A. Historical Evolution and Focus

The Research and Development Branch (R&D) was established within the California Department of Motor Vehicles (DMV) in 1959. The goal of R&D has been to provide guidance in the design and evaluation of departmental programs, particularly those relating to driver licensing and traffic safety. In contrast to other departmental staff functions, R&D undertakes studies involving application of advanced quantitative techniques and the development and evaluation of experimental program alternatives. Through such studies, ineffective programs have been eliminated or replaced by more cost-beneficial alternatives.

The use of case-control evaluation designs distinguishes R&D from most other research and evaluation groups in state government and DMV's of other states. In an independent 1976 report (see footnote 1) the University of California (UC) reviewed the evaluation programs of all California state agencies. That report cited only two examples in which a California agency had utilized controlled evaluation procedures in making decisions about program effectiveness; one of these examples is included in the UC report as a model of program evaluation methodology and both examples were studies conducted by the California DMV R&D office. We offer the following quote from the conclusion of the UC study because it succinctly conveys R&D's program evaluation philosophy:

"Because control- or comparison-group program evaluation plans provide the clearest explanation about the causes of program-related changes, these evaluation plans are particularly helpful in policy making. Although many respondents believe that such designs are feasible, few departments have conducted these kinds of program evaluations. The reason for this absence seems to relate more to a lack of understanding of the benefits of these kinds of evaluations than to a political or procedural problem hindering their use. Control- and comparison-group program evaluations could be implemented more often in California state departments, and evaluators should be trained in their usefulness."

A more recent (1987) independent appraisal of the California DMV's R&D Branch appears in a major policy monograph and literature synthesis by Traffic Injury Research Foundation of Ontario, Canada:see footnote 2)

"Model driver-improvement systems have been based on 'rational considerations and trends in empirical research'. Experts have stressed, however, the requirement to replicate prior research and evaluation. Developing effective driver-improvement systems in different jurisdictions, with different driving populations, during different time periods is an interactive, evolutionary process, a process in which research and evaluation become indispensable. For example, only an integral research/evaluation component will allow administrators and staff to measure cost-effectiveness over time and allow rational system self-corrections. Each jurisdiction should consider the example of California's Department of Motor Vehicles, which has implemented a driver-improvement system representing the state of the art."

Table of Contents | Next