Synonyms

Traffic light violation

Introduction

Red light running (RLR) is a violation of traffic rules that endangers the offender as well as other road users. A red light isolation is established when a driver fails to stop at a red signal indication. The crash risk induced by this reckless behavior calls for public intervention. Indeed, RLR can be considered as an external effect imposed upon road crash victims. A social cost is associated with this illegal behavior. The existence of this external effect finds its origin in the fact that the road is used as a common. Several tools are available for policy makers to internalize those costly consequences and enforce traffic light obedience.

Although there is a sizable scientific literature dealing with RLR, researchers generally focus on the predictors of such behavior and the impact of the countermeasures from an engineering and psychological perspective. The economic approach is quite inexistent, so that it is impossible to determine the appropriate intervention and be insured the public policy is efficient. It neglects also the economic dimension of driver’s choice. This entry proposes an economic approach which provides a new perspective for this issue and gives an account of the specialized literature.

An economic approach is required because red light violation as an external effect calls for a public intervention. However, the cost of intervention has to be proportional to the social cost of violations. In other words, the cost element constitutes the crucial dimension not only for framing countermeasures but also for understanding the offender’s choice.

The first section provides an economic explanation of the need for traffic light at intersection through a game theory approach emphasizing upon the needs of cooperation and fairness. The second section deals with an analysis of RLR from the driver’s choice perspective. The consequences of RLR are identified in section “Consequences of Traffic Light Violations”. Section “Regulation of Red Light Violations” reviews the available possibilities for enforcing this safety rule and regulating efficiently the RLRs through the economist’s lens.

Signalized Intersections

Cooperation and Safety at Intersections

Without traffic regulation, the situation of two drivers reaching the intersection from opposite directions can be represented as a noncooperative game (Table 1). If both players stop, they suffer a time loss (−t); if they both go, they suffer a large loss related to a road crash (−a), with a > t > 0. If they make different choices, the stopping driver suffers the time loss, while the other driver breaks even. In this game, the Pareto optimum is reached when drivers take different strategies.

Traffic Lights Violations, Table 1 Game matrix of drivers at intersection
Full size table

In that case, there is no dominant strategy. If the adverse driver is expected to stop (resp. go), the best decision is to go (resp. stop). The absence of unique equilibrium can lead to a suboptimal situation (the crash or the mutual stop). By forcing one driver to stop, traffic light can be considered as an exogenous intervention that imposes one Pareto optimal equilibrium. Then traffic light can find a justification from the economic perspective by making cooperation possible between drivers. Here a better cooperation is related with traffic safety.

Traffic Management: Fairness and Sustainable Cooperation

Different types of traffic regulation are possible for solving the coordination problem at intersection such as putting a stop sign or giving priority to the right. However, the two aforementioned solutions systematically give priority to the same users. For the sake of fairness and efficiency of traffic regulation, traffic light alternates the two equilibria. Drivers from the non-priority roads are not systematically penalized, and waiting time is shared between drivers coming from opposite direction. Thus, traffic light ensures fairness and reciprocity among users. It is an important feature. Indeed, fairness and reciprocity can motivate cooperative behavior (Fehr and Schmidt 2006) and can therefore contribute to driver obedience to this traffic rule. It also makes it a sustainable one.

Traffic management is another main objective of traffic light regulation. Traffic lights are sometimes used for ramp metering, in order to make new entrants wait during congested periods. Others are used for regulating speed in urban areas. Such regulation aims at producing a smoother and calmer traffic among the road users and avoiding congestion. A driver pays a limited period of waiting time in order to enjoy larger gains with a reduced total driving time. Although such traffic management tools aim at reducing road crashes and time wastes, and ultimately the related social costs, not all drivers always abide to these rules. In a sense, RL violation can be considered as a free riding activity. The free rider would like to benefit from traffic safety and management without participating to its funding by obeying the rule. It is also a departure from the Pareto optimal equilibrium defined previously.

Traffic Light Violations

Despite of the related crash risk, red light running (RLR) is not an uncommon event. Retting et al. (1998) observed an average of three violations per hour on two urban intersections in Arlington. Carnis and Kemel (2012) reported very similar results from a field investigation for 24 traffic lights in Nantes (France) and showed that RLR characteristics vary with the type of sites, users, and contexts.

Individual Choice

A RLR is mainly an individual choice. Indeed, the classical school of economics of crime assumes that traffic offenders choose whether to violate the law or not by following utility maximization rules (Becker 1968). For red light violation, the elements of this type of decision can be presented in a decision matrix (Table 2). Stopping at the traffic light results in a sure time loss (−Ct), whereas the consequence of a red light violation is uncertain and event contingent. It depends on the presence of another vehicle crossing the intersection and enforcement (a police officer or camera). If one of these events occurs while the driver decided to run the light, a cost is suffered: (−Ca) for the crash cost and (−Cf) for the cost of a fine, with Ca > Cf > Ct > 0. Ca can also include a penalty for causing the crash by red light violation.

Traffic Lights Violations, Table 2 Matrix of go/stop individual decision at red light
Full size table

There is no dominant strategy and the decision depends on users’ beliefs and preferences. Decision under uncertainty is classically modeled by expected utility. This model assumes that decision makers assign subjective probabilities to events and subjective utilities to consequences and choose the alternative that maximizes the mathematical expectation of their utility. Normalizing the utility with U(0) = 0, a driver is expected to run the light if U(−Ct) < pf × U(−Cf) + pa × U(−Ca), where pf (resp. pa) is the subjective probability of being fined (resp. responsible of a road crash). RLR is thus expected to vary across individuals and contexts depending on attitudes and perceived risks. This framework predicts also that RLR decreases when Pf, Pa, Cf, or Ca increases or when Ct decreases.

Empirical studies bring evidence for most of these predictions. The literature shows that increasing detection probability (by the mean of red light cameras, for instance) reduces RLR (Council et al. 2005). Moreover Carnis and Kemel (2012) show that violation rates are higher during night and low-traffic time periods when crash risk is lower.

The impact of red light duration on RLR was highlighted by Retting et al. (2008). When waiting time is too long, drivers fail to respect it. Guidelines recommend not having red light durations that exceed 2 min (CERTU 2010). Carnis et al. (2012) report field data showing that most RLRs occur during the very first seconds of the red phase, when the violation is the most profitable in terms of avoided waiting time.

Heterogeneity is also expected between users, because of the diversity of individual preferences. Retting et al. (1999) compare authors of RLR accidents to those of other accident types. Males are overrepresented among this population. Red light violators are also younger and more likely to be intoxicated. Porter and England (2000) observed a relationship between RLR and safety-belt use. Propensity to abide to red light also depends on the vehicle type (Carnis and Kemel 2012).

Coping with Dilemma and Interactions Between Drivers

The decision to respect the rule must be taken in a very short amount of time. Indeed, drivers must make the go/stop decision within a few seconds. Because of the urgency dimensions of the decision and drivers’ cognitive limits, illegal actions can sometimes be taken by mistake (Depken and Sonora 2009). Decision to go or stop at light also requires the driver to analyze the situation because the presence of closely following vehicles must be checked. If the decision to stop is likely to trigger a rear-end collision, decision to run the light must be taken. Consequently, illegal decision can be followed in particular situation for avoiding harm and costly consequences. Those aspects are not generally accounted for by the economics of crime framework that assumes that decision makers have time to choose, face clear-cut situations, and feature perfect cognitive capabilities.

The dilemma that faces the driver approaching light received an important attention in the literature (Elmitiny et al. 2010; Papaioannou 2007). The situation in which the driver is unable to stop safely or crossing the intersection at the green light is called the dilemma zone. The dilemma zone is related to the duration of amber light and the approaching drivers’ speed. Shortening this time increases RLR because drivers are not averted that the light will switch. Increasing this time increases the dilemma zone and may increase the number of drivers running amber light. Drivers exceeding speed limits are more likely to run amber and red light. Therefore, the dilemma zone does not only puzzle drivers but network managers as well.

Decision to run or not the red light is not only individual, but it is also impacted by other drivers’ behavior. The choice to commit a RLR takes into consideration the presence of other (preceding or following) drivers. For instance, drivers are more likely to run a light when a preceding driver did so (Elmitiny et al. 2010, p. 110). Observing that the preceding user runs the light may provide valuable information for decision that enforcement is low or nonexistent. Even though following behavior can be rational, it also creates risk of rear-end crash if the preceding driver decides to stop at the traffic light.

Consequences of Traffic Light Violations

Safety Consequences

Red light violation is a major concern for the policy makers because of the number of road crashes and victims involved (McGee and Eccles 2003). From the economist standpoint, road crashes are interpreted as an external effect related to the common use of the road network.

Moreover, the urban intersection implies mainly the involvement of vulnerable users (pedestrians, bicyclers, and motorcyclists). It means also the collision is characterized by a true asymmetrical dimension in terms of vulnerability between the involved users in a traffic collision (for instance, vehicle vs. pedestrian).

Large-scale studies evaluating the prevalence of RLR are not common. Retting et al. (1999) report that accidents occurring at intersections represent 27% of all injury crashes in the United States. Accidents due to RLR are however less frequent. Over the 1992–1996 period, RLR crashes represented 3% of all fatal crashes and 7% of injury crashes on urban roads. Compared to the prevalence of violations, these figures suggest that the collision probability in case of RLR is much lower than one could expect. According to the economic approach, violators may also decide to run red light when the traffic conditions and the visibility minimize crash risk. Carnis et al. (2012) observed that 90% of violations occur in the first two seconds of the red phase, when all lights of the intersection are red.

Paradoxically, a sizable part of intersection crashes derive from red or amber light stopping. Rear-end accidents are indeed not infrequent at signalized intersection. Their number has been found to increase after traffic light camera deployment (Erke 2009). Another frequent type of accident occurring at signalized intersection relates to left turns. Wang and Abdel-Aty (2006) estimate that these accidents rank third after rear-end and angle crashes for 1531 intersections in the state of Florida.

Traffic Regulation Consequences

Traffic regulation is another major objective for installing traffic lights. Therefore, consequences in terms of generated congestion and the related time losses have to be assessed. Time losses can be generated by two types of traffic light violations. First, when the traffic light regulates road access, failure to respect the red light disturbs traffic flow and increases congestion. In this case, the contribution of the marginal violator to congestion is small, but the overall effect can be important when violations are numerous. Second, when RLR occurs during a dense traffic condition, the RL runner can be stuck in the middle of the intersection and can totally freeze traffic on all junctions.

A better respect of red light can help in limiting congestion and save environmental costs related to air pollution. We are not aware of any study evaluating the impact of RLR on time losses, nor environmental costs, due to increased congestion, even if they have to be taken into consideration from an economic perspective.

Regulation of Red Light Violations

Red light violation is a source of external effects. It generates a social cost (mainly associated with the crash costs (material damages and injuries)), which requires internalization. Internalization of this external effect calls for an intervention aiming at the reduction of costs borne by the victims. To mitigate the consequences related to those illegal behaviors, the policy maker defines and implements a public policy. This social regulation intervention can be achieved by two different categories of policies: enforcement and other interventions.

The Enforcement Policy

Enforcing the Highway Code

Becker’s seminal works on the economics of crime show that illegal behavior can be mitigated by implementing an efficient policy of control and punishment (Becker 1968). Both the enforcer and the enforcement authority are concerned by the economic approach to crime. Efficiency of this enforcement policy requires taking into consideration the cost of intervention (respectively the relative costs of detection and punishment) and the social loss related to the harmful consequences of red light violations which could be reduced by deterrence. At the society level, it then becomes possible to determine an optimal deterrent policy associated with an optimal punishment (in terms of intensity of detection and severity of sanction) and an optimal number of violations. Consequently, it is rational from the economic perspective not to enforce all RLRs.

Different Techniques of Production

Different ways exist to enforce traffic light regulation. The traditional approach rests upon the manual detection of offenders by police officers, who monitor and intercept the offenders. This procedure is very costly in terms of time, because it requires a permanent supervision and numerous police officers to be able to catch the offender. In economic terms it is a labor-intensive technique of production. In practice, red light regulation was not especially enforced, because of its high unitary enforcement costs.

Since the mid-1980s, red light cameras (RLCs) have been replacing progressively the traditional enforcement method. This technique of detection can be considered as capital intensive and makes possible a systematic supervision of all the drivers, while minimizing the costs of labor intervention. Automation of traffic safety enforcement is a major trend of those last years, which has to be considered for understanding the spreading of such public programs.

When compared with the traditional approach, the RLC program appears as an efficient way for enforcing the regulation. It presents twofold economic advantages. It reduces substantially the cost of detection and punishment at a given level of traffic. The picture of the offender is automatically processed. The offender is identified through his license plate and receives his traffic infringement notice at home. It permits also to increase substantially the level of detection and punishment. Thousands or millions of tickets can be processed according to the limits of the computer system. In France, the number of RLR tickets was multiplied by 8 after the introduction of RLC. Introduction of RLC programs can be conceived as an innovation lowering the average cost of deterrence and making possible a stricter enforcement of red light regulation by generating scale economies. This cost killing effect explains probably why so many jurisdictions implemented such programs for securing the signalized road intersection (Carnis 2010).

Do RLCs Reduce Crashes?

While several contributions conclude to a positive contribution of RLC by reducing road injuries (Council et al. 2005; McGee and Eccles 2003), others show more debatable effects and question their impact. RLC would yield positive side effects with potential spillover impacts of RLC for other intersections and negative ones by increasing rear-end crashes and all category crashes (Hallmark et al. 2010; Vanlaar et al. 2014). However the gains associated with the reduction of right-angle injury crashes would largely compensate the costs related to the increase of rear-end crashes. More problematic are the recent conclusions of several contributions showing the insignificant impact of RLCs for reducing road crashes (Erke 2009; Høye 2013), contributions which were nevertheless criticized by other scholars putting in question their meta-analysis approach (Lund et al. 2009).

Cost-Benefit Evaluation is Needed

An economic approach to red light violation and regulation becomes particularly necessary when such a public intervention yields opposite and potential adverse side effects. It constitutes a prerequisite for concluding about the economic efficiency of such programs for reducing road injuries at signalized intersection. Proceeding to the economic assessment of RLC programs requires a comparison between advantages and costs. However, only few studies investigated the economic side of red light violations. More problematic is the finding of a careful literature review showing the quasi-generalized absence of economic assessment of RLC programs and rigorous evaluation of safety impacts, so that it is impossible to conclude that such programs are efficient and to determine the scope of the internalization policy (Langland-Orban et al. 2014). In fact, the present evaluative practices of RLC programs reflect both the complexity of evaluation process (non-replication of experiences in controlled laboratory conditions) and the costs of collecting and analyzing the data. It seems also to reflect that policy makers sometimes look for intervention whatever may be their impact or cost, when facing the risk of human injuries.

Public-Private Partnership for RLR Enforcement

The economic approach is particularly relevant when programs are not directly managed by governments. There are several procurement alternatives. Some of them could associate private operator, while some governments outsource the operation of the program (FHA 2005). The total or partial outsourcing of such social regulation activity raised some new issues concerning the possibility for contractor to manipulate the control activity and illegal use of the collected data (CSA 2002). Such situation is typically a principal-agent situation with asymmetrical information. Indeed, one agent is usually more informed than the other and can modulate its efforts. This contractual dimension emphasizes the necessity of a well-designed contract to be insured that private and public interests are aligned (Travis and Baxandall 2011). Indeed, while governments are more interested in maximizing their return in terms of safety impact (public safety hypothesis), the private firms are more concerned by the maximization of profit. Those considerations are quite important, because it could influence the location of radars and their impact for public safety.

Another interesting issue is related to the different payment options for the contractor. Fixed-price payments, fixed monthly payments, per citations payments, payments depending on time worked and materials used, and mixed payments are alternative possibilities. However, it is not clearly determined which type of payment is the most advantageous for the government and the most efficient in terms of welfare for the society. Nevertheless there are strong incentive implications for the different agents, especially here for the policy maker and the contractor. Unfortunately this dimension was not investigated.

RLC programs have to be considered also from the institutional perspective (Carnis 2010). Outsourced programs, contractual dimensions, and financial considerations are important characteristics. A more recent trend fires on the RLC programs. More and more governments turned off their RLC because of uncertain impacts in terms of traffic safety already mentioned. Between 2011 and 2013, it is estimated that 200 RLC programs were turned off in the United States (Slone 2014). The policy maker is reluctant to let a program continue while it could increase rear-end injury accidents and potentially jeopardize human lives. Moreover the court system becomes less supportive of such control system: the judge dismisses charges more often because of erroneous readings and identification of some license plates by the program, which questions its reliability. Another consideration concerns the reduction of revenues associated with the RLC while the costs are increasing. Moreover some citizens assimilate RCL fines as a new tax imposed upon the road user. Garrett and Wagner (2009) concluded that sustained municipalities obey revenue motives concerning traffic enforcement and tickets. RLC would not be exclusively concerned with public safety.

Other Policies

Providing Drivers With Better Information

The drivers’ decision of stopping or running the red light depends on beliefs and preferences. However most of the time the driver is not perfectly informed about the risks involved. Consequently, education and awareness campaign can play a useful role in providing accurate information related with the risk the driver faces when taking an adverse decision (FHA 2005). While education and awareness campaigns are conceived here as a provision for helping him in taking a correct decision, it presents a cost. Such campaigns have to be calibrated so that the costs and returns are in a same magnitude.

Nudging Drivers

Behavior change can also be achieved through nudging. This policy consists of framing the decision context. More precisely, the policy maker can design an environment that induces a promoted behavior. Thaler and Sunstein (2008) provide an illustrative example of such policy in the road safety field. The use of stripes can reinforce the visibility of a potential danger of a portion of a curve for instance (Thaler and Sunstein 2008, pp. 37–39). A review of possible applications of nudges to traffic safety policies is proposed by Avineri (2014). Regarding RLR, installing countdown systems can impact drivers’ behavior and appears as a typical nudging intervention.

No Traffic Light, No Traffic Light Violation

Red light violation depends also on some Highway Code adaptations and road infrastructure context. For instance, in the United States, users are generally allowed to a right turn during red light as long as they leave priority to other vehicles. During low-traffic hours (e.g., night hours), traffic light can be turned out and the right of way applies, avoiding the unnecessary waiting time at the green light phase. In France, an experiment tests the impact of giving to the bicycle user the possibility to cross the intersection at a red light provided that the priority is given to the opposite coming vehicle.

Reframing the context of the driver decision can also require bringing some modifications to the road infrastructure. Engineer investigations showed the influence of the average daily traffic volume, the number of traffic lanes, the left-turn lanes, and the speed limit regulation (Langland-Orban et al. 2014). A radical solution for regulating red light violation would consist of removing signalized intersection. In that case, RLR would disappear because the road infrastructure design makes them impossible. In some ways, it could constitute a kind of situational crime prevention approach. Concretely, it would consist of modifying the access to the road section through ramps or implementing roundabouts.

However such interventions can be very costly, especially in an urban context. Again a reasonable economic approach would consist of comparing costs and gains of different alternatives. This approach permits also to avoid the funnel approach by enlarging the problematic to other issues such as mobility and pollution considerations, emphasizing that red light violation prevention cannot be reduced only to public safety consideration.

Conclusion

The economic approach provides a consistent framework for understanding RLR. It is able to account for both users and policy makers decisions and highlight possible alternatives for intervention. It can also explain why it can be rational for a driver to commit a RLR under certain circumstances, but also why red light regulation is not enforced in some cases.

Economic variables are not only at play for explaining the way drivers choose in particular situations, but the economic consequences related to road crash and traffic congestion have also to be considered for understanding the role of traffic light. Economic valuation appears as a true alternative to the engineering perspective for understanding this issue and promotes different analysis and solutions.

Regulation of RLR is achievable and requires a calibrated enforcement policy. RLC program is a possible solution for enforcing traffic safety rules, but an economic approach is needed for designing correctly the public intervention, which could be assimilated to a particular productive process. Communication campaign, awareness program, and infrastructure modification are other available solutions. Nudging policy appears also as an interesting perspective that can be built upon a behavioral law and economic approach, providing new insights for designing traffic safety rules and enforcement policies.