prepared for routine physical inspections by reviewing railroads'
track inspection records to identify areas where defects are likely
to occur.  In addition, opportunities for performing this review have
declined as railroads have centralized and automated their track
inspection records.  Although the Department of Transportation
indicated in its written response to our draft report that it
considers reviews of photocopied or electronic records equal in value
to reviews of original records, the FRA and state inspectors with
whom we spoke were not aware of this policy.  By clarifying this
policy and by issuing a technical bulletin on the importance of
reviewing records before physically inspecting track, FRA could
facilitate and encourage inspectors' compliance with the FRA
regulation requiring record reviews. 

FRA faces a difficult challenge in revising the track safety
standards for excepted track.  Industry associations contend that
railroads would abandon lines if excepted track were required to meet
a Class I standard because the costs to upgrade this track would
exceed the revenues that its use could generate.  Current regulations
restrict inspectors' ability to enforce a minimal level of safety on
excepted track.  While inspectors can cite railroads for violating
operating deficiencies on excepted track, they cannot issue
violations for specific track defects.  Even for track defects that
constitute a safety danger, the track enforcement manual provides
inspectors with no specific guidance on actions they can or cannot
take.  In addition, the railroads are not required to report
designations of excepted track to FRA. 

In our opinion, the increase in the number of accidents and defects,
together with the examples of abuse of the intent of the excepted
track provision cited by FRA, provide strong evidence that changes to
the track safety regulations are needed.  We acknowledge the railroad
industry's concerns about the cost of upgrading excepted track to
Class I standards.  However, we view the revision of guidance on
excepted track as a necessary first step toward helping FRA enforce
the current rule.  Although further restrictions to the excepted
track standards may result from FRA's rulemaking, we believe that, at
a minimum, FRA needs to maintain a higher safety threshold for
excepted track than it is able to do under the existing regulations. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 5:2

To overcome the problems we identified in FRA's rail safety
inspection program and to ensure that the nation's railroad tracks
are safe, we recommend that the Secretary of Transportation direct
the FRA Administrator to take the following actions: 

  To improve the reliability of the NIP data, (1) establish a pilot
     program in one FRA region to gather the data on the volume of
     passenger and hazardous materials traffic needed to complete the
     RIP program and correct the deficiencies in the accident/injury
     data base and (2) clarify the role of the ATIP vehicle and
     target its use to areas with limited federal or state inspector
     resources. 

  Continue to fund training for track inspectors. 

  Clarify the extent to which photocopied or electronic inspection
     records constitute an acceptable basis for reviewing railroads'
     compliance history and writing defect or violation reports, and
     emphasize, through a technical bulletin, the importance of
     reviewing railroads' track inspection records before physically
     inspecting track. 

To strengthen the current regulations governing the excepted track
provision and improve safety on excepted track, we recommend that the
Secretary of Transportation direct the FRA Administrator to take the
following actions: 

  Require railroads to notify FRA when they have classified track
     segments as excepted. 

  Allow railroads to apply the excepted track rule only to track that
     meets the rule's original intent (track located on certain yard
     and little-used branch lines on comparatively level terrain and
     track located outside populated or residential areas). 

  Provide guidance to track inspectors on options available when
     excepted track deficiencies constitute an imminent threat of
     derailment or another safety hazard. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 5:3

DOT provided written comments on a draft of this report.  These
comments and our responses appear in appendix I. 

Overall, DOT appreciated our recognition of the significant strides
that FRA has made in developing and implementing a systematic
approach for allocating its track inspection resources.  DOT
concurred fully with five of our draft report's proposed
recommendations--one on funding training, three on strengthening
excepted track regulations, and one on coordinating track safety
programs--and it concurred in part with our three remaining
recommendations--one on gathering NIP data, one on using the ATIP
vehicle, and one on using photocopied records. 

Specifically, DOT supported the continuation of funding to train
track inspectors, and it agreed with all of our recommendations for
improving safety on excepted track.  DOT noted that FRA would
consider our recommendations on excepted track in preparing a Notice
of Proposed Rulemaking for revising track safety standards.  DOT also
concurred with a proposed recommendation in the draft report that
FRA's Office of Policy and Office of Safety coordinate their
respective NIP and Rail Network programs.  In their comments, DOT
stated that FRA has been coordinating these programs.  They stated
that the Office of Policy is using data provided by the Office of
Safety to build the network.  Similarly, the Office of Safety uses
some of the results of the rail network model in the NIP. 
Accordingly, we have deleted this recommendation from our final
report. 

DOT concurred in part with our proposed recommendations for gathering
data for the NIP on the volume of passenger and hazardous materials
traffic, for clarifying the role of the ATIP vehicle and targeting
its use to areas with relatively few inspectors, and for allowing
inspectors to review photocopied or electronic inspection records and
requiring them to review these or original inspection records before
physically inspecting track.  In addition, DOT commented on our
position that FRA should include state as well as federal track
inspectors in its staffing allocation model--a recommendation that we
first made in 1990. 


      DOT'S COMMENTS AND OUR
      RESPONSE ON IMPROVING NIP
      DATA
-------------------------------------------------------- Chapter 5:3.1

While agreeing with our report's conclusion that FRA could further
improve its track inspection program, DOT argued that gathering and
maintaining data on the volume of passenger and hazardous materials
traffic carried on individual routes would be prohibitively expensive
and was not necessary to deploy track inspectors effectively. 
According to DOT, the data in the NIP are adequate to establish
general inspection goals for states and railroads, and FRA and state
track inspectors have the experience needed to make site-specific
inspection decisions.  Our review showed, however, that the NIP goals
do not give inspectors the information they need to identify
high-risk routes and, as a result, the inspectors question the NIP's
usefulness and reliability.  Thus, despite DOT's arguments, we
continue to believe that additional data are needed to make the NIP
goals meaningful for inspectors in the field.  Nevertheless, we
acknowledge that collecting route-specific data as we originally
recommended may place too great a strain on FRA's resources, and we
have therefore modified our draft report to recommend that FRA test
the collection of site-specific data in one FRA region. 


      DOT'S COMMENTS AND OUR
      RESPONSE ON THE ROLE OF THE
      ATIP VEHICLE
-------------------------------------------------------- Chapter 5:3.2

According to DOT, the role of the ATIP vehicle is to prioritize track
inspectors' activities and not, as we suggest, to supplement
inspections.  Nevertheless, the railroad industry uses ATIP vehicles
to inspect track, and NTSB recommended that FRA use the vehicles to
supplement inspectors' activities.  Furthermore, as we determined
during our review, the ATIP vehicle generates data--including the
number of miles covered and the number of defects identified--that
FRA had not analyzed before we conducted our review but could analyze
routinely to help set inspection priorities. 


      DOT'S COMMENTS AND OUR
      RESPONSE ON THE USE OF
      PHOTOCOPIED REPORTS
-------------------------------------------------------- Chapter 5:3.3

In responding to our recommendation that FRA establish a policy
allowing FRA and state inspectors to review photocopied or electronic
records of track inspection reports, DOT stated that FRA inspectors
may review photocopied reports, which the agency considers equal in
value to original reports.  However, DOT did not say whether it
considered photocopied inspection reports acceptable as a basis for
writing defect or violation reports.  Since the FRA inspectors we
talked to understood that they were required to review original
inspection reports before writing defect or violation reports, we
believe that FRA needs to clarify the extent to which its inspectors
can rely on photocopied or electronic inspection reports. 

While agreeing with us that inspectors can derive useful information
about a railroad's compliance history by reviewing inspection reports
before physically inspecting track, DOT indicated that such reviews
were not always practicable because records of inspections were not
maintained at all locations where inspections of track were required. 
DOT further indicated that FRA would issue a technical bulletin
reemphasizing the integral role of record reviews in the inspection
process.  We concur with FRA's approach and have modified the
recommendation in our draft report to reflect our concurrence. 


      DOT'S COMMENTS AND OUR
      RESPONSE ON THE USE OF STATE
      RESOURCES
-------------------------------------------------------- Chapter 5:3.4

Finally, DOT commented on our suggestion that FRA include state
inspectors in its staffing allocation model and allocate its own
inspectors as needed to provide complete and uniform coverage
throughout a region.  In DOT's view, FRA is itself obliged to inspect
track throughout a region, without reference to states' efforts. 
Furthermore, according to DOT, an allocation policy such as we
suggested would lead FRA to assign federal inspectors to
nonparticipating states so that participating states would, in
effect, be subsidizing nonparticipating states.  Participating states
would then be likely to withdraw from the program, and their
resources would no longer be available to support FRA's inspection
efforts. 

We agree with DOT that FRA needs to ensure that railroads are
maintaining the quality of their track nationwide.  We do not agree
that some states may perceive the inclusion of state inspectors in
FRA's staffing allocation model as leading ultimately to the
disproportionate application of federal inspection resources in some
nonparticipating states.  States have established their own
inspection plans because they want independent assurance that the
industry is complying with track safety standards.  If FRA were to
assign fewer federal inspectors to inspect track in a state, it is
unlikely that the state would respond by reducing its own inspection
coverage.  Moreover, FRA's goal is to provide for uniform inspection
coverage.  The current inspection plan, which does not include state
inspectors, could establish disproportionately higher coverage in
states that provide state inspectors. 




(See figure in printed edition.)Appendix I
COMMENTS FROM THE DEPARTMENT OF
TRANSPORTATION
============================================================ Chapter 5



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on the Department of
Transportation's letter dated February 8, 1994. 


   GAO COMMENTS
---------------------------------------------------------- Chapter 5:4

1.  We agree with the overall framework that DOT has established to
respond to the Federal Railroad Safety Authorization Act and our past
reports.  We agree with DOT that FRA needs to move toward a more
quantitative approach to inspections yet still rely on the sound
judgment of its inspectors to make site-specific inspection
decisions.  However, a premise underlying DOT's overall framework is
that inspectors need better data to identify which routes pose the
highest safety risks.  The formulas in the NIP repeatedly emphasize
two critical risk factors--the volume of passenger and the volume of
hazardous materials traffic.  Both FRA and the industry want to
ensure that routes carrying these types of traffic are adequately
maintained to prevent accidents that will injure passengers or expose
populated areas to chemical risks.  However, the NIP goals do not
give FRA inspectors the information they need to identify these
high-risk routes.  As we stated in our report, regional
inspectors--the intended users of the inspection goals--have
questioned the usefulness and reliability of the NIP output because
the goals have not matched the inspectors' professional knowledge or
have not reflected the adjustments requested by the regions.  If the
intended users of the NIP goals question their credibility, then the
goals need to be improved. 

The difference between our views and DOT's centers on the type and
amount of data needed to render the NIP goals useful to the
inspectors in the field.  While DOT sees no need for change, we
believe that additional and more accurate data are needed to prevent
the expenditure of scarce agency resources to generate goals that
inspectors do not always use.  We acknowledge that collecting
site-specific data in all FRA regions, as we originally recommended,
could stretch the agency's resources.  We have, therefore, modified
our draft report to recommend that, at a minimum, FRA establish a
pilot program to collect site-specific data in one FRA region.  From
its experience in one region, FRA could then extrapolate the costs of
collecting site-specific information nationwide.  As part of this
pilot program, FRA could ask regional inspectors to indicate (1) what
information would best help them target inspections to high-risk
track and (2) how they could collect site-specific data.  In
addition, DOT's Bureau of Transportation Statistics could assist FRA
in defining cost-effective methods for collecting and analyzing the
needed information.  We believe that these collective efforts and
pilot results could give FRA answers to the questions raised in our
report. 

2.  We agree that a stable federal inspection presence is needed to
ensure that railroads are maintaining the quality of the nation's
tracks.  We do not agree that some states may perceive the inclusion
of state inspectors in FRA's staffing allocation model as leading
ultimately to the disproportionate application of federal inspection
resources in some nonparticipating states.  States have established
their own inspection plans because they want independent assurance
that the industry is complying with track safety standards. 
Moreover, FRA's argument needs to be weighed against two key factors. 
First, the scope of FRA's inspection framework is national; including
all resources--both state and federal--in the staffing allocation
model would provide a clearer picture of the areas needing attention. 
Second, the feasibility of shifting federal inspectors to states that
do not have comparable state inspectors has already been demonstrated
by FRA's Region 3.  As noted in the report, Region 3 shifted federal
inspections from Florida, which had three state inspectors, to
Georgia and Kentucky, which had no state inspectors.  We believe that
Region 3's flexible approach logically extends FRA's progress in
coordinating federal and state inspections.  In our view, integrating
inspection resources would further enhance communication and
coordination between federal and state regulators. 

3.  DOT concurred with our proposed recommendation that FRA's Office
of Policy and Office of Safety coordinate their respective NIP and
Rail Network programs.  In response to DOT's assurance that FRA has
been coordinating these programs, we deleted this recommendation from
our final report. 

4.  DOT agreed with all of our recommendations for improving safety
on excepted track.  DOT noted that FRA would consider our
recommendations on excepted track in preparing a Notice of Proposed
Rulemaking for revising track safety standards. 

5.  We have added information in the report to cite the additional
actions that FRA has taken to improve the accuracy of the railroads'
accident and injury reporting. 

6.  We based our recommendation that the ATIP vehicle be used to
supplement inspections on three facts:  (1) the industry uses ATIP
vehicles to conduct inspections, (2) NTSB concluded that ATIP
vehicles should be used to supplement inspectors' activities, and (3)
the data obtained from the ATIP vehicle are clearly underutilized. 
The last point, in particular, is based on evidence we gathered
during our review.  For example, the statistics cited in the report
on the number of defects the ATIP vehicle identified in 1991 and the
number of miles it covered were generated by FRA only after we
requested this information.  Until we made our request, FRA had not
analyzed ATIP data--data that conceivably could help it set
inspection priorities.  In addition, as we stated in the report, the
NIP, staffing allocation model, and Quality Improvement Program
provide the framework for setting FRA's inspection priorities. 
However, we found no evidence that describes how the ATIP vehicle
fits into this new inspection framework.  FRA spends approximately
$500,000 each year to support the ATIP contract.  At a minimum, using
the vehicle to supplement the work of the agency's track inspectors
would provide the agency with some concrete benefits. 

7.  We have updated the report to indicate that, for fiscal year
1994, FRA will provide track training courses. 

8.  Since the FRA inspectors we talked to understood that they were
required to review original inspection reports before writing defect
or violation reports, we believe that FRA needs to clarify the extent
to which its inspectors can rely on photocopied or electronic
inspection reports.  DOT further indicated that FRA would issue a
technical bulletin reemphasizing the integral role of record reviews
in the inspection process.  We concur with FRA's approach and have
therefore modified our proposed recommendation to reflect our
concurrence. 

9.  We have deleted the reference to the staffing allocation model as
an incomplete system.  FRA had stated on several occasions that the
model would not be completed until the NIP was complete.  Now that
the model is complete, FRA has the opportunity to integrate state
inspection resources into its staffing allocation decisions. 

10.  We have changed the discussion of the waybill sample to show how
FRA used the ICC information in the NIP. 

11.  We removed information on the Illinois Central railroad, since
this information was generated using a data base that FRA
subsequently corrected in response to findings in our draft report. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix II


   RESOURCES, COMMUNITY, AND
   ECONOMIC DEVELOPMENT DIVISION,
   WASHINGTON, D.C. 
-------------------------------------------------------- Appendix II:1

Allen Li, Associate Director
Ronnie E.  Wood, Assistant Director
Deborah L.  Justice, Assignment Manager
Marianne E.  Bradshaw, Staff Evaluator
Elizabeth R.  Eisenstadt, Reports Analyst


   CHICAGO REGIONAL OFFICE
-------------------------------------------------------- Appendix II:2

Joseph A.  Christoff, Assistant Director
Melvin J.  Koenigs, Evaluator-in-Charge
Bonnie M.  Pignatiello, Staff Evaluator
David I.  Lichtenfeld, Site Senior
Ruthann R.  Balciunas, Technical Adviser
John Zarem, Technical Adviser

RELATED GAO PRODUCTS

Railroad Safety:  Weaknesses in FRA's Safety Program
(GAO/T-RCED-91-32, Apr.  11, 1991). 

Railroad Safety:  Weaknesses Exist in FRA's Enforcement Program
(GAO/RCED-91-72, Mar.  22, 1991). 

Financial Management:  Internal Control Weaknesses in FRA's Civil
Penalty Program (GAO/RCED-91-47, Dec.  26, 1990). 

Railroad Safety:  FRA's Staffing Model Cannot Estimate Inspectors
Needed for Safety Mission (GAO/RCED-91-32, Nov.  21, 1990). 

Improvement Needed in FRA's Safety Inspection Program
(GAO/T-RCED-91-2, Oct.  5, 1990). 

Railroad Safety:  New Approach Needed for Effective FRA Safety
Inspection Program (GAO/RCED-90-194, July 31, 1990). 

Railroad Safety:  More FRA Oversight Needed to Ensure Rail Safety in
Region 2 (GAO/RCED-90-140, Apr.  27, 1990). 

Improvements Needed in FRA's Hazardous Materials Inspection and
Safety Reporting Programs (GAO/T-RCED-90-35, Feb.  28, 1990). 

Railroad Safety:  DOT Should Better Manage Its Hazardous Materials
Inspection Program (GAO/RCED-90-43, Nov.  17, 1989). 

Improvements Needed in DOT's Hazardous Materials Rail Safety Program
(GAO/T-RCED-90-13, Nov.  7, 1989). 

Railroad Safety:  FRA Needs to Correct Deficiencies in Reporting
Injuries and Accidents (GAO/RCED-89-109, Apr.  5, 1989). 

Rail Safety:  States' Reaction to Proposed Elimination of Inspection
Funding (GAO/RCED-87-84FS, Feb.  6, 1987). 

