The paper uses multivariate regression analysis to examine the relationship of state driver's licence renewal requirements, population characteristics and driving conditions to state driving licensure rates in the USA. It examines licensure rates for the years 1985-1989, using data provided by the Federal Highway Administration. All-age and senior licensed drivers are examined separately. The dependent variable in the regression analysis is the number of licensed drivers in a state as a percentage of state population; all except two US states were included. A logistic regression method is used, to avoid biased estimates for proportional data. Driving conditions variables and habits variables were included in the analyses, as well as measures of licence renewal requirements. It was found that all these variables have systematic relationships to licensure rates. Licence renewal tests are associated with lower senior driving licensure rates. Greater driving intensity, higher percentage of seniors, and higher income per head are associated with higher licensure rates. Miles driven per head, income per head, percentage young, truck registrations, and the urban-rural mix of driving are useful predictors of state licensure rates.
Abstract