UC Berkeley

Traffic crashes are a major cause of preventable deaths and pose significant social and economic burdens. Substantial reductions in alcohol-related fatal traffic crashes were observed in the U.S. during the 1980s and 1990s (Voas & Fell, 2011). However, progress on reducing alcohol-related traffic deaths has slowed in recent years and remains a substantial problem (Bergen, Shults, Beck, & Qayad, 2012; Fell & Voas, 2013; Williams, 2006). Further reductions in alcohol-related traffic fatalities may require a better understanding of how people make decisions about travel choice in this context and how these decisions are shaped. Among drinking and driving efforts, there has been little focus on understanding travel choice and the attitudes towards those travel alternatives. Further, lower risk perceptions of being caught and/or being in a crash have been associated with a greater risk for various measures of drinking and driving (Bertelli & Richardson, 2008; Dionne, Fluet, & Desjardins, 2007; Guppy, 1993; McCarthy, Lynch, & Pederson, 2007; Turrisi & Jaccard, 1992; Voas, Holder, & Gruenewald, 1997). Some have suggested that enforcement works indirectly through risk perceptions (Bertelli & Richardson, 2008). It has also been suggested that the effects of checkpoint enforcement of these policies are temporary (Turrisi & Jaccard, 1992) and that traffic injuries may be more salient than enforcement for modifying risk perceptions (Beck, Yan, Wang, Kerns, & Burch, 2009). However, little is known about how perceived crash risk and objective crash risk are correlated. Further, understanding the behavioral impact of laws requires testing enforcement and perceptions (Bertelli & Richardson, 2008).

This study utilized data from a roadside survey that took place on weekend nights June-August 2012 in 9 jurisdictions throughout California. This study reports on 2 analysis cohorts in an effort to understand the perceptions of risk, drinking and driving, and transportation mode choice and how these are shaped by the environment. The first cohort consists of 689 respondents who reported drinking alcohol in the past year and who completed a mode choice survey for a recent drinking occasion outside the home ("choice" cohort). The second consists of 1,147 respondents who reported perceptions of risk for drinking and driving and resided in California ("risk" cohort). This study reports on a particularly relevant segment of the population for this issue, weekend nighttime drivers.

While drinking and driving may be viewed as a failure to act in one's best interest, it is acknowledged that a variety of risks and benefits, in addition to biases and emotions, are associated with decisions. Using a revealed preference approach to reduce cognitive incongruence, a survey was administered to identify travel choice home for a recent drinking occasion. To better understand the impact of residential accessibility (i.e. access to amenities and destinations) and perceptions of risk on travel preferences and choices, hybrid choice modeling was conducted. Choice models are flexible and allow decision makers to weight aspects of choice differently, assuming utility maximization. Hybrid choice modeling incorporates psychological factors with discrete choice modeling to better represent the choice process (Ben-Akiva et al., 1999; Ben-Akiva et al., 2002). In the first cohort, nearly a third reported being a drinking-driver for a recent drinking occasion.

In an effort to better understand how risk perceptions are shaped, the relationship of environmental cues and risk perceptions for drinking and driving were evaluated and how these, in turn, impact drinking and driving behavior were explored using population average models. In the second cohort of abstainers and drinkers, 4% reported driving at least once in the past year when they had had too much to drink. In addition, 19% were somewhat accepting of drinking and driving and 66% perceived high arrest likelihood for drinking and driving.

The role of driving is deeply embedded in our society. Reducing reliance on personal driving has numerous health implications, including, traffic safety. However, the results of this research suggest that some forms of the built environment may encourage driving. The results also suggest that the environment may influence drinking and driving through different mechanisms and that different perceptions of risk influence pathways to drinking and driving. Local deterrence efforts may reduce drinking and driving among those most likely to be concerned about enforcement. In addition, the local environment may provide cues for risk perceptions or indicate populations where the norms are less permissive of drinking and driving. Perceived level of service (i.e. travel time and travel cost) and convenience and safety attitudes towards travel alternatives impacted mode choice in this study. Among the sample who drink alcohol, perceived high arrest likelihood impacted attitudes towards travel alternatives. Among a subset of drivers most likely to drink outside the home, acceptance of drinking and driving was associated with drinking and driving while perceptions of arrest were not.

To this author's knowledge, this is one of the few, if only, studies to evaluate both the environmental and perceived risk impact on travel choice and transportation attitudes in the context of drinking alcohol outside one's home. There are nuances to how people make decisions. The results of the present study add to the previous research by examining attitudes towards other travel alternatives as part of the behavioral process. Understanding human behavior is crucial for developing effective community, public health, and transportation planning. This research has implications for city and transportation planning, risk communication, and enforcement. Transportation planning could improve level of service for leisure, evening trips and increase the cost of parking to discourage car use in accessible areas. However, with the potential for greater alcohol consumption, the displacement of alcohol problems in communities should be considered. Enforcement and sobriety checkpoints require substantial resources. Traffic crashes may be a salient cue to impact risk perceptions and supplement efforts. Social media could be leveraged for both risk communication and for facilitating ride sharing. Regularly assessing risk perceptions can assist in benchmarking traffic safety efforts and in understanding how drivers, including new cohorts, respond to these efforts.