The Dubious Primary Seat-Belt Bill
by Ari Armstrong, January 31, 2006
There are some superficial similarities between the "primary" seat-belt bill -- which would allow police to pull people over just for not wearing their seat belts -- and the smoking ban. Both proposals are labeled "nanny statist" by their opponents. Both are supposed to increase safety by forcibly altering people's behaviors. Both have been introduced again and again in the Colorado legislature, and both have repeatedly failed. (See my reports for 2000 and 2002.) If either measure passes, it's likely to stay on the books indefinitely. Thus, opponents of the measure have to win anew every year, but advocates need win only once.
But there's also an important difference between the measures. Smoking bans are clear violations of property rights. However, seat-belt laws (of any variety) apply to use of a vehicle on government property, and so the case against them must be a little different.
Nobody has a problem with mandatory seat-belt rules on airplanes and rollercoasters. Those things are privately owned, and so we properly recognize the right of the owners to set policy for their use. (I think seat belts on airplanes are also regulated by the FAA, but even on a completely free market I suppose that airlines would implement stringent seat-belt rules.) If you don't want to wear a seat-belt on a rollercoaster, you won't be allowed on the ride, and that's perfectly fine. By extension, owners of private roads should be free to set policies for the use of those roads, including seat-belt rules.
However, the fact that seat-belt rules by private entities are legitimate does not imply that seat-belt rules pertaining to government-run roads are legitimate. (I do think there's room for prosecuting parents for child abuse who endanger their children in vehicles, but primary seat-belt laws apply to adults.) People are forced to pay for roads through taxes. Traffic laws are set statutorily and enforced by armed agents of the government. And people's rights must be protected against undue government intrusion.
In his January 28 column for the Rocky Mountain News, Dave Kopel summarizes the main arguments against primary seat-belt laws: "seat-belt stops could be employed discriminatorily, and that data from other states show that forcing some people to buckle up gives them an excess sense of security, so that they drive more aggressively, and cause more accidents."
Kopel explains, "Current law forbids seat-belt-based traffic stops for persons over 16 years old; the police may issue tickets if they discover a beltless adult during a traffic stop for another purpose."
Kopel rightly complains that media reports of the primary seat-belt bill have not included the arguments of opponents. Both the News and The Denver Post published articles on January 25 that gave voice only to the advocates of the bill.
At least Bianca Prieto's story for the News ("Effort grows to 'enhance' seat belt law") offers the basic information about the bill: "Last week, Rep. Fran Coleman, D-Denver, and Sen. Peter Groff, D-Denver, introduced House Bill 1125, which would make failure to wear a seat belt a primary offense, allowing police to pull a driver over for not buckling up."
Jeffrey Leib's story in the Post ("Parents of car-crash victims rally for seat-belt") mostly covered the same ground. He writes, "Fighting back tears, Pam Sallee of Delta described losing her 17-year-old daughter, Brittney Hudson, when the unbelted teenager, driving alone, lost control of her Toyota Celica and was ejected in a rollover accident just before Thanksgiving. To save lives, she urged Colorado legislators to pass the recently introduced primary seat-belt measure."
But this raises two issues that the newspapers ignore. First, it's already widely known that seat belts save lives. It's also already against the law to drive without wearing a seat belt. So would a "primary" law have made any difference in the girl's behavior? (Surely parents have some responsibility to instill in their children safe driving habits.) Second, while seat belts help protect some people in the event of a crash, they don't always save lives.
Leib continues, "Behind the parents on the Capitol's west steps, supporters of the primary seat-belt bill had lined up 282 pairs of shoes to represent the number of those killed in 2004 Colorado traffic accidents who were not wearing safety belts."
Of course, the fact that 282 people died while not wearing seat belts does not prove that a primary seat-belt law would have prevented those 282 deaths. Many of those people would not have worn a seat belt regardless of the law, and some who did wear a seat belt still would have died. But the Denver news pages have no use for critical analysis when they have sensational stories about blood and tears.
In a reasonable world, a primary seat-belt bill would have to pass at least five tests before becoming law (not that I suffer any illusions that the legislature is reasonable).
1. How much do primary seat-belt laws increase the wearing of seat belts? The effect would have to be significant, and due to the law rather than some incidental factor, in order to support the law.
2. How much do primary seat-belt laws result in drivers wearing seat belts but otherwise driving more dangerously? There's a common example in economics of how car safety impacts driving. If you drive with a metal spike coming out of the steering column, you'll drive very carefully indeed. Conversely, as cars become safer, some people compensate by driving less carefully. The increased safety from more seat-belt wearing would have to significantly offset the decreased safety from more dangerous driving.
3. Do primary seat-belt laws result in more abuses of police power, or do they interfere with personal choices in a way that's inappropriate for government? If so, then they might not be justified regardless of any potential increase in traffic safety.
4. Are limited police resources best spent on enforcing primary seat-belt laws, or are there more important uses for those resources? (See William Niskanen's article.) If a police offers pulls over somebody for suspected violation of the primary seat-belt law while a drunk driver passes, that hardly seems likely to increase traffic safety.
5. Can safer driving be encouraged by less-intrusive means, such as through non-profit advocacy?
Advocates of primary seat-belt laws have trouble passing those tests, though they make good headway on the first point.
At least the Post published Ed Quillen's January 29 column that criticized the bill. Quillen notes that "enforcement would be sporadic." He does raise a good point in favor of the bill, though, despite ultimately arguing against it: "[I]f there are people sliding all over a spinning car, the driver may not be able to regain control..." In other words, belted drivers are better able to remain in control of their vehicles, thus protecting other drivers, whereas unbelted passengers might contribute to drivers losing control of their vehicles. However, it's unclear (at least to me) in what percent of crashes this might be relevant. My sense is that most crashes are caused by something other than an unbelted person contributing to the loss of a vehicle's control. However, this would be an interesting point to explore further.
Brian Schwartz sent me a few interesting links. Sam Peltzman writes that "mandated devices" for car safety produced benefits that are "entirely offset by a combination of more accidents and more fatalities to nonoccupants such as pedestrians, bicyclists, and motorcyclists, whom the devices did not protect."
W. Kip Viscusi and Ted Gayer argue that "hazard warnings" should "provide new and honest information in an unbiased manner." Such warnings "can be among the most beneficial interventions in that they foster improved market performance without imposing any regulatory burdens."
On the other hand, a study by economists Alma Cohen and Liran Einav came up with mostly supportive results for mandatory seat belt laws: "Controlling for the endogeneity of seat belt usage, we find that it decreases overall traffic fatalities. The magnitude of this effect, however, is significantly smaller than the estimate used by the National Highway Traffic Safety Administration. Testing the compensating behavior theory, which suggests that seat belt use also has an adverse effect on fatalities by encouraging careless driving, we find that this theory is not supported by the data."
However, the finding with respect to the "compensating behavior theory" is incomplete, because the authors look only at fatalities, even though it's reasonable to think that seat-belt laws disproportionately impact fatalities of drivers but non-fatal injuries of others. (To their credit, the authors try to account for state differences in driving dangers. More dangerous conditions could prompt more seat-belt usage but also more accidents.)
The writers also state, "We find that whereas a mandatory seat belt law with secondary enforcement increases usage rate by about 11 percentage points, a mandatory seat belt law backed up by primary enforcement increases usage rate by about 22 percentage points... [W]e estimate that if all states now having secondary enforcement were to switch to primary enforcement, the national usage rate would shoot up from the current 68 percent to about 77 percent, producing an annual saving of about 500 lives."
The study also seems to account for some factors incidental to the law itself, such as increased media exposure of traffic fatalities among non-belt wearers. For example, three states "passed, repealed, and reinstated again" mandatory seat-belt laws, and "usage rates dropped drastically with the repeal of the laws and increased immediately after the laws were reinstated."
Another implication of the study, though, is that factors other than seat-belt laws are even more important. For example, a secondary law in North Dakota is associated with an increase in seat-belt use -- from 32 percent to 42 percent, and use in 1998 (several years after passage of the law) was 40 percent. Yet New Hampshire, the single state without any mandatory seat-belt law, had a seat-belt use rate of 56 percent in 1998. Passage of seat-belt laws resulted in more seat-belt use in almost all cases. However, in some states, use went up a little, while in other states it went up a lot. Furthermore, years later, relative to the rates "immediately after the law," use went up in some states and down in others. This suggests that things like public awareness and attitudes and enforcement play a big role in seat-belt usage.
As the final page of the report indicates, seat-belt usage tended to increase several years before and several years after the passage of a law, with a larger spike concurrent with passage of the law. This suggests that seat-belt laws were passed during periods when people were already increasing their use of seat belts. (Ohio saw a spike in use with passage of a law, then lost ground for a couple of years, then saw slow increases for several years.) This suggests that long-term trends are due to something other than legislation.
So, to take the findings of the study at face value, some people change their behavior with respect to seat belts because of reasons other than the law, but some people change their behavior specifically because of the law, presumably to avoid the risk of getting pulled over.
That's an astounding outcome, in that this latter group of people apparently wear their seat belts only to avoid getting pulled over, not to protect their own lives. This seems to mean that those people, absent a legal/financial incentive, consciously avoid wearing seat belts, despite the known risks, simply because they value not wearing a seat belt more than they value the added safety. Offhand, I can think of no benefit to not wearing a seat belt, other than that it takes a second to put it on and it is mildly restrictive. Such "costs" strike me as so trivial that I can't imagine not wearing a seat belt.
However, it's really none of the government's businesses to dictate to people what risks they accept. For example, skiing, rock-climbing, and playing football are also dangerous (as are hundreds of other activities), yet no one thinks it odd that, in these areas, people are and should be permitted to assume risks in order to obtain personal pleasure.
Rounding out Schwartz's links is the Wikipedia entry, which argues in favor of the theory of "compensating behavior" based on international studies.
As an aside, another of Schwartz's suggestions is that seat belt policy is properly a matter for insurance companies to decide. I think it makes sense for an insurance company to offer protection only when the occupant uses a seat belt. However, this seems to conflict with existing law, such as statutes mandating coverage. I think a good case could be made that existing law "externalizes" much of the harm of driving dangerously and thus discourages seat-belt use and safe driving generally. In addition to insurance regulations, welfare benefits and bankruptcy laws might also contribute to the problem. If people don't have to bear the full consequences of driving dangerously, obviously they'll be encouraged to drive more dangerously.
I found (more or less at random) a variety of arguments in favor of primary seat-belt laws.
A report from Minnesota's Department of Public Safety isn't very convincing. It raises the possibility that increased fines, enforcement, and publicity are factors besides the law itself that can impact outcomes. The report relies on dubious statistics from the National Highway Traffic Safety Administration. The report claims that primary seat-belt laws do not increase discriminatory traffic stops, but there are two main problems with this finding. First, we can expect that states with "primary" laws are generally more "progressive," and thus they might suffer less discrimination even if the primary seat-belt law exacerbates the problem. Second, the problem involves class, age, and lifestyle choices, not just race (as the report assumes).
Mothers Against Drunk Driving (MADD) seems to have overstepped its original mandate, as it now advocates primary seat-belt laws. Significantly, MADD argues that "85 percent of the costs of crashes are borne by society, not by the individuals in the crash." In other words, because the government has taken on the task of paying for people's health care, therefore, government should also regulate people's behavior and force them to be safer. It's easy to see where the logic of that argument leads.
What is obvious is that wearing a seat belt in conjunction with general safe driving practices will protect your life and safety. (While I'm a safe driver now, I regret that in my younger years I was not as conscientious.) Regardless of what the law says, you should wear your seat belt when you drive and adopt other safe driving habits.
While it appears that primary seat belt laws do persuade more people to buckle up, it's clear that non-legal factors are more important. Given the intrusive, potentially abusive nature of primary seat-belt laws, given the possibility of "compensating behavior" by some drivers, and given the ability to influence behavior through less-intrusive means, I continue to oppose primary seat-belt laws. However, such laws are not as unambiguously bad as are laws that directly interfere with private property.
The more important issue (that is for that reason off the political radar) is to evaluate how statutes enable drivers to push the costs of reckless driving onto others. That's a fairly complicated matter. Of course, the political problem of seat-belt use arises mostly because the government owns the roads. While there is a reasonable alternative to that, hardly anyone is interested in contemplating it.