The data say U.S. highways are safer than ever, but there are lies, damned lies, and safety statistics.
We the motorists of America have been preening ourselves the last few weeks because of the good news about significantly reduced carnage on our highways. And that’s as it should be: safety belt usage is up, drunken driving is down, cars and highways are safer, and-critically-people are driving less.
Among the facts cited by Obama’s new Secretary of Transportation, Ray LaHood, in his April 6 announcement for the National Highway Traffic Safety Administration (NHTSA) was this: “The nation also saw the lowest fatality rate ever recorded in 2008 at 1.28 fatalities per hundred million vehicle miles traveled, down from 1.36 in 2007.”
Now I can’t hold Secretary LaHood accountable for these numbers; I’d be willing to bet that if he bothered to check the numbers before he signed off on the NHTSA news release put in front of him for approval, he was told the fatality rate came from a brother federal agency, the Federal Highway Agency (FHWA). The FHWA is the government agency that since the 1940s has been supplying Washington and the nation with vehicle miles traveled (VMT) which are used for a whole host of government actions-highway expenditure allocations, environmental estimates of emissions (“tons per mile”) and fatality rates.
The fatality rate in particular is used to compare the U.S. experience with that of other countries, as reported through the International Road Traffic Accident Data Base (IRTAD). It also tells us, in a pseudo-scientific way, what our report card is on highway safety, better than the raw number of fatalities which is affected by a whole lot of factors including weather, enforcement and, well, luck.
So what? you are wondering.
Well, would you believe that the FHWA estimates the vehicle miles traveled on two bases: its consistent approach over sixty some years and data collected from 3,600 electronic or pressure plate traffic counters spread out among the states. These more-or-less-invisible collectors count traffic in five-minute intervals for 12 consecutive months. And, by the way, these are not the rubber hoses you may have noticed across roadways, which instead are used by local traffic agencies to determine needs for widening, traffic lights and so on.
Thus, from its counters sampling a relatively few roadway traffic patterns, the FHWA estimated that Americans drove 3 trillion miles in 2008, and that number was divided into the number of fatalities counted by NHTSA to calculate the fatality rate. My point is that traffic sampling by the FHWA is not the same as the NHTSA’s hard numbers for deaths, as collected from police agencies. Perhaps FHWA also takes into consideration the trends in federal gasoline taxes collected, new car registrations, scrappage rates and other numbers, some hard and some soft, for verification of VMT trends. But if so, the agency spokesman didn’t mention when I called for clarification.
Yet this data is widely used, as noted, for Washington’s and numerous industries’ policy decisions-and for anti-auto propaganda in the case of environmental issues.
Although sampling is widely accepted as a means of collecting data for trends in such matters as TV programming, desirability of consumer products and voter tastes, it seems odd in this computer age to depend on it the way our government does.
Or have I just made a mountain out of molehill?