Factors Affecting the Relative Incidence of Uninsured Motorists Claims
Ma, Yu-Luen, Schmit, Joan T., Journal of Risk and Insurance
Estimates of uninsured drivers in the United States hover around 15 percent (NAII, 1998). With extensive concern about insurance affordability and its effect on economic advancement (Joint Economic Committee, 1998) and with expansion of U.S. compensation systems to other nations, consideration of factors that affect the extent to which drivers choose to insure or not is valuable. This article addresses the effect of enforcement mechanisms for purchase of required insurance on the degree to which drivers choose not to insure. Results indicate that higher levels of enforcement stringency relate to lower levels of uninsured drivers. Lower levels of poverty and populations living in metropolitan areas are also related to lower levels of uninsured drivers, while no-fault states are associated with higher levels of uninsured motorists.
Automobile insurance represents nearly 50 percent of property/liability insurance premium volume in the United States and similarly large percentages throughout the globe (Insurance Information Institute, 1997). In part because of its market representation, automobile insurance has been the focus of numerous research projects. Much of this work has been oriented toward identification of factors that affect price and availability of the product, including consideration of the effect of regulation and market structure. No-fault laws and other tort reforms similarly have received extensive attention. Analyses of regulatory effects primarily focus on rating limitations and systems designed to provide coverage for high-risk drivers. Another factor appears relevant but receives scant attention: that of regulations requiring purchase of insurance and how these requirements affect the size of the uninsured motorist population.
In 1927, Massachusetts passed the first U.S. compulsory insurance law. Today, 43 states and the District of Columbia follow some version of compulsory law,  while the remaining seven states have financial responsibility laws.  Among states with compulsory insurance laws, various enforcement techniques exist. The authors test for relationships between various forms of insurance requirement enforcement laws and the relative incidence of uninsured motorists (UM) claims, using data from U.S. state experience. The results have implications for regulatory efforts designed to protect individuals involved in automobile accidents. 
Automobile insurance markets have been given significant attention in the literature, including consideration of no-fault laws, regulatory constraints, and market structure. Additionally, a large body of theoretical work associated with insurance purchase decisions exists. Issues directly associated with UM, however, have received only limited attention beyond several trade reports and a number of articles on alternatives to individually purchased policies ("pay-at-the-pump" options).
The question of optimal insurance purchase decisions relates to UM through the option to forgo insurance despite rules to the contrary. Sinn (1982) considered the issue by recognizing the unbounded nature of liability exposures (in contrast with property exposures) relative to the bounded nature of assets. This disconnect between potential losses and available assets yields rational decisions to forgo insurance, an idea also modeled by Keeton and Kwerel (1984). Huberman, Mayers, and Smith (1983) further demonstrate that bankruptcy potential offers a rationale for imposition of insurance limits, a form of selecting against the purchase of insurance.
Through simulation, Dahlby (1983) demonstrates that as the price of coverage increases, the percentage of drivers who purchase insurance decreases. Smith and Wright (1992) add to the discussion by providing evidence of a strong relationship between the price of automobile insurance and the relative size of the UM population, both in terms of price affecting the percentage of UM and in terms of the number of UM affecting price. …