Southeastern United States fatal crash study / prepared by Karen Dixon

FINAL SUMMARY REPORT
Southeastern United States Fatal Crash Study
Prepared By: Karen Dixon, Ph.D., P.E.
The Georgia Institute of Technology
October 2005
Prepared for the Georgia Department of Transportation

TECHNICAL REPORT STANDARD TITLE PAGE

1.Report No.: FHWA-GA-05-9815

2. Government Accession No.:

3. Recipient's Catalog No.:

4. Title and Subtitle: Southeastern United States Fatal Crash Study

5. Report Date: October 2005

6. Performing Organization Code:

7. Author(s): Karen Dixon, Ph.D., P.E. Georgia Institute of Technology
9. Performing Organization Name and Address: Georgia Institute of Technology School of Civil and Environmental Engineering Atlanta, GA 30332-0355

8. Performing Organ. Report No.: 9815
10. Work Unit No.:
11. Contract or Grant No.:

12. Sponsoring Agency Name and Address: Georgia Department of Transportation Office of Materials and Research 15 Kennedy Drive Forest Park, GA 30297-2534

13. Type of Report and Period Covered: Final; 1999-2003
14. Sponsoring Agency Code:

15. Supplementary Notes: Prepared in cooperation with the U.S. Department of Transportation, Federal Highway Administration.

16. Abstract: A significant safety issue in the United States is the substantial number of vehicle related crashes. The number of fatal
crashes in the southeastern portion of the U.S. (States of Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, and Tennessee) is disproportionately higher than those for the entire country. In general, the eight states collectively report approximately 26-percent of the total annual number of fatal automobile-related crashes in the U.S. On average, the southeastern states experience an additional 30 fatalities per million vehicle miles traveled than the U.S. average. The Federal Highway Administration and the eight southeastern states initiated a joint research effort for the region to study this observed over-representation of fatal crashes.
Findings of the study suggest that improved features such as widening shoulders, enhancing delineation, and protecting the clear zone would substantially reduce these fatal crashes. Some of the researchers recommended that additional procedures and policies may be an appropriate countermeasure for wide-scale improvements. Countermeasures (physical as well as political) were explicitly recommended to address two-lane rural roads, safety restraint use and fixed-object crashes.
A supplemental finding was the presence of extensive pavement edge drop-offs for fatal crash sites in at least two of the participating states. As this observation occurred as a result of field inspection and was not initially identified as a target problem, it was not studied in great detail for this research effort but merits special comment since it is potentially a significant finding of the study.

17. Key Words:

18. Distribution Statement:

19. Security Class. (report): Unclassified
Form DOT 1700.7 (8-69)

20. Security Class. (page): 21. Number of Pages:

Unclassified

53

i

22. Price:

TABLE OF CONTENTS
List of Tables ...................................................................................................................... ii List of Figures ..................................................................................................................... ii Executive Summary ............................................................................................................ 1 Introduction......................................................................................................................... 2 Project Objectives ............................................................................................................... 4 Research Administration and Participants .......................................................................... 6 Study Focus......................................................................................................................... 7
Two-lane Rural Roads .................................................................................................... 8 Safety Restraint Use........................................................................................................ 9 Driver Education and Licensing ................................................................................... 11 Commercial Vehicle Operations................................................................................... 12 Fixed-object Crashes..................................................................................................... 13 Speeding........................................................................................................................ 13 Supplemental Finding Pavement Edge Drop-offs ......................................................... 15 Project Status .................................................................................................................... 17 Alabama ........................................................................................................................ 18 Florida ........................................................................................................................... 19 Georgia.......................................................................................................................... 19 Kentucky ....................................................................................................................... 20 Mississippi .................................................................................................................... 21 North Carolina .............................................................................................................. 21 South Carolina .............................................................................................................. 22 Tennessee...................................................................................................................... 23 Summary of Findings and Conclusions ............................................................................ 24 Future Research ................................................................................................................ 24 References......................................................................................................................... 25 Appendix........................................................................................................................... 26
i

LIST OF TABLES
Table 1. Southeast and United States Fatal Crash Summary (1996 2003) ..................... 3 Table 2. Southeast and United States Fatality Rates (1996 2003).................................. 3 Table 3. Participating States and Researchers ................................................................... 7 Table 4. Percent Rural Crash Locations (1996 2003)..................................................... 8 Table 5. Rural Two-Lane Crash Percentage for 2000 ....................................................... 9 Table 6. Percent Fatal Crash Drivers Not Utilizing Safety Restraints (1996 2003) ..... 10 Table 7. Young Drivers ( 20 Years Old) and Crash Percentage for 2000 .................... 11 Table 8. Percent Large Trucks in Fatal Crashes (1996 2003)....................................... 12 Table 9. Percent Speeding-Related Fatalities (1996 2003)........................................... 14 Table 10. Project Status Summary Table......................................................................... 18
LIST OF FIGURES
Figure 1. Crash Direction for 54 Georgia Non-State-System Sites ................................. 16
ii

Executive Summary
This study evaluated the high number of fatal crashes in the southeastern United States in an effort to determine what may be unique to the region resulting in this disproportionate statistic. This eight state region selected a research agency to help with their respective state evaluation. Each participating state research team identified unique issues appropriate for their jurisdiction. Due to the variety of analysis procedures and identified objectives from their respective state representatives, the individual conclusions dramatically varied for each state. In general, seven of the eight states determined that the rural, two-lane road condition is the source of the elevated fatal crashes in the region. Improved features such as widening shoulders, enhancing delineation, and protecting the clear zone were identified consistently during the countermeasure analyses. Some of the researchers recommended to their state representatives that additional procedures and policies may be an appropriate countermeasure for wide-scale improvements. Of the six perceived topical areas identified at a kickoff meeting involving all eight states, countermeasures (physical as well as political) were explicitly recommended to address two-lane rural roads, safety restraint use and fixed-object crashes.
A supplemental finding for this study was the presence of extensive pavement edge drop-offs for fatal crash sites in at least two of the participating states. As this observation occurred as a result of field inspection and was not initially identified as a target problem, it was not studied in great detail for this research effort but merits special comment since it is potentially a significant finding of the study.
1

Introduction
A significant safety issue in the United States is the substantial number of vehicle related crashes. In particular, death due to injuries sustained in an automobile crash is the leading cause of death for persons between the ages of 2 and 33 years old (1).
The number of fatal crashes in the southeastern portion of the United States is disproportionately higher than those for the entire country. Table 1 depicts an eight year summary of the number of fatal crashes for the eight southeastern states of Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, and Tennessee. In general, the eight states collectively report approximately 26-percent of the total annual number of fatal automobile-related crashes in the United States. Table 2 includes the individual state fatality rates from 1996 to 2003. On average, the southeastern states experience an additional 30 fatalities per million vehicle miles traveled than the United States average. The Federal Highway Administration (FHWA) and the eight southeastern states initiated a joint research effort for the region to study this observed over-representation of fatal crashes.
This study is complete and this summary report provides an overview of the study participants, their role in the project, and the varying results available.
2

Table 1. Southeast and United States Fatal Crash Summary (1996 2003)

State

1996

1997 1998 1999 2000

2001 2002 2003

8 Year Total

AL

1,024 1,050 958 992 910 900 931 899

7,664

FL

2,496 2,528 2,548 2,629 2,733 2,714 2,810 2,874 21,332

GA

1,402 1,405 1,414 1,314 1,380 1,471 1,362 1,463 11,211

KY

734 774 766 724 721 762 810 845

6,136

MS

695 741 842 832 846 704 769 786

6,215

NC

1,329 1,290 1,433 1,350 1,408 1,360 1,427 1,375 10,972

SC

821 798 912 944 948 962 949 904

7,238

TN

1,120 1,104 1,110 1,169 1,177 1,126 1,058 1,091 8,955

SE Sub-Total

9,621 9,690 9,983 9,954 10,123 9,999 10,116 10,237

79,723

Total U.S.* 37,494 37,324 37,107 37,140 37,526 37,862 38,491 38,252 301,196

Percent of U.S.* Total

25.7%

26.0%

26.9%

26.8%

27.0%

26.4%

26.3%

26.8%

26.5%

* U.S. values include the 50 states and the District of Columbia

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

Table 2. Southeast and United States Fatality Rates (1996 2003) Fatalities are Shown per 100 Million Vehicle Miles Traveled

State

1996 1997 1998 1999 2000 2001 2002 2003

8 Year State Average

AL

2.2 2.2 1.9 2.0 1.8 1.8 1.8 1.7

1.9

FL

2.1 2.1 2.1 2.1 2.0 1.9 1.8 1.7

2.0

GA

1.8 1.7 1.6 1.5 1.5 1.5 1.4 1.5

1.6

KY

2.0 1.9 1.8 1.7 1.8 1.8 2.0 2.0

1.9

MS

2.7 2.7 2.8 2.7 2.7 2.2 2.4 2.3

2.6

NC

1.9 1.8 1.9 1.7 1.6 1.7 1.7 1.6

1.7

SC

2.3 2.2 2.3 2.4 2.3 2.3 2.2 2.0

2.3

TN

2.1 2.0 1.9 2.0 2.0 1.9 1.7 1.7

1.9

SE

Weighted 2.1 2.0 2.0 1.9 1.9 1.8 1.8 1.7

1.9

Average**

U.S.* Average

1.7 1.6 1.6 1.5 1.5 1.5 1.5 1.5

1.6

* U.S. values include the 50 states and the District of Columbia ** Weighted Average is the Entire Number of Fatalities (all eight states) divided by the
Entire Number of 100 Million Vehicle Miles Traveled (all eight states)

Source: National Center for Statistics & Analysis, Traffic Safety Facts State Traffic Data (1996 through 2003)

3

Project Objectives
This research project had one overall goal: To quantify the influence of various statistically significant factors contributing to fatal crash occurrence through coordinated in-depth studies in the eight southeast states of Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, and Tennessee.
On November 17-18, 1998 participants and representatives from each state convened at a kick-off meeting in Georgia to review proposed research objectives, discuss the potential framework of the studies, and define specific topical areas. This meeting included extensive discussion about the targeted study focus for individual states. Each state has unique problems and the research performed by each research group needed to address the issues identified by the state for which that group was to perform the research effort.
Two overall research objectives were identified to help guide the individual development of the research scope for each state. These two general objectives were:
To identify and quantify the impact of the "top 6 safety concerns" (or a subset thereof) within the states, and
To identify countermeasures for reducing fatal crashes and/or quantify the effect of various countermeasures (for the top 6) when and where possible.
The meeting participants also discussed a general framework for the research effort. Though each state transportation department directed the required research effort for that state and incorporated unique goals and content for their research, in general the kick-off meeting participants identified four potential research tasks suitable for a broad safetyoriented research framework. These four research tasks included:
4

1. Regional Fatal Crash Description. This effort was aimed at quantifying road safety differences between the southeastern (SE) and nonsoutheastern United States. The purpose of this task was to identify and quantify over-representation of crashes in the SE (by certain characteristics), and to identify the largest raw numbers of crash types. This research effort helped identify and quantify high safety concern topical areas for reference and comparison purposes.
2. Fatal Crash Causal Analysis of Two-Lane Rural Roads. This effort involved the cooperation of all participating states for data collection, and involved one or more of the research teams for analysis efforts. The state representatives discussed evaluating a statistically random sample of 150 fatal crashes for this task.
3. Countermeasure Identification. The purpose of this effort was to identify and carefully review past literature of countermeasure effectiveness relevant to SE safety concerns. The focus was to synthesize the results into succinct and substantive practical results.
4. Targeted Before-After Studies. The purpose of this effort was to conduct retroactive or new before-after studies to evaluate specific countermeasure effectiveness.
In general, the top 6 topical areas discussed by the participants included: Two-lane rural roads, Safety restraint use,
5

Driver education and licensing, Commercial vehicle operations, Fixed-object crashes, and Speeding.
Not all state representatives felt all 6 topical areas were relevant to their specific jurisdiction. Following the kick-off meeting, each state and respective research team met to discuss their specific research questions, level of participation in the effort, and proposed final product. The appendix includes summary slides from the kick-off meeting, an example of slides for one of the state briefings (Georgia), and two sets of slides summarizing the proposed research evaluation methodologies. These evaluation methods included regional fatal crash summary statistics, fatal crash causal analysis for a collection of randomly selected fatal crashes, countermeasure identification, and a targeted before-after study using expert evaluation combined with statistical analysis.
Research Administration and Participants
The FHWA Atlanta Resource Center provided project oversight and guidance under the direction of Mr. Frank Julian. The FHWA coordinated with representatives of the eight southeastern states individually to help organize and coordinate the pooled-fund research efforts. Each state department of transportation then contracted with a local university to perform the associated research activity. The participating states and university contacts are summarized in Table 3.
6

State Alabama Florida Georgia
Kentucky Mississippi North Carolina South Carolina Tennessee

Table 3. Participating States and Researchers

University
Auburn University (with sub-contract to University of Alabama) University of South Florida Center for Urban Transportation Research Georgia Institute of Technology
University of Kentucky Transportation Center
Mississippi State Transportation Research Center University of North Carolina Highway Safety Research Center Clemson University
University of Tennessee

Individuals Responsible for Research
Dr. Brian L. Bowman (Auburn), Dr. David Brown (University of Alabama) Ms. Patricia Turner
Dr. Simon Washington (now with Arizona State University) and Dr. Karen Dixon (now with Oregon State University) Mr. Kenneth R. Agent Mr. Jerry G. Pigman Dr. Nikiforos Stamatiadis Dr. James W. Epps (retired)
Mr. James K. Lacy (now with North Carolina Department of Transportation) Dr. David B. Clarke (now with University of Tennessee) Mr. Matthew Cate

Study Focus
To further narrow the scope of the study, researchers at the Georgia Institute of Technology evaluated the distribution of the fatal crashes to determine if a specific road type, crash type, or location may be appropriate for a targeted study. Georgia Tech team members also specifically evaluated, where feasible, the six perceived topical areas discussed at the project kick-off meeting.

7

Two-lane Rural Roads

Table 4 demonstrates that approximately 64-percent of the southeastern crashes occurred at rural locations (average for years 1996 to 2003). It is important to note that the eight participating states have varying definitions of "rural" and "urban" unique to each state, so this variable is restricted to the state-by-state rural designation. In addition, the state of Florida experienced a substantially smaller number of rural crashes (approximately 45percent) than the remaining seven southeastern states. As shown in Table 4, if Florida is removed from the rural analysis, the total average percent of rural crashes for the other southeastern states is approximately 71-percent compared to the United States value of approximately 59-percent.
Table 4. Percent Rural Crash Locations (1996 2003)

8 Year

1996 1997 1998 1999 2000 2001 2002 2003 Average

State

[%] [%] [%] [%] [%] [%] [%] [%]

[%]

AL

67.9 62.4 63.6 69.9 69.6 70.0 73.8 66.6

68.0

FL

43.3 44.2 44.8 45.5 46.8 46.3 44.0 48.5

45.4

GA

58.3 57.2 59.3 58.3 55.5 56.3 59.9 55.6

57.6

KY

75.3 79.8 77.7 80.0 75.6 75.3 75.7 77.3

77.1

MS

83.3 98.8 98.3 99.0 99.6 99.7 99.0 75.3

94.1

NC

69.8 68.3 63.9 60.4 63.2 67.6 72.0 73.2

67.3

SC

91.2 89.3 84.4 85.4 87.1 89.1 89.1 89.6

88.2

TN

60.8 57.6 62.7 63.6 65.3 63.9 62.5 58.6

61.9

Percent Rural

for All SE 63.2 63.5 64.0 64.5 64.7 64.9 65.7 63.6

64.3

Crashes

Percent Rural

for SE Crashes 70.2 70.3 70.6 71.4 71.3 71.8 74.1 69.5

71.2

(excluding FL)

Percent Rural

for All U.S.* 57.1 58.4 59.4 60.0 58.1 58.6 59.1 58.5

58.7

Crashes

* U.S. values include the 50 states and the District of Columbia

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

8

Table 5 depicts crashes for the year 2000 and further emphasizes the disparity

between the Florida crash locations and those for the remaining seven states. Whereas

only 38.2-percent of the Florida crashes occurred at two-lane rural roads in the year 2000,

the percentage occurring for similar roads in the other southeastern states ranged from

51.2-percent (Georgia) up to 77.7-percent (Mississippi). Due to the high representation

of crashes on two-lane rural roads in seven of the states, the FHWA and state

representatives chose to narrow the evaluation to fatal crashes on rural two-lane roads.

As may be expected, the State of Florida elected to withdraw from the rural two-lane

study and perform an independent study relevant to their specific safety concerns.

Table 5. Rural Two-Lane Crash Percentage for 2000

Number Rural

Percent Rural

Two-Lane Road Total Number of

Two-Lane

State

Crashes

Crashes

Crashes

AL

607

910

66.7%

FL

1,044

2,733

38.2%

GA

707

1,380

51.2%

KY

537

721

74.5%

MS

657

846

77.7%

NC

778

1,408

55.3%

SC

705

948

74.4%

TN

733

1,177

62.3%

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

Safety Restraint Use The proper use of restraint systems reduces the likelihood of fatal injury to front-
seat car occupants by 45 percent (3). In general, states with primary seat belt laws experience higher seat belt usage rates than those with secondary laws. In the southeast, the states of Alabama (enacted 1999), Georgia (enacted 1996), and North Carolina

9

(enacted 1985) have a primary seat belt law. On July 1, 2004 the State of Tennessee enacted a primary seat belt law.
Table 6 shows the percent of drivers involved in fatal crashes who did not utilize safety restraint systems. The National Highway Safety Administration estimated overall shoulder seat belt use rates of 79 percent for 2003, 71 percent in 2000, and 67 percent in 1999 (3). On average, approximately 35.7 percent of the U.S. drivers involved in fatal crashes for an eight year period did not use their safety restraints. The states with a primary seat belt law demonstrated higher use by drivers of safety restraints than did the states without a similarly enforceable law. The eight southeastern states collectively exhibited almost seven percent more fatal crashes than the national average for drivers who did not wear the required safety restraints.

Table 6. Percent Fatal Crash Drivers Not Utilizing Safety Restraints (1996 2003)

8 Year

1996 1997 1998 1999 2000 2001 2002 2003 Average

State

[%] [%] [%] [%] [%] [%] [%] [%]

[%]

AL

53.9 49.1 49.2 49.2 44.6 35.9 41.5 37.7

45.1

FL

37.1 35.7 36.7 36.8 40.0 36.9 36.3 35.3

36.9

GA

42.1 39.5 31.9 34.7 33.2 28.9 29.6 29.4

33.7

KY

52.7 49.7 49.7 52.2 48.2 49.0 48.4 46.8

49.6

MS

67.3 61.0 58.5 60.8 58.1 52.7 53.8 55.9

58.5

NC

26.1 26.8 23.5 25.4 28.5 26.8 29.3 23.7

26.3

SC

42.3 40.4 41.6 43.7 42.5 43.9 43.5 42.6

42.6

TN

52.7 51.0 51.8 51.8 52.1 46.9 45.1 43.5

49.4

Percent for

All SE

46.8 44.2 42.9 44.3 43.4 40.1 40.9 39.4

42.8

Crashes

Percent for

All U.S.* 38.5 37.8 36.6 37.1 35.4 34.3 34.2 32.0

35.7

Crashes

*U.S. values include the 50 states and the District of Columbia

Note: Shaded regions represent primary seat belt laws in effect for the entire year.

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

10

Driver Education and Licensing Many crashes are attributed to the age or experience of the driver. Approximately
6.4 percent of the licensed drivers in the United States are between the ages of 15 and 20 years old, yet 14 percent of drivers involved in fatal crashes in the United States were in this age group in 2003. In fact, approximately 18 percent of all police-reported crashes involve young drivers (4). Table 7 demonstrates a typical distribution of young drivers in the southeastern states for the year 2000. In general, these states have fewer young drivers involved in fatal crashes than the entire United States; however, the number of young drivers is still disproportionate to the number of licensed drivers in this same age group. Strategies for reducing this over-representation may include expanded driver education or modified driver license procedures where the drivers gradually receive increased levels of responsibility before receiving an unrestricted driver's license.

Table 7. Young Drivers ( 20 Years Old) and Crash Percentage for 2000

Number Young Total Number of Percent Young

Drivers involved in Drivers in Fatal Drivers in Fatal

State

Fatal Crashes

Crashes

Crashes

AL

215

1,363

15.8%

FL

559

4,266

13.1%

GA

286

2,149

13.3%

KY

174

1,082

16.1%

MS

188

1,236

15.2%

NC

275

2,162

12.7%

SC

160

1,411

11.3%

TN

259

1,741

14.9%

Note: Shaded regions indicate states with Intermediate Stages in licensing procedures.

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

and 2002 Insurance Institute for Highway Safety, Highway Loss Data Institute

11

Commercial Vehicle Operations In the United States one out of every nine traffic fatalities results from a collision
with a large truck. The extreme differences in vehicle size result in the vehicle other than the truck sustaining considerable damage in a collision. In addition, approximately 75 percent of the fatalities involving large trucks are occupants of the vehicle other than the truck (6). Table 8 depicts the number of large trucks involved in fatal crashes for the southeastern states during the eight year period from 1996 to 2003. As shown in the table, large truck crashes in the southeast occur, on average, at a rate similar to that of the entire United States.

Table 8. Percent Large Trucks in Fatal Crashes (1996 2003)

8Year

1996 1997 1998 1999 2000 2001 2002 2003 Average

State [%] [%] [%] [%] [%] [%] [%] [%] [%]

AL

9.0 10.2 10.2 9.4 11.2 10.5 9.0 10.7 10.0

FL

7.1 7.2 7.7 7.8 7.1 7.7 7.9 7.7

7.5

GA

9.9 10.0 8.9 10.5 9.6 9.8 9.2 9.1

9.6

KY

8.2 9.3 8.1 8.8 8.9 8.2 9.3 9.1

8.7

MS

8.5 8.9 8.3 9.1 9.5 8.4 6.5 5.9

8.1

NC

8.0 9.7 10.3 8.9 8.3 8.9 7.7 7.4

8.7

SC

8.0 7.5 8.7 8.7 6.1 7.6 6.6 7.1

7.5

TN

9.7 7.7 7.9 9.4 9.0 7.5 8.3 7.0

8.3

Percent for

All SE 8.4 8.6 8.7 8.9 8.4 8.5 8.1 8.0

8.5

Crashes

Percent for

All U.S.* 8.3 8.5 8.7 8.6 8.6 8.3 7.8 8.0

8.4

Crashes

* U.S. values include the 50 states and the District of Columbia

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

12

Fixed-object Crashes The high percent of crashes on rural two-lane highways in the southeast are
represented by numerous run-off-the-road crashes. In general, the rolling terrain in the southeast coupled with narrow right-of-way, adjacent wooded areas, and extreme horizontal road curvature combine to result in a large number of fixed-object crashes. Fixed objects can include trees, utility poles, walls, and other rigid items including roadside barrier. The designation in a police crash report for impact with a fixed object is either a "first harmful" event or a "most harmful" event and is based on the reporting police officer's interpretation of the crash condition. As a result, the data reported in fatal crash databases is, at best, subjective. For this reason, database summaries for fixedobject crashes can provide strong indications of fixed object problems but should not be used as a definitive indicator of this type of problem. The best (and certainly more costly) method to evaluate fixed object crashes is by physical site examination combined with a critical review of the individual crash report.
Speeding Driving too fast for appropriate road conditions is a common cause for crashes.
Speeding can create a serious problem for the single-car crash because it is difficult to correct the direction of errant vehicles at high speeds. In multi-car crashes, the larger speed differential contributes to the crash severity. NHTSA estimates that speeding is a contributing factor in approximately 31 percent of all fatal crashes (7). Unfortunately, many state crash reports do not have an appropriate method for determining pre-crash speed. In Georgia, for example, there is no requirement to report estimated vehicle speed
13

but instead the reporting officer must indicate posted speed limit (which is unlikely to represent the travel speed of the vehicles involved in the crash). The reporting officer may elect to report that one of the factors contributing to the crash was "driving too fast for road conditions." This variable, however, is highly subjective. Table 9 depicts speedrelated fatal crash statistics for an eight year period (1996 to 2003). Based on the values shown in Table 9, the southeastern states have speeds below the national average. This finding may simply be a factor of police report techniques rather than factually based on actual crash conditions. Most of the state representatives involved in this research project were convinced that speed is a significant factor in many fatal crashes.

Table 9. Percent Speeding-Related Fatalities (1996 2003)

8 Year

1996 1997 1998 1999 2000 2001 2002 2003 Average

State [%] [%] [%] [%] [%] [%] [%] [%] [%]

AL

37.8 35.1 36.1 35.8 37.1 35.3 39.8 46.9 38.0

FL

26.2 27.6 21.7 18.0 17.5 17.6 17.8 17.0 20.4

GA 23.3 22.6 21.2 21.1 22.2 21.1 20.6 20.5 21.6

KY 30.1 28.8 23.8 26.4 20.6 18.5 19.6 13.1 22.6

MS 17.3 27.2 23.1 22.2 23.3 19.0 26.2 19.5 22.2

NC 35.6 34.7 34.7 37.7 35.3 37.5 38.2 37.0 36.3

SC

46.7 49.4 47.3 47.1 29.3 46.3 47.0 42.4 44.4

TN

26.8 26.4 25.2 28.2 24.5 23.0 24.9 22.8 25.2

Percent

for All SE

29.8 30.4 27.9 27.8 24.9 26.0 27.3 25.5

27.5

Crashes

Percent

for All U.S.*

31.0 31.1 30.1 30.3 29.5 30.5 32.0 31.4

30.7

Crashes

* U.S. values include the 50 states and the District of Columbia

Source: National Center for Statistics and Analysis, Fatality Analysis Reporting System

14

Supplemental Finding Pavement Edge Drop-offs
During the site evaluation process, representatives from the Federal Highway Administration and the State of Georgia observed a common theme for non-state-owned roads that were the sites of fatal crashes. At many of these locations, the edge of the pavement was not flush with the adjacent ground. In many cases, the pavement was characterized by a height differential of several inches. This drop-off appeared to be due to erosion as well as pavement maintenance overlays. In addition, rutting was often located adjacent to the road, particularly in the vicinity of roadside mailboxes. Since the specific evaluation of edge drop-offs was not one of the initial project objectives and much of the field work was completed at the time of this observation, only the states of North Carolina and Georgia had an opportunity to evaluate the extent of the drop-off problem.
In Georgia, the researchers only inspected non-state-owned and maintained sites since the state-owned roads had been previously inspected using a video library owned by GDOT. As a result, team members reviewed photographs and site inspection reports for the 69 sites not located on the Georgia state-system. At 55 percent of these sites, drop-offs or edge rutting was present. Upon review of the crash causal analysis, 21 of these 38 crashes included the edge drop-off as one of the direct causal factors to the fatal crash. North Carolina researchers also reviewed their site photographs and reports and determined that drop-offs and edge rutting occurred at 47 percent of their 150 crash locations.
15

Figure 1. Crash Direction for 54 Georgia Non-State-System Sites

STRAIGHT

Cross-Over Exit

Right Exit

9 due to Drop-Off 8 others Sub-Total = 17

1 due to Drop-Off 4 others Sub-Total = 5

Total = 22

CURVE TO LEFT

Cross-Over Exit (Inside)
2 due to drop-off 5 others Sub-Total = 7

Right Exit (Outside)
1 due to drop-off 4 others Sub-Total = 5

Total = 12

CURVE TO RIGHT

Cross-Over Exit (Outside)

Right Exit (Inside)

5 due to drop-off 10 others Sub-Total = 15

3 due to drop-off 2 others Sub-Total = 5

Total = 20

16

Review of the Georgia drop-off crashes in further detail indicated that day versus night and wet versus dry conditions were distributed evenly. Figure 1 demonstrates the road configuration and vehicle pattern at the Georgia pavement edge drop-off crash locations. Often the right wheels of a vehicle would run off the pavement and the driver would over-correct in an effort to re-direct the vehicle. This driver reaction often resulted in a cross-over exit.
The Georgia researchers reviewed the police crash reports to determine if a crash due to pavement edge drop-offs can be identified from crash data or reports. There were not any consistent variables to point to the drop-off problem. Common police report comments at these locations included:
"For reasons unknown." "... traveled with passenger side tires on the shoulder." "... came back on to the roadway and overcorrected and went into a
broadside skid..." "The driver ...steered back onto the roadway and lost control of the vehicle."
The pavement edge drop-off problem appears to be an extensive issue for rural two-lane highways and merits additional focused research based on the preliminary findings of this study.
Project Status
Seven of the eight states completed their evaluations. Each state research team evaluated issues pertinent to their region. Table 10 shows a summary of individual state
17

status. This table also shows the internet address of the final report, if available, via the internet.

Table 10. Project Status Summary Table

State AL FL
GA
KY MS NC SC TN

Current Status Complete Complete
Complete
Complete Complete Complete Complete Incomplete

Final Report Web Address (if available) ---
http://www11.myflorida.com/research-center/Completed_Proj/ Summary_SF/FDOT_BD158_rpt.pdf
http://www.dot.state.ga.us/dot/construction/materials-research/ b-admin/research/onlinereports/r-RP9905.pdf
http://www.ktc.uky.edu/Reports/KTC_01_11_SPR211_00_2F.pdf ---
http://www.hsrc.unc.edu/pdf/2002/sefatal_fr.pdf -----

Alabama The State of Alabama elected to perform a regional fatal crash description
evaluation using the comprehensive Critical Analysis Reporting Environment (CARE) traffic analysis database at the University of Alabama. In addition, researchers at Auburn (the lead university for this effort) collected crash information for 150 randomly selected rural, two-lane roads and provided this information to Georgia Tech for future research and analysis. These efforts represented research tasks #1 and #2 as identified at the project kickoff meeting. Alabama researchers have completed their research study and submitted a final report to the Alabama Department of Transportation.

18

Florida Florida researchers used a multi-step process to identify safety issues important to
the State of Florida. They divided problem areas into the categories of behavior, environment, vehicle, and engineering. Their efforts were based on data from the National Highway Traffic Safety Administration's Fatality Analysis Reporting System (FARS). The top three problems for drivers were determined to be shoulders, vehicles turning left just prior to crash, and crashes involving a drunk driver. Florida proposed research for future efforts will evaluate policies and programs as well as design standards and practices and determine how they differ from other states and the influence of these differences on the crash condition.
Georgia Georgia researchers participated in the regional fatal crash description, fatal crash
causal analysis of two-lane rural roads, and countermeasure identification tasks (Tasks #1, 2, & 3 as identified at the project kickoff meeting). Representatives from the Georgia Department of Transportation asked the research team to focus on possible engineering countermeasures that could be implemented. The analysis performed by the Georgia team included a statistical analysis whereby microscopic crash causal analysis and countermeasure assessments were combined with historic data to determine the most effective countermeasures feasible for two-lane rural roads. The microscopic analysis was based on 150 randomly selected rural two-lane road fatal crashes from 1997. Five specific countermeasures were recommended for future implementation strategies to combat these crashes. These countermeasures included:
1. Addition of advisory speed signs or other speed controls,
19

2. Geometric alignment improvements, 3. Widening of lanes/pavement widths, 4. Adding and/or widening graded/stabilized shoulders, and 5. Widening/improvement of clear zones. Future research efforts by Georgia Tech will include the development of predictive models based on the 150 fatal crash databases provided by several of the participating states. Georgia Tech submitted a final Georgia report to the Georgia Department of Transportation in 2002.
Following completion of the Georgia final report, the research team performed an additional evaluation of pavement edge drop-offs. A summary of these findings was previously included in this report.
Kentucky Researchers from Kentucky evaluated 150 fatal crashes from two-lane rural roads
(Task #2 of the kickoff meeting) and recommended countermeasures to reduce the number and severity of crashes for this road type. Crashes were from the years 1996 through 1998. The researchers determined that the effect of the enactment of a mandatory safety belt law had the greatest potential to reduce fatalities on two-lane rural roads. For roadway related countermeasures (excluding work zone devices), the addition of shoulder or centerline rumble strips and the installation of chevron signs at horizontal curves were determined to be the most likely to reduce the fatal crash frequency or severity. Kentucky submitted a final research report to the Kentucky Transportation Cabinet in May of 2001.
20

Mississippi Researchers for Mississippi performed a summary analysis to determine the
common characteristics of fatal crashes in the State of Mississippi. They prepared a final report summarizing general fatal crash characteristics and submitted the report to the Mississippi Department of Transportation. In addition, Mississippi researchers collected crash data for 150 fatal crashes on rural, two-lane roads. They submitted this database to Georgia Tech for future analysis.
North Carolina North Carolina researchers performed a causal chain analysis on 150 North
Carolina fatal crashes at two-lane, rural roads. In addition to this analysis they developed a ranked list of candidate safety countermeasures that could reduce the frequency or severity of these crashes. The research team identified twelve ranked countermeasures suitable for future consideration due to their proposed influence on the crash condition. These ranked countermeasures include:
1. Clear Zone Improvements Traversable Drainage Structure, 2. Install or Upgrade Guardrail, 3. Geometric Realignment, 4. Enforce Speed Limits, 5. Improve Sight Distance without Geometric Realignment, 6. Clear Zone Improvements Remove Fixed Object, 7. Clear Zone Improvements Widen Clear Zone, 8. Warning Sign, 9. Clear Zone Improvements Flatten Side Slope,
21

10. Improve Shoulder Add or Widen Graded or Stabilized Shoulder, 11. Widen Travel Lanes / Pavement Width, and 12. Improve Longitudinal Shoulder Pave Existing Graded Shoulder of Suitable
Width. The North Carolina researchers have provided the fatal crash database with 150 two-lane, rural roads to Georgia Tech for future analysis and submitted a final report to the North Carolina Department of Transportation in January 2002.
Following completion of the North Carolina final report, the research team performed an additional overview evaluation of pavement edge drop-offs. A summary of these findings was previously included in this report.
South Carolina The State of South Carolina (SC) research team, in conjunction with their SCDOT
and FHWA sponsors, chose to perform a fatal crash causal analysis and evaluate the potential safety improvements based on a list of 30 safety countermeasures. They performed this analysis using the framework proposed by the Georgia Tech team. The SC team assigned a societal cost of $3 million for each fatal injury and calculated potential benefits based on this value. The results of this research are included in their Final Report titled "Fatal Crashes on Rural Secondary Highways."
Estimated societal benefits for prospective countermeasures ranged from $0 to $846.5 million. The SC research team identified eleven countermeasures that would potentially result in cost savings over $200 million. These ranked items and their associated 1998 estimated societal cost benefits included:
22

1. Enforce Speed Limits -- $846,489,893 2. Remove Fixed Object -- $603,443,270 3. Rumble Strips -- $565,734,375 4. Pave Existing Graded Shoulder of Suitable Width -- $464,325,490 5. Widen and Pave Existing Shoulder -- $441,165,024 6. Add or Widen Graded or Stabilized Shoulder -- $416,745,068 7. Geometric Realignment (Horizontal, Vertical, Intersection) -- $330,430,540 8. Install or Upgrade Guardrail -- $247,837,236 9. Flatten Side Slope $228,232,109 10. Relocate Fixed Object -- $222,316,082 11. Warning Sign -- $207,366,237
Three countermeasures the SC research team determined would have had little or no influence on the studied fatal crashes were improved access management, wider clear zones, and traversable drainage structures.
Tennessee Tennessee researchers did not complete this research effort. They successfully
identified and visited 150 two-lane rural fatal crash sites, but did not progress further on this project.
23

Summary of Findings and Conclusions
Each participating state research team identified unique issues appropriate for their jurisdiction. Due to the variety of analysis procedures and identified objectives from their respective state representatives, the individual conclusions dramatically varied for each state. In general, seven of the eight states determined that the rural, two-lane road condition is the source of the elevated fatal crashes in the region. Improved features such as widening shoulders, enhancing delineation, and protecting the clear zone were identified consistently during the countermeasure analyses. Some of the researchers recommended to their state representatives that additional procedures and policies may be an appropriate countermeasure for wide-scale improvements. Of the six perceived topical areas identified at the kickoff meeting, countermeasures (physical as well as political) were explicitly recommended to address two-lane rural roads, safety restraint use and fixed-object crashes.
A supplemental finding for this study was the presence of extensive pavement edge drop-offs for fatal crash sites in at least two of the participating states. As this observation occurred as a result of field inspection and was not initially identified as a target problem, it was not studied in great detail for this research effort but merits special comment since it is potentially a significant finding of the study.
Future Research
Several of the research teams identified future research needed to further understand the crash condition in their state. In addition, the Georgia Department of
24

Transportation will direct a study using the available state crash databases (150 fatal twolane rural roads per state) to further determine the feasibility of predicting crash conditions and to understand the differences in the road conditions or the individual states and how these differences influence safety.
The pavement edge drop-off supplemental finding indicates that future research regarding these drop-offs and methods for addressing this common problem is warranted.
References
1. National Highway Traffic Safety Administration, Traffic Safety Facts 2003 Overview, Report DOT HS 809 767, Washington, DC.
2. National Center for Statistics and Analysis, Traffic Safety Facts -- State Traffic Data (1996 2003), National Highway Traffic Safety Administration, Washington, DC.
3. National Highway Traffic Safety Administration, Traffic Safety Facts 2003 Occupant Protection, Report DOT HS 809 765, Washington, DC.
4. National Highway Traffic Safety Administration, Traffic Safety Facts 2003 Young Drivers, Report DOT HS 809 774, Washington, DC.
5. Insurance Institute for Highway Safety, U.S. Licensing Systems for Young Drivers, Laws as of October 2002, Arlington, VA.
6. National Highway Traffic Safety Administration, Traffic Safety Facts 2003 Large Trucks, Report DOT HS 809 763, Washington, DC.
7. National Highway Traffic Safety Administration, Traffic Safety Facts 2003 Speeding, Report DOT HS 809 771, Washington, DC. 25

Appendix
26

Summary Slides from Kick-off Meeting
27

28

29

Sample Slides for State (Georgia) Briefing
30

31

32

33

34

35

36

37

38

Slides for Proposed Research Evaluation Method I
39

40

41

Slides for Proposed Research Evaluation Method II
42

43

44

45

46

47

48

49