Tuesday, October 13, 2009

Do Zero Tolerance Laws Really Work?

This is a question to which most researchers would answer "yes," due to numerous empirical studies as well as in theory, but many have had their doubts. Enter Darren Grant, an economist at Sam Houston State University, who did a recent study that shows that there are good reasons to doubt the effectiveness of such laws.

Zero Tolerance (ZT) laws are what they sound like: laws that prohibit people (under 21 in this case) from driving with any measurable amout of alcohol in one's blood. These laws were justified by studies that found that the relative risk of a fatal crash for 16-20 year old drivers begins rising at a lower BAC, and more rapidly, than for those over 21.  By 1998, all 50 states and DC have passed such laws. Many states did so voluntarily, but the rest were forced by Congress to do so or lose highway funding (sound familiar?) and capitulated. ZT laws had to set the BAC limit for drivers under 21 at 0.02 or less in order to pass muster for Congress.  And until now, virtually every study has supported their effectiveness.

Grant's study found that ZT laws had essentially no effect on traffic fatality rates or the BAC distribution of fatalites. Data was from 1988-2000, during which time all states had a 21 drinking age. Several control variables and fixed effects were accounted for. Whether a state had a "partial" ZT law (a lower BAC limit for youth that was either >0.02 or had an age limit below 21) before adopting a "full" ZT law (satisfying the federal mandate), which several did, was also accounted for. Similar effects were observed on daytime and nighttime fatalities, despite the fact that most alcohol-related fatalities are concentrated at night. In addition, effects were compared to drivers over 21, who should not be affected by the ZT laws, but such effects were in fact similar, which suggests a spurious relationship due to unknown coincident factors. Finally, the fractions of total fatalities that were zero BAC, low BAC, and high BAC drivers were equivalent before and after a ZT law was passed, further casting doubt on its effectiveness.

What are we to make of all this? Previous studies were either pre/post studies that omitted key variables, or longitudinal studies that were otherwise less thorough than this one. This study appears to trump the rest, and it is thus very tempting to say that zero tolerance = zero intelligence.

That being said, we at 21 Debunked still recommend that if the drinking age is lowered to 18, which we strongly advocate for all 50 states, that the ZT laws generally remain as is, with an age limit of 21. Better yet, ZT laws ought to be strengthened by making the BAC limit 0.02 for the first x number of years of licensed driving for all ages, or age 21, whatever is longer. Several countries with lower drinking ages (Canada, Germany, Netherlands, etc.) interestingly set ZT age limits higher than the legal drinking age and/or base it on the number of years of licensed driving one has, with the latter making even more sense.  Puerto Rico, with a drinking age of 18, also recently adopted a 0.02 limit for those under 21. And there are good reasons for America maintaining some type of "full" ZT law even after the drinking age is lowered.
  • ZT laws send a strong message that alcohol and driving don't mix, and removing them may be seen by some as implying otherwise.
  • Inexperience in drinking as well as driving, regardless of age, can be a deadly combination.
  • If the drinking age was 18, one would have three years of legal drinking experience and up to five years of driving experience at 21.
  • ZT laws would be a good precautionary measure against the alleged adverse effects of lowering the drinking age, and would help assuage fears about such effects.
  • Federal law requires ZT laws as a condition of highway funding. It's bad enough to have to deal with that kind of coercion with the drinking age, let alone additional funding losses.
  • Some studies find positive externalities from ZT laws, such as reduced suicide rates, which may or may not be causally related. 
  • ZT laws have strong public support, and frankly, it would be politically impossible to lower the drinking age without maintaining such laws.
However, we recommend that laws that set the BAC limit to 0.00 be changed to 0.02 to prevent false positives, and violating a ZT law (but not violating the adult BAC limit) should not be a criminal offense. For those who violate adult BAC limits, there should be graduated penalties that are proportional to BAC, eliminating the current perverse incentives resulting from reduced marginal penalties for each additional drink above the 0.08 threshold. And those penalties should be tougher than the current ones as well. In addition, we also recommend lowering the adult BAC limit of 0.08 down to 0.05, like it is in many other countries, since impairment begins well before 0.08. In addition, preliminary breath testers (i.e. handheld ones) should be not be sufficient evidence; that's what evidential breath testers are for.

But most importantly, we should never lose sight of the fact that the average BAC in an alcohol-related fatality is 0.16 in general and 0.14 for drivers under 21. Moreover, only 5.8% and 5.4% of total fatalities, for all ages and ages 16-20, respectively, involve a BAC below 0.08--and most of such deaths would probably still have occurred in the absence of booze.  And drivers age 21-24 are the most overrepresented in alcohol-related fatal crashes (at all BACs) as well.  Thus, the highest priority in the fight against impaired driving should be catching and deterring those very high BAC drivers of all ages (who are responsible for the lion's share of such deaths), rather than strategize on how best to catch young drivers with BACs below 0.05 (who are responsible for only a small fraction). In other words, we need to see the forest for the trees.

No comments:

Post a Comment