Estimating blood alcohol content from a national telephone survey.

One of the best methods for estimating blood alcohol content (BAC) in the driving population is through roadside breathtesting surveys that measure directly the BAC levels of drivers. However, there are barriers to conducting roadside breathtesting surveys in some states in the US. An alternative method for assessing BAC in the driving population may be to estimate BAC from self-reports of alcohol-involved driving collected via telephone surveys. This paper documents the results of estimating BAC in the driving population based on self-reported drinking, sex, and weight from a 1997 national telephone survey of US adult residents. These results are then discussed in light of findings from roadside surveys.

A comparison of two direct-observation methods for measuring daytime safety belt use.

This study compared two methods for measuring front-outboard shoulder-belt use: looking into vehicles when they stopped at a traffic control device (SVDO) and looking into vehicles as they traveled along a traffic corridor (MVDO). The reliability of the latter method has been questioned and certain surveys, such as the National Occupant Protection Use Survey (NOPUS), use both methods to estimate safety belt use. In one experiment, the methods were compared on overall belt use rates and reliability. A second experiment investigated the effect that vehicle speed had on an observer's ability to measure accurately belt use using the MVDO method. The results showed that daytime belt use rates between methods were nearly identical and inter-method reliability was quite high, indicating that front-outboard shoulder-belt use can be measured identically with either method. The second experiment showed that measurement accuracy was not affected by vehicle speeds of up to 60 mph and that overall accuracy was above 95%.

Field effectiveness of two restraint systems: the 3-point manual belt versus the 2-point motorized-shoulder/manual lap belt.

The research question addressed in this paper is whether or not occupant death, injury, and ejection outcomes differ between vehicles equipped with 3-point manual belts versus 2-point motorized-shoulder/manual-lap (motorized/manual) belt systems. Census crash data sets for the states of Washington and Texas, and the Fatal Accident Reporting System (FARS) were subset to provide data on front-outboard occupants of Ford Escorts involved in crashes in calendar years 1981-1991. Logistic regression analyses showed that occupants of vehicles equipped with the motorized/manual system experienced 11.7% to 26.4% fewer K- or A-level injuries than occupants of vehicles equipped with the 3-point system. Similar analyses of FARS data showed lower ejection rates for occupants of vehicles with the motorized/manual system in both rollover and nonrollover crashes. The installation of the motorized/manual system provided a substantial safety benefit over the manual 3-point system in the time periods examined.

Developing policies for automated speed enforcement: a survey of Michigan opinions.

This paper describes select results of a pilot test of automated speed enforcement devices (ASEDs) in Michigan. A mail response survey of drivers (1,209 respondents) showed the general public favors use of ASEDs in select situations, particularly in school zones, in areas where traffic enforcement is dangerous for police, for heavy trucks, and in construction zones. The survey also showed opposition to ASED use on freeways, on bridges, and on "all roads." In general, observed speeders and persons who reported having multiple citations in the previous two years were in greater opposition to the use of ASEDs than the general population.

The effect of enforcement upon service of alcohol to intoxicated patrons of bars and restaurants.

Laws prohibiting the service of alcohol to already intoxicated patrons of bars and restaurants are seldom enforced. Following introduction of an enforcement effort in Washtenaw County, Michigan, observed refusals of service to "pseudopatrons" simulating intoxication rose from 17.5% to 54.3%, declining eventually to 41.0%. At the same time, the percentage of those arrested drunk drivers coming from bars and restaurants declined from 31.7% to 23.3%. In a comparison county, refusals of service rose to a significantly smaller extent, from 11.5% to 32.7%, while the percentage of DWIs coming from bars and restaurants showed no significant changes. Service refusals were related to volume of business and numbers of intoxicated patrons in an establishment at the time of observation, while numbers of arrested DWIs was related to the nature of the establishment's clientele, policies, and practices. While enforcement of alcohol service laws offers a potentially cost beneficial means of reducing highway crashes, replication across additional jurisdictions is needed.

Increasing safety belt use in a secondary enforcement state: evaluation of a three-county special enforcement program.

Special enforcement programs to increase belt use present a unique challenge in states with secondary enforcement laws. This paper reports findings from an evaluation of a combined public information and education/enforcement program to increase restraint use along a highly traveled corridor in three Michigan counties. This program (US-31 SAVE) was successful in increasing belt use at least temporarily along the special enforcement corridor. Observed belt use increased from a baseline rate of 56.7% to 65.1% during the program before slipping slightly to 62.7% after the intensive enforcement and PI&E declined (all differences significant at p less than .05). This paper details program and evaluation activities and suggests future research needs to better understand the most effective mix of public information and enforcement efforts in states with secondary belt use laws.

Estimating costs of traffic crashes and crime: tools for informed decision making.

Traffic crashes and crime both impose significant economic and social burdens through injury and loss of life, as well as property damage and loss. Efforts to reduce crashes and crime often result in competing demands on limited public resources. Comparable and up-to-date cost data on crashes and crime contribute to informed decisions about allocation of these resources in important ways. As a first step, cost data provide information about the magnitude of the problems of crashes and crime by allowing us to estimate associated dollar losses to society. More importantly, cost data on crashes and crime are essential to evaluating costs and benefits of various policy alternatives that compete for resources. This paper presents the first comparable comprehensive cost estimates for crashes and crime and applies them to crash and crime incidence data for Michigan to generate dollar losses for the state. An example illustrates how cost estimates can be used to evaluate costs and benefits of crash-reduction and crime-reduction policies in making resource allocation decisions. Traffic crash and selected index crime incidence data from the calendar year 1988 were obtained from the Michigan State Police. Costs for crashes and index crimes were generated and applied to incidence data to estimate dollar losses from crashes and index crimes for the state of Michigan. In 1988, index crimes in Michigan resulted in $0.8 billion in monetary costs and $2.4 billion in total monetary and nonmonetary quality-of-life costs (using the willingness-to-pay approach). Traffic crashes in Michigan resulted in $2.3 billion in monetary costs and $7.1 billion in total monetary and nonmonetary quality-of-life costs, nearly three times the costs of index crimes. Based on dollar losses to the state, the magnitude of the problem of traffic crashes clearly exceeded that of index crimes in Michigan in 1988. From a policy perspective, summing the total dollar losses from crashes or crime is of less importance than understanding the costs and benefits of various policy alternatives. This paper therefore concludes with an example of how our cost estimates can be used to compare the costs and benefits of competing policies.

Effects of the 65 mph speed limit on injury morbidity and mortality.

Effective December 1987 and January 1988, the maximum speed limit on rural limited access highways in Michigan was raised from 55 mph to 65 mph. This study examined the effects of the raised limit on injury morbidity and mortality. A multiple time-series design was used, comparing roads where the speed limit was raised with roads where the limit remained unchanged. Data were collected on numbers and rates of automobile crashes, injuries, and deaths from January 1978 through December 1988. Time-series intervention analyses were conducted to estimate effects associated with the speed limit change while controlling for long-term trends, seasonal cycles, and other patterns. Statistical controls were also included for major factors known to influence crash and injury rates. Results revealed significant increases in casualties on roads where the speed limit was raised, including a 19.2% increase in fatalities, a 39.8% increase in serious injuries, and a 25.4% increase in moderate injuries. Fatalities also increased on 55 mph limited access freeways, suggesting that the 65 mph limit may have spillover effects on segments of freeways where the limit was not changed. No significant changes in fatalities or injuries were found on other types of roads. The increased convenience of reduced travel time with the higher speed limit is obtained at a significant cost in terms of injury morbidity and mortality.

Public opinion on alcohol policies.

We conducted a survey of a probability sample of adult residents in the State of Michigan, assessed public support for various alcohol and drinking-driving policy changes, examined how opinions on various policies clustered, and compared Michigan results with national surveys on these issues. Results showed high levels of public support for raising alcohol excise taxes (82%), prohibiting concurrent sales of alcohol and gasoline (74%), administratively suspending drivers licenses of those over the legal alcohol limit (67%), limiting the numbers of alcohol outlets via government regulation (63%), and lowering the legal alcohol limit for drivers to .05 g/dl (55%). Support for these policies was found across all income and education categories, but was significantly lower among frequent heavy drinkers. Policies with lower levels of public support include liability of commercial servers and social hosts, and stricter limits on hours of alcohol sales. Results are clearly relevant to current policy debates, and point to the need for further research on how opinions concerning alcohol policies are interrelated.

Are there really shortcuts? Estimating seat belt use with self-report measures.

We examined the utility of estimating rates of automobile seat belt use with self-report measures. Self-report measures overestimate belt use rates compared to observational surveys of the same population. Laws mandating seat belt use did not substantially affect the degree to which self-reports are upwardly biased. We found self-report measures overestimate observed belt use by 8.9 to 19.4 percentage points or by a factor of 1.2 to 2. Our best estimate is that self-reported seat belt use rates be discounted by 12 percentage points to estimate actual belt use rates.

Macroeconomic conditions and alcohol-impaired driving.

Several distinct bodies of literature indicate that economic factors influence alcohol consumption and driving behavior, including: econometric studies of effects of income and price on drinking, studies of booming and depressed communities, studies of the effects of recession on mental health and studies of the effects of the business cycle on driving patterns. The core research questions of this study were: (1) Does the state of the economy affect the rate of fatal motor vehicle crashes involving drinking drivers? and (2) Is the relationship between the state of the economy and motor vehicle fatalities mediated by effects of economic conditions on amount of alcohol consumed and amount of driving? Data were collected on multiple indicators of economic conditions, alcohol consumption, vehicle miles traveled and rates per population of fatal crashes in the United States on a monthly basis from 1976 through 1985. Nonlinear time-series modeling methods were used to estimate both direct and indirect effects and both coincident and lagged relationships. Results showed that economic conditions significantly influence fatal crash rates both directly and via changes in aggregate amount of alcohol consumed and aggregate amount of driving. However, these interrelationships at multiple lags are complex and not yet fully understood.

An experimental test of risk compensation: between-subject versus within-subject analyses.

This study examined parameters under which risk compensation in driving can occur following the use of safety belts. Risk compensation theories hypothesize that if individuals use safety belts, they will drive in a more risky manner than if they do not use safety belts due to an increased perception of safety. Although the existence of risk compensation in driving has been debated in the literature for many years, the current study was the first experimental analysis of this theory that permitted a controlled examination of both between-subject and within-subject effects. This study required subjects to drive a 5-hp. go-kart around an oval track either buckled or unbuckled in the first of two phases of 15 driving trials. After the first phase the safety condition was switched for half the subjects (i.e., the safety belt was removed from subjects using it or was used by subjects who previously did not use it). Dependent measures included latency for each lap, deviations from the prescribed lane, and perceived safety while driving. The amount of time it took for subjects to travel to the go-kart track and their safety belt use during that trip was also measured. Risk compensation theory was not supported in the between-subject analyses of the research data; however, some within-subject comparisons did demonstrate risk compensation. Subjects who switched from not using the safety belt to using it increased driving speed during the second phase significantly more than subjects who used the safety belt during both driving phases. The study suggested that the occurrence of risk compensation is dependent upon individuals being able to compare the sensations using a safety belt with those of not using a safety belt. Risk compensation did not manifest itself in between-subject studies because this comparison could not take place. The implications of this study to driving automobiles on multi-user roadways is discussed. Suggestions for research to further expand the knowledge about how and when risk compensation occurs are also provided.