Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
45 Conclusions Strong state SBU laws are the foundation from which other successful programs can be implemented. Initial laws are largely responsible for elevating usage in the United States from below 15% in 1983 to over 50% by 1992. Early laws have been asso- ciated with a median 9% reduction in fatalities and a median 13% reduction in serious injuries. Since 1992, upgrades in 18 jurisdictions (through February of 2007) have been associated with increases in observed usage that have averaged 13 to 16 percentage points. More importantly, recent upgrades have been found to have had an impact on higher risk motorists, such as young males, rural drivers, occupants of pickup trucks, and drinking drivers. Based on existing studies, primary law upgrades have been associated with a median 7% to 8% reduction in fatalities. Cost savings estimates are as high as $300 million per year, based on impacts reported for some states. If current second- ary law states would enact primary law upgrades, they would likely achieve about a 10 percentage point increase in usage among occupants killed in crashes, a 6% to 7% reduction in occupant deaths and injuries, and an annual cost savings of about $100 million per year. The combination of pri- mary law upgrades and continued implementation of HVE provides the greatest potential for current secondary laws to achieve high usage rates, both observed and among fatal crash victims. HVE efforts have also been associated with significant increases in usage at the local, state, and national levels. At the local level, the Elmira, New York, benchmark program resulted in an immediate 28-point increase in observed SBU. On the statewide level, the North Carolina benchmark program was associated with an immediate 16-point increase in observed usage, a 7% reduction in driver occupant deaths, and cost savings estimated to be as high as $300 million per year. While more recent programs have had more modest impacts, they also have been of lesser intensity and have reached much lower levels of enforcement awareness, compared with the benchmarks. Recent statewide efforts, when supported with paid and earned media, have attained median increases of 7 to 9 percentage points. There is consistent circumstantial evidence to suggest that publicized increases in fines (to approximately $50 plus court costs) and/or imposing penalty points for safety belt violations, if publicized, would increase the impact of cur- rent legislation and enforcement efforts. The research in this area is not nearly as strong as it is for legislation or enforce- ment, but it is consistent. Further, there is evidence of increas- ing public support for larger fines and possibly for penalty points. The past 10 years have witnessed the implementation of a series of successful statewide and national enforcement mobilizations (sometimes implemented in conjunction with primary law upgrades) that have been associated with increases in the national usage rate from about 61% in 1996 to 82% in 2005, an increase of more than 21 points. However, usage remains lower among high-risk nighttime drivers and pas- sengers and it appears that the rate of increase in observed daytime usage is slowing. This suggests a need to reexam- ine all possible means for getting high-risk occupants to buckle up. One of the highest risk groups is represented by occupants who are on the roadways late at night, often trav- eling to or from drinking establishments. Additional empha- sis has recently been placed on reaching this target group. While there have not been a sufficient number of studies of the potential impact of nighttime enforcement efforts, the handful of available studies suggest that such efforts, when accompanied by sufficient publicity, can have an impact on late-night road users, including bar patrons. Two important findings from state and local studies are that: a) daytime enforcement may not affect late-night SBU, and b) nighttime enforcement efforts may affect late-night usage without affecting daytime use. C H A P T E R 5 Conclusions and Discussion
Discussion One objective of this report and the review from which it emanated is to suggest how states at various levels of SBU might make additional progress in increasing usage, either among the population as a whole or, more importantly, among high-risk and crash-involved occupants. As indicated, the larger review suggests that the basic foundation for acceptably high usage rates is a law that allows for standard/primary enforce- ment and a reasonable penalty. With such a law and with sub- stantial media and publicity, enforcement and sanctions have the greatest potential for impact. Currently, in the United States, there is a mixture of primary and secondary laws, modestly high and very low fines, and lit- tle use of penalty points for safety belt law violations. On the other hand, most states have been participating at least annu- ally in enforcement mobilizations and, as a result of such par- ticipation, some of these states have acquired observed use rates of 90% or greater. After participating in HVE mobiliza- tions for several years, some states are experiencing smaller increases associated with such efforts and, as a result, there has been some concern whether or not additional progress in these states is possible, particularly those with usage rates of 85% to 90% and higher. A review of observed usage and usage among crash victims suggests that there is still much progress to be made and much to be gained in terms of reductions in deaths, injuries, and related costs. This position is based on: a) the known effective- ness of legislative and enforcement efforts; b) the potential effectiveness of increased and publicized sanctions; and c) the relatively low rates of usage among crash victims in most states. In those states with the highest usage rates, both observed and among victims, it is likely that there will need to be a refocus of programs to better target the highest risk motorists and perhaps more attention paid to measuring progress in terms of usage among occupants involved in potentially fatal crashes. Of course, any state could benefit from such a refocus, regardless of its current usage rate. A Taxonomy Based on Observed Use Rates Observed SBU has been the primary measure by which most programs have been evaluated over the past 2 decades. Thus, this examination begins with a taxonomy of states based on observed usage. Using statewide reported use rates for 2006,98 there were 11 states with usage of 90% or greater in that year.99,100,101 The median rate in these 11 states was 91% (range: 90% to 96%). All had standard/primary enforce- ment laws and all appear to have participated regularly in recent CIOT mobilizations by intensifying enforcement (with an estimated average of about 11 citations per 10,000 resi- dents per week) and implementing paid media to publicize their enforcement activities (expending an estimated average of 3¢ to 7¢ per resident, per week, based on the various cam- paigns reviewed). These data seem to suggest that there is little potential for further impact in these states. However, a closer look at the median usage rate among occu- pants killed in 2005 (the last year for which FARS data were available at this writing) shows that the median usage among crash victims in these high-use states was only 54% (range: 44% to 68%).102 This translates to a median UPFC of about 71% (with a range of 60% to 81%).103 In three of these states (Hawaii, New Mexico, and Georgia) the FARS rate was less than 50% and the UPFC was less than 67%. Thus, about a third of occupants involved in potentially fatal crashes in these states were still not buckled up. Three of the 11 states (Oregon, Michigan, and California) had FARS rates of 60% or greater (UPFC >75%). Oregon had a FARS rate of 68%, which translated to a UPFC of about 81%. Thus, in these states, about 20% of those involved in potentially fatal crashes were not buckled at the time of the crash. While this group is likely to be very difficult to reach, any impact on its members (as measured by changes in UPFC) is likely to have a direct impact on deaths and serious injuries. As a group, these 11 states (particularly Oregon, Michigan, and California) may have to shift their focus, tactics, and meas- urement approach to impact this high-risk group of holdouts (i.e., the approximately 20% to 30% of crash-involved occu- pants who remain unbuckled). Programmatically, this will likely require enforcement that is more specifically directed at the high-risk driver, who is frequently on the road late at night and often after drinking alcohol. Highly publicized 46 98Based upon SBU in 2006âUse Rates in the States and Territories, Traffic Safety Facts: Crash Stats. Report DOT-HS-810-690. National Highway Traffic Safety Administration, 2007. 99Reported rates are rounded to the nearest whole percentage point. 1002006 usage is coded as not available for three states in the source doc- ument. Two of these states (Arizona and Nevada) reported usage rates greater than 90% in 2005; the third state (Pennsylvania) reported 83% usage in 2005. 101These 11 states with 90%+ usage in 2006 (and their rates) were: Washington (96%); Michigan and Oregon (94%); California and Hawaii (93%); Maryland (91%); and Texas, Georgia, New Jersey, Iowa, and New Mexico (90%). 102Based on 2005 FARS data, which were the latest data available at this writing. 103UPFC in this section is estimated for each group, based on an overall 52% effectiveness estimate for safety belts against deaths (i.e., aggregat- ing usage for all vehicle types and seating position) in 2005.
increases in fines would likely add to the impact in these states, as would the publicized use of penalty points for con- victions.104 From an evaluation standpoint, these states may have to place more emphasis on measuring impact in terms of increases in nighttime usage, usage among crash victims, and/or UPFC. A second group of 18 jurisdictions, those with observed usage between 80% and 89%, includes 17 states and the District of Columbia.105 The median observed rate for these states was 84% in 2006. As expected, however, use among crash victims was considerably lower and there was much variation from state to state, with a median usage rate of only 44% (range: 37% to 55%). The median UPFC was about 63% (range: 55% to 72%). Thus, between 28% and 45% of those involved in potentially fatal crashes in these states were not buckled upâ leaving substantial room for additional gains. Ten of these 18 jurisdictions had standard/primary laws and eight had secondary laws. West Virginia, Delaware, Alabama, and the District of Columbia had FARS rates under 40% and UPFC of less than 59%. In the eight states with secondary laws, top priority should be placed on obtaining a standard/primary enforcement law. This would likely result in an immediate 10-point increase in usage among crash victims (i.e., the FARS rate). Continued conventional enforcement and media efforts would likely result in additional increases. As in the previ- ous category of states, increased usage among high-risk groups would be likely with a focus on nighttime enforcement and/or a publicized increase in fine levels (and/or license demerit points). Also, as with the previous group of states, evalua- tion should include measurement of usage among nighttime road users and/or those involved in fatal and serious injury crashes. All of the remaining states, with the exception of Arizona, Nevada, and Pennsylvania (for which 2006 rates were not avail- able in the source document), currently have observed rates of less than 80%. With primary laws in place, continued enforce- ment (day or night), and fines of at least $30, each of these states should be capable of reaching an observed rate of 85% to 90%. If any would choose to focus on high-risk nighttime drivers and drinking drivers at this time, even higher rates would likely result. A Taxonomy Based on FARS Usage and UPFC One problem with the above taxonomy is the substantial variation in use rates among victims (FARS rates) within each category of observed usage. Assuming that FARS use is accu- rately and reliably recorded, a more relevant taxonomy would be based on FARS data itself. Table 11 shows five groups of jurisdictions based on 2005 usage among crash victims and among the hypothetical population of occupants involved in fatal crashes (UPFC).106 This table also includes relevant infor- mation with regard to law type,107 fine level,108 and observed usage for 1998 (the first full year of mobilization efforts) and 2005.109 Note that while the national observed usage rate was 80% in 2005, the FARS rate was 45% and UPFC was about 63%. Using the FARS rate, the top tier of states includes three primary law states with rates of 60% or greater. (See Appen- dix F for a complete array of states by observed and FARS usage rates.) They are Oregon (68%), Michigan (63%), and Califor- nia (77%). The median UPFC among these states is 78%, rang- ing from 77% in California to 81% in Oregon. All three states had 2005 observed usage rates of 93%. Perhaps the most rele- vant point is that, while observed rates in these states suggest that only 7% of motorists were unbuckled, UPFC rates suggest that between 19% and 23% of persons involved in potentially fatal crashes were unbuckled. Still, as was suggested for states with such high rates, further progress will likely require more specific targeting of high-risk groups, likely nighttime enforce- ment, and monitoring of FARS and nighttime usage rates (see Appendix F for observed/FARS usage distributions). The same is true for the next group, which consists of eight primary law states that had 2005 FARS rates between 50% and 60%. Median FARS usage was 54% (range: 50% to 55%) and median UPFC was 70% (range: 67% to 71%). These rates sug- gest that about 30% of crash-involved occupants were unbuck- led (range: 28% to 37%). Observed usage in this group ranges from 78% in Alaska to 95% in Washington. Based on FARS 47 104While there is reasonably strong public support for primary enforce- ment and increases in fines, there is less support for the assessment of penalty points. 105From highest to lowest usage, jurisdictions that reported 80% to 89% observed usage were: Utah, North Carolina, and West Virginia (89%); Illinois (88%); Delaware (86%); the District of Columbia (85%); Indi- ana, Oklahoma, and Connecticut (84%); Minnesota, Alaska, New York, and Alabama (83%); Vermont and Ohio (82%); Florida (81%); and Colorado and Idaho (80%). 106The UPFC values in this table are estimates, based on a national-level esti- mate of the overall effectiveness of safety belts against deaths (E) across all vehicle types, as calculated in Appendix A. Ideally, effectiveness (E) would be calculated individually for each state, using the state-specific distribu- tion of involvement in fatal crashes by vehicle type and seating position. 107Law type status is as of September 2006. The source is Traffic Tech # 317: Primary Enforcement Saves Lives: The Case for Upgrading Sec- ondary Safety Belt Laws. Traffic Safety Facts. National Highway Traffic Safety Administration, 2006. 108Fine level does not include court costs. Some fines have wide ranges. Minimum fine is shown in this table. Potential for assessing penalty points on license indicated by asterisk. Source of these data is Glass- brenner, SBU in 2005âUse Rates in the States and Territories. Traffic Safety Facts. Research Note, 2005. 109Observed usage is also from Glassbrenner (2005); rates are rounded to nearest full percentage point. Rates for Georgia, Iowa, and Minnesota were updated based on data from NHTSAâs Traffic Safety Facts, March 2007.
and UPFC rates, these states (like Oregon, Michigan, and Cali- fornia) may need to redirect their enforcement, publicity, and penalty efforts to more effectively impact the behavior of young males, nighttime drivers, and drinking drivers. Such a focus should directly impact usage among crash-involved drivers. A side benefit of such an approach would likely be a reduction in alcohol-related deaths, since about 70% of occupants killed in alcohol-related crashes are currently unbuckled. Another ben- efit would be that these states could provide a model for other states still implementing more traditional enforcement efforts. The third group includes 18 states with FARS rates between 40% and 49% (median: 45%). The median UPFC in this group was about 62% (range: 58% to 66%). An average of 38% of crash-involved occupants in this group were unbuckled (range: 34% to 42%). By comparison, observed usage ranged from 74% (in Georgia) to 95% (in Hawaii). This could be considered a transition group, within which there is still room for progress using more general enforcement and publicity efforts. Even if their strategies do not change, many of these states will likely have to monitor usage among crash victims in order to docu- ment further progress. All would benefit from more focused, intense, and publicized enforcement, combined with tougher and more visible sanctions. The fourth group of jurisdictions includes 17 states and the District of Columbia, all of which have usage rates among vic- tims that are below 40% (median 35%; range: 30% to 39%). The median UPFC in this group is 53% (range: 46% to 57%). Observed usage among these jurisdictions ranges from 70% in South Carolina to 94% in Arizona. Clearly, there is much room for additional impact in these states, even with current daytime enforcement efforts. Eleven still have secondary laws and at least six have very low fines of only $10. As a group, there is much potential for impact among these states.110 The final grouping includes states with FARS rates of less than 30% and UPFC rates of 40% to 44%. Three of these four states have secondary laws with fines of under $25. Nearly 48 State La w Fine ($) FARS Use (Est. ) UPFC Obs. Use 1998â2005 State La w Fine ($) FARS Use (Est. ) UPFC Obs. Use 1998â2005 I. States with 60% + Use Among Victims (FARS) IV. States with 30% to 39% Use Among Victim s OR P $94 68% 82% 83% 93% AL P $25 39% 57% 52% 82% MI P $25 63% 78% 70% 93% AZ S $10 39% 58% 62% 94% CA P $20 62% 77% 89% 93% TN P $10+ 39% 57% 57% 74% WI S $10 38% 55% 62% 73% II. States with 50% to 59% Use Among Victims DC P $50* 38% 55% 80% 89% NY P $50* 55% 71% 75% 85% WV S $25 38% 56% 57% 85% AK P $15 55% 72% 57% 78% PA S $10 37% 53% 68% 83% IA P $25 54% 71% 77% 87% WA P $101 54% 71% 79% 95% RI P $75 35% 52% 59% 75% MD P $25 54% 70% 83% 91% KY P $25 35% 53% 54% 67% NJ P $42 53% 69% 63% 86% VA S $25 35% 52% 74% 80% TX P $25+ 52% 70% 74% 90% WY S $10+ 34% 54% 50% N/A NC P $25 50% 67% 77% 87% MO S $10 34% 51% 60% 77% MA S $25 33% 50% 51% 65% III. States with 40% to 49% Use Among Victim s KS S $10 33% 51% 59% 69% HI P $45 49% 66% 81% 95% AR S $25 32% 50% 53% 68% NM P $25* 49% 68% 83% 90% NE S $25 32% 50% 65% 79% IL P $25 49% 66% 65% 86% SC P $10 32% 49% 65% 70% VT S $10 48% 63% 63% 85% NH S $25 30% 46% N/A IN P $25 47% 65% 62% 81% NV S $25 47% 65% 76% 95% V. States with 20% to 29% Use Among Victim s MN S $25 47% 64% 64% 84% MT S $20 27% 44% 73% 80% CT P $37 45% 62% 70% 82% MS P $25 27% 43% 58% 61% UT S $15+ 45% 62% 67% 87% SD S $20 26% 41% 46% 69% GA P $15 44% 61% 74% 90% ND S $20 25% 40% 40% 76% DE P $25+ 44% 60% 62% 84% State Law Fin e FARS UPFC Obs. Use ME S $25+ 43% 58% 61% 76% CO S $15 43% 61% 66% 79% ID S $10 43% 61% 57% 76% OH S $25 42% 59% 61% 79% FL S $30 41% 58% 57% 74% OK P $20 40% 58% 56% 83% LA P $25+ 40% 58% 66% 78% State Law Fin e F ARS (Est.) Obs. Use Notes: - FARS rate is calculated on known use; unknown use is omitted from rate calculation - Observed use and fine data from Glassbrenner (2005) - UPFC calculated by vehicle type; then aggregated (See Appendix A for example of procedure ) - Asterisk in fine column signifies points also assessed - See Appendix F for scatter plot Table 11. Categorization of jurisdictions by usage among victims of fatal crashes (FARS Use) in 2005: Law type, fine level, UPFC, and observed use also included. 110It should be noted that among these states is South Carolina, which implemented a benchmark CIOT program in 2000 and enacted a pri- mary law upgrade in 2005. South Carolina had a FARS rate of only 32% in 2005. Clearly, there have been obstacles to sustaining a high usage rate in this state.
60% of occupants involved in fatal crashes in these four states are unbuckled. In summary of this examination of state usage rates, both observed and among crash victims, there is substantial room for impact in nearly all states if more focus is placed on high- risk occupants who are most likely to be involved in fatal crashes. They include those traveling at night and those trav- eling in vehicles driven by drinking drivers. States with lower FARS rates (e.g., under 40%) should reexamine their efforts to determine why, in spite of reasonably high observed usage rates, more crash victims are not buckled up. As a postscript, it is important to reiterate that, as the infor- mation in Appendix C suggests, there is a strong relationship between alcohol use and the nonuse of safety belts. While it cannot be definitively stated that these are the same groups of road users, there certainly appears to be a great degree of over- lap. Considering the fact that about two-thirds of the victims of alcohol-related crashes are in the drinking driverâs vehicle (mostly the drinking driver himself) and the fact that about 70% are unbuckled at the time of the crash, there is consider- able potential for addressing the alcohol-related fatality issue by getting drinking drivers and their passengers to buckle up. Since both problems peak at nearly identical times (i.e., between midnight and 3 a.m.), there is opportunity for coordinated enforcement efforts. From the standpoint of reducing alcohol-related deaths, a focus on nighttime enforcement could impact the problem by: a) increasing deterrence (since safety belt nonuse is a more observable offense than impaired driving) and b) simply increasing protection among occupants involved in alcohol- related crashes. From the standpoint of reducing unrestrained deaths, focusing enforcement on one of the highest risk, lowest use groups (i.e., drinking drivers and their passengers) would seem to provide maximum potential for directly affecting those most likely to be involved in a crash. Further, such an approach would facilitate manpower and staffing problems that enforcement agencies may face when mounting nighttime enforcement activities. 49
