Chapter Two. Selecting Interventions

This chapter explains our approach to selecting interventions to be included in the tool. Our first task consisted of identifying evidence-based interventions to be studied in the course of the project. The project evaluates interventions that seek to change driver and passenger behavior, such as wearing seat belts and helmets, avoiding drunk driving, and obeying speed limits and traffic signals. We did not consider any interventions that make vehicles safer in crashes without any action by the driver (such as air bags) or those that change the design of roadways.

Given the nature of the project (i.e., assisting states in determining the most cost-effective interventions), we proposed the following criteria to select these interventions:

  • the likely magnitude of the effect (i.e., potential for high impact). This study focused on interventions with sufficient empirical evidence to suggest a significant reduction in the number of motor vehicle–related injuries or deaths. We first identified interventions that seemed to have a solid evidence base and later reviewed specific studies to determine the reduction in injuries and crashes (see Chapter Four for more details).
  • the number of states that have implemented the intervention (i.e., potential for broader implementation). If an intervention is already in widespread use, the potential for additional impact is limited. The study included interventions that have been implemented in no more than two-thirds of the states; most have been implemented in fewer than half the states.
  • states’ ability to implement an intervention. Because the goal of this project was to assist state decisionmaking, interventions that would be implemented exclusively at the federal or local level would be ranked low.

Following our initial review of the NHTSA-sponsored Countermeasures That Work report (University of North Carolina [UNC] Highway Safety Research Center, 2011),4 which describes and assesses 131 interventions, we identified a preliminary list of interventions that meet the three criteria.5 The Countermeasures report ranked all interventions on a scale of 1 to 5, with 5 considered the most effective. We retained only the 38 interventions that rated 4 or 5. Of these, we eliminated 13 whose use was deemed “high” (that is, all or almost all states already use these interventions), five that would be implemented by actors other than states, and another five that would be difficult to study either because they are seldom implemented as stand-alone measures or for which data appear to be limited. We made iterative changes to the initial list in conjunction with CDC staff, bringing the number of interventions to 12; details of those changes are provided in Appendix A.

As part of a follow-up project, we added two other interventions. Again we began with the Countermeasures That Work report (which had, by that point, been updated; see Goodwin, Kirley, et al., 2013) but looked for promising policies addressing four areas that were not part of the original 12 interventions: older drivers, increased fines, bicyclists and pedestrians, and cell phone and texting bans. According to research not cited in Goodwin, Kirley, et al., 2013, we determined that in-person license renewal appeared to be an effective policy and that higher seat belt fines met our original criteria.

The final list is shown in Table 2.1. Brief definitions are provided below, and fact sheets with more detail on each intervention (including evidence of effectiveness) are provided in Appendix B. In addition, in Chapter Three, we describe how we operationalized the interventions for the purposes of estimating the implementation costs.


Table 2.1 Interventions for Analysis
Name of Intervention (Short Name) Description Effectiveness Use Comments
Automated red-light camera enforcement (red-light camera) Automated red-light camera enforcement, more commonly called red-light cameras, is used to capture an image of a vehicle whose driver fails to stop for a red light. Tickets are generally sent to offenders by mail. 5 Medium Retained despite some conflicting evidence about effectiveness
Automated speed-camera enforcement (speed camera) Automated speed-camera enforcement, often called speed cameras, captures an image of a vehicle whose driver is driving in excess of the posted speed limit. Unlike red-light cameras, which are deployed only at intersections, mobile speed cameras are often used to cover multiple road segments. 5 Medium None
Alcohol interlocks Alcohol interlocks, also called ignition interlocks, are devices that prevent a vehicle from starting until the driver has blown into a tube and determined that his or her BAC is below the allowable level set by the state (0.02 in most jurisdictions). This intervention calls for interlocks to be installed on the vehicles of convicted repeat DWI offenders, as well as high-BAC and first offenders, depending on state legislation. 5 Medium Legal in all states; states vary in whether they are mandatory and under which circumstances.
Sobriety checkpoints At a sobriety checkpoint, teams of police officers stop cars at a specific location to check drivers for alcohol levels. States generally publicize such events to discourage drivers from drinking, particularly during times when drunk driving is more common than usual (such as holiday weekends). 5 Medium Legal in most states but limited use
Saturation patrols Saturation patrols consist of an increased police presence in selected locations where they patrol the area looking for suspicious driving behavior. In contrast to sobriety checkpoints, they do not stop every vehicle. 4 High Legal in all states; can be used in states that prohibit checkpoints
Bicycle helmet laws for children (bicycle helmet) To reduce the likelihood of trauma to the head and its related consequences, bicycle helmet laws mandate the use of helmets by children while they are riding bicycles. 5 Medium None
Universal motorcycle helmet laws (motorcycle helmet) This law requires all motorcyclists, regardless of age or experience level, to wear a helmet the meets safety standards set by DOT. These laws contrast with partial helmet laws, which typically apply only to riders below a certain age. 5 Medium None
Primary enforcement of seat belt laws (primary enforcement of seat belt laws) States with seat belt laws vary in their enforcement. A primary law allows police to ticket an offender exclusively for not wearing a seat belt. A secondary law allows police to write a ticket for not wearing a seat belt only if the driver has been pulled over for a different offense. 5 Medium None
High-visibility enforcement for seat belts and child restraint laws (seat belt enforcement campaign) High-visibility enforcement is a technique that combines intense enforcement over a fixed period (for example, one or two weeks) with a publicity campaign. A campaign focused on restraint use generally includes all forms of restraints: seat belts, child safety seats, and booster seats.a 5 Medium This combines four previously separate interventions that all had the same ratings.
License plate impoundment This intervention requires a driver who has been convicted of DWI to surrender the vehicle’s license plate, which is either impounded or destroyed. In some jurisdictions, the license plate is not physically removed; rather, officers place stickers on the license plate to indicate that it is invalid. The stickers are designed so that, if someone tries to remove them, they leave a visible pattern on the plate. Because it is relatively easy for police to observe whether a vehicle has a license plate or the stickers, this intervention deters convicted DWI offenders from driving that vehicle. 4 Medium None
Limits on diversion and plea agreements (limits on diversion) Although all states have penalties for DWI, many states have additional programs that allow some offenders to be diverted out of the normal procedures or to plead guilty to a lesser offense and receive a lighter sanction. These programs are most often targeted at first-time offenders, with the goal of reducing the DWI case load by diverting people who are thought to be unlikely to reoffend. Limits on diversion and plea agreements would increase the number of DWI arrestees convicted of more-serious DWI-related charges. 4 Medium None
Vehicle impoundment This intervention results in the vehicle of a DWI offender being confiscated for a period of time and stored in a public impound lot. An offender can either reclaim or surrender his or her vehicle when the impoundment period ends. 4 Medium Ratings based on all vehicle and license plate sanctions combined
In-person license renewal This intervention requires all drivers over age 70 to renew their driver’s licenses in person at a department of motor vehicles instead of using mail-in or online renewal. 2 Medium None
Higher seat belt fines This intervention adds $75 to a state’s existing fine, which represents a significant increase over existing seat belt fines in most states. 4 Low None

SOURCES: Effectiveness and use ratings from UNC Highway Safety Research Center, 2011; Goodwin, Kirley, et al., 2013.

NOTE: BAC = blood alcohol concentration. DWI = driving while intoxicated. DOT = U.S. Department of Transportation. State terminology varies; a DWI charge against a drunk driver is the same as a charge of driving under the influence (DUI). For the sake of consistency, this report uses DWI. The short name is the same as the intervention name in the tool.

a Child restraint includes both child car seats and booster seats, For simplicity, we refer generally to child restraints.


All of these interventions can be implemented at the state level. In some cases, though, the state may pass a law allowing the use of the intervention (e.g., red-light cameras) but leave it to individual jurisdictions within the state to decide whether to implement it. For the purposes of this tool, we assume a statewide implementation of the interventions. If the state is not considering full implementation, it will need to scale the estimates accordingly. The tool allows the user to conduct analyses using different assumptions about the costs and effects of the intervention if he or she wants to consider a lower level of implementation.



  1. This was the most recent version of the report when we conducted our search.
  2. Although we also reviewed interventions cataloged in Governors Highway Safety Association, undated, and Child Injury Prevention Tool, undated, none of these additional interventions was selected.