More About Guam

As you already know, much to our chagrin Guam was in the news for hastily raising the drinking age to 21 in July 2010.  That makes them the first part of the USA to change the drinking age in over two decades.  They were in the news again recently in August.  The first is that they will actually get tougher on DUI by requiring a mandatory overnight jail stay and will prosecute cases within 48 hours, instead of the former policy of "catch and release" that made it such a joke before.  (This we certainly applaud, by the way.) The second was the fact that the arrest rate for DUI had been skyrocketing since 2007, especially for younger drivers.  Aside from being the major impetus for the latest change in DUI criminal procedure, this fact was also used by some to retrospectively justify the drinking age hike to 21.

But the latter claim does not stand up to closer scrutiny.  In fact, it falls flat on its face.  The stats from the Guam Police Department show the following numbers, in a population of about 175,000 residents:

Year	Total DUI Arrests	Under 21	%  Under 21
2007	790	58	7.3%
2008	677	70	10.3%
2009	1146	116	10.1%
2010 (first half)	382	41	10.7%

Clearly arrests have risen for all ages, and doubled for those under 21 in two years, though the share of arrests under 21 has essentially plateaued since 2008, after jumping from 2007 to 2008.  Back in 2005, it was only 6%.  However, arrest rates can be quite deceiving, as the table of fatalities below so clearly shows:

Year	Total Fatalities	Alcohol Related	Under 21	% Under 21
2007	24	11	1	9%
2008	8	5	0	0%
2009	13	9	1	11%
2010	10	2	0	0%
Avg.	15.7	7.7	0.57	7.4% (3.7%)

Here we see a very different picture indeed.  It does not appear that alcohol-related fatalities have been rising for any age group.  Quite the opposite in fact, a whopping 64% decrease overall, and thus the reason for rising arrests is most likely greater enforcement and targeting of younger drivers, as opposed to more drunk driving. You read right that in 2008 and 2010, there have actually been zero traffic fatalites of those under 21.  The 2010 data only include the first half of the year (up to June 30), during which the drinking age was still 18, so one can thus project 20 total deaths and 4 total alcohol-related deaths for the whole year, and either zero or one death under 21, had the status quo remained. 

As for the percentage under 21, since alcohol involvement is not given for the under 21 data, we assumed the worst (that all of them involved booze) and calculated the number of under-21 deaths as a percentage of total alcohol-related deaths.  This gives 7.4%, but if we assume that half of the under-21 deaths involve booze (a reasonable estimate given the all-ages data), we get a mere 3.7%.  Thus, drivers under 21 are overrepresented in arrests, but underrepresented in fatalities.  Put another way, even if all under-21 drinking was to somehow magically disappear, over 96% of the deaths would most likely still occur.

How does this compare with the rest of the nation, where the drinking age has been 21 since 1988?  Well, research shows that in 2008, drivers under 21 accounted for 12% of total fatalities and 13% of alcohol-impaired fatalities.  Clearly worse than Guam by any measure, but remember that 21-24 year olds are the worst of all in terms of overrepresentation in drunk driving deaths, a fact that is true in almost every developed nation in the world regardless of drinking age.  Thus, these data are hardly a ringing endorsement for a 21 drinking age.