Traffic violator school research report

By: Michael Finigan

Date: 1995-04-09

This study was commissioned by the Santa Clara County, California Municipal Court, with the co-operation of the California Department of Motor Vehicles. For more information and a complete copy of the study, write or call:
Dr. Michael Finigan, Northwest Professional Consortium. 17725 Hillsdale Drive, West Linn, Oregon 97068. Phone (503) 635-9896.

Introduction

This study, involving over 35,000 drivers, is the largest study of a Traffic Violator School Program ever undertaken using a scientifically valid research design. The California Department of Motor Vehicles approved the original research design, provided data from driver records, and provided technical assistance and peer review on a number of occasions. While the final report reflects many of the suggestions from the California DMV, the conclusions of the study are entirely the responsibility of the author.

This study was designed to examine two issues. The first is the curriculum effect -the effectiveness of the NTSI curriculum (Levels I and II) in reducing subsequent recidivism, particularly accidents, for its participants when compared to a no-treatment control group. The second is the program effect -the effectiveness of the NTSI program in reducing subsequent accidents for its participants compared to drivers who are eligible to attend the program but who chose not to take the TVS dismissal route and to pay the fine for the violation instead.

The sample for the study began to be collected in July, 1989, and was completed in July, 1990. This document is a summary of the report issued to the Santa Clara County Municipal Court on this research.

Summary of previous research

Studies on Traffic Violator School Programs (TVS) using strong research designs are few in number. Although driver improvement schools have been in existence in the United States since the 1930s, research on them is a relatively recent phenomenon.

The studies that examine specific programs such as AARP, DDC, and NTSI continually find significant effects of the programs in violation recidivism compared to no-treatment control groups. Conversely, several studies conducted by the California DMV on the industry as a whole have found that traffic programs collectively have no or even negative effect.

Methodology

Research Design: This study was conducted on NTSI's level I and Level II course curriculum as provided in Santa Clara County. Level I is an eight-hour course generally intended for drivers who have not attended a Traffic Violator School (TVS) program in the previous three years and is by far the most frequently attended course both in Santa Clara County and nationwide. Level II is a 12-hour advanced course for divers who are more frequent offenders.

Curriculum Effect: The study used an experimental design employing a computer to randomly assign NTSI classes into treatment and control (no treatment) conditions. Those drivers in the treatment group received the normal NTSI curriculum while those in the control group were dismissed shortly after arriving for the course.

Program Effect: In order to test the entire program's effect compared to available options, a comparison group of drivers was selected from those who incurred tickets during the same 1989-1990 period but opted to pay the fine. This provided a test of the effect of the program in the contest of alternatives.

Sample Size: Because of the interest in accident statistics as an outcome measure, a large sample size was necessary. After excusing some drivers for various reasons, a total of 11,162 treatment group drivers, 11,397 control group drivers and 12,283 comparison group drivers were identified, totalling 34,842 sample subjects.

Data Source-Driver Records: The study looked at driver records for treatment, control and comparison group drivers at intervals of six months, 12 months, 24 months and 36 months subsequent to their attendance/completion date. The citation date was used as the reference date for the comparison group who did not choose the TVS option. The period of driving exposure was the same for each driver in the study with the DMV providing data for the three-year period prior to the reference date as well as subsequent data.

Results

Both the randomly assigned treatment group and the randomly assigned control group were found to be equivalent with no pre-existing differences, a clear indication that the randomization process was successful and that any subsequent differences in driver records could be attributed to the NTSI curriculum.

Subsequent Accidents (Level I-Eight-Hour Curriculum):
The NTSI Level I graduates showed a 14% reduction in subsequent accidents compared to the no-treatment control group six months after the completion of their course. As a subgroup, female NTSI Level I graduates had an accident rate 26% lower than the no-treatment control group females six months after the course.

Subsequent Convictions: One of the original rationales for the existence of TVS programs was their potential to reduce subsequent traffic violations which, in addition to increasing traffic safety, would help to relieve clogged court dockets. This data suggests that the NTSI curriculum does have this effect as NTSI Level I graduates show a respectable 12% reduction in subsequent convictions six months after the completion of their course. This reduction in subsequent convictions continues with a 10% reduction at one year and 7% reduction (not statistically significant) at two years after the course. After three years the effect is insignificant.

Level II (12-Hour Curriculum): NTSI also has a Level II curriculum designed for those drivers who have already completed a Level I TVS program within the past three years. This is a long curriculum with more emphasis on dealing with the higher risk driver. Samples for this portion of the study were drawn exactly as in the Level I research.

Recidivism: Level II Curriculum: The data indicates that the impact of the Level II program on subsequent accidents is quite dramatic, reducing subsequent accidents by 16% over the no-treatment control group six months after the completion of the program. Even after two years, the decrease in accidents associated with the NTSI group is a respectable 9%. The fact that the Level II program does better than the Level I program may not be surprising. More frequent violators have more room for improvement. In fact, when those drivers with "clean" prior records are removed from the analysis, the NTSI Level I program participants have 16% fewer subsequent accidents than the no-treatment control group members, a figure identical to the 16% found for the Level II group. This suggests that both the Level I and Level II programs may be similar in their effectiveness with drivers who have prior violations on their records.

Subsequent Convictions-Level II: The NTSI Level I graduates show a 9% reduction in subsequent convictions compared to the no-treatment control group six months after the completion of their course. This reduction in subsequent convictions continues eight months and one year after the course. The NTSI Level II graduates also show an 11% reduction in subsequent negligent operator points compared to the no-treatment control group six months after the completion of their course. This reduction in subsequent negligent operator points continues for at least two years following the course. The fact that there is a larger reduction in negligent operator points than in violations suggests that the NTSI program is particularly adept in reducing the more serious, traffic safety-related violations.

The program effect: TVS graduates compared to the conviction group

Another concern is whether a TVS program is more effective than the alternative of simply paying the fine. This is called the program effect and represents an evaluation of the entire TVS program.

However, because these program effect results must depend on a quasi-experimental design without random assignment to the treatment and comparison groups, it is always possible that some other factors may account for the results. ¹

The only way to provide definitive results for the program evaluation would be to conduct a study in which offending drivers would be randomly assigned into treatment and fine-paying conditions (as they were for the curriculum effect results given in the previous section). Random assignment into NTSI or fine-paying groups would require judges to mandate to offenders either an NTSI course or a fine. Both judicial resistance and logistic difficulties would make this problematic. Nonetheless, quasi-experimental approaches such as the one presented in this report are commonly used when experimental approaches are impractical or undesirable and the results accepted as valid.

The analysis of covariance model used by the researcher examined accidents and violations and compared the NTSI group with the fine-paying group with the pre-existing differences of gender, age, and prior driving records as control variables. A summary of the program effect results follows:

1. The NTSI program participants had 12% fewer accidents six months after their course than a comparison group of drivers who, during the same period, paid their fines rather than attend traffic school. At one year, there was an 8% reduction in accidents and a 14% reduction in police-reported accidents (often the most serious accidents).

2. The NTSI program participants had 12% fewer convictions at six months and 9% fewer at one year.

3. The NTSI program participants had 21% fewer subsequent licence suspensions at both six months and one year than the fine-payers. Since TVS dismissals that are masked are not assessed negligent operator points, some of the effect may be an artifact of the lower neg-op count. Nonetheless, since one of the prime reasons for licence suspensions is failure to provide adequate insurance, the NTSI program may be helping to increase the number of drivers who drive with adequate automobile insurance.

If we control for the number of incident sin the three years prior to the citation, we find the following pattern. For those drivers who have three or fewer incidents in the prior three year period, those who completed the NTSI Level I class have 12% fewer subsequent accidents (in the six months following the class) than the fine-paying drivers with the same number of prior incidents (three or less).

Conclusion

This research was designed to test both the curriculum and program effects of the NTSI traffic school option. Drivers who completed the NTSI Level I and Level II programs had better subsequent driving records, including fewer subsequent accidents, than the no-treatment control groups. For the Level I groups, this resulted in 12% fewer accidents and convictions at the six month mark and 10% fewer convictions at one year.

The Level II results were more dramatic indicating a 16% reduction in subsequent accidents over the no-treatment control group six months after completion. Even after two years this decrease remained at a very respectable 9%.

This data suggests that the TVS program in Santa Clara County does have a positive effect in reducing the risk of accidents in the county. Since over 950,000 drivers have graduated from NTSI in the years of its existence in Santa Clara County, it can be estimated from the accident reduction effect found in these data that approximately 5,640 accidents, 135 injuries and 45 fatalities may have been prevented by the program since its inception.²

When added to the fact that traffic fatalities in Santa Clara County have decreased by 35% over the past 10 years (as opposed to a 17% decrease in California as a whole) and that the county continues to have the lowest traffic fatality rate of the eight largest California counties, it seems plausible that the NTSI program has had an impact in minimizing traffic injuries and fatalities.

Footnotes

1. For instance, it is possible that those who, for whatever reason, attend the traffic school represent a differently motivated group than those who, for whatever reason (choice or simple lack of understanding of the option) do not attend. Or perhaps there are differences in driving exposure between the two groups. These are difficult items to measure and while we have some data on both these issues for the NTSI group (from the questionnaire), we have no comparable data for the fine-paying group.

2. Although no specific fatality reductions were found in this study for the NTSI program, fatalities are relatively rare and difficult to be captured adequately even by a study of this size. We are assuming that over time, a proportion of the lesser number of accidents in Santa Clara County would include injuries and fatalities.