About 43,000 people died and another 290,000 were seriously injured on the nation's roads in 2006. To reduce these numbers, the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU) nearly doubled funding for the Federal Highway Administration's (FHWA) Highway Safety Improvement Program (HSIP). SAFETEA-LU added requirements for states to develop strategic highway safety plans that include four key elements and to publicly report on at least the top 5 percent of hazardous locations on all of their public roads. The act also set aside funds for a legacy rail-highway crossing program and a new high-risk rural road program. As requested, GAO examined (1) states' implementation of HSIP following SAFETEA-LU, (2) HSIP results to date, and (3) FHWA's guidance and assistance to states. GAO analyzed plans from 25 states, including 19 randomly selected states and 6 states that GAO visited. GAO also interviewed FHWA and state safety officials.

All states adopted strategic highway safety plans, and the 25 state plans that GAO analyzed addressed the 4 key elements added by SAFETEA-LU, although states lacked some of the crash data and analysis capabilities described in the law. GAO's analysis showed that the 25 states (1) involved multidisciplinary safety stakeholders; (2) defined areas of safety emphasis through analyses of state fatality data using crash data analysis systems; (3) identified strategies and projects to address these emphasis areas through infrastructure improvements, behavioral approaches, and emergency medical services; and (4) provided for overall and individual project evaluations. However, many of the 25 states lacked components of the prescribed crash data analysis systems, such as a system for locating crashes and roadway data for local roads. FHWA is developing such a system for the states, but many states lack necessary data for local roads because they do not maintain or operate them. Without the prescribed components, states cannot conduct some of the safety analysis defined by SAFETEA-LU or report to FHWA on their most hazardous locations on all public roads, determine appropriate remedies, and estimate costs--all requirements added by SAFETEA-LU. While FHWA has set a deadline for states to develop the capability to locate crashes on all public roads, it has not done so for roadway data. Because states were not required to submit their strategic highway safety plans to FHWA until October 2007, they have not had sufficient time to implement and evaluate their HSIP strategies and projects; hence, it is too soon to evaluate HSIP results carried out after SAFETEA-LU. However, two of HSIP's statutory funding provisions may not be aligned with some states' safety priorities contained in their strategic plans. First, FHWA data show that most states have not used a new flexible funding provision that allows states to allocate some HSIP funds for behavioral approaches or emergency medical services. Some states may be reluctant to use this provision, according to state officials we interviewed, partly due to an HSIP certification requirement that all state highway safety infrastructure needs have been met. Second, the rail-highway crossing set-aside program does not target a key safety priority ofsome states and provides significant funding to some crossing areas that have relatively few fatalities. Better alignment of federal funding with state priorities in their strategic plans could help ensure that HSIP funding best addresses those priorities. Lastly, as states implement the high-risk rural roads program, they are hindered by limited data on rural roads and crashes, which are needed to identify qualifying roadways and appropriate remedies. FHWA provided comprehensive guidance and training to assist states in preparing their strategic highway safety plans, and participated in states' strategic safety planning processes. FHWA's guidance to states on reporting their most hazardous locations took states' data limitations into account and gave states latitude in defining the methodology and scope of their 5 percent reports. Consequently, these reports vary in content and format and may not increase public awareness of highway safety as intended.