Sunstein: A cost-benefit analysis of making safer cars
Can a federal regulation be worthwhile if its monetized costs exceed its monetized benefits?
Consider the Department of Transportation's decision last week to require rear-visibility cameras in essentially all new cars starting in 2018. The regulation will eventually affect millions of Americans, increasing the safety of these behind their cars, particularly children.
Sound good? Here's the problem: The Department of Transportation projects that the new regulation will cost between $546 million and $620 million - and perhaps as much as $924 million. That will mean an additional cost, per car, of $44 (for cars already equipped with a suitable visual display) to $142 (for cars without such a display).
At the same time, the government projects benefits of only $265 million to $396 million, with a high-end estimate of $595 million. (Disclosure: As administrator of the Office of Information and Regulatory Affairs from 2009 to 2012, I worked on this rule in its earlier stages.)
Since the early 1980s, both Republican and Democratic presidents have required executive agencies to give careful consideration to the costs and benefits of regulations, and to proceed only if the benefits justify the costs (unless Congress has required otherwise). The Obama administration has been intensely committed to cost-benefit analysis. To date, very few of its regulatory requirements have generated monetized costs in excess of monetized benefits.
Why, then, did the government proceed with a rule that, by its own projection, has net costs of at least $200 million? There are two answers.
The first involves the Department of Transportation's obligation to comply with the law. In 2007, Congress enacted the Cameron Gulbransen Kids Transportation Safety Act, a law designed to reduce the risk of backover crashes, which result in about 267 deaths and 15,000 injuries each year. About a third of the fatalities involve children under the age of 5. The act requires the agency "to expand the required field of view to enable drivers of a motor vehicle to detect areas behind the motor vehicle."
Under the law, the Department of Transportation can consider sensors and additional mirrors, but its own evidence suggested that both of these would have little or no effect in reducing backover crashes. For that reason, the department concluded that a less costly approach would not be consistent with the law's requirements.
The department's second answer was less straightforward and more interesting. It insisted that the monetized figures do not tell us everything we need to know about the benefits of the regulation. The governing executive order, issued by President Barack Obama in 2011, allows agencies to consider "values that are difficult or impossible to quantify, including equity, human dignity, fairness, and distributive impacts."
On this count, the department emphasized four points. First, about one-third of the deaths involve very young children. While it used its usual $9 million figure to value a "statistical life" (as it is called), it emphasized that some studies find a significantly higher number for young children.
Second, the department noted that in some cases, drivers have been the direct cause of their own children's death - an unimaginably horrific outcome that "is not fully or adequately captured in the traditional measure of the value of a statistical life."
Third, the department stressed that many of those killed or injured in backover crashes are disabled or old, and they are not in a good position to protect themselves.
Finally, the department said that it had not quantified all the benefits that drivers will receive, including simplified parking and increases in ease and convenience.
All this is pretty sensible. To be sure, a hard-line approach to cost-benefit analysis would demand that the Department of Transportation be more disciplined in its discussion of unquantifiable benefits, and it would insist on more serious efforts at quantification.
For example, some research suggests that for children, the value of a statistical life is around $18 million. If the department had used that number, would the total benefits number get close to the costs?
And perhaps it would be possible to quantify the monetary benefits of an easier driving experience - and to see whether that figure would materially increase the department's benefits estimate.
To reduce the risk of unhelpful or dumb regulation, cost- benefit analysis is exceptionally important. Everyone should agree that agencies should not defeat the purpose of that analysis by invoking unquantifiable benefits. But in the context of rear visibility, the Department of Transportation acted quite reasonably, pointing to values that are not easy to turn into monetary equivalents but cannot be considered irrelevant to its overall assessment.
No one wants to pay more for a new car. But if the result is to save some children's lives and to prevent a significant number of serious accidents, the added cost is worth incurring.
See more about comments here.