Why Are Traffic Deaths on the Rise? (Part Deux)

After decades of a massive secular decline in traffic deaths, reaching an all-time record low in 2014 per VMT as well as per capita, such deaths have been creeping up again since then.  2015 and 2016 both saw national increases in fatalities, and while 2017 saw a slight decrease from 2016, the number of deaths still remains stubbornly higher than it was before the increase.  In fact, 2015-2016 is the largest two-year jump in deaths in half a century.  So why has progress stalled and begun to reverse in recent years?

The list of most likely factors includes the following:

• Lower gas prices
• An improving economy since the Great Recession
• An increase in distracted driving (and walking), primarly from smartphones
• Higher speed limits than in the past 
• Infrastructure in disrepair from decades of gross neglect
• Slacking on traffic safety improvements in general since the early 1990s

All of these things are true, and all of them are known to be correlated with traffic casualties.  Other factors are involved as well, to be sure, but these are the big ones.  The first three are the proximal causes, while the last three are the more distal ones.

Of course, drunk driving and not wearing seatbelts remain rather persistent contributors to the number of these deaths, but such behaviors remain far lower than they were decades ago.  Nevertheless, they remain at dangerous levels, and it is apparently a bit too early to feel safe in that regard.  And with real alcohol prices at record lows today and alcohol consumption on the rise for the past two decades, there is definitely a cause for concern in that regard.

What about drugged driving, then?  Is it really on the rise, like some have claimed?  Perhaps, but it may simply be that we are getting better at detecting it rather than an actual increase.  Or perhaps it is a bit of both.  The opioid epidemic certainly doesn't make the roads any safer, with such doped-up drivers nodding off behind the wheel.  And contrary to the anti-legalization folks, there does not seem to be any firm link between cannabis legalization and traffic fatalities.  In fact, some studies have found decreases in highway deaths following cannabis liberalization, due to an apparent substitution with alcohol (and perhaps opioids as well).  As for the specious claim of preliminary evidence linking the increase in pedestrian deaths in some legalization states in the first half of 2017 with legalization, that does not really pass the smell test because 1) not all legalization states even saw any increase during that time, 2) small numbers tend to fluctuate wildly, and 3) why would cannabis legalization only affect pedestrian deaths and not other traffic deaths during that time as well?

One thing is for sure.  Whether this spike in traffic casualties is a short-term blip or the start of a longer-term trend (which will only be known in hindsight), it should be a major wake-up call that we clearly cannot afford to be complacent about it any longer.  The USA has seriously lagged behind other industrialized countries for decades in terms of progress on traffic safety (all of which have lower drinking ages than we do, interestingly enough), and we need to catch up, yesterday.  That includes the safety of pedestrians and cyclists as well, who have borne the brunt of the recent increase in traffic deaths.  And even when vehicle miles (or kilometers) traveled are taken into account, the USA still has either higher fatality rates and/or has seen less progress since the 1980s compared with nearly all other industrialized (and even semi-industrialized) nations.

A short list of things we can do include:

• Crack down on drunk driving, drug-impaired driving, reckless driving, and distracted driving--yesterday.
• Stiffen the penalties for hit-and-run crashes--yesterday.
• Reduce speed limits, especially on side streets, arterials, and smaller highways--yesterday.
• Install speed cameras and red-light cameras in more places (but be sure to also lengthen the yellow lights and double-reds to prevent it from backfiring with more rear-enders).
• Raise the gas tax by a penny per week until it is at least 50 cents/gal higher than now ("A Penny for Progress") and/or implement a carbon tax-and-dividend scheme.
• Raise the alcohol taxes significantly as well (note how those taxes have been lagging behind inflation for decades in most states).
• Design roads/streets to be more pedestrian-friendly and bike-friendly as well.
• Invest more in public transportation, as well as "safe-rider" programs as well. 
• Rebuild America's neglected and crumbling infrastructure, generating millions of new jobs in the process.
• And last but not least, make the road test harder like it is in many other countries, and make driver's licenses easier to lose for serious and/or repeated traffic violations.

After all, if it saves even ONE life, it's worth it, right?  Thought so.  So what are we waiting for?

Which Ancillary Laws Actually Work to Reduce Teen Drinking? (Spoiler Alert: Almost None)

The legal drinking age (at least in terms of purchase and public possession) has been 21 in all 50 states and DC since 1988 (notwithstanding the Louisiana Loophole through 1995), but states and localities have differed since then on the "ancillary" laws that are used to prop up the 21 drinking age.  These laws include, among others:

• Laws against private possession and/or consumption 
• Internal possession laws
• Laws against furnishing alcohol to people under 21
• Dram shop liability laws
• Social host liability laws (civil or criminal)
• Keg registration 
• Use alcohol and lose your license ("use and lose" laws)
• Zero-tolerance laws for DUI
• Fake ID laws and ID scanner laws
• Age limits for serving alcohol

Most if not all of these laws have been studied to some extent, with mixed results and even some serious plot twists.  Most of these studies have been conducted by MADD-affiliated researchers such as James C. Fell, Robert Voas, and Ralph Hingson, so they should probably be taken with at least a grain of salt, if not a whole pound.  But even with this potentially significant bias, most of these laws were found to have either no effect, inconsistent effects, very small effects, or even perverse effects in terms of "alcohol-related" traffic fatalities among young people.

If these laws (and by extension, the 21 drinking age itself) did work as intended, one would expect the effectiveness of these laws to show up not only in traffic fatality statistics (which are the tip of a very large iceberg), but also in surveys of teen drinking as well, especially when recent data are studied.  A 2014 study done by Vanessa H. Sacks et al. of Child Trends examined the relationship between 14 different ancillary laws (and alcohol taxes) and both current drinking (any in the past 30 days) and "binge" drinking (5+ drinks "within a couple of hours") among high school students from 2005-2011 as reported on the Youth Risk Behavior Survey (YRBS) of those years.  The 14 laws studied were the following:

• Laws against consumption of alcohol by people under 21
• Exceptions to possession laws
• Exceptions to furnishing laws
• Social host liability laws 
• BAC limits for people under 21 (i.e. zero-tolerance laws)
• Use-and-lose laws (strength measured in three different ways)
• ID scanners in retail locations
• Keg registration
• Age limits for serving alcohol 
• Beer excise taxes in 2005
• Distilled spirits excise taxes in 2005

And the results?  All but three such laws had no significant or even perceptible effect individually on either current drinking or "binge" drinking rates.  The only three that did have any statistically significant correlation were:

• Beer taxes were significantly and negatively associated with both current drinking and "binge" drinking rates.  No surprise there, as this dovetails nicely with the reams of evidence that have found similar results.  (Distilled spirits taxes showed no correlation either way, but that apparently null result is likely due to multicollinearity since beer and liquor taxes are highly correlated with each other).
• Keg registration laws were significantly and negatively correlated with current drinking, but not "binge" drinking, in one of two models.  This result should probably be interpreted with caution though, given how multiple other studies have found a positive correlation between keg registration and "alcohol-related" youth traffic fatalities.
• Use-and-lose laws (i.e. driver's license penalties for mere possession or consumption of alcohol) showed a positive correlation with current drinking, that is, such laws seemed to perversely increase teen drinking.
• And the real kicker: after controlling for drinking rates in previous years and the number of ancillary laws in previous years, states with a greater total number of such policies perversely had higher rates of both current and "binge" drinking.  But drinking rates in previous years did not predict the number of policies in place in later years, thus ruling out the possibility of reverse causation.

Thus, while the researchers caution that these results alone are not definitive enough to establish causation, it is quite clear that such results certainly cast serious doubt on the effectiveness of such laws, and by extension the 21 drinking age itself, in reducing teen drinking and related problems.  With the notable exception of alcohol taxes, the effects of such laws are most likely negligible or even perverse overall.  That, of course, dovetails rather nicely with Miron and Tetelbaum (2009), who found that raising the drinking age to 21 seems to have had "only a minor impact on teen drinking".  And we couldn't agree more.