This article was downloaded by: [Universitat Politècnica de València] On: 28 October 2014, At: 07:45 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Traffic Injury Prevention Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/gcpi20
The Effects of Primary Enforcement Laws and Fine Levels on Seat Belt Usage in the United States a
b
b
c
James L. Nichols , A. Scott Tippetts , James C. Fell , Angela H. Eichelberger & Philip W. d
Haseltine a
Nichols & Associates, Vienna, Virginia
b
Pacific Institute for Research and Evaluation, Calverton, Maryland
c
Insurance Institute for Highway Safety, Arlington, Virginia
d
Haseltine Safety Consulting, LLC, Manteo, North Carolina Published online: 27 May 2014.
Click for updates To cite this article: James L. Nichols, A. Scott Tippetts, James C. Fell, Angela H. Eichelberger & Philip W. Haseltine (2014) The Effects of Primary Enforcement Laws and Fine Levels on Seat Belt Usage in the United States, Traffic Injury Prevention, 15:6, 640-644, DOI: 10.1080/15389588.2013.857017 To link to this article: http://dx.doi.org/10.1080/15389588.2013.857017
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Traffic Injury Prevention (2014) 15, 640–644 C Taylor & Francis Group, LLC Copyright ISSN: 1538-9588 print / 1538-957X online DOI: 10.1080/15389588.2013.857017
The Effects of Primary Enforcement Laws and Fine Levels on Seat Belt Usage in the United States JAMES L. NICHOLS1, A. SCOTT TIPPETTS2, JAMES C. FELL2, ANGELA H. EICHELBERGER3, and PHILIP W. HASELTINE4 1
Nichols & Associates, Vienna, Virginia Pacific Institute for Research and Evaluation, Calverton, Maryland 3 Insurance Institute for Highway Safety, Arlington, Virginia 4 Haseltine Safety Consulting, LLC, Manteo, North Carolina
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
2
Received 11 February 2013, Accepted 15 October 2013
Background:Seat belt use in the United States increased from 11 percent in 1979 to 86 percent in 2012. Most of this increase has been attributed to seat belt laws, primary law upgrades, and highly visible enforcement. There has been less research on the effect of fines on seat belt usage. Methods: We examined law type and fine levels as predictors of seat belt use among fatally injured occupants of passenger vehicles from 1997 through 2008 using data from the Fatality Analysis Reporting System (FARS). Fine levels used were the statutory maximum base fines for a first offense. Results: Having a primary enforcement law was associated with a 9 to 10 percentage point increase in seat belt use. An increase in the fine amount, from the current median level of $25 to a level of $60, was associated with a 3 percentage point increase in usage. An increase in fine from $25 to $100 was associated with a 6- to 7-point increase. Such increases were in addition to the effects of a shift from secondary to primary enforcement. Discussion: The results of this study suggest that, in addition to current emphases on primary law upgrades and high-visibility enforcement of seat belt usage, increasing fine levels provides another viable strategy for increasing seat belt use. In addition, based on these results, states should consider publicizing such increases just as they publicize enforcement efforts. Keywords: seat belt usage, primary enforcement laws, seat belt enforcement mobilizations, seat belt citations, fine levels
Introduction Increases in Seat Belt Use Based on the results of daytime observational surveys, seat belt use in the United States has increased by more than 70 percentage points over the past 3 decades, from 11 percent in 1979 to 86 percent in 2012 (Phillips 1980; Pickerell and Ye 2008). Nearly all of this increase has occurred since 1984, when New York enacted the first mandatory seat belt use law and nationwide usage was 14 percent (Goryl and Cynecki 1985). The most abrupt increases in use occurred from 1984 through 1987 (from 14% to 42%; +14 percentage points), when 30 states and the District of Columbia enacted seat belt laws. Since 1990, much of the belt use increase, from 49 percent to more than 80 percent, was likely associated with a combination of highvisibility enforcement (HVE) and primary law upgrades, the
Managing Editor David Viano oversaw the review of this article. Address correspondence to James C. Fell, Pacific Institute for Research and Evaluation, 11720 Beltsville Drive, Suite 900, Calverton, MD 20705. E-mail: [email protected]
latter of which provide police officers with the authority to stop a vehicle and issue a citation based solely on an observed seat belt violation. Secondary enforcement laws require an officer to first observe some other traffic law violation before stopping a motorist for seat belt nonuse. High-Visibility Enforcement HVE mobilizations are characterized by vigorous enforcement of seat belt laws combined with publicity about the enforcement. These mobilizations, which typically involve intensified enforcement for 2 weeks, were first initiated statewide and nationwide in the early 1990s, but nationwide efforts became more common after 1996. Key enforcement events since 1990 include (a) a 2-year national enforcement program (i.e., the 70% by ‘92 program) launched in 1991; (b) the first statewide Click It or Ticket (CIOT) program launched in North Carolina in 1993; (c) more than 20 statewide Occupant Protection Special Traffic Enforcement Program grants and demonstrations implemented from 1993 through 1998; (d) the nationwide Operation ABC (America Buckles Children) mobilizations launched in 1997; (e) a series of statewide and regional CIOT programs implemented from 2000 through 2002;
Effects of Laws on Seat Belt Use
641
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
Fines
Fig. 1. Effects of laws and fines on seat belt use.
and finally (f) National CIOT mobilizations that replaced Operation ABC in 2003 and have continued through 2013. Nearly all of these efforts have been formally shown to be associated with relatively immediate increases in public awareness of enforcement activity and with significant increases in observed seat belt usage. Perhaps the best example of a high gain in seat belt usage is provided by the 16 percentage point statewide increase in observed seat belt use associated with the 2-year (1993–1994) CIOT program in North Carolina (Williams et al. 1996). Over time, however, baseline usage rates have increased and, as a result, the magnitude of observed gains has declined (Nichols et al. 2007; Nichols et al. 2009; Solomon et al. 2007; Solomon et al. 2002; Tison et al. 2008). Considerable evidence from state (and local) demonstrations suggest that, although much of the effect associated with HVE is retained for several months, usage nearly always declines over time (e.g., see Solomon et al. 1999). Such “decay” can be minimized by repeated waves of enforcement. Seat Belt Law Upgrades After 1990, there was both an increased emphasis on HVE and important changes in seat belt laws. Twenty-three states upgraded their laws to allow for primary (i.e., standard) enforcement procedures. California implemented the first upgrade, which was followed by 22 additional upgrades through June 2009. The greatest long-term increases in observed seat belt usage among the 15 states that participated in early HVE programs were found in states that had upgraded their seat belt laws (Nichols and Ledingham 2008). This suggests that the combination of primary law upgrades and HVE was a powerful factor for increased usage. As with HVE programs, the evidence shows that upgrades in seat belt laws implemented from 1990 through 1999, when baseline usage rates were lower, had a greater effect on observed usage than upgrades implemented since 2000. However, there is some indication that the more recent law upgrades increased belt usage among front-seat occupants killed in crashes, at least to the same extent, as earlier upgrades (Nichols and Ledingham 2008). One advantage of a law upgrade, relative to an HVE mobilization, is that it is always in effect, whereas a mobilization must be reinitiated periodically to maintain its effectiveness.
Relative to law upgrades and high-visibility enforcement, the impact of increasing fines to increase seat belt use has had relatively little focus. Statutory fines have increased somewhat over the past 15 years, with the average such fine increasing from about $20 in 1995 to slightly more than $30 in 2009. The median statutory fine level, however, has remained at $25. Data from an early study of safety belt laws showed that, when a fine was in effect when a primary law was implemented, observed seat belt use increased by an average of about 38 percentage points. When no fine was in effect, such as during a warning-only period, the average increase was only 23 percentage points—about the same as the change associated with a secondary law with or without any sanctions in effect (Campbell et al. 1987). Winnicki (1995) reported that the largest increases in seat belt usage among occupants of passenger vehicles involved in potentially fatal crashes (a rate that is based on data obtained from the Fatality Analysis Reporting System [FARS] and on the estimated effectiveness of seat belts when they are worn) were in primary law states where fines became effective within 4 months after implementation of the law. The median increase was 21 percentage points in those states. Winnicki also reported that each $10 increase in fine level was associated with about a 7.4 percent increase in usage among occupants involved in potentially fatal crashes. A third example of the potential impact of fines was provided by the state of Washington when it enacted a primary law upgrade in 2002. This upgrade, coupled with highly publicized enforcement efforts and increased sanctions for all traffic law violations, significantly increased the observed belt use. The new fine for a seat belt violation was $101, compared to the previous fine of $86. This increase in fines was publicized extensively, including messages on road signs. Salzberg and Moffat (2004) concluded that the publicized increase in fines likely played a role in the observed increase in belt use in Washington. The most recent study of the effect of fine levels on seat belt use comes from a 2005 study by Houston and Richardson. These researchers examined annual statewide observed seat belt use from 1991 through 2001, using time series, crosssection analyses similar to our approach in this study. They found primary laws to be associated with a 9.1 percentage point increase in observed use, compared with secondary laws, and they found that fine level was associated with higher use. They estimated that the $25 median-level fine was associated with a 3.8 percentage point increase in seat belt use compared to no fine, apart from any effect associated with enforcement type. For this study, our main objective was to determine the relative influence of primary laws and fines on seat belt use rates, while controlling for levels of enforcement and media based on available documentation, primarily data collected during annual mobilizations. Recent studies have confirmed that the type of law, intensity of law enforcement, and media publicizing the enforcement effort are important factors related to increased seat belt use. Fewer studies, however, have looked at the role of fines in increasing seat belt use.
642 Method
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
Periods of Study The first study period examined included the years from 1997 through 2002, which were characterized by primary law upgrades in 11 states and by annual enforcement efforts organized and supported by the Air Bag and Seat Belt Safety Campaign (i.e., Operation ABC Mobilizations). The second study period was from 2003 through 2008, which included primary law upgrades in 8 states and annual CIOT mobilizations, supported with innovative grant funds from the Transportation Equity Act for the 21st Century. During the first period, baseline usage rates were lower than during the second period and, although the legislative and enforcement activity seemed comparable, more complete enforcement data were available during the latter period. Whereas the first period was limited to the number of law enforcement agencies participating in the mobilization and reporting their activities, the second period benefited from having citation rates issued during the mobilizations. These data were available for nearly every state and study year. Separate models and analyses were developed for each period, as well as for both periods combined.
Nichols et al. killed in passenger vehicles (i.e., noncommercial, nonindustrial vehicle body types, excluding motorcycles and off-road vehicles). These data were obtained from the NHTSA’s FARS. A more complete study was published as an NHTSA research report by the NHTSA (Nichols et al. 2010). It included analyses using both observed seat belt use and seat belt use among occupants of passenger vehicles involved in fatal crashes. In part because the outcomes were similar for both analyses, only the analyses of usage among unbuckled occupants in fatal crashes are reported in this article. Data regarding fine levels, enforcement agency participation, citations, and media expenditures during the mobilization periods were obtained from a comprehensive review of historical reports provided by the NHTSA, the Governor’s Highway Safety Association, and the Insurance Institute for Highway Safety. Analyses
The baseline for the first study period was the 2 years (1995–1996) immediately preceding the first study period. The baseline for the second study period was the 3 years (2000–2002) immediately preceding the second study period. These were the last 3 years before the start of Transportation Equity Act for the 21st Century funding for CIOT mobilizations. Thus, 13 years of FARS data (1995–2007) were included in this longitudinal-panel regression analyses for each of the 50 states plus the District of Columbia (referred to as the 51 states) included in this study. This resulted in a matrix of 408 state–year cells), including 2 years of baseline, for the first period and 459 state–year data cells for the second period, including the 3 baseline years. Because the baseline years for study period 2 (2000–2002) overlapped with the intervention years for study period 1, there were 663 total state–years of FARS data included in the overall data set used for the analyses.
Longitudinal-panel regression analysis, sometimes called cross-sectional time series analysis, was used to determine the association of each predictor variable with the outcome measure (i.e., seat belt use among occupants involved in fatal crashes). In this analytic method, predictor variables can vary across time (within a state) and across states (within a unit of time, such as from year to year). Separate sources of variance are estimated for each of these 2 dimensions (i.e., temporal and panel). Temporally correlated errors and values within a state are accounted for by a combination of autoregressive and moving average parameters (where significant) that are always of lag = 1, because the data set consists of annual data points. Stepwise selection of predictor variables, using P ≤ .05 as a selection criterion for variable entry and P > .10 for removal, was used to guard against potential collinearity and model overfit issues. Variables that were not statistically significant were removed to ascertain the best-fit model; however, those predictors were still assessed for effect size and probability value against the final best-fit model. The fine levels, percentage of enforcement agency participation in May mobilizations, and CIOT citation rates were tested as the predictor variables in the panel regression analyses, along with law type and per capita media expenditures.
Independent and Dependent Variables
Results
Baselines for the Two Periods of Study
During these 2 periods, we examined the association of various predictor variables with levels of seat belt usage among frontseat occupants killed in crashes. All predictors were coded as annual measures. They included (a) the law type (i.e., primary enforcement, secondary enforcement, and no seat belt law; both periods); (b) statutorily defined fine levels for seat belt violations, in dollar amounts (both periods); (c) the percentage of state enforcement agencies that participated in seat belt mobilizations each year (both periods); (d) number of seat belt citations issued per 10,000 residents during May CIOT mobilizations (available during the second period only); and (e) per capita media expenditures for May CIOT mobilizations (available during the second period only). Law type data and fine data were obtained from listings provided by the NHTSA over time. The dependent variable selected for this article was the buckled percentage of front-seat occupants older than age 8
Effect on Seat Belt Usage Rates among Passenger Vehicle Occupants Killed Table 1 shows the parameters and results associated with all analyses using seat belt use among occupants killed as the dependent variable for both periods examined (1997–2002 and 2003–2008). It shows that all 3 variables (primary law, fine level, and agency participation) were significantly and positively associated with increases in the dependent variable. The estimates of effect size are shown in column 1 and described below. Law Status Tables 1 and 2 show that, across both study periods, primary seat belt laws were associated with higher usage rates than secondary laws. During the first period, which was the Operation
Effects of Laws on Seat Belt Use
643
Table 1. Results of regression analyses: seat belt use among vehicle occupants killed Parameter
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
1st Period: Primary seat belt law Fines (square root)∗ Agency participation Intercept AR(1) 2nd Period: Primary seat belt law Fines (square root)∗ Agency participation Intercept AR(1)
Estimate
SE
df
t
P
Lower
Upper
.0915
.0128
190.85
7.14
Traffic Injury Prevention Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/gcpi20
The Effects of Primary Enforcement Laws and Fine Levels on Seat Belt Usage in the United States a
b
b
c
James L. Nichols , A. Scott Tippetts , James C. Fell , Angela H. Eichelberger & Philip W. d
Haseltine a
Nichols & Associates, Vienna, Virginia
b
Pacific Institute for Research and Evaluation, Calverton, Maryland
c
Insurance Institute for Highway Safety, Arlington, Virginia
d
Haseltine Safety Consulting, LLC, Manteo, North Carolina Published online: 27 May 2014.
Click for updates To cite this article: James L. Nichols, A. Scott Tippetts, James C. Fell, Angela H. Eichelberger & Philip W. Haseltine (2014) The Effects of Primary Enforcement Laws and Fine Levels on Seat Belt Usage in the United States, Traffic Injury Prevention, 15:6, 640-644, DOI: 10.1080/15389588.2013.857017 To link to this article: http://dx.doi.org/10.1080/15389588.2013.857017
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Traffic Injury Prevention (2014) 15, 640–644 C Taylor & Francis Group, LLC Copyright ISSN: 1538-9588 print / 1538-957X online DOI: 10.1080/15389588.2013.857017
The Effects of Primary Enforcement Laws and Fine Levels on Seat Belt Usage in the United States JAMES L. NICHOLS1, A. SCOTT TIPPETTS2, JAMES C. FELL2, ANGELA H. EICHELBERGER3, and PHILIP W. HASELTINE4 1
Nichols & Associates, Vienna, Virginia Pacific Institute for Research and Evaluation, Calverton, Maryland 3 Insurance Institute for Highway Safety, Arlington, Virginia 4 Haseltine Safety Consulting, LLC, Manteo, North Carolina
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
2
Received 11 February 2013, Accepted 15 October 2013
Background:Seat belt use in the United States increased from 11 percent in 1979 to 86 percent in 2012. Most of this increase has been attributed to seat belt laws, primary law upgrades, and highly visible enforcement. There has been less research on the effect of fines on seat belt usage. Methods: We examined law type and fine levels as predictors of seat belt use among fatally injured occupants of passenger vehicles from 1997 through 2008 using data from the Fatality Analysis Reporting System (FARS). Fine levels used were the statutory maximum base fines for a first offense. Results: Having a primary enforcement law was associated with a 9 to 10 percentage point increase in seat belt use. An increase in the fine amount, from the current median level of $25 to a level of $60, was associated with a 3 percentage point increase in usage. An increase in fine from $25 to $100 was associated with a 6- to 7-point increase. Such increases were in addition to the effects of a shift from secondary to primary enforcement. Discussion: The results of this study suggest that, in addition to current emphases on primary law upgrades and high-visibility enforcement of seat belt usage, increasing fine levels provides another viable strategy for increasing seat belt use. In addition, based on these results, states should consider publicizing such increases just as they publicize enforcement efforts. Keywords: seat belt usage, primary enforcement laws, seat belt enforcement mobilizations, seat belt citations, fine levels
Introduction Increases in Seat Belt Use Based on the results of daytime observational surveys, seat belt use in the United States has increased by more than 70 percentage points over the past 3 decades, from 11 percent in 1979 to 86 percent in 2012 (Phillips 1980; Pickerell and Ye 2008). Nearly all of this increase has occurred since 1984, when New York enacted the first mandatory seat belt use law and nationwide usage was 14 percent (Goryl and Cynecki 1985). The most abrupt increases in use occurred from 1984 through 1987 (from 14% to 42%; +14 percentage points), when 30 states and the District of Columbia enacted seat belt laws. Since 1990, much of the belt use increase, from 49 percent to more than 80 percent, was likely associated with a combination of highvisibility enforcement (HVE) and primary law upgrades, the
Managing Editor David Viano oversaw the review of this article. Address correspondence to James C. Fell, Pacific Institute for Research and Evaluation, 11720 Beltsville Drive, Suite 900, Calverton, MD 20705. E-mail: [email protected]
latter of which provide police officers with the authority to stop a vehicle and issue a citation based solely on an observed seat belt violation. Secondary enforcement laws require an officer to first observe some other traffic law violation before stopping a motorist for seat belt nonuse. High-Visibility Enforcement HVE mobilizations are characterized by vigorous enforcement of seat belt laws combined with publicity about the enforcement. These mobilizations, which typically involve intensified enforcement for 2 weeks, were first initiated statewide and nationwide in the early 1990s, but nationwide efforts became more common after 1996. Key enforcement events since 1990 include (a) a 2-year national enforcement program (i.e., the 70% by ‘92 program) launched in 1991; (b) the first statewide Click It or Ticket (CIOT) program launched in North Carolina in 1993; (c) more than 20 statewide Occupant Protection Special Traffic Enforcement Program grants and demonstrations implemented from 1993 through 1998; (d) the nationwide Operation ABC (America Buckles Children) mobilizations launched in 1997; (e) a series of statewide and regional CIOT programs implemented from 2000 through 2002;
Effects of Laws on Seat Belt Use
641
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
Fines
Fig. 1. Effects of laws and fines on seat belt use.
and finally (f) National CIOT mobilizations that replaced Operation ABC in 2003 and have continued through 2013. Nearly all of these efforts have been formally shown to be associated with relatively immediate increases in public awareness of enforcement activity and with significant increases in observed seat belt usage. Perhaps the best example of a high gain in seat belt usage is provided by the 16 percentage point statewide increase in observed seat belt use associated with the 2-year (1993–1994) CIOT program in North Carolina (Williams et al. 1996). Over time, however, baseline usage rates have increased and, as a result, the magnitude of observed gains has declined (Nichols et al. 2007; Nichols et al. 2009; Solomon et al. 2007; Solomon et al. 2002; Tison et al. 2008). Considerable evidence from state (and local) demonstrations suggest that, although much of the effect associated with HVE is retained for several months, usage nearly always declines over time (e.g., see Solomon et al. 1999). Such “decay” can be minimized by repeated waves of enforcement. Seat Belt Law Upgrades After 1990, there was both an increased emphasis on HVE and important changes in seat belt laws. Twenty-three states upgraded their laws to allow for primary (i.e., standard) enforcement procedures. California implemented the first upgrade, which was followed by 22 additional upgrades through June 2009. The greatest long-term increases in observed seat belt usage among the 15 states that participated in early HVE programs were found in states that had upgraded their seat belt laws (Nichols and Ledingham 2008). This suggests that the combination of primary law upgrades and HVE was a powerful factor for increased usage. As with HVE programs, the evidence shows that upgrades in seat belt laws implemented from 1990 through 1999, when baseline usage rates were lower, had a greater effect on observed usage than upgrades implemented since 2000. However, there is some indication that the more recent law upgrades increased belt usage among front-seat occupants killed in crashes, at least to the same extent, as earlier upgrades (Nichols and Ledingham 2008). One advantage of a law upgrade, relative to an HVE mobilization, is that it is always in effect, whereas a mobilization must be reinitiated periodically to maintain its effectiveness.
Relative to law upgrades and high-visibility enforcement, the impact of increasing fines to increase seat belt use has had relatively little focus. Statutory fines have increased somewhat over the past 15 years, with the average such fine increasing from about $20 in 1995 to slightly more than $30 in 2009. The median statutory fine level, however, has remained at $25. Data from an early study of safety belt laws showed that, when a fine was in effect when a primary law was implemented, observed seat belt use increased by an average of about 38 percentage points. When no fine was in effect, such as during a warning-only period, the average increase was only 23 percentage points—about the same as the change associated with a secondary law with or without any sanctions in effect (Campbell et al. 1987). Winnicki (1995) reported that the largest increases in seat belt usage among occupants of passenger vehicles involved in potentially fatal crashes (a rate that is based on data obtained from the Fatality Analysis Reporting System [FARS] and on the estimated effectiveness of seat belts when they are worn) were in primary law states where fines became effective within 4 months after implementation of the law. The median increase was 21 percentage points in those states. Winnicki also reported that each $10 increase in fine level was associated with about a 7.4 percent increase in usage among occupants involved in potentially fatal crashes. A third example of the potential impact of fines was provided by the state of Washington when it enacted a primary law upgrade in 2002. This upgrade, coupled with highly publicized enforcement efforts and increased sanctions for all traffic law violations, significantly increased the observed belt use. The new fine for a seat belt violation was $101, compared to the previous fine of $86. This increase in fines was publicized extensively, including messages on road signs. Salzberg and Moffat (2004) concluded that the publicized increase in fines likely played a role in the observed increase in belt use in Washington. The most recent study of the effect of fine levels on seat belt use comes from a 2005 study by Houston and Richardson. These researchers examined annual statewide observed seat belt use from 1991 through 2001, using time series, crosssection analyses similar to our approach in this study. They found primary laws to be associated with a 9.1 percentage point increase in observed use, compared with secondary laws, and they found that fine level was associated with higher use. They estimated that the $25 median-level fine was associated with a 3.8 percentage point increase in seat belt use compared to no fine, apart from any effect associated with enforcement type. For this study, our main objective was to determine the relative influence of primary laws and fines on seat belt use rates, while controlling for levels of enforcement and media based on available documentation, primarily data collected during annual mobilizations. Recent studies have confirmed that the type of law, intensity of law enforcement, and media publicizing the enforcement effort are important factors related to increased seat belt use. Fewer studies, however, have looked at the role of fines in increasing seat belt use.
642 Method
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
Periods of Study The first study period examined included the years from 1997 through 2002, which were characterized by primary law upgrades in 11 states and by annual enforcement efforts organized and supported by the Air Bag and Seat Belt Safety Campaign (i.e., Operation ABC Mobilizations). The second study period was from 2003 through 2008, which included primary law upgrades in 8 states and annual CIOT mobilizations, supported with innovative grant funds from the Transportation Equity Act for the 21st Century. During the first period, baseline usage rates were lower than during the second period and, although the legislative and enforcement activity seemed comparable, more complete enforcement data were available during the latter period. Whereas the first period was limited to the number of law enforcement agencies participating in the mobilization and reporting their activities, the second period benefited from having citation rates issued during the mobilizations. These data were available for nearly every state and study year. Separate models and analyses were developed for each period, as well as for both periods combined.
Nichols et al. killed in passenger vehicles (i.e., noncommercial, nonindustrial vehicle body types, excluding motorcycles and off-road vehicles). These data were obtained from the NHTSA’s FARS. A more complete study was published as an NHTSA research report by the NHTSA (Nichols et al. 2010). It included analyses using both observed seat belt use and seat belt use among occupants of passenger vehicles involved in fatal crashes. In part because the outcomes were similar for both analyses, only the analyses of usage among unbuckled occupants in fatal crashes are reported in this article. Data regarding fine levels, enforcement agency participation, citations, and media expenditures during the mobilization periods were obtained from a comprehensive review of historical reports provided by the NHTSA, the Governor’s Highway Safety Association, and the Insurance Institute for Highway Safety. Analyses
The baseline for the first study period was the 2 years (1995–1996) immediately preceding the first study period. The baseline for the second study period was the 3 years (2000–2002) immediately preceding the second study period. These were the last 3 years before the start of Transportation Equity Act for the 21st Century funding for CIOT mobilizations. Thus, 13 years of FARS data (1995–2007) were included in this longitudinal-panel regression analyses for each of the 50 states plus the District of Columbia (referred to as the 51 states) included in this study. This resulted in a matrix of 408 state–year cells), including 2 years of baseline, for the first period and 459 state–year data cells for the second period, including the 3 baseline years. Because the baseline years for study period 2 (2000–2002) overlapped with the intervention years for study period 1, there were 663 total state–years of FARS data included in the overall data set used for the analyses.
Longitudinal-panel regression analysis, sometimes called cross-sectional time series analysis, was used to determine the association of each predictor variable with the outcome measure (i.e., seat belt use among occupants involved in fatal crashes). In this analytic method, predictor variables can vary across time (within a state) and across states (within a unit of time, such as from year to year). Separate sources of variance are estimated for each of these 2 dimensions (i.e., temporal and panel). Temporally correlated errors and values within a state are accounted for by a combination of autoregressive and moving average parameters (where significant) that are always of lag = 1, because the data set consists of annual data points. Stepwise selection of predictor variables, using P ≤ .05 as a selection criterion for variable entry and P > .10 for removal, was used to guard against potential collinearity and model overfit issues. Variables that were not statistically significant were removed to ascertain the best-fit model; however, those predictors were still assessed for effect size and probability value against the final best-fit model. The fine levels, percentage of enforcement agency participation in May mobilizations, and CIOT citation rates were tested as the predictor variables in the panel regression analyses, along with law type and per capita media expenditures.
Independent and Dependent Variables
Results
Baselines for the Two Periods of Study
During these 2 periods, we examined the association of various predictor variables with levels of seat belt usage among frontseat occupants killed in crashes. All predictors were coded as annual measures. They included (a) the law type (i.e., primary enforcement, secondary enforcement, and no seat belt law; both periods); (b) statutorily defined fine levels for seat belt violations, in dollar amounts (both periods); (c) the percentage of state enforcement agencies that participated in seat belt mobilizations each year (both periods); (d) number of seat belt citations issued per 10,000 residents during May CIOT mobilizations (available during the second period only); and (e) per capita media expenditures for May CIOT mobilizations (available during the second period only). Law type data and fine data were obtained from listings provided by the NHTSA over time. The dependent variable selected for this article was the buckled percentage of front-seat occupants older than age 8
Effect on Seat Belt Usage Rates among Passenger Vehicle Occupants Killed Table 1 shows the parameters and results associated with all analyses using seat belt use among occupants killed as the dependent variable for both periods examined (1997–2002 and 2003–2008). It shows that all 3 variables (primary law, fine level, and agency participation) were significantly and positively associated with increases in the dependent variable. The estimates of effect size are shown in column 1 and described below. Law Status Tables 1 and 2 show that, across both study periods, primary seat belt laws were associated with higher usage rates than secondary laws. During the first period, which was the Operation
Effects of Laws on Seat Belt Use
643
Table 1. Results of regression analyses: seat belt use among vehicle occupants killed Parameter
Downloaded by [Universitat Politècnica de València] at 07:45 28 October 2014
1st Period: Primary seat belt law Fines (square root)∗ Agency participation Intercept AR(1) 2nd Period: Primary seat belt law Fines (square root)∗ Agency participation Intercept AR(1)
Estimate
SE
df
t
P
Lower
Upper
.0915
.0128
190.85
7.14
