When comparing U.S. road fatality figures, use the count to see how many people died and the rate to compare deaths relative to how much people drove. Before drawing either conclusion, check that the figures refer to the same time period, geography and type of fatality.
Match the unit before looking at the trend
Count crashes and people separately. One fatal crash may result in more than one death, so a table of fatal crashes and a table of fatalities can give different totals for the same place and year.
For a comparison of deaths relative to travel, check the denominator. A common NHTSA measure uses fatalities per 100 million vehicle miles traveled. A place can record more deaths and a lower rate if vehicle travel is sufficiently higher. The count shows the toll in lives. The rate puts that toll in the context of travel volume.
Check which crashes are in the total
The Fatality Analysis Reporting System covers qualifying motor vehicle crashes across the 50 states, the District of Columbia and Puerto Rico. Its inclusion rules require a motor vehicle on a trafficway normally open to the public and the death of an occupant or someone outside the vehicle within 30 days of the crash. Check the specific table's geography before assuming it includes every area in the reporting system.
Compare like with like
Start by writing down each table's unit, year and geography. For a rate, add the travel measure and confirm it covers the same place and period as the fatality count. If the figures use different boundaries, such as a county count and a statewide rate, describe each within its own boundaries. For a particular trip, use a source that covers the route and decision you need to make.
