Transcript Overview of MADD: Reasons for Success

Which Underage
Drinking Laws Have
Been Effective?
James C. Fell
Senior Research Scientist
Pacific Institute for Research & Evaluation
August 25, 2014
The Problem
Young Drivers and Alcohol
Graduated Driver Licensing
Underage Drinking
Minimum Legal Drinking Age 21 [MLDA-21]
Background
MLDA-21 Law Components
Prior Research
Current Research
Methods
Strengths of the MLDA-21 Laws
Results of Current Research
Should the MLDA Remain at 21?
The Problem
What are some of
the leading causes
of death for young
people?
Leading Causes of Death
Ages 16-20: 2009
Rank
Cause
Percent of Number
Deaths
of Deaths
1
Motor Vehicle Crashes
28%
3,947
2
Homicide
17%
2,386
3
Suicide
14%
1,948
4
Other Injury (falls,
drowning, etc.)
10%
1,446
5
Cancer
5%
714
6
Heart Disease
3%
423
All Other Causes Combined
22%
3,016
TOTAL
100%
13,880
Leading Causes of Death
Ages 16-20: 2007
Rank
Cause
Percent of
Deaths
Number of
Deaths
1
Motor Vehicle Crashes
34%
5,360
2
Homicide
17%
2,711
3
Suicide
11%
1,784
4
Accidental Poisoning
8%
1,178
5
Cancer
5%
724
6
Heart Disease
3%
396
All Other Causes (falls,
drowning, etc.)
22%
3,543
TOTAL
100%
15,696
Young Drivers
and Alcohol
The current problem
and what is working
to reduce their crash
risk
Young Drivers
Aged 15-20 in 2012

More than one out of four (28%) deaths of young
people aged 16-20 are from motor vehicle crashes.

More than one out of five (22%) young drivers aged
15-20 involved in fatal crashes had been drinking.

Almost three out of four (71%) young drinking
drivers killed in crashes were not wearing safety
belts.

Half (50%) of the underage 21 drinking drivers
involved in fatal crashes were speeding.
Young Driver
Over-Involvement in Fatal Crashes
Ages 15-20
13.6% of All Drivers
Involved in Fatal Crashes
13% of All Alcohol- Involved
Drivers in Fatal Crashes
14
12
10
6.4% of All
Licensed Drivers
8
6
4
2
0
Young Licensed Drivers Young Drivers Involved in Young Alcohol-Involved
Fatal Crashes
Drivers in Fatal Crashes
Source: NHTSA Traffic Safety Facts, 2003
The Problem
The problem is particularly acute
for 16-year-old drivers
16-year-old drivers have crash rates that
are 3 times that of 17 year olds, 5 times
that of 18 year olds, and even twice that
of 85 year olds.
 15% of 16 year olds involved in fatal
crashes in 2012 had been drinking
(BACs>=.01)

What are some of
the reasons for the
high crash rates of
young drivers?
The Research: Risk Factors

Inexperience – Especially the first few months of
licensure

Late night driving (without supervision)

The number of teen passengers

The presence of alcohol (and other drugs) in young
drivers

The low rate of seat belt usage

Type of vehicle driven

Distractions (especially cell phones)

Attention Deficit and Hyperactivity Disorder (ADHD)
Crashes per Million Miles by
Driver Age in the U.S.
50
40
30
20
10
0
50- 5516 17
17 18
18 19
19 2020- 2525- 3030- 3535- 4040- 4545- 5055- 6060- 6565- 7070- 7575- 8080- 85+
85+
16
DRIVER AGE
Percent of Licensed Drivers in Crashes in
One Year by Age and Gender in the U.S.
50
male
female
40
30
20
10
0
16 17
17 18
18 19
19 2020- 2525- 3030- 3535- 4040- 4545- 5050- 5555- 6060- 6565- 7070- 7575- 8080- 85+
85+
16
DRIVER AGE
Crash Rates in the U.S. by Driver Age and
Passenger Presence Per 10,000 trips
6
5
4
3
2
1
0
1
2
ages 16-17
3+
0
1
2
3+
ages 18-19
number of passengers
0
1
2
ages 30-59
3+
Graduated Driver
Licensing
Graduated Driver Licensing (GDL)
in the U.S.
Premise:

Permit beginning drivers to acquire on-the-road
driving experience under low-risk conditions

Adopt special restrictions and requirements after
initial learning phase

Allow young drivers to “graduate” to full license
privileges after successfully completing certain
stages
Graduated Driver Licensing
in the U.S.
Stage 1: Learner’s Permit – Model Components










Minimum age requirement
Vision/knowledge tests
Basic skills training
Licensed adult (at least age 21) required in vehicle at all
times
Teenage passenger limitations
All occupants must wear safety belts
No alcohol or drugs (zero tolerance: BAC<.02)
Crash- and conviction-free for at least six months
Parental certification of practice hours
Distinctive permit from other licenses
Graduated Driver Licensing
in the U.S.










Stage 2: Intermediate or Provisional License – Model
Components
Complete Stage 1
Minimum age requirement
Behind-the-wheel test
Advanced driver education training
Teen passenger restrictions
All occupants must wear safety belts
Nighttime driving restrictions: Licensed adult required in vehicle
for all late night driving (when drinking most likely to occur)
Zero tolerance (no alcohol or drugs)
Crash- and conviction-free for at least 12 consecutive months
Distinctive license from full license
Graduated Driver Licensing
in the U.S.
Stage 3: Full Licensure



Complete Stage 2
Minimum age requirement
Zero tolerance (no alcohol or
drugs) up to age 21
Graduated Driver Licensing
Evaluations of Effectiveness in the U.S. and Canada
Florida
9% reduction in crashes for drivers aged 16-17
California
5% reduction in crashes; 10% reduction in
convictions for drivers aged 16-17
Oregon
16% reduction in crashes for 16-17 year old male
drivers
North
Carolina
26% reduction in crashes for 16 year old drivers
Michigan
27% reduction in crashes for 16 year old drivers
Nova
Scotia
24% reduction in crashes for 16 year old drivers
Ontario
31% reduction in crashes for drivers aged 15-19
Graduated Driver Licensing
Insurance Institute for Highway Safety Ratings:
“Good” GDL Law if it has 5 out of the following 7
components:

Minimum age for learner’s permit

Mandatory waiting period to advance to intermediate stage

Minimum hours of supervised driving

Minimum age for intermediate license

Nighttime restriction in intermediate stage

Passenger limitation in intermediate stage

Minimum age for full unrestricted license
Graduated Driver Licensing
National Studies in the U.S.

Dee, et al., (2005) – Significant 4.4% reduction in 15- to 17year-old driver deaths attributable to GDL laws when covariate
laws included

Chen, et al., (2006) – GDL associated with an 11% lower fatal
crash involvement rate for 16 year olds; 16-21% reduction with
GDL programs with 5 of 7 key components

Morrisey, Grabowski, Dee & Campbell (2006) – “Good” GDL
programs reduce fatalities among 15- to 17-year-old drivers by
19%; “Fair” GDL programs reduce nighttime fatal crash
involvements by 13% but no effect on daytime fatal crashes;
“Marginal” GDL programs had no effect.

Vanlaar, et al., (2009) – Meta-analysis showed
significant reduction of fatal crashes of 16-yearold drivers, but not 17, 18, or 19 year old drivers
Graduated Driver Licensing
National Studies in the U.S.

McCartt, et al., (2009) – GDL laws rated as “Good” had
30% lower fatal crash rate for 15- to 17-year-old drivers
compared to GDL laws rated as “Poor”

Lyon, Pan & Li (2012) – a strict GDL permit stage was
associated with a 58% reduction in fatal crash risk by 16year-old drivers compared to a lenient permit stage.
Intermediate stage associated with a 44% reduction in
fatal crashes but little effect on injury crashes. Strongest
effects were seen for crashes with passengers, followed
by nighttime crashes, and then overall crashes.
Graduated Driver Licensing
Three National Studies in the U.S.: First Study Results
Fell, Jones, Romano, Voas (2011)
Traffic Injury Prevention, 12(5), 423-431

States that adopt a basic GDL law can expect a decrease
of 7-11% in the proportion of 16- to 17-year-old drivers
involved in fatal crashes relative to 21- to 25-year-old
drivers depending upon their other existing laws (zero
tolerance; primary belt; .08 BAC; use and lose)

States that adopt “Good” GDL laws can expect decreases
of 8-13% in the proportion of 16- to 17-year-old drivers
involved in fatal crashes relative to 21- to 25-year-old
drivers depending upon their other existing laws (zero
tolerance; primary belt; .08 BAC; use and lose)
Graduated Driver Licensing
Three National Studies in the U.S.: Second Study Results
Fell, Todd, Voas (2011):
Journal of Safety Research, 42, 283-290

GDL nighttime restrictions reduce 16- to 17-year-old
driver involvements in nighttime fatal crashes by 10%

Nighttime restrictions reduce 16- to 17-year-old
drinking drivers in nighttime fatal crashes by 13%.

Passenger restrictions reduce 16- to 17-year-old driver
involvements in fatal crashes with teen passengers by 9%
Graduated Driver Licensing
Three National Studies in the U.S.: Third Study Results
Romano, Fell, Voas (2011)
Annals of Advances in Automotive Medicine, 55, 51-61

Found GDL effects for White, African-Americans and
Asian young drivers but not Hispanic young drivers

GDL had no apparent effect on speeding-related fatal
crashes of novice drivers

The observed differences involving young Hispanic drivers
might be related to factors other than GDL (e.g., SES
causing young Hispanics not to drive as much as Whites).
Graduated Driver Licensing Effectiveness
Reductions in 16-17 year old drivers in fatal crashes
Fell, Jones, Romano,
Voas, 2011
-7% GDL
when
other
laws
present
-11% when
GDL alone
GDL effects
Fell, Todd,
Voas, 2011
-13% for
drinking
-10% in
nighttime 16-17
year
fatal
crashes old driver
in
nighttime
fatal
crashes
Nighttime restriction
effects
Fell, Todd,
Voas, 2011
-9% in fatal
crashes with
teen
passengers
Passenger
limitation effects
Graduated Driver Licensing
Summary






GDL is a rational way to deal with novice driver problems
GDL allows beginning drivers to gradually gain
experience under low-risk conditions
GDL provides incentives to drive safely to graduate to full
licensure
Surveys indicate substantial support for GDL by parents
GDL significantly reduces crashes involving 16- to 17year-olds
GDL can save thousands of lives in the future if all States
adopt it and if it is extended to all young drivers under age
21.
So, has Graduated
Driver Licensing
been a total
success?
Graduated Driver Licensing
Recent National Study
Masten, Foss, Marshall (2011)
Journal of the American Medical Association, 306(10), 1098-1103

“Good” GDL programs associated with lower incidence of
fatal crashes for 16-year-old drivers by 26%, resulting in
about 1,348 fewer fatal crashes.

However, “Good” GDL programs were associated with a
12% increase in fatal crashes by 18-year-old drivers,
resulting in about 1,086 more fatal crashes.

Lead author (Masten) speculated that the increase in 18year-old fatal crashes was due to them delaying licensure
in the “Good” GDL states because of the strong restrictions.
Graduated Driver Licensing
Additional Recent Analyses
(based upon Fell, Romano & Voas, 2013)
RESULTS



Any GDL law reduces fatal crash involvements of 16year-old drivers by 17% (p<.05) saving 1,945 lives over
the years.
“Good” GDL laws reduce fatal crash involvements of 16year-old drivers by 20% (p<.05) saving 2,347 lives over
the years.
However, 18-year-old fatal crash involvements increased
by 11% (p<.05) associated with “Good” GDL laws,
resulting in 2,724 fatalities over the years. Why?
Graduated Driver Licensing
Possible Reasons for the Conflict:




Drivers aged 18-19 are skipping GDL phases and driving
at a later age---reducing their experience
Drivers aged 18-19 are exhibiting more risk-taking
behaviors (e.g. impaired driving; lack of safety belt use;
distracted driving)
Drivers aged 18-19 have increased exposure to risk for a
fatal crash (more late night driving; driving on high speed
roads; driving more miles)
Drivers aged 18-19 who have gone through two phases of
GDL lack experience under high-risk conditions because
of GDL restrictions
Graduated Driver Licensing
Further Research is Needed





Should States require GDL up to age 21? Only one State
does (New Jersey).
Has the license status and license type changed over the
years for 16-, 17-, and 18-year-old drivers compared to
the same ages in fatal crashes before GDL?
Have 18-year-old drivers delayed licensure? In which
states? Why?
What other components of GDL laws are effective?
Minimum ages, holding periods, practice hours, etc.?
Which GDL components affect impaired driving the most?
Underage Drinking
What percent of 1820 year olds in the
U.S. report drinking
alcohol?
Some Young People Do Drink, But a
Smaller Percent Compared to Adults: 2002
Drinking Among Youth
(past 30 days)
12- to 14-year-olds
18- to 20-year-olds
5% 1%
10%
24%
18%
94%
49%
72%
27%
15- to 17-year-olds
Drinking Occasions
0
1 to 4 5 or more
Source: NSDUH, 2002
Some Young People Do Drink, But a
Smaller Percent Compared to Adults: 2012
Drinking Among Youth
(past 30 days)
12- to 14-year-olds
18- to 20-year-olds
3% 1%
5%
15%
19%
55%
96%
80%
26%
15- to 17-year-olds
Drinking Occasions
0
1 to 4 5 or more
Source: NSDUH, 2012
But, Young Drinkers Drink More Heavily
than Adult Drinkers: 2002
Comparison of drinking patterns for adult and
underage drinkers (past 30 days)
100%
35%
80%
28%
50%
57%
60%
40%
65%
72%
50%
43%
20%
0%
12- to 14- year-olds
15- to 17- year-olds
Non-bingers
18- to 20- year-olds
Bingers
Source: NSDUH, 2002
Adults
But, Young Drinkers Drink More Heavily
than Adult Drinkers: 2012
Comparison of drinking patterns for adult and
underage drinkers (past 30 days)
Non-bingers
Bingers
Source: NSDUH, 2012
Relative Risk* of Being Involved
in a Fatal Crash by BAC
Driver Age
.05 - .079
BAC
.08 - .099
>.15
16-20
6.24
12.61
490.41
21-34
4.78
8.74
200.03
35+
4.03
6.89
111.94
*Risk relative to BAC=.00 for same age group

Relative risks are the same for men and women at a given BAC.
Relative risk for 16-20 year old women are now the same as 1620 year old men at a given BAC (a change from 1996).
Source: Voas, Torres, Romano, Lacey, JSAD, (2012)
Relative Risk of Being Killed as a Driver
in a Single Vehicle Crash
(Relative to BAC =.00)
RELATIVE RISK BY AGE
16 - 20
BAC
Both Male/Female
21 - 34
35+
Both Male/Female
Both Male/Female
.10+
15 - 68
14 - 38
15 - 29
.08
.06
7 - 21 times 7 - 14 times 7 - 11 times
.05
.04
3 - 5 times
3 - 4 times
.02
.00
Source: Voas, Torres, Romano, Lacey (2012)
3 - 4 times
Minimum Legal
Drinking Age
Laws
Minimum Legal Drinking Ages
Around the World
Australia – 18
Canada – 18
France – 16
Italy – 16
Japan – 20
Sweden – 18
Switzerland – 14
Great Britain – 16
USA – 21
Venezuela – 18
Source: WHO
Background
Minimum Legal Drinking Age 21
(MLDA-21)
History in the United States

Minimum Legal Drinking Age (MLDA) Laws
established in the States after the Repeal of
Prohibition in 1933
(21st Amendment to the US Constitution)

Most States set the MLDA at 21
(e.g., 11 States set and stayed at 21):
AR (1935); CA (1933); IN (1934); KY (1938);
NV (1935); NM (1934); ND (1936); OR (1933);
PA (1935); UT (1935); WA (1934)
Minimum Legal Drinking Age 21
History in the United States

Voting age lowered from 21 to 18 in 1971 (26th
Amendment to the US Constitution)

Many States began to lower the drinking age to
18 or 19 in the early 1970s

Studies in the 1970s and 1980s showed
significant increases in alcohol-related fatal
traffic crashes involving youth aged 16-20 in
States that lowered their drinking age
Minimum Legal Drinking Age 21
History in the United States



Mothers Against Drunk Driving (MADD) founded
in 1980
President’s Commission on Drunk Driving
established in 1982
MADD, President’s Commission, Members of
Congress recommended a National Uniform
Drinking Age Law set at 21 in order to reduce
youthful alcohol-related fatalities and eliminate the
“blood borders”
MLDA 21 Milestones
President Ronald
Reagan signs federal
21 Minimum Drinking
Age law in 1984.
All 50 States plus DC
adopt age 21 as
minimum legal
drinking age (illegal to
possess alcohol and
illegal to purchase
alcohol) by 1988 or
risk the withholding of
a portion of Federal
Highway Construction
funds.
US Minimum Legal Drinking Ages as of December 31, 1969
35 states had 21 as the drinking age
WA
NH
ND
MT
VT
MN
ME
SD
OR
ID
WI
WY
NY
MI
MA
IA
NE
IL
NV
UT
CO
NJ
MO
KS
PA
OH
IN
WV
KY
CA
DE
VA
RI
CT
MD
TN
OK
AZ
NC
AR
NM
LA
DC
SC
MS
AL
GA
TX
FL
AK
21 years old
20 years old
18 years old
HI
US Minimum Legal Drinking Ages as of December 31, 1975
Only 12 states had 21 as the drinking age
WA
NH
ND
MT
VT
MN
ME
SD
OR
ID
WI
WY
NY
MI
MA
IA
NE
IL
NV
UT
CO
NJ
MO
KS
PA
OH
IN
WV
KY
CA
DE
VA
RI
CT
MD
TN
OK
AZ
NC
AR
NM
LA
DC
SC
MS
AL
GA
TX
FL
AK
21 years old
20 years old
19 years old
18 years old
HI
US Minimum Legal Drinking Ages as of July 1, 1988
All states had 21 as the drinking age
WA
NH
ND
MT
VT
MN
ME
SD
OR
ID
WI
WY
NY
MI
MA
IA
NE
IL
NV
UT
CO
NJ
MO
KS
PA
OH
IN
WV
KY
CA
DE
VA
MD
TN
OK
AZ
NC
AR
NM
LA
SC
MS
AL
GA
TX
FL
AK
HI
DC
RI
CT
What Have Been the
Effects of the Minimum
Legal Drinking Age 21
in the United States?
What Have Been the Effects of
the Minimum Legal Drinking
Age 21 in the United States?

Reduction in alcohol consumption by those <21

Reduction in drinking driver fatal crashes by
those <21

Reduction in alcohol-related homicides, suicides,
unintentional injuries by those <21
Research

Studies in the 1980s and 1990s show that
raising the drinking age saves lives

Studies from the National Highway Traffic
Safety Administration (NHTSA) show that
MLDA 21 laws reduce 18- to 20-year-old driver
involvements in fatal crashes by 13% (Arnold,
1985; Womble, 1989; Kindelberger, 2005)

MLDA-21 laws estimated to save 800-900
lives each year in reductions in traffic fatalities
Drivers Over Age 21 Involved
in Fatal Crashes, 1982-2004
Sources: FARS; NHTSA
+38%
Sober drivers age 21+
Drinking drivers age 21+
-33%
Drivers Under Age 21 Involved in
Fatal Crashes, 1982-2004
Sources: FARS; NHTSA
+9%
Sober drivers under age 21
Drinking drivers under age 21
-62%
Other than illegal to
possess and illegal to
purchase alcohol if you
are under age 21, what
other MLDA-21 laws are
there in the States?
MLDA-21
Law
Components
20 Key Components of Underage
Drinking Laws in the United States
MLDA 21 Law Components
CORE LAWS:
 Apply to Youth
 Possession
 Purchase/attempt to purchase
EXPANDED LAWS:
 Apply to Youth
 Consumption
 Internal possession
 Use and lose driving privileges
 Use of fake ID illegal
 Apply to Youth Driving
 Zero tolerance
 GDL with night restrictions
# States with Law
51
48
35
9
40
51
51
51
20 Key Components of Underage
Drinking Laws in the United States
MLDA 21 Law Components

Apply to Providers
 Furnishing/selling
 Age 21 for on-premises
Server (all 3 beverage types)
 Age 21 for on-premises
Bartender (all 3 beverage types)
 Age 21 for off-premises Seller
 Keg registration
 Beverage Service Training
 Retail Support Provisions for Fake ID
 Hosting underage drinking parties
 Dram Shop Liability
 Social Host Civil Liability
# States with Law
51
13
24
23
31
38
45
28
45
33
20 Key Components of Underage
Drinking Laws in the United States
MLDA 21 Law Components
Apply to Manufacturers of Fake ID
 Transfer/production of Fake ID illegal
 Apply to State
 State control of alcohol
(at least 1 beverage)
# States with Law

24
11
20 Key Components of Underage
Drinking Laws in the United States

Utah has all 20 MLDA-21 Components

Kentucky has only 9 out of the 20 laws

Georgia has 11 of the laws

Only 5 laws have been adopted by all 50
States and DC
5 MLDA-21 Laws Adopted by all States
and DC in the United States
Illegal to possess alcohol
 Illegal to use a fake ID to purchase alcohol
 Zero Tolerance (BAC>.02) for driving
 Graduated Driver Licensing System
 Illegal to furnish alcohol to persons under
age 21

Georgia’s MLDA-21 Laws











Possession
Purchase
Use & Lose
Fake ID Use
Zero Tolerance for Driving
GDL with Night Restrictions
Furnishing
Keg Registration
Fake ID Retail Support
Dram Shop
Social Host Civil Liability
Prior Research
from PIRE
Prior Research
We evaluated 6 of the 20 MLDA-21 laws:
The Impact of Underage Drinking Laws
on Alcohol-Related Fatal Crashes
of Young Drivers
By Fell, Fisher, Voas, Blackman & Tippetts
Alcoholism: Clinical and Experimental Research
Vol. 33, No. 7, July 2009
Results
MLDA 21 Law Effects on Ratio of <Age 21
Drinking Drivers to Non- Drinking Drivers
in Fatal Crashes (1982-2004)
20%
Core Laws
Possession/
Purchase
Use & Lose
Keg
Zero Tolerance
Registration
Driving
GDL with
Night Restriction
12%
(p<.001)
10%
No Effect
0%
-5%
(p=.026)
-10%
-20%
-16%
(p<.001)
-5%
(p=.015)
Results
Traffic Safety Law Effects on the Ratio of Underage 21
Drinking Drivers to Non-Drinking Drivers
in Fatal Crashes
(1982-2004)
.08 BAC
Per Se
.10 BAC
Per Se
Seat Belt
Law
ALR
10%
0%
-3%
(p=.041)
-8%
-10%
(p=.002)
-7%
( p =.065)
-5%
(p =.024)
Results
Traffic Safety Law Effects and Enforcement
on the Ratio of Over Age 25 Drinking Drivers
to Non-Drinking Drivers in Fatal Crashes
(1982-2004)
10%
.08 BAC
Per Se
ALR
.10 BAC
Per Se
Sobriety
Checkpoint
Frequency
0%
-1.4%
(p=.004)
-6%
(p<.001)
-10%
-4%
(p<.001)
-4%
(p=.042)
Current
Research
Methods
MLDA-21 Laws
We selected for evaluation:
 Three False Identification (FID) laws:




(1) FID for minors;
(2) FID for suppliers
(3) FID for retailer support; and
Two social host laws:


(4) Social Host Prohibition (SHP); and
(5) Social Host Civil Liability (SHCL).
MLDA-21 Laws
Fake ID laws
(1)
for minors prohibited the use of false identification cards
by minors (i.e., minors under age 21);
(2)
for suppliers prohibited a person from lending,
transferring to, or producing false identification for an
underage person (i.e., suppliers); and
(3)
for retailer support established policies which allowed
one or more of the following for retailers of alcohol
outlets: the use of identification scanners, distinctive
licenses for those under age 21, seizure of identification
cards deemed false, or the right to sue an underage
person that uses false identification to purchase alcohol
(i.e., retailer support).
MLDA-21 Laws
Social Host Laws
1. Social Host Civil Liability (SHCL) laws involve the
liability faced by noncommercial alcohol servers for injuries or
damages caused by their intoxicated or underage drinking
guests. In states with SHCL laws, injured third parties are
able to sue social hosts (as well as the minor who caused the
crash) for monetary damages. SHCL law is established by
statute or by a state court through common law.
MLDA-21 Laws
Social Host Laws
2. Social Host Prohibitions (SHP) involve underage
furnishing and host party criminal policies which can result in
penalties imposed by the state (e.g., fines, jail, etc.).
Data Sources
Legal Research

Alcohol Policy Information System (APIS) developed and
maintained by the National Institute on Alcohol Abuse and
Alcoholism (NIAAA); Westlaw; HeinOnline
Crash Analyses

Fatality Analysis Reporting System (FARS) developed and
maintained by the National Highway Traffic Safety
Administration (NHTSA) for the years 1982-2010
Model
The model we used is composed of:
1.
Three FID laws (i.e., minor, retailer support, and supplier);
2.
Two versions of the social host laws (i.e., social host prohibitions (SHP)
and social host civil liability (SHCL) laws)
3.
Three drinking-and-driving laws (0.08 legal limit, 0.10 legal limit, and
ALR);
4.
Two driving-safety laws (use of sobriety checkpoints and seatbelt use
laws);
5.
Two variables determining economic strength (employment rate and
VMT by state and year);
6.
Per capita beer consumption by state and year;
7.
Ratio of underage 21 drinking drivers to underage 21 non-drinking
drivers in fatal crashes; and
8.
Ratio of drinking drivers to non-drinking drivers in fatal crashes among
drivers aged 26 years and older as a comparison.
Structural Equation Modeling
(SEM)
Strengths of the
Laws
Scoring
Fake Identification – Minor
Scoring Criteria
Fake Identification law
License sanction procedure
Weight Point Values
+1.0 point for presence of the law
+2.0 points for administrative
sanctions
OR
+2.0 points for administrative and
judicial sanctions
OR
+1.0 points for judicial sanctions
only
Scoring
Fake Identification – Supplier
Scoring Criteria
Lending, transfer or sale of false IDs
criminalized
Manufacturing and distribution of
false IDs criminalized
Weight Point Values
+1.0 point if yes
+1.0 point if yes
Scoring
Fake Identification – Retailer Support
Scoring Criteria
Incentives for retailers to use
scanners
Use of distinctive licenses
Seizure of suspicious ID by retailer
permitted
Right to sue minor
Affirmative defense
Weight Point Values
+1.0 point if yes
+2.0 points if yes
+1.0 point if yes
+1.0 point if yes
-1.0 point for general affirmative
defense
0.0 points for specific affirmative
defense or none
Scoring
Scoring Criteria
Weight Point Values
Social Host Prohibitions
Type of statute
+2.0 points for general
OR
+1.0 point for specific
Underage guest actions triggering violation +1.0 point for possession
+1.0 point for consumption
+1.0 point for intention to possess or
consume
Property type covered by law
+1.0 point for residence
+1.0 point for outdoor area
+1.0 point for other areas
Knowledge standard
+2.0 points for negligence
+1.0 point for knowledge
0 points for overt act
Preventive actions available to offender
-1.0 if yes
Exceptions to law
-1.0 point for family
-1.0 point for resident of household
-1.0 point for other
Scoring
Social Host Civil Liability
Common law or common law and statutory +3.0 points
law
Statutory law only
+3.0 points
Limitations to the law
-1.0 point for limitation on who may be
sued
-1.0 point for limitation on elements or
standards of proof
Which of the 5 MLDA-21
laws have had a
significant effect on
underage drinking
driver fatal crashes?
Results
Effects of five MLDA-21 laws on the ratio of <age 21
drinking drivers to <age 21 nondrinking drivers involved
in fatal crashes (Source: FARS 1982-2010)
Regression Weights
and Significance Level
for Direct Effects Including Strength of Fake Identification and Social Host MLDA-21 Laws
[values in bold and italics indicate a significant effect]
Model including strength of FID
and SHP laws
Predictor
Estimate
S.E.
p-value
Outcome = under 21 FARS ratio
Alcohol-impaired driving laws
-.160
.019
<.001
Driving safety laws
-.151
.007
<.001
Economic strength
.001
.003
.866
FID minor laws
-.003
.006
.564
FID supplier laws
-.010
.005
.030
FID retailer laws
-.004
.003
.218
SHP laws
-.001
.001
.937
SHCL laws
-.006
.003
.054
Beer consumption
.181
.022
<.001
Conclusions


Only 24 States have Fake ID Supplier Laws
and only 33 States have Social Host Civil
Liability Laws
If all States adopt fake ID supplier laws
(effect size 1.0%) an additional 30 lives per
year would be saved—of which 14 lives are
currently being saved in the 24 States that
have the law.
Conclusions

Our prior study (Fell et al., 2009) estimated
that the two core MLDA-21 laws (possession
and purchase) and the zero tolerance law for
driving (BAC>.02 illegal) are saving 732
lives per year.

Use & Lose laws are saving an additional 132
lives annually and if all states adopted Use &
Lose laws an estimated 165 lives could be
saved per year.
Conclusions

An estimated 878 lives per year are
being saved by these five effective laws.
If all states adopted all five laws, up to
927 lives could be saved each year.
What does NHTSA
officially say about
MLDA-21 Laws?
Conclusions
Future Research

Eleven of the 20 MLDA-21 Laws have now
been evaluated for their effectiveness on
underage drinking driver fatal crashes.

Under a NIAAA grant [R03 AA022446] we
are currently evaluating:




Dram Shop Laws
State Alcohol Control Laws
Responsible Beverage Service Training Laws
Eventually all 20 MLDA-21 Laws with
strengths will be evaluated.
Acknowledgements
■
This study was conducted under a grant from the
National Institute on Alcohol Abuse and Alcoholism
(NIAAA) [R21 AA019539]. We thank Mr. Gregory Bloss of
NIAAA for his excellent guidance and Dr. Ralph Hingson
of NIAAA for his comments and suggestions.
■
The scoring of the 20 underage drinking laws was
conducted under a NIAAA grant [R03 AA022446] and a
grant from the Office of Juvenile Justice and
Delinquency Prevention (OJJDP) [2012-AH-FX-0005].
■
This study is scheduled to be published in Traffic Injury
Prevention in October 2014.
Should the
Minimum Legal
Drinking Age
Remain at 21?
Have you heard
about the Amethyst
Initiative led by
college president
John McCardell?
John M. McCardell, Jr.
Director of Choose Responsibility
President Emeritus of Middlebury College
and current Vice Chancellor of
The University of the South
in Sewanee, TN
MLDA-21 Does Not Make Sense
At 18, you can vote, get married, serve in
the military, smoke, etc.
 Why can’t you drink at 18?
 MLDA-21 is frequently and easily broken
 Alcohol has become a forbidden fruit for
those under 21

John McCardell
MLDA-21 Promotes Binge Drinking
Promotes drinking by <21 in clandestine
locations without supervision
 Reason for a 50% increase in binge
drinking by college students
 Reason for a significant increase in alcoholrelated off-road underage 21 deaths
 Reason for increase in underage 21 alcohol
poisoning deaths
John McCardell

MLDA-21 Not a Traffic Safety Issue
Anymore
More youth die in off-road accidents than
on-road
 Canada had a similar decrease in underage
21 drinking driver fatal crashes and they
have MLDAs of 18 and 19
 Youth in European countries with lower
drinking ages don’t have the problems we
do in the US
John McCardell

Solution
Repeal the Federal Law
 Lower the MLDA to 18
 Require 40 hours of alcohol education to
obtain a “drinking license”
 License revoked for any violations
 This will Promote Responsible drinking
 Drinking will occur in supervised
environments
John McCardell

Amethyst Initiative
135 College Presidents have signed the
initiative to take a harder look at the
MLDA-21
 Keep in mind: This is only 3-4% of the
3500 college presidents in the U.S. who
have signed off

Initiated by John McCardell
Is Dr. McCardell
right?
Evidence of MLDA 21 Law
Effectiveness









Williams, Zador, Harris, Karpf, 1983
Arnold, 1985
Decker, Graitcer, Schaffner, 1988
Womble, 1989
O’Malley & Wagenaar, 1991
Toomey, Rosenfeld & Wagenaar, 1996
Voas, Tippetts & Fell, 2003
Ponicki, Gruenewald & LaScala, 2007
Many Others
Reviews of the Literature on MLDA 21



Wagenaar & Toomey, 2002

Examined all of the research from 1960 to 2000 when
states lowered the drinking age and then raised the
drinking age

Every one of the 46 high-quality studies that found an
effect found that the MLDA 21 saves lives
Shults, et al, 2001

Centers for Disease Control reviewed all of the high
quality peer-reviewed published studies on drinking ages

Concluded:

Lowering the MLDA = crash increase of 10%

Increasing the MLDA = crash decrease 16%
One of the best studied public health laws in history
30-day Alcohol Use by Youth Under Age
21: Minimum Drinking Age Effect
13% Decline
Years Before and After MDA was
Raised
Source: O’Malley & Wagenaar (1991)
Evidence of MLDA 21 Law
Effectiveness
Drinking drivers over age 21 involved in fatal crashes:
the decrease between 1982 and 2004 was
-33%
Drinking drivers under age 21 involved in fatal crashes:
the decrease between 1982 and 2004 was
-62%
MLDA 21 accounted for much of the difference
(Hedlund, et al., 2001)
Canada

Experienced similar decreases in underage 21
drinking drivers in fatal crashes even though
MLDA was 18-19

However, drinking drivers over age 21
experienced almost the same rate of decrease

Therefore, factors other than the MLDA
accounted for the Canadian decrease

Remember, in the U.S., the <21 drinking driver
rate decreased twice as much as the >21
drinking driver rate
More Evidence of Effectiveness
Additional studies have shown a decrease in other
injuries associated with MLDA 21:

Decrease in homicides by 18-20 year-olds
o


Jones, Pieper, and Robertson, 1992
Decrease in suicides by 18-20 year-olds
o
Links, 2000
o
Birckmayer and Hemenway, 1999
Decrease in other unintentional injuries by 1820 year-olds
o
Jones, Pieper, and Robertson, 1992
Do European youth
drink responsibly?
The European Drinking Age Myth
A higher percentage of young people from a majority of
European countries in 2007 report:





Experimenting with alcohol
Drinking in the past year
Drinking in the past 30 days
Heavy episodic drinking
Intoxication
26 countries show higher rates of adolescent
intoxication compared to the US rate.
9 countries show lower rates.
Prevalence of Intoxication in Past Year - 2007
Adolescents Aged 15-16
Sources: Hibell, et al., 2009; Johnston, O’Malley, Bachman, & Schulenberg, 2008
First Intoxication Before Age 13 - 2007
Adolescents Aged 15-16
Sources: Hibell, et al., 2009; Johnston, O’Malley, Bachman, & Schulenberg, 2008
Drunkenness Among 15-Year-Olds,
2005-6 (Drunk at least twice in life)
Source: O.E.C.D.; Currie et al. (2008)
July 17, 2008
French Combat Youth
Binge-Drinking
By Bruce Crumley / Paris
50% increase in 15-24
year olds hospitalized for
excess drinking since 2004
50% of 17 year olds
reported being drunk in the
past month
Alcohol is the leading factor
in deaths among French
youth
December 17, 2008
Binge Drinking Comes
to France
By David Chazan
Binge drinking is now on the
rise among young people
France used to take pride in
initiating children into the art of
sipping wine with their parents
But these days it does not
seem to work
The government recognizes
the problem and plans to raise
the legal age for buying alcohol
from 16 to 18 next year
How long has binge
drinking in U.S.
colleges been going
on?
Binge Drinking

Straus & Bacon (1953): Drinking in College –
discusses binge drinking since 1920s

Straus & Bacon (1953) estimated 65% of
college students drank in past 30 days in 1950

College Alcohol Survey (2007) estimated 70%
of college students drank in past 30 days

College binge drinking is cultural and not
caused by the MLDA-21
Is College Binge Drinking Increasing?
Percent of College Students Who Report
Being Drunk in the Past 30 Days
Source: Monitoring the Future, College Students and Adults, 1975-2006, NIH
Are Young People Binge
Drinking More?
5+ drinks in one session past 2 weeks
19-20 Year Olds
Source: Monitoring the Future, 1975-2006, NIH
Underage Binge Drinking:
1975-2007
[% who had 5+ drinks in a row in previous 2 weeks]
Source: Monitoring the Future 2007
PERCENT
60
40
20
1 2 th G ra d e
1 0 th G ra d e
8 th G ra d e
0
‘7 5 ‘7 7 ‘7 9 ‘8 1 ‘8 3 ‘85 ‘8 7 ‘8 9 ‘9 1 ‘9 3 ‘9 5 ‘9 7 ‘9 9 ‘0 1 ‘0 3 ‘0 5 ‘0 7
YEAR
How many young
people under age 21
die of alcohol
poisoning each
year?
Alcohol-Attributable Deaths
Youth Under Age 21 – 2001
(Source: CDC, Midanik, et al, MMWR, 2004)










Motor Vehicle Crashes
Homicides
Suicides
Poisoning (drug and alcohol)
Alcohol Poisoning
Drowning
Fires
Falls
Etc.
TOTAL
2,075
1,227
480
234
18
125
41
37
……….
4,492
Number of Deaths Attributable to the Harmful Effects of
Excessive Alcohol Use Among 18-20 and 21-24 Year Olds:
Alcohol Poisoning 2001-2006
Sources: CDC; Alcohol-Related Disease Impact (ARDI) Software;
Alcohol Attributable Deaths (AAD)
28
Alcohol Attributable
Deaths (Alcohol
Poisoning)
30
24
21
10
10
12
25
21 - 24 Year Olds
19
20
9
24
11
10
10
18 - 20 Year Olds
0
2001
2002
2003
2004
2005
2006
Number of Deaths Attributable to the Harmful Effects of
Excessive Alcohol Use Among 18-20 and 21-24 Year Olds:
Non-Highway Injury Deaths 2001-2006
Alcohol Attributable Deaths
(Non-Highway)
Sources: CDC; Alcohol-Related Disease Impact (ARDI) Software;
Alcohol Attributable Deaths (AAD)
300
267
269
255
258
255
260
141
21 - 24 Year
Olds
143
200
150
151
142
143
18 - 20 Year
Olds
100
0
2001
2002
2003
2004
2005
2006
Surgeon General’s Call to Action to
Prevent and Reduce Underage
Drinking, 2007
Reducing Underage Drinking:
A Collective Responsibility
National Academy of Sciences, Institute of Medicine, 2004
Not one page, not one
paragraph, not one word
even suggests that
lowering the legal drinking
age is part of the solution
Has the MLDA 21 Been Effective?
Yes.

40-50 peer-reviewed studies say it has

Research also shows that when the drinking age
is lowered, alcohol related problems increase:
traffic fatalities, suicides, STDs, assaults, etc.

Raising the drinking age reduces problems

Lowering the drinking age increases problems
Does the MLDA 21 eliminate
underage drinking? No.
Should It Remain at 21?
Yes.
Those who start drinking at age 18 vs. those
who start at age 21:

15% become alcohol dependent vs. 9%

2.4 times the odds of being in a motor vehicle
crash as a drinking driver

1.6 times the odds of being in a fight after
drinking
Has the MLDA 21 Been Harmful?
No.
There is no evidence of associations with
drinking age 21 and the following:

Increases in alcohol poisoning

Binge drinking in uncontrolled, clandestine
environments

Alcohol-related injuries of any kind
What are some of
the arguments
people use to
advocate lowering
the drinking age?
Argument: “If I’m old enough to vote and
go to war, I should be old enough to drink.”
Response:
Many rights have different ages of initiation:
Driver’s license
Age 16
Vote
Age 18
Serve in the Military
Age 18
License to carry handgun
Age 21
Rent a Hotel Room
Age 21
Rent a Car
Age 25
Serve in the House
Age 25
Elected President
Age 35
The brain is not fully developed until about age 25.
Argument: “If people cannot drink until
they are 21, will they just drink more
when they reach 21?”
Response:
■
Research shows that they will not drink more
once they turn 21
■
Early legal access to alcohol is associated
with higher rates of drinking later in life
■
Those who wait until they are 21 to start
drinking have significantly fewer alcoholrelated problems later in life
Argument: “Since 19 and 20-year-olds
drink anyway, wouldn’t it be better to let
them drink in controlled settings?”
Response:
■ Are bars “safe” and “controlled”?
−
Many individuals stopped for drinking and driving
come from a bar
−
79% of licensed establishments will sell to an
obviously intoxicated person
■
MLDA 21 saves many 19- and 20-year-old lives
■
We need to enforce age-21 law in all settings
New Zealand Natural Experiment

It has been 30 years since the last state in the
U.S. lowered its drinking age.

If a state were to lower its drinking age under
the current conditions—would youthful crash
injuries increase?

New Zealand lowered its drinking age from
20 to 18 in 1999, providing an opportunity
to study such a reduction.
New Zealand Study Conclusions
(Kypri, et al., 2006)

The size of the effect for 18- to 19-year-olds (increases
of 12% for males and 51% for females in traffic crash
injuries) was remarkable given the legal exceptions to
the pre-1999 law and its poor enforcement.

There were large “trickle-down” effects for 15to 17-year-olds (increases of 14% for males and 24% for
females).

There were significantly more alcohol-involved crashes
among 15- to 19-year-olds than would have occurred
had the purchase age not been reduced to 18.
No Evidence that Lowering MLDA to 18
for those who Complete an Alcohol
Education Course will Reduce
the Problem

Driver education courses did not reduce crashes
for beginning drivers

DARE education program did not reduce drug
usage by youth

How can we expect an 18 year old to “choose
responsibility” when they are drinking alcohol?
They actually lose their inhibitions when under
the influence.
College Interventions (IOM, 2004)

Campuses should adopt comprehensive
evidence-based approaches:





Screening and brief interventions
Limit alcohol availability and access for underage
students
Consistent enforcement of laws and policies
Universal educational approaches as well as
selective and indicated approaches
NIAAA and SAMHSA should continue to fund
evaluation of college-based programs and should
maintain list of evidence-based programs.
Promoting a College or University Climate
that De-Emphasizes the Role of Alcohol:
Strategies that Work
Dr. Jonathan Gibralter
President, Frostburg State University
Policies and Procedures

Identified campus and community policies and
practices addressing alcohol use and abuse

Addressing off-campus behavior as violations
of University policy—in Student Code of
Conduct

Increased communications with two other local
colleges

Are bringing High Schools into the mix also.
Results

Reduction in off-campus citations

Number of repeat offenders has decreased
dramatically—almost no third offenses.

Built awareness that off-campus parties also
consist of non-students

Improved relations with community

Addressed high-risk events

Students understand the message

Off-campus events involve fewer people
What can we do
over the next 5
years to further
reduce underage
drinking?
Strategies That Show Evidence
of Reducing Underage Drinking

Strengthen the Laws

Better Enforcement of the Laws (against the
youth and the providers)

Educate Parents

Increase the excise tax and the price of alcohol
– consumption goes down (10% increase in
price results in a 5% decrease in consumption)

Delay the age of onset of drinking

De-normalize binge drinking
U.S. Public Supports MLDA-21
 The
public is in favor of keeping the MLDA at 21
 Latest
Gallup Poll showed 74% opposed lowering
the MLDA to 18
 Virtually
every major public health organization
supports the MLDA-21 law:
AMA
CDC
CADCA
 Only
NHTSA
GHSA
NTSB
CSPI
IIHS APHA
MADD
NSC
SADD
135 college presidents (3-4%) out of 3500
colleges signed the Amethyst Initiative
New Brain Research Validates 21

Alcohol on the developing brain can
interfere with learning and memory

The hippocampus in an adolescent who
drinks heavily can be up to 10% smaller

Teenage drinkers score worse on
vocabulary and memory tests and are
more likely to perform poorly in school
147
149
Why Should the Minimum
Legal Drinking Age Stay at 21?

Minimum drinking age 21 laws save 800-900
lives per year in reductions in traffic fatalities
involving young drivers.

Early onset of drinking increases the risk for
future alcohol abuse problems, crashes and
assaults.

European countries with lower drinking ages
experience higher percentages of youth that
report intoxication in the past month.

Emerging medical research shows that the
brain is not fully developed until about age
25 and excessive drinking by youth under age
21 may cause brain damage as well as reduce
brain function.
47%
QUESTIONS?
Strategies for Discussion
 What strategies can you offer to change social
norms around binge drinking?
 What strategies can you offer to avoid adverse
circumstances for someone else’s excessive
drinking?
 Knowing the drinking culture on local college
campuses, what recommendations would you
make to the college administration to reduce the
adverse effects of underage drinking? To law
enforcement?
Contact Information
James C. Fell, M.S.
Pacific Institute for Research and Evaluation (PIRE)
11720 Beltsville Dr., Suite 900
Calverton, MD 20705-3111
301-755-2746
[email protected]