Regulatory Evaluation of Electronic Logging

Devices and Hours of Service Supporting

Documents Final Rule

Regulatory Impact Analysis

Final Regulatory Flexibility Analysis

Unfunded Mandates Analysis

November 2015

i

EXECUTIVE SUMMARY This regulatory impact analysis (RIA) provides an assessment of the costs and benefits of requiring motor carriers and their commercial motor vehicle (CMV) drivers to use electronic logging devices (ELDs), formerly referred to as electronic on-board recorders (EOBRs), to record driving and other duty status periods. This RIA accompanies the Agency’s final rule. This rule is required by MAP-21, and the Federal Motor Carrier Safety Administration (FMCSA) believes that increasing the use of ELDs will improve compliance with the hours of service (HOS) rules and improve safety by decreasing the risk of fatigue-related crashes attributable to violations of the applicable HOS regulations. ELDs are electronic devices that automatically record driving time and facilitate electronic recording of other duty status categories, and provide the same information currently collected on paper records of duty status (RODS).

The ELD final rule follows the supplemental notice of proposed rulemaking (SNPRM) published in March 2014, which proposed requirements that superseded those proposed in an earlier notice of proposed rulemaking (NPRM). Because the Agency’s 2010 final rule providing technical specifications for ELDs was vacated, the SNPRM proposed new technical specifications for ELDs and addressed the issue of ELDs being used by motor carriers to harass drivers. The ELD final rule makes changes to the technical requirements proposed in the SNPRM. Specifically, the final rule simplifies the data transfer options and exempts vehicles with a pre-2000 model year from the ELD requirement. Neither change had a significant impact on the cost or benefit estimates for the remaining options in the RIA for the final rule.

Other modifications to the final rule analysis, however, did result in moderate changes to the cost and benefit estimates for the final rule from what was included in the SNPRM. For example, the purchase price of the ELD was reduced to reflect the most up-to-date prices consistent with the technical requirements of the final rule, the population estimates were adjusted to update the universe of drivers subject to the requirements of the final rule, and equipment requirements for inspectors were adjusted to no longer include QR scanners. The population changes had the effect of increasing costs, while adjustments to the ELD purchase price and equipment needs resulted in a decrease in costs. Overall, the total costs are slightly lower costs than what was projected in the SNPRM. In addition, the total benefits of the final rule increased due to updated wage estimates and adjustments to the projection of the cost of a crash. This resulted in an increase in the overall net benefits for the final rule from what was proposed in the SNPRM. These revisions are discussed in more detail throughout the RIA.

There are two alternatives for an ELD to meet the data transfer requirements in the final rule. The telematics alternative requires the ELD to have wireless Web services and email as the primary mechanisms for data transfer. The local transfer alternative requires an ELD to have USB 2.0 and Bluetooth capabilities for the primary data transfer method. In addition to the primary mechanisms, both options require the ELD to provide a printout or display with the required information as a back-up method for data transfer. The implementation plan for the final rule expects that inspectors would have one telematics means of data transfer, email or wireless Web services, and one local transfer option, USB 2.0 or Bluetooth.

ii

The RIA for the final rule retains two of the four options put forward in the SNPRM:

• Option 1: ELDs are mandated for all CMV operations subject to 49 CFR part 395.1 Under this option 4.27 million drivers would be subject to the rule, requiring the purchase of approximately 3.5 million ELDs or upgrades to an existing fleet management system (FMS).

• Option 2: ELDs are mandated for all CMV operations where the driver is required to complete RODS under 49 CFR 395.8. Under this option 3.51 million drivers would be subject to the rule, requiring the purchase of approximately 2.8 million ELDs or upgrades to an existing FMS.

In the RIA for the final rule, FMCSA adopts a slight variation of Option 2 from the SNPRM. Based on comments received on the SNPRM, Options 3 and 4 are not included in the final rule. Unlike the SNPRM, to provide a backup means of accessing data FMCSA will require either a display or printout regardless of the specific data transfer technologies required, thus rendering Options 3 and 4 unnecessary. In response to comments received to the SNPRM, the specific data transfer technologies required under today’s rule are simplified, with technologies such as QR Codes and TransferJet eliminated. In the SNPRM, the required data transfer technologies were the same across the four options presented in the SNPRM, with the only differences being the population the rule would apply to and a specific requirement for the ability to print out data. In today’s rule, the now more simplified set of required data transfer technologies are the same across the two options presented, with the only difference being the population. The change in data transfer technologies from the SNPRM does not affect the per unit cost of the ELD. However, in today’s rule the purchase price of the ELD was reduced from that used in the SNPRM, to reflect the most up-to-date prices consistent with the technical requirements of the final rule. This change in data transfer technologies from the SNPRM also simplifies and enhances uniformity of enforcement. For purposes of comparison, the analysis from the SNPRM, including Options 3 and 4, is available in the docket for this rulemaking.

To evaluate Options 1 and 2, FMCSA gathered cost information from publicly available marketing material and through communication with fleet management systems (FMS) vendors. Although the prices of some models have not significantly declined in recent years, manufacturers have been introducing less expensive FMS in-cab units and support systems with fewer features (for example, they do not include real time tracking and routing), as well as in-cab units that resemble a stand-alone ELD. The Agency bases its calculations in this RIA on the Mobile Computing Platform (MCP) 50 produced by Omnitracs (formerly Qualcomm), which is the largest manufacturer (by market share) of FMS in North America.2 While this analysis is not

1 49 CFR part 395 establishes the hours of service requirements for drivers. 2 Omnitracs (formerly Qualcomm) has one of the largest market shares in the industry with over 2,500 customers and 270,000 vehicles using its technology in North America and Latin America. In September 2014 Omnitracs announced its merger with XRS Corporation, another ELD industry leader: Omnitracs LLC 2013 Merger Announcement, Securities and Exchange Commission Form 8K. http://www.sec.gov/Archives/edgar/data/854398/ 000085439814000030/form8k-fy14q4xmergerannoun.htm. Accessed November 17, 2014.

iii

an endorsement of Omnitracs' products, the Agency believes that Omnitracs' large market share makes the MCP50 FMS an appropriate example of current state-of-the-art, widely available devices with ELD functionality. FMCSA also examined cost information from several other vendors, and found that the MCP50, with an annualized cost3 of $419 per CMV when all installation, service, repairs, and hardware costs are considered, falls in the middle of the range for other ELDs—$166 to $667 per CMV on an annualized basis. The $419 annualized cost estimate for the MCP50 is less than the annualized cost in the SNPRM of $495, primarily because the estimated purchase price declined from $799 in the SNPRM to $500. The purchase price of $500 is the minimum price for an MCP50 that could meet the ELD requirements.

The Agency also carefully considered the VDO RoadLog device produced by the Continental Corporation, which, through its VDO subsidiary, has a 60 percent share of the electronic tachograph4 market in the European Union (EU) and more than 6 million electronic tachographs or ELD-like devices in use worldwide.5 Continental has recently begun offering the RoadLog in the North American market, and the Agency believes that the overall capacity and market share of this corporation may allow it to influence the U.S. ELD market. As discussed below, the Agency has found that basing costs on the MCP50, the VDO RoadLog, or several other devices, all lead to positive net benefits of this rulemaking. Although carrier preferences and device availability prevent FMCSA from more precisely estimating costs, the Agency is confident that this rule’s costs will be lower than its benefits.

This analysis also evaluates the costs and benefits of improvements in motor carrier compliance with the underlying HOS rules through the use of ELDs. To evaluate compliance costs, the Agency has updated its assessment of the cost to come into compliance with the HOS rules to account for changes in factors such as inflation, changes in the HOS violation rate that preceded the mandate for ELD use, and the vehicle miles traveled by CMVs. To evaluate safety benefits, the Agency examined several types of analysis and has used its judgment to select a conservative result for the number of crashes and fatalities avoided by ELD use.

The costs and benefits are detailed later in the RIA and the methods by which they were derived are also discussed. The major elements that contribute to the overall net benefits are shown below in Table 1. This table summarizes the figures for the Agency’s selected option, which also has the highest net benefits, Option 2.

3 For a summary of the concept of annualization of costs and benefits, see the document Regulatory Impact Analysis: Frequently Asked Questions (FAQs) published by the Office of Management and Budget (OMB). February 7, 2011. http://www.whitehouse.gov/sites/default/files/omb/assets/OMB/circulars/a004/a-4_FAQ.pdf. Accessed June 9, 2014. 4 Electronic tachographs are on-board recording devices used on commercial vehicles outside the U.S. in areas of the world such as the European Union (EU), Japan, Australia and elsewhere to record commercial vehicle and driver activity measures such as vehicle speed, vehicle distance traveled, and driver activity type. 5 About VDO. http://www.vdoroadlog.com/about-vdo/. 2014. Accessed November 17, 2014.

iv

Table 1. Summary of Annualized Costs and Benefits (7% Discount Rate)

Cost Element

Annualized Total Value (2013 $

Millions) Notes

New ELD Costs $1,032.2 For all long-haul (LH) and short-haul (SH) drivers that use RODS, to pay for new devices and FMS upgrades.

Automatic On-Board Recording Device (AOBRD) Replacement Costs $2.0

Carriers that purchased AOBRDs for their CMVs and can be predicted to still have them in 2019 and would need to replace or update them with ELDs.

Enforcement Equipment Costs $1.3 The final rule does not require inspectors to purchase QR code scanners. Instead, inspectors would have Bluetooth capability and USB 2.0.

Enforcement Training Costs $1.6 Costs include travel to training sites, as well as training time, for all inspectors in the first year and for new inspectors each year thereafter.

CMV Driver Training Costs $8.0 Costs of training new drivers in 2017, and new drivers each year thereafter.

HOS Compliance Costs $790.4 Extra drivers and CMVs needed to ensure that no driver exceeds HOS limits.

Total Costs $1,836

Benefit Element

Annualized Total Value (2013 $

Millions) Notes Paperwork Savings (Total of three parts below) $2,437.6

(1) Driver Time $1,877.2 Reflects time saved as drivers no longer have to fill out and submit paper RODS.

(2) Clerical Time $433.9 Reflects time saved as office staff no longer have to process paper RODS.

(3) Paper Costs $126.6 Purchases of paper logbooks are no longer necessary

Safety (Crash Reductions) $572.2

Although the predicted number of crash reductions is lower for SH than LH drivers, both should exhibit less fatigued driving if HOS compliance increases. Complete HOS compliance is not assumed.

Total Benefits $3,010

Net Benefits $1,174

The Agency has clarified its supporting document requirements, recognizing that ELD records serve as the most robust form of documentation for on-duty driving periods. FMCSA neither increases nor decreases the burden associated with supporting documents. These adopted changes are expected to improve the quality and usefulness of the supporting documents retained, and will consequently increase the effectiveness and efficiency of the Agency’s review

v

of motor carriers’ HOS records during on-site compliance reviews, thereby increasing its ability to detect HOS rules violations. The Agency is currently unable to evaluate the impact changes to supporting documents requirements will have on crash reductions. Table 2 summarizes the analysis. The figures presented are annualized using 7 percent and 3 percent discount rates.

Table 2. Annualized Costs and Benefits (2013 $ millions)

Option 1: All HOS

Drivers Option 2: (Adopted) RODS

Drivers Only 3% 7% 3% 7%

Total Benefits $3,150 $3,124 $3,035 $3,010

Safety (Crash Reductions) $694 $687 $579 $572 Paperwork Savings $2,456 $2,438 $2,456 $2,438

Total Costs $2,298 $2,280 $1,851 $1,836

New ELD Costs $1,348 $1,336 $1,042 $1,032

AOBRD Replacement Costs $2 $2 $2 $2 HOS Compliance Costs $936 $929 $797 $790 CMV Driver Training Costs $9 $10 $7 $8 Enforcement Training Costs $1 $2 $1 $2

Enforcement Equipment Costs $1 $1 $1 $1

Net Benefits $852 $844 $1,184 $1,174

The estimated benefits of ELDs do not differ greatly between the options, and the paperwork savings are identical for both options. The Agency estimates zero paperwork burden from operations exempt from RODS, so ELDs can only reduce the paperwork burden of RODS users, which are included in both options. Safety benefits are higher when all regulated CMV operations are included in the ELD mandate (Option 1), but the marginal costs (ELD costs plus compliance costs) of including these operations are more than 3½ times higher than the marginal benefits. Option 1 would add SH drivers who do not use RODS, have better HOS compliance, and much lower crash risk from HOS noncompliance. For the SH non-RODS subgroup, FMCSA’s analysis indicates that ELDs are not a cost-effective solution to improving the HOS compliance of SH non-RODS drivers. This result is consistent with that of past ELD analyses. Option 2, which requires ELDs for paper RODS users, provides the largest net benefits, about $1,174 million per year (annualized using a 7 percent discount rate). Net benefits are positive for Option 1, but are less than those estimated for Option 2. Crash reductions associated with the options are also substantial. The final rule, which adopts Option 2, is expected to prevent 1,844 crashes and 26 fatalities annually. Review of Trucks Involved in Fatal Accidents (TIFA) data from 2005 to 2009 supports this analysis: variables indicating that the driver of the CMV was drowsy, sleepy, asleep, or fatigued are coded for crashes that caused an average of 85 deaths per year in that period. An average of nine crashes per year in TIFA were associated with fatigue and

vi

drivers exceeding drive time limits. Additional factors were at play in most of these events, but the removal of some substantial fraction of fatigued driving should provide some benefit. Table 3 shows the total costs, benefits, and net benefits of rule options 1 and 2 over the 10-year analysis time period. These figures are presented on an undiscounted (0 percent) basis and on a discounted basis at both 3 percent and 7 percent rates.

Table 3. Cumulative Costs, Benefits, and Net Benefits over 10 Years (2013 $ millions)

Option 1 Option 2 Discount Rate 0% 3% 7% 0% 3% 7% Benefits $31,708 $27,680 $23,479 $30,544 $26,665 $22,619 Costs $23,117 $20,191 $17,138 $18,622 $16,259 $13,794 Net Benefits $8,591 $7,489 $6,341 $11,922 $10,406 $8,825

Estimated crash reductions due to the final rule are summarized in Table 4.

Table 4. Estimated Reductions in Crashes

Option 1: All HOS Drivers Option 2: RODS Drivers Only Crashes Avoided 2,217 1,844 Injuries Avoided 675 562 Lives Saved 31 26

vii

Technical Report Documentation Page 1. Report No.

2. Government Accession No.

3. Recipient's Catalog No.

4. Title and Subtitle Regulatory Evaluation of Electronic Logging Devices and Hours of Service Supporting Documents Final Rule

5. Report Date November 2015

6. Performing Organization Code

7. Author(s) Brian Preslopsky, Steve McGonegal, Brian Seymour, Kenneth Blackman,

Katie Relihan, David Simon, Marina Krylova, Brendan O’Connor

8. Performing Organization Report No.

9. Performing Organization Name and Address Econometrica, Inc.

7475 Wisconsin Avenue, Suite 1000 Bethesda, MD 20814

10. Work Unit No. (TRAIS)

11. Contract or Grant No.

12. Sponsoring Agency Name and Address U.S. Department of Transportation

Federal Motor Carrier Safety Administration Office of Analysis, Research, and Technology

1200 New Jersey Ave. SE Washington, DC 20590

13. Type of Report and Period Covered

Final Report

14. Sponsoring Agency Code FMCSA

15. Supplementary Notes Contracting Officer’s Technical Representative: Olu Ajayi

16. Abstract This regulatory impact analysis (RIA) provides an assessment of the costs and benefits of requiring motor carriers to use electronic logging devices (ELDs) to track driving and duty times. The Federal Motor Carrier Safety Administration (FMCSA) is issuing a final rule, which this RIA accompanies, to improve compliance with the hours of service (HOS) regulations (49 CFR part 395) for commercial motor vehicle drivers (CMVs). ELDs track driving time and other activities electronically, and provide largely the same information currently collected on paper records of duty status.

17. Key Words electronic logging devices, ELDs, hours of service, driver

fatigue, crash avoidance

18. Distribution Statement No restrictions

19. Security Classif. (of this report) Unclassified

20. Security Classif. (of this page) Unclassified

21. No. of Pages 170

22. Price

Form DOT F 1700.7 (8-72) Reproduction of completed page authorized.

viii

TABLE OF CONTENTS EXECUTIVE SUMMARY ......................................................................................................................... I

ABBREVIATIONS, ACRONYMS, AND SYMBOLS ..................................................................... XVII

1. INTRODUCTION ............................................................................................................................. 1

1.1 BACKGROUND ....................................................................................................................... 1

1.2 HISTORY OF ELD RULEMAKING ACTIVITIES ................................................................ 2

1.3 DESCRIPTION OF FINAL RULE ........................................................................................... 3

2. REGULATORY ANALYSIS ........................................................................................................... 7

2.1 EXECUTIVE ORDER 12866 (REGULATORY PLANNING AND REVIEW) ..................... 7

2.2 IDENTIFICATION OF THE PROBLEM AND THE NEED FOR THE RULE ..................... 7

2.3 OPTIONS CONSIDERED ........................................................................................................ 8

2.4 SCOPE AND PARAMETERS OF THE ANALYSIS .............................................................. 8

2.4.1 Presentation of Figures ................................................................................................. 8

2.4.2 Labor Costs .................................................................................................................. 8

2.4.3 Crash Costs ................................................................................................................. 11

2.5 BASELINE FOR ANALYSIS ................................................................................................ 13

2.5.1 Numbers of Carriers and Drivers ............................................................................... 13

2.5.2 HOS Compliance Costs ............................................................................................... 16

2.5.3 Methodology for ELD Effectiveness ......................................................................... 16

2.5.4 Methodology for Safety Benefits ............................................................................... 16

2.5.5 Current AOBRD/ELD and FMS Use ......................................................................... 19

2.5.6 Total ELDs Required.................................................................................................. 20

3. COSTS OF FINAL RULE .............................................................................................................. 25

3.1 ELD DEVICE COSTS ............................................................................................................ 25

3.1.1 New ELD Costs .......................................................................................................... 25

ix

3.1.2 Annualized ELD Cost ................................................................................................ 26

3.1.3 ELD Monitoring on Current FMS .............................................................................. 27

3.1.4 AOBRD Replacement Costs ...................................................................................... 28

3.2 NEW EQUIPMENT FOR ROADSIDE INSPECTORS ......................................................... 29

3.3 TRAINING COSTS ................................................................................................................ 30

3.3.1 Driver and Carrier Office Staff Training .................................................................... 31

3.4 HOS COMPLIANCE COSTS ................................................................................................. 32

3.4.1 Estimate of Costs of Full HOS Compliance ............................................................... 32

3.4.2 HOS Compliance Costs of ELDs ............................................................................... 34

3.5 VENDOR REGISTRATION COSTS ..................................................................................... 36

3.6 COSTS TO FEDERAL GOVERNMENT .............................................................................. 36

4. BENEFITS ....................................................................................................................................... 37

4.1 PAPERWORK SAVINGS ...................................................................................................... 37

4.1.1 Driver Time Savings .................................................................................................. 37

4.1.2 Clerical Time Savings ................................................................................................ 38

4.1.3 Paper Cost Savings ..................................................................................................... 38

4.1.4 Total Paperwork Savings............................................................................................ 38

4.2 SAFETY BENEFITS .............................................................................................................. 38

4.2.1 Violations and Predicted Crash Reductions ............................................................... 39

4.2.2 Adjusted Crash Reductions ........................................................................................ 44

4.2.3 SH Crash Reductions.................................................................................................. 45

4.2.4 Average Crash Reductions over Both Samples .......................................................... 46

5. RESULTS OF ANALYSIS ............................................................................................................. 49

5.1 CALCULATIONS FOR OPTION 2 ....................................................................................... 50

5.2 CALCULATIONS FOR OPTION 1 ....................................................................................... 51

x

6. SENSITIVITY ANALYSES ........................................................................................................... 52

6.1 ALTERNATE VALUES OF STATISTICAL LIFE ............................................................... 52

6.2 LOWER COST ELD ............................................................................................................... 52

6.3 LOWER CRASH REDUCTIONS .......................................................................................... 53

6.4 DIFFERENCES IN DRIVER TIME SAVINGS .................................................................... 54

7. REGULATORY FLEXIBILITY ANALYSIS .............................................................................. 56

7.1 INTRODUCTION ................................................................................................................... 56

7.2 NEED FOR AND OBJECTIVES OF THIS RULE ................................................................ 56

7.3 PUBLIC COMMENT ON THE IRFA, FMCSA ASSESSMENT AND RESPONSE ........... 57

7.4 FMCSA RESPONSE TO COMMENTS BY THE CHIEF COUNSEL FOR ADVOCACY OF THE SMALL BUSINESS ADMINISTRATION ON THE IRFA ......................................... 57

7.5 DESCRIPTION AND NUMERICAL ESTIMATE OF SMALL ENTITIES AFFECTED BY THE RULEMAKING ............................................................................................................. 57

7.6 DESCRIPTION OF REPORTING, RECORDKEEPING AND OTHER COMPLIANCE REQUIREMENTS OF THE RULE ............................................................................................................................ 60

7.7 STEPS TO MINIMIZE ADVERSE ECONOMIC IMPACTS ON SMALL ENTITIES ........ 61

7.8 DESCRIPTION OF STEPS TAKEN BY A COVERED AGENCY TO MINIMIZE COSTS OF CREDIT FOR SMALL ENTITIES ....................................................................................................................... 61

8. UNFUNDED MANDATES REFORM ANALYSIS ..................................................................... 62

APPENDIX A: IMPACT OF ELDS ON ROADSIDE INSPECTION VIOLATIONS FOR FIVE COMPANIES .................................................................................................................................. 63

A.1 METHODOLOGY .................................................................................................................. 63

A.2 RESULTS ................................................................................................................................ 64

A.3 LOWER BOUND OF ELD EFFECTIVENESS MEASURES ............................................... 66

A.4 HOURS OF SERVICE VIOLATIONS CITED IN TABLES 45 AND 46 ............................................... 68

APPENDIX B: FORECASTS OF ELD AND FMS USE ................................................................... 71

APPENDIX C: PRICE AND AVAILABILITY OF ELDS ............................................................... 76

xi

C.1 ELD VENDORS ..................................................................................................................... 76

C.2 DOMINANT FIRMS .............................................................................................................. 77

C.3 SMALL/EMERGING FIRMS ................................................................................................ 78

C.4 INTERNATIONAL FIRMS .................................................................................................... 79

C.5 COSTS OF SAMPLE DEVICES ............................................................................................ 79

APPENDIX D: COMPLIANCE COSTS OF THE HOS RULES ..................................................... 82

D.1 BASIS FOR 2003 RULE AND DERIVATION OF PRE-2003 STATUS QUO .................... 82

D.1.1 Survey Data Used in Status Quo ................................................................................ 82

D.1.2 Driver Schedule Simulation Methods and Results ..................................................... 84

D.1.3 Rolling Work and Sleep Schedules ............................................................................ 86

D.1.4 Dispatching Simulation .............................................................................................. 86

D.1.5 Results of Rolling Work and Sleep Schedule Analysis .............................................. 88

D.2 COMPLIANCE COSTS .......................................................................................................... 89

D.2.1 Labor Cost Changes ................................................................................................... 89

D.2.2 Non-Wage Costs ........................................................................................................ 98

D.3 RESULTS OF 2003 HOS RULE ANALYSIS ..................................................................... 105

D.4 CHANGES TO COST AND BENEFIT CALCULATIONS FROM THE 2005 HOS RIA . 106

D.4.1 Work Patterns ........................................................................................................... 106

D.4.2 Use of New Provisions of the 2003 HOS Rule ........................................................ 110

D.4.3 Carrier Operation Simulations .................................................................................. 112

D.4.4 LH Compliance Costs .............................................................................................. 114

D.4.5 Total Costs of the 2005 HOS Rule Changes ............................................................ 116

APPENDIX E: ROADSIDE INTERVENTION MODEL ............................................................... 117

E.1 BACKGROUND ................................................................................................................... 117

E.2 METHODOLOGY ................................................................................................................ 118

xii

E.2.1 Input Data Selection ................................................................................................. 118

E.2.2 Assignment of Crash Risk Reduction Probabilities ................................................. 118

E.2.3 Calculation of Benefits ............................................................................................. 119

E.3 FY 2009 INTERVENTION MODEL RESULTS ................................................................. 120

E.3.1 National Level Estimates.......................................................................................... 120

E.3.2 State-Level Estimates ............................................................................................... 122

E.4 CONCLUSION ..................................................................................................................... 134

E.5 IMPROVEMENTS TO THE INTERVENTION MODEL METHODOLOGY ................... 136

E.5.1 Development of the Intervention Model .................................................................. 136

E.5.2 Improvements to the Intervention Model ................................................................. 136

E.5.3 Correction of Violations ........................................................................................... 136

E.5.4 Multiple Violations per Intervention ........................................................................ 138

E.5.5 Crash Risk Reduction Associated with Violations ................................................... 139

E.5.6 Indirect Effects ......................................................................................................... 145

E.5.7 Comparison of Results between Intervention Models 3.0 and 2.0 ........................... 147

APPENDIX F: VALIDATING SAFETY BENEFIT MODEL ....................................................... 150

xiii

LIST OF FIGURES Figure 1. Forecast of FMS Use without ELD Rule ..................................................................................... 72 Figure 2. Forecast of Voluntary ELD Use without ELD Rule .................................................................... 72 Figure 3. U.S. Firms Annual Sales for 2013 (Million $) ............................................................................ 78 Figure 4. Total Wage Curve for Non-Union Drivers .................................................................................. 93 Figure 5. Marginal and Average Wage Curves for Non-Union Drivers ..................................................... 93 Figure 6. Operational Cycle of HOS Model ............................................................................................. 113 Figure 7. Estimated Number of Crashes Avoided and Lives Saved Trends ............................................. 121 Figure 8. Distributions of Multiple Violations in Post-Crash and Non-Crash Inspections ....................... 139 Figure 9. Violation Group Crash-Risk Probability Estimation Equation.................................................. 141 Figure 10. Equation to Estimate Numerator of Figure 9 ........................................................................... 141 Figure 11. Final Calculation of Risk Reduction for Violation (j) Equation .............................................. 145 Figure 12. Intervention Model: Direct Effects Versions 3.0 and 2.0 ........................................................ 148 Figure 13. Intervention Model: Indirect Effects Version 2.0 .................................................................... 148 Figure 14. Intervention Model Results: Comparison of Versions 3.0 and 2.0 .......................................... 149

xiv

LIST OF TABLES Table 1. Summary of Annualized Costs and Benefits (7% Discount Rate) ................................................... iv Table 2. Annualized Costs and Benefits (2013 $ millions) ............................................................................ v Table 3. Cumulative Costs, Benefits, and Net Benefits over 10 Years (2013 $ millions) ............................. vi Table 4. Estimated Reductions in Crashes ................................................................................................... vi Table 5. Past and Current ELD Rule Estimated Costs and Benefits per ELDa, Adjusted to 2013 $ ................ 3 Table 6. Wage, Time, and Labor Costs (2013 $) .......................................................................................... 10 Table 7. Fractions of VSL for AIS ................................................................................................................. 11 Table 8. Costs of CMV Crashes (2013 $) ..................................................................................................... 12 Table 9. Costs of CMV Crashes (2013 $) ..................................................................................................... 13 Table 10. Carriers and Drivers Subject to the Rule ..................................................................................... 15 Table 11. Estimated Crash Reductions for HOS Violations ......................................................................... 18 Table 12. Current and Projected ELD and FMS Use without an ELD Mandate ........................................... 20 Table 13. Estimates of LH ELD Use, Without ELD Mandate ........................................................................ 21 Table 14. Estimates of SH ELD Use, Without ELD Mandate ........................................................................ 22 Table 15. Estimates of Industry ELD Use without Mandate ....................................................................... 23 Table 16. Costs of ELDs and Other Hardware (2013 $) ............................................................................... 25 Table 17: ELD Purchase Payment Estimates (2013 $) ................................................................................. 27 Table 18. Replacement Costs of AOBRDs (2013 $ millions) ........................................................................ 29 Table 19. Cost of Roadside Enforcement Equipment (2013 $ millions) ..................................................... 30 Table 20. Enforcement Training Costs ........................................................................................................ 31 Table 21. CMV Driver Training Costs .......................................................................................................... 32 Table 22. Costs of Improved HOS Compliance per ELD (2013 $) ................................................................ 32 Table 23. Compliance Costs of HOS Rules (Current Industry Behavior to Full Compliance) (millions) ...... 34 Table 24. Derivation of Annual Costs of Improved HOS Compliance per ELD (2013 $).............................. 35 Table 25. Annual Paperwork Savings per Driver (2013 $) .......................................................................... 37 Table 26. Safety Benefits per ELD (2013 $) ................................................................................................. 39 Table 27. Crash Risk Reductions from Sample of Voluntary LH ELD Users ................................................. 41 Table 28. Crash Risk Reductions from Sample of Mandated LH ELD Users ................................................ 43 Table 29. Adjusted Crash Reduction per LH CMV, Voluntary ELD Use ....................................................... 44 Table 30. Adjusted Crash Reduction per LH CMV, Mandatory ELD Use ..................................................... 45 Table 31. Crash Reduction per SH CMV ...................................................................................................... 46 Table 32. Average Safety Benefits per ELD (2013 $) ................................................................................... 47 Table 33. Summary of Annualized Costs and Benefits (2013 $ millions) .................................................... 49 Table 34. Option 2 Calculations Costs and Benefits (7% Discount Rate, 2013 $ Million) ........................... 50 Table 35. Option 1 Calculations Costs and Benefits (7% Discount Rate, 2013 $ Million) ........................... 51 Table 36. Summary of Options 1 and 2 with High and Low VSL (2013 $ millions) ..................................... 52 Table 37. Summary of All Options with Lowest Cost ELD (2013 $ millions) ............................................... 53 Table 38. Summary of All Options with Lowest Crash Reduction (2013 $ millions) ................................... 54 Table 39. Summary of Option 2 with Different Driver Time Savings (2013 $ millions) .............................. 55 Table 40. SBA Size Standards for Selected Industries (2014 $) ................................................................... 58

xv

Table 41. Estimates of Numbers of Small Entities ...................................................................................... 60 Table 42. Annualized Net Expenditures by Private Sector (2013 $ millions) .............................................. 62 Table 43. Pre-ELD and Post-ELD Installation Period Definitions for Carriers Evaluated ............................. 64 Table 44. Driver OOS Rates, HOS OOS Rates, and HOS Violation Rates for Voluntary and Mandated

Carriers before and After ELD Installation (1-Yr. Evaluation Periods) ....................................... 65 Table 45. Distribution of HOS Violations for Pre- and Post-ELD Inspections, for Voluntary Carriers ......... 66 Table 46. Distribution of HOS Violations for Pre- and Post-ELD Inspections, for Mandated Carriers ........ 66 Table 47. Adjusted ELD Effectiveness, Voluntary Group ............................................................................ 67 Table 48. Adjusted ELD Effectiveness, Mandatory Group .......................................................................... 68 Table 49. Estimated ELD and FMS Adoption ............................................................................................... 74 Table 50. Major ELD Vendors ...................................................................................................................... 77 Table 51. Comparison of ELD Cost Estimates (2013 $) ............................................................................... 81 Table 52. Work Week Length for UMTIP and Modeled Drivers ................................................................. 85 Table 53. Driver Schedule Proportion Matrix for Status Quo ..................................................................... 86 Table 54. Relative Driver Productivity from Dispatch Analysis ................................................................... 87 Table 55. Generation of Proportion of Schedules that Roll ........................................................................ 89 Table 56. Regression Results for Truck Driver Wage Relationship, 1995-2000 .......................................... 91 Table 57. Predicted Annual Wages for Different Hours/Week ................................................................... 92 Table 58. Index of Truck Driver Employment and Real Wages ................................................................... 95 Table 59. Employment Levels for Blue-Collar Occupations (thousands) .................................................... 96 Table 60. Summary of Labor Assumptions ................................................................................................. 98 Table 61. Assumptions Used to Model Motor Vehicle Equipment Costs ................................................. 100 Table 62. Assumptions Used to Model Parking Space Construction and Maintenance Costs ................. 102 Table 63. Cost of Driver Churn (2013 $).................................................................................................... 105 Table 64. Changes in Drivers Needed for Full Compliance with HOS Limits ............................................. 106 Table 65. Annual Direct Cost Changes for Full Compliance (2013 $ millions) .......................................... 106 Table 66. Daily Driving and On-Duty Hours--Averages ............................................................................. 108 Table 67. Average Weekly Hours and Days Worked ................................................................................. 108 Table 68. Driving Hours per Tour of Duty ................................................................................................. 108 Table 69. On-duty Hours per Tour of Duty (Field Survey Only) ................................................................ 109 Table 70. On-duty Hours in 8-day Periods ................................................................................................ 109 Table 71. Incidence of Splitting ................................................................................................................. 111 Table 72. Incidence of Splitting-- Team and Solo ...................................................................................... 111 Table 73. Percentage of Work Days with 11th Hour Use ......................................................................... 112 Table 74. Schedule Output of HOS Model ................................................................................................ 114 Table 75. LH Unit Costs for HOS Rule Changes (2013 $) ........................................................................... 116 Table 76. Total Costs of the 2005 HOS Rule (2013 $) ............................................................................... 116 Table 77. Program Activity FY 2007–09 .................................................................................................... 120 Table 78. Program Effectiveness FY 2007–09 Using Intervention Model 3.0 ........................................... 120 Table 79. Intervention Model Version 3.0 Estimated Program Benefits, CY 2004–FY 2009 .................... 122 Table 80. Program Exposure: U.S. Domiciled vs. Non-U.S. Domiciled Carriers, FY 2009 .......................... 122 Table 81. Program Effectiveness: U.S. Domiciled vs. Non-U.S. Domiciled Carriers, FY 2009 ................... 123

xvi

Table 82. Roadside Inspection Program Estimated Benefits by Reporting State, FY 2009 ...................... 125 Table 83. Traffic Enforcement Program Estimated Benefits by Reporting State, FY 2009 ....................... 127 Table 84. Roadside Inspection Program Estimated Benefits by Domicile State and Country, FY 2009 ... 130 Table 85. Traffic Enforcement Program Estimated Benefits by Domicile State and Country, FY 2009 .... 132 Table 86. Historical Results for Intervention Model, CY 2001–05 and FY 2006–09 ................................. 135 Table 87. Vehicle Maintenance Violation Correction Rates in Version 3.0 .............................................. 137 Table 88. Driver Fitness Violation Correction Rates in Version 3.0 .......................................................... 138 Table 89. Duration of Crash Risk Reduction by Violation Type ................................................................. 140 Table 90. Crash Risk Reductions for Violation Groups in Intervention Model 3.0 ................................... 142 Table 91. Change in Carrier Performance (Estimated Crashes Avoided) ................................................. 146 Table 92. Probability of a Carrier Sustaining Improved Performance ...................................................... 146 Table 93. Indirect Effects Estimates Based on Sustained Improvement by Carriers ................................ 147 Table 94. Comparison of Estimated Safety Benefits (2013 $) .................................................................. 150

xvii

ABBREVIATIONS, ACRONYMS, AND SYMBOLS

Acronym Definition

AIS abbreviated injury scale

AOBRD automatic on-board recording device

BLS Bureau of Labor Statistics

CMV commercial motor vehicle

CPS Current Population Survey

DFACS Driver Fatigue, Alertness and Countermeasures Study

ECEC Employer Costs for Employee Compensation

EOBR electronic on-board recorder

ELD electronic logging device

EU European Union

TIFA Trucks Involved in Fatal Accidents

FMCSA Federal Motor Carrier Safety Administration

FMCSR Federal Motor Carrier Safety Regulation

FMS fleet management systems

GDP gross domestic product

HHI Herfindahl-Hirschman Index

HM hazardous material

HOS hours of service

IIHS Insurance Institute for Highway Safety

IPI International Parking Institute

LH long-haul

LTL less than full truckload

MCP Mobile Computing Platform

xviii

Acronym Definition

NAICS North American Industry Classification System

NATSO National Association of Truck Stop Owners

NERA National Economic Research Associates

NPRM Notice of Proposed Rulemaking

OES Occupational Employment Statistics

OMB Office of Management and Budget

OOIDA Owner-Operator Independent Drivers Association

OOS out of service

OST Office of the Secretary of Transportation

OTR over-the-road

PRA Paperwork Reduction Act

RIA regulatory impact analysis

RODS record of duty status

SBA Small Business Administration

SH short-haul

SNPRM Supplemental Notice of Proposed Rulemaking

TIFA Trucks Involved in Fatal Accidents

TL Truckload

UMTIP University of Michigan Trucking Industry Program

USDOT U.S. Department of Transportation

VIN vehicle identification number

VMT vehicle miles traveled

VSL value of a statistical life

1

1. INTRODUCTION

1.1 BACKGROUND

The goal of the Federal hours of service (HOS) regulations (49 CFR part 395) is to promote safe driving of commercial motor vehicles (CMVs) by limiting on-duty driving time, thereby ensuring that drivers have adequate time to obtain rest. FMCSA (“the Agency”) conducts regular checks at the roadside and performs in-depth compliance reviews to ensure that drivers are operating within the HOS limits. The Agency’s roadside inspection data from the Motor Carrier Management Information System (MCMIS) indicate that in 2012, 5.5 percent of inspections that examined HOS resulted in at least one out-of-service (OOS) violation of the HOS rules. An OOS order issued for a violation found during a roadside inspection immediately prevents a CMV or driver from further operation until the violation has been rectified, thereby reducing its risk of being involved in a crash. In about 3.8 percent of year 2012 inspections that examined HOS, drivers were issued OOS citations for missing, incomplete, improper, or fraudulent records of duty status (RODS). These percentages have decreased slightly from when the last major change to the HOS rules, which increased cumulative on-duty hour limits and daily off-duty periods, went into effect in 2004. This also may be due to some lagging effects of lower demand for truck shipping that accompanied the last recession and that reduced the tendency for some drivers to work over the capacity set by their HOS limits.

Surveys have also shown that many CMV drivers violate HOS limits, and that many falsify their paper RODS to give the appearance of legal operation.6,7 An online survey, conducted by United Safety Alliance, Inc., revealed that over 78 percent of its respondents (CMV drivers) believe that the most common, deliberate HOS violation is logging time as off-duty when actually on duty. The survey also discovered that “77 percent of the respondents admitted to deliberately violating the HOS regulations in the past and 55 percent said they were still currently deliberately violating the rules.” Furthermore, survey respondents were asked to estimate how many days per month they were in violation. Respondents admitted to intentionally violating HOS regulations an average of 6 days per month, and unintentionally 5 days per month, either by accident, oversight, or honest mistake.8 Possible bias in these survey data prevents our using it in our numerical analysis.

Current data from roadside inspections indicate that over the time periods covered in these surveys, between 6 and 7 percent of inspections in which RODS were checked resulted in an HOS OOS order. Clearly there has been a disparity between the total number of HOS violations and the number that roadside inspectors have been able to detect by examining paper RODS. As

6 McCartt, A. T., L. A. Hellinga, and M. G. Solomon, “Work Schedules Before and After 2004 Hours-of-Service Rule Change and Predictors of Reported Rule Violations in 2004: Survey of Long-Distance Truck Drivers,” 2005. Proceedings of the 2005 International Truck and Bus Safety and Security Symposium, Alexandria, VA. Available from: http://trid.trb.org/view/2005/C/1156406, as of Feb. 4, 2014. 7 Commercial Vehicle Safety Alliance. Roadcheck 2007 Results Show Safety Improvements Are Needed. June 29, 2007. http://www.cvsa.org/news/2007_press.php. Accessed December 12, 2013. 8 Ol’Blue, USA. United Safety Alliance, Inc. 2006 HOS Survey Results. Online self-administered survey conducted August 15 – October 31, 2006. http://www.olblueusa.org/news/jan07.html. Accessed December 12, 2013.

2

stated, as of 2012, the HOS OOS rate was 5.5 percent. Although inspectors are finding a slightly lower percentage of violations, the number they are not able to detect likely still remains high, although it may have fallen slightly as well, in proportion to the cited violation rate. The National Transportation Safety Board and safety advocacy groups have recommended mandatory use of electronic logging devices (ELDs) as a way to increase compliance with HOS regulations.

1.2 HISTORY OF ELD RULEMAKING ACTIVITIES

The Automatic On-Board Recording Devices (AOBRD) regulation (49 CFR 395.15) was published in September 1988. In order to be considered a compliant device, an AOBRD must be integrally synchronized with specific operations of the CMV in which it is installed (see 49 CFR 395.2, Definitions, “automatic on-board recording device”). Use of AOBRDs is voluntary.

On April 5, 2010, the Agency issued a final rule that addressed the limited, remedial use of electronic on-board recorders (now called ELDs) for motor carriers with significant HOS violations (75 FR 17208), informally referred to as EOBR 1. The April 2010 final rule required the limited, remedial use of ELDs by any motor carrier found, during a single compliance review, to have a 10 percent violation rate for any HOS regulation listed in a new Appendix C of 49 CFR part 385. The final rule required ELDs on all of the motor carrier’s CMVs for a period of 2 years. The compliance date for the final rule was to have been June 4, 2012.

The Owner-Operator Independent Drivers Association (OOIDA) challenged the final rule in the United States Court of Appeals for the Seventh Circuit. OOIDA raised several concerns relating to ELDs, including potential use for driver harassment. On August 26, 2011, the court vacated the entire final rule. The Court held that, contrary to statutory requirements, the Agency failed to address the issue of driver harassment, including how ELDs might be used to harass drivers and ways to ensure that ELDs were not used for this purpose. As a result of the Court’s decision, carriers relying on electronic devices to monitor HOS compliance are still governed by the Agency’s rules regarding the use of AOBRDs (49 CFR 395.15). The requirements for AOBRDs, set forth in 49 CFR 395.15, were not affected by the Seventh Circuit’s decision. Also as a result of the decision, no supporting document paperwork relief was granted to motor carriers voluntarily using ELDs. In addition, on May 14, 2012, FMCSA rescinded the final rule published on April 5, 2010, entitled ‘‘Electronic On-Board Recorders for Hours-of-Service Compliance,’’ along with a technical amendment that had been published on September 13, 2010.

On February 1, 2011, FMCSA initiated the EOBR 2 rulemaking to propose a broader mandate for ELD use, coupled with a reduction in supporting document retention requirements (76 FR 5537-5555). Although FMCSA had received comments during the EOBR 1 rulemaking recommending expanding the reach of the rule beyond the limited number of motor carriers, the limited scope of that NPRM, focusing on remedial directives, prevented the Agency from doing so in the April 2010 final rule. The 2011 EOBR NPRM proposed to require all motor carriers currently required to document their drivers’ HOS with RODS to use ELDs, as defined in the April 2010 final rule. The 2011 NPRM also proposed reducing the number of supporting documents motor carriers using ELDs would be required to maintain by eliminating the need for supporting documents to verify driving time. FMCSA also proposed to require all motor carriers,

3

both RODS and timecard users, to systematically monitor their drivers’ compliance with the HOS requirements. One option of the 2011 NPRM also included passenger carriers and bulk hazardous material haulers. However, the elimination of safety gaps and HOS violations resulting from the rule were not large enough to compensate for the cost of ELDs for those segments of the industry, and subsequent analyses did not include that option. The last action that influenced development of the ELD rule was the passage of the MAP-21 legislation in 2012; which required ELDs for certain CMVs and resulted in the Agency publishing a supplemental notice of proposed rulemaking (SNPRM) in March 2014.

The rulemaking options have evolved over time, and cost and benefit estimates have been updated accordingly. The four iterations of the ELD rulemakings with FMCSA-adopted options are compared in Table 5.

Table 5. Past and Current ELD Rule Estimated Costs and Benefits per ELDa, Adjusted to 2013 $

EOBR1 1x10 Remedial

Directive (2010) EOBR2 Option 1

RODS Users (2011)

ELD SNPRM Option 2 RODS

Users (2014)

ELD Final Rule Option 2 Rods Users (2015)

Average ELD Costs $539 $715 $411 $332 Average HOS Compliance Costs

$725 $180 $260 $219

Average Paperwork Savings

$548 $886 $705 $805

Average Safety Benefits

$1,152 $331 $170 $186

Notes: a These values represent an annual average cost or benefit per driver for the drivers affected by the rule, in order to demonstrate the various costs and benefits across the four iterations of this rulemaking. For example, the value of $332 for “Average ELD Costs” under Option 2 of the ELD Final Rule represents the total aggregate annualized ELD costs as presented earlier in Table 1 of $876.2 million, divided by the average annual number of affected drivers under Option 2 over the 2017 through 2026 analysis time period (which averages to 3,069,192 affected drivers but varies somewhat for each individual year due to the pre-2000 model year exemption).

The underlying HOS regulations, the CMV, motor carrier, and driver populations, and many other factors have changed during this time, which contributes to the broad range of values for costs and benefits shown.

1.3 DESCRIPTION OF FINAL RULE

This regulatory impact analysis (RIA) provides an assessment of the costs and benefits of requiring motor carriers to use ELDs to record driving, on-duty, off-duty, and sleeper berth time. This requirement would improve CMV drivers’ compliance with the HOS limits. ELDs are electronic devices that automatically record driving time and facilitate electronic recording of other categories of duty status, and provide information similar to that currently collected on RODS. Use of ELD technology is intended to significantly reduce or eliminate false or erroneous

4

driving time records, and reduce false and erroneous on-duty, off-duty, and sleeper-berth entries.9

The ELD final rule is intended to improve CMV safety and reduce the paperwork burden by increasing the use of ELDs within the motor carrier industry, which FMCSA believes will improve HOS compliance, and thereby reduce the number of crashes related to CMV driver fatigue associated with violations of the HOS rules. This final rule establishes technical specifications for ELDs, simplifies supporting document requirements, defines which drivers and carriers are required to use ELDs, and addresses the issue of ELDs being used by motor carriers to harass drivers.

Among other changes, the final rule makes two changes to the technical specifications for ELDs from what was included in the SNPRM. Specifically, the final rule simplifies the data transfer options and exempts vehicles with a pre-2000 model year from the ELD requirement. Neither change had a significant impact on the cost or benefit estimates for the remaining options in the RIA for the final rule.

Other modifications to the final rule analysis, however, did result in moderate changes to the cost and benefit estimates for the final rule from what was included in the SNPRM. For example, the purchase price of the ELD was reduced to reflect the most up-to-date prices consistent with the technical requirements of the final rule, the population estimates were adjusted to update the universe of drivers subject to the requirements of the final rule, and equipment requirements for inspectors were adjusted to no longer include QR scanners. The population changes had the effect of increasing costs, while adjustments to the ELD purchase price and equipment needs resulted in a decrease in costs. Overall, the total costs are slightly lower costs than what was projected in the SNPRM. In addition, the total benefits of the final rule increased due to updated wage estimates and adjustments to the projection of the cost of a crash. This resulted in an increase in the overall net benefits for the final rule from what was proposed in the SNPRM. These revisions are discussed in more detail throughout the RIA.

There are two alternatives for an ELD to meet the data transfer requirements in the final rule. The telematics alternative requires the ELD to have wireless Web services and email as the primary mechanisms for data transfer. The local transfer alternative requires an ELD to have the USB 2.0 or Bluetooth capabilities for the primary data transfer method. In addition to the primary mechanisms, both options require the ELD to provide a printout or display with the required information as a back-up method for data transfer. The final rule expects that inspectors will have one telematic means of data transfer—email or wireless Web services, and one local transfer option—USB 2.0 or Bluetooth.

9 The ELDs record some but not all aspects of CMV use automatically, which is why false entries would be reduced but not eliminated. Time and location of the CMV at the point of the driver’s entry of a change of duty status are recorded, which will help to verify the entered information. Also, the driver cannot enter off duty or sleeper berth status while the truck is moving.

5

This RIA examines two options:

• Option 1: ELDs are mandated for all CMV operations subject to 49 CFR part 395 (Which defines hours of service regulations).

• Option 2: ELDs are mandated for all CMV operations where the driver is required to complete RODS under 49 CFR 395.8 (Where hours of service must be recorded in paper logs or some other way).

The Agency gathered information from publicly available marketing material and contact with fleet management systems (FMS) vendors about the costs and features of ELDs currently on the market. Although the prices of specific models have not significantly declined in recent years, manufacturers have been introducing less expensive FMS units with fewer features (for example, not including real time tracking and routing), and units resembling a stand-alone ELD are now available. The Agency has chosen to base its calculations in this RIA on the Mobile Computing Platform (MCP) 50 produced by Omnitracs, which is the largest manufacturer of FMS in North America. This analysis is not an endorsement of Omnitracs’ products, but the Agency believes that its large market share makes the MCP50 FMS an appropriate example of current state-of-the-art, widely available devices with ELD functionality. FMCSA also examined cost information from several other vendors and found that the MCP50, when all installation, service, and hardware costs are considered, falls roughly into the middle of the price range of FMS with ELD capabilities. The Agency also carefully considered the VDO RoadLog device produced by the Continental Corporation, which, through its VDO subsidiary, has a 60 percent share of the electronic tachograph market in the EU and more than 6 million electronic tachographs or ELD-like devices in use worldwide.10 Continental has recently begun offering the RoadLog as a near stand-alone ELD-like device in the North American market, and the Agency believes that the overall capacity and market share of this corporation may allow it to influence the U.S. ELD market. As discussed below, the Agency has found that basing costs on the MCP50, the VDO RoadLog, or several other devices, all lead to positive net benefits of this rule.

Although carrier preferences and limited device availability prevent FMCSA from more precisely estimating costs, the Agency is confident that this rule’s costs will be lower than its benefits. The Agency determined that the change in data transfer requirements does not affect the unit cost of ELDs. Most ELDs currently on the market already meet the data transfer requirements. However, the change in data transfer alternatives does reduce the cost of roadside enforcement equipment because inspectors are no longer required to obtain QRcode scanners. Appendix C of this RIA contains additional discussion of the availability and prices of ELD-class devices and systems.

The Agency has been closely monitoring the market for enhanced AOBRDs and units that currently appear to meet the core requirements for ELDs. They fall into three categories: AOBRDs that electronically log HOS in compliance with 49 CFR 395.15, FMS that track the assets of large fleets for logistics and management purposes (and whose functions are unregulated except for AOBRD features; some of these devices may come close to meeting the core ELD requirements), and ELDs that would come close to complying with the core requirements of the final rule. Almost all electronic HOS recording units currently available or in 10 About VDO. http://www.vdoroadlog.com/about-vdo/. 2014. Accessed November 17, 2014.

6

use are applications within FMS that offer mobile communications and tracking and are integrally synchronized with the CMV, which maintains compliance with 49 CFR 395.15. The latter is the most critical hardware requirement for the ELDs. However, devices that function only, or primarily, as ELDs, are now coming onto the market, but many carriers required to acquire and use ELDs will likely obtain new FMS that offer ELD functionality to take advantage of having several features bundled within one device. At least one GPS manufacturer offers an ELD feature with its unit for carriers that want to limit themselves to just this technology. To estimate costs that carriers would bear to equip CMVs with ELDs, FMCSA examined prices, fees, and availability of specific units. This is also discussed in detail in Appendix C. The Agency has determined that a median-priced FMS from a large manufacturer can be used as a reasonable baseline for estimating device costs, but will evaluate in a sensitivity analysis the costs of this rule should stand-alone ELDs become widely available. Also, as it did in analysis performed for prior ELD rules, the Agency will continue to recognize that carriers already using FMS will have a low or no cost upgrade path to acquiring ELD functionality. These carriers will simply begin using their devices’ logging feature or upgrade their devices to this functionality.

7

2. REGULATORY ANALYSIS

2.1 EXECUTIVE ORDER 12866 (REGULATORY PLANNING AND REVIEW)

FMCSA has determined that this rulemaking is an economically significant regulatory action under Executive Order (E.O.) 12866,11 Regulatory Planning and Review, as supplemented by E.O. 13563 (76 FR 3821, January 21, 2011). It also is significant under Department of Transportation regulatory policies and procedures because the economic costs and benefits of the rule exceed the $100 million annual threshold and because of the substantial congressional and public interest concerning the crash risks associated with driver fatigue.

This regulatory analysis:

• Identifies the target problem, including a statement of the need for the action. • Discusses alternative approaches considered. • Defines the scope and parameters of the analysis. • Defines the baseline. • Defines and evaluates the costs and benefits of the action and the main alternatives

identified by the analysis. • Compares the costs and benefits. • Interprets the cost and benefit results.

This regulatory analysis is the synthesis of research conducted specific to ELDs, the body of research FMCSA has accumulated on the effects of driver fatigue, the costs and benefits of each provision of the HOS rules, and estimates of the paperwork burden of RODS, HOS supporting documents, and ELD usage. It refers to several other analytical reports produced by the Agency. Several appendices that follow the main body of this RIA summarize the research from these other reports that were used in this analysis.

2.2 IDENTIFICATION OF THE PROBLEM AND THE NEED FOR THE RULE

FMCSA has determined that there are significant violations to the HOS rules that can lead to increased CMV driver fatigue and pose an unacceptable risk to the public. Additionally, the MAP-2112 legislation includes a requirement that CMVs engaged in interstate commerce and required to keep RODS be equipped with ELDs. ELDs are electronic devices that automatically record driving time and facilitate electronic recording of other categories of duty status, providing information similar to that currently collected on paper RODS. Use of ELD technology is intended to significantly reduce or eliminate false or erroneous driving time records, and is also expected to reduce false and erroneous on-duty, off-duty, and sleeper-berth entries. Using technology to improve recording of CMV driver activity can reduce fatigue by helping carriers to prevent drivers from exceeding driving time and related on-duty time limits as well as by preserving off-duty time for drivers to recover. Use of ELDs could also assist roadside 11 Exec. Order No. 12866, Fed. Reg. 58, No. 190 (September 30, 1993): 51735-51744. 12 MAP-21 §32301

8

inspectors and safety investigators in more rapidly detecting those HOS violations that do occur so that they can cite noncompliant carriers and drivers and hopefully change their behavior by doing so. In addition, the use of ELDs can significantly reduce drivers’ time spent logging their hours and handling and transmitting these logs, and carriers’ time spent handling these logs. The monetary value of these time savings may exceed the costs of the devices.

2.3 OPTIONS CONSIDERED

FMCSA is mandating the installation and use of ELDs for the majority of motor carrier operations. In the SNPRM, the Agency evaluated four options, but based on public comments the Agency evaluated two options in the final rule:

• Option 1: ELDs are mandated for all CMV operations subject to 49 CFR part 395. Under this option 4.27 million drivers would be subject to the rule, requiring the purchase of approximately 3.5 million ELDs or upgrades to an existing fleet management system (FMS).

• Option 2: ELDs are mandated for all CMV operations where the driver is required to complete RODS under 49 CFR 395.8. Under this option 3.51 million drivers would be subject to the rule, requiring the purchase of approximately 2.8 million ELDs or upgrades to an existing FMS.

2.4 SCOPE AND PARAMETERS OF THE ANALYSIS

2.4.1 Presentation of Figures This analysis follows guidance issued by the Office of Management and Budget (OMB) in its Circular A-4.13 The main discussion of the analysis presents annualized costs and benefits discounted at a 7 percent rate over a 10-year period that begins with an anticipated 2017 compliance date of the rule.14 Carriers currently using earlier ELD technology will be required to ensure that their devices are fully compliant, replacing equipment if necessary, 2 years after the compliance date. A summary of annualized costs and benefits is presented with 3 percent discount rate and undiscounted amounts for each of the 10 years of the analysis. Per OMB guidance, because the annual impact of this rule exceeds $1 billion, a formal quantitative analysis of uncertainty is also presented.

2.4.2 Labor Costs FMCSA computes its estimates of labor costs using data gathered from several sources. Labor costs comprise wages, fringe benefits, and overhead. Fringe benefits include paid leave, bonuses and overtime pay, health and other types of insurance, retirement plans, and legally required 13 Office of Management and Budget. Circular A-4. September 17, 2003. http://www.whitehouse.gov/omb/circulars_a004_a-4/. Accessed February 4, 2014. 14 All figures are presented in year 2013 dollars, the last complete year for which price index data are available. Figures not originally expressed in year 2013 dollars have been updated using the gross domestic product (GDP) deflator published by the Bureau of Economic Analysis (more information available at: http://research.stlouisfed.org/fred2/categories/21, and http://www.bea.gov/itable/).

9

benefits (Social Security, Medicare, unemployment insurance, and workers compensation insurance). Overhead includes any expenses to a firm associated with labor that are not part of employees’ compensation; this typically includes many types of fixed costs of managing a body of employees such as manager and human resource staff salaries or payroll services. The economic costs of labor to a firm should include the costs of all forms of compensation and labor related expenses. For this RIA, it has been calculated relative to total compensation (wages plus fringe benefits).

FMCSA has recognized in past RIAs for ELD rules that many CMV drivers are not paid for time spent filling out or forwarding their RODS to carriers. This is especially true for drivers who are paid by the mile. Although administrative activities might be part of their duties, a reduction in time spent with paper RODS provides a benefit in the form of extra leisure time, a benefit that is not paid by their employers. The Agency assumes drivers value this leisure time at the same amount that they accept in exchange for it, that is their base wage plus fringe benefits, which is lower than the full labor cost borne by their employers. This assumption is not necessarily valid for drivers who are paid for this administrative time—the firm’s labor cost, a higher figure, would be appropriate for these drivers—but due to lack of precise data on how drivers are compensated, this RIA uses this lower cost for all driver time savings. Consequently, the benefit of driver time savings may be underestimated. It is also possible that drivers internalize the costs/opportunity costs of their nondriving, unpaid on-duty time into their decision to accept a given wage per mile driven.

The primary source for wages is the current median hourly wage data (May 2013) from the Bureau of Labor Statistics (BLS) Occupational Employment Statistics (OES).15 Even though private carriers would be classified under the industries of their parent companies, FMCSA has assumed that staff working for their motor carrier operations would be paid market wages similar to those of staff working in the trucking industry. (Moreover, wages for certain office occupations would be distorted by those employed in similar occupations in their primary, non-CMV operations.) Because manager and clerical wages are similar for both truck and passenger transportation, this RIA uses wages for the trucking industry. However, for bus and truck mechanics and drivers, because these occupations are more specifically defined, this RIA uses all-industry wages.

BLS does not publish data on fringe benefits for specific occupations, but it does for the broad industry groups in its Employer Costs for Employee Compensation (ECEC) release. This RIA uses an average hourly wage of $22.95 and average hourly benefits of $12.99 for private industry workers in “transportation and warehousing”16 to estimate that fringe benefits are equal to 57 percent ($12.99 ÷ $22.95) of wages. For overhead, the Agency used industry data gathered for the Truck Costing Model developed by the Upper Great Plains Transportation Institute, North

15 Bureau of Labor Statistics (BLS). Occupational Employment Statistics (OES). May 2013. http://www.bls.gov/oes/current/oessrci.htm. Accessed November 13, 2014. 16 Bureau of Labor Statistics (BLS). Table 10: Employer costs per hour worked for employee compensation and costs as a percent of total compensation: Private industry workers, by industry group, June 2014. http://www.bls.gov/news.release/ecec.t10.htm. Accessed November 13, 2014.

10

Dakota State University.17 Research conducted for this model found an average cost of $0.107 per mile of CMV operation for management and overhead, and $0.39 per mile for labor, indicating an overhead rate of 27 percent ($0.107 ÷ $0.39).

For estimates of labor costs for roadside inspectors, FMCSA updated wage figures from its Cost of Roadside Inspections from year 2006 to year 2013 dollars.18 This report gathered annual salary and fringe benefits data from 18 States (another 5 States provided only total compensation, that is wage plus benefits, data), which FMCSA used to create an hourly wage by dividing by 2,080 hours (assuming 52 40-hour work weeks per year). A national average was computed by weighting these by the number of inspectors these States reported to the Motor Carrier Safety Assistance Program for fiscal year 2010. The result was a base wage of $26 per hour and fringe benefits equal to 32 percent of wages. The research for this report also gathered data on overhead, which was provided by six States. The weighted average overhead rate was 16 percent.

Table 6 summarizes the wage, time, and labor cost estimates used in this RIA. Several driver wage estimates are presented for the sake of comparison, although the Heavy and Tractor-Trailer Truck Driver wage is used throughout. FMCSA believes almost all truck drivers affected by this rulemaking would work in this occupation group, and wages for Bus Drivers, Transit, and Intercity, are nearly identical. As shown, this analysis uses $31 per hour for the driver time cost, $42 per hour for the truck and bus mechanic labor cost, $36 per hour for the motor carrier clerical labor cost, and $40 per hour for the roadside inspector labor cost.

Table 6. Wage, Time, and Labor Costs (2013 $)

Occupation

BLS Occupation

Code NAICS

Mean Hourly Wage

Fringe Benefits Overhead

Hourly Cost

Heavy and Tractor-Trailer Truck Drivers 53-3032 All Industry $19.68 57% N/A $31

Light Truck or Delivery Services Drivers 53-3033 All Industry $16.10 57% N/A $25

Bus Drivers, Transit and Intercity 53-3021 All Industry $18.63 57% N/A $29

Bus and Truck Mechanics and Diesel Engine Specialists 49-3031 All Industry $21.21 57% 27% $42

Information and Record Clerks, All Other 43-4199 48400 $18.05 57% 27% $36

Roadside Inspectors N/A N/A $26.00 32% 16% $40

17 Berwick, Farooq. Truck Costing Model for Transportation Managers. North Dakota State University. Upper Great Plains Transportation Institute. 2003. Appendix A, pp. 42-47. http://ntl.bts.gov/lib/24000/24200/24223/24223.pdf. Accessed June 18, 2012 18 Econometrica, Inc. Roadside Inspection Costs. Report to the Federal Motor Carrier Safety Administration. October 11, 2007. http://ntl.bts.gov/lib/51000/51300/51343/Roadside-Inspection-Costs-Oct2007.pdf Accessed February 4, 2014.

11

2.4.3 Crash Costs FMCSA followed guidance from the USDOT Office of the Secretary of Transportation (OST) for valuing fatalities and injuries,19 and data from a 2006 report on the cost of CMV crashes20 for medical and emergency service costs (inflated to year 2013 dollars using the Consumer Price Index for Medical Costs published by the BLS) and property damage costs (inflated to year 2013 dollars using the GDP deflator). Fatalities are evaluated at a $9.20 million (year 2013 dollars) value of statistical life (VSL), which is the figure provided in the guidance published by OST on June 13, 2014. As described in this guidance, this VSL is based upon the prior year’s VSL adjusted for the annual change in the Consumer Price Index for All Urban Consumers (CPI-U) multiplied by the change in the BLS measure of real income growth known as the Median Usual Weekly Earnings (MUWE) taken to a 1.0 power to reflect the most recent OST guidance on the estimated elasticity of VSL with respect to increases in real income:

VSL2013 = VSL2012 x ((CPI-U2013 ÷ CPI-U2012) × ((MUWE2013 ÷ MUWE2012)1.0))

Per that same guidance, injuries are valued according to their abbreviated injury scale (AIS) scores at the fractions of VSL shown below in Table 7.

Table 7. Fractions of VSL for AIS

AIS Level Severity Fraction of VSL Value of VSL Fraction

(2013 $ millions)

AIS 1 Minor 0.003 $0.03

AIS 2 Moderate 0.047 $0.43

AIS 3 Serious 0.105 $0.97

AIS 4 Severe 0.266 $2.45

AIS 5 Critical 0.593 $5.46

AIS 6 Unsurvivable 1.000 $9.20

FMCSA’s crash incident data rely on police accident reports that do not record AIS—this scoring requires the expertise of a medical professional—but do include the police assessment of injuries. However, FMCSA’s Large Truck Crash Causation Study21 contains both AIS scores and police-assessed injury severities. The numbers of CMV crashes and crash victims from the

19 U.S. Department of Transportation. Office of the Undersecretary for Policy. Treatment of the Value of Preventing Fatalities and Injuries in Preparing Economic Analyses. June 13, 2014. https://www.transportation.gov/sites/dot.gov/files/docs/VSL_Guidance_2014.pdf. Accessed December 4, 2014. 20 Zaloshnja, E. and Miller, T. (2006). Unit Costs of Medium and Heavy Truck Crashes. Final Report for the Federal Motor Carrier Safety Administration (FMCSA) and the Federal Highway Administration (FHWA). http://ai.volpe.dot.gov/carrierresearchresults/pdfs/crash%20costs%202006.pdf. Accessed March 23, 2015. Disaggregated figures from Table 2 of that report were adjusted for inflation and to conform to current DOT VSL guidance. Averages presented here use the current distribution of no injury, injury, and fatal crashes. 21 U.S. Department of Transportation. Federal Motor Carrier Safety Administration. Report to Congress on the Large Truck Crash Causation Study. March 2006. http://www.fmcsa.dot.gov/sites/fmcsa.dot.gov/files/docs/ltccs-2006.pdf. Accessed February 4, 2014.

12

National Highway Transportation Safety Administration’s (NHTSA) General Estimates System and Fatality Analysis Reporting System22 were used to compute weighted average costs for non-injury, injury, fatal, and all crashes, as shown in Table 8. The raw figures for the crash costs were drawn from Zaloshnja and Miller,23 adjusted to conform to the June 2014 guidance on the VSL and to account for inflation. Safety benefits were evaluated using the estimated number of crashes avoided multiplied by the average cost per crash of $272,000.

Table 8. Costs of CMV Crashes (2013 $)

Maximum Severity Medical

Emergency Services

Property Damages

Lost Productivity

from Roadway

Congestion

Emissions and

Additional Fuel from Roadway

Congestion Fatalities

and Injuries Total No Injury & Unknown $1,000 $0 $7,000 $13,000 $1,000 $50,000 $72,000

Injury $28,000 $0 $15,000 $16,000 $1,000 $393,000 $453,000 Fatal $71,000 $2,000 $16,000 $42,000 $3,000 $10,885,000 $11,019,000 All $7,000 $0 $9,000 $14,000 $1,000 $241,000 $272,000

Sources: Zaloshnja, E. and Miller, T. (2006). Unit Costs of Medium and Heavy Truck Crashes. Final Report for the Federal Motor Carrier Safety Administration (FMCSA) and the Federal Highway Administration (FHWA). http://ai.volpe.dot.gov/carrierresearchresults/pdfs/crash%20costs%202006.pdf. Accessed March 23, 2015. Disaggregated figures from Table 2 of that report were adjusted for inflation and to conform to current DOT VSL guidance. Averages presented here use the current distribution of no injury, injury, and fatal crashes. Hagemann, G., Hymel, K., Klauber, A., Lee, D., Noel, G., Pace, D., and Taylor, C. Delay and Environmental Costs of Truck Crashes. 2013. U.S. Department of Transportation, Volpe Center.DOT/Volpe Center. Cambridge, MA. http://ntl.bts.gov/lib/48000/48200/48200/Crash-Costs-Final-Report.pdf. Accessed December 4, 2014.

June 2014 guidance from the USDOT Office of the Secretary of Transportation (OST) directs agencies to use a 1.18 percent annual growth rate to project the VSL to future years.24 Table 9 shows projected VSL and the average cost of a crash through 2026, the end of the 10-year analysis time period. Safety benefits are evaluated in each year using the appropriate projected future cost of a CMV crash. 22 http://www.nhtsa.gov/Data/National+Automotive+Sampling+System+(NASS)/NASS+General+Estimates+System, http://www.nhtsa.gov/FARS. Accessed February 4, 2014. 23 Zaloshnja, E. and Miller, T. (2006). Unit Costs of Medium and Heavy Truck Crashes. Final Report for the Federal Motor Carrier Safety Administration (FMCSA) and the Federal Highway Administration (FHWA). http://ai.volpe.dot.gov/carrierresearchresults/pdfs/crash%20costs%202006.pdf. Accessed March 23, 2015. Disaggregated figures from Table 2 of that report were adjusted for inflation and to conform to current DOT VSL guidance. Averages presented here use the current distribution of no injury, injury, and fatal crashes. 24 U.S. Department of Transportation. Office of the Secretary of Transportation. Guidance on Treatment of the Economic Value of a Statistical Life – 2014 Adjustment. June 13, 2014. Pg. 1 and pp. 8-9. https://www.transportation.gov/sites/dot.gov/files/docs/VSL_Guidance_2014.pdf. Accessed July 18, 2014.

13

Table 9. Costs of CMV Crashes (2013 $)

Year VSL Average CMV

Crash Cost 2013 $9,200,000 $272,000 2014 $9,308,600 $275,000 2015 $9,418,400 $278,000 2016 $9,529,500 $281,000 2017 $9,642,000 $283,000 2018 $9,755,800 $286,000 2019 $9,870,900 $290,000 2020 $9,987,400 $292,000 2021 $10,105,200 $296,000 2022 $10,224,400 $299,000 2023 $10,345,100 $302,000 2024 $10,467,200 $305,000 2025 $10,590,700 $308,000 2026 $10,715,700 $312,000

2.5 BASELINE FOR ANALYSIS

2.5.1 Numbers of Carriers and Drivers Crash risks from fatigue and HOS violations vary by length of haul, as recognized in the Federal Motor Carrier Safety Regulations (FMCSR). The HOS record-keeping requirements for short-haul (SH) are less stringent than those for long-haul (LH) operations, subject to certain conditions. SH drivers are generally not required to keep RODS if they work fewer than 12 hours a day, begin and end their work days at the same location, and operate within a 100-air mile radius (under the provisions of 49 CFR 395.1(e)(1)) or operate certain property-carrying CMVs within a 150-air mile radius (under the provisions of 49 CFR 395.1(e)(2)).25

FMCSA analyzed data from its Motor Carrier Management Information System (MCMIS) as of December 27, 2013, to estimate the numbers of carriers, drivers, and CMVs affected by this rule. From these data, FMCSA estimates that it currently regulates about 539,000 interstate passenger carriers, interstate property carriers, and intrastate hazardous materials carriers. These carriers currently maintain 4,604,000 CMVs and employ 4,409,000 drivers.26 Many of these CMVs are used in operations not subject to the HOS rules, and many operations subject to the HOS rules are exempt from using RODS under the SH provisions. FMCSA has assumed that carriers will acquire enough ELDs for each one of their drivers, but not for every CMV the carrier operates. Many CMVs are used in operations that will not require a driver to use RODS. The Agency has also assumed that school bus drivers will usually operate under an exemption from the FMCSRs, including the HOS rules, and that the number of times they are not exempt will fall under the minimum threshold for ELD use.

25 These SH drivers are allowed to substitute time-cards for RODS. 26 U.S. Department of Transportation, Federal Motor Carrier Safety Administration. 2014 Pocket Guide to Large Truck and Bus Statistics. October 2014 Update. Table 1-7, p. 11.

14

In the 2014 Pocket Guide to Large Truck and Bus Statistics, FMCSA estimated that as of January 2014 there are 2.2 million intrastate CMV drivers.27 Of those intrastate drivers, 880,000 work for carriers regulated directly by FMCSA. Of these 880,000 drivers, 82.3 percent are short-haul and do not keep RODS.28 The remaining 156,000 are long-haul drivers that the Agency assumes keep RODS. The 1,320,000 intrastate drivers who work for intrastate carriers (that do not carry hazardous materials) are not regulated directly by FMCSA. MCMIS does not systematically record data on intrastate drivers in this group. However, the Agency assumes that most are not long-haul drivers required to keep RODS. As a conservative estimate, FMCSA assumes that the number of intrastate RODS drivers working for intrastate carriers is 156,000—the same as the number of intrastate RODS drivers working for FMCSA-regulated carriers. This assumption implies that about 89 percent of intrastate drivers working for intrastate carriers are not required to keep RODS, slightly higher than the 82.3 percent of intrastate drivers working for FMCSA-regulated carriers. FMCSA assumes that drivers for intrastate carriers are more likely to drive within a SH air-mile radius. Based on these assumptions, FMCSA estimates that 312,000 intrastate drivers are required to keep RODS and are subject to this rulemaking.

Table 10 shows the Agency’s current estimate of drivers and carriers subject to the HOS and RODS requirements. LH drivers working for for-hire carriers are generally paid by the mile, and FMCSA has assumed that the SH drivers working for these carriers would also be paid this way rather than on an hourly or time-card basis. The exemption from RODS for SH drivers is only applicable if time-cards are available as alternative documentation for driving and on-duty time. Consequently, the Agency has assumed that all drivers at for-hire carriers will be required to keep RODS. However, the Agency assumes that SH drivers working for passenger carriers and private property carriers will be paid hourly, and therefore will keep time-cards that document driving and on-duty time in lieu of RODS. These drivers generally operate local private delivery trucks, private service vehicles, taxis and limousines, mini-buses, or passenger vans.

27 U.S. Department of Transportation, Federal Motor Carrier Safety Administration. 2014 Pocket Guide to Large Truck and Bus Statistics. October 2014 Update. Page 7, and Figure 1-4 on p. 9. 28 FMCSA, Motor Carrier Management Information System (MCMIS) registration data as of December 27, 2013, and the FMCSA 2014 Pocket Guide to Large Truck and Bus Statistics, October 2014 Update.

15

Table 10. Carriers and Drivers Subject to the Rule

Type of Entity

Type of Operation

TOTAL

For-Hire Private General Freight

Specialized Freight Passenger Property Passenger

CARRIERS SUBJECT TO THE RULE

FMCSA-Regulated Carriers 176,000 152,000 8,000 197,000 6,000 539,000

DRIVERS SUBJECT TO THE RULE

Long-Haul RODS Drivers (Intrastate HM & all Interstate)

1,223,000 474,000 65,000 480,000 18,000 2,260,000

Long-Haul RODS Drivers (Intrastate)29

82,000 78,000 20,000 128,000 4,000 312,000

Short-Haul RODS Drivers (Intrastate HM & all Interstate)

495,000 298,000 0 0 0 793,000

Total Drivers Subject to ELD Requirement in Final Rule30

1,800,000 850,000 85,000 608,000 22,000 3,365,000

DRIVERS NOT SUBJECT TO THE RULE

Short-Haul Drivers Subject to HOS, but Exempt from RODS31

0 0 86,000 551,000 13,000 650,000

Source: FMCSA, Motor Carrier Management Information System (MCMIS) registration data as of December 27, 2013, and the FMCSA 2014 Pocket Guide to Large Truck and Bus Statistics, October 2014 Update.

The number of drivers is an input into the calculation of FMCSA’s cost and benefits estimate and a significant contributor to the magnitude of the rule’s impact. However, the effects of different sources of costs and benefits are not uniform for all drivers and CMVs. For example, the safety benefits of the HOS rules are highest for LH drivers, those most at risk to suffer from inadequate rest because of longer continuous daily driving periods, irregular wake and sleep schedules, and nights spent away from home. LH carriers are the most likely to already be using FMS that include ELD functionality to efficiently route their drivers or to track their whereabouts. SH operations accrue smaller safety benefits from the HOS rules, and those SH carriers exempt from

29 Intrastate long-haul RODS drivers include intrastate long-haul RODS drivers who work for interstate or intrastate carriers. 30 Only these 3,365,000 drivers currently required to keep RODS will be subject to the final rule (which adopted Option 2 of the RIA). For purposes of evaluating Option 2 (the adopted option), FMCSA used a driver population based on the estimated number of RODS drivers in 2017 using an industry growth rate of 1.078%, and then estimated the number of these drivers that will need to either obtain an ELD or upgrade an existing FMS (approximately 2.8 million drivers). See Table 15 for more information. 31 These 650,000 drivers are not subject to the ELD requirements of the final rule (which adopted Option 2 of the RIA). However, these drivers were included in Option 1 of the RIA. Under Option 1, 4,015,000 drivers are subject to the rule (comprised of the 3,365,000 drivers subject to Option 2, plus the 650,000 short-haul drivers subject to HOS but exempt from RODS), approximately 3.3 million of which would be affected by the rule by needing to obtain ELDs or upgrade an existing fleet management system (FMS). As with Option 2, for Option 1 an industry growth rate of 1.078% annually was applied to the driver population.

16

paper RODS requirements likely have no voluntary ELD use and would not benefit from the elimination of paper RODS afforded by ELDs.

2.5.2 HOS Compliance Costs To construct the current baseline estimate of compliance costs, the Agency began with the baseline level of noncompliance with the 60- and 70-hour rule, the 10- and 11-hour driving time limit (for motorcoach and truck drivers, respectively), the 14- and 15-hour rule, and the 34-hour restart. Also, the Agency considered the changes that took effect on July 1, 2013; the limit on the use of the 34-hour restart to once per week, and the requirement for a 30-minute break during the work shift (49 CFR 395.3).

2.5.3 Methodology for ELD Effectiveness ELD data are not continuously monitored by carriers or enforcement staff for violations, and even if they were, not all HOS violations, such as those not related to driving time, would be documented. Consequently, the Agency cannot assume that ELD use would lead to perfect compliance with the HOS regulations. Compliance costs and safety benefits were therefore reduced to reflect the limitations of the devices. FMCSA calculates its measure of ELD effectiveness based on the percentage change in HOS rule violation rates in roadside inspections. The Agency constructed this estimate using data from two carriers that had voluntarily installed advanced-AOBRDs (devices that for the most part meet the ELD technical specifications) and three carriers that agreed in court settlements to install advanced-AOBRDs in lieu of paying civil penalties for past violations. This result attempts to account for possible bias in the reduction in HOS OOS violations for these carriers—particularly, the carriers in this sample may have a larger than average impact to their violations from ELD use because all were attempting to reduce violations anyway. FMCSA constructed a confidence interval around the estimated percent reduction and used the lower bound, 45 percent, in this analysis. Appendix A contains the full details of how this figure was calculated.

2.5.4 Methodology for Safety Benefits The Agency was not able to construct statistically significant measures of safety improvements for carriers that installed ELDs by directly examining the crash data of these carriers because a crash is a rare occurrence for an average CMV. According to FMCSA’s 2012 Large Truck and Bus Crash Facts report, on average, only about 3.4 percent of CMVs are involved in crashes each year.32 This makes estimating the crash reduction using only the small sample of CMVs used in the effectiveness measure difficult. ELDs, by reducing HOS violations, reduce the crash risk for the entire CMV population; only by evaluating this reduction in crash risk against the total population of CMVs can FMCSA derive meaningful results.

To construct an indirect estimate of the safety benefits of ELD use, the Agency combined data from two different sources: data on reductions in HOS OOS violations in carriers that installed

32 U.S. Department of Transportation. Federal Motor Carrier Safety Administration. Large Truck and Bus Crash Facts 2012. June 2014. http://www.fmcsa.dot.gov/safety/data-and-statistics/large-truck-and-bus-crash-facts-2012, Accessed December 4, 2014. Calculated as 4,053 fatal crashes (Table 1) plus 89,000 injury crashes (Table 2) plus 295,000 property damage only crashes (Table 3), divided by 11,423,889 registered CMVs (Table 1).

17

ELDs and data on crash risks associated with HOS OOS violations from the Agency’s model of the effectiveness of roadside inspections and traffic stops. Crash risk is reduced during a period after a CMV driver has been cited for an HOS OOS violation. The Agency estimated this crash risk reduction over a 30-day period after each citation to capture the immediate effects of specific episodes of fatigue being avoided and short-term secondary effects, including deterrence effects. The Agency uses these estimated crash reductions as a proxy for those eliminated by ELD use. ELDs do not eliminate all HOS violations. To measure the number and types of violations eliminated, FMCSA examined data from the settlement agreement carriers and the voluntary users.

FMCSA has been estimating the crash risk reduction for every FMCSR violation in its Roadside Intervention Model.33 The Roadside Intervention Model uses a risk-based approach to estimate the benefits of the roadside inspection program. This model estimates crash risk reductions from traffic enforcement and roadside inspections using data from all traffic stops and roadside inspections. An OOS order issued for a violation found during a roadside inspection immediately prevents a CMV or driver from further operation until the violation has been rectified, thereby minutely reducing its risk of being involved in a crash in the future. The intervention model also estimates a secondary effect beyond the period immediately following a violation, implicitly capturing the short-term deterrence effect. For HOS violations, the duration of those effects was estimated to be 30 days.

ELDs can be thought to function in a similar manner, although not by placing a driver OOS per se, but by preventing many HOS violations from occurring in the first place. A crash risk reduction estimate can be applied to the same violation reductions used to calculate the ELD effectiveness measure. It must be noted that the Model measures safety improvements that wane over discrete short-term periods, but ELDs may lead to continuous improvements in driver HOS compliance because the truck will always be carrying the ELD. Consequently the crash risk reductions from ELD use would likely be higher than that those associated with violations cited during a roadside inspection or a traffic stop.

Table 11 shows the crash risk reduction estimates from the FY2009 Roadside Intervention Model. These were estimated using a sample period from January 2005 through September 2009. The table shows the annual crash reduction for each violation cited scaled to 240 working days per year per driver.34 The figures represent crash reductions associated with remediated violations and secondary short-term deterrence effects following a violation. Since fractional crashes do not occur, crash reductions less than one should be thought of in terms of the number of violations that occur for each crash associated with that violation. For example, an 11-hour rule35 violation has an estimated crash risk reduction of 0.02496, which indicates that 1 in 40 (1

33 U.S. Department of Transportation. Federal Motor Carrier Safety Administration. FMCSA Safety Program Effectiveness Measurement: Intervention Model Fiscal Year 2009. http://ntl.bts.gov/lib/48000/48100/48198/FMCSA-Intervention-FY-2009.pdf. Accessed February 4, 2015. 34 The Roadside Intervention Model estimates that crash reductions persist for 30 days after an HOS violation is issued due to drivers being deterred from committing another violation in that period. Because ELD’s in principle permanently deter many violations, one can apply the estimated 30-day effect to an average 240 working days per driver by multiplying the output from the model by a factor of 8. 35 49 CFR 395.3(a)(1) (2001).

18

÷ 0.02496) of these violations are associated with a crash, or that FMCSA would have eliminated 1 crash when 40 11-hour rule violations are eliminated.

Table 11. Estimated Crash Reductions for HOS Violations36

Violation Description Crash

Reduction 395.13(d), 395.13(d)(1), 395.13(d)(2) Driving after being declared out-of-service 1.37784 395.15(b), 395.15(b)(5), 395.15(c), 395.15(d)(1), 395.15(f), 395.15(g), 395.15(h)(3), 395.15(i)(5)

On-board recording device information requirements not met 0.02952

395.3(a)(1) Requiring or permitting driver to drive more than 11 hours 0.02496

395.3(a)(2) Requiring or permitting driver to drive after 14 hours on duty 0.02496

395.8(a) No driver’s record of duty status 0.02952 395.8(a)(1) Other Log/Form and Manner 0.00521 395.8(a)(2) Incomplete/Wrong Log 0.02952 395.8(c) Other Log/Form and Manner 0.00521 395.8(d)(1) Other Log/Form and Manner 0.00521 395.8(d)(2), 395.8(d)(4), 395.8(d)(5), 395.8(d)(6), 395.8(d)(7), 395.8(d)(8), 395.8(d)(9), 395.8(d)(10), 395.8(d)(11)

Other Log/Form and Manner 0.02952

395.8(e) False report of driver’s record of duty status 0.05088 395.8(f)(1), 395.8(f)(2), 395.8(f)(3), 395.8(f)(4), 395.8(f)(5), 395.8(f)(6), 395.8(f)(7), 395.8(f)(9), 395.8(f)(9), 395.8(f)(10), 395.8(f)(11), 395.8(f)(12)

Driver’s record of duty status not current 0.02952

395.8(g) Other Log/Form and Manner 0.02952 395.8(h)(1), 395.8(h)(2), 395.8(h)(4), 395.8(h)(5) Other Log/Form and Manner 0.02952

395.8(i) Incomplete/Wrong Log 0.02952 395.8(j)(2) Other Log/Form and Manner 0.02952 395.8(k)(1), 395.8(k)(2) Incomplete/Wrong Log 0.02952 Source: FMCSA Safety Program Effectiveness Measurement: Intervention Model Fiscal Year 2009. Crash reduction figures adjusted from 30 to 240 driver working days.

In addition to the analysis of safety benefits in this rule, FMCSA commissioned a study to evaluate the effects of ELD use.37 The Safety Study used a different methodology and data set than the RIA, but also found evidence that ELDs increase safety. Thus, while the benefit

36 http://www.fmcsa.dot.gov/regulations/title49/part/395 37 Hickman, Jeffrey S., Matthew C. Camden, Feng Guo, Namoi J. Dunn, and Richard J. Hanowski. Evaluating the Potential Safety Benefits of Electronic Hours-of-Service Recorders. Final Report. April 2014. http://ntl.bts.gov/lib/51000/51800/51846/13-059-Evaluating_the_Potential_Safety_Benefits_of_Electronic_HOS--Full_Report.pdf. Accessed December 19, 2014.

19

estimates in the RIA do not depend on the Safety Study, the Safety Study provides corroborative evidence that ELDs provide safety benefits relative to paper logs.

The Safety Study focused on estimating the effects of ELDs on outcome measures of safety such as crash rates rather than process measures such as violation rates and fatigue. The Study found a significant reduction in the overall crash rate and the preventable crash rate for trucks with ELDs compared to trucks without ELDs. Due to limited data, the study could not evaluate the effect of ELDs on DOT-reportable and fatigue-related crashes. However, DOT-reportable and fatigue related crashes are included in the overall and preventable crash data.

2.5.5 Current AOBRD/ELD and FMS Use Many carriers already use AOBRDs and FMS. Their capabilities and the carrier’s use of their functions vary. As it did in RIAs prepared for previous ELD rulemakings, the Agency excludes any voluntary use of AOBRDs or ELDs from its analysis. The specifications for AOBRDs have been in the FMCSRs since 1988, and although some sectors of the motor carrier industry (specifically, many private motor carriers of property)38 use AOBRDs or FMS (with AOBRD or ELD functionality) extensively, the analysis in the SNPRM estimated that the adoption rate by motor carriers, as of 2014, is only approximately 19 percent for LH operations and 5.6 percent for SH operators. Appendix B contains a detailed discussion of the estimates of AOBRD use.

This RIA uses an updated model for forecasting voluntary use of ELDs or AOBRDs and FMS without any ELD mandate. Details on how this forecast was constructed are presented in Appendix B of this RIA. The Agency forecasted voluntary use only for LH operations, and based on data observed in 2005, assumed that voluntary use for SH operations would be one-third that of LH. FMCSA also assumed that without the regulation, voluntary use of ELDs would eventually approach 100 percent.39 As a consequence, the Agency assumes that SH voluntary use will never exceed 33 percent. SH CMVs operate close enough to their terminals that HOS monitoring and CMV tracking are not problematic, and those features of these devices are not as useful. Table 12 presents the forecasts of voluntary ELD and FMS use that would occur without the ELD mandate. FMCSA has been monitoring data on ELD use coming primarily from manufacturers to assess the performance of the model. David Kraft, Qualcomm’s Director of Industry Affairs, hosted a webinar on August 15, 2012, in which he reported that his company estimates approximately 500,000 ELDs were currently in use, up from 200,000 units in 2009. During a follow-up telephone conversation, Mr. Kraft further explained that this figure was based on estimates of Qualcomm’s and its largest competitors’ market shares, and that the estimate fell into a range from 420,000 to 610,000 units. FMCSA’s model estimated 485,000 ELDs in use at the beginning of 2013 (see the second to last column of Table 15), slightly below Qualcomm’s range. However, FMCSA’s model estimated 284,000 ELDs in use as of 2009, which was higher than Qualcomm’s estimate, and FMCSA’s 2005 data on AOBRD use indicated

38 Leavitt, Wendy. “HOS and electronic logs: fleets hold lead.” FleetOwner. October 28, 2009. http://fleetowner.com/management/news/hos-electronic-logs-1028/. Accessed June 30, 2010. 39 FMCSA estimates that 100 percent adoption would occur around 2090. However, for the period of this analysis, voluntary adoption would only reach 43%. See Table 12 below and Figure 1 in Appendix B for additional information.

20

that nearly 200,000 devices were already in use in 2005. While there is currently no definitive data on ELD use, the data available from multiple sources, despite diverging in particular estimates, are following similar trends.

Table 12. Current and Projected ELD and FMS Use without an ELD Mandate

Year ELD/AOBRD

LH ELD/AOBRD SH LH FMS (including those

with ELD) SH FMS (including

those with ELD) 2013 17% 6% 41% 14%

2014 19% 6% 43% 14%

2015 20% 7% 45% 15%

2016 22% 7% 47% 16%

2017 23% 8% 49% 16%

2018 25% 8% 51% 17%

2019 27% 9% 53% 18%

2020 29% 10% 55% 18%

2021 31% 10% 57% 19%

2022 33% 11% 59% 20%

2023 35% 12% 60% 20%

2024 37% 12% 62% 21%

2025 39% 13% 64% 21%

2026 41% 14% 66% 22%

2027 43% 14% 67% 22%

As Table 12 shows, FMCSA estimates that by the 2017 compliance date, 23 percent of LH CMVs and 8 percent of SH CMVs will have some kind of electronic HOS recording device. In addition, it estimates that 49 percent of CMVs will have voluntarily installed an FMS device. Many of these FMS devices will already have electronic HOS recording capabilities, and the Agency believes, according to past research, those that do can have this recording function activated for a relatively low additional monthly cost.

2.5.6 Total ELDs Required The estimates presented in the previous two sections are used to produce estimates of the number of ELDs that would need to be produced to meet the Agency’s mandate. A large portion of the ELD supply will be met by existing FMS which, as discussed in the next section, significantly reduces the cost of this rule because the FMS upgrade is estimated to be significantly cheaper than the purchase of any new device. FMCSA used the BLS job outlook estimate for 2012-2022 for heavy and tractor-trailer truck drivers to estimate an annual growth for LH and SH drivers.40 Based on this projected change, FMCSA applied an annual growth rate of 1.078% for LH and

40 Bureau of Labor Statistics. Occupational Outlook Handbook - Heavy and Tractor-trailer Truck Drivers. http://www.bls.gov/ooh/transportation-and-material-moving/heavy-and-tractor-trailer-truck-drivers.htm

21

SH drivers. Tables 13 and 14 present the estimates of new ELDs and upgradeable FMS needed for the LH and SH drivers presented in Table 10.

Table 13. Estimates of LH ELD Use, Without ELD Mandate

Year ELD LH

(a)

FMS LH (b)

LH (=LH Drivers,

Table 10) (c)

LH with Non-ELD FMS

(d) = ((b - a) × c)

LH without ELD or FMS

(e) = ((1- a) × c - d)

LH with ELDs

(f) = (a × c)

2013 17% 41% 2,572,000 619,000 1,513,000 440,000

2014 19% 43% 2,599,726 640,000 1,478,000 482,000

2015 20% 45% 2,627,751 660,000 1,440,000 528,000

2016 22% 47% 2,656,078 678,000 1,402,000 576,000

2017 23% 49% 2,684,711 693,000 1,365,000 627,000

2018 25% 51% 2,713,652 706,000 1,326,000 681,000

2019 27% 53% 2,742,905 717,000 1,288,000 738,000

2020 29% 55% 2,772,474 725,000 1,249,000 799,000

2021 31% 57% 2,802,361 731,000 1,210,000 862,000

2022 33% 59% 2,832,570 733,000 1,172,000 928,000

2023 35% 60% 2,863,106 734,000 1,133,000 997,000

2024 37% 62% 2,893,970 731,000 1,095,000 1,068,000

2025 39% 64% 2,925,167 726,000 1,057,000 1,142,000

2026 41% 66% 2,956,700 718,000 1,020,000 1,219,000

2027 43% 67% 2,988,573 708,000 983,000 1,298,000

22

Table 14. Estimates of SH ELD Use, Without ELD Mandate

Year ELD SH

(a)

FMS SH (b)

SH RODS Drivers (=SH RODS

Drivers, Table 10) (c)

SH RODS Drivers with Non-ELD

FMS (d) = ((b - a) × c)

SH RODS Drivers without ELD or FMS (e) = ((1- a) × c - d)

SH RODS Drivers with

ELDS (f) = (a × c)

2013 6% 14% 793,000 64,000 684,000 45,000

2014 6% 14% 801,549 66,000 686,000 50,000

2015 7% 15% 810,189 68,000 688,000 54,000

2016 7% 16% 818,923 70,000 690,000 59,000

2017 8% 16% 827,751 71,000 692,000 64,000

2018 8% 17% 836,674 73,000 694,000 70,000

2019 9% 18% 845,694 74,000 696,000 76,000

2020 10% 18% 854,810 75,000 698,000 82,000

2021 10% 19% 864,025 75,000 700,000 89,000

2022 11% 20% 873,339 75,000 703,000 95,000

2023 12% 20% 882,754 75,000 705,000 102,000

2024 12% 21% 892,270 75,000 707,000 110,000

2025 13% 21% 901,889 75,000 709,000 117,000

2026 14% 22% 911,611 74,000 712,000 125,000

2027 14% 22% 921,438 73,000 715,000 133,000

23

Table 15. Estimates of Industry ELD Use without Mandate

Year RODS

Drivers

RODS Drivers

with Non-ELD FMS

RODS Drivers without

ELD or FMS RODS Drivers without ELDs

SH Non RODS

Drivers (no ELD or

FMS)

RODS Drivers with

ELDs

RODS Drivers without

ELDs Driving Pre-2000

CMVs

2013 3,365,000 683,000 2,197,000 2,880,000 650,000 485,000

2014 3,401,275 706,000 2,163,275 2,869,275 657,007 532,000

2015 3,437,940 728,000 2,127,940 2,855,940 664,090 582,000

2016 3,475,001 748,000 2,092,001 2,840,001 671,248 635,000

2017 3,512,462 764,000 2,057,462 2,821,462 678,484 691,000 159,000

2018 3,550,326 779,000 2,020,326 2,799,326 685,799 751,000 120,000

2019 3,588,599 791,000 1,983,599 2,774,599 693,191 814,000 85,000

2020 3,627,284 800,000 1,946,284 2,746,284 700,664 881,000 62,000

2021 3,666,386 806,000 1,909,386 2,715,386 708,217 951,000 47,000

2022 3,705,910 808,000 1,874,910 2,682,910 715,852 1,023,000 38,000

2023 3,745,859 809,000 1,837,859 2,646,859 723,569 1,099,000 32,000

2024 3,786,240 806,000 1,802,240 2,608,240 731,369 1,178,000 26,000

2025 3,827,055 801,000 1,767,055 2,568,055 739,253 1,259,000 20,000

2026 3,868,311 792,000 1,732,311 2,524,311 747,222 1,344,000 16,000

2027 3,910,011 781,000 1,698,011 2,479,011 755,277 1,431,000 12,000

To assess costs associated with obtaining a new ELD, FMCSA used the 2017 population of RODS drivers without ELDs as well as the number of new drivers, based on a 1.078% annual growth rate, that would need to obtain ELDs during the period of this analysis. This represents the number of ELDs that would need to be purchased as a result of this rule. A number of drivers already have FMS devices or AOBRDs that can be made ELD compliant, and therefore those drivers are not assigned the full new-device cost estimate for this provision. Drivers with an FMS device are assigned an FMS upgrade cost that is significantly lower than the cost of a new ELD. Drivers with an AOBRD are assigned a retrofit cost, discussed in more detail in Section 3.1.4. For Option 2, there are approximately 2.02 million RODS drivers without an ELD and 764,000 RODS drivers with an FMS device that will need to be upgraded to be ELD compliant. Option 1 includes the driver population subject to Option 2 (3.51 million) plus the 678,000 short-haul drivers subject to HOS but exempt from RODS, as presented earlier in Table 10. The total estimate for drivers needing to obtain a new ELD or upgrade an existing FMS under Option 1 is approximately 3.5 million drivers. In response to issues raised by commenters to the SNPRM that ELDs might not be compatible with older CMVs, FMCSA is excepting drivers operating pre-2000-model-year CMVs. These amount to a small number of drivers and CMVs, as shown in the last column of Table 15. As these older vehicles are retired, drivers will have to acquire ELDs for their newer replacement vehicles. This represents a continual increase in the percentage of RODS drivers using ELDs,

24

from about 95 percent in 2007, to over 99.5 percent in 2027. The costs and benefits of this increase in ELD coverage have been accounted for in this analysis. SH drivers who do not use RODS require special consideration in this analysis. FMCSA has assumed, due to their RODS exemption and operating in such close proximity to their terminals, that they do not have ELDs and are unlikely to use more sophisticated FMS that would be capable of being upgraded to be compliant with the ELD final rule. Consequently, including this subpopulation in the mandate would likely require carriers to purchase an ELD for each one of these vehicles. FMCSA, however, will not require ELD use for those SH carriers that do not use RODS more than 8 days out of any 30-day period. In addition, because drivers of these vehicles do not use RODS, they would not gain any paperwork reduction benefit by using ELDs.

25

3. COSTS OF FINAL RULE

3.1 ELD DEVICE COSTS

The Agency has calculated annualized costs of $419 for a new ELD purchase, and $93 per CMV with FMS capable of being upgraded to be compliant with the ELD final rule. The modest reduction in device cost reflects the most up-to-date prices consistent with the technical requirements of the final rule. Some carriers that have already adopted AOBRDs will have to replace their older devices in 2019 (2 years after the effective date of the final rule). FMCSA estimates that the annualized cost of replacing an older AOBRD is $128 per unit. Table 16 summarizes these costs, and the three sections that follow explain in detail what assumptions the Agency made and how these costs were calculated.

Table 16. Costs of ELDs and Other Hardware (2013 $)

Purchase Installation

Replace in Year 5 & Year 10

(Purchase & Instal-

lation) 41 Monthly

Fees

Total Annualized Costs

(7% Discount Rate) ELD with USB42 $560 $2043 $580 $0 $166 ELD With Telematics44 $500 $84 $626 $20 $419

Upgrade FMS $845 $93

3.1.1 New ELD Costs The Agency is basing its estimate of ELD purchase cost on one FMS that offers electronic HOS monitoring and features similar to those of most devices the Agency researched. Cost information was gathered from the vendor’s marketing material.46 The $500 purchase price of this unit is similar to that of other units, and its manufacturer has the largest share of the FMS market. Therefore such a unit is representative of the most abundantly available types of FMS that offer HOS monitoring. Appendix C contains a broader discussion of the types of ELDs

41 FMCSA assumes than an ELD requires 1 hour to remove, and the new ELD 2 hours to install: 3 × $42 = $126. 42 Based on cost of Continental’s VDO RoadLog. Purchase price includes $500 for device including USB driver key, and $60 cable to connect to Engine Control Module (ECM). Device is portable and can be simply plugged into the ECM or hardwired into engine. Only device with built in printer. No monthly fees required without telematics. Adding telematics capabilities requires the purchase of a RoadLog connect driver key for $243. RoadLog connect will work with smartphone free app. http://www.partdeal.com/gauges/roadlog.html. Accessed September 14, 2014. 43 The installation cost is only $20 for this device—the VDO RoadLog—because it is portable and can be installed in about 15 minutes. 44 Based on Omnitracs MCP50. The price of the device decreased from $799 in the SNPRM to $500 currently. 45 Based on Omnitracs MCP50 $28 per month fees for Compliance Plan and $20 per month fees for Core Plan, the basic FMS plan without. http://fleet.omnitracs.com/rs/omnitracsllc1/images/LCL1147_03-14_ApplicationBundles_and_DataPlans_Brochure.pdf . Accessed November 14, 2014. 46 Omnitracs MCP50 EOBR Plan. http://www.omnitracs.com/eobr-plan. Accessed November 13, 2014.

26

currently available, their purchase costs, and any other fees they might have.47 FMCSA assumes that this device will remain operational for about 4.5 to 5 years before it needs to be replaced.48 This assumption is based on the manufacturer’s warranty material (3 years) and Internal Revenue Service guidance on depreciating computing equipment (full depreciation is reached in year 5).49 A salesperson for this vendor indicates it generally covers all repairs within a 3-year initial warranty period, and that it has been charging $300 to $500 per repair outside of the warranty period. All told, FMCSA assumes that using an ELD will entail hardware purchase or repair costs in year 1 of $500, year 5 of $500, and year 10 of $500.50

Two other costs related to the device are considered in the analysis. The first is the cost of installation. The vendor of the device FMCSA is using in this analysis states that installation would take no more than 2 hours per vehicle. FMSCA calculates labor costs of this activity to be $42 per hour, for a total installation cost of $84. The second is a monthly service fee. For this unit the monthly fee is $20, which will result in an annualized cost of $232.

In this analysis, the Agency assumes that carriers own or have access to computer technology, such as personal computers, tablet devices, or smartphones, as the use of one or more of these types of devices is thought to be nearly ubiquitous among businesses.51 Therefore, the analysis did not consider an estimate of the cost of purchasing these types of new equipment.

3.1.2 Annualized ELD Cost FMCSA estimates a total annualized cost for an ELD of $419, which will be used throughout the rest of this analysis. Amortizing purchase and repair costs—$500 for purchase plus $84 for installation in year 1, and $500 for repair or purchase plus $126 for uninstallation and installation in years 5 and 10—at a 7 percent discount rate into an annual amount results in an annualized cost of $187 for hardware. The RIA prepared for the SNPRM used an annualized device cost of $495, which FMCSA acknowledged was on the high end of the range of costs of existing units. The difference between the SNPRM estimate and the estimate for the final rule is largely due to the reduction in the purchase price from $799 to $500. As discussed in appendix C, several more

47 The estimate of a $500 purchase price that is used in this RIA is based not on the electronic logging device with the lowest unit cost as presented in Table 51 of Appendix C, but instead is based on the unit cost of the most widely utilized and available exemplar electronic logging device presented in Table 51 in Appendix C, the Omnitracs MCP50, which is considered an appropriate example of a current state-of-the-art, widely available device with ELD functionality. Furthermore, rather than attempt to estimate what percentage of those acquiring an ELD under the final rule would purchase a device with a unit cost less than, or even greater than, the Omnitracs MCP50, the $500 unit price of the Omnitracs MCP50 was used for this analysis. This can be considered a conservative assumption, given that there may be minimally compliant ELDs available for a lesser unit cost. 48 FMCSA assumes that if a device needs to be repaired, the truck in which it is installed will not need to be placed out of service. FMCSA assumes that repairs could be done during routine maintenance, or at a time when the truck was otherwise non-operational. 49 Department of Treasury. Internal Revenue Service (IRS). Instructions for Form 4562: Depreciation and Amortization. http://www.irs.gov/pub/irs-pdf/i4562.pdf. Accessed February 4, 2014. 50 Cost information for the Qualcomm Mobile Computing Platform (MCP) 100 was acquired through a conversation with Qualcomm sales representative, Angelo Matera, and FMCSA during a meeting on May 12, 2010. 51 Even among smaller carriers such as owner-operators, recent data suggest that approximately 85% already own a computer. http://www.ooida.com/OOIDA%20Foundation/RecentResearch/OOfacts.asp. Accessed November 12, 2014.

27

devices with a price range extending down to $166 on an annualized basis are now produced. These units have lower prices because they have fewer FMS functions, not because of declines in manufacturing and component costs. For users who do not require additional functionality, they could be expected to utilize a lower priced device that is available and meets their particular needs at the time of purchase. ELDs comprise several technologies that have been available for many years, and FMCSA is skeptical that these technologies will become much cheaper.

The impact that this rule will have on carriers, particularly small carriers required to purchase an item of capital equipment, is of some concern. FMCSA assumes that some carriers will not be able to purchase an ELD outright; we do not have specific information that indicates this is too large a burden, but we acknowledge that we have no information about the typical small carrier’s cash flow. When making this type of purchase, some companies finance the acquisition. For small amounts (less than 10 devices) a bank or similar financial institution may not be interested in creating a loan. Other sources of financing are available to small businesses. FMCSA believes a purchase agreement with an ELD manufacturer would, if the carrier has good credit, be financed at competitive market-based interest rates. The interest rates that we use in our analysis already, 3 percent and 7 percent, can serve as a realistic range of competitive interest rates. Other avenues might include corporate credit cards or a lease. The table below shows a range of yearly payments that a carrier might face when financing the purchase of one ELD. We use a range of prices from $200 to $800, and 3 percent and 7 percent interest rates. We assume that the term of financing would be 5 years, since that is the assumed lifespan of the devices.

Table 17: ELD Purchase Payment Estimates (2013 $)

Device Cost Interest Rate Yearly Payment $200 3% $43.67

7% $48.78 $800 3% $174.68

7% $195.11

3.1.3 ELD Monitoring on Current FMS Many FMS currently offer an HOS monitoring feature. Vendors of these devices have told the Agency that they are already modifying the functionality of these devices to comply with the technical requirements of EOBR 1. Almost all FMS currently sold offer the two most important hardware features necessary for ELDs, location tracking and synchronization with the CMV engine. The Agency believes that most upgrades will be relatively inexpensive software changes. Although HOS monitoring is often included in a single monthly fee, some manufacturers offer a la carte pricing for that particular feature. Two vendors contacted for EOBR 1 quoted prices of $5 or $8 per month for this additional service,52 and JJ Keller currently has a $15 price differential for FMS with the ELD feature. The result of amortizing the additional monthly fee of $8 at a 7 percent discount rate to a single annual amount is $93 per year.

52 RouteTracker charges an additional $5 for HOS functionality and Qualcomm charges $8. Qualcomm has since dropped its a la carte pricing for this function.

28

3.1.4 AOBRD Replacement Costs This final rule requires motor carriers using AOBRDs to refit or upgrade these devices to meet the ELD performance standards within 2 years of the compliance date of this rule. FMCSA believes that many of these devices would have to be replaced with new ELDs. Any noncompliant devices that are still in operation 2 years after the compliance date will have to be replaced. The Agency sets a compliance date of 2017, so older noncompliant devices would have to be replaced by 2019, 5 years or more into the lifecycles of these devices.

3.1.4.1 Number of Obsolete AOBRDs in 2019 FMCSA’s estimates for the use of HOS recording devices extend back to 1990, and they estimate that 569,000 AOBRDs will have been deployed by the compliance date of 2017. However, the Agency believes that an ELD or AOBRD will have an average operational life of about 5 years, and by 2019, many of these devices would have been replaced anyway. Nevertheless, the Agency recognizes that some devices will be kept operational.

The costs of replacing AOBRDs are broken into four components: removal of the AOBRD, disposal of the AOBRD, initial purchase of the ELD, and initial installation of the ELD. Service fees that were accounted for with new ELD purchases are not included here because they represent existing costs of operating an AOBRD.

Failure times of electronic devices can be modeled by using the Weibull model, which is comprised of three parameters (shape, scale and location).53 These parameters allow one to control the distribution of device failures by adjusting the location of the density, or when most failures occur (shape); the scale, which can be thought of as the size of the peak when most failures occur; and location, or the point in time when peak failures occur. However, the exponential model may be used to model failure times during the normal (useful) life of an electronic device, when the failure rate is almost constant,54 as is assumed to be the case for AOBRDs, and is a special case of the Weibull model, where the shape parameter is equal to one and the location parameter is equal to zero. Setting these parameters this way is what creates a model with a constant failure rate across time.

Using the exponential model, along with a mean time to failure of 5 years, the probability of an AOBRD purchased in 2015 being operational in 2019 is 45 percent. The Agency’s technology adoption model forecasts that 33,450 AOBRDs will have been purchased in 2015, meaning over 15,000 of those new units will still be operational in 2019. For 2014, 34,000 AOBRDs are estimated to be sold, with about 45 percent operational by 2019. The Agency continued this analysis back several years to estimate that 79,319 older AOBRD units would need to be 53 Sharifi M, Memariani A, Noorossanah R., “Real Time Study of a k-out-of-n System: n Identical Elements with Increasing Failure Rates.” Iranian Journal of Operations Research, vol 1, no. 2, 2009, pp. 56-67. http://www.sid.ir/en/VEWSSID/J_pdf/115720090205.pdf. Accessed December 18, 2013. Klein, J.P., Moeschberger, M.L., “Survival Analysis: Techniques for Censored and Truncated Data.” New York Springer-Verlag, 2009. http://sistemas.fciencias.unam.mx/~ediaz/Cursos/Estadistica3/Libros/0a9X.pdf. Accessed December 18, 2013. 54 Retterath, B., Venkata, S.S., Chowdhury, A.A., “Impact of time-varying failure rates on distribution reliability.” International Journal of Electrical Power and Energy Systems, Vol. 27, Issue 9-10. November - May, 2005. Pages 682-688. http://www.sciencedirect.com/science/article/pii/S0142061505000876 . Accessed January 15, 2014.

29

replaced in 2019. Some current AOBRDs might be upgraded to compliance with this rule via a software patch, eliminating the need for these units to be replaced.

3.1.4.2 Costs of Replacing AOBRDs FMCSA has conservatively based its estimate of the cost of replacing an AOBRD in 2019 on its baseline ELD price estimate of $500. The Agency also accounted for the cost of uninstalling the AOBRD, which it estimates to take 50 percent as long as installation, along with the cost of the new ELD. Based on the baseline installation cost of $84, the cost of labor to replace an AOBRD is $126. The units installed in 2019 are repaired/replaced again in 2024. As discussed in the Environmental Assessment accompanying this RIA, environmental disposal costs are not significant and therefore have not been included.

Applying the exponential model to the predicted adoption of AOBRDs through 2015 allows for an estimate of the number of AOBRDs that will remain operational in 2019 and need replacement at that time. The model estimates that 79,319 AOBRDs in place in 2019 were installed in 2015 or earlier. Because many of the ELDs currently on the market already meet the major technical specifications of the final rule, the Agency assumed 20 percent of the ELDs will need to be replaced as of 2019. As shown in Table 18, on an undiscounted basis this replacement cost totals $9.9 million in 2013 dollars.

Table 18. Replacement Costs of AOBRDs (2013 $ millions)

Purchase Year AOBRDs

Purchased Survival

Rate

AOBRDs remaining in

2019 prior to 2007 80,119 3.0% 2,404

2007 25,520 9.1% 2,315 2008 33,450 11.1% 3,706 2009 25,520 13.5% 3,454 2010 25,520 16.5% 4,218 2011 33,450 20.2% 6,753 2012 51,040 24.7% 12,586 2013 33,450 30.1% 10,075 2014 51,040 36.8% 18,777 2015 33,450 44.9% 15,030 Total 79,319

Per-Unit Replacement and Installation Cost in 2019 $626 Percent Units Needing to be Replaced 20% Total Undiscounted Cost in 2019 (millions) $9.9

3.2 NEW EQUIPMENT FOR ROADSIDE INSPECTORS

FMCSA is setting technical standards that it believes would create flexibility for roadside inspectors to be able to receive RODS information from ELDs electronically and to review these

30

records. The inspectors will need a Bluetooth adapter to ensure access to the data. Typical costs for a Bluetooth adapter and USB adapter are $20 and $75, respectively.55,56 Additionally, each inspection point will need a printer, which the Agency estimates will cost approximately $200. The Agency estimates that equipping all 11,000 roadside inspectors with this equipment, valued at about $295 per inspector, will have an annualized cost of $1.3 million. These costs represent conservatively high estimates, because inspectors may already possess these adapters, and inspectors are only required to have Bluetooth or USB capabilities. The Agency believes that there are no additional costs to support the option of wireless and Web email services.

This hardware will not be used for every inspection of drivers’ RODS, although these devices will likely get heavy use in weigh stations and roadside traffic stops. Consequently, FMCSA believes that replacing these devices will need to occur more frequently than replacing ELD hardware used in the CMVs, and assumes a 3-year operational life for these devices , which comports with IRS guidance on the useful life for calculating depreciation of laptop and desktop computers and related components.57 As shown in Table 19, the annualized cost of this equipment is estimated to be $1.3 million.

Table 19. Cost of Roadside Enforcement Equipment (2013 $ millions)

Purchase Annualized Cost Year 1 $3.25 $0.43 Replacement in Years 4, 7 and 10 $3.25 $0.88 Total $12.98 $1.31

3.3 TRAINING COSTS

The Agency considered the costs of training drivers, motor carriers, and roadside inspectors on ELD technology. There are about 11,000 State and Federal employees who would require 8 hours of training. The cost of travel to training centers will vary by person, but FMCSA assumes an average cost of about $200 per person. Training would be conducted in the first year for the 11,000 existing inspectors, and for the new inspectors for years 2-10. FMCSA estimates an inspector turnover rate of 16%58. Table 20 below summarizes the training costs for inspectors.

By the effective date of this rule, most drivers will already have interacted with ELDs, FMSs, or user interfaces similar to those utilized by ELDs; and FMCSA assumes that each driver will require one-half hour of training. Table 21 below summarizes the training costs for CMV drivers.

55 See https://m.cdwg.com/shop/products/StarTech.com-Mini-USB-Bluetooth-2.1-Adapter-Class-1-EDR-Wireless-Network/2633660. Accessed February 4, 2014. 56 See http://www.cdw.com/shop/products/Apricorn-Aegis-Secure-Key-USB-flash-drive-4-GB/2620738.aspx. Accessed February 4, 2014. 57 Department of Treasury. Internal Revenue Service (IRS). Internal Revenue Manual. IT Equipment. http://www.irs.gov/irm/part1/irm_01-035-006.html. Accessed February 4, 2014. 58 The turnover rate is based on the 2010-2014 average for Government separations. Source: Bureau of Labor Statistics. Job Openings and Labor Turnover Survey. http://data.bls.gov/pdq/querytool.jsp?survey=jt. Accessed March 18, 2015.

31

Table 20. Enforcement Training Costs

Travel Cost per Person

Hourly Labor Cost

Hours per

Person

Training Cost per Person Employees

Training Cost (2013 $ millions)

Annualized Training Cost

(2013 $ millions)

Current Inspectors (Year 1) $200 $40 8 $528 11,000 $5.8 $0.8

New Inspectors (16% per Year, Years 2-10) $200 $40 8 $528 1,760 × 9

years $8.4 $0.8

Inspectors: Total $200 $40 8 $528 26,840 $14.2 $1.6

3.3.1 Driver and Carrier Office Staff Training This final rule does not mandate specific training requirements for drivers in connection with ELDs. However, implicit in the requirement that drivers use ELDs is the notion that the driver be familiar with their use, so driver training costs have been estimated.

EOBR 1 assumed that there would be a need for 30 minutes of training for each trucker with an office employee for the first year when an ELD is used. No costs were recognized for the following years because they would be replacing paper log training with this ELD training, cancelling the extra costs. In EOBR 1, the companies were to be required to use ELDs for 2 years as a response to their HOS violations. New carriers would be subjected to the remedial directive each year, so there would be some training costs in ELDs each year.

In this analysis FMCSA assumes the same 30 minutes of training time and uses the hourly labor cost of $31 for the driver, and a cost of $8 per driver for trainers (estimated by applying the 27 percent for overhead discussed in Section 2.4.2 above, which captures general and administrative fixed costs, such as training). If all the drivers active in 2017 would be trained that year, then during the following years the training expense will only be for new drivers that have had no exposure to ELDs in any previous setting—drivers moving from one carrier to another will be able to use the new ELDs, even if they are slightly different. The average growth between 2010 and 2012 in heavy and tractor-trailer truck driver employment was 3.0 percent. FMCSA considers these people to be the new drivers and uses this figure to estimate the number of new RODS drivers that will need to be trained each year.

Option 1 assumes an initial population of 3.30 million drivers in 2017 when this rule will become effective. The initial cost of training those drivers would be $59.48 million, and $20.9 million in training expense for new drivers over the following 9 years, for a total of $80.37 million over all years. Using a 7 percent discount rate, the annualized cost of 10 years of driver ELD training is $9.4 million per year. Under Option 2, the initial population required to use ELDs in 2017 will be 2.66 million drivers. The initial training costs will be $47.83 million in the first year of the rule, and $16.97 million over the following 9 years, for a total of $64.80 million over all years. The annualized cost for Option 2 is $8.0 million.

The cost of training carrier office staff is estimated to be minimal. ELD service providers provide Web-based access to electronic driver RODS which, as discussed in Section 4.1 below, reduces many administrative tasks. FMCSA assumes that carrier supervisors and office staff would require little to no formal training to learn how to interact with a new Web site. Those carriers

32

that already use FMS will already be familiar with the Web sites of their service providers. Staff of other carriers will likely learn quickly by doing, and become more proficient accessing electronic RODS.

Table 21. CMV Driver Training Costs

Hourly Labor Cost

Hours per

Person

Training Cost per Person Drivers

Training Cost (2013 $ millions)

Annualized Training Cost

(2013 $ millions)

Option 2: Current Drivers (Year 1) $39 0.5 $19.5 2,657,299 $47.83 $6.1

Option 2: New Drivers (Years 2-10) $39 0.5 $19.5 104,763× 9

years $16.97 $1.4

Option 2: Total $39 0.5 $19.5 3,600,167 $64.80 $8.0 Option 1: Current Drivers (Year 1) $39 0.5 $19.5 3,304,167 $59.48 $7.6

Option 1: New Drivers (Years 2-10) $39 0.5 $19.5 129,007× 9

years $20.90 $1.8

Option 1: Total $39 0.5 $19.5 4,465,228 $80.37 $9.4

3.4 HOS COMPLIANCE COSTS

The Agency's estimate of ELD effectiveness (discussed in Section 2.5.3) indicates that ELDs significantly reduce the HOS non-compliance problem, but do not lead to full compliance. Drivers operating CMVs in which carriers have voluntarily installed ELDs will still have HOS rule violations, although, on average, much fewer than those who use paper RODS. FMCSA calculated the current cost of closing the HOS compliance gap and divided this into costs attributable to voluntary ELD and AOBRD users (which this rule would not affect) and costs attributable to CMVs without voluntarily installed ELDs and AOBRDs. These results are used to calculate the average cost of increased HOS compliance, shown in Table 22, attributable to each ELD that would be installed as a result of this rule.

Table 22. Costs of Improved HOS Compliance per ELD (2013 $)

LH SH Annual Cost to Achieve Full HOS Compliance to ELD Non-Users $635 $429 Annual Cost for Increased HOS Compliance from ELD Use (45% Effectiveness) $286 $193

3.4.1 Estimate of Costs of Full HOS Compliance For this final rule, FMCSA estimated the costs of eliminating current noncompliance with the HOS based on estimates of compliance costs of changes to the HOS rules from 2003 to the present. Appendix D contains a more detailed discussion of how these compliance costs were originally estimated and how the Agency has adjusted them.

Although the HOS rules have undergone many changes in past years, the first major revision occurred with the 2003 HOS final rule. For that rule, FMCSA examined survey data on driver

33

schedules to estimate how well the industry was already complying with the changes to drive time and on-duty time that would go into effect in 2004. Based on the number of drivers that were currently operating in excess of the future rule limits, FMCSA estimated the number of new drivers and CMVs that would be needed to redistribute the workload such that no driver would exceed the new limits. Costs considered included the labor supply elasticity and whether the need for additional drivers would lead to an increase in wages; costs associated with recruiting and hiring drivers; and purchase, maintenance, and storage costs of new CMVs. The Agency found that the last two categories were significant, and estimated that the annual cost of full compliance would be $881 million for all LH operations and $400 million for all SH operations (both in year 2000 dollars). This compliance cost reflects the incremental additional expenditures that would have been needed in the industry to bring it into full 100% compliance with the HOS rule from an already high preexisting level of compliance. The $881 million cost for all LH operations and $400 million cost for all SH operations (both in year 2000 dollars) that are presented in the top row of Table 23 measure the incremental additional cost to bring the then-current non-compliant drivers into compliance. Roadside inspections of the driver population in 2004 established a 93.5 percent59 compliance rate, based on the percentage of drivers placed OOS (6.5 percent) for an HOS violation. These inspections were performed during the first full year that rule was in effect. The drivers placed OOS for an HOS violation are in the population of non-compliant drivers that would be impacted by the estimated cost to come into full compliance. Soon thereafter, the Agency made small modifications to the rule that became effective in 2005 and resulted in an increase in annual LH compliance costs of $13 million (year 2004 dollars). To assess the cost of eliminating all current noncompliance with the HOS rules, the Agency made three adjustments to these compliance cost estimates:

• Adjusted these figures to year 2013 dollars using the GDP deflator. • Adjusted these figures for increases in CMV vehicle miles traveled (VMT), which we use

as a proxy for overall industry growth and CMV activity, under the assumption that, ceteris paribus (holding other factors constant), compliance costs are proportional to CMV activity.

• Adjusted these figures for observed improvements in HOS compliance as measured by HOS OOS violation rates from roadside inspections.

These adjustments were necessary to accurately estimate current compliance costs. The cost estimates from 2003 needed to be adjusted for inflation and scaled to reflect industry growth; assuming that new companies fail to comply with HOS rules, the costs that they would incur to come into compliance should be accounted for. The observed HOS OOS violation rates at roadside inspections demonstrated increased levels of compliance over time. Between 2004 and 2012 compliance increased from 93.5 percent to 94.5 percent (that is, the percentage of inspections with HOS OOS violations declined from 6.5 percent to 5.5 percent). The attached compliance costs (expenditures needed to bring companies up to 100 percent compliance)

59 As discussed in Appendix D, compliance costs were estimated from survey data on driver work schedules. This allowed for the degree of non-compliance (for example, the number of hours over daily drive time limits) to be incorporated. Roadside inspections only note the presence of an HOS OOS violation, not how severe it is.

34

therefore are assumed to have decreased by 15 percent during that time, in conjunction with the decline in HOS OOS violation rates (1 – (5.5% ÷ 6.5%) = 15%).60 Table 23 shows the steps and final results of these adjustments. The costs presented here reflect the total costs for all current LH and SH drivers to become compliant with HOS regulations, updated from estimates made for the 2003 HOS rule. VMT data are collected by the Federal Highway Administration and published by FMCSA in Large Truck and Bus Facts.61 The year 2010 publication shows about a 30 percent increase in CMV (truck plus bus) VMT from 2003 or 2005 to 2010. Also, HOS OOS violation rates have dropped about 15 percent since the 2003 HOS rule went into effect in 2004, so holding the level of CMV activity constant, the level of HOS noncompliance is about 85 percent as large as it was in 2004. The current compliance cost baseline reflects that 15 percent compliance improvement. The Agency estimates that, all told, the annual cost to eliminate all current HOS noncompliance is $1,260 million for LH and $565 million for SH.

Table 23. Compliance Costs of HOS Rules (Current Industry Behavior to Full Compliance) (millions)

Description of Cost LH SH Total 2003 HOS Rule Baseline (2000 $) $881 $400 $1,281 2003 Baseline, Inflation Adjusted (2013 $) $1,125 $511 $1,636 2003 Baseline, Adjusted for Industry Growth (2013 $) $1,463 $664 $2,127 2005 HOS Rule Changes ($2013) $13 $0 $13 2005 Changes, Inflation Adjusted (2013 $) $15 $0 $15 2005 Changes, Adjusted for Industry Growth (2013 $) $20 $0 $20 Current Compliance Baseline, No Compliance Improvement (2013 $) $1,483 $664 $2,147 Current Compliance Cost Baseline, 15% Compliance Improvement (2013 $) (Current Compliance Baseline × 85%) $1,260 $565 $1,825

3.4.2 HOS Compliance Costs of ELDs The compliance costs estimate described in the previous section resulted from a move to full HOS compliance from a baseline that included CMVs and carriers that already use ELDs. To show the compliance costs attributable only to the new ELDs that would be placed into service as a result of this rule, we isolate that particular impact: the current users of ELDs have fewer HOS violations than the rest of the population. The first step is to calculate the compliance cost per CMV, which is implicitly a weighted average of CMVs with and without ELDs:

60 Roadside inspectors generally target carriers and drivers most likely to have violations. For this reason, non-compliance rates at roadside inspections may be higher than those for the overall population. Because that bias is present in both periods, the percent change may still be an accurate estimate of the decrease in non-compliance. However, if the severity of HOS violations has also decreased, the compliance improvement may be understated, resulting in over-estimates of current compliance costs. 61 U.S. Department of Transportation. Federal Motor Carrier Safety Administration. Large Truck and Bus Facts 2010. August 2012 https://www.fmcsa.dot.gov/sites/fmcsa.dot.gov/files/docs/LTCC_Report_LargeTruckandBusCrashFacts2010.pdf . Accessed December 18, 2013.

35

EQUATION #1: Compliance Cost per CMV = Compliance Cost per CMV with ELD × Percentage of CMVs with ELD + Compliance Cost per CMV without ELD × Percentage of CMVs without ELD

Since we have estimated ELD effectiveness in reducing violations to be 45%, the compliance cost for a CMV with an ELD can be expressed as 100%-45% (55%) of the compliance cost for a CMV without an ELD and substituted into equation 1:

EQUATION #2: Compliance Cost per CMV = ((100% - 45% ELD Effect) × Compliance Costs per CMV without ELD × Percentage of CMVs with ELD) + (Compliance Costs per CMV without ELD × Percentage of CMVs without ELD)

Explained another way, a CMV equipped with an ELD would bear 55 percent (100 percent minus the 45 percent ELD effect) of the cost, and a CMV without an ELD would completely close the compliance gap. The compliance cost per CMV without an ELD can be solved by rearranging equation 2:

EQUATION #3: Compliance Costs per CMV without ELD = Cost of Full HOS Compliance ÷ Number of CMVs ÷ (55% × Percentage of CMVs with ELD + Percentage of CMVs without ELDs)

For LH, the result of this calculation uses $1,260,000,000 as the cost of full HOS compliance, 2,137,000 LH CMVs, 16 percent of LH CMVs with ELDs, and 84 percent of LH CMVs without ELDs. The resulting calculation is:

EQUATION #4: $1,260,000,000 ÷ 2,137,000 ÷ ((55% × 16%) + 84%) = $635 With an effectiveness of 45 percent, the average cost of improved compliance per new ELD used is $286 (45% × $635) for LH. Table 24 below presents the steps and results of this analysis for both LH and SH.

Table 24. Derivation of Annual Costs of Improved HOS Compliance per ELD (2013 $)

LH SH Total Costs of Full Compliance (millions) $1,260 $565 Total CMVs 2,137,000 1,335,000 Weighted Average Compliance Cost per CMV $590 $423 ELD Users 336,000 37,000 Percent ELD Users (as of 2012) 16% 3% ELD Non-Users 1,801,000 1,298,000 Percent Non-ELD Users 84% 97% Unweighted Average Compliance Cost per CMV $635 $429 Compliance Costs per CMV for Each New ELD $286 $193

36

3.5 VENDOR REGISTRATION COSTS

ELD device vendors will also be required to register with the Agency. Registration is expected to be a relatively short process that would require approximately 4 hours of clerical staff time for each vendor to complete and confirm, assuming that vendors register an average of four ELD devices in the first year. Appendix C indicates that there are as many as 20 U.S. and 2 foreign ELD vendors that would have to complete the registration process. Using the fully-loaded hourly wage rate of $36 for motor carriers from Table 6, initial vendor registration costs are estimated at $3,168 total for all 22 vendors on a preliminary basis. Updating a registration can be expected to be completed in about half the time as completing an initial registration, so total vendor registration costs are estimated to be $1,584 annually on a preliminary basis.62 FMCSA assumes that testing is part of manufacturers’ product developments costs, so there are no testing costs included here.

3.6 COSTS TO FEDERAL GOVERNMENT

FMCSA is currently developing eRODS software in conjunction with its State partners. During an inspection, the eRODS software system will receive, analyze, and display ELD data in a way that can be efficiently used by authorized safety officials. However, because FMCSA included a requirement that ELDs present a RODS grid-graph display on a screen or a printout, enforcement officials can still review the display of information to enforce HOS rules without eRODS. FMCSA estimates that developing eRODS software will cost FMCSA between $250,000 and $500,000. Maintenance of the software and incoming data will be accommodated in the duties of existing employees. Annualized at a 7 percent rate over the 10-year analytical horizon of this rule, the Agency estimates that development and maintenance of eRODS will cost between $80,000 and $120,000 per year. FMCSA does not deem this as a significant cost of the overall rule.

62 This is a preliminary estimate pending completion of an Agency evaluation of the labor burden associated with the registration requirement. This preliminary estimate is not included in the overall estimate of compliance costs. However, the impact on the overall costs is expected to be negligible in any event.

37

4. BENEFITS

4.1 PAPERWORK SAVINGS

The use of ELDs will significantly reduce the paperwork and recordkeeping burden associated with the HOS regulations. Drivers will benefit the most: ELDS will greatly reduce time spent completing RODS and eliminate time that some drivers currently spend forwarding RODS to their employers while away from the motor carriers’ terminals. Comments received in response to the NPRM for the EOBR 1 rule suggest that carriers often do not actually save any money because drivers are not always compensated for completing these tasks. Regardless, the Agency has long recognized in the estimates it prepares for the HOS Information Collection Request supporting statements that the largest portion of burden falls directly on drivers. With ELD use, carriers will accrue clerical time savings and will avoid altogether most of the cost of purchasing paper log books. The Agency has clarified its supporting document requirements, recognizing that ELD records serve as the most robust form of documentation for on-duty driving periods, however the final rule neither increases nor decreases the burden associated with supporting documents, and therefore no paperwork or recordkeeping burden changes related to supporting documents are estimated either for drivers or carriers. Table 25 summarizes the paperwork and recordkeeping savings for each driver using an ELD.

Table 25. Annual Paperwork Savings per Driver (2013 $)

Driver Filling Out RODS

Driver Submitting RODS Clerk Filing RODS

Elimination of Paper Log Books

Total Paperwork Savings

$558 $65 $144 $38.5 $805

4.1.1 Driver Time Savings ELDs do not fully eliminate driver time spent logging HOS. Although changes from on-duty driving to on-duty not-driving are logged automatically by monitoring the CMVs’ motion, a driver will still have to interact with these devices to log in at the beginning of the work shift, to log out at the end of the work shift, and to change duty status. The Agency estimates that each driver fills out an average of 240 RODS per year. ELDs are estimated to reduce the amount of time drivers spend logging their HOS by 4.5 minutes per RODS. ELDs are also assumed to completely eliminate the time drivers spend filing or forwarding their RODS to the carriers, which the Agency estimates takes 5 minutes and occurs 25 times per year. These estimates match those used in the currently approved HOS Information Collection Request supporting statement. Total annual estimated time savings per driver are 18 hours for filling out the RODS (4.5 minutes × 240 RODS ÷ 60 minutes per hour) and 2.08 hours for forwarding RODS (5 minutes × 25 occurrences ÷ 60 minutes per hour). These result in annual labor or time cost savings per driver of $558 (18 hours × $31 per hour) plus $65 (2.08 hours × $31 per hour).63

63 Please see Section 6.4 for further discussion of driver time savings.

38

4.1.2 Clerical Time Savings Because electronic RODS will likely be automatically transmitted and stored on a secure Web site, carrier clerical staff will no longer have to handle these documents. The Agency had estimated that it took carrier clerical staff only a minute to file each RODS, and ELDs will completely eliminate that task, resulting in annual time savings of 4 hours per driver (1 minute × 240 RODS ÷ 60 minutes per hour). The labor cost saving of clerical time per driver is $144 (4 hours × $36 per hour). The burden associated with carriers reviewing a portion of the RODS of their drivers to ensure that they are complete and accurate is assumed to neither decrease nor increase under the final rule.

4.1.3 Paper Cost Savings Vendors, such as JJ Keller, sell bound packets containing a month’s worth of paper RODS for about $3.50.64 ELDs will largely eliminate the need for these materials, resulting in annual cost savings per driver of $38.5 (11 monthly log books × $3.50). Drivers are required to have a back-up paper ROD, in the event the ELD is not functioning. Therefore, FMCSA assumed drivers would still require 1 monthly log book per year.

4.1.4 Total Paperwork Savings The two options discussed in the final rule include two different possible populations of drivers. The total paperwork savings, however, are realized by the former RODS users who are present in both groups and are the same number of drivers in both groups. The $805 per driver benefits a total of approximately 3.07 million drivers,65 for an annualized savings of $2.44 billion over the 10-year period being considered.

4.2 SAFETY BENEFITS

FMCSA applied the violation reduction data from the two voluntary adopters and three motor carriers that entered into settlement agreements requiring ELD use to the crash reduction estimates from the intervention model to construct forecasts of the total number of crashes avoided each year per ELD installed. Although total crashes for this group declined by 2 while the number of CMVs they operated increased by 1,800, the sample is still too small to yield a statistically significant measure of the difference in crashes, measured as the number of crashes per CMV operated.66 One would need a sample of carriers with 7 to 10 times the number of CMVs as there are in FMCSA’s current sample. By contrast, the intervention model has estimated its crash risk reductions using all violations and crashes that occurred for the entire

64 Cost information for a basic RODS from http://www.jjkeller.com/webapp/wcs/stores/servlet/product_Driver's-Daily-Log-Book-with-No-DVIR--Stock_10151_-1_10551_73998. Accessed June 19, 2012. An average of 8 drivers is employed by each motor carrier, requiring 12 log books per year, or 96 total per carrier per year. JJ Keller’s volume pricing for up to 100 log books is $3.50. 65 The annual number of affected drivers under Option 2 over the 2017 through 2026 analysis time period averages to approximately 3.07 million affected drivers but varies somewhat for each individual year due to the pre-2000 model year exemption. 66 A better measure of exposure to crash risk is CMV mileage, but FMCSA does not have these data. The number of CMVs is an acceptable proxy if they all have similar VMT.

39

motor carrier industry. The accuracy of FMCSA’s safety benefit estimates is highly contingent on the assumptions the Agency has made about the applicability of these data.67 The results were adjusted to use the same lower bound of the ELD effectiveness measure that was used for the compliance cost estimates. Table 26 presents the results of the safety benefits analysis.

Table 26. Safety Benefits per ELD (2013 $)

11-hour

Rule 14-hour

Rule

Over 60 Hours/7 Days or 70 Hours/8

Days

Missing, Incomplete, Improper, or

Fraudulent RODS Total Crash Reduction per Million LH CMVs, Industry Weighted Average 132.2 135.9 24.4 381.8 674.3

Crash Cost (millions), Industry Weighted Average $0.277 $0. 277 $0. 277 $0. 277 $0.277

Safety Benefit per LH ELD, Industry Weighted Average $37 $37 $7 $106 $187

Crash Reduction per Million SH CMVs, Industry Weighted Average 90.3 92.9 16.7 260.9 460.7

Crash Cost (millions), Industry Weighted Average $0.277 $0. 277 $0. 277 $0. 277 $0.277

Safety Benefit per SH ELD, Industry Weighted Average $25 $26 $5 $72 $128

4.2.1 Violations and Predicted Crash Reductions Based on an analysis of companies using ELDs, FMCSA estimated the reduction of HOS violations ELD use produced. The number and types of HOS violations eliminated by ELD use was based on actual violation data from five motor carriers that installed advanced AOBRDs. Data on 8,545 roadside inspections conducted on 5,792 CMVs were used in this analysis. FMCSA calculated a 95 percent confidence interval around these violation reduction estimates, and then applied the figure at the lower bound to the rest of the analysis. In doing so, FMCSA took care not to overestimate safety benefits.

As discussed in Section 2.5.4, FMCSA used data from the Roadside Intervention Model to assign a crash risk to each type of HOS violation and estimate the reduction in crashes when the violations are reduced. The crash risks are estimated in the Roadside Intervention Model using data from roadside inspections and traffic enforcements. The crash risk probability uses information from inspections both when a violation was and was not recorded, and after crashes. The Crash Risk Probability, which expresses the likelihood of a crash with a violation, is the ratio of the number of crashes with a recorded violation to the number of days that drivers drove with that violation. The risk probability prior to any intervention then is simply the ratio of the number of times that a crash occurred when drivers were in violation to the total number of times that drivers were in violation. This provides the relationship in the data between violations and crashes. 67 FMCSA assumes that violations detected by inspections can be linked to the risk of crashes--a reduction in violations leads to a lower risk of crashing. Please see Appendix E.

40

The crash risk reduction is calculated using the crash risk probability and the data showing how often drivers are caught with that violation. If drivers are more often seen to commit the violation, then reducing the violation with the requirement of ELD use should have a greater impact reducing the number of crashes. The crash risk reduction uses the crash risk probability and the estimated duration of the effect (how long after drivers are caught do they change their behavior towards safer practices?) to calculate the reduction in crash risk. This method of estimation assumes that ELDs are responsible for the reduction in the violations cited after they have been installed. The Roadside Intervention Model estimates crash risk reductions from traffic enforcement and roadside inspections using data from all traffic stops and roadside inspections. These were estimated using a sample period from January 2005 through September 2009, which contained 9.7 million interventions.

The difference in the number of violations multiplied by the estimated crash reduction per violation cited yields an estimate of the number of crashes avoided through LH ELD use. FMCSA, however, has made some adjustments to scale back these results to reflect more conservative figures. Table 27 and Table 28 show the output from these calculations. The results are presented as crashes avoided per million inspections. This is used as a proxy for crashes avoided per million CMVs based on the assumption that each inspection represents a snapshot of a unique CMV and a unique driver. However, a particular driver or CMV may have been inspected more than once. It is likely that drivers who are cited for a violation will alter their behavior, and those with multiple inspections may, depending on the amount of time between inspections, receive fewer citations after their first inspection. However, the data indicate that multiple inspections of the same driver or CMV exist in about the same proportions in both the pre- and post-periods and did not significantly impact the comparison between periods. It should be noted that the number of crashes in our data from ELD users is small relative to the number of CMVs—1,265.3 per million CMVs in the voluntary group, or 0.001 per CMV, and 1,361.6 per million CMVs in the mandated (by the settlement agreement) group, or 0.001 per CMV—which highlights the difficulty in estimating safety benefits of ELDs from a small sample of ELD installations.

41

Table 27. Crash Risk Reductions from Sample of Voluntary LH ELD Users

Part 395 Sec-tion

Description of Violation

Annual-ized

Crash Re-

ductions

(a)

Pre-ELD Viola-tions (b)

Pre-ELD Crash Risk (c)

[a × b]

Pre-ELD Inspec-

tions (d)

Pre-ELD Crash Risk per Million

CMVs (e)

[(c ÷ d) × 106]

Post-ELD Viola-tions

(f)

Post-ELD Crash Risk

(g) [a × f]

Post-ELD Inspec-tions

(h)

Post-ELD Crash Risk per Million

CMVs (i)

[(g ÷ h) × 106]

Un-adjusted Crash

Reduction per Million CMVs

(j) [e – i]

3(a)(1) 11-Hour 0.02496 18 0.45 1,767 254.3 1 0.02 4,334 5.8 248.5

3(a)(2) 14-Hour 0.02496 23 0.57 1,767 324.9 10 0.25 4,334 57.6 267.3

3(b) Property: 60/70 Hour 0.02496 5 0.12 1,767 70.6 0 0.00 4,334 0.0 70.6

5(b) Passenger: 60/70 Hour 0.02496 2 0.05 1,767 28.3 0 0.00 4,334 0.0 28.3

8 Form and Manner 0.02952 0 0.00 1,767 0.0 0 0.00 4,334 0.0 0.0

8(a)

RODS show status for every 24 hour period

0.02952 3 0.09 1,767 50.1 0 0.00 4,334 0.0 50.1

8(e) Incomplete or false logs

0.05088 17 0.86 1,767 489.5 17 0.86 4,334 199.6 289.9

8(f)(1)

RODS current to last change of duty status

0.02952 0 0.00 1,767 0.0 0 0.00 4,334 0.0 0.0

8(k)(2) RODS retained for last 7 days

0.02952 19 0.56 1,767 317.4 1 0.03 4,334 6.8 310.6

15(b) Duty Status Record Display.

0.02952 0 0.00 1,767 0.0 0 0.00 4,334 0.0 0.0

42

Part 395 Sec-tion

Description of Violation

Annual-ized

Crash Re-

ductions

(a)

Pre-ELD Viola-tions (b)

Pre-ELD Crash Risk (c)

[a × b]

Pre-ELD Inspec-

tions (d)

Pre-ELD Crash Risk per Million

CMVs (e)

[(c ÷ d) × 106]

Post-ELD Viola-tions

(f)

Post-ELD Crash Risk

(g) [a × f]

Post-ELD Inspec-tions

(h)

Post-ELD Crash Risk per Million

CMVs (i)

[(g ÷ h) × 106]

Un-adjusted Crash

Reduction per Million CMVs

(j) [e – i]

15(c) Driver and Carrier Information

0.02952 0 0.00 1,767 0.0 0 0.00 4,334 0.0 0.0

15(g) AOBRD Data Retrieval 0.02952 0 0.00 1,767 0.0 0 0.00 4,334 0.0 0.0

Total 87 2.71 1,767 1,535.1 29 1.17 4,334 269.7 1,265.3

43

Table 28. Crash Risk Reductions from Sample of Mandated LH ELD Users

Part 395 Sec-tion

Description of Violation

Annual-ized

Crash Re-

duction (a)

Pre-ELD Viola-tions

(b)

Pre-ELD Crash Risk (c)

[a × b]

Pre-ELD Inspec-tions

(d)

Pre-ELD Crash Risk per Million

CMVs (e)

[(c ÷ d) × 106]

Post-ELD Viola-tions

(f)

Post-ELD Crash Risk

(g) [a × f]

Post-ELD Inspec-

tions (h)

Post-ELD Crash Risk per Million

CMVs (i)

[(g ÷ h) × 106]

Un-adjusted Crash

Reduction per Million CMVs

(j) [e – i]

3(a)(1) 11-Hour 0.02496 16 0.40 1,134 352.2 1 0.02 1,310 19.1 333.1 3(a)(2) 14-Hour 0.02496 22 0.55 1,134 484.2 1 0.02 1,310 19.1 465.2 3(b) Property:

60/70 Hour 0.02496 3 0.07 1,134 66.0 0 0.00 1,310 0.0 66.0

5(b) Passenger: 60/70 Hour

0.02496 0 0.00 1,134 0.0 0 0.00 1,310 0.0 0.0

8 Form and Manner

0.02952 0 0.00 1,134 0.0 0 0.00 1,310 0.0 0.0

8(a) Form and Manner

0.02952 1 0.03 1,134 26.0 2 0.06 1,310 45.1 -19.0

8(e) Form and Manner

0.05088 12 0.61 1,134 538.4 2 0.10 1,310 77.7 460.7

8(f)(1) Form and Manner

0.02952 0 0.00 1,134 0.0 0 0.00 1,310 0.0 0.0

8(k)(2) Form and Manner.

0.02952 3 0.09 1,134 78.1 1 0.03 1,310 22.5 55.6

15(b) Duty Status Record Display.

0.02952 0 0.00 1,134 0.0 0 0.00 1,310 0.0 0.0

15(c) Driver and Carrier Information.

0.02952 0 0.00 1,134 0.0 0 0.00 1,310 0.0 0.0

15(g) AOBRD Data Retrieval 0.02952 0 0.00 1,134 0.0 0 0.00 1,310 0.0 0.0

Total 57 1.75 1,134 1,545.0 7 0.24 1,310 183.4 1,361.6

44

4.2.2 Adjusted Crash Reductions The Agency is wary of overstating the effectiveness of ELDs in this analysis, and has therefore applied the same adjustments to its crash reduction estimates as it did to its ELD effectiveness measure. Specifically, crash reductions were re-estimated to use the lower bound of a 95 percent confidence interval around the estimated reduction in violations (a calculation that reduced overall ELD effectiveness from 86 or 89 percent to 45 percent). Data on 8,545 roadside inspections conducted on 5,792 CMVs were used in this analysis. These calculations were discussed in Section 4.2.1 above and are explained in full detail in Appendix A. FMCSA calculated the reduction in HOS violation rates for several broad categories of HOS violations at the mean difference in before and after violation rates, and at the lower bound of this difference. A percentage change in the HOS violation rate reduction from the lower bound compared to the mean was then computed and applied to the safety benefits per million CMVs to create a more conservative measure of crash reduction. For example, the 11-hour drive time rule violation rate reduction for voluntary adopters was measured at 98 percent at the mean and 52 percent at the lower bound. Thus, using the lower bound reduces the estimated effectiveness in preventing these types of violations by 47 percent (100% - (52% ÷ 98%)). The number of crashes avoided under this category is then adjusted to reflect this, from 248.5 crashes avoided per million CMVs to 131.1 crashes avoided per million CMVs (248.5 – (47% × 248.5)). Table 29 and Table 30 show the calculations of the adjusted crash reductions for the voluntary and mandated groups.

Table 29. Adjusted Crash Reduction per LH CMV, Voluntary ELD Use

11-hour

Rule 14-hour

Rule

60 Hours/7 Days or 70

Hours/8 Days

Missing, Incomplete, Improper, or Fraudulent

RODS Total Source Pre-ELD Crash Risk 0.45 0.57 0.17 1.51 2.71 Table 26, C Pre-ELD CMVs Inspected 1,767 1,767 1,767 1,767 1,767 Table 26, D Pre-ELD Crash Risk per Million CMVs 254.3 324.9 98.9 857.0 1,535.1 Table 26, E

Post-ELD Crash Risk 0.02 0.25 0.00 0.89 1.17 Table 26, G Post-ELD CMVs Inspected 4,334 4,334 4,334 4,334 4,334 Table 26, H Post-ELD Crash Risk per Million CMVs 5.76 57.59 0.00 206.39 269.74 Table 26, I

Unadjusted Crash Risk Reduction per Million CMVs

248.5 267.3 98.9 650.7 1,265.3 Table 26, J

Mean ELD Effectiveness 98% 82% 100% 81% 86% Appendix A Lower Bound (LB) ELD Effectiveness 52% 40% 26% 49% 45% Appendix A

Percent Reduction in Effectiveness=Crash Risk Reduction Adjustment

53% 49% 26% 60% (LB ÷ Mean)

45

11-hour

Rule 14-hour

Rule

60 Hours/7 Days or 70

Hours/8 Days

Missing, Incomplete, Improper, or Fraudulent

RODS Total Source

Adjusted Crash Risk Reduction per Million CMVs

131.1 130.7 25.8 392.4 679.9

Crash Risk Reduction Adjustment × Unadjusted Crash Risk

Table 30. Adjusted Crash Reduction per LH CMV, Mandatory ELD Use

11-hour

Rule 14-hour

Rule

60 Hours/7 Days or 70

Hours/8 Days

Missing, Incomplete, Improper, or Fraudulent

RODS Total Source Pre-ELD Crash Risk 0.40 0.55 0.07 0.73 1.75 Table 27, C Pre-ELD CMVs Inspected 1,134 1,134 1,134 1,134 1,134 Table 27, D Pre-ELD Crash Risk per Million CMVs 352.2 484.2 66.0 642.5 1,545.0 Table 27, E

Post-ELD Crash Risk 0.02 0.02 0.00 0.19 0.24 Table 27, G Post-ELD CMVs Inspected 1,310 1,310 1,310 1,310 1,310 Table 27, H Post-ELD Crash Risk per Million CMVs 19.05 19.05 0.00 145.28 183.39 Table 27, I

Unadjusted Crash Risk Reduction per Million CMVs

333.1 465.2 66.0 497.3 1,361.6 Table 27, J

Mean ELD Effectiveness 95% 96% 100% 73% 89% Appendix A Lower Bound ELD Effectiveness 45% 54% -13% 19% 35% Appendix A

Percent Reduction in Effectiveness/Crash Risk Reduction Adjustment

47% 56% -13% 26% (LB ÷ Mean)

Adjusted Crash Risk Reduction per Million CMVs

157.8 261.4 -8.6 128.5 539.0

Crash Risk Reduction Adjustment × Unadjusted Crash Risk

4.2.3 SH Crash Reductions As discussed above, the number and severity of crashes related to HOS violations is lower for SH operations. Although FMCSA inspection data does record information on specific vehicles and drivers, it does not record, nor can roadside inspectors determine, whether the driver and CMV were engaged in LH or SH operations at the time of inspection, making it impossible to calculate HOS violation rates specific to SH operations from FMCSA’s complete inspection data set. Moreover, the samples of carriers that installed ELDs pertain to LH use. For these reasons,

46

the Agency could not construct direct measures of SH crash reductions from ELD use. However, one possibility would be to assume that the ratio of SH compliance costs ($193 per SH ELD) to LH compliance costs ($286 per ELD) per ELD installed is indicative of the relative safety benefits. Under this assumption, SH safety benefits are estimated to be about 67 percent as large ($193 ÷ $286) as LH benefits per ELD. The resulting crash risk reduction estimates for SH CMVs are presented in Table 31.

Table 31. Crash Reduction per SH CMV

11-hour

Rule 14-hour

Rule

Over 60 Hours/7

Days or 70 Hours/8

Days

Missing, Incomplete, Improper, or Fraudulent

RODS Total

Adjusted Crash Risk Reduction per Million LH CMVs, Voluntary ELD Use 131.1 130.7 25.8 392.4 679.9

Adjusted Crash Risk Reduction per Million LH CMVs, Mandatory ELD Use 157.8 261.4 -8.6 128.5 539.0

Relative Compliance Cost per SH ELD Installed 67% 67% 67% 67% 67%

Crash Risk Reduction per Million SH CMVs, Voluntary ELD Use 88.5 88.2 17.4 264.8 458.9

Crash Risk Reduction per Million SH CMVs, Mandatory ELD Use 106.5 176.4 -5.8 86.7 363.8

4.2.4 Average Crash Reductions over Both Samples The data presented in the preceding sections are from two samples, one comprising two carriers that voluntarily installed ELDs, and a second comprising three carriers required to install ELDs after having been identified with excessive HOS rule violations. Although the number of carriers for which data were available is small, the total number of inspections of these carriers was large enough to allow for statistically significant estimates of changes in HOS violations. The two carrier groups may be representative of the bounds of ELD users, safety conscious voluntary ELD users versus carriers with high HOS violation rate. ELDs were slightly less effective in reducing violations and crash risk for the second group. Much of this can be attributed to the differences in the reduction in the 49 CFR 395.8 (fraudulent or improper RODS) violations. Although these types of violations were drastically reduced for both samples, the reduction is 7 percentage points lower at the means and 11 percentage points at the lower bounds for the mandated group. This result would seem to challenge past assumptions the Agency has made (in EOBR 1) that an ELD mandate as a remedial option for carriers with relatively high HOS violation rates would result in relatively larger safety benefits per CMV than those for average carriers. The Agency is continually gathering data on ELD use, but the results so far seem to indicate that, because ELDs can only directly record drive time, drivers could continue to deliberately make other improper entries (for example, logging sleeper-berth when they were actually on duty). Based on these results, FMCSA assumes that average carriers will more successfully discipline their drivers (or discourage fraudulent entries) and have more success in using ELDs.

47

For EOBR 1, FMCSA estimated the broadest definition of carriers that would fall under a remedial mandate would affect 70,000 LH CMVs and drivers. This represents about 4 percent of the LH CMVs and drivers that did not have ELDs at the time. FMCSA believes the carriers that have installed ELDs under a remedial directive are representative of this 4 percent of the driver population. FMCSA weighted the crash reduction results of the mandated sample by this 4 percent, and the results from the voluntary sample by 96 percent, to construct a weighted average that covers all types of carriers. The results of this analysis and the final results for safety benefit per LH or SH ELD installation are shown in Table 32 below.

Table 32. Average Safety Benefits per ELD (2013 $)

11-hour

Rule 14-hour

Rule

Over 60 Hours/7

Days or 70 Hours/8

Days

Missing, Incomplete, Improper, or Fraudulent

RODS Total Crash Reduction per Million LH CMVs, Voluntary Use 131.1 130.7 25.8 392.4 679.9

Crash Cost (millions), Voluntary Use $0.277 $0.277 $0.277 $0.277 $0.277 Safety Benefit per LH ELD, Voluntary Use $36 $36 $7 $109 $188

Crash Reduction per Million SH CMVs, Voluntary Use 88.5 88.2 17.4 264.8 458.9

Crash Cost (millions), Voluntary Use $0.277 $0.277 $0.277 $0.277 $0.277 Safety Benefit per SH ELD, Voluntary Use $25 $24 $5 $73 $127

Crash Reduction per Million LH CMVs, Mandatory Use 157.8 261.4 -8.6 128.5 539.0

Crash Cost (millions), Mandatory Use $0.277 $0.277 $0.277 $0.277 $0.277 Safety Benefit per LH ELD, Mandatory Use $44 $72 -$2 $36 $149

Crash Reduction per Million SH CMVs, Mandatory Use 106.5 176.4 -5.8 86.7 363.8

Crash Cost (millions), Mandatory Use $0.277 $0.277 $0.277 $0.277 $0.277 Safety Benefit per SH ELD, Mandatory Use $29 $49 -$2 $24 $101

Crash Reduction per Million LH CMVs, Industry Weighted Average 132.2 135.9 24.4 381.8 674.3

Crash Cost (millions), Industry Weighted Average $0.277 $0.277 $0.277 $0.277 $0.277

Safety Benefit per LH ELD, Industry Weighted Average $37 $38 $7 $106 $187

Crash Reduction per Million SH CMVs, Industry Weighted Average 89.2 91.7 16.5 257.7 455.1

Crash Cost (millions), Industry Weighted Average $0.277 $0.277 $0.277 $0.277 $0.277

Safety Benefit per SH ELD, Industry Weighted Average $25 $25 $5 $71 $126

48

As shown, FMCSA calculated an average safety benefit of $187 per LH ELD and $126 per SH ELD. The Agency also recalculated benefits and net benefits using the lower per-ELD safety benefit figure from the sample of mandated installations ($149 for LH and $101 for SH). These results are presented in Section 6, the Sensitivity Analyses.

49

5. RESULTS OF ANALYSIS Table 33 summarizes the results of the analysis. Option 2, which represents the final rule, has the highest net benefits. Extending the ELD mandate to drivers not required to maintain RODS, reduces the net benefits of Option 1. All of these drivers are required to incur the cost of purchasing and maintaining ELDs, but do not receive any paperwork reduction benefits.

Table 33. Summary of Annualized Costs and Benefits (2013 $ millions)

Option 1: All HOS Drivers Option 2: (Adopted) RODS

Drivers Only 3% 7% 3% 7%

Total Benefits $3,150 $3,124 $3,035 $3,010

Safety (Crash Reductions) $694 $687 $579 $572 Paperwork Savings $2,456 $2,438 $2,456 $2,438

Total Costs $2,298 $2,280 $1,851 $1,836

New ELD Costs $1,348 $1,336 $1,042 $1,032

AOBRD Replacement Costs $2 $2 $2 $2 HOS Compliance Costs $936 $929 $797 $790 CMV Driver Training Costs $9 $10 $7 $8

Enforcement Training Costs $1 $2 $1 $2

Enforcement Equipment Costs $1 $1 $1 $1

Net Benefits $852 $844 $1,184 $1,174

50

5.1 CALCULATIONS FOR OPTION 2

The RIA presents the calculations for the rulemaking Option 2 in Table 34. This option is presented first because it is the adopted option and provides the highest net benefits.

Table 34. Option 2 Calculations Costs and Benefits (7% Discount Rate, 2013 $ Million)

Year

Costs Benefits

Net Benefits ELDs

AOBRD Replacemen

t HOS

Compliance Driver

Training Enforcement

Training Enforcement Equipment Safety

Paperwork Reduction

2017 $875.2 $0.0 $692.6 $47.8 $5.8 $3.2 $479.3 $2,139.3 $993.9

2018 $873.4 $0.0 $681.8 $2.2 $0.9 $0.0 $476.9 $2,104.8 $1,023.4

2019 $859.7 $8.7 $664.3 $1.8 $0.8 $0.0 $469.7 $2,049.6 $984.0

2020 $834.5 $0.0 $640.2 $1.6 $0.8 $2.6 $456.8 $1,974.5 $951.7

2021 $803.0 $0.0 $612.7 $1.4 $0.7 $0.0 $443.6 $1,889.2 $915.0

2022 $768.7 $0.0 $584.0 $1.3 $0.7 $0.0 $427.2 $1,800.3 $872.9

2023 $734.6 $6.6 $555.9 $1.2 $0.6 $2.2 $410.3 $1,713.2 $822.5

2024 $701.6 $0.0 $528.9 $1.1 $0.6 $0.0 $394.4 $1,629.9 $792.0

2025 $669.1 $0.0 $502.7 $1.0 $0.5 $0.0 $378.7 $1,548.7 $754.0

2026 $637.1 $0.0 $477.1 $1.0 $0.5 $1.8 $363.1 $1,469.8 $715.3 Total of Discounted Values

$7,756.9 $15.3 $5,940.2 $60.3 $11.9 $9.8 $4,299.9 $18,319.2 $8,824.8

Annualized Values 1032.2 2.0 790.4 8.0 1.6 1.3 572.2 2437.6 $1,174.3

51

5.2 CALCULATIONS FOR OPTION 1

The RIA presents the calculations for the rulemaking Option 1 in Table 35.

Table 35. Option 1 Calculations Costs and Benefits (7% Discount Rate, 2013 $ Million)

Year

Costs Benefits

Net Benefits ELDs AOBRD

Replacement HOS

Compliance Driver

Training Enforcemen

t Training Enforcement Equipment Safety

Paperwork Reduction

2017 $1,149.2 $0.0 $817.4 $59.5 $5.8 $3.2 $577.6 $2,139.3 $681.8

2018 $1,140.2 $0.0 $803.4 $2.7 $0.9 $0.0 $573.9 $2,104.8 $731.6

2019 $1,117.4 $8.7 $781.7 $2.2 $0.8 $0.0 $564.6 $2,049.6 $703.3

2020 $1,081.4 $0.0 $752.7 $1.9 $0.8 $2.6 $548.3 $1,974.5 $683.5

2021 $1,038.2 $0.0 $719.9 $1.7 $0.7 $0.0 $531.8 $1,889.2 $660.6

2022 $992.0 $0.0 $685.7 $1.5 $0.7 $0.0 $512.1 $1,800.3 $632.4

2023 $946.4 $6.6 $652.4 $1.5 $0.6 $2.2 $491.3 $1,713.2 $594.9

2024 $902.5 $0.0 $620.4 $1.4 $0.6 $0.0 $472.2 $1,629.9 $577.1

2025 $859.5 $0.0 $589.4 $1.3 $0.5 $0.0 $453.3 $1,548.7 $551.3

2026 $817.3 $0.0 $559.2 $1.2 $0.5 $1.8 $434.5 $1,469.8 $524.3 Total of Discounted Value

$10,044.0 $15.3 $6,982.3 $74.8 $11.9 $9.8 $5,159.6 $18,319.2 $6,340.8

Annualized Values 1336.5 2.0 929.1 10.0 1.6 1.3 686.6 2437.6 $843.7

52

6. SENSITIVITY ANALYSES

6.1 ALTERNATE VALUES OF STATISTICAL LIFE

June 2014 USDOT OST guidance recommends that Agencies evaluate alternative safety benefits using low and high alternative estimates of the VSL of $5.2 million and $13.0 million, respectively, in addition to the primary recommended VSL figure of $9.2 million.68 FMCSA has evaluated all safety benefits by multiplying the estimated number of crashes avoided by the average cost of a crash. Although the medical, emergency services, property damage, and delay cost components were kept constant, the cost of injuries and fatalities were scaled according to the VSL. In the main body of the analysis, the Agency used a VSL of $9.2 million to calculate safety benefits (and an associated crash cost of $277,000) and uses a lower VSL of $5.2 million and a higher VSL of $13.0 million to calculate alternatives. The cost of an average crash associated with these alternative VSL are $167,000 and $337,000, respectively.

Table 36. Summary of Options 1 and 2 with High and Low VSL (2013 $ millions)

Option1: All HOS

Option1: Low VSL

Option 1: High VSL

Option 2: RODS Only

Option 2: Low VSL

Option 2: High VSL

Discount Rate 7% 7% 7% 7% 7% 7% Total Benefits $3,124 $2,826 $3,400 $3,010 $2,761 $3,240

Safety (Crash Reductions) $687 $388 $963 $572 $323 $802 Paperwork Savings $2,438 $2,438 $2,438 $2,438 $2,438 $2,438

Total Costs $2,280 $2,280 $2,280 $1,836 $1,836 $1,836

ELD Costs $1,336 $1,336 $1,336 $1,032 $1,032 $1,032

AOBRD Replacement Costs $2 $2 $2 $2 $2 $2

HOS Compliance Costs $929 $929 $929 $790 $790 $790

Driver Training Costs $10 $10 $10 $8 $8 $8

Enforcement Training Costs $2 $2 $2 $2 $2 $2

Enforcement Equipment Costs $1 $1 $1 $1 $1 $1

Net Benefits $844 $545 $1,120 $1,174 $925 $1,404

6.2 LOWER COST ELD

The Agency recomputed costs and net benefits using the price of the lowest cost (in annualized terms) ELD on which it was able to gather information, $166 for the VDO RoadLog69 device. As discussed above, FMCSA did not use this device due to its manufacturers’ low market share in North America and because fleet owners have typically demanded ELDs bundled with other

68 U.S. Department of Transportation. Office of the Secretary of Transportation. Guidance on Treatment of the Economic Value of a Statistical Life – 2014 Adjustment. June 13, 2014. Pg. 1 and pp. 10-11. https://www.transportation.gov/sites/dot.gov/files/docs/VSL_Guidance_2014.pdf. Accessed July 18, 2014. 69 A description of the VDO RoadLog is available from: http://www.vdoroadlog.com/. Accessed February 4, 2014.

53

FMS services to gain productivity benefits (not evaluated in this RIA) from these other services. However, this device is the only minimally compliant ELD of which the Agency is aware.

Table 37. Summary of All Options with Lowest Cost ELD (2013 $ millions)

Option 1: All

HOS

Option 1: All HOS with Lower

ELD Cost Option 2: RODS

Only

Option 2: RODS Only With

Lower ELD Cost

Discount Rate 7% 7% 7% 7%

Total Benefits $3,124 $3,124 $3,010 $3,010 Safety (Crash Reductions) $687 $687 $572 $572 Paperwork Savings $2,438 $2,438 $2,438 $2,438

Total Costs $2,280 $1,536 $1,836 $1,271

ELD Costs $1,336 $592 $1,032 $467

AOBRD Replacement Costs $2 $2 $2 $2

HOS Compliance Costs $929 $929 $790 $790

Driver Training Costs $10 $10 $8 $8

Enforcement Training Costs $2 $2 $2 $2

Enforcement Equipment Costs $1 $1 $1 $1 Net Benefits $844 $1,588 $1,174 $1,739

6.3 LOWER CRASH REDUCTIONS

FMCSA has analyzed HOS violation data from only five carriers to assess the effectiveness of ELDs. Although the number of violations produces a large sample capable of producing statistically significant results, the samples of voluntary and mandated ELD users are considered separately and result in different estimates of violation and crash reductions. FMCSA recalculated benefits and net benefits using the lowest estimate of crash reduction (based on data only from mandated ELD users): 539.0 crashes per million LH CMVs and 368.3 crashes per million SH CMVs. As shown in Table 38, Options 1 and 2 still have positive net benefits with lower crash reductions.

54

Table 38. Summary of All Options with Lowest Crash Reduction (2013 $ millions)

Option 1: All

HOS Option 2: RODS

Only

Option 1:with Lower

Crash Reduction

Option 2: with Lower

Crash Reduction

Discount Rate 7% 7% 7% 7%

Total Benefits $3,124 $3,010 $2,986 $2,895

Safety (Crash Reductions) $687 $572 $549 $457

Paperwork Savings $2,438 $2,438 $2,438 $2,438

Total Costs $2,280 $1,836 $2,280 $1,836

ELD Costs $1,336 $1,032 $1,336 $1,032

AOBRD Replacement Costs $2 $2 $2 $2

HOS Compliance Costs $929 $790 $929 $790

Driver Training Costs $10 $8 $10 $8

Enforcement Training Costs $2 $2 $2 $2

Enforcement Equipment Costs $1 $1 $1 $1

Net Benefits $844 $1,174 $706 $1,059

6.4 DIFFERENCES IN DRIVER TIME SAVINGS

The Agency has estimated large paperwork savings from ELDs, approximately $2,438 million per year (annualized using a 7% discount rate). Because FMCSA’s data currently indicate that the cost of improved HOS compliance will exceed its safety benefits if applied to all carriers, these savings increase the benefits of the rule such that certain policy options, such as Option 2, are estimated to have positive net benefits. The estimate of paperwork savings, however, is extremely sensitive to the estimated amount of time drivers will save each day using ELDs rather than completing paper RODS. FMCSA assumes that the time spent recording duty status will drop from 6.5 minutes per day (the estimate used in the Agency’s Information Collection Request Supporting Statement) to 2 minutes per day. Although it seems extremely unlikely that no time will be saved, or that it will increase, FMCSA considered different time savings and evaluated them for its adopted option, Option 2. The results of this analysis are presented in Table 39.

55

Table 39. Summary of Option 2 with Different Driver Time Savings (2013 $ millions)

Minutes Saved per RODS 4.5 min 3.5 min 2.5 min 0 min

Total Benefits $3,010 $2,645 $2,269 $1,331 Safety (Crash Reductions) $572 $572 $572 $572 Paperwork Savings $2,438 $2,073 $1,697 $759

Total Costs $1,836 $1,836 $1,836 $1,836

New ELD Costs $1,032 $1,032 $1,032 $1,032 AOBRD Replacement Costs $2 $2 $2 $2

HOS Compliance Costs $790 $790 $790 $790

Driver Training Costs $8 $8 $8 $8

Enforcement Training Costs $2 $2 $2 $2

Enforcement Equipment Costs $1 $1 $1 $1

Net Benefits, Option 2 $1,174 $809 $434 -$505

56

7. REGULATORY FLEXIBILITY ANALYSIS

7.1 INTRODUCTION

The Regulatory Flexibility Act of 1980, Pub. L. 96-354, 94 Stat. 1164 (5 U.S.C. 601-612), as amended by the Small Business Regulatory Enforcement Fairness Act of 1996 (Pub. L. 104–121, 110 Stat. 857, March 29, 1996) and the Small Business Jobs Act of 2010 (Pub. L. 111-240, September 27, 2010), requires Federal agencies to consider the effects of the regulatory action on small business and other small entities and to minimize any significant economic impact. The term “small entities” comprises small businesses and not-for-profit organizations that are independently owned and operated and are not dominant in their fields, and governmental jurisdictions with populations of less than 50,000. Accordingly, DOT policy requires an analysis of the impact of all regulations on small entities, and mandates that agencies strive to lessen any adverse effects on these businesses.

A Regulatory Flexibility Analysis must contain the following:

1) A statement of the need for, and objectives of, the rule. 2) A statement of the significant issues raised by the public comments in response to the

IRFA, a statement of the assessment of the agency of such issues, and a statement of any changes made in the proposed rule as a result of such comments.

3) The response of the agency to any comments filed by the Chief Counsel for Advocacy of the Small Business Administration in response to the proposed rule, and a detailed statement of any change made to the proposed rule in the final rule as a result of the comments.

4) A description of and an estimate of the number of small entities to which the rule will apply or an explanation of why no such estimate is available.

5) A description of the projected reporting, recordkeeping, and other compliance requirements of the rule, including an estimate of the classes of small entities which will be subject to the requirement and the type of professional skills necessary for preparation of the report or record.

6) A description of the steps the agency has taken to minimize the significant economic impact on small entities consistent with the stated objectives of applicable statutes, including a statement of the factual, policy, and legal reasons for selecting the alternative adopted in the final rule and why each of the other significant alternatives to the rule considered by the agency which affect the impact on small entities was rejected.

7) For a covered agency, as defined in section 609(d)(2), a description of the steps the agency has taken to minimize any additional cost of credit for small entities.

7.2 NEED FOR AND OBJECTIVES OF THIS RULE

The Agency is issuing this final rule to mandate the use of ELDs by the majority of CMV operations. The objective is to reduce the number of crashes caused by driver fatigue that could have been avoided had the driver complied with the HOS rules. The legal basis for this rule is described in the preamble to the final rule.

57

The Agency is required by statute (MAP-21) to adopt regulations requiring that CMVs operated in interstate commerce, operated by drivers required to keep RODS, be equipped with ELDs. FMCSA amends part 395 of the FMCSRs to require the installation and use of ELDs for CMV operations for which RODS are required. CMV drivers are currently required to record their HOS (driving time, on- and off-duty time) in paper RODS, although some carriers have voluntarily adopted an earlier standard for HOS recording using devices known as AOBRDs. The HOS regulations are designed to ensure that driving time, one of the principal “responsibilities imposed on the operators of commercial motor vehicles,” does “not impair their ability to operate the vehicles safely” (49 U.S.C. 31136(a)(2)). Driver compliance with the HOS rules helps ensure that “the physical condition of commercial motor vehicle drivers is adequate to enable them to operate the vehicles safely” (49 U.S.C. 31136(a)(3)). FMCSA believes that properly designed, used, and maintained ELDs will enable motor carriers to track their drivers’ on-duty driving hours accurately, thus preventing regulatory violations or excessive driver fatigue.

Improved HOS compliance will prevent commercial vehicle operators from driving for long periods without opportunities to obtain adequate rest. Sufficient rest is necessary to ensure that a driver is alert behind the wheel and able to respond appropriately to changes in the driving environment.

Substantial paperwork and recordkeeping burdens are also associated with HOS rules, including time spent by drivers filling out and submitting paper RODS and time spent by motor carrier staff reviewing, filing, and maintaining these RODS. ELDs will eliminate clerical tasks associated with the RODS and significantly reduce the time drivers spend recording their HOS. These paperwork reductions offset most of the costs of the devices.

7.3 PUBLIC COMMENT ON THE IRFA, FMCSA ASSESSMENT AND RESPONSE

Although public comment on the SNPRM for this rule was extensive, there were no comments specific to the Initial Regulatory Flexibility Analysis.

7.4 FMCSA RESPONSE TO COMMENTS BY THE CHIEF COUNSEL FOR ADVOCACY OF THE SMALL BUSINESS ADMINISTRATION ON THE IRFA

The FMCSA did not receive comments from the Chief Counsel for Advocacy of the Small Business Administration on the IRFA included with the SNPRM for this rule.

7.5 DESCRIPTION AND NUMERICAL ESTIMATE OF SMALL ENTITIES AFFECTED BY THE RULEMAKING

The motor carriers regulated by FMCSA operate in many different industries, and no single Small Business Administration (SBA) size threshold is applicable to all motor carriers. Most for-

58

hire property carriers operate under North American Industrial Classification System70 (NAICS) code 484, truck transportation (see: http://www.bls.gov/iag/tgs/iag484.htm), although some for-hire carriers categorize themselves as “express delivery services” (NAICS 492110) or “local delivery” (NAICS 492210) or operate primarily in other modes of freight transportation. As shown in Table 40 below, the SBA size standard for truck transportation and local delivery services is currently $27.5 million in revenue per year and 1,500 employees for express delivery services. For other firms in other modes that may also be registered as for-hire motor carriers, the size standard is 500 or 1,500 employees. As Table 40 also shows, for-hire passenger operations that FMCSA regulates have a size standard of $15 million in annual revenue.

Table 40. SBA Size Standards for Selected Industries (2014 $)

NAICS Codes NAICS Industry Description

Annual Revenue (millions) Employees

481112 and 481212 Freight Air Transportation 1,500

482111 Line-Haul Railroads 1,500

483111 through 483113 Freight Water Transportation 500

484110 through 484230 Freight Trucking $27.5

492110 Couriers and Express Delivery 1,500

492210 Local Messengers and Local Delivery $27.5

485210 through 485510 Bus Transportation $15.0

445110 Supermarkets and Grocery Stores $32.5

452111 Department Stores (except Discount Department Stores)

$32.5

452112 Discount Department Stores $29.5

452910 Warehouse Clubs and Superstores $29.5

452990 Other General Merchandise Stores $32.5

453210 Office Supplies and Stationery Stores $32.5

236115 through 236220 Building Construction $36.5 237110 Water and Sewer Line and Related Structures

Construction $36.5

237120 Oil and Gas Pipeline and Related Structures Construction

$36.5

237130 Power and Communication Line and Related Structures Construction

$36.5

237210 Land Subdivision $27.5

237310 Highway, Street, and Bridge Construction $36.5

237990 Other Heavy and Civil Engineering Construction $36.5

238110 through 238990 Specialty Trade Contractors $15.0

70 More information about NAICS is available at: http://www.census.gov/eos/www/naics/. Accessed February 4, 2014.

59

NAICS Codes NAICS Industry Description

Annual Revenue (millions) Employees

111110 through 111998 Crop Production $0.75

112111 Beef Cattle Ranching and Farming $0.75

112112 Cattle Feedlots $7.5

112120 Dairy Cattle and Milk Production $0.75

112210 Hog and Pig Farming $0.75

112310 Chicken Egg Production $15.0

112320 through 112990 All Other Animal Production $0.75

113310 Logging 500

211111 through 213111 Oil and Gas Extraction and Mining 500

This rulemaking will also affect private motor carriers. These carriers use CMVs they own or lease to ship their own goods (such as a motor carrier that is operated by a retail department store chain to distribute goods from its warehouses to its store locations) or in other regulated transportation activities related to their primary business activities (for example, dump trucks used by construction companies). The latter category also includes the provisions of passenger transportation services not available to the general public. FMCSA does not have NAICS codes for motor carriers and therefore cannot determine the appropriate size standard to use for each case. As shown, the size standards vary widely, from $0.75 million for many types of farms to $36.5 million for building construction firms. For for-hire motor carriers, FMCSA examined data from the 2007 Economic Census71 to determine the percentage of firms that have revenue at or below SBA’s thresholds. Although boundaries for the revenue categories used in the Economic Census do not exactly coincide with the SBA thresholds, FMCSA was able to make reasonable estimates using these data. According to the Economic Census, about 99 percent of trucking firms had annual revenue less than $25 million; the Agency concluded that the percentage would be approximately the same using the SBA threshold of $27.5 million as the boundary. For passenger carriers, the $15 million SBA threshold falls between two Economic Census revenue categories, $10 million and $25 million. The percentages of passenger carriers with revenue less than these amounts were 96.7 percent and 98.9 percent. Because the SBA threshold is closer to the lower of these two boundaries, FMCSA has assumed that the percent of passenger carriers that are small will be closer to 96.7 percent, and is using a figure of 97 percent.

For private carriers, the Agency constructed its estimates under the assumption that carriers in the 99th percentile in terms of number of CMVs of for-hire property carriers will be large. In the case of for-hire property carriers, we assumed that carriers in the 97th percentile will also be large. That is, any company large enough to maintain a fleet large enough to be considered a large truck or bus company will be large within its own industry. This could overestimate the 71 U.S. Census Bureau. 2007 Economic Census. http://www.census.gov/econ/census/ . Accessed December 18, 2013.

60

number of small, private carriers. However, the Agency is confident that no small private carrier would be excluded. The Agency found that for property carriers, the threshold was 194 CMVs, and that for passenger carriers, it was 89 CMVs. FMCSA identified 195,818 small private property carriers (99.4 percent of this group), and 6,000 small private passenger carriers (100.0 percent of this group).

The table below shows the complete estimates of the number of small carriers. All told, FMCSA estimates that 99.1 percent of regulated motor carriers are small businesses according to SBA size standards.

Table 41. Estimates of Numbers of Small Entities

For-Hire General Freight

For-Hire Specialized

Freight For-Hire

Passenger Private

Property Private

Passenger Total Carriers 176,000 152,000 8,000 197,000 6,000 539,000 Percentage of Small Carriers 98.9% 98.9% 97.0% 99.4% 100.0% 99.1%

Number of Small Carriers 174,064 150,328 7,760 195,818 6,000 533,970

7.6 DESCRIPTION OF REPORTING, RECORDKEEPING AND OTHER COMPLIANCE REQUIREMENTS OF THE RULE

FMCSA believes that implementation of the final rule will not require additional reporting, recordkeeping, or other paperwork-related compliance requirements beyond what are already required in the existing regulations. In fact, the final rule is estimated to result in paperwork savings, particularly from the elimination of paper RODS. Furthermore, the carriers will experience compensatory time-saving or administrative efficiencies as a result of using ELD records in place of paper RODS. The level of savings will vary with the size of the carrier implementing the systems (larger carriers generally experience greater savings).

Under current regulations, most CMV drivers are required to fill out RODS for every 24-hour period. The remaining population of CMV drivers is required to fill out time cards at their workplace (reporting location). Motor carriers must retain the RODS (or timecards, if used) for 6 months. FMCSA estimates annual recordkeeping cost savings from this rule of about $805 per driver. This comprises $558 for a reduction in time drivers spend completing paper RODS and $65 submitting those RODS to their employers; $144 for motor carrier clerical staff to handle and file the RODS; and $38 for elimination of expenditures on blank paper RODS for drivers. One of the options discussed in the final rule (Option 1) would extend the ELD mandate to carrier operations that are exempt from the RODS requirements. Paperwork savings would not accrue to drivers engaged in these operations.

61

7.7 STEPS TO MINIMIZE ADVERSE ECONOMIC IMPACTS ON SMALL ENTITIES

Of the population of motor carriers that FMCSA regulates, 99 percent are considered small entities under SBA’s definition. Because small businesses constitute a large part of the demographic the Agency regulates, providing exemptions to small business to permit noncompliance with safety regulations is not feasible and not consistent with good public policy. The safe operation of CMVs on the Nation’s highways depends on compliance with all of FMCSRs. Accordingly, the Agency will not allow any motor carriers to be exempt from coverage of the final rule based solely on a status as a small entity.

The Agency recognizes that small businesses may need additional information and guidance in order to comply with the regulation. To improve their understanding of the rule, FMCSA intends to conduct outreach aimed specifically at small businesses. FMCSA will conduct webinars and other presentations upon request as needed and at no charge to the participants. These will be held after the final rule has published and before the rule’s compliance date. To the extent practicable, these presentations will be interactive. Their purpose will be to describe in plain language the compliance and reporting requirements so they are clear and readily understood by the small entities that will be affected.

ELDs can lead to significant paperwork savings that can offset the costs of the devices. The Agency, however, recognizes that these devices entail an up-front investment that can be burdensome for small carriers. At least one vendor, however, provides free hardware and recoups the cost of the device over time in the form of higher monthly operating fees. The Agency is also aware of lease-to-own programs that allow carriers to spread the purchase costs over several years. Nevertheless, the typical carrier will likely be required to spend about $584 per CMV to purchase and install ELDs.72 In addition to purchase costs, carriers will also likely spend about $20 per month per CMV for monthly service fees. We have estimated that the replacement of the original ELD will cost an additional $42 to uninstall the first device above the $584 to purchase and install the replacement, for a total of $626, occurring in both Year 5 and Year 10 of the analysis time period.73

7.8 DESCRIPTION OF STEPS TAKEN BY A COVERED AGENCY TO MINIMIZE COSTS OF CREDIT FOR SMALL ENTITIES

FMCSA is not a covered agency as defined in section 609(d)(2) of the Regulatory Flexibility Act, and has taken no steps to minimize the additional cost of credit for small entities.

72 This $584 represents $500 for acquisition and 2 hours of labor cost at $42 per hour for installation ($84 total for installation). See details presented earlier in Section 3.1.1.

73 See details presented earlier in Section 3.1.1.

62

8. UNFUNDED MANDATES REFORM ANALYSIS The Unfunded Mandates Reform Act of 1995 requires Agencies to evaluate whether an Agency action would result in the expenditure by State, local, and tribal governments, in the aggregate, or by the private sector, of $151 million or more (which is $100 million in 1995, adjusted for inflation) in any 1 year and, if so, to take steps to minimize these unfunded mandates. As Table 42 shows, this rulemaking will result in private sector expenditures in excess of the $151 million threshold for all of the options. Gross costs, however, are expected to be offset by savings from paperwork burden reductions.

The Agency is required by statute to adopt regulations requiring that CMVs operated in interstate commerce, operated by drivers required to keep RODS, be equipped with ELDs (49 U.S.C. 31137). To the extent this rule implements the direction of Congress in mandating the use of ELDs, a written statement under the Unfunded Mandates Reform Act is not required.74 However, the Agency provides its projection of the annualized costs to the private sector in Table 42 below. Additionally the Agency’s adopted option provides the lowest cost and highest net benefits of the options considered.

Table 42. Annualized Net Expenditures by Private Sector (2013 $ millions)

Cost or Savings Category Option 1 Option 2

New ELD Costs $1,336 $1,032

AOBRD Replacement Costs $2 $2

HOS Compliance Costs $929 $790

CMV Driver Training Costs $10 $8

TOTAL COSTS $2,278 $1,833

TOTAL SAVINGS (PAPERWORK) $2,438 $2,438

NET EXPENDITURE BY PRIVATE SECTORa- -$160 -$605

Notes: a A negative value for “Net Expenditure by Private Sector” represents a net savings to the private sector, with total

savings exceeding total costs.

74 Unfunded Mandates Reform Act of 1995. Title II, Section 201. http://www.gpo.gov/fdsys/pkg/PLAW-104publ4/pdf/PLAW-104publ4.pdf. Accessed March 23, 2015.

63

APPENDIX A: IMPACT OF ELDS ON ROADSIDE INSPECTION VIOLATIONS FOR FIVE COMPANIES

Five motor carriers, currently using ELDs on all of their vehicles, were evaluated to assess the impact of the ELD installation on roadside inspection violations. Three of the motor carriers were required to have their ELDs installed in lieu of paying fines as a result of settlement agreements with FMCSA. The other two carriers voluntarily installed ELDs on all of their vehicles and currently have service agreements with Qualcomm, Inc. The five carriers are identified in this appendix by letter: A–C are the three that used ELDs because of their settlement agreements, and D–E voluntarily installed them.

A.1 METHODOLOGY

The data available to be analyzed were limited. For this analysis, the Agency wanted to evaluate carriers that had at least 1 year of post-ELD installation inspection history. Several carriers have been required to install ELDs on their vehicles as a result of Settlement Agreements (SA), but most of these carriers have installation deadlines that are less than a year old and thus do not have a full year’s worth of “post-installation” inspection history. Out of all carriers with SAs requiring ELD installations, only three were required to have 100 percent of their fleets equipped with ELDs more than 12 months ago and, thus, have a full year or more of both pre-installation and post-installation inspection data.

In addition to the data available from the three carriers with settlement agreements, an ELD vendor provided data on two carriers that have ELD service contracts with it. For one of these carriers, the vendor provided the Vehicle Identification Numbers (VINs) for all of the power units in its fleet and the date that each vehicle had its ELD installed. For this particular company, only vehicles that had their ELDs installed more than 1 year ago were considered for this analysis. The second carrier installed ELDs on its entire fleet on roughly the same date, December 31, 2008, thus meeting the criteria for 1 or more years of post-installation inspection history.

For this analysis, the roadside inspection safety performance of all carriers was assessed for both the pre-ELD installation period and the post-ELD installation period. For the carriers with settlement agreements, the post-installation period began with the date the carriers were mandated to have all of their ELDs installed and ended 1 year later. The pre-installation period extended from the date of the settlement agreement and went back 1 year. Thus in the case of the settlement agreement carriers, there was a small gap of several months between the end of the pre-ELD installation period and the beginning of the post-ELD installation period. For the two carriers that voluntarily installed ELDs, this gap did not exist. However, for the first voluntary adopter, which did not install ELDs on all of its vehicles at the same time, the pre- and post-installation dates differed for each vehicle. Each roadside inspection for this carrier was classified as a pre- or post-ELD installation inspection, based on the inspected vehicle’s installation date.

Pre- and post-installation period OOS rates and violation rates were calculated for each company and in the aggregate, based on all inspections identified as belonging to each period. Since there

64

may be differences in the safety impact of ELD installation for carriers required to install them and for carriers installing them voluntarily, the two groups of carriers were evaluated separately.

Table 43 summarizes the ELD installation information for each of the five carriers.

Table 43. Pre-ELD and Post-ELD Installation Period Definitions for Carriers Evaluated

Firm Source of

Data

Pre/Post Periods

Determined By

Settlement Date

ELD Install Date

Pre-ELD Installation

Period

Post-ELD Installation

Period

A Settlement Agreement

SA date and installation

deadline 10/29/10 3/31/11 10/29/09-

10/29/10 3/31/11 – 3/31/12

B Settlement Agreement

SA date and installation

deadline 12/29/10 4/1/11 12/29/09-

12/29/10 4/1/11 – 4/1/12

C Settlement Agreement

SA date and installation

deadline 2/24/10 5/1/10 2/24/09 –

2/24/10 5/1/10 – 5/1/11

D ELD Vendor Each vehicle’s install date n/a By vehicle By vehicle By vehicle

E ELD Vendor Company-

wide install date

n/a 12/31/08 12/31/07 – 12/31/08

12/31/08 – 12/31/09

The following three metrics were used to evaluate the effect of the ELDs in the post-installation period: driver OOS rate, HOS OOS rate, and HOS violation rate. These metrics are defined as follows:

• Driver OOS rate—The percentage of driver inspections at the roadside that resulted in the driver being placed out of service due to one or more driver-related violations of the FMCSRs.

• HOS OOS rate—The percentage of driver inspections at the roadside that resulted in the driver being placed out of service due to one or more driver-related violations of the FMCSRs, and at least one of these violations pertained to HOS.

• HOS violation rate—The average number of HOS violations found per driver inspection.

A.2 RESULTS

Table 44 lists the driver OOS rates, HOS OOS rates, and HOS violation rates for both the pre- and post-installation periods for the two carriers that voluntarily installed ELDs, as well as the three carriers that were mandated to do so. The table indicates that all metrics improved dramatically in the post-ELD installation period. The combined HOS OOS rate for the “voluntary” carriers decreased by 86 percent (from 4.5 percent to 0.6 percent). The combined HOS OOS rate for the “mandated” carriers decreased by 90 percent (from 4.0 percent to 0.4 percent). The overall driver OOS rates fell in a similar manner during the post-installation period, although this drop is mainly attributable to the drop in the HOS OOS rate.

65

The average number of HOS violations per inspection fell from roughly 0.5 violations per inspection to 0.05 violations for the “voluntary” carriers and from roughly 0.4 to 0.04 violations per inspection for the “mandated” carriers.

Table 44. Driver OOS Rates, HOS OOS Rates, and HOS Violation Rates for Voluntary and Mandated Carriers before and After ELD Installation (1-Yr. Evaluation Periods)

Carrier Type/Name

Pre-Period

Inspect-ions

Post-Period

Inspect-ions

Pre-Period Driver OOS

Rate

Pre-Period

HOS OOS Rate

Pre-Period

HOS Viols per

Insp.

Post-Period Driver OOS Rate

Post-Period

HOS OOS Rate

Post-Period

HOS Viols per

Insp. VOLUNTARY

D 1,222 3,928 5.4 % 4.8 % 0.53 1.4 % 0.7 % 0.06 E 487 406 3.7 % 3.7 % 0.31 0.3 % 0.0 % 0.03

TOTAL VOLUNTARY 1,709 4,334 4.9 % 4.5 % 0.47 1.3 % 0.6 % 0.05 MANDATED

A 789 1031 4.9 % 3.7 % 0.32 1.1 % 0.2 % 0.03 B 176 135 5.1 % 5.1 % 0.41 0.8 % 0.8 % 0.03 C 169 144 4.1 % 4.1 % 0.44 1.4 % 1.4 % 0.06

TOTAL MANDATED 1,134 1,310 4.9 % 4.0 % 0.35 1.1 % 0.4 % 0.04

Table 45 and Table 46 show the frequency distribution of specific violations found for the “voluntary” and mandated carriers, respectively, during the pre- and post-ELD installation periods. The violations are distributed over the sections of part 395, Hours of Service of Drivers. The tables indicate that the installation of the ELD does not eliminate RODS and maximum driving time violations, although they are dramatically reduced. The most common violation, § 395.15(g), in the post-ELD period for the mandated carriers, pertains to having an instruction sheet onboard for the ELD. FMCSA investigated the driver OOS rates and HOS OOS rates from all roadside inspections conducted in 2008-2010. In 10.3 million inspections done during that time interval, the driver OOS rate was 5.7%, and the HOS OOS rate was 4.0%. FMCSA notes that the population inspected in that time already included some carriers that used AOBRDs, FMS, or some other electronic driver logs, however this number is unknown. While these comparisons do not speak to the impact of ELDs they do place the violation rates of the sampled carriers in context.

66

Table 45. Distribution of HOS Violations for Pre- and Post-ELD Inspections, for Voluntary Carriers

Part 395 Section75

Pre-ELD Violations Frequency

Pre-ELD Violations

Percent Part 395 Section76

Post-ELD Violations Frequency

Post-ELD Violations

Percent 8(f)(1) 292 36.73% 8(e) 10 20.41% 8 164 20.63% 8(f)(1) 8 16.33% 3(a)(2) 132 16.60% 15(g) 7 14.29% 3(a)(1) 88 11.07% 3(a)(1) 7 14.29% 8(e) 63 7.92% 8 5 10.20% 8(k)(2) 32 4.03% 3(a)(2) 5 10.20% 3(b) 11 1.38% 15(c) 4 8.16% 5(b) 6 0.75% 8(k)(2) 2 4.08% 8(a) 5 0.63% 15(b) 1 2.04% 15(b) 2 0.25% TOTAL 49 100.00% TOTAL 795 100%

Table 46. Distribution of HOS Violations for Pre- and Post-ELD Inspections, for Mandated Carriers

Part 395 Section77

Pre-ELD Violations Frequency

Pre-ELD Violations

Percent Part 395 Section78

Post-ELD Violations Frequency

Post-ELD Violations

Percent 8(f)(1) 125 31.25% 15(g) 16 33.33% 3(a)(2) 84 21.00% 8 9 18.75% 8 82 20.50% 8(e) 7 14.58% 3(a)(1) 48 12.00% 8(a) 4 8.33% 8(e) 40 10.00% 8(f)(1) 4 8.33% 8(k)(2) 8 2.00% 15(b) 2 4.17% 3(b) 6 1.50% 3(a)(1) 2 4.17% 8(a) 3 0.75% 3(a)(2) 2 4.17% 15(g) 2 0.50% 8(k)(2) 2 4.17% TOTAL 400 100% TOTAL 48 100.00%

A.3 LOWER BOUND OF ELD EFFECTIVENESS MEASURES

The data on mandated and voluntary ELD users revealed significant reductions in HOS violations and OOS rates, but neither type of carrier is typical of a randomly selected carrier from the population. For both groups, ELD installation was done to address HOS compliance problems, either because they were noted by FMCSA or because the carrier flagged them itself. In other words, albeit for different reasons, all of these carriers have higher than average vigilance regarding HOS compliance. Although carriers that do not already use ELDs might not 75 Please see the list of violations and descriptions at the end of Appendix A for details of the violations. 76 Please see the list of violations and descriptions at the end of Appendix A for details of the violations. 77 Please see the list of violations and descriptions at the end of Appendix A for details of the violations. 78 Please see the list of violations and descriptions at the end of Appendix A for details of the violations.

67

willfully violate the HOS rules, they may be less vigilant or indifferent to these rules. The benefits of reducing HOS violations do not occur automatically; carriers must review and act on the data. For these reasons, ELDs may be less effective at reducing HOS violations for many carriers than they are for carriers that installed these devices specifically to improve their HOS compliance.

In order to construct a more conservative estimate of ELD effectiveness, FMCSA first examined OOS violations. In Table 47, we discuss four hours-of-service violations (49 CFR 395.3). The 14-hour rule refers to the maximum length of time that a driver may be in service, driving or otherwise, before being required to have a 10-hour break. The 11-hour rule is the maximum time during that 14-hour shift that the driver can actually be driving. The over 60 hours/7 days or 70 hours/8 days rule refers to how long the driver can be on duty, during 7 or 8 days, before being required to have at least a 34-hour break. Carriers that are not exceptionally conscientious about safety and do not have significant noncompliance problems will most likely focus on these violations because they cause disruptions to their operations. Not only are a driver and CMV stopped for at least 10 hours, but also the carrier must remedy the situation and report back to the jurisdiction issuing the OOS citation. Next, the Agency constructed a 95 percent confidence interval around the average reductions in HOS OOS violations and assumed, in order not to overestimate the effectiveness of ELDs, that typical carriers would reduce their violations at the lower bounds of those confidence intervals. Table 47 and Table 48 show the results of this analysis.

Table 47. Adjusted ELD Effectiveness, Voluntary Group

11-hour

Rule 14-hour

Rule

Over 60 Hours/7 Days or 70 Hours/8

Days

Missing, Incomplete, Improper, or

Fraudulent RODS Pre-ELD OOS Violations 18 23 7 39 Pre-ELD Inspections 1,767 1,767 1,767 1,767 Pre-ELD OOS Rate 1.0% 1.3% 0.4% 2.2% Standard Error, Pre-ELD OOS Rate 0.2 0.3 0.1 0.3 Post-ELD OOS Violations 1 10 0 18 Post-ELD Inspections 4,334 4,334 4,334 4,334 Post-ELD OOS Rate 0.0% 0.2% 0.0% 0.4% Standard Error, Post-ELD OOS Rate 0.0 0.1 0.0 0.1 Difference in OOS Rates at Mean 1.0 1.1 0.4 1.8 Standard Error of Difference 0.2 0.3 0.1 0.4 Difference in OOS Rates at Lower Bound of 95% Confidence Interval 0.5 0.5 0.1 1.1

Percent Improvement at Lower Bound 52% 40% 26% 49%

68

Table 48. Adjusted ELD Effectiveness, Mandatory Group

11-hour

Rule 14-hour

Rule

Over 60 Hours/7 Days or 70 Hours/8

Days

Missing, Incomplete, Improper, or

Fraudulent RODS Pre-ELD OOS Violations 16 22 3 16 Pre-ELD Inspections 1,134 1,134 1,134 1,134 Pre-ELD OOS Rate 1.4% 1.9% 0.3% 1.4% Standard Error, Pre-ELD OOS Rate 0.4 0.4 0.2 0.4 Post-ELD OOS Violations 1 1 0 5 Post-ELD Inspections 1,310 1,310 1,310 1,310 Post-ELD OOS Rate 0.1% 0.1% 0.0% 0.4% Standard Error, Post-ELD OOS Rate 0.1 0.1 0.0 0.2 Difference in OOS Rates at Mean 1.3 1.9 0.3 1.0 Standard Error of Difference 0.4 0.4 0.2 0.4 Difference in OOS Rates at Lower Bound of 95% Confidence Interval 0.6 1.0 0.0 0.3

Percent Improvement at Lower Bound 45% 54% -13% 19%

For voluntary users, the average effectiveness at the lower bound, weighted by the number of violations, was 45 percent. For the mandatory group, it was 35 percent. Because average carriers do not have above-average HOS violation rates, the Agency assumes that the slightly higher level of effectiveness is more appropriate to use in this analysis. Consequently, all safety benefits and compliance costs were estimated based on ELDs eliminating 45 percent of HOS violations.

A.4 HOURS OF SERVICE VIOLATIONS CITED IN TABLES 45 AND 46

Violations are tabulated as recorded by field personnel during inspections or other stops. Some offenses were recorded as violations of Part 395 Section 8, without information on which sub-section of 395.8.

395.3(a)(1) Start of work shift. A driver may not drive without first taking 10 consecutive hours off duty 395.3(a)(2) 14-hour period. A driver may drive only during a period of 14 consecutive hours after coming on duty following 10 consecutive hours off duty. The driver may not drive after the end of the 14-consecutive-hour period without first taking 10 consecutive hours off duty. 395.3(b) No motor carrier shall permit or require a driver of a property-carrying commercial motor vehicle to drive, nor shall any driver drive a property-carrying commercial motor vehicle, regardless of the number of motor carriers using the driver's services, for any period after—

(1) Having been on duty 60 hours in any period of 7 consecutive days if the employing motor carrier does not operate commercial motor vehicles every day of the week; or (2) Having been on duty 70 hours in any period of 8 consecutive days if the employing motor carrier operates commercial motor vehicles every day of the week.

69

395.5(b) No motor carrier shall permit or require a driver of a passenger-carrying commercial motor vehicle to drive, nor shall any driver drive a passenger-carrying commercial motor vehicle, regardless of the number of motor carriers using the driver's services, for any period after—

(1) Having been on duty 60 hours in any 7 consecutive days if the employing motor carrier does not operate commercial motor vehicles every day of the week; or (2) Having been on duty 70 hours in any period of 8 consecutive days if the employing motor carrier operates commercial motor vehicles every day of the week.

395.8 Violations recorded without information about which subsection of 395.8 was violated. 395.8(a) Except for a private motor carrier of passengers (nonbusiness), every motor carrier shall require every driver used by the motor carrier to record his/her duty status for each 24 hour period using the methods prescribed in either paragraph (a)(1) or (2) of this section. 395.8(e) Failure to complete the record of duty activities of this section or § 395.15, failure to preserve a record of such duty activities, or making of false reports in connection with such duty activities shall make the driver and/or the carrier liable for prosecution. 395.8(f)(1) Entries to be current. Drivers shall keep their records of duty status current to the time shown for the last change of duty status. 395.8(k)(2) The driver shall retain a copy of each record of duty status for the previous 7 consecutive days which shall be in his/her possession and available for inspection while on duty. 395.15(b) Information requirements.

(1) Automatic on-board recording devices shall produce, upon demand, a driver's hours of service chart, electronic display, or printout showing the time and sequence of duty status changes including the drivers' starting time at the beginning of each day. (2) The device shall provide a means whereby authorized Federal, State, or local officials can immediately check the status of a driver's hours of service. This information may be used in conjunction with handwritten or printed records of duty status, for the previous 7 days. (3) Support systems used in conjunction with on-board recorders at a driver's home terminal or the motor carrier's principal place of business must be capable of providing authorized Federal, State or local officials with summaries of an individual driver's hours of service records, including the information specified in § 395.8(d) of this part. The support systems must also provide information concerning on-board system sensor failures and identification of edited data. Such support systems should meet the information interchange requirements of the American National Standard Code for Information Interchange (ANSCII) (EIARS-232/CCITT V.24 port (National Bureau of Standards “Code for Information Interchange,” FIPS PUB 1-1)). (4) The driver shall have in his/her possession records of duty status for the previous 7 consecutive days available for inspection while on duty. These records shall consist of information stored in and retrievable from the automatic on-board recording device, handwritten records, computer generated records, or any combination thereof. (5) All hard copies of the driver's record of duty status must be signed by the driver. The driver's signature certifies that the information contained thereon is true and correct.

395.15(c) The duty status and additional information shall be recorded as follows:

(1) “Off duty” or “OFF”, or by an identifiable code or character; (2) “Sleeper berth” or “SB” or by an identifiable code or character (only if the sleeper

70

berth is used); (3) “Driving” or “D”, or by an identifiable code or character; and (4) “On-duty not driving” or “ON”, or by an identifiable code or character. (5) Date; (6) Total miles driving today; (7) Truck or tractor and trailer number; (8) Name of carrier; (9) Main office address; (10) 24-hour period starting time (e.g., midnight, 9:00 a.m., noon, 3:00 p.m.) (11) Name of co-driver; (12) Total hours; and (13) Shipping document number(s), or name of shipper and commodity.

395.15(g) On-board information. Each commercial motor vehicle must have on-board the commercial motor vehicle an information packet containing the following items:

(1) An instruction sheet describing in detail how data may be stored and retrieved from an automatic on-board recording system; and (2) A supply of blank driver's records of duty status graph-grids sufficient to record the driver's duty status and other related information for the duration of the current trip.

71

APPENDIX B: FORECASTS OF ELD AND FMS USE To forecast ELD and FMS use, the Agency fitted a model to the few data points it had on the percentage of CMVs equipped with these devices. The Agency picked 1990 as a starting point: around 5-6 years after the first AOBRDs were introduced and FMS79 first came on the market. Those early FMS would have had none of the functionality required for HOS logging as defined in EOBR 1, but the Agency assumes that early adopters will have upgraded their FMS to those with fully compliant ELD functions by the time the rule would become effective. The Agency relied on a simple model for technology adoption called a diffusion model. A study by Ryan and Gross (1943)80 on farmers’ adoption of hybrid seed technology is widely recognized as the starting point for modern research into the technological diffusion process. A more formal mathematical analysis of this earlier study was presented by Griliches (1957).81 The general premise is that the total percentage of new technology adopters will follow a sigmoidal shape (which may be produced by functions such as the logistic function and cumulative distribution function of the normal distribution) extending over time. A model that produces such a shape is the Bass Diffusion Model (Bass, 1969).82 This model has been in general use for decades. It has successfully been applied to products in a variety of markets, for both specialized and widely used new technology.83 Though different types of models exist, a special case of the Bass model, appropriate for when the adoption rate is zero at time zero (Mahajan and Wind, 1985),84 was used. Using 1990 as an approximate starting point, with zero adoption rates, the Agency fitted diffusion curves to the 1990 starting point and other available data points on FMS and ELD use. The curve for FMS use is given by:

𝐹𝐹𝐹𝐹𝐹𝐹 𝑈𝑈𝑈𝑈𝑈𝑈(𝑡𝑡) = 1− 𝑒𝑒−(0.065215)𝑡𝑡

1+3.979171𝑒𝑒−(0.065215)𝑡𝑡

and that for ELD use is given by:

𝐸𝐸𝐸𝐸𝐸𝐸 𝑈𝑈𝑈𝑈𝑈𝑈(𝑡𝑡) = 1− 𝑒𝑒−(0.086363)𝑡𝑡

1+29.504884𝑒𝑒−(0.086363)𝑡𝑡

79 Reference for Business Encyclopedia of Business, 2nd ed. SIC 4123 trucking except local. http://www.referenceforbusiness.com/industries/Transportation-Communications-Utilities/Trucking-Except-Local.html. Accessed November 27, 2009. 80 Ryan, B. and Gross, N. (1943). “The diffusion of hybrid seed corn in two Iowa communities.” Rural Sociology. 8(1), p. 15-24. 81 Griliches, Z. (1957). “Hybrid Corn: An Exploration in the Economics of Technological Change.” Econometrica, 25 (4): 501-522. http://www.jstor.org/stable/1905380. Accessed December 18, 2013. 82 Bass, F. M. (1969). “A New Product Growth for Model Consumer Durables.” Management Science, 15 (January), 215-227. http://www.uvm.edu/~pdodds/teaching/courses/2009-08UVM-300/docs/others/everything/bass1969a.pdf. Accessed December 18, 2013. 83 Dodson, Joe (n.d.). “New Product Forecasting: The Bass Model.” University of Washington Faculty Web Server, May 15, 2013. http://faculty.washington.edu/sundar/NPM/BASS-Forecasting%20Model/Bass%20Model%20Technical%20Note.pdf. Accessed December 18, 2013. 84 Mahajan, V. and Y. Wind (1985). Innovation Diffusion Models of New Product Acceptance: A Reexamination. http://digitalrepository.smu.edu/cgi/viewcontent.cgi?article=1127&context=business_workingpapers. Accessed February 4, 2014.

72

In both models, t represents the number of years elapsed since 1990. Figures 1 and 2 display the fitted curves for FMS and ELD use, respectively.

Figure 1. Forecast of FMS Use without ELD Rule

Figure 2. Forecast of Voluntary ELD Use without ELD Rule

73

Table 49 presents the percentages for FMS and ELD used from each of the diffusion model curves and the final estimates of technology adoption used in the analysis. To fit these curves, the Agency used several data sources that it had analyzed for the EOBR 1 rule. Survey data that Cantor collected from 415 large carriers in 2005 indicate that about 8 percent of LH CMVs were equipped with AOBRDs.85 FMCSA’s 2007 field survey of its own compliance reviews found that 5 percent of LH CMVs were equipped with AOBRDs, but data from carriers undergoing compliance reviews may be biased because these carriers may be less likely to voluntarily adopt safety technology, a finding corroborated by Cantor. A Volpe Center study estimated that from 2003 to 2005, the number of LH CMVs with FMS that could be upgraded to ELDs grew from 20 percent in 2003 to 25 percent in 2005, and from 5 percent to 8 percent for SH CMVs.86 The Volpe study indicated that the percentage of SH operations using FMS was about one-third the percentage of LH operations using this technology (8 percent for SH versus 25 percent for LH), and the Agency assumed this ratio to be the same for AOBRD use. Qualcomm hosted a webinar where an estimate of 500,000 AOBRDs in use was presented, which equates to about a 12 percent adoption rate as of 2012.87 The Agency’s model predicts a 16 percent AOBRD adoption rate for LH and 5 percent for SH, or a weighted average of about 13 percent for all CMVs. During a personal conversation after this webinar, Qualcomm indicated that this number was not 85 Cantor, D.E. et al. (2009). “Do Electronic Logbooks Contribute to Motor Carrier Safety Performance.” Journal of Business Logistics 30(1), 203-23. Abstract and availability information available at http://trid.trb.org/view.aspx?id=891607. Accessed February 4, 2014. 86 U.S. Department of Transportation, Federal Motor Carrier Safety Administration. Prepared by U.S. Department of Transportation, Volpe Center. Recommendations Regarding the Use of Electronic On-Board Recorders (EOBRs) for Reporting Hours of Service. 2005. Available in the docket: FMCSA-2004-18940-0351. 87 Qualcomm, Inc. "From the Desk of Dave Kraft: Update on the Upcoming ELD Mandate and its Impact." http://www.omnitracs.com/qualcomm-webinar-desk-dave-kraft-update-upcoming-eobr-mandate-and-its-impact. Accessed February 4, 2014.

74

based on any scientific research and could be “as low as 420,000 or as high as 610,000.”88 These numbers correspond to a range of 10 percent to 14 percent. The Agency’s estimate, although somewhat low as compared to anecdotal evidence, indicates that its technology adoption model is performing very well. The Agency does not have sufficient data to estimate separate SH technology adoption curves and assumes that technology adoption for SH will continually be one-third that of LH for any point on the diffusion curves.

Table 49. Estimated ELD and FMS Adoption

Year ELD LH ELD SH LH FMS SH FMS 1990 0% 0% 0% 0% 1991 0% 0% 1% 0% 1992 1% 0% 3% 1% 1993 1% 0% 4% 1% 1994 1% 0% 6% 2% 1995 2% 1% 7% 2% 1996 2% 1% 9% 3% 1997 3% 1% 10% 3% 1998 3% 1% 12% 4% 1999 4% 1% 14% 5% 2000 4% 1% 16% 5% 2001 5% 2% 17% 6% 2002 6% 2% 19% 6% 2003 6% 2% 21% 7% 2004 7% 2% 23% 8% 2005 8% 3% 25% 8% 2006 9% 3% 27% 9% 2007 10% 3% 29% 10% 2008 11% 4% 31% 10% 2009 12% 4% 33% 11% 2010 13% 4% 35% 12% 2011 14% 5% 37% 12% 2012 16% 5% 39% 13% 2013 17% 6% 41% 14% 2014 19% 6% 43% 14% 2015 20% 7% 45% 15% 2016 22% 7% 47% 16% 2017 23% 8% 49% 16% 2018 25% 8% 51% 17% 2019 27% 9% 53% 18% 2020 29% 10% 55% 18%

88 Telephone conversation with Dave Kraft, director of industry affairs for Qualcomm Enterprise Services, August 21, 2012.

75

Year ELD LH ELD SH LH FMS SH FMS 2021 31% 10% 57% 19% 2022 33% 11% 59% 20% 2023 35% 12% 60% 20% 2024 37% 12% 62% 21% 2025 39% 13% 64% 21% 2026 41% 14% 66% 22% 2027 43% 14% 67% 22%

76

APPENDIX C: PRICE AND AVAILABILITY OF ELDS

C.1 ELD VENDORS

Understanding the market structure of the ELD industry is crucial to determining how firms and consumers interact, the supply and demand for the devices, and their prices. The rulemaking requiring ELDs to be used by certain carriers resulted in an expansion of the ELD manufacturing industry, with small niche firms developing to meet the anticipated increase in demand. In 2010, when EOBR 1 was published, Qualcomm Incorporated dominated the ELD manufacturing industry, and FMCSA used Qualcomm devices to analyze the expected costs of the devices. Since then, the market landscape has expanded and is operating as monopolistic competition—led by Omnitracs in the United States as the dominant firm—with a competitive fringe made up of nine smaller firms within the United States. Some of the most notable changes in the market landscape have been sales and mergers among some of the top ELD producing companies. Qualcomm sold its Omnitracs fleet management technology division to Vista Equity Partners for $800 million in 2013 and is still operating under the Omnitracs LLC name. Also, in September 2014 Omnitracs acquired XRS Corporation (formerly XATA) for approximately $178 million. As noted by Omnitracs CEO John Graham at the merger’s announcement, “This agreement to purchase XRS represents another milestone in the renewed growth of Omnitracs and further demonstrates our continued commitment to retaining and advancing our position as the leading software provider within the transportation and logistics industry.”89 As a result of the merger, Omnitracs now holds 48 percent of the U.S. market, with JJ Keller far behind with only 11 percent market share, and Rand McNally and Zonar each with 9 percent market shares. Continental AG and Teletrac are also large dominant firms; however, they operate on a global scale with Continental controlling 70 percent of the global market share. The industry is exceptionally fragmented. While Qualcomm has had a dominant market share in producing ELDs in the past—controlling approximately 30 percent of the market—the recent mergers and acquisitions have allowed Omnitracs to grab more monopoly power.

89 Omnitracs to Acquire XRS, Corp. Press Release. September 2, 2014. https://xrscorp.com/blog/news/omnitracs-acquire-xrs-corporation/

77

Table 50. Major ELD Vendors

Company Share of U.S.

Market Annual Sales (2013 $

millions) Employees Omnitracs (includes XRS) 48.03% $105.60 NA JJ Keller 11.20% $24.63 1,300 Rand McNally 9.82% $21.60 200 Zonar 9.69% $21.30 NA Inthinc 7.23% $15.90 85 Fleetmatics 6.69% $14.70 35 Networkfleet 5.87% $12.90 95 PeopleNet (Trimble company) 0.68% $1.50 25 Fleetboss 0.68% $1.50 15 GPS Fleet Solutions (d.b.a. Fleetistics) 0.11% $0.25 5

C.2 DOMINANT FIRMS

The ELD market has a CR-4 measure of 78.74 percent, meaning that the four largest firms in the industry hold almost 79% of the total market share. The industry is highly concentrated, likely due to the nature of the product and characteristics of customers. ELDs are typically sold to motor carriers with fleets of vehicles, not to individual drivers, so if an ELD vendor is able to enter a contract with a large trucking fleet or several fleets, it can easily obtain a large share of the market and dominate sales. Once firms obtain customers with the initial purchase of the device, they can take advantage of the high switching costs consumers face. The dominant firms can keep their current customers for future purchases, due to fleet expansion or updating the ELD systems. Trucking companies are likely to keep the same system for their entire fleet and not switch companies when new systems are needed to prevent the need for additional training and integration costs in switching systems.

78

Figure 3. U.S. Firms Annual Sales for 2013 (Million $)

C.3 SMALL/EMERGING FIRMS

In recent years, there has been a growing presence of small/emerging firms entering the ELD market. The new competitors are producing low-cost terrestrial-based and satellite-based Fleet Management Systems (FMS), creating more product differentiation and allowing for more consumer choice in the market. Advancements in technology over the past 3 years (e.g., the use of Tablet devices and the ease of developing and downloading smartphone/Tablet Apps) has allowed other firms to expand and emerge with new electronic logging solutions. Firms are offering more options for fleets to access and record their information, and the in-cab monitor with paper printouts is no longer the standard. Instead, most of the ELD Vendors either offer their own Tablet device, which is typically removable from the cab, and/or offer an App for download, so drivers are able to use their own smartphone or Tablet to record their information.

The presence of these smaller firms can work to drive the large firms to lower prices or offer a wider range of products.

These small firms are also taking advantage of the cost savings of using smartphone technology to run the monitoring programs. For example, UDrove and BigRoad Apps operate through the WiFi and cellular data networks to track and allow drivers to record their information. While these are not yet fully compliant with DOT regulations, a representative at BigRoad revealed it will soon be releasing a device that connects to the engine and is FMCSA compliant. With the

79

engine connectivity, the Big Road App will automatically enter logging information, generate HOS charts, calculate driver hours remaining, and issue alerts for any DOT hours of service violations.

C.4 INTERNATIONAL FIRMS

Domestically, there are five key competitors firms with annual sales over $40 million; internationally, there are several large competitors firms in Europe and Mexico. These competitors are offering new value-added products and services similar in many cases to our existing or developing technologies. The major international competitors include Continental AG (Germany), Web Tech Wireless (Canada), and Teletrac (Europe and North America). Continental AG and Teletrac are both large, dominant players in their regions with $45.88 billion and $19.12 billion in overall revenues respectively.

C.5 COSTS OF SAMPLE DEVICES

The device examined in the analysis for cost estimates is the Omnitracs MCP50 FMS that uses terrestrial communications. Omnitracs has the largest market shares of FMS; although its devices are slightly more expensive, devices with costs close to that of the MCP50 may be the most abundant in the market when the compliance phase-in begins.

FMCSA has limited information on the ability of vendors to produce adequate supplies of ELDs by the compliance date of the rulemaking. However, there is currently sufficient depth in the market to suggest that vendors would be able to meet demand. The Agency estimates that about one-half of LH operations currently use FMS or ELDs and is forecasting that about three-quarters of LH operations will have been using these devices absent this rulemaking. It is also estimated that about 15 percent of SH operations use these devices and that this will have grown to about 25 percent without this rule. Vendors are currently able to meet demand, and the Agency believes that they have already been anticipating growth in ELD and FMS markets. However, the sudden growth at the time this rule goes into effect will be much larger than what vendors have been recently experiencing. FMCSA estimates that LH FMS (with and without ELD functionality activated) will have grown 5 to 6 percent per year in the 5 years leading up to the effective date of this rule. When the rule becomes effective, FMCSA estimates that producers would at least have to double their output to meet demand. Given industry awareness of a future ELD mandate, it is possible that manufacturers will have anticipated this demand.

Table 51 presents additional comparisons of device costs. The cost of producing devices may decline over time, but the Agency is uncertain, given the surge in demand caused by the compliance date, whether these cost savings will be passed on to purchasers in the near term. The Agency has found that many manufacturers do not disclose costs of extra required hardware in their promotional material, such as a smartphone (usually running the Android operating system) as a user interface or the hardware needed to connect to a CMV on-board diagnostics port such as a J1939 port. FMCSA tried to make reasonable assumptions about these other costs. The price of smartphones was especially problematic because the full cost of the phone is often embedded in monthly service fees that a customer is required to pay over a contract, typically 2 years long. FMCSA found the median price for an Android smartphone to be $495. However, as of

80

December 2013, Android phones comprised 51.5 percent of the smartphone market,90 so the Agency assumed that only the remaining 48.5 percent of drivers would need to purchase these phones, which brings the average cost paid by carriers down to $240 ($495 × 48.5%). This cost represents an industry-average expense—about half will have a compatible phone, and about half must buy a new phone to use with their ELD. FMCSA also assumed that the amount of data used by a smartphone used as an ELD would require a data plan costing $50 per phone, which conforms to the lowest cost noncontract plans that include unlimited data; this cost was adjusted to $24 ($50 × 48.5%) because current users of these phones would already be paying this fee. Installation for most devices was assumed to take 30 minutes at a labor cost of $ 42 per hour, resulting in $21 per device.

90 https://www.comscore.com/Insights/Press-Releases/2014/2/comScore-Reports-December-2013-US-Smartphone-Subscriber-Market-Share. Accessed November 19, 2014.

81

Table 51. Comparison of ELD Cost Estimates (2013 $)

Device Purchase/ Activation Install Monthly Fees

Extra Hardware

Monthly Fees for

Extra Hardware Repair/Replacement

Annualized Purchase,

Installation, Replacement

Annual-ized Fees

Total Annual-

ized Cost

JJ Keller (Compliance Tablet)

$199 ELD $21 $45 $240 phone

$24 data plan

$299.00 (compliance tablet)

$155 $198 $353

JJ Keller (BYOD) $199 ELD $21 $45 $240 phone

$24 data plan

$166.30 (smartphone) $86 $198 $284

RAND McNally TND 760

$299 ELD/ GPS $21 $19.95 (6 months free,

then $19.95/month)

$119.99-699.99

$299 replace $149 $232 $381

Continental VDO RoadLog91

$499 ELD $21 $0 $60 $499 replace $166 $0 $166

Omnitracs MCP50

$500ELD/ FMS $84 $20/month $126 labor +$500 replacement

$187 $232 $419

Omnitracs MCP110, 200

$500 ELD $84 $39.95/month $126 labor + $500 replacement

$187 $480 $667

FMS Upgrade $8 $0 $96 $96

91 http://www.ryderfleetproducts.com/continental-auto-3290.95200100/e-log-driver-log-vdo-roadlog-intro-bundle-p-cnn-329095200100. Accessed November 20, 2014. Price with ELD only without fleet management software.

82

APPENDIX D: COMPLIANCE COSTS OF THE HOS RULES

Most of the content of this appendix has been taken verbatim from the RIAs prepared for the 2003 and 2005 HOS rules. This appendix provides detail about FMCSA’s methodology for estimating costs and benefits. FMCSA has not undertaken a comprehensive survey of drivers to measure the level of noncompliance with the HOS rules since enactment of the 2003 HOS rule. The Agency does not attempt to directly measure the costs and benefits of the increased HOS compliance that is expected with the adoption of ELDs. Instead, FMCSA starts with the level of noncompliance that was found when drivers were surveyed prior to 2003. Then the Agency replicates portions of the analysis that was performed to estimate costs and benefits of the changes in the 2003 and 2005 rules to account for the changes in the HOS rules and changes in drivers’ compliance with the rules. All sources and citations were accurate when the RIAs quoted in this appendix were originally published. FMCSA, however, cannot guarantee the current availability of data and publications from outside sources.

D.1 BASIS FOR 2003 RULE AND DERIVATION OF PRE-2003 STATUS QUO

This section summarizes the survey data and analytical techniques used to evaluate the baseline level of non-compliance with the HOS rules. Survey data provided information on drivers’ actual schedules, which were than evaluated against schedules that fully complied with the pre-2003 and 2003 HOS rules. Simulated adjustments were made to any schedules that were found to be noncompliant to bring them to the minimal level of compliance. These adjustments were then translated into a redistribution of hours for a given driver and across drivers to estimate compliance costs and safety benefits.

The status quo level of compliance was derived from a simulation of driver work and rest schedules calibrated with data from two driver surveys. After this status quo was established, FMCSA determined what changes to driver schedules would have to occur should 100 percent of noncompliant drivers move toward full compliance with the pre-2003 HOS rules, and then what changes would occur if it were the 2003 HOS rules that drivers were complying with. Results for costs and benefits for the pre-2003 baseline and the effects of the 2003 rules are the starting points for the compliance cost and benefits calculations estimated in this rule.

D.1.1 Survey Data Used in Status Quo

University of Michigan Trucking Industry Program (UMTIP) Driver Surveys (1997-1999), By Dale Belman et al., with the University of Michigan Institute for Social Research The first wave of the UMTIP data collection effort resulted in 573 long surveys completed by truck drivers at 19 mid-western truck stops between July and October of 1997. The second phase of the driver survey, conducted between summer 1998 and spring 1999, used the same methodology and essentially the same questions at 12 truck stops and increased the sample size to over 1,019 valid observations. Truck stops were chosen based on the number of overnight parking spaces available, which gives a measure of traffic volume. The probability sampling technique employed ensures that the selected truck stops match the distribution of overnight

83

parking spaces by both state and size category. A potential respondent was interviewed if he or she reported being a truck driver, possessed a CDL, and was driving a tractor trailer at the time of the interview.

The variables of interest contained in the UMTIP data set included hours spent sleeping, working and driving in the 24 hours leading up to the interview, hours worked in the last pay period,92 and detailed variables concerning the timing and/or duration of activities during the last completed trip (for example, waiting for a dispatch, loading/unloading, or driving). For descriptive statistics and cross-tabs, FMCSA used sample weights to account for sampling bias due to the size of the truck stops selected as survey locations.93

In cooperation with the authors of the UMTIP driver survey, FMCSA studied customized statistical outputs for particular subsets of the population surveyed. These subsets were designed to match, as closely as possible and where appropriate, the industry segments determined to reflect the most relevant profile for the present regulatory impact analysis. In particular, this data set provided several useful variables on driver type (owner-operator, employee, union, non-union, and teams), industry segment (for-hire, private carriage, truckload [TL] and less-than-truckload [LTL]), range of operations (local, regional, LH) and size of firm, which enabled comparative analyses of many different sub-groups of drivers. For comparison with other data sets, it was useful to study miles driven in the past year and miles driven on a “typical run.94”

The UMTIP driver survey provides a representative picture of certain segments of the regional and LH over-the-road (OTR) truck driver population. The survey team did not intend to capture every aspect of the trucking industry; rather, it was the express intent of the authors to sample regional and LH OTR truck drivers. For example, the authors clearly state that local pickup and delivery drivers are underrepresented. Analysis of their data further indicates that those SH drivers are not a representative sub-sample of the SH population. The survey design addressed the potential for bias by applying randomization techniques to the choice of truck stops, the choice of potential respondents, the day of the week, and the time of day. Subsets of the driver population, to the extent possible, were analyzed separately to ensure that dissimilar subsets were not grouped together. Advantages of the UMTIP driver survey are that it captures the portion of the driving population that would be most affected by the regulations, it offers a rich range of information about its subjects, and its limitations are transparent.

“Effects of Sleep Schedules on Commercial Motor Vehicle Driver Performance,” 2000, By Balkin et al. (Walter-Reed Army Institute of Research) The Walter Reed Commercial Motor Vehicle field study gathered sleep patterns via wrist actigraphy and self-reported sleep logs from 25 LH and 25 SH drivers over 2 to 3 weeks. The data was entered into the Walter Reed Sleep Performance Model. Participants were recruited through flyers at truck stops and through word-of-mouth and were required to hold a CDL. SH

92 Drivers’ responses pertaining to pay period were standardized to 7 days. 93 Weights were constructed and used according to the procedures in Belman, D., Monaco, K., and Brooks, T. (1998). “Let it Be Palletized: A Portrait of Truck Drivers’ Work and Lives,” University of Michigan Trucking Industry Program. 94 The UMTIP data were compared with a Walter Reed Field Study. See D.1.1

84

and LH drivers were differentiated based on whether they were able to return home at the end of work periods to sleep.

The variables of interest in the Walter Reed study were the date of observation, the time spent on duty and off duty each day, and the time of day and duration of each sleep period. This study represents the most accurate information available regarding truckers’ exact sleep routines by time of day. This study does not suffer from problems of other data sets that underestimate the proportion of drivers with extreme schedules because they do not sufficiently differentiate among types of truckers or their differing work schedules over the past 7 days by aggregating truckers’ schedules over time and across groups. Moreover, with exception of Walter Reed, most studies ask subjects for their subjective view of how much they are sleeping, incremented in one-hour units, and do not differentiate between time in bed versus actual sleep. The strength of the Walter Reed study resides in the precise nature of its sleep measurements and the fact that they were carried out over a relatively long period of time.

D.1.2 Driver Schedule Simulation Methods and Results UMTIP provides insufficient raw data to completely enumerate driver schedules over time. For instance, the UMTIP surveys ask drivers about hours worked over the last 24 hours rather than for a full week or for averages over time. The Walter Reed Field Study provided more information across days but is a small sample that was not randomly drawn and was insufficient to determine whether work/sleep schedules shift under current or other options. FMCSA gathered descriptive statistics describing the distribution of schedule types from these studies in order to model 25-day schedules representative of those found in the real world.

In order to model representative schedules, the Agency first estimated the distribution across individuals of the average number of hours worked per day using the average number of hours worked in a 24-hour period for LH OTR drivers excluding team drivers95 from UMTIP. The average number of hours worked is required because the relationship between truckers’ schedules and sleep is estimated and available only for sleep with on-duty hours. For modeling hours worked per day, a large distribution of average number of hours worked per day was generated from 100 random numbers with mean and standard deviation values taken from the UMTIP and Walter Reed data to represent a distribution of average number of hours worked per day. The random numbers are drawn from a normal distribution with a mean value equal to the mean number of hours worked per 24-hour period for LH OTR drivers from UMTIP (11.37 hours per day). The standard deviation of the random numbers is equal to the standard deviation across LH drivers in the Walter Reed Field Study of their average number of hours worked per day (1.88 hours).96 Rather than model 100 separate schedules with only slight differences in average length of work day, the analysis groups them into four bins representing average work day lengths around 9, 11, 13, and 15 hours on duty on average in a 24-hour period (excluding full days off duty). These bin values are chosen to divide the distribution such that the middle two values (11 and 13) represent about a third of the distribution under current compliance

95 Team drivers are less likely to be affected by the rule changes, and were not modeled. 96 This figure excludes days in which the driver was on duty for fewer than two hours. The UMTIP was preferred for the average number of hours worked per day due to its larger sample size combined with its more random sampling approach.

85

levels, with the remainder divided about equally between the other two values.97 This provides the average number of hours worked per 24-hour period worked for the current HOS rules under current compliance (status quo scenario).

Next, average number of hours worked per 24-hour period is used to simulate the number of hours worked per 8-day work period under the pre-2003 HOS rules. This is derived from the frequency distribution from UMTIP of number of days worked in the last 7-day pay period.98 Interviews with industry experts indicate that most OTR LH drivers follow an 8-day work schedule. In order to make this distribution based on 7-days apply to the pre-2003 compliance baseline, the UMTIP distribution is scaled up from days worked in 7 days to days worked in 8 days. Because the mean, median, and modal OTR driver worked 5 days in the 7-day period, it is assumed that, from each group, 5 of every 7 would work another day in an 8-day period.99 The majority of drivers in the resulting distribution worked 5 to 8 of the last 8 days. Those who work 4 or fewer days in 8 are not expected to be affected by changes in HOS rules. The analysis is simplified by reducing the number of schedules to model by combining into one bin those who worked 4 or fewer days within the 8-day period. This group, modeled as working 3 days in 8, represents 12 percent of the trucker population. Table 52 displays the original distribution of workweek lengths as well as the rescaled and simplified distributions.

Table 52. Work Week Length for UMTIP and Modeled Drivers

Driver Distribution

1 Day per

Week

2 Day per

Week

3 Day per

Week

4 Day per

Week

5 Day per

Week

6 Day per

Week

7 Day per

Week

8 Day per

Week UMTIP 7-Day Distribution 1% 3% 5% 13% 35% 24% 21% 0 Distribution Rescaled to 8 Days 0% 1% 3% 7% 19% 31% 23% 15% Simplified Distribution 0% 0% 12% 0% 19% 31% 23% 15%

No information is available in UMTIP to estimate directly the proportion of the driver population by both hours worked in 24 hours and the number of days worked. The proportion of drivers who worked on average around 9, 11, 13, or 15 hours in 24-hours is multiplied by the proportion who worked 3, 5, 6, 7, or 8 days in an 8-day schedule. This does not account for possible negative correlation between these two variables, although this is no evidence of a strong negative relationship.100 The matrix of proportions of truckers working various hours-per-day and days- 97 Nine hours is approximately the average number of hours worked in 24-hours for truckers modeled as having worked up to 10.5 or fewer hours on average. Eleven hours is approximately the average number of hours worked in 24-hours for truckers modeled as having worked 10.5 up to 12 hours on average. Thirteen hours is approximately the average number of hours worked in 24-hours for truckers modeled as having worked 12 up to 14 hours on average, and 15 hours is approximately the average number of hours worked in 24-hours for truckers modeled as having worked 14 or more hours on average. 98 This distribution is based on only those OTR drivers who reported fewer than 24 hours of combined work and sleep. 99 There is no evidence of a negative correlation of number of days worked across weeks. For this reason, it is assumed that the number of days worked in the seven-day period covered in the survey represents the average for the respective proportion of the trucker population. 100 Walter Reed field study LH driver data revealed a small negative correlation of –0.05 between hours worked and days worked in a seven-day period. This would translate into a difference of less than a third of a day shorter workweek for those averaging 15 hours of work per day than those averaging 9 hours per day.

86

per-eight has been termed the “driver schedule proportion matrix” and any individual cell within the matrix a “driver proportion cell.” The driver-schedule proportion matrix is generated first to reflect current compliance levels with HOS rules. This matrix is shown for the status quo in Table 53. The percentages in the cell that represents three days of work in an 8-day period and nine hours of work in 24 is simply the product of 12 percent and 14 percent, or 2 percent.

Table 53. Driver Schedule Proportion Matrix for Status Quo

Hours Work/ 24

Hours

Modeled Hours/24

Distribution

Simplified Hours/24

Distribution 1 2 3 4 5 6 7 8 7 1% 0% -- -- -- -- -- -- -- -- 8 3% 0% -- -- -- -- -- -- -- -- 9 10% 14% -- -- 2% -- 3% 4% 3% 2%

10 14% 0% -- -- -- -- -- -- -- -- 11 20% 34% -- -- 4% -- 7% 11% 8% 5% 12 15% 0% -- -- -- -- -- -- -- -- 13 18% 33% -- -- 4% -- 6% 10% 8% 5% 14 9% 0% -- -- -- -- -- -- -- -- 15 6% 19% -- -- 2% -- 4% 6% 4% 3% 16 4% 0% -- -- -- -- -- -- -- --

The proportion matrix was adjusted to show fully compliant schedules under the different numbers of hours worked per week allowed under the pre-2003 and 2003 HOS rules. Driver proportion cells were truncated to reflect daily and weekly limits allowed under current HOS rules. That is, if a group of drivers work too many hours per day or week, those drivers are added to a cell that complies with HOS rules. Driver proportion cells that allow too many hours per day are carried down to the cell with the next lower number of hours that meets the daily limits. If the total number of hours per week worked for that driver proportion cell remains above that allowed under a given HOS rule option, hours worked per day are reduced within the same number of days per 8-day period by shifting the proportion in that cell to the cell with the next lower number of hours per day working. If the cell already is in the nine hours of work per 24-hours cell and still is over the threshold, the proportion of hours in that cell is shifted to a cell in which drivers work fewer days per week.

D.1.3 Rolling Work and Sleep Schedules The modeled runs from a dispatching simulation are used to predict the extent to which, under the pre-2003 and 2003 HOS rules, drivers’ primary sleeping time (and, thus, their whole sleep-work cycle) steadily moves, or rolls, over a series of days or remains fixed over time.

D.1.4 Dispatching Simulation The following table summarizes the results of the dispatch simulations. The numbers in the cells in the table are index measures that reflect relative productivity of drivers and are different for the for-hire and private cases. In the for-hire case, the index reflects both number of loads moved and length of haul. Changing HOS rules affects both the number of loads carried and the

87

distances covered by a for-hire, truckload-company. If such a company can make longer hauls with a given set of resources, it is likely that it will; longer moves for the same tonnage mean more revenue. In the real world, a company might use fewer resources to carry the same number of loads the same distance in response to less restrictive HOS rules. Either response would reflect the same productivity gain. Simulation analyses of for-hire operations use a fixed level of resources, so the output varies with a change to the HOS rules.

The for-hire index is a composite, weighted one-third according to loads moved per vehicle and two-thirds according to distance moved per vehicle. Productivity measures based on delivered orders per driver per week and miles per driver per week can differ somewhat if the length of haul differs under different options. After reviewing the factors that would cause the length of haul to vary, and the relative contributions of driving to non-driving activities to producing value for shippers, FMCSA developed a composite measure of productivity that weights miles per driver per week twice as heavily as orders per driver per week. This weighting scheme, which was not intended to be precise, was based roughly on the ratio of driving time to non-driving time for a LH TL shipment.

Selection of private-carriage scenarios posed some problems. One can postulate a set of basic patterns for private-carriage operations, but there is no empirical basis for allocating shares of private-carriage activity among various patterns. Nor is it possible to specify a set of patterns and assert that they account for all, or almost all, of private movement. The analysis focused on just two private scenarios, the “national one-to-few” case and the “regional one-to-many.” The national one-to-few case could be a manufacturer shipping from one factory to a few regional-hub distribution centers (DCs) from which large numbers of stores or other DCs are served. The regional-hub DCs might be owned by the manufacturer or by its customers. The one-to-many case may be thought of as one of those regional DCs, shipping on to stores and/or lower-tier DCs as the case may be. In the private cases, output is fixed and variation in the level of resources used to produce that output is captured. Most of that variation takes the form of more or less intensive use of the same set of tractors and drivers. The index reflects the number of loads a driver could deliver in a standardized, fully employed week.

Table 54. Relative Driver Productivity from Dispatch Analysis

Pre-2003 Rules Full Compliance

Pre-2003 Rules Status Quo 2003 Rules

For-hire scenarios Short regional 100 112 109 Long regional 100 115 112 LH 100 115 114

Private scenarios Regional one-to many 100 105 109 National one-to-few 100 110 115

Estimates of total vehicle miles of travel for TL and private carriers and numbers of drivers, presented elsewhere, are used to convert these productivity changes into costs (or benefits) of rule changes. It would be incorrect, however, to use these index numbers directly for that purpose. Not all companies have operations that are always pressing against HOS limits; many

88

do not, as indicated by anecdotal evidence and from the findings from the UMTIP driver survey of hours actually worked. The impacts on for-hire TL were scaled back to 46 percent of the result from direct application of these indices and to 35 percent in the case of private carriage. The difference in these percentages simply reflects the fact that TL operations are more likely to be pushing against the HOS rules than are private operations.

D.1.5 Results of Rolling Work and Sleep Schedule Analysis The total change in time of day between the beginning and end of a working week schedule at which an OTR driver begins his or her sleeping break period was calculated. For example, if a driver begins a route at 8 am on the first day of the route and ends at 4 am on the last day of the route before an extended (multiple-day) break, the schedule is defined to have rolled backwards by four hours. The threshold at which a schedule qualifies as having work-sleep cycles that rolled is a difference of at least two hours over a route for a backward-rolling schedule and three hours over a route for a forward-rolling schedule. Our analysis suggested that less than two-hour change in sleeping time over a driving route would not result in in a significant change in schedules. This means that if compliance with HOS regulations forces the driver of a given schedule to sleep two hours longer, the carrier will still schedule the route in the same way as they had before they were compliant. Because initial tests of the model indicated that a schedule that rolls forward adds about two-thirds of the incremental crash probability as a schedule rolling backwards, the forward-rolling threshold was set to an equivalent to three hours. The likelihood of rolling was analyzed separately for the pre-2003 rules relative to full compliance and the 2003 rules, and for regional and LH operations. The number of hours a schedule rolls was also measured separately in order to down-weight schedules that roll for fewer hours than modeled.

For the current rule fully enforced, the analysis found 3 of 13 regional schedules and 7 of 11 LH to shift. The majority of these were rolling backwards. The regional schedules rolled on average 2 hours and the LH over 10 hours over the driving period. These driving periods varied in their length depending on the limits for each option and trip lengths for the driver types. FMCSA assumed that the proportion of OTR drivers is split evenly between regional and LH companies and combined the calculated proportion of regional and LH drivers whose schedules roll into an overall weighted average. That is, even though it is not expected that all of the drivers with sleeping times that roll shift by a total of 10 hours, the modeling uses a weighted proportion that would be equivalent to the proportion whose sleep periods would shift back 10 hours. For the current rule fully enforced, the result is a weighted average of 34 percent of drivers rolling backwards an average of 10 hours (given a five-day route). For the 2003 rules, 2 of 8 regional schedules and 7 of 10 LH roll backwards.101

101 This includes two schedules rolling backwards nearly two hours.

89

Table 55. Generation of Proportion of Schedules that Roll

Data Element

Pre-2003 Rules, Fully Enforced

LH

Pre-2003 Rules, Fully Enforced

Regional 2003 Rules, Fully

Enforced LH

2003 Rules, Fully Enforced

Regional

# of schedules 11 13 10 13

# of schedules that roll 7 3 7 2

% rolling 64% 23% 70% 15%

# of hours roll 10 2 3 2

% rolling (relative to rolling 10 hours)*

64% 5% 21% 3%

% of total population 50% 50% 50% 50%

Overall relative % rolling* 34% 12%

*This percentage is relative to a hypothetical option in which all schedules roll by 10 hours. The number of hours schedules roll forward are treated as equivalent to half that of rolling back.

D.2 COMPLIANCE COSTS

This section summarizes the results of applying the analysis of driver schedules to costs of achieving full compliance with HOS rules. Additional drivers (and CMVs) would have to be added to shift work away from over-utilized drivers, which entail additional labor, overhead, and capital costs. The analysis also examined the total labor pool available to the motor carrier industry, not just the driver population, and concluded that there were sufficient numbers of workers available to shift into CMV driver jobs. Given the current under-employment in blue collar occupations, the Agency believes this conclusion still holds.

Compliance costs with the HOS rules are those costs associated with changes that carriers would need to make to their operations to achieve full compliance with the HOS rules. Noncompliance is the result of over-utilization of both drivers and CMVs beyond what is allowed under the HOS rules. Assuming carriers maintain the same level of work, additional drivers and equipment would have to be acquired to complete these loads without exceeding the HOS limits.

D.2.1 Labor Cost Changes Compliance with HOS rules is expected to result in changes in labor productivity for truck drivers, leading to changes in driver labor demand. The analysis uses the changes in labor productivity obtained from simulations of trucking routes for the various options and translates them to dollar impacts based on the labor supply relationships for truck drivers.

Issues related to the truck driver wage equation, as a function of job and employee characteristics are discussed first, followed by a discussion of the labor supply elasticity for truck drivers. Components of the indirect labor costs that are associated with driver wage costs are also analyzed to complete the discussion on various aspects of the labor cost changes.

90

D.2.1.1 Estimate of Driver Wage Function To analyze the labor costs of the different options, the Agency examined the relationship between hours worked and wages earned for truck drivers. The issue of individual driver wages is important because it is one dimension of the cost of the HOS regulations. As hours of work are shifted from drivers who currently work very long hours to newly hired drivers, the cost implications for carriers depend on the employment costs per additional hour of work by existing drivers relative to the costs for new hires.

D.2.1.2 Data, Methodology, and Results The primary data source for the analyses carried out in the following section is the Current Population Survey (CPS), a household-based survey conducted by the BLS every month. The 2003 HOS RIA used annual CPS data compiled by the National Bureau of Economic Research (NBER) that gave earnings and hours worked for a randomly chosen sample. FMCSA combines annual data from 1995 to 2000 and models the wage equation for non-union truck drivers only.

The wage equation was estimated for truck drivers based on their demographic information and job characteristics. It was hypothesized that the wage earned by truck drivers depends on their hours worked,102 along with their occupational experience, and used dummy variables to capture whether they are high school graduates, married, sex, race, whether they are in the for-hire industry, as well as dummies to control for year and regional effects. The details of all the variables used in the regression, including descriptive statistics and the estimated coefficients are presented in Table 56.

102 The square of the number of hours worked captures a nonlinear effect of hours on wages.

91

Table 56. Regression Results for Truck Driver Wage Relationship, 1995-2000

Variable Coefficient Mean S.D. Min Max Natural Log Hours Worked 4.12 3.78 0.149 2.4 4.09

(5.68) Natural Log Hours Worked Squared -0.398 14.32 1.14 5.74 16.76

(-4.2) High School Diploma Earned 0.122 0.784 0.412 0 1

(13.01) Occupational Experience (Age-Years of Ed.-6) 0.023 21.28 11.93 1 71

(20.5) Experience Squared (Age-Years of Ed.-6)^2 -0.0004 595.1 591.24 1 5041

(-17.97) Married 0.076 0.634 0.482 0 1

(9.115) Gender 0.186 0.962 0.191 0 1

(9.46) Race (0=White; 1=Non-White) -0.027 0.14 0.347 0 1

(-2.414) For Hire Trucker 0.115 0.36 0.48 0 1

(14.34) Constant -0.393 --- --- --- ---

(-0.28) Notes: t-statistics in parentheses Number of observations = 11,017 Regression model includes region and year control dummies

All the variables of interest have the expected signs and most are statistically significant (except for some of the region and year control dummies).103 Of particular importance are the coefficients associated with the 2 hours worked variables and the distribution of wages implied. Table 57 presents predicted wages for different levels of weekly hours worked for the sample on non-union truck drivers only. Based on the total wage relationship, the model predicts that the average 50 hour/week driver earns $28,307, a 60 hour/week driver makes $33,588, and a 70 hour/week driver makes $38,022 annually.

103 The R-squared of the model is low (27 percent) but consistent with the evidence in the literature - see for e.g., Rose (1987), Hirsch (1988).

92

Table 57. Predicted Annual Wages for Different Hours/Week

Hours/week Predicted Annual Wage (2000 $)

40 $ 22,149

45 $ 25,336

50 $ 28,307

55 $ 31,058

60 $ 33,588

65 $ 35,907

70 $ 38,022

75 $ 39,947

80 $ 41,692

Figure 4 shows the implied relationship for the total annual wages and Figure 5 shows the average and marginal wages as a function of the hours worked.

93

Figure 4. Total Wage Curve for Non-Union Drivers

Figure 5. Marginal and Average Wage Curves for Non-Union Drivers

94

The results indicate that the total annual wages for drivers is an increasing function of hours worked that increases at a decreasing rate. This implies that the marginal cost to the firm of an additional hour of driver labor diminishes constantly as the hours of work increases. The specific shape of the total wage curve also ensures that the average wage curve is below the marginal wage for a significant part of the distribution. The downward slope of the marginal cost curve implies that as the number of hours worked by drivers under the HOS rules are curtailed, the cost savings to companies from cutting down HOS from drivers is less than the increase in cost due to the hiring of new drivers, that is, giving an hour of work to a new driver costs more than the savings from taking an hour of work away from an over utilized driver who would exceed the HOS limits. Another implication of the slope of the marginal wage curve is that every hour of driver labor does not cost the same for the trucking company. This is because of the “nonstandard” labor-leisure choice faced by truck drivers. While they are on the road, they are willing to work an extra hour for a lower marginal wage (and cost to the firm), to maximize their earnings potential, in part because the value of leisure time out on the road is low. As drivers work more and more hours, the shape of the marginal wage curve implies that the cost to the firm for the extra hours declines gradually.

D.2.1.3 Supply of Drivers Another aspect of the labor costs of the different HOS options is related to the issue of the labor supply curve in the market for truck drivers. The shape of the labor supply curve determines the impact that changes in labor demand would have on the wage rates for truck drivers, and this is expressed as the elasticity of labor supply. As discussed, there is evidence in the literature to suggest that trucking is a very competitive industry with relatively free entry and exit and that its market labor supply curve is quite elastic. This is because trucking is considered a low-skill job with relatively low fixed costs. A small change in labor demand from additional drivers needed for full compliance with the HOS rules would not lead to any substantial changes in wage rates.

D.2.1.4 Previous Studies Rose (1987)104 contends that the truck driver labor supply curve, especially for the non-union TL sector should be highly elastic. This is because “truck driving is a low-skill occupation with considerable turnover.” She also discusses the fact that there is a large pool of drivers outside of the regulated interstate trucking industry who perform the same type of job—owner-operators, private carriage drivers and delivery drivers. She argues thus that the labor supply curve should be highly elastic for this occupation as a whole. Hirsch (1988)105 also makes the same argument that truck driver labor supply is likely to be highly elastic, given the fact that it is considered a

104 Rose, Nancy. 1987. "Labor Rent Sharing and Regulation: Evidence from the Trucking Industry." Journal of Political Economy, Vol. 95, No. 6 (December), pp. 1146-78. http://economics.mit.edu/files/4315. Accessed February 5, 2014. 105 Hirsch, Barry. 1988. "Trucking Regulation, Unionization, and Labor Earnings, 1973-85." Journal of Human Resources, Vol. 23, No. 2 (Spring), pp. 296-319. Abstract and availability information at: http://www.jstor.org/discover/10.2307/145831?uid=3739936&uid=2129&uid=2&uid=70&uid=4&uid=3739256&sid=21103463086263. Accessed February 5, 2014.

95

low-skilled job. Engel106 (1998) argues that the high turnover rate in trucking, especially in the TL sector, indicates that this occupation has a highly elastic labor supply curve and provides an easy entry to new truck drivers. The author further argues that these high rates of turnover also indicate that trucking is a job that is difficult to perform over extended periods. There is evidence in the literature that the trucking industry, especially the TL sector, suffers from significant driver shortage. The details about how this can potentially impact the issue of market labor supply are discussed below.

Other studies that looked at the issue of labor supply elasticity in general (not for trucking only) have come up with estimates ranging between 2 and 5. (See Lettau (1994), Eberts and Stone (1992)). These studies introduce a spatial dimension to the analysis by looking at local labor markets and therefore are not directly comparable to the analysis here. Nevertheless, these estimates provide a “benchmark” for labor supply elasticity values. Lettau,107 for example, argues that empirical studies that look at local area labor demand-labor supply relationships, find that “an area’s elasticity of labor supply is between 2.0 and 5.0.” Eberts and Stone108 use a recursive model to identify the labor supply and demand relationships in local labor markets using CPS data. They find a labor supply elasticity of 4.9 using a five-period lag structure.

D.2.1.5 Evidence from Industry Data Analysis of historical employment data on truck drivers confirms the view held by experts on this industry that the market labor supply for truck drivers is relatively elastic. FMCSA constructed indices of truck driver employment and real wages using BLS OES data. Table 58 shows the trend of employment and real wages 2003–2013. Although driver employment has grown about 14 percent over this period, real mean and median wages have declined between 5 and 7 percent, and have declined in all but 2 years (2007 and 2009) of this period.

Table 58. Index of Truck Driver Employment and Real Wages

Year Employment Index Real Mean Wage Index Real Median Wage Index 2003 87.15 106.19 108.75 2004 87.70 104.19 106.59 2005 89.71 103.32 105.46 2006 91.73 102.50 104.46 2007 94.77 103.08 104.95 2008 97.55 102.35 104.03 2009 98.75 104.09 105.68 2010 98.86 102.96 104.09 2011 99.29 100.75 101.35 2012 100.00 100.00 100.00 2013 101.31 99.98 99.84

106 Engel, Cynthia. “Competition Drives the Trucking Industry”. Monthly Labor Review. April 1998. http://www.bls.gov/mlr/1998/04/art3full.pdf. Accessed February 5, 2014. 107 Lettau, Michael K., “Wage Adjustments in Local Labor Markets: Do the Wage Rates in all Industries Adjust?” Office of Economic Research, Bureau of Labor Statistics. May 1994. http://www.bls.gov/ore/abstract/ec/ec940070.htm . Accessed February 5, 2014. 108 Eberts, Randall W., and Joe A. Stone. “Wage and Employment Adjustment in Local Labor Markets”. W.E. Upjohn Institute, 1992. http://ideas.repec.org/b/upj/ubooks/wea.html. Accessed February 5, 2014.

96

These data support the idea that employment growth for this occupation has not been “significantly impeded by wage movements.”

D.2.1.6 The Pool of Available Truck Drivers Wage elasticity for new drivers should ideally be considered in the context of potential truck drivers. Since hiring new drivers would mean shifting or attracting workers from other competing sectors to trucking, the Agency analyzed the existing labor pool for blue-collar workers to see where the new drivers could come from. Table 59 gives the number of blue-collar workers in some of the industries that could supply additional truck drivers needed to comply with the new rules.

Table 59. Employment Levels for Blue-Collar Occupations (thousands)

Occupational Categories 2000 1995 1990 1985 Mechanics and repairers, except supervisors 4,652 4,173 4,221 4,209 Construction trades, except supervisors 5,153 4,372 4,545 4,143 Extractive occupations, including oil well drillers, explosive and mining occupations

84 90 111 141

Precision Production occupations, including metalworking, woodworking, food, textile, apparel and furnishings

1,665 1,686 1,691 1,709

Operators, fabricators, and laborers 18,319 18,068 18,071 16,816 Fabricators, assemblers, and hand working occupations 2,070 2,059 1,978 1,833 Farm occupations, except managerial 847 862 964 1,064 Related agricultural occupations (except Supervisors) 1,094 1,024 1,014 765 Forestry and logging occupations (except Supervisors) 89 116 123 96 Transportation and material moving occupations (except Supervisors and Truck Drivers)

2,382 2,209 2,165 2,057

Handlers, equipment cleaners, helpers, and laborers (except Supervisors)

5,429 4,976 4,973 4,431

Janitors and cleaners 2,233 2,071 2,222 2,049 Public transportation attendants 127 94 100 65 Baggage porters and bellhops 42 43 39 20 Total Other Blue Collar 44,186 41,843 42,217 39,398 Truck Drivers 3,088 2,861 2,627 2,414

Source: Current Population Survey The data in the above table indicate that truck drivers account for about 5 to 6 percent of a total blue-collar population.109 Most of the occupations listed above can be considered similar to truck driving in terms of attracting people. This is one reason to believe that there is a large labor pool of blue-collar workers from which to attract potential new truck drivers as a result of the HOS options. This fact, coupled with the historical trends on wage movements, suggests that changes in labor demand would not lead to substantial wage effects due to the high labor supply elasticity.

109 The definition of blue-collar workers is broader than that used by BLS to include occupations that can potentially supply new truck drivers.

97

Some analysts believe the high turnover rates in this industry are not driven by a shortage of drivers. According to a study done by the Upper Great Plains Transportation Institute110 in 1990, and quoted in a report on “What Matters to Drivers”111 published in 1997, the trucking industry does not suffer from a “shortage of drivers” to hire from. The study claims the fact that this industry has been able to sustain such high driver turnover rates over the years is an indication that the problem is not one of labor shortage, but a lack of human resource strategy to take advantage of the available driver pool.

Also, the truck driver population in the U.S. under the current conditions is predominantly middle-aged white males. The average age of drivers in the CPS sample is 39 years (for both males and females), with 96 percent of the population being male. However, according to a study done by The Gallup Organization,112 females, non-whites (or minorities) and those who have less than 15 years of experience most likely see trucking as a good occupational choice. There is a growing segment of the labor force that has remained untapped to increase the pool of drivers. Improving the working conditions of drivers and making their job characteristics consistent with other competing occupations would be one way to attract this previously unused portion of the labor force.

D.2.1.7 Turnover and its Impact on Driver Labor Another issue that is related to this and could have a potential impact on labor supply is that of turnover. Evidence suggests that this industry, particularly the TL sector, has been plagued by very high rates of turnover.113

The Gallup study argues that between 1994 and 2005, the industry would have needed to hire an additional 403,000 drivers/year (even before the 2003 HOS rule). Of these, about 320,000 (or 80 percent) would be because of “churning” or internal turnover, drivers leaving one company to go to another, because of their dissatisfaction with the present job and pay. Another 34,000 (or 8 percent) would be needed to account for growth in the industry. And the remaining 48,000 (or 12 percent) would be needed because of attrition, retirement and external turnover.

The study also notes five specific job attributes that can predict overall job satisfaction for truck drivers:114

• Steadiness of work (i.e., consistent driving assignments). • Genuine care of managers for their drivers.

110 Rodriguez, J.M. and Griffin, G.C., (1990). "The Determinants of Job Satisfaction of Professional Drivers." Journal of the Transportation Research Forum, 2, pp. 453-464. Abstract and availability information at: http://trid.trb.org/view.aspx?id=312585. Accessed February 5, 2014. 111 See Penneau, B. and Smits, R. (1997). What Matters to Drivers. J. J. Keller & Associates, Inc., December 1997. 112 See “Empty Seats and Musical Chairs: Critical Success Factors in Truck Driver Retention.” Prepared by The Gallup Organization for ATA Foundation. October 1997. http://www.atri-online.org/research/results/musical_chairs.pdf . Accessed February 5, 2014. 113 Turnover rates are the highest in the TL sector – close to 100 percent—which also has the worst pay structure – see Belzer (1995). 114 See “Empty Seats and Musical Chairs: Critical Success Factors in Truck Driver Retention”, Prepared by the Gallup Organization for the ATA Foundation, October 1997, page 4.

98

• Pay. • Support from company while on the road/ Number of hours of work.

Any improvement in the work schedules of drivers that makes driving comparable to other competing occupations could reduce driver turnover. Pay structure is also important. Since pay is also listed as an important reason for the lack of satisfaction, any changes in the rules that results in a reduction in pay for drivers could increase driver turnover. However, the net effect of better work schedules and lower overall wages is unknown. Based on the issues discussed above and the evidence from previous literature and data on truck driver labor supply, the Agency assumes a labor supply elasticity of 5 to measure the impact on wages as a result of a change in demand for drivers. An elasticity of 5 is consistent with the view held by industry analysts that trucking is a fairly low-skill, easy entry job. Although there seems to be very limited research on truck driver’s market labor supply models that are directly relevant for the 2003 HOS analysis, an elasticity measure of 5 is reasonable. Changes in labor demand due to compliance with HOS rules should not have large impacts on wage rates.

D.2.1.8 Summary of Labor Cost Assumptions Using the methodology discussed above, labor costs of compliance with HOS rules are calculated separately for LH and SH. Using the labor productivity changes for these segments as inputs, FMCSA calculated the changes in the driver population that would be needed to maintain the same VMT. Using the relationships derived from the labor supply curve for individual truck drivers, as well as for the market labor supply, FMCSA calculated the avoided (from reducing hours of over utilized drivers) and new (from giving hours to new drivers) labor costs and the overhead for the drivers.

Table 60. Summary of Labor Assumptions

LH SH

Current Driver Population 1,500,000 1,500,000

Labor Supply Elasticity 5 5

Overhead Labor Cost Proportional to Driver Labor 4.0% 4.0%

New Driver Fringe Share 31% 31%

D.2.2 Non-Wage Costs A change in the number of drivers required to conduct trucking activity requires a complementary change in the fleet size and supporting infrastructure. FMCSA identified methods for estimating the cost of new power units, trailers, parking spaces, and maintenance and insurance costs of this equipment based on a review of the literature and relevant databases, as well as on the conclusions of its industry experts. It was assumed that no new docking facilities or change in mileage-based costs occurs since no direct change in the number of deliveries or in VMT is assumed, that is the amount of total work conducted by the trucking industry is held as fixed.

99

D.2.2.1 Trucking Equipment The method used to estimate the change in the number of tractors and trailers incorporates two countervailing impacts from the change in labor productivity. The first is the obvious change in the number of trucks associated with the incremental change in the number of drivers. The second has to do with the fact that a change in the number of tractors and trailers, under an assumption of no direct change in overall fleet VMT, changes the life of the entire fleet of tractors and trailers, including the life of newly purchased trucks. For example, consider the case of lower labor productivity requiring more drivers and, hence, tractors and trailers. The existing fleet is now driving less to maintain the same VMT, meaning that the average life of each truck is longer. On average, this will translate into lower vehicle replacement based on the change in the number of trucks relative to the initial fleet.

For purposes of illustration, the preceding example will be worked through a representative case in which 10,000 new drivers are hired by trucking companies in response to lower labor productivity, where the initial fleet size is 1,500,000 drivers. Table 61 summarizes the key assumptions used in the analysis. FMCSA assessed and incorporated a number of assumptions made by National Economic Research Associates (NERA)115 (2001).

The change in motor vehicle expenditures has two implications for the economy that are estimated for purposes of conducting the regional economic analysis. First, the direct expenditures on equipment are estimated based on the starting year of the policy, the anticipated reaction time by trucking firms, and the anticipated life of assets adjusted for the change in fleet size. Second, the assumption is made that firms would not simply bear the swings in capital costs in each year. Instead, firms would finance the costs over the amortization schedule at a reasonable weighted-average cost of capital.

The money to cover these transactions is assumed to come from personal consumption in the regional economic framework. In year one of an amortization schedule, consumers provide the cash to cover the change in purchases of new vehicles, net of the first year’s principal and interest payments. In subsequent years, consumers receive the remaining principal and interest payments associated with the original loan in the first year. The principal and interest costs represent an increase in the production costs facing trucking-related sectors in the economy.

115 Mark Berkman, Jesse David, Michael Liu, and Alison Pan (2000). “A Review of the Federal Motor Carrier Safety Administration’s Economic Analysis for its Proposed Hours of Service Standard.” Prepared for the American Trucking Association by National Economic Research Associates (NERA). August 3.

100

Table 61. Assumptions Used to Model Motor Vehicle Equipment Costs

Variable Units Name Assumption Source

Ratio of Tractors to Drivers Ratio Ratio_TD 0.75 NERA116

Ratio of Trailers to Tractors Ratio Ratio_TT 1.00 FMCSA

Cost of New Tractor (2004 $) $Nominal Tractor_Cost $95,000 FMCSA/NERA117

Cost of New Trailer (2004 $) $Nominal Trailer_Cost $20,000 FMCSA/NERA

Amortization Period Years Truck_Amort 5 FMCSA

Cost of Capital (% Return on Assets)*

Percent Truck_Cap 14.00% FMCSA

Phase In Period Years Truck_Phase 2 FMCSA

Initial Ratio of Trailers to Trucks Ratio Initial_RTT 2.50 NERA118

Average Tractor Life Years Tractor_Life 7 NERA

Average Trailer Life Years Trailer_Life 10 FMCSA

*Pre-tax revenue requirement based on weighted-average cost of capital assuming a debt-to-equity ratio of 1 (50% debt/50% equity), a pre-tax bond rate of 8%, and a 20% pre-tax equity rate. Returning to the example, the steps involved in estimating the changes in the demand for motor vehicle equipment are as follows (t is used to represent time by year and MM for millions):

1. Estimate the Number, Timing, and Cost of New Tractors and Trailers New Tractors = Ratio_TD * New Drivers = 0.75 * 10,000 = 7,500

New Trailers = Ratio_TT * New Tractors = 1.00 * 7,500 = 7,500

For t = 0 to Truck_Phase, New Tractorst = New Tractors/Truck_Phase = 7,500/2 = 3,750

For t = 0 to Truck_Phase, New Trailerst = New Trailers/Truck_Phase = 7,500/2 = 3,750

Tractor Costt = New Tractorst * Tractor_Cost = 3,750 * $95,000/1MM = $356.25 MM

Trailer Costt = New Trailerst * Trailer_Cost = 3,750 * $20,000/1MM = $75.00 MM

2. Estimate the Change in Asset Life of Tractors and Trailers

116 NERA evaluated ratios of trucks to drivers based on estimates of 1.14 from an ATA Driver Comparison Study (August 2000), an ATA survey of members at 1.18, and a study by Belzer, Michael, Hours of Service Impact Assessment, University of Michigan Transportation Research Institute, March 5, 1999. 117 NERA evaluated truck valuations and expected lives based on information from Martin Labbe Associates and an ATA report of costs between $60,000 and $150,000 per truck. The Agency revised up to $95,000 per tractor based on discussions with its industry experts. 118 NERA’s estimate based on ATA’s Motor Carrier Annual Report (1998) reporting 737,339 owned and leased trailers and semi-trailers and 294,658 owned and leased truck-tractors. The Agency examined the 2000 TTS National Motor Carrier Directory database that revealed ratios in the 1.95 to 2.30 range.

101

Adj. Tractor Life = Tractor_Life * [1 + New Tractors / (New Tractors + Initial Tractor Inventory)] = 7 * [1 + 7,500 / (7,500 + 1,500,000)] = 7.034829 Years

Initial Trailer Inventory = Initial Tractor Inventory * Initial_RTT = 1,500,000 * 2.5 = 3,750,000

Adj. Trailer Life = Trailer_Life * [1 + New Trailers / (New Trailers + Initial Trailer Inventory)] = 10 * [7,500 / (7,500 + 3,750,000)] = 10.01996 Years

3. Estimate the Replacement Timing and Cost for New Tractors and Trailers Tractor Costt+INT(Adj. Tractor Life) = Tractor Costt = $356.25 MM

Trailer Costt+INT(Adj. Tractor Life) = Trailer Costt = $75.00 MM

4. Estimate the Change in Existing Annual Fleet Replacement ∆ Annual Tractor Repl. = Initial Tractor Inventory*[(1/Tractor_Life)-(1/Adj. Tractor Life)] = 1,500,000 * [(1/10)-(1/10.01996)] = -1,061 Tractors/Year

∆ Annual Tractor Repl. Cost = ∆ Annual Tractor Repl. * Tractor_Cost = -1,061 * $95,000 = -$100.79 MM/Year

∆ Annual Trailer Repl. = Initial Trailer Inventory*[(1/Trailer_Life)-(1/Adj. Trailer Life)] = 3,750,000 * [(1/10)-(1/10.01996)] = -747 Trailers/Year

∆ Annual Trailer Repl Cost. = ∆ Annual Tractor Repl. * Tractor_Cost = -747 * $95,000 = -$14.94 MM/Year

5. Estimate Capital Payments for Each Year’s Change in Investment Over Time For each year, calculate aggregate net change in Capital Cost required across Steps 1 to 4

Annuitization Factor = [Truck_Cap] / [1-(1/((1+Truck_Cap)^Truck(Trailer)_Amort))] = 29.128%

Capital Payment = Annuitization Factor * Capital Cost (associated with a given year’s investment—–need to aggregate capital payments across multiple years’ investments according to the amortization life and year)

Capital Paymentt=1 = ($326.25 MM + $75.00 MM - $100.79 MM - $14.94 MM)*29.128% = $83.168 MM/Year for 5 Years

D.2.2.2 Parking Space Construction and Maintenance A change in the number of tractor-trailer sets would require that additional parking spaces be available at terminals. The construction and maintenance of new parking spaces requires both an up-front capital expenditure in the first year followed by annual maintenance costs in subsequent years. The capital expenditures would be capitalized and amortized as a cost to the trucking sector with financing assumed to be substituted for personal consumption as with equipment expenditures. Unlike trucks, no offsetting change in the life of existing parking spaces occurs,

102

because the life of a parking space is expected to be longer than the 10-year horizon under consideration in this analysis.

The assumptions used in the analysis are documented in Table 62. Information on parking space requirements for tractor-trailer sets was taken from the National Association of Truck Stop Owners (NATSO) and for auto parking spaces from International Parking Institute (IPI). Maintenance costs for auto spaces were assumed to occur at the same ratio as truck space maintenance to capital cost.

Table 62. Assumptions Used to Model Parking Space Construction and Maintenance Costs

Variable Units Name Assumption Source Ratio of Tractor/Trailers Per Acre Ratio TPA_Ratio 18.00 NATSO119, Ratio of Tractors to Drivers Ratio Ratio_TD 0.75 NERA Max. # of New Trucks at Terminal Percent Terminal_Max 75% FMCSA Capital Cost Per Acre for Trucks $Nominal TPA_Cost $100,000 NATSO O&M Cost Per Acre for Trucks $Nominal TPA_OM $10,000 NATSO Drivers Parking at Terminal Percent Terminal_TD 75% FMCSA Capital Cost Per Space for Autos $Nominal APS_Cost $1,500 IPI120 O&M Cost Per Space for Autos $Nominal APS_OM $150 FMCSA Cost of Capital (% Return on Assets)* Percent Truck_Cap 14.00% FMCSA Amortization Period Years Pkg_Amort 10 FMCSA Average Life of Parking Spaces Years Pkg_Life 20 FMCSA

*Pre-tax revenue requirement based on weighted-average cost of capital assuming a debt-to-equity ratio of 1 (50% debt/50% equity), a pre-tax bond rate of 8 percent, and a 20 percent pre-tax equity rate Not all new tractor-trailer sets would be at the terminal at any given point in time, where Terminal_Max summarizes the proportion of additional spaces to sets. In addition to parking spaces for new tractor-trailer sets, additional spaces must be constructed for new drivers to park at truck stops, rest areas, or terminals while en route. The ratio of new drivers parking at work to new tractor-trailer sets is accounted for in variable Terminal_TD. The installation and maintenance costs are estimated based on the following steps:

119 Costs per space based on Scott Imus, National Association of Truck Stop Owners. http://www.natso.com/for_members/government_downloads/truckparking_solutions2001.doc: Assume 18:1 ratio of tractor/trailer sets per acre, $100K construction costs per acre, and $8,000-$10,000 per year in annualized maintenance costs; translates to $5,555 in capital costs and $555 per year in maintenance costs per tractor/trailer set. These values are consistent with estimates of incremental, rest-area pull-off parking space construction costs of $5,000-$7,000 per space based on information derived from truck stop operators and a national rest area database in U.S. Department of Transportation, Federal Highway Administration (1996). Commercial Driver Rest & Parking Requirements: Making Space for Safety – Final Report. Report No. FHWA-MC-96-0010. May 1996. Table III-2, p. 97. http://www.fhwa.dot.gov/publications/research/safety/commercial.pdf. Accessed February 5, 2014. 120 The International Parking Institute states that, “Surface facilities can be built for $1,500 per space in most cases.” http://www.parking.org/index.html. Accessed February 4, 2014. http://wwww.fsfarchitects.com/ . Accessed February 4, 2014. Estimates the ratio of spaces per acre at 80:1 to 100:1 at a cost of approximately $1,000 per space. It was assumed that 10 percent of construction was maintenance.

103

Step 1: Estimate the Number of Parking Spaces Required New Tractor-Trailer Spaces = New Drivers * Ratio_TD * Terminal_Max = 10,000 * 0.75 * 75% = 5,625

New Auto Spaces = New Tractor-Trailer Spaces * Terminal_TD = 5,625 * 75% = 4,219

Step 2: Estimate the New Tractor-Trailer Set Capital & Maintenance Costs TT Set Pkg. Capital Cost = New Tractor-Trailer Spaces * TPA_Cost / TPA_Ratio = 5,625 * $100,000 / 18 = $31.25 MM in Year 1

TT Set Pkg. Maint. Cost = New Tractor-Trailer Spaces * TPA_OM / TPA_Ratio = 5,625 * $10,000 / 18 = $3.125 MM in Years 2+

Step 3: Estimate the New Auto Parking Capital & Maintenance Costs Auto Pkg. Capital Cost = New Auto Spaces * APS_Cost = 4,219 * $1,500 = $6.328 MM in Year 1

Auto Pkg. Maint. Cost = New Auto Spaces * APS_OM = 4,219 * $150 = $0.6328 MM in Years 2+

Step 4: Estimate Capital Payments Calculate aggregate net change in Capital Cost required across Steps 1 and 2 in Year 1

Annuitization Factor = [Truck_Cap] / [1-(1/((1+Truck_Cap)^Truck(Trailer)_Amort))] = 19.171%

Capital Payment = Annuitization Factor * Capital

Capital Paymentt=1 = ($31.25 MM + $6.328 MM)*19.171% = $7.204 MM/Year for 10 Years

D.2.2.3 Insurance Additional tractor-trailer sets have value whether they are on the road or not. Though incremental insurance costs are predominantly associated with changes in the VMT, a portion of the insurance cost is associated with the intrinsic value of the change in the capital stock represented by the change in the number of tractor-trailer sets. NERA estimated a value of $2,549 per new driver per year in insurance costs based on ATA data, and the 2003 HOS rule analysis used this estimate.121 Industry experts estimated that perhaps 25 percent of this cost is associated with the intrinsic value of the truck, and is therefore a fixed cost per CMV. The remainder is assumed to be variable with changes in VMT. FMCSA assumed that no direct change would occur in the variable portion of insurance costs since overall VMT is assumed to remain the same. The end result is a change of $637.25 per driver per year, or, $6.3725 million per year.

121 Based on data from Motor Carrier Financial and Operating Statistics, 1998.

104

D.2.2.4 Maintenance Analogous to the issue of insurance, additional tractor-trailer sets require some regular maintenance whether they are on the road or not. Though incremental maintenance costs are predominantly associated with changes in the VMT, a portion of the maintenance cost is associated with regular safety inspections and other routine, scheduled maintenance represented by the change in the number of tractor-trailer sets. Industry experts estimated a value of $8,500 per new tractor-trailer per year in maintenance costs and that perhaps 25 percent of this cost is associated with fixed maintenance costs per truck. The remainder is assumed to be variable with changes in VMT. The portion of the maintenance cost associated with new trucks is negated by changes in VMT in the rest of the fleet to yield no direct net change in VMT, and, hence, no change in the variable portion of maintenance costs. The end result is a change of $2,125 per truck per year, or $21.25 million per year.

D.2.2.5 Recruitment The need for more or fewer drivers would have an impact on recruitment costs associated with the hiring of new drivers. Rodriquez, et al. (1998),122 surveyed 15 LH, for-hire trucking firms to determine the average costs associated with driver turnover, or churn. The study estimated an average cost to firms of $5,423, as summarized by cost category in the first three columns of Table 63. The Agency excluded the costs and profits from idle equipment and the production loss due to churn since equipment costs are explicitly modeled and VMT is assumed to remain constant. It also assumed that 75 percent of the advertising and staff labor costs are based on fixed annual budgets, where only 25 percent is variable based on the change in number of new drivers to be recruited. Firms are likely to have exhausted the most important marketing channels to them given the high industry churn rate. Administrative support (Staff Labor) is assumed to be characterized by a high degree of automation, with the 25 percent assumption used to cover the fact that some back-office recruiting labor was included in this category by Rodriquez, et al. The rest of the costs are assumed to be fully variable with a change in the number of new drivers. The result is an incremental cost of $1,610 per hire as compared to the average of $5,423 based directly on the survey results.

122 Rodriguez, J., Kosir, M, Lantz, B., Griffin, G., Glatt, J., “The Costs of Truckload Driver Turnover,” Upper Great Plains Transportation Institute, North Dakota State University, April 2000. http://www.ugpti.org/pubs/pdf/SP146.pdf . Accessed February 5, 2014. Average values for churn-related costs ranged from $6,400 to $8,600 per hire, with an average of $8,234 per hire. However, ICF evaluated the averages as reported in the study for each cost category and the average number hires to determine the value of $5,423 per hire. Values in other surveys reviewed by NERA ranged from $2,000 to $20,000.

105

Table 63. Cost of Driver Churn (2013 $)

Average Costs Per Firm Total Cost Per Hire* 25% Per Hire % Fixed

Advertising $446,000 $340 $85 75%

Staff Labor $1,063,000 $811 $203 75%

Testing Fees $193,000 $147 $147 0%

Recruitment Fees $580,000 $442 $442 0%

Orientation Fees $323,000 $246 $246 0%

Training Fees $543,000 $414 $414 0%

Referral/Sign On Bonus $94,000 $72 $72 0%

Costs for Idle Equipment $2,313,000 $1,764

Lost Profits Due to Idle Equipment $705,000 $538

Production Loss Due to Turnover $849,000 $648

Totals $7,109,000 $5,423 $1,610

*Based on survey average of 1,311 hires per firm from Rodriquez, et al. (1998). The cost per driver is multiplied by the change in the number of drivers in the first year. In subsequent years, the cost per hire is multiplied by the change in the number of new drivers times the assumed churn rate for drivers. Estimates of churn rates vary from 25 percent to over 100 percent depending on the survey. The 2003 HOS RIA employed the churn rate of 25 percent assumed by NERA (2001).123 The cost in Year 1 is $1,610 * 10,000 drivers, or $16.10 million. The cost in subsequent years is equal to $16.10 times the churn rate of 25 percent, or $4.02 million per year.

D.3 RESULTS OF 2003 HOS RULE ANALYSIS

Table 64 shows the results for the estimates of the change in the number of drivers, the primary determinant of HOS compliance costs.

123 NERA claims this is a conservative assumption based on survey data indicating 25 percent churn per quarter for TL and 4 percent for NTL carriers, or ~75 percent annually across all driver types (Trucking Activity Report, June 2000, ATA). FMCSA also reviewed a compendium of surveys compiled by J.J. Keller & Associates (“What Matters to Drivers,” Neenah, WI, 1997) which generally confirms the range of estimates cited by NERA.

106

Table 64. Changes in Drivers Needed for Full Compliance with HOS Limits

Pre-2003 2003 LH Percent Change 8.1% 4.2% SH Percent Change 0.7% 1.4% LH Number 121,500 63,000 SH Number 10,800 21,300 Total 132,300 84,300

The direct costs relative to the status quo are shown in Table 65. This table shows the costs of the current rules with full compliance in the fourth column from the right. Because there would be costs for compliance with the pre-2003 rules, the costs of the current rules are higher relative to the status quo than relative to the pre-2003 rule with full compliance.

Table 65. Annual Direct Cost Changes for Full Compliance (2013 $ millions)

Cost Category Pre-2003 2003

LH Driver Labor Cost $1,185 $550 Other Costs $769 $332 Total Costs $1,954 $882

SH Driver Labor Cost $143 $233 Other Costs $90 $168 Total Costs $232 $400

Total Costs, LH and SH $2,187 $1,282

D.4 CHANGES TO COST AND BENEFIT CALCULATIONS FROM THE 2005 HOS RIA

In the 2005, HOS rule, FMCSA changed the regulations to constrain the use of sleeper berths to ensure that each sleeper berth period is at least 8 hours, and is supplemented by a two-hour break that may be outside the sleeper berth. At that time, the Agency also implemented an exemption from maintaining RODS for certain SH operations, which generates an ongoing paperwork savings. However, compliance costs and safety benefits to SH were unaffected.

D.4.1 Work Patterns Patterns of working by drivers in the different sectors of the trucking industry formed the basis for the 2005 HOS rule analysis. In particular, the analysis focused on intensity of effort; this may be thought of as the degree to which drivers work close to the limits imposed by the HOS rules. This can refer to hours worked (on-duty hours) in a week and in a day, hours driven in a day, days worked, and days off in a week. These measures are important for analysis of both productivity and safety effects of rule changes.

107

D.4.1.1 Data Sources The measures of work patterns and intensity were based on several data sources. There were four sets of data on current experience (under the 2003 HOS rule): data provided by Schneider National on some aspects of its operations; data from OOIDA based on a survey of its members; a survey of private carriers carried out by Professor Stephen Burks of the University of Minnesota; and data collected by FMCSA (the “field survey”). The Schneider, OOIDA, and Burks data were gathered with the express purpose of obtaining information on use of three aspects of the new rule: the 11th hour, restarts, and split sleeper periods. Each of these sources is focused on a different sector of the industry.

In addition to the above data, information was used from nine private interviews with carriers, eight small TL firms and one small LTL firm.

Schneider Schneider’s data are for a large TL firm and cover approximately 16,000 drivers. They were taken from company records for August and October of 2004. [(Schneider National Logbook Summary), January 2005. FMCSA 2004-19608-1996.]

OOIDA OOIDA data are based on owner-operators and a few company drivers for TL firms. OOIDA posted a survey form on its Web site asking drivers for information on use of the 2003 rule provisions in June 2004. The data used here are based on responses from 1,223 drivers.

Burks Professor Burks mailed a survey form to private carriers asking for information on their drivers’ use of the new-rule features in June 2004. He received usable responses from 29 firms covering 3,311 drivers. [See Burks, S.V., “A Survey of Private Fleets on Their Use of the Three New ‘Hours of Service Features,’” 2004. In the docket at: FMCSA-2004-19608-1991.]

FMCSA Field Survey The field-survey data largely represent company drivers with small TL companies. In terms of distribution of company size, this makes sense; most TL companies are quite small. In the field survey, 86 percent of for-hire, TL/OTR companies have fewer than 25 tractors. Viewed in terms of TL company size, the field survey is a representative sample, but these small companies account for a fairly small share of TL/OTR VMT, about 17 percent. LTL firms and private carriers are sparsely represented in the field survey.

These data, based on drivers’ log books, were obtained from companies in the course of compliance reviews or safety audits. Data cover 542 drivers with 269 firms in the period July 2004 to January 2005. For each driver, data for one month of operation were collected.

D.4.1.2 Average Hours per Day—On-Duty and Driving Two basic measures of work are daily hours of driving and total work, the latter term including all on-duty time, both driving and other work. The field survey and the Schneider data provide information on driving time per tour; only the field survey provides data on on-duty hours per tour. The field survey provides some information on local drivers; the Schneider data do not distinguish between local and OTR operations.

108

A basic assumption in the calculation is that a day is equivalent to a tour of duty. While there are exceptions, most drivers work one shift in a day. A tour of duty comprises the time from the driver’s start of work to end of work, including driving, other on-duty, and off-duty time. Results are in Table 66. That the numbers for driving hours for Schneider and OTR drivers from the field survey are so close enhances confidence in these numbers, even though the Schneider data include local service along with OTR operation.

Table 66. Daily Driving and On-Duty Hours--Averages

Field Survey Schneider Driving 7.7 7.6 On-duty 9.2 N/A

D.4.1.3 Average Hours and Days of Work Per Week For OTR drivers, a typical measure of work is number of hours in 8 days, which shows how close drivers work to the 70-hour limit for 8 days. A more complete understanding of drivers’ work patterns, though, is revealed by examining data on days worked per week. Both the field survey and the Schneider data give hours worked in 8 days, 62 hours for Schneider drivers, 59 hours for field-survey drivers.124 Some intermediate steps are required to convert these numbers to days per week. They are first divided by 9.2, the field survey figure for on-duty hours per tour of duty, to obtain hours worked per 8 days and then make a further adjustment to obtain hours worked per seven days. These results are presented in Table 67.

Table 67. Average Weekly Hours and Days Worked

Field Survey Schneider On-duty hours/8 days 59 62 Days worked per week 5.6 5.9

D.4.1.4 Intensity of Effort Examining only the average hours of driving and hours and days of work might lead one to conclude that all drivers work well within the limits imposed by the HOS rules. Many drivers work and drive longer hours than the averages, and this analysis relies on the percentages of drivers that work close to the limits for estimating productivity and safety effects. The following table shows the distributions of daily driving and on-duty hours and in 8-day periods.

Table 68. Driving Hours per Tour of Duty

Driving Hours Schneider Percentage

of Tours Field Survey Percentage of

Tours OOIDA Percentage of

Tours 11 10.7 16.2 28.0 10 15.5 16.1 N/A

9 16.4 11.2 N/A <9 57.4 56.5 N/A

124 For both data sources, the analysis discarded all drivers with fewer than 50 hours of work in 8 days on the grounds that they were not driving full-time in the period covered.

109

The OOIDA survey asked for frequency of use of the eleventh hour but did not otherwise ask about driving hours. The figure presented here was calculated from the underlying survey data. It is worth noting that the on-duty hours show a pattern relative to the 14-hour limit different from that of the driving hours relative to the 11-hour limit. Drivers are driving 10 or more hours in more than 25.0 percent of their workdays while reporting 13 or more on-duty hours for only 8.0 percent of days. The latter number suggests that drivers are generally taking 2 hours of break in a 14-hour tour or their normal work shifts are shorter than 14 hours. The Agency suspects that both are true. Inaccurate logging of on-duty hours could also be a factor.

Table 69. On-duty Hours per Tour of Duty (Field Survey Only)

On-duty Hours Percentage of Tours 14 2.7 13 5.5 12 13.2 11 15.6

<11 63.0

Table 70. On-duty Hours in 8-day Periods

On-duty Hours

Schneider Company Percentage of 8-day

Periods

Schneider Leased Percentage of 8-day

Periods

Field Survey All TL Percentage of 8-day

Periods >64 41.0 41.3 26.3

60-64 23.5 25.7 16.6 50-59 35.6 33.1 57.1

Table 69 and Table 70 show that, while daily on-duty hours tend to “bunch” away from the limit, multi-day on-duty hours bunch close to the limit, closer, indeed, than is the case for driving hours. Table 68 and Table 69 highlight differences in behavior between company drivers and owner-operators. While driving hours show a marked difference, the difference in multi-day hours is slight. Some of this could be accounted for by the fact that OOIDA data include some owner-operators working on their own authority; those in the Schneider data are all leased. Regarding differences in the average on-duty hours listed in Table 69, it should be noted that there are few owner-operators in the field-survey data; the higher percentage of 11th-hour use from the field survey, as compared with Schneider, suggests that smaller companies may push harder than larger ones, insofar as the driving limit is concerned. The OOIDA data on the 11th hour could be seen as part of such a pattern, especially if one thinks the own-authority owner-operators are using the 11th hour heavily. However, the multi-day hours show the reverse pattern. For 65.0 percent of reported instances, Schneider’s drivers have over 59 hours; from the field survey, the comparable number is 43.0 percent. This might suggest that a big company does not schedule as close to the driving limits as a smaller company might but enjoys greater success in marketing and, thus, is able to keep its drivers moving more consistently. There could, of course, be other explanations.

110

One must be wary of reaching too far in drawing inferences from these data. To the extent that data from different sources show consistent patterns, one can use this information in analysis with some confidence. One pattern that comes through consistently is that the preponderance of OTR drivers and trucking firms are not operating at, or close to, the HOS limits. Approximately 25 to 30 percent of drivers are driving more than 9 hours regularly and 25 to 40 percent of drivers are regularly working more than 64 hours in 8 days. The industry experts consulted for this analysis agreed that this is an accurate general view of industry operations.

D.4.2 Use of New Provisions of the 2003 HOS Rule The analysis examined the use of three aspects of the 2003 rule: restarts, the 11th hour, and the split sleeper-berth provision. The data come from the sources already mentioned: Schneider, OOIDA, Burks, and the FMCSA field survey.

D.4.2.1 Restarts All four of data sources reported on use of restarts. OOIDA reported that almost 90 percent of drivers used the restart at least some of the time.125 Burks reported that private carrier drivers used the restart on 61.0 percent of their runs.126 Neither OOIDA nor Burks, however, reported on length of restarts.

A driver using the restart provision may not be taking only the minimum 34 hours for the restart period. Schneider and the field survey both reported a high level of use of restarts and gave information on the length of restarts. In Schneider’s data, only 2.0 percent of restarts were only 34 hours. Depending on the reporting period, one-quarter to one-third of the restarts were 44 hours or fewer. Forty-three percent were 58 hours or fewer. Schneider showed a bi-modal distribution with peaks at 39 and 62 hours. Presumably, the former reflects cases in which the driver has taken 1 full day off, plus a few hours from the preceding and following days; the latter would reflect 2 full days off. The field survey shows that 33.0 percent of restarts were 44 hours or fewer. This comports well with the Schneider data. On this basis, one can say that at least one-third of restarts are short enough to bring a productivity gain. Using the alternative method of the moving 8-day period, drivers would usually have to stay off more than 44 hours before returning to work.

Anecdotal information on company attitudes towards restarts is that they like the provision and find some productivity gain even though drivers are staying off more than 34 hours. Managers seem hesitant to demand a return to work after 34 hours, except in unusual situations. It may, of course, be the case that taking only 34 hours off would not fit with the work schedule of many drivers, that is, there would not be anything for them to do at the 35th hour. For example, the 35th hour might come at 3:00 AM, and the company might have no use for the driver until 8:00 AM. When a TL driver comes off his restart, his first task is to pick up a new load; the hour at which the company needs his services would be set by the requirements of the shipper of that first load.

125 John H. Siebert, “A Survey of Owner-Operators and Company Drivers on their Use of Three New ‘Hours of Service Features,’” OOIDA Foundation, September 15, 2004. 126 Stephen V. Burks, A Survey of Private Fleets on their Use of Three New ‘Hours of Service Features,’” September 15, 2004.

111

D.4.2.2 Split Sleeper Berths Data clearly indicate that most drivers never split, and those that do so only do so occasionally. Schneider data for October 2004 show 97 percent of drivers never splitting and only 0.4 percent splitting “regularly.” Before the new rule, Schneider did not allow solo drivers to split at all and has only allowed them to split on an 8-and-2 basis under the new rule. The data from OOIDA and the field survey show many more drivers splitting occasionally but few splitting frequently.

Table 71. Incidence of Splitting

Splitting Frequency Field Survey OOIDA 0 times per month 66% 55% 1–4 times per month 20% 20% 0–4 times per month (sum of above rows) 86% 75% Average percent splitting per day 6% 13%

The Burks data show a higher percentage of frequent splitting, although they are not directly comparable with those from the field survey and OOIDA. They suggest that 52.0 percent of drivers split four or fewer times a month with the rest splitting more frequently. It is not clear why private drivers would split more frequently than others. There might be a higher percentage of teams in Burks’s data; evidence suggests that teams split more frequently than solo drivers.127 The data in the following table come from an Insurance Institute for Highway Safety (IIHS) survey of drivers at weigh stations in Pennsylvania and Oregon and from FMCSA’s Driver Fatigue, Alertness and Countermeasures Study (DFACS).

The IIHS and DFACS findings agree that teams split more than solo drivers. There is some anecdotal evidence that the incidence of splitting by teams is higher than that found by IIHS and DFACS. Several comments to the HOS rule docket suggested higher percentages than these and also indicated that team splitting is generally balanced; that is, sleeper periods and driving periods are about equal at 4-6 hours each.128 The IIHS/DFACS findings for solo drivers sometimes splitting are lower than those from OOIDA and the field survey.

Table 72. Incidence of Splitting-- Team and Solo

IIHS DFACS Solo 24% 22% Team 47% 52%

Data on splitting clearly show that splitting for most solo drivers occurs on an occasional and opportunistic basis. They do not build splitting into their operating routines. When they do take a split period in the sleeper, they go right back to the ten-hour rest at the next rest period. This does suggest that most drivers find the limited rest period unsatisfactory and use it only to avoid some 127 There is also some ambiguity in the Burks survey which asked for the percentage of “runs” using the splitting rule but did not define runs. It appears that some respondents interpreted “run” to be a multi-day period. The field survey data contained cases in which drivers reported splitting, though one of the “split” rest periods exceeded 10 hours. These cases were discarded, but this shows the danger of inaccurate logging of split sleeper periods. 128 Docket 19608; see comments by Yellow-Roadway, FedEx, CR England, Overnite, ATA, MCFA.

112

other problem. An unexpected period of congestion would be one example. However, routine splitting is probably part of the daily operation of many teams.

D.4.2.3 The 11th Hour Table 73. Percentage of Work Days with 11th Hour Use

Schneider OOIDA Field Survey Burks 10.4 28.0 16.2 31.0

Field-survey numbers for compliant drivers only. Burks data might overstate 11th hour use because drivers reported 11th hour use for “runs,” which may encompass an entire trip spanning multiple workdays.

Data show that the 11th hour is definitely being used. Comparing Schneider data to OOIDA and the field survey implies that big companies use it less often than small companies or owner-operators. The analysis uses the assumption that usage is heavier for smaller firms and sets 25 tractors as the demarcation point between big and small TL companies. FMCSA estimates that 40 percent of TL VMT is from small companies and 60 percent is from large companies.

D.4.3 Carrier Operation Simulations In the simulation model used to assess impacts on the more complex types of carrier operations, a truck’s progress is tracked in a computer program as the driver moves between origin and destination points, choosing new loads at the end of each run from a set of choices randomly selected from a data base representative of inter-county shipment patterns. The driver’s choices are made on the basis of which loads feasibly can be picked up and delivered within specified windows, given the limits imposed by the need to stop and rest. Within feasible choices, the driver is assumed to choose (or be assigned) the load that is most advantageous in terms of its contribution to its productivity. Because the HOS rules affect which loads can be delivered, and change the amount of time that can be devoted to driving, the model is able to estimate impacts on productivity, and the accompanying changes in typical schedules.

113

Figure 6. Operational Cycle of HOS Model

The model starts at the user defined home terminal. Out of 20 randomly generated origin-destination pairs, it chooses the pair that best fits its schedule as well as maximizes its productivity. Then it moves to the origin terminal, waits until the pick-up window time, and loads the shipment. It then drives to the destination terminal, waits until the delivery window time, and unloads the shipment. At this point, the model again analyzes another set of 20 origin-destination pairs and repeats the same procedure prescribed above for the time duration defined by the user. The movement of the truck in the model is constrained by HOS rules (i.e., all required rest periods) allowing the user to compare different facets of HOS rules with assumption of full compliance. At the end of the simulation, the model yields an output that shows how the CMV operator behaved at each time of day. The truck’s movement following the operation cycle is recorded in the schedule output table that reveals what the truck was doing at each time of the day each day during the whole simulation duration. The duration of simulation is defined by the user so the model can generate up to one year’s worth of the schedule table. Table 74 is a snapshot of the schedule table showing only a small portion of it. The table actually has over 40 columns providing details such as time of the day, day of the week, driving status, load status, origin county, destination county, cumulative duty, driving, and rest hours.

114

Table 74. Schedule Output of HOS Model

D.4.4 LH Compliance Costs The analysis of costs recognizes that the different provisions of the options would affect carrier operations in complex and interacting ways. It also recognizes that these effects would depend strongly on the carriers’ baseline operating patterns, which vary widely across this diverse industry. To produce a realistic measurement of the options’ impacts, then, FMCSA divided the industry into broad segments, collected information on operations within these segments, and then created a model of carrier operations as they are affected by HOS rules. The variety of operational patterns made it necessary to limit the analysis to the most important cases.

The model was first loaded with data representative of shipping patterns and carrier cost structures and tested to ensure that it could realistically simulate typical lengths of haul, empty mile ratios, and productivity. It was then set up to cover most important cases, under constraints representing the options, and used to simulate carrier operations under different conditions. The Agency then analyzed the data representing the simulated operations, using changes in miles driven as a measure of productivity impacts. Output measures from individual runs were weighted to give a realistic representation of the affected industry, including the drivers’ use of the most important provisions. The weighted changes in productivity from this procedure were then used to estimate the cost increases imposed on the industry by the options, using an analysis of the changes in wages and other costs likely to result from changes in productivity. These productivity-related costs were combined with transition costs associated with shifting to new rules to produce estimates of total social costs.

For representative carriers in each of several carrier size categories, the financial impact of the HOS rule was estimated in terms of the change in net income (in $2004) to the carrier, as well as a change in profits as a fraction of operating revenues. The approach used to estimate these impacts involved the development of a pro forma financial model of firms of different sizes confronted by changes in productivity, wages, and prices. Financial impacts were estimated under two assumptions about prices of trucking services: unchanged prices (representing the short run), and prices after industry-wide cost changes have been passed through to consumers.

Trip Day

Time of Day

(24HR Format)

Day of the

WeekStatus

Load Status

Origin Destn

Driving HRS Until

Arrival

Load Time

Unload Time

Dummy Rest

Dummy On-duty

Dummy Driving

Cumulative On-Duty HRS

Cumulative

Driving HRS

1 7.00 MON REST EMPTY 17,031 17,031 1.00 2.00 3.00 0 0 0 - - 1 7.50 MON DRIVE EMPTY 17,031 17,031 0.50 2.00 3.00 0 1 1 0.50 0.50 1 8.00 MON DRIVE EMPTY 17,031 17,031 - 2.00 3.00 0 1 1 1.00 1.00 1 8.50 MON LOAD FULL 17,031 5,045 11.00 2.00 3.00 0 1 0 1.50 1.00 1 9.00 MON LOAD FULL 17,031 5,045 11.00 2.00 3.00 0 1 0 2.00 1.00 1 9.50 MON LOAD FULL 17,031 5,045 11.00 2.00 3.00 0 1 0 2.50 1.00 1 10.00 MON LOAD FULL 17,031 5,045 11.00 2.00 3.00 0 1 0 3.00 1.00 1 10.50 MON DRIVE FULL 17,031 5,045 10.50 2.00 3.00 0 1 1 3.50 1.50 1 11.00 MON DRIVE FULL 17,031 5,045 10.00 2.00 3.00 0 1 1 4.00 2.00 1 11.50 MON DRIVE FULL 17,031 5,045 9.50 2.00 3.00 0 1 1 4.50 2.50 1 12.00 MON DRIVE FULL 17,031 5,045 9.00 2.00 3.00 0 1 1 5.00 3.00 1 12.50 MON DRIVE FULL 17,031 5,045 8.50 2.00 3.00 0 1 1 5.50 3.50

115

The 2005 rule resulted in small adverse financial impacts (reduced profits) on most carriers directly related to the magnitude of the drop in labor productivity. The results in terms of profit impacts relative to revenues seem to suggest very small impacts for firms across the wide range of size categories examined, including both large and small entities. The threshold for impacts considered to be of moderate size is generally taken to be 1 percent of revenues, and the average impacts of the rule fall well below that magnitude.

D.4.4.1 Core Cost Components A significant portion of the total cost is driver labor costs. Changes in the number of hours drivers can work or drive were translated to changes in driver’s labor productivities using the simulation model explained above. These changes were then used to calculate the additional number of drivers needed to achieve full compliance. Changes in the number of drivers were then translated into labor cost changes using the estimated wage-hours worked functional relationship for truck drivers used in the 2003 HOS RIA. The changes made in the 2005 HOS rules were also evaluated with respect to the same non-wage costs considered in the 2003 HOS RIA.

The results of the simulation model described above showed a 3.9 percent decrease in productivity for LH drivers (a 7.0 percent decrease for regional less a 3.1 percent productivity increase for LH OTR) using split sleeper berths as a result of the elimination of the sleep sleeper berth provision in the 2005 HOS rule. These percentages are calculated for the simulated drivers; to extrapolate to the real driver population, these figures were weighted according to the fraction of total VMT attributable to these drivers, resulting in a productivity loss of 0.042 percent (a 0.08 percent decrease for regional less a 0.038 percent productivity increase for LH OTR).

The 2003 HOS rule analysis found that a 3.9 percent increase in LH labor productivity from the 2003 rules could be valued at about $1 billion, or about $275 million per percentage point (referred to as the “unit cost”) in year 2000 dollars. The 2005 rule analysis updated this figure to $298 million per percentage point of productivity year in 2004 dollars using the GDP deflator. Converting the total cost changes to a unit cost number, as is done here, is possible because the 2003 HOS analysis showed that there was a linear relationship between changes in driver labor productivity and the associated costs. Table 75 presents the breakdown of LH unit costs.

116

Table 75. LH Unit Costs for HOS Rule Changes (2013 $)

Unit: Change in Labor Demand 1% Change in Number of Drivers 15,000

Driver Labor Cost $176

Avoided Labor Wages -$429

Avoided Labor Benefits -$26

New Labor Wages $482

New Labor Benefits $149

Other Costs $121

Non-driver Labor $7

Trucks $50

Parking $15

Insurance $11

Maintenance $19

Recruitment $20

Total $298

D.4.5 Total Costs of the 2005 HOS Rule Changes Table 76. Total Costs of the 2005 HOS Rule (2013 $)

Change in LH Productivity 0.042% Change in Annual Unit Labor Costs due to Productivity Impact=0.042*$298 $13

117

APPENDIX E: ROADSIDE INTERVENTION MODEL FMCSA has employed parameters from its Roadside Intervention Model129 (which was published April 2013) to estimate crash reductions from HOS violations that are avoided. In the model FMCSA directly measures the relationship between crashes and violations using roadside inspection, traffic enforcement, and safety data. This represents a major improvement in the Agency’s estimates of the safety benefits of ELD use. This appendix is a summary of that report. All sources and citations were accurate when the research report quoted in this appendix was originally published. FMCSA, however, cannot guarantee the current availability of data and publications from outside sources.

E.1 BACKGROUND

The Roadside Inspection and Traffic Enforcement Programs are two key Federal Motor Carrier Safety Administration (FMCSA) safety programs. The Roadside Inspection Program consists of roadside inspections performed by qualified safety inspectors following the guidelines of the North American Standard, developed by FMCSA and the Commercial Vehicle Safety Alliance (CVSA). Most roadside inspections are conducted by the States under the Motor Carrier Safety Assistance Program (MCSAP). This program has six levels of inspections, including a vehicle component, a driver component, or both. The Traffic Enforcement Program is composed of two distinct activities: a traffic stop as a result of a moving violation and a subsequent roadside inspection.

FMCSA developed an analytic model to measure the effectiveness of roadside inspections and traffic enforcements in terms of crashes avoided, injuries avoided, and lives saved. This model is known as the Roadside Intervention Effectiveness Model. In this model, traffic enforcements and roadside inspections are considered interventions.

The model is based on the premise that interventions resulting in the correction of vehicle and driver violations, specifically roadside inspections and traffic enforcements, contribute to a reduction in crashes. The model associates each violation of the FMCSRs and Federal Hazardous Material Regulations with a specific crash probability. Using these probabilities, the number of crashes avoided as a result of correcting these violations can be estimated.

Additionally, the Intervention Model provides FMCSA management with information to address the Government Performance and Results Act of 1993 (GPRA), which requires Federal agencies to measure the effectiveness of their programs as part of the budget cycle process. It also provides FMCSA and State safety program managers with a quantitative basis for optimizing the allocation of safety resources in the field.

129 FMCSA Safety Program Effectiveness Measurement: Intervention Model Fiscal Year 2009. April 2013.

118

E.2 METHODOLOGY

The Intervention Model is based on the premise that the Roadside Inspection and Traffic Enforcement Programs contribute to the reduction of crashes by discovering vehicle and/or driver violations during interventions (roadside inspections and traffic enforcements). When these violations are corrected as the result of interventions, it reduces the probability that the vehicles/drivers will be involved in subsequent crashes.

Since the occurrence of a single violation implies a certain degree of crash risk, each inspection that uncovers and corrects at least one violation can be interpreted as reducing crash risk. The model expresses this risk reduction in terms of the elimination of specific crash probabilities associated with each violation corrected. For an individual intervention, the reduction in crash risk depends on the number and type of violations found. By summing the crash risk probabilities for all violations corrected over all inspections, the model estimates the number of crashes avoided as a result of the Roadside Inspection and Traffic Enforcement Programs.

E.2.1 Input Data Selection One fiscal year (FY) (defined as October 1 of the previous year through September 30 of the FY referenced) of intervention data is extracted from the Motor Carrier Management Information System (MCMIS) database. This database contains roadside inspection information compiled from Federal and State safety agencies, including violations (if any) cited during interventions. While inspections are not required to have violations associated with them, in practice, about two-thirds of all interventions do find one or more violations. The violation data are the key component in the model, as they represent the defects identified and subsequently corrected as a result of the two programs.

E.2.2 Assignment of Crash Risk Reduction Probabilities The model assumes that observed deficiencies (i.e., violations) discovered at the time of an intervention can be converted into crash risk probabilities. This assumption is based on the premise that detected violations represent varying degrees of mechanical or judgmental faults and, further, that some are more likely than others to play a contributory role in motor carrier crashes.

An improved methodology was developed for determining the crash risk associated with violations in Intervention Model Version 3.0 and implemented in FY 2008. The improved methodology uses applicable results of related FMCSA research, including the Violation Severity Assessment Study (VSAS),130 as well as research performed for the Agency’s Compliance, Safety, Accountability (CSA) initiative. The revised methodology is based on sound safety data and statistical approaches, as well as input from subject matter experts when empirical data are not available.

130 Violations Severity Assessment Study and Carrier Safety Measurement System (CSMS), available in the docket: FMCSA-2004-18898-0210.

119

The Version 3.0 methodology introduced the concept of a violation group as developed by the CSA initiative.131 A violation group is defined as a set of violations similar in nature and having equal crash risks. The model assumes that correcting a violation associated with a particular violation group during an intervention reduces the risk of a subsequent crash by a finite amount, equal to the crash risk probability (CRP) associated with that group.

The model employs three estimates in developing the crash risk reduction probability for a violation group:

• The crash risk for violations in the group is defined as the likelihood that the unsafe behavior associated with the violation contributes to a crash during a commercial motor vehicle (CMV) daytrip. (A “daytrip” is defined as a CMV’s travel during 1 day.)

• The duration of the reduction in crash risk when a violation in the group is identified at the roadside and corrected. The duration of the risk reduction varies according to the violation group to which the violation is assigned.

• The correction rate for violations in the group that are corrected as a result of the intervention.

A preliminary crash risk reduction for a violation group is calculated from the product of the CRP and the violation group’s duration. The preliminary crash risk reduction is then multiplied by a violation correction rate to produce the final crash risk reduction for each violation in the violation group. The violation correction rate adjusts for the reality that not all violations are corrected within the required time period. Preliminary research indicates that only 69.9 percent of Vehicle Maintenance violations and 68.8 percent of Driver Fitness violations are corrected within the allotted time.132 The violation correction rate thus decreases the magnitude of the crash risk reduction used in the model to account for violations not corrected. For a more detailed discussion of crash risk, duration, and correction rates and their derivations, see Appendix A of the FY 2009 Intervention Model Report.

E.2.3 Calculation of Benefits To produce an estimate of the annual number of crashes avoided due to inspections, the model first determines the number of inspections for each violation group in which a violation was recorded during the FY. The inspection count is then multiplied by the final crash risk reduction associated with the violation group, yielding the estimate of annual crashes avoided. Lastly, the estimated crashes avoided are added up across all violation groups to produce an estimate of the total annual crashes avoided during the FY.

Once the number of crashes avoided is totaled for all inspections during the year, the model then computes the number of lives saved and injuries avoided as a result of those crashes avoided. Average numbers of fatalities per crash, injuries per crash, and injuries per fatal crash are computed using MCMIS data for all crashes in the United States for the year. These averages are

131 For more information about how the CSA initiative groups safety violations, see the Safety Measurement System Methodology at http://csa.fmcsa.dot.gov/Documents/SMSMethodology.pdf. Accessed February 4, 2014. 132 See the SMS Factsheet for descriptions and examples of Vehicle Maintenance and Driver Fitness violations: http://csa.fmcsa.dot.gov/Documents/SMS_factsheet.pdf. Accessed February 4, 2014.

120

then multiplied by the number of crashes avoided to estimate the number of lives saved and injuries avoided due to the inspections.

E.3 FY 2009 INTERVENTION MODEL RESULTS

Total crashes avoided, total lives saved, and total injuries avoided as a result of roadside inspection and traffic enforcement activities performed during FY 2009 were estimated by the Intervention Model. The results are presented at the national and State levels. Beginning in FY 2006, the Intervention Model was implemented to estimate benefits from roadside interventions by FY; previous years were implemented by calendar year (CY). As a result, estimates of benefits for years 2005 and earlier are shown by CY.

E.3.1 National Level Estimates Table 77 provides a breakdown of the program activity at the national level for the current analysis year (FY 2009) and the 2 years prior (FY 2007 and FY 2008). Program activity was higher in FY 2009 than in the 2 previous years. The number of interventions performed increased by about 1.1 percent from FY 2008, roadside inspections rose by 65,152 (2.4 percent), and traffic enforcements decreased by 25,253 (3.3 percent).

Table 77. Program Activity FY 2007–09

Interventions FY 2007 FY 2008 FY 2009

Roadside Inspections 2,616,868 2,723,576 2,788,728

Traffic Enforcements 752,649 756,169 730,916 Total 3,369,517 3,479,745 3,519,644

Table 78 presents the estimated benefits of the two programs over the past 3 years. The model estimates that the Roadside Inspection Program prevented 8,149 crashes in FY 2009, while the Traffic Enforcement Program prevented 8,789, for a total of 16,939 crashes avoided. The number of crashes avoided decreased from FY 2008 to FY 2009, even as the total number of interventions increased, because the proportion of inspections resulting in no violations also increased (from 32 percent to 34 percent). Because more roadside inspections found no violations, the average number of violations per inspection decreased from 2.14 in 2008 to 2.07 in 2009. Traffic enforcement interventions are an exception: they prevented more crashes per intervention in FY 2009 than in FY 2008, but prevented fewer crashes overall because the total number of these interventions decreased.

Table 78. Program Effectiveness FY 2007–09 Using Intervention Model 3.0

Estimated Intervention Benefits FY 2007 FY 2008 FY 2009

Crashes Avoided Due to Roadside Inspections 8,101 8,464 8,149

Crashes Avoided Due to Traffic Enforcements 8,769 9,053 8,789

Total Crashes Avoided 16,870 17,517 16,939

Injuries Avoided Due to Roadside Inspections 5,222 5,381 5,206

121

Estimated Intervention Benefits FY 2007 FY 2008 FY 2009

Injuries Avoided Due to Traffic Enforcements 5,652 5,755 5,615

Total Injuries Avoided 10,874 11,136 10,821

Lives Saved Due to Roadside Inspections 307 304 276

Lives Saved Due to Traffic Enforcements 332 325 297

Total Lives Saved 639 629 573

Figure 7 shows the trends in estimated crashes avoided and lives saved from CY 2004 to FY 2009.133 All estimates prior to FY 2009 were recalculated as necessary using the most recent Intervention Model (Version 3.0) to provide a historical time series compatible with FY 2009 estimates. In FY 2009, the number of lives saved decreased from the previous years, while the number of crashes avoided remained relatively unchanged. Complete Version 3.0 results from CY 2004 to FY 2009 are shown in Table 79.

Figure 7. Estimated Number of Crashes Avoided and Lives Saved Trends

133 For 2004–05, data are only available by CY; from 2006 onward, data are organized by FY.

122

Table 79. Intervention Model Version 3.0 Estimated Program Benefits, CY 2004–FY 2009

Estimated Intervention Results CY 2004 CY 2005 FY 2006 FY 2007 FY 2008 FY 2009

Number of Roadside Inspections 2,210,842 2,193,954 2,372,802 2,616,868 2,723,576 2,788,728 Number of Traffic Enforcements 802,798 826,951 900,260 752,649 756,169 730,916 Total Number of Interventions 3,013,640 3,020,905 3,273,062 3,369,517 3,479,745 3,519,644 Crashes Avoided Due to Roadside Inspections 7,353 7,575 7,593 8,101 8,464 8,149

Crashes Avoided Due to Traffic Enforcement 8,467 9,205 9,422 8,769 9,053 8,789

Total Crashes Avoided 15,820 16,780 17,015 16,870 17,517 16,938 Injuries Avoided Due to Roadside Inspections 5,362 5,252 5,090 5,222 5,381 5,206

Injuries Avoided Due to Traffic Enforcement 6,174 6,382 6,316 5,652 5,755 5,614

Total Injuries Avoided 11,535 11,634 11,405 10,874 11,136 10,820 Lives Saved Due to Roadside Inspections 284 282 287 307 304 275

Lives Saved Due to Traffic Enforcement 327 342 357 332 325 297

Total Lives Saved 611 624 644 639 629 572

E.3.2 State-Level Estimates The model’s flexibility makes it possible to examine the results with finer detail, such as benefits by reporting State or by carrier domicile State. State-level totals are presented by both reporting State and country of domicile (U.S. vs. non-U.S.).

E.3.2.1 Estimates by Country of Domicile (U.S. vs. Non-U.S.) This section summarizes a comparison between carriers domiciled in the United States and those carriers domiciled outside of the United States.

Table 80 shows the number of roadside inspections and traffic enforcements performed on U.S. and non-U.S. domiciled carriers during FY 2009.

Table 80. Program Exposure: U.S. Domiciled vs. Non-U.S. Domiciled Carriers, FY 2009

Interventions U.S. Domiciled Non-U.S. Domiciled Roadside Inspections 2,508,240 279,847 Traffic Enforcements 712,829 18,027 Total Interventions 3,221,069 297,874

Table 81 compares the effectiveness of interventions conducted in FY 2009 on carriers domiciled in the U.S. and non-U.S. domiciled carriers. Because the exposure for U.S. domiciled carriers is more than 10 times that for non-U.S. domiciled carriers, the U.S. carriers have many more inspections; as a result, their crashes avoided, injuries avoided, and lives saved are all much higher.

123

Table 81. Program Effectiveness: U.S. Domiciled vs. Non-U.S. Domiciled Carriers, FY 2009

Types of Benefits

Estimated Benefits:

U.S.

Estimated Benefits: Non-U.S.

Estimated Benefits per 1,000

Interventions: U.S.

Estimated Benefits per 1,000 Interventions:

Non-U.S. Crashes Avoided Due to Roadside Inspections 6,768 1,375 2.70 4.91

Crashes Avoided Due to Traffic Enforcements 8587 201 12.05 11.13

Total Crashes Avoided 15,355 1,576 4.77 5.29 Injuries Avoided Due to Roadside Inspections 4,324 878 1.72 3.14

Injuries Avoided Due to Traffic Enforcements 5486 128 7.70 7.11

Total Injuries Avoided 9,810 1,006 3.05 3.38 Lives Saved Due to Roadside Inspections 229 47 0.09 0.17

Lives Saved Due to Traffic Enforcements 290 7 0.41 0.37

Total Lives Saved 519 54 0.16 0.18

To provide a more interesting basis for comparison between U.S. and non-U.S. domiciled carriers, Table 81 also includes the estimated program benefits per 1,000 interventions. From that analysis, it is clear that non-U.S. carriers had a somewhat higher rate of crashes avoided as a result of roadside inspections (4.91 compared to 2.70 crashes avoided per 1,000 roadside inspections), while the rates of crashes avoided per traffic enforcement are fairly similar for U.S. and non-U.S. carriers (12.05 and 11.13 crashes avoided per 1,000 traffic enforcements). This observation suggests that similar numbers and kinds of violations were found during traffic enforcement interventions on U.S. and non-U.S. domiciled carriers, but a roadside inspection on a non-U.S. domiciled carrier was likely to find more violations or more severe violations than one on a U.S. domiciled carrier.

E.3.2.2 Estimates by Reporting State Table 82 provides roadside inspection results, and Table 83 provides traffic enforcement results, by reporting State (as well as for Federal staff), for interventions conducted in all 50 States, the District of Columbia, and the U.S. territories (American Samoa, the Northern Mariana Islands, and Puerto Rico). Both tables provide intervention counts and total estimated benefits (crashes avoided, injuries avoided, lives saved).

Since activity levels vary widely from State to State, these tables include the number of benefits provided per 1,000 interventions (per 1,000 roadside inspections in Table 82; per 1,000 traffic enforcements in Table 83). This analysis can illuminate interesting trends about the effectiveness of interventions in different States. For example, while Texas has both a large number of roadside inspections and a large number of crashes avoided as a result of those inspections, the State also has highly effective inspections: its 5.33 crashes avoided per 1,000 inspections is nearly twice the national average of 2.94. Roadside inspections and traffic enforcements performed by Federal staff are likewise highly effective, with 5.59 crashes avoided per 1,000

124

roadside inspections and 20.41 crashes avoided per 1,000 traffic enforcements (compared to the national average of 12.02).

125

Table 82. Roadside Inspection Program Estimated Benefits by Reporting State, FY 2009

Reporting State

Total Interventions

Initiated

Number of Roadside

Inspections

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives Saved

Est. Crashes Avoided

per 1,000 Inspections

Est. Injuries Avoided

per 1,000 Inspections

Est. Lives Saved per 1,000

Inspections Alabama 43,559 36,922 106.07 67.76 3.59 2.87 1.84 0.10 Alaska 9,926 8,626 19.14 12.23 0.65 2.22 1.42 0.08 Arizona 69,452 45,080 196.43 125.49 6.64 4.36 2.78 0.15 Arkansas 41,953 32,894 95.54 61.03 3.23 2.90 1.86 0.10 California 523,903 458,312 541.45 345.90 18.31 1.18 0.75 0.04 Colorado 56,458 44,782 175.07 111.84 5.92 3.91 2.50 0.13 Connecticut 18,796 12,403 63.92 40.83 2.16 5.15 3.29 0.17 Delaware 5,010 2,922 7.99 5.10 0.27 2.73 1.75 0.09 District of Columbia 6,725 4,746 4.49 2.87 0.15 0.95 0.60 0.03 Federal 126,587 124,637 697.32 445.47 23.59 5.59 3.57 0.19 Florida 101,735 81,053 199.25 127.29 6.74 2.46 1.57 0.08 Georgia 99,232 76,775 304.05 194.24 10.28 3.96 2.53 0.13 Hawaii 5,440 4,657 7.20 4.60 0.24 1.55 0.99 0.05 Idaho 12,903 6,155 24.79 15.84 0.84 4.03 2.57 0.14 Illinois 77,774 52,967 152.51 97.43 5.16 2.88 1.84 0.10 Indiana 95,524 39,953 144.96 92.61 4.90 3.63 2.32 0.12 Iowa 57,602 37,948 150.02 95.84 5.07 3.95 2.53 0.13 Kansas 52,303 43,293 106.04 67.74 3.59 2.45 1.56 0.08 Kentucky 87,760 61,680 111.46 71.20 3.77 1.81 1.15 0.06 Louisiana 50,929 33,530 130.63 83.45 4.42 3.90 2.49 0.13 Maine 23,814 20,403 55.00 35.14 1.86 2.70 1.72 0.09 Maryland 106,630 87,534 190.80 121.89 6.45 2.18 1.39 0.07 Massachusetts 18,477 8,157 25.06 16.01 0.85 3.07 1.96 0.10 Michigan 71,922 39,406 131.25 83.85 4.44 3.33 2.13 0.11 Minnesota 42,354 25,557 81.33 51.96 2.75 3.18 2.03 0.11 Mississippi 42,114 40,759 83.01 53.03 2.81 2.04 1.30 0.07 Missouri 85,001 58,144 152.63 97.51 5.16 2.63 1.68 0.09 Montana 45,152 41,469 81.17 51.85 2.75 1.96 1.25 0.07

126

Reporting State

Total Interventions

Initiated

Number of Roadside

Inspections

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives Saved

Est. Crashes Avoided

per 1,000 Inspections

Est. Injuries Avoided

per 1,000 Inspections

Est. Lives Saved per 1,000

Inspections Nebraska 37,255 28,013 81.22 51.89 2.75 2.90 1.85 0.10 Nevada 32,097 23,536 64.62 41.28 2.19 2.75 1.75 0.09 New Hampshire 12,085 9,542 26.73 17.08 0.90 2.80 1.79 0.09 New Jersey 42,655 33,883 98.41 62.87 3.33 2.90 1.86 0.10 New Mexico 121,636 91,090 143.30 91.54 4.85 1.57 1.00 0.05 New York 111,725 97,587 238.60 152.43 8.07 2.44 1.56 0.08 North Carolina 90,610 73,020 124.18 79.33 4.20 1.70 1.09 0.06 North Dakota 14,706 13,210 19.42 12.41 0.66 1.47 0.94 0.05 Ohio 85,586 69,145 160.62 102.61 5.43 2.32 1.48 0.08 Oklahoma 29,562 18,734 56.21 35.91 1.90 3.00 1.92 0.10 Oregon 56,288 45,660 136.30 87.07 4.61 2.99 1.91 0.10 Pennsylvania 81,478 64,665 162.41 103.75 5.49 2.51 1.60 0.08 Rhode Island 2,662 1,613 7.16 4.57 0.24 4.44 2.83 0.15 South Carolina 52,261 39,035 124.62 79.61 4.22 3.19 2.04 0.11 South Dakota 29,212 24,480 61.46 39.26 2.08 2.51 1.60 0.08 Tennessee 69,586 51,346 82.12 52.46 2.78 1.60 1.02 0.05 Texas 370,667 356,375 1,900.54 1,214.13 64.28 5.33 3.41 0.18 Utah 40,642 32,700 103.85 66.34 3.51 3.18 2.03 0.11 Vermont 9,575 7,239 24.61 15.72 0.83 3.40 2.17 0.11 Virginia 43,711 32,497 97.34 62.18 3.29 3.00 1.91 0.10 Washington 108,113 77,492 191.85 122.56 6.49 2.48 1.58 0.08 West Virginia 33,735 23,099 43.11 27.54 1.46 1.87 1.19 0.06 Wisconsin 36,204 22,308 97.27 62.14 3.29 4.36 2.79 0.15 Wyoming 19,259 14,264 46.01 29.39 1.56 3.23 2.06 0.11 U.S. Territories 9,299 7,431 18.76 11.98 0.63 10.75 2.52 1.61 Total 3,510,345 2,781,297 8,130.54 5,194.07 275.00 2.94 1.87 0.10

127

Table 83. Traffic Enforcement Program Estimated Benefits by Reporting State, FY 2009

Reporting State Total Initiating Interventions

Number Traffic Enforcements

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives

Saved

Est. Crashes Avoided

per 1,000 Enforcements

Est. Injuries Avoided

per 1,000 Enforcements

Est. Lives Saved per 1,000

Enforcements Alabama 43,559 6,637 82.04 52.41 2.77 12.36 7.90 0.42 Alaska 9,926 1,300 16.74 10.69 0.57 12.88 8.22 0.44 Arizona 69,452 24,372 345.46 220.69 11.68 14.17 9.06 0.48 Arkansas 41,953 9,059 110.93 70.87 3.75 12.25 7.82 0.41 California 523,903 65,591 485.67 310.26 16.43 7.40 4.73 0.25 Colorado 56,458 11,676 126.30 80.68 4.27 10.82 6.91 0.37 Connecticut 18,796 6,393 95.02 60.70 3.21 14.86 9.49 0.50 Delaware 5,010 2,088 19.08 12.19 0.65 9.14 5.84 0.31 District of Columbia 6,725 1,979 29.63 18.93 1.00 14.97 9.57 0.51 Federal 126,587 1,950 39.79 25.42 1.35 20.41 13.04 0.69 Florida 101,735 20,682 256.47 163.84 8.67 12.40 7.92 0.42 Georgia 99,232 22,457 317.29 202.70 10.73 14.13 9.03 0.48 Hawaii 5,440 783 7.09 4.53 0.24 9.05 5.79 0.31 Idaho 12,903 6,748 80.80 51.62 2.73 11.97 7.65 0.40 Illinois 77,774 24,807 241.62 154.36 8.17 9.74 6.22 0.33 Indiana 95,524 55,571 555.55 354.90 18.79 10.00 6.39 0.34 Iowa 57,602 19,654 315.28 201.41 10.66 16.04 10.25 0.54 Kansas 52,303 9,010 130.24 83.20 4.41 14.46 9.23 0.49 Kentucky 87,760 26,080 358.52 229.03 12.13 13.75 8.78 0.47 Louisiana 50,929 17,399 167.09 106.74 5.65 9.60 6.13 0.32 Maine 23,814 3,411 44.25 28.27 1.50 12.97 8.29 0.44 Maryland 106,630 19,096 254.79 162.77 8.62 13.34 8.52 0.45 Massachusetts 18,477 10,320 136.90 87.46 4.63 13.27 8.47 0.45 Michigan 71,922 32,516 393.92 251.65 13.32 12.11 7.74 0.41 Minnesota 42,354 16,797 294.35 188.04 9.96 17.52 11.19 0.59 Mississippi 42,114 1,355 18.98 12.13 0.64 14.01 8.95 0.47 Missouri 85,001 26,857 401.21 256.31 13.57 14.94 9.54 0.51 Montana 45,152 3,683 45.80 29.26 1.55 12.44 7.94 0.42

128

Reporting State Total Initiating Interventions

Number Traffic Enforcements

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives

Saved

Est. Crashes Avoided

per 1,000 Enforcements

Est. Injuries Avoided

per 1,000 Enforcements

Est. Lives Saved per 1,000

Enforcements Nebraska 37,255 9,242 77.52 49.52 2.62 8.39 5.36 0.28 Nevada 32,097 8,561 89.37 57.09 3.02 10.44 6.67 0.35 New Hampshire 12,085 2,543 40.45 25.84 1.37 15.91 10.16 0.54 New Jersey 42,655 8,772 112.27 71.72 3.80 12.80 8.18 0.43 New Mexico 121,636 30,546 357.12 228.14 12.08 11.69 7.47 0.40 New York 111,725 14,138 173.41 110.78 5.87 12.27 7.84 0.42 North Carolina 90,610 17,590 287.48 183.65 9.72 16.34 10.44 0.55 North Dakota 14,706 1,496 16.51 10.55 0.56 11.04 7.05 0.37 Ohio 85,586 16,441 155.41 99.28 5.26 9.45 6.04 0.32 Oklahoma 29,562 10,828 127.11 81.20 4.30 11.74 7.50 0.40 Oregon 56,288 10,628 116.42 74.37 3.94 10.95 7.00 0.37 Pennsylvania 81,478 16,813 221.05 141.21 7.48 13.15 8.40 0.44 Rhode Island 2,662 1,049 16.39 10.47 0.55 15.62 9.98 0.52 South Carolina 52,261 13,226 185.96 118.80 6.29 14.06 8.98 0.48 South Dakota 29,212 4,732 65.53 41.86 2.22 13.85 8.85 0.47 Tennessee 69,586 18,240 206.99 132.23 7.00 11.35 7.25 0.38 Texas 370,667 14,292 143.35 91.58 4.85 10.03 6.41 0.34 Utah 40,642 7,942 119.08 76.07 4.03 14.99 9.58 0.51 Vermont 9,575 2,336 29.98 19.15 1.01 12.83 8.20 0.43 Virginia 43,711 11,214 145.48 92.94 4.92 12.97 8.29 0.44 Washington 108,113 30,621 381.37 243.63 12.90 12.45 7.96 0.42 West Virginia 33,735 10,636 97.11 62.04 3.28 9.13 5.83 0.31 Wisconsin 36,204 13,896 156.19 99.78 5.28 11.24 7.18 0.38 Wyoming 19,259 4,995 66.97 42.78 2.27 13.41 8.56 0.45 U.S. Territories 9,299 1,868 29.88 19.10 1.00 16.00 10.22 0.54 Total 3,519,644 730,916 8789.21 5614.84 297.27 12.02 7.68 0.41

129

E.3.2.3 Estimates by Carrier State of Domicile Table 84 and Table 85 provide detailed roadside inspections and traffic enforcement results, respectively, organized by carrier domicile State for interventions conducted on carriers registered in all 50 States, the District of Columbia, and U.S. territories, as well as Canada, Mexico, and other countries. The number of benefits provided per 1,000 interventions (per 1,000 roadside inspections in Table 84; per 1,000 traffic enforcements in Table 85) is again included to provide a scale for comparison between States with different levels of activity.

Approximately 10 times as many roadside inspections were performed on carriers domiciled in Nebraska (41,075) as carriers domiciled in Delaware (4,104), but the benefits per 1,000 inspections performed on carriers from each State were similar: 2.61 crashes and 1.67 injuries were avoided, and 0.09 lives were saved for every 1,000 inspections on Nebraska-domiciled carriers; for Delaware-domiciled carriers, these numbers were 2.63 crashes and 1.68 injuries avoided, and 0.09 lives saved.

It is also possible, in Table 84 and Table 85, to see more details of the effectiveness of roadside inspections and traffic enforcements on non-U.S. domiciled carriers. Table 85 confirms that the effectiveness of traffic enforcements is similar for U.S. and non-U.S. domiciled carriers. While the average for U.S. domiciled carriers was 12.05, carriers domiciled in Canada had 10.86 crashes avoided per 1,000 enforcements, and Mexico domiciled carriers had 11.84 crashes avoided per 1,000 enforcements. Non-North American carriers did have a noticeably higher rate of 17.09 crashes avoided per 1,000 enforcements. Table 84, on the other hand, shows some interesting differences in the results of roadside inspections on carriers domiciled in various non-U.S. countries. Canada has 10.52 crashes avoided per 1,000 inspections, and other non-North America countries average 6.10 crashes avoided per 1,000 inspections, which are greater than the U.S. domiciled average of 4.91 crashes avoided per 1,000 inspections. Mexico, on the other hand, has a much lower average number of crashes avoided per 1,000 inspections: 2.54.

130

Table 84. Roadside Inspection Program Estimated Benefits by Domicile State and Country, FY 2009

Carrier State Total Initiating Interventions

Number Roadside

Inspections

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives Saved

Est. Crashes Avoided

per 1,000 Inspections

Est. Injuries Avoided

per 1,000 Inspections

Est. Lives Saved

per 1,000 Inspections

Alabama 54,479 43,754 126.14 80.58 4.27 2.88 1.84 0.10 Alaska 7,845 6,694 16.45 10.51 0.56 2.46 1.57 0.08 Arizona 90,526 71,728 165.44 105.69 5.60 2.31 1.47 0.08 Arkansas 60,419 47,069 115.40 73.72 3.90 2.45 1.57 0.08 California 436,014 362,682 549.41 350.98 18.58 1.51 0.97 0.05 Colorado 44,902 34,485 115.37 73.70 3.90 3.35 2.14 0.11 Connecticut 12,325 8,513 28.24 18.04 0.96 3.32 2.12 0.11 Delaware 5,413 4,104 10.79 6.89 0.36 2.63 1.68 0.09 District of Columbia 1,533 1,203 1.91 1.22 0.06 1.59 1.01 0.05 Florida 147,447 115,783 338.32 216.13 11.44 2.92 1.87 0.10 Georgia 100,395 75,379 256.13 163.62 8.66 3.40 2.17 0.11 Hawaii 4,837 4,131 6.82 4.36 0.23 1.65 1.06 0.06 Idaho 18,896 14,040 42.28 27.01 1.43 3.01 1.92 0.10 Illinois 127,303 91,703 252.27 161.16 8.53 2.75 1.76 0.09 Indiana 91,744 62,921 175.05 111.83 5.92 2.78 1.78 0.09 Iowa 69,335 49,615 125.31 80.05 4.24 2.53 1.61 0.09 Kansas 40,915 30,231 89.81 57.37 3.04 2.97 1.90 0.10 Kentucky 53,560 37,706 86.92 55.53 2.94 2.31 1.47 0.08 Louisiana 38,716 29,357 119.94 76.62 4.06 4.09 2.61 0.14 Maine 14,255 11,453 30.00 19.17 1.01 2.62 1.67 0.09 Maryland 53,874 43,218 94.57 60.41 3.20 2.19 1.40 0.07 Massachusetts 27,698 18,198 52.95 33.83 1.79 2.91 1.86 0.10 Michigan 85,465 57,475 177.17 113.18 5.99 3.08 1.97 0.10 Minnesota 74,971 51,594 121.72 77.76 4.12 2.36 1.51 0.08 Mississippi 28,290 22,655 66.92 42.75 2.26 2.95 1.89 0.10 Missouri 84,484 62,228 146.90 93.84 4.97 2.36 1.51 0.08 Montana 18,884 16,060 38.23 24.42 1.29 2.38 1.52 0.08 Nebraska 54,580 41,075 107.06 68.39 3.62 2.61 1.67 0.09 Nevada 14,524 11,866 35.16 22.46 1.19 2.96 1.89 0.10

131

Carrier State Total Initiating Interventions

Number Roadside

Inspections

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives Saved

Est. Crashes Avoided

per 1,000 Inspections

Est. Injuries Avoided

per 1,000 Inspections

Est. Lives Saved

per 1,000 Inspections

New Hampshire 9,474 7,007 21.14 13.50 0.72 3.02 1.93 0.10 New Jersey 65,535 51,404 144.52 92.32 4.89 2.81 1.80 0.10 New Mexico 22,776 17,282 43.33 27.68 1.47 2.51 1.60 0.09 New York 75,357 60,621 179.53 114.69 6.07 2.96 1.89 0.10 North Carolina 88,890 68,228 154.25 98.54 5.22 2.26 1.44 0.08 North Dakota 13,390 10,796 26.56 16.97 0.90 2.46 1.57 0.08 Ohio 103,587 79,680 181.55 115.98 6.14 2.28 1.46 0.08 Oklahoma 43,109 32,391 109.65 70.05 3.71 3.39 2.16 0.11 Oregon 45,251 36,576 86.81 55.46 2.94 2.37 1.52 0.08 Pennsylvania 112,806 89,043 198.38 126.73 6.71 2.23 1.42 0.08 Rhode Island 3,846 2,574 9.60 6.13 0.32 3.73 2.38 0.12 South Carolina 44,350 33,457 113.93 72.78 3.85 3.41 2.18 0.12 South Dakota 15,574 11,670 33.94 21.68 1.15 2.91 1.86 0.10 Tennessee 88,541 67,428 148.65 94.96 5.03 2.20 1.41 0.07 Texas 329,289 292,716 1,266.55 809.12 42.84 4.33 2.76 0.15 Utah 47,523 36,258 95.59 61.07 3.23 2.64 1.68 0.09 Vermont 5,163 4,100 12.50 7.99 0.42 3.05 1.95 0.10 Virginia 44,966 33,719 85.75 54.78 2.90 2.54 1.62 0.09 Washington 84,145 64,094 160.57 102.58 5.43 2.51 1.60 0.08 West Virginia 20,267 15,005 31.60 20.19 1.07 2.11 1.35 0.07 Wisconsin 77,804 54,957 136.46 87.18 4.62 2.48 1.59 0.08 Wyoming 6,618 4,987 16.32 10.43 0.55 3.27 2.09 0.11 U.S. Territories 9,179 7,327 18.6 11.88 0.63 8.86 5.66 0.30 Canada 91,239 77,562 138.05 88.18 4.68 10.52 6.72 0.35 Mexico 205,566 201,350 1227.26 784.02 41.49 2.54 1.62 0.09 Non-North America 1,069 935 9.84 6.28 0.33 6.10 3.89 0.21 Unknown 701 641 5.68 3.63 0.19 1.78 1.14 0.06 Total 3,457,320 2,788,728 8149.29 5,206.02 275.62 2.92 1.87 0.10

132

Table 85. Traffic Enforcement Program Estimated Benefits by Domicile State and Country, FY 2009

Carrier State Total Initiating Interventions

Number Traffic Enforcements

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives

Saved

Est. Crashes Avoided

per 1,000 Enforcements

Est. Injuries Avoided

per 1,000 Enforcements

Est. Lives Saved per 1,000

Enforcements Alabama 54,479 10,725 127.98 81.76 4.33 11.93 7.62 0.40 Alaska 7,845 1,151 15.16 9.68 0.51 13.17 8.41 0.44 Arizona 90,526 18,798 234.05 149.52 7.92 12.45 7.95 0.42 Arkansas 60,419 13,350 152.79 97.61 5.17 11.44 7.31 0.39 California 436,014 73,332 703.63 449.50 23.80 9.60 6.13 0.32 Colorado 44,902 10,417 123.33 78.79 4.17 11.84 7.56 0.40 Connecticut 12,325 3,812 51.18 32.70 1.73 13.43 8.58 0.45 Delaware 5,413 1,309 18.14 11.59 0.61 13.86 8.85 0.47 District of Columbia 1,533 330 4.46 2.85 0.15 13.52 8.64 0.45 Florida 147,447 31,664 403.95 258.06 13.66 12.76 8.15 0.43 Georgia 100,395 25,016 339.02 216.58 11.47 13.55 8.66 0.46 Hawaii 4,837 706 6.51 4.16 0.22 9.22 5.89 0.31 Idaho 18,896 4,856 65.02 41.54 2.20 13.39 8.55 0.45 Illinois 127,303 35,600 417.70 266.84 14.13 11.73 7.50 0.40 Indiana 91,744 28,823 328.98 210.16 11.13 11.41 7.29 0.39 Iowa 69,335 19,720 256.07 163.59 8.66 12.99 8.30 0.44 Kansas 40,915 10,684 133.67 85.39 4.52 12.51 7.99 0.42 Kentucky 53,560 15,854 202.45 129.33 6.85 12.77 8.16 0.43 Louisiana 38,716 9,359 107.65 68.77 3.64 11.50 7.35 0.39 Maine 14,255 2,802 32.21 20.58 1.09 11.50 7.34 0.39 Maryland 53,874 10,656 140.34 89.65 4.75 13.17 8.41 0.45 Massachusetts 27,698 9,500 139.32 89.00 4.71 14.67 9.37 0.50 Michigan 85,465 27,990 340.01 217.21 11.50 12.15 7.76 0.41 Minnesota 74,971 23,377 331.11 211.52 11.20 14.16 9.05 0.48 Mississippi 28,290 5,635 66.32 42.37 2.24 11.77 7.52 0.40 Missouri 84,484 22,256 273.69 174.84 9.26 12.30 7.86 0.42 Montana 18,884 2,824 35.15 22.46 1.19 12.45 7.95 0.42 Nebraska 54,580 13,505 133.33 85.18 4.51 9.87 6.31 0.33 Nevada 14,524 2,658 31.82 20.33 1.08 11.97 7.65 0.41

133

Carrier State Total Initiating Interventions

Number Traffic Enforcements

Est. Crashes Avoided

Est. Injuries Avoided

Est. Lives

Saved

Est. Crashes Avoided

per 1,000 Enforcements

Est. Injuries Avoided

per 1,000 Enforcements

Est. Lives Saved per 1,000

Enforcements New Hampshire 9,474 2,467 37.52 23.97 1.27 15.21 9.72 0.51 New Jersey 65,535 14,131 177.25 113.23 6.00 12.54 8.01 0.42 New Mexico 22,776 5,494 64.00 40.89 2.16 11.65 7.44 0.39 New York 75,357 14,736 183.94 117.51 6.22 12.48 7.97 0.42 North Carolina 88,890 20,662 296.27 189.27 10.02 14.34 9.16 0.48 North Dakota 13,390 2,594 34.60 22.10 1.17 13.34 8.52 0.45 Ohio 103,587 23,907 261.94 167.34 8.86 10.96 7.00 0.37 Oklahoma 43,109 10,718 142.19 90.84 4.81 13.27 8.48 0.45 Oregon 45,251 8,675 100.36 64.11 3.39 11.57 7.39 0.39 Pennsylvania 112,806 23,763 272.31 173.96 9.21 11.46 7.32 0.39 Rhode Island 3,846 1,272 18.47 11.80 0.62 14.52 9.28 0.49 South Carolina 44,350 10,893 160.19 102.33 5.42 14.71 9.39 0.50 South Dakota 15,574 3,904 51.13 32.66 1.73 13.10 8.37 0.44 Tennessee 88,541 21,113 246.50 157.47 8.34 11.68 7.46 0.40 Texas 329,289 36,573 412.75 263.68 13.96 11.29 7.21 0.38 Utah 47,523 11,265 143.25 91.51 4.85 12.72 8.12 0.43 Vermont 5,163 1,063 13.89 8.87 0.47 13.07 8.34 0.44 Virginia 44,966 11,247 142.67 91.14 4.83 12.69 8.10 0.43 Washington 84,145 20,051 247.43 158.07 8.37 12.34 7.88 0.42 West Virginia 20,267 5,262 59.01 37.70 2.00 11.21 7.16 0.38 Wisconsin 77,804 22,847 254.83 162.79 8.62 11.15 7.13 0.38 Wyoming 6,618 1,631 22.07 14.10 0.75 13.53 8.65 0.46 U.S. Territories 9,179 1,852 29.78 19.03 1.00 16.08 10.28 0.54 Canada 91,239 13,677 148.50 94.85 5.01 10.86 6.94 0.37 Mexico 205,566 4,216 49.92 31.88 1.67 11.84 7.56 0.40 Non-North America 1,069 134 2.29 1.47 0.07 17.09 10.97 0.52 Unknown 701 60 1.14 0.73 0.04 19.00 12.17 0.67 Total 3,519,644 730,916 8,789.24 5,614.86 297.26 12.02 7.68 0.41

134

E.4 CONCLUSION

The Roadside Inspection and Traffic Enforcement Programs are two of FMCSA’s most powerful safety tools. By continually examining the results of these programs, FMCSA can ensure that they are executed effectively and are producing the desired safety benefits. Results for individual States can be examined and compared to provide guidance on how to allocate safety resources. The total national results show the scale of Roadside Inspection and Traffic Enforcement Programs and the magnitude of their effects on highway safety: in 2009, 2,781,297 roadside inspections and 730,916 traffic enforcements were conducted. Together, it is estimated that in FY 2009, these interventions saved approximately 570 lives and prevented 10,800 injuries by averting almost 17,000 crashes. Over the past 9 years, it is estimated that these two programs have saved almost 6,000 lives (Table 86).

135

Table 86. Historical Results for Intervention Model, CY 2001–05 and FY 2006–09

Intervention Results CY 2001 CY 2002 CY 2003 CY 2004 CY 2005 FY 2006 FY 2007 FY 2008 FY 2009 Number of Roadside Inspections 2,050,786 2,253,070 2,215,669 2,210,842 2,193,954 2,372,802 2,616,868 2,723,576 2,788,728 Number of Traffic Enforcements 695,619 760,094 791,116 802,798 826,951 900,260 752,649 756,169 730,916 Total Number of Interventions 2,746,405 3,013,164 3,006,785 3,013,640 3,020,905 3,273,062 3,369,517 3,479,745 3,519,644 Crashes Avoided Due to Roadside Inspections

6,658 7,218 7,176 7,353 7,575 7,593 8,101 8,464 8,149

Crashes Avoided Due to Traffic Enforcement

7,263 8,115 8,251 8,467 9,205 9,422 8,769 9,053 8,789

Estimated Total Crashes Avoided 13,921 15,333 15,427 15,820 16,780 17,015 16,870 17,517 16,938 Injuries Avoided Due to Roadside Inspections

5,050 5,458 5,456 5,362 5,252 5,090 5,222 5,381 5,206

Injuries Avoided Due to Traffic Enforcement 5,509 6,136 6,274 6,174 6,382 6,316 5,652 5,755 5,614 Estimated Total Injuries Avoided 10,559 11,594 11,730 11,535 11,634 11,405 10,874 11,136 10,820 Lives Saved Due to Roadside Inspections 331 346 317 284 282 287 307 304 275 Lives Saved Due to Traffic Enforcement 361 389 364 327 342 357 332 325 297 Estimated Total Lives Saved 691 735 681 611 624 644 639 629 572

136

E.5 IMPROVEMENTS TO THE INTERVENTION MODEL METHODOLOGY

E.5.1 Development of the Intervention Model The Intervention Model is part of FMCSA’s efforts to address requirements of the GPRA, which requires Federal agencies to measure the effectiveness of their programs as part of the budget cycle process. Work on FMCSA Program Performance Measures resulted in an initial model for assessing the effectiveness of roadside inspections. After a review panel made recommendations for improvement in a 1998 Volpe Center report entitled “OMC Safety Program Performance Measures,” the initial model evolved into Intervention Model 1.0, which produced estimates of crashes and injuries avoided and lives saved due to the Roadside Intervention and Traffic Enforcement programs for the years 1998, 1999, and 2000. Subsequently, several improvements to the 2001–07 model yielded Version 2.0, which was used to calculate benefits for CY 2001–05 and FY 2006–07. For FY 2008, Intervention Model 3.0 was developed in an effort to improve the methodology in several areas as described below.

E.5.2 Improvements to the Intervention Model Ten years of experience with the Version 2.0 methodology revealed the need for updating the model. Intervention Model 3.0 includes improvements in four areas. The improvements are data-driven modifications, based on empirical data that provide a more realistic basis for estimating the effects of the intervention program:

• Addressed the fact that not all violations recorded during interventions are actually corrected.

• Eliminated the multiplicative factor that was applied when multiple violations were identified during an inspection.

• Implemented a new way of determining the crash risks associated with violations. • Eliminated the calculation of indirect effects.134

E.5.3 Correction of Violations Intervention Model 2.0 assumed that all violations cited during an intervention were corrected either before the driver resumed operating the vehicle or shortly after the completion of the daytrip in which the intervention occurred. Therefore, Intervention Model 2.0 did not consider the possibility that noncorrection of violations might lessen the crash risk reduction resulting from recording violations during interventions. 134 Indirect effects estimated in Intervention Model 2.0 were the byproducts of the carriers’ increased awareness of FMCSA programs and their potential consequences if steps were not taken to ensure and/or maintain higher levels of safety. Indirect effects were essentially changes in carriers’ safety behavior during the year following their exposure to the interventions. However, re-analysis of past years’ data revealed that indirect effects probably were only a small fraction of direct effects, and that estimates of indirect effects by Version 2.0 were overstated. With the introduction of CSA and the revision of intervention program effectiveness models, it was determined that Intervention Model 3.0 would estimate only the immediate (direct) effects of Roadside Inspections and Traffic Enforcements, leaving the estimation of longer-term (indirect) effects to other CSA effectiveness models.

137

Based on a study examining vehicles undergoing a second inspection within 7 days of the first, correction rates were determined for vehicle maintenance and driver fitness violations. The study did not support the assumption that 100 percent of these violations are corrected within the regulatory time period. On average only 69.94 percent of vehicle maintenance-related violations and 68.82 percent of driver fitness-related violations were corrected within the allotted time. Table 87 and Table 88 show the correction rates that are used in Intervention Model 3.0. In the Version 3.0 methodology, the overall estimates of the effectiveness of roadside inspections and traffic enforcements are multiplied by these factors to account for violations that are not corrected.

Table 87. Vehicle Maintenance Violation Correction Rates in Version 3.0

Violation Group ID Violation Group Description Correction Rate (%)

18 Brakes Out of Adjustment 70.30

19 Brakes, All Others 79.36

20 Coupling Devices 92.76

21 Exhaust Discharge 81.72

22 Fuel Systems 91.62

23 Lighting 60.81

24 Steering Mechanism 81.82

25 Suspension 88.77

26 Tires 66.56

27 Wheels, Studs, Clamps, Etc. 71.37

28 Windshield/Glass/Markings 73.16

29 Cab, Body, Frame 90.69

30 Inspection Reports 69.93

31 Vehicle Jumping OOS 95.45

32 Other Vehicle Defect 64.63

33 Emergency Equipment 74.49

34 Tire vs. Load 92.73

46 Clearance Identification Lamps/Other 57.18

Total All Violations 69.94

138

Table 88. Driver Fitness Violation Correction Rates in Version 3.0

Violation Group ID Violation Group Description Correction Rate (%)

11 Driver Qualification 70.92

12 Endorsements & Vehicle Group 85.04

13 Medical Certificate 63.67

14 Physical 92.79

15 Multiple License 92.86

48 Fitness Jumping OOS 100.00

Total All Violations 68.82

E.5.4 Multiple Violations per Intervention Intervention Model 2.0 used a multiplicative factor to augment avoided risk when multiple violations at the same risk level occurred in an inspection. For example, if a roadside inspection recorded three violations considered to be of equal risk, the model multiplied the CRP of each violation by a factor of three, and then totaled the augmented risks to produce the total risk avoided. If the three violations in the inspection did not have equal risk levels, then the multiplicative factor was not applied.

An analysis was conducted to investigate whether multiple violations occurred more frequently in post-crash inspections than in non-crash inspections. It found that although the distribution of the number of violations per inspection in post-crash inspections was slightly skewed toward higher numbers of violations than in non-crash inspections, the differences were minimal. Figure 8 shows the two distributions.

139

Figure 8. Distributions of Multiple Violations in Post-Crash and Non-Crash Inspections

This analysis does not justify the multiplicative factor used by Intervention Model 2.0 for multiple violations in the same risk category. Inspections with two to four violations are more likely to be non-crash inspections than post-crash inspections. The occurrence of two to four violations in an inspection does not indicate that it is more likely to be a post-crash inspection than not. (The mean number of violations per inspection is 2.3, and more than 85 percent of inspections have four or fewer violations.) A multiplicative factor designed to augment estimated crash risk would be inappropriate when four or fewer violations indicate no additional crash risk. Furthermore, inspections with five or more violations are only slightly more likely to be post-crash, so a multiplicative factor would be small and only appropriate for 15 percent of inspections. Thus, it was decided to eliminate the multiplicative factor from Intervention Model 3.0.

E.5.5 Crash Risk Reduction Associated with Violations Intervention Model 2.0 assumed that recording a violation in a roadside inspection or traffic enforcement reduced the risk of a subsequent crash by a finite probability associated with the inherent potential of that violation to contribute to a CMV crash. The violations were grouped into five risk categories, each with a CRP representing the probability or likelihood that a violation would lead to a crash if left uncorrected. The relative weights were based on a 1998 study, Risk-based Evaluation of Commercial Motor Vehicle Roadside Violations: Process and Results.135 Industry experts and the author’s analysts converted these relative weights into crash risk reduction probabilities for use in the Intervention Model.

An improved method was developed for determining the crash risk reduction associated with violations in Intervention Model 3.0. The improved methodology uses FMCSA research, including the VSAS, as well as research performed for the CSA initiative. The 135 Cycla Corporation. Risk-based Evaluation of Commercial Motor Vehicle Roadside Violations: Process and Results. July 1998. http://permanent.access.gpo.gov/gpo6567/tb99-005.pdf. Accessed February 6, 2014.

140

underlying revised methodology is based on sound safety data and statistical approaches, relying to the minimum degree possible on expert opinion and assumptions for data not readily available in useable form.

The Version 3.0 methodology assumes that recording a violation associated with a particular violation group during an intervention reduces the risk of a subsequent crash by a finite probability equal to the inherent potential of that violation group to contribute to a CMV crash, based on empirical data. The concept of a BASIC violation group was first developed by the VSAS. A violation group consists of similar violations thought to have equal crash risk. The violation groups are listed in Table 89.

Table 89. Duration of Crash Risk Reduction by Violation Type

Violation Type Duration

Unsafe Driving Fatigued Driving Improper Loading

30 days, based on several studies suggesting that a citation deters the behavior for 30–90 days.

Controlled Substance and Alcohol

90 days, length of license suspension for cited driver, assumes sober driver would replace him.

Vehicle Maintenance Observable violation: 7 days. Non-observable violation: 37 days, recommended inspection interval by several vehicle maintenance organizations.

Driver Fitness 45 days, mean time between CVSA inspections, assumes driver could have violation immediately following CVSA inspection until the exemption period is up.

The model employs three estimates in developing the crash risk reduction probability for a violation group:

• The crash risk for violations in the group is defined as the likelihood that the

unsafe behavior associated with the violation will contribute to a crash during a CMV daytrip (a “daytrip” is defined as a CMV’s travel during 1 day). The crash risks for each violation group are calculated using data from the VSAS and CSA research, as well as the MCMIS database.

• The duration of the reduction in crash risk when a violation in the group is identified at the roadside and corrected. The duration of the risk reduction varies according to the violation group to which the violation is assigned.

• The correction rate for violations in the group that are corrected as a result of the intervention.

On any given day, there are CMVs on the road that exhibit unsafe behaviors associated with violation groups. Roadside inspections and traffic enforcements intercept “non-crash daytrips,” revealing these unsafe behaviors and recording the associated violations. CMV daytrips that end in crashes (“crash daytrips”) are frequently subjected to post-crash inspections, revealing violations that existed prior to the crash. Non-crash daytrips intercepted in roadside inspections and traffic enforcements and their violations are considered to be a representative sample of all non-crash daytrips, and crashes

141

undergoing post-crash inspections and their violations are considered to be a representative sample of all crash daytrips.

E.5.5.1 Estimation of Crash Risk To estimate the CRP for a particular violation group (viol (j)), the percent of daytrips with a violation in that violation group resulting in a crash is calculated as shown in Figure 9.

Figure 9. Violation Group Crash-Risk Probability Estimation Equation

Estimating the numerator of Figure 9 is straightforward, since all the data needed are in MCMIS. That calculation is shown in Figure 10.

Figure 10. Equation to Estimate Numerator of Figure 9

Estimating the denominator of Figure 9 requires data in addition to that supplied by MCMIS, while the total number of non-crash daytrips is derived from other sources.136 Figure 11 shows the equation for estimating the denominator.

Figure 11. Denominator Determination Equation to Estimate CRP

136 www.fhwa.dot.gov/policy/ohim/hs06/pdf/vm1.pdf and www.fmcsa.dot.gov/documents/tb99-002.pdf. Accessed February 4, 2014.

142

Table 90. Crash Risk Reductions for Violation Groups in Intervention Model 3.0

Violation Group

Estimated Number of Non-Crash

Daytrips (Denominator)

Estimated Number of

Crashes (Numerator)

Estimated Likelihood of a

Crash (N/D)

Duration of Crash Risk Reduction (Daytrips)

Crash Risk Reduction

Careless Driving 184,915,747 26,039.53 0.000141 30 0.004224 Reckless Driving 184,915,747 5,139.75 0.000028 30 0.000834 Speeding-Related 184,915,747 14,432.49 0.000078 30 0.002341 HM-Related 184,915,747 177.48 0.000001 30 0.000029 Other Driver Violations 184,915,747 74,228.28 0.000401 30 0.012038 392.2 Driver 184,915,747 96,931.23 0.000524 30 0.015718 Hours 135,147,842 14,113.03 0.000104 30 0.00312 False Log 45,214,552 9,569.60 0.000212 30 0.00636 Incomplete/Wrong Log 501,678,937 61,875.83 0.000123 30 0.00369 Jumping OOS/Driving Fatigued 799,215 4,614.42 0.005741 30 0.17223 EOBR-Related 521,501 63.89 0.000123 30 0.00369 Driver Qualification 30,223,328 6,311.11 0.000209 45 0.009405 Endorsements & Vehicle Group 50,894,361 9,072.66 0.000178 45 0.00801 Medical Certificate 212,070,741 31,378.06 0.000148 45 0.00666 Physical 6,308,906 582.13 0.000092 45 0.00414 Multiple License 712,945 184.58 0.000259 45 0.011655 Fitness Jumping OOS 4,847 7.1 0.001463 45 0.065835 Alcohol 3,976,205 3,464.37 0.000871 90 0.07839 Drugs 1,484,049 1,476.61 0.000994 90 0.08946 Alcohol Jumping OOS 75,608 42.59 0.000563 90 0.05067 Brakes Out of Adjustment 598,114,058 76,330.43 0.000128 37 0.004736 Brakes, All Others 1,273,227,151 98,405.12 0.000077 37 0.002849 Coupling Devices 32,839,240 8,194.14 0.000249 7 0.001743 Exhaust Discharge 210,994,572 12,245.08 0.000058 37 0.002146 Fuel Systems 64,210,788 8,085.10 0.000126 37 0.004662 Lighting 820,016,680 76,271.72 0.000093 7 0.000651

143

Violation Group

Estimated Number of Non-Crash

Daytrips (Denominator)

Estimated Number of

Crashes (Numerator)

Estimated Likelihood of a

Crash (N/D)

Duration of Crash Risk Reduction (Daytrips)

Crash Risk Reduction

Steering Mechanism 137,599,046 10,576.05 0.000077 37 0.002849 Suspension 206,889,516 25,957.88 0.000125 37 0.004625 Tires 517,702,925 70,627.25 0.000136 7 0.000952 Wheels, Studs, Clamps, Etc. 612,694,671 60,948.60 0.000099 7 0.000693 Windshield/Glass/Markings 216,312,307 21,605.01 0.0001 7 0.0007 Cab, Body, Frame 151,690,832 23,567.58 0.000155 7 0.001085 Inspection Reports 237,913,373 36,928.13 0.000155 37 0.005735 Vehicle Jumping OOS 670,547 159.35 0.000238 37 0.008806 Other Vehicle Defect 536,736,891 72,589.81 0.000135 37 0.004995 Emergency Equipment 437,084,631 41,599.71 0.000095 37 0.003515 Tire vs. Load 39,856,664 3,967.07 0.0001 37 0.0037 Clearance Identification Lamps/Other 668,424,583 54,809.29 0.000082 7 0.000574 392.2 Vehicle 1,474,311,922 155,755.68 0.000106 37 0.003922 Load Securement 232,189,190 38,907.47 0.000168 30 0.00504 Other Cargo 78,886,501 12,454.75 0.000158 30 0.00474 Fire Hazard 525,171 41.94 0.00008 30 0.0024 Markings 28,696,023 1,593.54 0.000056 30 0.00168 Cargo Protection 3,131,643 478.06 0.000153 30 0.00459 Documentation 16,323,309 1,098.70 0.000067 30 0.00201 HM Route 169,000 25.16 0.000149 30 0.00447 Fraudulent Behavior 18,778 0 0 30 0 Package Integrity 5,248,076 436.13 0.000083 30 0.00249 HM Other 3,383,022 251.61 0.000074 30 0.00222 Package Testing 4,485,457 385.8 0.000086 30 0.00258 HM Shipper 27,526,351 1,710.96 0.000062 30 0.00186

144

E.5.5.2 Estimation of Duration The recording of a violation group results in a risk reduction for a greater duration than the daytrip during which the intervention occurs. Studies have found evidence that drivers’ behaviors are affected long after they are caught for unsafe driving practices. And vehicle safety deficiencies not discovered in inspections would have continued to exist on active CMVs for many more trips. The duration of the risk reduction varies according to the violation group to which the violation is assigned. The rationales for the reductions are described below.

For violations in the Unsafe Driving, Fatigued Driving, and Improper Loading categories, the corrected violation is assumed to remain corrected for 30 daytrips. These categories are treated similarly in the Intervention Model, since the same violation may occur independently on successive days. For example, a driver who improperly secured goods on one daytrip could improperly secure the next shipment of goods as well. Two recent studies137 suggest that the effect of a traffic citation on a driver is to deter the driver from that behavior for the next 1–3 months. Assuming the same effect would occur when a CMV driver is given a violation in one of these three categories in a traffic enforcement or roadside inspection, a duration of 1 month (to be conservative) is applied when these violations occur in the Intervention Model.

For a drug- or alcohol-related violation, the crash reduction is assumed to be 90 daytrips. The minimum driver’s license suspension in all States for such a violation is 90 days.138 By using a 90-day duration in the Intervention Model, it is assumed that a sober replacement driver performs the trips the cited driver would have performed for the duration of the suspension.

For vehicle maintenance- or driver fitness-related violations, the crash reduction is assumed to last for more than one daytrip. The reasoning is that if the inspection revealing the vehicle maintenance or driver fitness violation had not occurred, then the vehicle or driver would have operated with that violation and its corresponding crash risk until another inspection or a crash occurred. For example, if an equipment violation, such as flawed brakes, had not been discovered during a roadside inspection, then the CMV would have continued to operate with the flawed brakes until the flaw was discovered during preventive maintenance or a roadside inspection, or until the brakes failed and caused a crash. In this case, the violation reduced the crash risk for a duration equal to the number of days between the time the flaw was detected at the roadside and the time it ultimately would have been discovered after one of these three latter events—assuming that the brakes remained fixed during this same time duration.

For a Vehicle Maintenance violation involving aspects of a vehicle that can be observed in a walk-around inspection, the duration is assumed to be seven daytrips. For other Vehicle Maintenance violations that require a more thorough inspection, the duration is assumed to be 37 137 Redelmeier, Donald A. , Robert J Tibshirani, and Leonard Evans; “Traffic-law enforcement and risk of death from motor-vehicle crashes: case-crossover study,” The Lancet, Vol 361, June 28, 2003. http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(03)13770-1/fulltext. Accessed February 6, 2014. www.thelancet.com. And “Traffic enforcement in Europe: effects, measures, needs and future,” The “Escape Project.” Technical Research Center of Finland. December 2002. http://www.transport-research.info/Upload/Documents/200409/20040909_144405_72910_escape.pdf. Accessed February 6, 2014. 138 Serenity Insurance. DUI penalties State by State. http://www.serenitygroup.com/dui-penalties/. Accessed February 6, 2014.

145

daytrips (8,000 miles at 216 miles per day on average for the U.S. CMV fleet). These durations represent inspection intervals recommended by vehicle maintenance groups, such as the Transportation Services Division of Maryland’s Department of Public Works and Transportation139; a vehicle buyers’ service140; or a trucking resource center.141

The duration for driver fitness violations is assumed to be the mean time between CVSA inspections. CVSA issues decals to commercial vehicles that pass safety inspections according to FMCSA and CVSA criteria. These decals exempt a vehicle from additional inspections for the remainder of the quarter in which it was inspected, or an average of 45 days.142 Assuming that a driver inspection would typically occur along with the vehicle inspection, a driver could operate for an average of 45 daytrips before a Driver Fitness violation that occurred immediately after an inspection was discovered in a subsequent inspection.

Table 89 (above) shows the durations of risk reduction used in the Intervention Model 3.0 for violation groups and the basis for their magnitudes.

E.5.5.3 Estimation of Crash Risk Reduction Figure 11 shows the final calculation for the Risk Reduction (RR) for violation (j).

Figure 11. Final Calculation of Risk Reduction for Violation (j) Equation

Table 90 (above) shows the crash risk reduction estimates using 4 years of MCMIS data (CY 2003–06).

E.5.6 Indirect Effects The premise behind including indirect effects of interventions was that once carriers had been exposed to interventions, they would change their future behavior. Version 2.0 calculated indirect effects by comparing a carrier’s performance in the base year to performance for the following year. Fewer violations per inspection in the following year compared to the base year would indicate improvement resulting from changed behavior. The Version 2.0 model translated the differences in the year-to-year violation rates into crashes avoided, using the same methodology as for the direct effects estimation, summing over all carriers that showed improvement and expressed as a percentage change from base year crashes avoided. This improvement rate was applied to those carriers for which there were insufficient inspection data as well. An overall estimate of indirect effect crashes avoided combined the results for both carriers with sufficient inspection data and carriers with insufficient inspection data.

139 St. Mary’s County Government Web site. http://www.co.saint-marys.md.us/dpw/prevemaint.asp. Accessed February 6, 2014. 140 Edmunds.com, “How to Do a Maintenance Inspection.” http://www.edmunds.com/how-to/how-to-do-a-maintenance-inspection.html?articleid=43792. Accessed February 6, 2014. 141 Boom Truck Resource Center, “Maintenance Schedule.” http://www.boomtruck.com/Default.aspx Accessed February 6, 2014. 142 Source: FMCSA Field Office personnel.

146

It was determined that the estimation of indirect effects would be eliminated from Intervention Model 3.0, based on research that attempted to answer two analyses questions.

E.5.6.1 Did significantly more carriers improve or decline in their performance between the base and the following year?

The intent of this analysis was to determine whether random year-to-year fluctuations in carrier performance were being credited as indirect benefits of the program. The analysis looked at each 2-year period between 2004 and 2008 inclusive, and compared the number of carriers whose year-to-year performance improved and the number of carriers whose performance declined, their average number of violations corrected during interventions, and the resulting estimated crashes avoided. As shown in Table 91, the difference in the number of carriers that improved and declined and their performance was not significant in any of the periods. The Version 2.0 indirect effects estimation looked only at carriers that improved and ignored those that declined in performance. If Version 2.0 had also considered those carriers with declining performance, there would have been much less net improvement, and a smaller indirect effect.

Table 91. Change in Carrier Performance (Estimated Crashes Avoided) 2-Year Period

Carriers That Improved

Average Improvement

Carriers That Declined

Average Decline

2004–05 23,717 1.01 25,280 -1.00

2005–06 25,507 1.02 25,877 -1.03

2006–07 29,463 1.02 24,521 -1.01

2007–08 28,015 0.99 26,959 -1.00

E.5.6.2 Did improvement in carrier performance last for more than 1 year? Using 2004 as a baseline, the analysis determined how many carriers sustained improved performance over the next 3 years. The carriers were grouped according to the number of inspections they received during the base year. Improvement was measured in terms of crashes avoided per inspection. Table 92 shows the probability of sustained improvement over the 3 years by the number of inspections in 2004 and the degree of improvement. It can be seen that carriers with the greatest number of inspections in 2004 had the greatest likelihood of maintaining improved performance during the next 3 years.

Table 92. Probability of a Carrier Sustaining Improved Performance for 3 Years by the Number of Inspections in 2004

Crashes Avoided

5–20 Inspections

21–35 Inspections

36–70 Inspections

70+ Inspections

0–0.25 12% 19% 21% 30%

0.25–5 18% 32% 38% 49%

0.5–75 24% 39% 50% 53%

0.75–1 29% 48% 58% 65%

1–2 41% 55% 65% 71%

147

Crashes Avoided

5–20 Inspections

21–35 Inspections

36–70 Inspections

70+ Inspections

2–3 50% 76% 76% 83%

3–4 57% 82% 100% 100%

4–5 70% 96% 100% 100%

5+ 81% 97% 100% 100%

If Version 2.0 had considered fluctuating performance by a carrier during the 3-year period after the base to indicate a random response to its intervention experience, and had considered only those carriers that sustained their improvements over 3 years to represent actual behavioral changes, then the probabilities in Table 93 would represent the proportion of carriers in each category that experienced indirect effects of interventions. When these probabilities are applied to CY 2004–07, Table 93 shows the corresponding indirect effects estimates, which are on average about 10 percent of the indirect effects generated from Version 2.0. These estimates are within the range of the direct effects estimates.

Table 93. Indirect Effects Estimates Based on Sustained Improvement by Carriers

Year Crashes Avoided

Number of Carriers with Sustained Improvement

Crashes Avoided

per Carrier

CY 2004 246 8,286 0.030

CY 2005 279 7,929 0.035

CY 2006 536 9,065 0.059

CY 2007 517 8,742 0.059

E.5.6.3 Consideration of Duration in Version 3.0 Crash Risk Reduction Estimation The improved crash risk reduction method for Intervention Model 3.0 encompasses indirect effects of interventions to some degree by including the concept of the duration of the crash risk reductions. Recall that correcting a violation results in a risk reduction for a greater duration than the daytrip during which the intervention occurs. This, in a sense, is an indirect effect. The durations in Table 89 can be considered measures of indirect effects of interventions that are much more straightforward and accountable than those produced by the Version 2.0 estimation method.

E.5.7 Comparison of Results between Intervention Models 3.0 and 2.0 Due to the changes to the methodology introduced to Intervention Model 3.0 in FY 2008, one cannot directly compare FY 2008 results with estimates for previous years based on Version 2.0. Most significantly, the method for determining the crash risk associated with violations in Version 3.0 is completely different from that used in previous versions. To provide a consistent time series of annual estimates for analytic purposes, Model 3.0 was run for several prior years, and the differences between results from the two models were compared.

148

Figure 12 shows that Version 3.0 direct effects estimates are consistently larger than those of Version 2.0, ranging from 3 to 12 percent.

Figure 12. Intervention Model: Direct Effects Versions 3.0 and 2.0

Figure 13 shows the magnitude of indirect effects as estimated by Version 2.0. As discussed in the previous section, recent analysis has shown these to have been inflated. Hence they were eliminated from Version 3.0.

Figure 13. Intervention Model: Indirect Effects Version 2.0

Finally, Figure 14 shows a comparison of the overall results of Intervention Models 2.0 and 3.0. Because indirect effects are not estimated in Version 3.0, Version 3.0 estimates are consistently

149

lower than those of Version 2.0, ranging from 84 to 91 percent of the latter. The elimination of indirect effects estimates and the implementation of the other modifications in Version 3.0 yield results that are based on a sounder and more defensible statistical footing.

Figure 14. Intervention Model Results: Comparison of Versions 3.0 and 2.0

150

APPENDIX F: VALIDATING SAFETY BENEFIT MODEL FMCSA examined Trucks Involved in Fatal Accidents data to validate its safety benefits model. For each year’s worth of data from NHTSA’s Fatality Analysis Reporting System (FARS), a census of all fatal crashes in the United States, the University of Michigan Transportation Research Institute, under FMCSA sponsorship, conducted survey research into those fatal crashes involving large trucks. Among the additional information collected are data on whether or not the large truck driver was fatigued and how many hours he or she had been driving, or, if the precise number of hours is not available, an indication of whether or not he or she was in violation of the 11-hour drive time limit. TIFA cannot be used for a complete comparison to the Agency’s safety benefit analysis because it is confined only to fatal crashes and does not collect data on other HOS violations. However, secondary data that even partially corroborate FMCSA’s estimate of safety can confirm the overall validity of the model.

Data for the 5 years from 2007 through 2011 indicate that about 1.23 percent of crashes are fatal. FMCSA applied this percentage to the crash reductions from the Roadside Intervention Model specific to enforcement of the 11-hour rule to estimate a reduction of 4.2 fatal crashes for Option 1, and 3.5 fatal crashes for Option 2.

To construct a comparison, FMCSA examined only TIFA crashes in which a driver was fatigued and in violation of the 11-hour rule. The most current TIFA data are available for 2010, but for that year, a redesign of driver-related factors in FARS necessitated the omission of the fatigue indicator. Instead, FMCSA used the 5 years from 2005 through 2009 for its analysis. Over that period, an average of 9.4 fatal crashes satisfying these filters occurred each year. One can assume that the ELD effectiveness measure, a 45 percent compliance improvement, would apply to the CMVs involved in these crashes. Consequently, TIFA data combined with the ELD effectiveness parameter indicates that 4.21 fatal crashes could be avoided each year with ELD use for Option 1, and 3.49 fatal crashes for Option 2. The full details of how these comparisons were calculated are shown in Table 94 below. Also, as shown in the last line of this table, the two sets of estimates were different by less than one percentage point. Not only does this confirm that the order of magnitude of the crash reduction estimated using the Intervention Model is correct, but also results in estimates that are remarkably close.

Table 94. Comparison of Estimated Safety Benefits (2013 $)

Option 1 Option 2 Intervention Model

LH Crashes Avoided 223.60 223.60 SH Crashes Avoided 115.88 58.57 Total Crashes Avoided 339.49 282.18 Percentage of Crashes That Are Fatal 1.23% 1.23% Fatal Crashes Avoided 4.18 3.47 Fatal Crash Cost (millions) $10.80 $10.80 Annual Safety Benefits from Fatal Crashes Avoided (millions) $45.15 $37.48

TIFA Data, 2005-2009 Annual Average Number of Fatal Crashes 9.4 9.4 ELD Effectiveness 45% 45%

151

Option 1 Option 2 Fatal Crashes Avoided 4.21 4.21 Adjustment for smaller SH Population 0% -17% Crashes Avoided 4.21 3.49 Fatal Crash Cost (millions) $10.80 $10.80 Annual Safety Benefits from Fatal Crashes Avoided (millions) $45.48 $37.70 Percent Difference, Intervention Model and TIFA 0.71% 0.57%