Elsevier

Accident Analysis & Prevention

Volume 81, August 2015, Pages 134-142
Accident Analysis & Prevention

Acceptance of drinking and driving and alcohol-involved driving crashes in California

https://doi.org/10.1016/j.aap.2015.04.035Get rights and content

Highlights

  • We conducted a cross-sectional survey of weekend, nighttime drivers in California.

  • We explored perceived drinking and driving risk and city alcohol-involved crashes.

  • At low proportions of crashes, lower acceptance of drinking and driving was observed.

  • At high proportions of crashes, lower acceptance of drinking and driving was observed.

  • Risk perceptions influence behavior for those most likely to drink outside the home.

Abstract

Background

Alcohol-impaired driving accounts for substantial proportion of traffic-related fatalities in the U.S. Risk perceptions for drinking and driving have been associated with various measures of drinking and driving behavior. In an effort to understand how to intervene and to better understand how risk perceptions may be shaped, this study explored whether an objective environmental-level measure (proportion of alcohol-involved driving crashes in one’s residential city) were related to individual-level perceptions and behavior.

Methods

Using data from a 2012 cross-sectional roadside survey of 1147 weekend nighttime drivers in California, individual-level self-reported acceptance of drinking and driving and past-year drinking and driving were merged with traffic crash data using respondent ZIP codes. Population average logistic regression modeling was conducted for the odds of acceptance of drinking and driving and self-reported, past-year drinking and driving.

Results

A non-linear relationship between city-level alcohol-involved traffic crashes and individual-level acceptance of drinking and driving was found. Acceptance of drinking and driving did not mediate the relationship between the proportion of alcohol-involved traffic crashes and self-reported drinking and driving behavior. However, it was directly related to behavior among those most likely to drink outside the home.

Discussion

The present study surveys a particularly relevant population and is one of few drinking and driving studies to evaluate the relationship between an objective environmental-level crash risk measure and individual-level risk perceptions. In communities with both low and high proportions of alcohol-involved traffic crashes there was low acceptance of drinking and driving. This may mean that in communities with low proportions of crashes, citizens have less permissive norms around drinking and driving, whereas in communities with a high proportion of crashes, the incidence of these crashes may serve as an environmental cue which informs drinking and driving perceptions. Perceptual information on traffic safety can be used to identify places where people may be at greater risk for drinking and driving. Community-level traffic fatalities may be a salient cue for tailoring risk communication.

Introduction

Alcohol-impaired driving was the cause of nearly 11,000 deaths in 2009, with an estimated $64 billion in associated social costs in 2008 (Compton and Berning, 2009, Shults et al., 2009). This behavior accounts for nearly a third of all U.S. traffic-related fatalities (31%; National Highway Traffic Safety Administration [NHTSA], 2013). Substantial improvements in alcohol-related fatal driving crashes were observed in the U.S. during the 1980s and 1990s (Bergen et al., 2012; NHTSA, 2010; Williams, 2006). It has been estimated that 44% of the reduction in alcohol-related traffic fatalities from 1982 to 2005 can be attributed to alcohol policies; a small percentage can be attributed to reductions in alcohol consumption; and a substantial proportion can be attributed to shifts in the demographic composition. In terms of demographics, there has been an increase in the proportion of female and older drivers who are at lower crash risk than younger men (Voas and Fell, 2011). Since the mid- to late 1990s progress in the reduction of alcohol-related traffic fatalities has stabilized and drinking and driving remains a substantial problem (Williams, 2006). Further progress in preventing alcohol traffic fatalities may require a better understanding of the factors that lead some people to drink and drive.

While drinking and driving could be viewed as an obviously poor choice, there are nuances to how people make decisions. A variety of risks and benefits, in addition to biases, can influence these types of decisions. For some people, benefits associated with drinking alcohol outside the home may far outweigh any perceived risk of drinking and driving. While those with lower educational attainment and minorities are more likely to drink and drive and to be arrested for drinking and driving (Dunaway et al., 2011, Gruenewald et al., 1996), many of the demographic differences may be attributed to differences in alcohol consumption patterns and drinking location preferences (Bergen et al., 2012, Birdsall et al., 2012, Chia et al., 2011, Dunaway et al., 2011, Gruenewald et al., 1996, Gruenewald et al., 2002).

Relevant to risk perceptions, decision theory suggests that the perceived probability of an outcome will have a direct impact on one’s decisions (Turrisi and Jaccard, 1992). The decision to drink and drive may be influenced by the perceived probabilities of the range of possible adverse consequences of such behavior. Differences in risk perception may explain some of the differences in drinking and driving behavior by age and sex. For example, younger drivers and males tend to have lower and less realistic risk perceptions compared with other groups (DeJoy, 1989, Finn and Bragg, 1986, Guppy, 1993).

In general, people adapt their behaviors according to perceived risk (Beck et al., 2009, Janz et al., 2002). Risk perception of being in a crash due to drinking and driving has been associated with various measures of drinking and driving (Bertelli and Richardson, 2008, McCarthy et al., 2007, Turrisi and Jaccard, 1992). Lower risk perceptions for drinking and driving have been associated with poorer driving records in adults and increased self-reported behavior among college students ((McCarthy et al., 2007, Turrisi and Jaccard, 1992). This was further illustrated in a national survey, where again, lower risk perception for drinking and driving were associated with a higher propensity for engaging in drinking and driving (Bertelli and Richardson, 2008).

While perceptions of risk for drinking and driving are associated with various measures of drinking and driving behavior (Bertelli and Richardson, 2008, Dionne et al., 2007, Guppy, 1993, McCarthy et al., 2007, Turrisi and Jaccard, 1992), little is known about how objective environmental-level crash risk measures and individual-level perceived risks are correlated. Some have suggested that traffic injuries may be more salient than enforcement efforts for altering risk perceptions. In bivariable county-level analyses in Maryland (Beck et al., 2009), impaired driving injury crash rates were positively correlated with concerns about drinking and driving as a traffic safety issue. In addition, higher alcohol-impaired fatality rates were positively associated with beliefs about levels of enforcement, which may deter behavior. Driving under the influence (DUI) citation rates were not associated with concerns about drinking and driving as a traffic safety issue or beliefs about levels of enforcement (Beck et al., 2009).

The aim of this paper was to understand how to intervene to reduce drinking and driving by examining how risk perceptions may be shaped by objective environmental-level measures of crash risk and by relaxing the linear assumption. A person’s knowledge of their environment is affected by both the external reality and their perceptions of that environment. Understanding how people respond to cues in their environment can inform tailored educational efforts. However, very few studies have examined the relationship between an objective environmental-level crash measure and corresponding risk perceptions. We extend previous findings by analyzing the relationship between residential city-level alcohol traffic crashes and individual-level risk perceptions while controlling for factors that may bias risk perceptions. We further add to the literature by examining how alcohol traffic crashes and related risk perception influence self-reported drinking and driving behavior.

Section snippets

Material and methods

The current study is nested within a broader project, the California Roadside Survey, which consisted of a roadside survey in 9 communities throughout California. The overall objective of the primary project was to estimate the prevalence of substance use and driving among California weekend nighttime drivers. This study took place June–August 2012.

Descriptive

The analysis cohort of 1147 participants resided in 170 California cities or places. The number of participants per city ranged from 1 to 182 with an average of 6.7 participants per city.

Frequency distributions for categorical variables and means and standard deviations for continuous variables are shown in Table 1, overall and stratified by low (lower quartile), moderate, and high (upper quartile) proportion of alcohol-involved crashes in a city. In this sample, the average age was 33, 40%

Discussion

In a sample of weekend nighttime drivers in California, nineteen percent were somewhat accepting of drinking and driving, four percent reported drinking and driving in the past year, two percent had a BAC of 0.05 g/dL or higher at the time of survey, and the average city proportion of drinking and driving traffic injury crashes was nearly twelve percent. In this exploratory study, a non-linear relationship between drinking and driving crash exposure and acceptance of drinking and driving was

Conclusion

To these authors’ knowledge, this is one of the few studies to evaluate the relationship between objective and subjective crash risk for drinking and driving. Traffic crashes are typically studied to address safety issues and perceived traffic safety is sometimes studied to understand barriers to active modes of transportation (Cho et al., 2009). It has also been suggested that perceptual information on traffic safety can also be used to identify “problems waiting to happen” (Schneider et al.,

Acknowledgements

The main study was a collaborative effort of the California Office of Traffic Safety, the University of California at Berkeley Safe Transportation Research and Education Center, and the Pacific Institute for Research and Evaluation and was funded as a part of the California Traffic Safety Program through the support of the California Office of Traffic Safety and the National Highway Traffic Safety Administration. Dr. MacLeod was supported by award number T32AA007240, Graduate Research Training

References (42)

  • K.H. Beck et al.

    The relationship between impaired driving crashes and beliefs about impaired driving: do residents in high crash rate counties have greater concerns about impaired driving?

    Traffic Inj. Prev.

    (2009)
  • Berkowitz, A., 2004. The Social Norms Approach: Theory, Research, and Annotated Bibliography. Retrieved from:...
  • A.M. Bertelli et al.

    The behavioral impact of drinking and driving laws

    Policy Stud. J.

    (2008)
  • W.C. Birdsall et al.

    Alcohol-impaired driving: average quantity consumed and frequency of drinking do matter

    Traffic Inj. Prev.

    (2012)
  • Centers for Disease Control and Prevention, 2011. Drinking and Driving: A Threat to Everyone. Retrieved from:...
  • G. Chia et al.

    Gasoline prices and their relationship to drunk-driving crashes

    Accid. Anal. Prev.

    (2011)
  • Compton, R., Berning, A., 2009. Results of the 2007 National Roadside Survey of alcohol and drug use by drivers....
  • G. Dionne et al.

    Predicted risk perception and risk-taking behavior: the case of impaired driving

    J. Risk Uncertainty

    (2007)
  • K. Dunaway et al.

    Alcohol-impaired driving

  • Enders, P.B., 2011. Binary_mediation. Retrieved from:...
  • D.C. Furr-Holden et al.

    Toward national estimates of alcohol use disorders among drivers: results from the National Roadside Survey Pilot Program

    Traffic Inj. Prev.

    (2009)
  • Cited by (18)

    • Deterring illegal smartphone use while driving: Are perceptions of risk information associated with the impact of informal sanctions?

      2022, Accident Analysis and Prevention
      Citation Excerpt :

      Associating the behaviour with drink-driving and cannabis use may be effective in shifting attitudes. The challenge, however, is to develop messages that provide meaningful, relevant (Urschler et al., 2019) and coherent signals to a specific group or individual (MacLeod et al., 2015). Yet, having the ability to target a particular message to a specific individual will have the greatest impact, and although we show differences between genders, empirical testing should acknowledge the variety of socio-demographic factors, such as class, race, previous experience and attitudes to risk, and the differences within those groups, that impact both perceived effectiveness of a risk message and the impact on perceptions of informal sanctions.

    • Objective and perceived risk in overtaking: The impact of driving context

      2021, Transportation Research Part F: Traffic Psychology and Behaviour
      Citation Excerpt :

      These patterns could highlight scenarios related to crash risk in overtaking, while user input can further explore their impact on the risk experienced. While the use of crash data together with survey responses has been previously applied in order to examine the relationship between objective and subjective risk under different conditions and driving contexts (e.g. Macleod, Karriker-Jaffe, Ragland, Satariano, Kelley-Baker, & Lacey, 2015), limited studies explored this relationship in the context of AVs. Driving conditions highlighted as risky through crash data can constitute test scenarios to explore perceived safety and overall user acceptance when combined with subjective risk evaluations.

    • The specific role of spatial orientation skills in predicting driving behaviour

      2020, Transportation Research Part F: Traffic Psychology and Behaviour
      Citation Excerpt :

      This datum, more than due to an increased attitude toward risk, might be associated with degradations in sensory, cognitive, and physical capabilities (e.g., Zahabi, Abdul Razak, Shortz, Mehta, & Manser, 2020). Gender is another predictor of risky driving behaviour: for instance, males usually drink more heavily and perceive lower risk while driving after drinking (MacLeod et al., 2015; Cordellieri et al., 2016; Navas et al., 2019). In addition, males show higher levels of sensation seeking – the tendency to engage in intense, novel, or potentially dangerous (arousing) activities – (for a meta-analysis see Cross et al., 2011; see also Lazarus et al., 2019; Lucidi, Bosco, Mallia, & Setti, 2020), whereas females are less tolerant to delayed rewards (Navas et al., 2019).

    • Detecting DUI (Non) deterrence: A macro-methodology to uncover “restrictive v permissive” county jurisdictions in California

      2019, Accident Analysis and Prevention
      Citation Excerpt :

      Historically, alcohol and drunk-driving are also topics of extreme divisiveness, and are still evident in the political geography of “wet/ dry/ moist” counties 80 years after Prohibition (Frendreis and Tatalovich, 2010) and in modern consumption and accident patterns (Baughman et al., 2001; Berman et al., 2000; Hilton, 1988, and Greenfield and Room, 1997). How the prevailing, local accident environment affects residents’ drinking-and-driving behaviors has only recently been tested (MacLeod et al., 2015). What does it take to induce (long-term) change in community norms?

    View all citing articles on Scopus
    View full text