Characteristics of Interstate Motorcoach Carriers With Elevated Rates of Crashes and Inspection Violations

Abstract

Widely-publicized fatal motorcoach crashes have caused public concern about their safety. This study estimated crash and violation rates among interstate motorcoach carriers based on 2005–2011 data obtained from the Federal Motor Carrier Safety Administration (FMCSA). Motorcoach carriers with relatively high crash and violation rates were compared with those with better safety records. The principal component analysis produced three orthogonal factors that captured the majority (63 percent) of the total variance in the data set. Motorcoach carriers operating 10 or fewer motorcoaches were more likely to be classified in both the high crash rate and the high inspection and violation rates group. Those carriers with 10 or fewer years in business were more likely to be classified in the high inspection and violation rates group. The vast majority of motorcoach carriers with problematic safety records were non-scheduled route providers (charters). Scheduled-service motorcoach carriers identified as providing at least occasional curbside service, defined as picking up or dropping off passengers at a place other than a traditional terminal at the origin or destination, had an increased risk of involvement in fatal crashes compared with other scheduled-service carriers (1.4 per 100 vehicles, 95% C.I.: 0.1–2.7 versus 0.2, 95% C.I.: 0.0–0.5). The data did not indicate whether crashes or violations occurred during the trips where curbside service was provided. These findings suggest that FMCSA and the states need to have the resources necessary for close monitoring of motorcoach carriers, particularly high-risk ones such as small and less experienced motorcoach carriers.

INTRODUCTION

Motorcoach safety has received increased public attention as a result of multiple fatal and serious crashes during 2011. According to the Federal Motor Carrier Safety Administration (FMCSA) (2009), “a motorcoach is a bus with integral construction designed for long-distance passenger transportation. It measures at least 35 feet long and can seat 30 or more passengers on an elevated passenger deck over a baggage compartment.”

Motorcoaches are not specifically identified in either the Fatality Analysis Reporting System (FARS) census of fatal crashes on U.S. public roads (maintained by National Highway Traffic Safety Administration (NHTSA)) or the Buses Involved in Fatal Accidents database that supplements FARS (maintained by University of Michigan Transportation Institute). FARS does refer to cross-country/intercity buses; such buses largely are motorcoaches.

National statistics on buses are useful for understanding motorcoach safety because different categories of buses have similar vehicle designs. The annual number of fatal crashes involving buses averaged 310 during 2004–2008 [Matteson et al., 2011]. Compared with passenger vehicles and light trucks, buses and other commercial motor vehicles have a higher likelihood of fatal crash involvement per registered vehicle. Federal Highway Administration (FHWA) [FHWA, 2009] and NHTSA (2010) data showed that, per 100,000 registered vehicles, 14.7 passenger cars, 19 light trucks (pickups and sport utility vehicles), 29.3 buses, 45.4 large trucks, and 69.8 motorcycles were involved in fatal crashes during 2008. Given involvement in a crash, occupants of buses are less likely to die than occupants of passenger vehicles. During 2009, the bus occupant fatality rate was 45 deaths per 100,000 crashes, compared with 251 deaths per 100,000 crashes for passenger car occupants.

Of all fatal bus crashes during 1999–2007, school buses were the most common type of passenger carrier (39 percent) followed by transit buses (33 percent), charter/tour buses (11.5 percent), other/unknown bus types (10 percent), and intercity buses (3.4 percent) [Blower et al., 2010]. Most road users who were fatally injured in bus crashes were occupants of other vehicles followed by non-occupants (pedestrians and bicyclists) and then bus occupants. The proportion of fatally injured road users who were bus occupants was higher for intercity and charter/tour buses than for school or transit buses.

The focus of this paper is on the type of bus, namely motorcoaches, that provides long-distance transport for 30 or more passengers.An average of 9 intercity motorcoaches and 38 charter motorcoaches were involved in fatal crashes each year during 2004–2008 [Matteson et al., 2011]. Although these numbers of fatal crashes are low relative to the numbers of fatal crashes involving large trucks, any crash involving a motorcoach has the potential for a large loss of life or large number of serious injuries.

Interstate motorcoach travel has been described as the most rapidly growing mode of transport in recent years [Schwieterman and Fischer, 2010]. The National Household Travel Survey, which the FHWA conducts every 5 to 7 years, estimated that 60 million person-trips occurred on intercity buses in 2009 and that 241 million person-trips occurred on charter and tour buses during 2009 [FHWA, 2011]. Accordingly, the number of people exposed to motorcoach travel either as motorcoach occupants or as other road users is considerable and growing.

One business model for carrying passengers has become popular and is commonly called curbside service [Klein, 2009; O’Toole, 2011]. It involves providing scheduled service from one city to another city or to a popular destination (e.g. casino) charging low fares and picking up or discharging passengers at places other than traditional bus terminals, including sidewalks or parking lots. Many motorcoach carriers established during the past decade provide this service and some of the more established motorcoach carriers, including Greyhound, also provide it. Most motorcoach carriers that engage in curbside operations also provide other services, including charter operations for private groups [United Motorcoach Association, 2011].

Questions have been raised about whether safety is being compromised by curbside service. One concern is that lower bus fares may be associated with fewer resources being directed toward vehicle maintenance and compliance with safety rules.

The FMCSA oversees the safety of interstate motor carriers, including trucking carriers and passenger-carrying carriers [U.S. Department of Transportation (USDOT), 2009]]. This includes regulating commercial interstate driver hours-of-service limits, alcohol and drug testing requirements, driver medical certification standards, minimum requirements for commercial driver’s licenses issued by individual states, requirements for freight carriers and certain types of brokers, and minimum standards for vehicle maintenance and roadworthiness. Monitoring of compliance with safety regulations is done in concert with the states, who receive financial assistance from FMCSA to conduct roadside inspections. The most comprehensive form of oversight is the compliance review, which involves a visit to headquarters and examination of operational records. Roadside inspections of driver logbooks, driver qualifications, and vehicle condition are an essential component of oversight. Compliance with the Federal Motor Carrier Safety Regulations and Hazardous Material Regulations is determined at inspections.

This study was one component of an investigation conducted by the National Transportation Safety Board (2011) concerning motorcoach safety. The research aimed to identify characteristics of motorcoach carriers with high crash rates and high safety violation rates. Curbside service and frequency of roadside inspections were of particular interest.

METHODS

Carrier, Crash, and Inspection Data

Motor carrier census information were obtained from FMCSA registration data for all motor carriers with USDOT numbers considered active as of April 22, 2011, including company name, address, and numbers and types of power units [USDOT, 2011a]. A list of carriers that are registered to carry passengers was extracted from a separate FMCSA database system [USDOT, 2011b]. By combining multiple FMCSA data files, we identified 4,172 interstate motorcoach carriers as those with interstate authority operating at least one motorcoach.

State crash records in the FMCSA database on motor carriers [USDOT, 2011c] were also extracted. A total of 908,394 records of crashes involving trucks or buses between January 2005 and March 2011 were extracted. Ten percent (80,527 crashes) were bus crashes of various injury severity levels (fatal, nonfatal injury, and property-damage-only). Buses were categorized as either large (seats for more than 15 people, including driver) or small (seats for 9 to 15 people, including driver). The majority (86%) of bus crash records involved large buses such as motorcoaches. These crash records were linked to the 4,172 interstate motorcoach carriers by USDOT number. However, a large number of crashes did not have the USDOT number listed. An attempt was made to link crashes without the USDOT number by the carrier names. However, this effort did not succeed in linking all crashes with carriers, so the crashes of interstate motorcoach carriers were undercounted. A total of 6,019 motorcoach crashes were included in the analysis.

Whenever rates were calculated, including crash rates, inspection rates, and inspection violation rates, an adjustment was made to account for the length of time in business relative to the data time frame. This was necessary because some carriers had only been in business for part of the periods that data were collected during 2005–2011. The term “adjusted rates” refers to this procedure.

With the adjustment, we calculated five measures per vehicle for each carrier: (1) adjusted overall crash rate; (2) adjusted injury or fatal crash rate; (3) adjusted fatal crash rate; (4) adjusted death rate; and (5) adjusted injured person rate (Table 1).

Table 1.

List of variables and time periods used for principal component analysis

Variable	Data period
1. Adjusted overall crash rate	2005 to 2011
2. Adjusted injury or fatal crash rate	2005 to 2011
3. Adjusted fatal crash rate	2005 to 2011
4. Adjusted death rate	2005 to 2011
5. Adjusted injured person rate	2005 to 2011
6. Adjusted motorcoach inspection rate	2007 to 2011
7. Adjusted inspection rate (all buses)	2009 to 2011
8. Adjusted motorcoach out-of-service (OOS) violation rate	2007 to 2011
9. Adjusted violation rate (all buses)	2009 to 2011
10. Adjusted out-of-service (OOS) violation rate (all buses)	2009 to 2011
11. Number of states where motorcoach inspections occurred	2009 to 2011
12. Percent of roadside motorcoach inspections that were routine	2009 to 2011

We obtained two sets of FMCSA inspection records. The first covered about four years (January 2007 to March 2011) of motorcoach inspection summary data file [USDOT, 2011d], but did not provide details about specific violations. We linked these records to the 4,172 interstate motorcoach carriers and derived four more measures: (1) adjusted motorcoach inspection rate per vehicle; (2) adjusted motorcoach out-of-service rate per vehicle; (3) number of states where motorcoach inspections occurred; and (4) percent of motorcoach inspections not triggered by moving traffic violations (Table 1). Inspection rates were included because safety violations cannot be discovered without inspections, so this was a method of adjusting for whether a passenger carrier had received inspections. Motorcoach inspection and out-of-service rates were adjusted by length in business and used number of motorcoaches as the denominator. The third variable was created to provide a proxy of the geographic scope of operations whereas the fourth variable was designed to estimate the percentage of regular roadside inspections among all inspections.

We provided the list of 4,172 interstate motorcoach carriers to FMCSA to extract detailed inspection records for a 24-month period, roughly April 2009 to March 2011. This database enabled us to look at the in-depth inspection and violation history for the carriers. Because this dataset covered a two-year period instead of the previously-used longer periods, we computed another adjusted violation rate (all buses) and another adjusted out-of-service rate (all buses). These two rates, combined with the measures described above, formed the core of the data used in the principal component analysis (PCA) (Table 1).

Independent of the PCA, we were able to categorize safety violations in the detailed inspection records into four main categories: (1) driver fitness; (2) fatigued driving; (3) vehicle maintenance; and (4) unsafe driving. Adjusted violation rates were computed for all four categories and adjusted out-of-service violation rates for driver fitness, fatigued driving, and vehicle maintenance were calculated per vehicle for each motorcoach carrier. The carriers that ranked among the worst 10 percent (that is, in the 90^th percentile) in each of these categories were identified. This additional step augmented the PCA results by further highlighting carriers with severe violation rates.

We also obtained information about the type of authority, based on the registration application, from FMCSA [USDOT, 2011d]. This was used to identify those carriers authorized to conduct regular route scheduled services available to all purchasing tickets, typically long-distance point-to-point services along major interstate corridors.

Classification of Motorcoach Carriers Based on Safety Records and Type of Service

A data matrix of the 4,172 carriers by 12 variables listed in Table 1 was developed to represent the overall crash and inspection experience. These variables reflected somewhat different aspects of the overall safety record of the carriers, but the variables were highly correlated with each other. In this study, we used the scree-plot approach to determine that three principal components should be retained. The absolute values of the PC scores associated with each carrier for the three components were compared and the carrier was then assigned to the component with the highest positive scores. Our focus was on those carriers with poor safety records, such as high crash rates or high inspection violation rates.

We categorized the 4,172 interstate motorcoach carriers into three sub-categories based on the type of service provided. Because no official categorization exists and FMCSA does not systematically classify the carriers, we used a qualitative approach with a focus on identifying those providing curbside service, as defined earlier. This classification process first identified carriers that provide scheduled regular route services from one city to another. Among these carriers, we used multiple online ticketing websites to further identify those providing curbside service by carefully evaluating the origins, destinations, and en-route stops of their scheduled runs [NTSB, 2011].

The operational characteristics, such as the size of the operation (in terms of number of motorcoaches), length in business, and type of services of the categories of carriers based on safety records were compared using difference of means and chi-square analyses. Mantel-Haenszel chi-square statistics was used when appropriate. Two-sided P-values of 0.05 were used to evaluate all statistical tests.

RESULTS

Operational Characteristics of Motorcoach Carriers

The majority of interstate motorcoach carriers had very few motorcoaches and drivers. Of 4,172 motorcoach carriers, 38 percent of all carriers operated only 1 motorcoach and 47 percent had 2 to 10 motorcoaches. Similarly, 24 percent of carriers only had 1 driver while 47 percent used 2 to 10 drivers. Motorcoaches made up a very large portion of the fleets; on average, 84% of all power units (vehicles) are motorcoaches. Most carriers were fairly young, with a median length in business of 8 years. As of March 2011, 46 percent of all carriers had been in business for 5 years or less, 38 percent had been in business for 6–20 years, while 16 percent had been in business for more than 20 years.

The 4,172 motorcoach carriers were registered across the United States, as shown in Figure 1a, which includes the lower 48 states and the District of Columbia (Alaska and Hawaii have 14 and 3 carriers, respectively). To show the relative concentration of these carriers, we computed carrier density per square kilometer using a kernel density method. The area with the highest concentration of carriers was the northeast corridor. Four other areas with high densities of carriers were Chicago, Atlanta, Los Angeles, and San Francisco.

Figure 1.

Locations of motorcoach carriers, carrier density, crash counts by county, and crash density

a. Locations of motorcoach carriers and carrier density per square kilometer

b. Crash counts by county (January 2005–March 2011) and crash density per square kilometer

Crashes involving large buses, including motorcoaches, during the January 2007 to March 2011 period occurred across the United States (Figure 1b). We also computed the crash density per square kilometer and highlighted the locations with very high crash densities with contours. Two distinctions are apparent when compared with Figure 1a. Although the North Carolina and Atlanta areas had high densities of registered carriers, they did not have high crash densities. In contrast, Denver had a high crash density but not a high registered carrier density.

Classification of Motorcoach Carriers Based on Safety Records

According to state crash data compiled by the FMCSA, 1,254 of the 4,172 motorcoach carriers (30%) had been involved in at least one reportable crash from January 2005 to December 2010. These motorcoach carriers were involved in a total of 6,019 crashes, 191 of which were fatal (resulting in 262 deaths) and 3,134 of which were nonfatal injury crashes (injuring 9,062 people). On average annually, motorcoach carriers were involved in 1,003 reported crashes, 32 fatal crashes resulting in 44 deaths, and 505 nonfatal crashes that injured at least one person. During January 2007 to March 2011, 64% of all motorcoach carriers had roadside inspections at least once; this included vehicles other than motorcoaches, as most carriers operated more than one type of vehicle. Table 2 shows the breakdown of these inspections by inspection level. Thirty-four percent of these inspections were driver-only, largely during traffic enforcement activities.

Table 2.

Distribution of motorcoach inspections by inspection level and percentage of motorcoach carriers receiving at least one inspection, 2007–11

Inspection Level	Inspections	% of Total
I: Full	19,089	20
II: Walk-around	19,147	21
III: Driver-only	31,696	34
V: Vehicle-only	22,989	25
All inspections	92,921	100

Using principal component analysis, the 4,172 by 12 data matrix produced 12 principal components, each of which independently explained a portion of the total variance. We examined the scree-plot and decided to retain only three principal components, which collectively captured 63 percent of total variance.

We examined the component matrix of the first three principal components. The values are correlation coefficients between the principal components and the original variables and they range from −1 to +1. By interpreting these values, we initially classified all 4,172 carriers into three components. Principal component 1, explaining 31 percent of the total variance, captured primarily the overall motorcoach and bus inspection and out-of-service violation rates. The motorcoach carriers with very high positive principal component scores in the first component had high rates of being inspected and also high out-of-service violation rates. These carriers did not necessarily have high crash rates. The second principal component explained 20 percent of the total variance and was largely capturing the overall crash experience, as expressed in crash rates and death and injured person rates. Therefore, those carriers with high positive principal component scores in this component generally had more crashes relative to their time in business. The third principal component only captured 12 percent of the total variance and is largely tied to the geographic scope of roadside motorcoach inspections and percent of roadside motorcoach inspections that were routine. Those carriers with very high positive scores in this component had their motorcoach inspected in multiple states, indicative of their geographic scope of business, and had higher percentages of their inspections conducted outside of traffic enforcement activities.

Since our focus was on safety records, we focused on the first two principal components. By comparing the PC scores for each carrier, we identified 233 carriers with very high positive PC scores for the second component (crash experience). These carriers were categorized as the high crash rate group. The second group of carriers included 353 carriers that had very high positive PC scores for the first component (very high inspection and violation rates). We then examined the violation rates and out-of-service violation rates in the driver fitness, fatigued driving, vehicle maintenance, and unsafe driving categories to further characterize those carriers in the first two principal component groups. All the remaining motorcoach carriers were collectively grouped into the third group. The ultimate results of these classification procedures produced three groups: (1) high crash rate group with at least one high violation rate (n=74); (2) high inspection and violation rates group (n=279); (3) other group (n=3,819). Carriers in both groups 1 and 2 had at least one of the seven violation and out-of-service violation rates ranked in the 90^th percentile.

Characteristics of Carriers with Elevated Crash Rates and Inspection Violation Rates

The mean adjusted overall crash rate during 2005–2011 was significantly higher, 93.8 per 100 vehicles, for the high crash rate group, whereas it was 22.6 for the high inspection and violation rates group and 7.7 for the other group. In the high crash rate group, the adjusted death rate was 14.5 per 100 vehicles and the injured person rate was 248.6 per 100 vehicles, both significantly higher than the high inspection and violation rates group.

As expected, the high inspection and violation rates group had significantly higher violation rates in driver fitness, fatigued driving, and vehicle maintenance. The carriers in the high crash rate group also had higher violation rates and out-of-service violation rates in these safety categories than those in the third (other) group. Having high inspection and violation rates was a marker for having significantly increased overall crash rates and combined injury and fatal crash rates for the 279 carriers in this group compared with the motorcoach carriers placed in the third group.

Table 3 shows the percent distribution of carriers classified by their safety records, numbers of motorcoaches, and number of years in business.

Table 3.

Percent distribution of motorcoach carriers by final safety and inspection classification, size of motorcoach fleet, and years in business

Motorcoach carrier principal component groups	No. of motorcoaches (χ2=16.5, p=0.0003)		No. of years in business (χ2=21,6, p<0.0001)
	≤10 (n=616)	>10 (n=3,556)	≤10 (n=1,765)	>10 (n=2,407)
High crash/high violation rate group (n=74)	96	4	57	43
High inspection and violation rates group (n=279)	91	9	71	29
Others (n=3,636)	85	15	57	43
All (n=4,172)	85	15	58	42

Ninety-six percent of carriers in the high crash rate group, 91% of carriers in the high inspection and violation rate group, and 85% of carriers in the other group had 1 to 10 motorcoaches (χ²=16.5, p=0.0003). In terms of length in business, the high inspection and violation rate group had the highest percent of carriers (71%) that had been in business for 10 years or less (χ²=21.6, p<0.0001).

Based on the registration information of the carriers and their memberships in one of the three groups based on the safety and inspection records, we computed the percent of registered carriers classified in the high crash rate group (Figure 2a) and the percent of registered carriers classified in the high inspection and violation rates group by state (Figure 2b). The state means were 1.5% and 4.3%, respectively, for the high crash and high inspection and violation groups. Figure 2 highlighted the states in which the percentages were greater than one standard deviation above the mean. Figure 2a shows no geographical pattern in terms of the percentage of carriers in the high crash rate group.

Figure 2.

Percent distribution of motorcoach carriers by state in the high crash rate group and high inspection and violation rates group.

However, Figure 2b shows distinct geographical patterns in terms of the percentage of carrier in the high inspection and violation rates group. Two large regions clearly emerged, one area encompassing Pennsylvania, Delaware, Maryland, West Virginia, and Ohio, while the other stretches across the southwest from California to Texas.

Table 4 shows the relationship between type of service provided and categorization by safety records. The 4,172 carriers were categorized either as scheduled curbside, scheduled conventional, and nonscheduled other (primarily charter). If there was evidence of a carrier providing scheduled curbside service at origin or destination points, such carrier was identified as a curbside carrier. If there was no evidence, but they provided scheduled service, they were classified as conventional carriers. Based on our research, we found evidence that 71 carriers provided at least one scheduled route run service that was considered curbside service, 51 carriers provided conventional schedule service, and the remaining 4,050 provided non-scheduled “other” services, presumably charter or tour group services. Although curbside carriers comprised 2% of all motorcoach carriers, they comprised 5% of carriers in the high crash rate group and 6% of carriers in the high inspection and violation rate group (χ²=33.4, p<0.0001).

Table 4.

Percent distribution of motorcoach carriers by final safety and inspection classification and type of services

Motorcoach carrier principal component groups	Type of service (Mantel-Haenszel χ2=33.4, p<0.0001)
	Curbside (n=71)	Conventional (n=51)	Nonscheduled Others (n=4,050)
High crash rate (n=74)	5	0	95
High inspection and violation rates group (n=279)	6	3	91
Others (n=3,819)	1	1	98
All (n=4,172)	2	1	97

Figure 3 shows that curbside carriers had a higher driver fitness OOS violation rate of 13.8 compared with 4.7 for conventional carriers as well as a higher fatigued driving OOS rate (16.7 versus 11.2), whereas conventional carriers had a slightly higher unsafe driving violation rate and vehicle maintenance OOS violation rate. Although motorcoach carriers providing curbside service had the highest crash rates per vehicle (see Figure 4), 70 of the 74 of the motorcoach carriers identified as having high crash rates were in the nonscheduled other (charter) category. Similarly, 254 of the 279 motorcoach carriers identified as having high inspection and violation rates were in the nonscheduled other (charter) category.

Figure 3.

Average violation rates and out-of-service violation rates by service type, April 2009 to March 2011

Figure 4.

Average rates of overall crash, fatal crash, and injured persons by service type, January 2005 to March 2011

DISCUSSION

In general motorcoach carriers provide a safe mode of travel. Crashes among all types of interstate motorcoach services, including those applying the curbside business model, are infrequent. Occupant fatalities in motorcoach are even less common and most deaths in fatal motorcoach crashes occurred in other vehicles. Bus occupants in an accident have a lower risk of dying than passenger vehicle occupants.

There are over 4,000 interstate motorcoach carriers in the U.S. and they are registered across the US in every state. There are higher concentrations of these carriers along the northeast corridor and around major metropolitan areas such as Chicago, Atlanta, Los Angeles, and San Francisco. These are also the areas where the emergence of the long-distance low-fare curbside business model occurred in the last decade [Schwieterman and Fischer, 2010].

Interstate motorcoach carriers tended to be fairly small. The carriers were also fairly young, with 58 percent of all carriers having 10 or fewer years of business experience.

Crashes involving buses operated by these motorcoach carriers occurred across the U.S. However, the areas with high densities of large bus crashes were not always in the same areas as those with high densities of registered carriers. The possibility of motorcoach carriers registering in states other than the ones in which most of their operations take place was noted by NTSB (2011). Further investigations are warranted to determine whether some motorcoach carriers are doing this to avoid government oversight.

We examined safety records of 74 carriers classified in the high crash rate group and another 279 carriers classified in the high inspection and violation rate group. Motorcoach carriers with 10 or fewer motorcoaches or 10 or fewer years of business experience tended to be classified more often in the high crash rate or high inspection and violation rate groups. One implication of this finding is that FMCSA and states need to have the resources necessary to closely monitor small and less experienced motorcoach carriers.

In addition, motorcoach carriers with high violation rates had significantly increased overall crash rates and combined injury and fatal crash rates. This is consistent with the findings for all motor carriers from an evaluation of FMCSA’s Compliance, Safety Accountability program to identify high-risk motor carriers [Green and Blower, 2011]. FMCSA is targeting motor carriers for intervention based on violation and crash rates.

Type of service also appeared to be related to safety experience. Higher percentages of carriers offering curbside services were classified into the high crash rate group and the high inspection and violation rate group. We also found that carriers offering curbside service had higher death and injured person rates and higher driver fitness and fatigue violation rates than the other two carrier groups. Per vehicle, carriers in the conventional scheduled-service group had somewhat higher unsafe driving violation rates and vehicle maintenance violation rates than the other carrier groups. It should be noted that 95% of the 74 carriers identified as having high crash rates were in the nonscheduled other (charter) category. This is not surprising, given that there are far more charter motorcoach carriers and they travel more miles than scheduled carriers.

In interpreting the findings, it is important to note that the rates were computed using different denominators, such as number of motorcoaches or number of buses operated by the carrier. Mileage would be a better exposure measure; however, the mileage data captured by the MCS-150 forms, required to be updated every two years, was often missing or out of date [NTSB, 2011].

The analyses of carriers by whether they were in the 90^th percentile for crashes or violations were useful for describing the details of violations. Vehicle maintenance out-of-service rates far exceeded rates for other types of violations. Carriers could be included in the high crash rate only if they had an inspection record during the 24-month period between April 2009 and March 2011; this resulted in having some carriers with high crash rates classified in the third (others) group.

Crash counts were underestimated by a substantial amount. Police-reported crashes are uploaded by states to the FMCSA database system. The majority of motorcoach carrier crash records do not include USDOT numbers and the names of the carriers can vary. FMCSA periodically documents the extent of crash underreporting in each state [Green and Matteson, 2010].

The classifications of carriers based on type of services were imperfect. There is no official definition and no systematic data collection for such information. Our efforts, which relied on Internet resources, may undercount carriers that provide regular scheduled services and more particularly those providing scheduled curbside services. Furthermore, the industry is dynamic. Many carriers add and remove routes and service types regularly.

Crash records and inspection information were tied to the carriers and not to a specific type of service. Although we found that carriers providing curbside service were more likely to be classified in the high crash rate group or the high inspection and violation rate group, we did not know if those crashes and violations occurred while providing curbside service. Therefore our findings apply to the overall safety experience of the carrier and not the specific type of service provided by the carrier.

Future research on motorcoach safety would benefit from having more accurate measures of exposure, including mileage and travel routes. The collection of mileage data relies upon how well FMCSA enforces the MCS-150 form updating requirement. A more in-depth investigation of the operational practices of motorcoach carriers with high crash or violation rates would be useful to understand why they have higher rates and to identify possible countermeasures. One research approach could involve open-ended interviews of drivers and motor carrier staff and detailed investigations of contributing factors to particular crashes and safety violations. Potential countermeasures could include additional regulations, additional contacts between government safety oversight agencies and motor carriers, and incentives to reward good safety records. Any countermeasures need to be rigorously evaluated to determine if they are working as intended.

CONCLUSION

Although motorcoaches provide a safe means of transport, some categories of motorcoach carriers have higher crash rates and inspection violation rates, including those with 10 or fewer motorcoaches, those with 10 or fewer years in business, and those that have evidence of providing curbside service. Within these categories, safety records vary. Among motorcoach carriers with high crash or violation rates, more than 90 percent were unscheduled other (charter) carriers. These findings suggest that FMCSA and the states need to have the resources necessary for close monitoring of motorcoach carriers, particularly high-risk ones such as small and less experienced motorcoach carriers.

In 2012, NTSB recommended that the FMCSA revise the MCS-150 reporting requirements, as specified in 49 CFR 390.19, to require that carriers report fleet mileage, by year, for the two previous calendar years and that the FMCSA should develop and implement a plan for consistent, nationwide enforcement.