Alcohol Prices, Consumption, and Traffic Fatalities
Author(s): Douglas J. Young and Agnieszka Bielinska-Kwapisz
Source: Southern Economic Journal, Vol. 72, No. 3 (Jan., 2006), pp. 690-703
Published by: Southern Economic Association
Stable URL: http://www.jstor.org/stable/20111841 .
Accessed: 10/06/2014 13:52
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
Southern Economic Association is collaborating with JSTOR to digitize, preserve and extend access to
Southern Economic Journal.
http://www.jstor.org
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
Southern Economic
Journal
2006, 72(3), 690-703
Alcohol Prices, Consumption,
Traffic Fatalities
Douglas
We
and Agnieszka
J. Young*
examine
the relationships
from 1982
across U.S.
between
taxes
between
prices,
previous
and
consumption,
studies have
traffic
fatalities
found
large, negative
price data suggest
using data
associations
and
alcohol
from measurement
estimators
alcohol
Some
fatalities.
used
little or no
However,
commonly
and fatalities.
These
result
prices
apparently
conflicting
findings may
error and/or endogeneity
in the price data, which
biases
least squares
ordinary
a finding of no price effects. Using
taxes as instrumental
alcohol
fatali
variables,
alcohol
connection
Bielinska-Kwapisz|
among
to 2000.
states
and
toward
ties are found
to be negatively
related to prices.
biases may
tively related to fatalities. However,
as instruments.
In addition,
still remain,
alcohol
is strongly posi
consumption
taxes are not entirely
suitable
because
JEL Classification: II, H2, C3
1. Introduction
Traffic
are
fatalities
a
have
to significantly
the potential
This
on
focuses
paper
cause
leading
of
death,
premature
among
particularly
under
people
is involved in about 40% of all traffic fatalities, and hence alcohol policies
35 years of age.1 Alcohol
reduce
the
links
fatality
between
rates.2
alcohol
prices,
and
consumption,
traffic
fatalities.
The
fundamental question is, How does the price of alcohol affect fatalities? Price is an important policy
variable,
since
actually
set by
Schematically,
Although
to price,
taken
taxes
there
and
is little
thus
Economic
about
the qualitative
the magnitudes
the estimated
theory
for some
beverages
predicts
that alcohol
are expected,
?
?
are Tax
Price
relationships
of
and
in price
increases
dispute
estimates
together,
and other measures,
authorities.
the hypothesized
of quantitative
whether,
by
control
alcohol
related
negatively
range
it is affected
magnitudes
nature
of
these
sense.
Some
step
states,
to reduce
Fatalities.
relationships,
there
a further
researchers,
for
be
fatalities.
??
and
is
price
will
consumption
paribus,
Consumption
at each
involved
make
ceteris
in some
is a wide
question
example,
about
have
estimated reduced form relationships based on tax and fatality data, ignoring the intermediate rela
tionships
between
taxes
and prices,
prices
and consumption,
and consumption
*
208A Linfield Hall, Montana
Department of Agricultural Economics and Economics,
59717-0292 USA; E-mail djyoung@montana.edu;
corresponding author.
t College of Business, Montana State University, Bozeman, MT 59717-3040 USA.
and fatalities.
State University,
Even
these
Bozeman, MT
This research was supported by the National Institute on Alcohol Abuse and Alcoholism
under grant R03 A A13264.
two anonymous reviewers, and co-editor Dek Terrell. All conclusions and any
Helpful remarks were provided by Jon Nelson,
errors are the sole responsibility of the authors.
Received June 2004; accepted March 2005.
See http://www.cdc.gov/nchs/data/nvsr/nvsr49/nvsr49_08.pdf,
2
Based on estimates from the National Highway Traffic
1
NCSA/TSFAnn/TSF2003EarlyEdition.pdf.
Estimates
http://www.niaaa.nih.gov/databases/crashO
1 .txt.
Table
10.
Safety Administration,
http://www-nrd.nhtsa.dot.gov/pdf/nrd-30/
are lower:
from the National Institute on Alcohol Abuse and Alcoholism
690
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
Alcohol Prices and Traffic Fatalities
studies
have
alcohol
taxes,
the few
However,
that taxes may
suggesting
studies
(2002)
find
indeed
have
a
that estimated
are relatively
There
estimates.3
and Bielinska-Kwapisz
Young
approach.
different
markedly
produced
that beverage
prices
substantial
price-fatality
few
on
impacts
a structural
for one with
and fatalities.
consumption
inconsistent
yielded
relationship
taking
than one
studies
rise more
results.4
A key empirical concern is the quality of the price data that is typically employed
and
fatalities,
consumption,
other
behaviors.
studies
Many
have
beer,
employed
in analyses of
are
data
to suffer
likely
from
error
measurement
substantial
and
(Young
and
wine,
prices collected by the American Chamber of Commerce Researchers Association
these
691
spirits
(ACCRA), but
Bielinska-Kwapisz
2003).5 In addition, beverage prices may be endogenous in the sense that higher demand may result in
higher market prices (Manning, Blumberg, and Moulton 1995).
Measurement error in the price data implies that the ordinary least squares (OLS) estimator is biased
and
inconsistent.
consistent.
may
alcohol
However,
excise
and
taxes,
both
models,
problems
beer,
in control
renders
the estimated
higher
taxes?including
state markups
also
prices
bias
how much
underestimate
substantially
of
endogeneity
Similarly,
In simple
prices
price
the OLS
estimator
elasticity
toward
discourage
and
wine,
and
consumption
unit
per
spirits
a set of
states?provides
biased
zero.
excise
and
That
traffic
taxes,
fatalities.
percentage
variables
instrumental
in
is, OLS
in
that,
principle, can resolve the problems with the price data. This paper applies instrumental variable (IV)
to the
techniques
estimation
substantial
provide
evidence
consumption
larger
and more
suggest
that
than
significant
taxes
are not
and
from OLS.
suitable
as
consumption-fatalities
error
of price
those
fully
and
measurement
of
IV estimates
and
data,
the price-fatalities
of
and/or
on
effects
consumption
a priori
However,
in both
endogeneity
considerations
the
are
fatalities
and
price
substantially
some
and
results
The
linkages.
evidence
instruments.
2. Methods
We
estimate
consumption),
that
models
regression
socioeconomic
per
implies
squares
4
capita.
and
nonnegative
3
variable,
dependent
fatalities
that
has
yit,
(Greene
2003,
is the
The
logistic
been
widely
the disturbance
the
characteristics,
=
y it
The
express
wit
section
as
rates
and
environment,
legal
functions
state
alcohol
of
and
year
(or
prices
fixed
effects:
\nPita + */f ? + u?+ vt + wit. (1)
transformation
logistic
transformation
used
fatality
restricts
in previous
is heteroskedastic,
studies.6
and
of
the
the
observed
predicted
the
However,
thus Equation
rate,
fatality
value
of
logistic
1 is estimated
rit,
fatalities
that
is
to be
transformation
by weighted
least
21.4.6).
For examples, Cook (1981); Chaloupka, Saffer, and Grossman
(1993); and Ruhm (1996) find large, negative relationships
between beer taxes and fatalities, whereas Dee (1999); Mast, Benson, and Rasmussen
(1999); and Young and Likens (2000)
find little or no relationship.
and Schenzler (1994) find that alcohol price is negative and significant at the 10% level in one of three
specifications, but that it is sensitive to the inclusion of time-fixed effects. Price effects in Young and Likens (2000) are small,
negative, and statistically insignificant.
5
The ACCRA data have been used in studies of alcohol consumption by Gruenewald, Ponicki, and Holder (1993); Kenkel
(1993); Manning, Blumberg, and Moulton
(1995); Beard, Gant, and Saba (1997); Grossman, Chaloupka, and Sirtalan (1998);
Sloan, Reilly,
and Nelson
6
(2003); studies of traffic accidents, homicides, suicides, and other deaths by Sloan, Reilly, and Schenzler (1994);
spouse abuse by Markowitz
(2000); alcohol-related motor vehicle fatalities by Young and Likens (2000); sales tax incidence by
Besley and Rosen (1999); and suicides by Markowitz, Chatterji, and Dave (2002), among others.
Young and Likens (2000) compare the logistic specification with a linear probability model
differences in the results.
and find that there are only small
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
692
J. Young and Agnieszka Bielinska-Kwapisz
Douglas
The price of alcohol, P, is the Stone price index, a geometric weighted average of the prices of beer,
and
wine,
with
spirits,
to each
equal
weights
share,
beverage's
oc,on the price variable in Equation
to the price
respect
of
For
alcohol.8
1 is approximately the elasticity of fatalities with
a coefficient
example,
on alcohol:7
expenditure
(2)
ln/?=?>lnpy,
j
The coefficient,
in national
9y,
of -0.1
a
that
implies
10%
in
increase
alcohol prices is associated with a 1% decline in fatalities. Similarly, when the natural logarithm of per
capita alcohol consumption is entered instead of price, the coefficient is approximately the elasticity of
with
fatalities
to consumption.
respect
If beverage prices are measured with error, then the price index will be correlated with the
disturbance term and theOLS estimator of the price coefficient is biased and inconsistent (Greene 2003,
section
If there
5.6).
a
is only
variable
single
to classical
subject
measurement
then
error,
the OLS
estimator is biased toward zero (attenuated). Similarly, if price is endogenous, it is correlated with the
disturbance term, and OLS estimates aweighted average of the demand (negative) and supply (positive)
coefficients.
Similar
Thus,
endogeneity
reasoning
The
in alcohol
if a
set
in pooled
prices
consumption
error
of
and
the OLS
bias
estimator
replaces
be
eliminated
can
variables
from
away
as a right-hand
price
can
endogeneity
instrumental
proper
Although
anti-alcohol
or
taxes
taxes
is the case,
data
to be correlated
In particular,
alcohol.
sentiment,
If this
time-series
are unlikely
toward
attitudes
alcohol.
cross-section
taxes
alcohol
unmeasured
toward
alcohol
also
be
side
by
found.
values.
negative
variable.9
standard
two-stage
and
Young
Bielinska
(2002) show that state and federal excise taxes andmarkups explain about 30% of the variation
Kwapisz
stronger
will
to measurement
due
methods
estimation
when
applies
biases
of prices
may
change
are not
proper
to that employed
similar
errors
with
taxes
over
may
be
higher
instrumental
in which
to changes
In addition,
variables.
study.10
reflect
they may
in states
in response
time
in this
of measurement,
there
is
in attitudes
taxes may
be correlated with other policies intended to reduce alcohol use, drunk driving, and/or fatalities, and
data limitations may make it impossible to explicitly include these policies as control variables. In this
taxes
case,
will
We
conclude
variables
to deal
7
be
for use
exogenous
correlated
as
state
that
with
with
the
disturbance
term,
and
taxes
thus
not
will
be
appropriately
instruments.11
and
federal
the measurement
excise
error
taxes
problem,
are
and markups
but may
not
be
likely
fully
to be
good
satisfactory
instrumental
to deal
with
data are reported quarterly for a varying number of cities in each state. We aggregate to the state level by
(2002, 2003).
computing simple averages of the city-quarter data. Details are described in Young and Bielinska-Kwapisz
8
The elasticity is exactly equal to (1 - r)a, which is very close to a, because 1 r = 0.999.
9
Consumption may be correlated with the disturbance term because of measurement error or because some unmeasured factors
The ACCRA
to both drinking and fatalities. For example, some youth engage in a number of risky behaviors that include
reckless
drinking,
driving, and/or unprotected sex (Dee and Evans 2001; Gruber 2001; Grossman, Kaestner, and Markowitz
2002). Alcohol consumption is then correlated with fatalities both because it is in fact a causal factor, and as a reflection of
are common
example, as
underlying attitudes toward risk. The latter correlation is spurious, in that a reduction in alcohol consumption?for
a result of higher prices?does
not change attitudes toward risk. These biases in OLS estimates of the effects of consumption
error biases the estimator toward zero, whereas spurious
on fatalities are likely to operate in opposite directions: Measurement
correlation makes the estimator too positive.
10
Beer taxes by themselves explain only about 5% of alcohol price variation, and
employing the broader set of tax variables on spirits and wine, and including
11
See Manning, Blumberg, and Moulton
(1995), footnote 4. Kubik and Moran
taxes are endogenous. Brown, Jewell, and Richer (1996) find that county-level
Eisenberg (2003) concludes
because of endogeneity.
that existing estimates of the effects of 0.08
therefore are less adequate as instruments than
state markups and percentage taxes.
(2003) provide evidence that changes in beer
prohibitions on alcohol sales are endogenous.
laws and graduated licensing programs are overstated
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
693
Alcohol Prices and Traffic Fatalities
of
problems
regressions
sentiment
toward
socioeconomic
and
alcohol
The
correlation.
endogeneity/spurious
include
variables,
over
its evolution
latter
state
time
and/or
dummies,
the sample
to the extent
is mitigated
problem
trends
that
the
control
that
for
period.
3. Data
The
data
data
are missing
price
states
the contiguous
include
for various
1,...,
(i=
in some
years
and
48)
the years
1982-2000
so the total number
states,
19), but
1,...,
(t=
is 869.
of observations
Separate
regressions are estimated for fatalities in the total population and the population aged 16-20, which we
term "teen"
a significant
display
Separate
6 p.m.
6 p.m. Monday-Friday,
involve
alcohol.
for one
account
fatalities
plus
of
closely
related
to one
or
prices
occur
As
even
these
sense,"
are
than
consumption
day
between
Fatalities
times."14
third of fatalities,
and
6 p.m.
these
weekend
night
in each
case.
weekend
6 a.m.
and
Friday
6 a.m.
less
account
to
night
for only
one
be more
should
fatalities
and
likely
weekend
indicates,
times
at other
fatalities
rates
Thus,
are much
la
fatality
between
occurring
Table
though
quarter
time.
to 6 a.m. Monday,
Sunday
"other
to "make
accident
Sunday.13
6 p.m.
fatalities
the
Both
one
by about
declining
on
alcohol,
6 a.m.
and
the results
to alcohol
involve
occurring
these
fourth
For
the week.
seventh
to
Saturday
those
term
We
computed
likely
1980s,
based
of the total population.
those
twice
the early
since
also
are most
and between
Saturday,
trend
are
rates
which
rates are about
fatality
downward
fatality
fatalities,
night
Teen
fatalities.12
times.
Per capita alcohol consumption declined steadily from the early 1980s tomid 1990s before leveling
off at about 2.2 gallons of pure ethanol per capita (Nelson 1997; Nephew et al. 2002). The real price of
alcohol generally declined except for an upward blip in 1991 when federal excise taxes increased.15
Extensive
drinking
have
educational
and
fatalities.
concluded
that
in part
measure,
measure
environment
by
a result.
The
several
changed
of dummy
last
adopted
educational
developed
a series
summing
often
in the
changes
eventually
as
declined
states
and
legal
laws were
belt
fatalities
because
enforcement,
and
efforts
Seat
variables
two
by
programs.
indicating
also may
other
laws
simultaneously,
We
form
affected
index
studies
are
of
a state has
to
harder
up
stepped
an
or not
whether
have
numerous
and
state,
every
of many
effects
laws
decades
hard-to
the
legal
(i) an open
container law, (ii) a preliminary breath test law, (iii) a dram shop law, (iv) an illegal per se blood
alcohol content (BAC) level of at least 0.1, (v) mandatory licensing action upon first conviction for
driving
the
under
innovations
are
influence
included
illegal per se BAC
(DUI),
as
separate
(vi)
and
variables
an
administrative
in order
to provide
per
se
law.16
evidence
Two
on
more
their
recent
effectiveness:
level of 0.08 or less and a server training law.17 In addition,
legal
an
teen fatality
12
Fatality data were provided by NHTSA from the FARS database (http://www.nhtsa.dot.gov/).
13
An alternative approach is to use NHTSA's measures of alcohol-related fatalities. However, these data are imputations based
on samples of drivers who were actually tested for alcohol, and these samples varied greatly over time and across states (http://
Selection bias may be significant: Drivers were more likely to be tested if beer
www.niaaa.nih.gov/databases/crash06.txt).
cans were found in the back seat. The current method has been used by Dee (1999) and other researchers.
14
There were three instances of zero fatalities in the original data for teens during other times: North Dakota in 2000 and Rhode
Island in 1993 and 1994. Since the logistic transformation can not be computed with a zero fatality rate, we assumed that one
fatality occurred in each instance.
15
The price of alcohol is adjusted for national inflation using the CPI and expressed relative to the overall ACCRA cost of living
in each state. It therefore represents the cost of alcohol relative to other goods, expressed in dollars of 2000 purchasing power.
averages of ACCRA prices closely follow the CPI price index for alcohol "at home." The price of alcohol "away
from home" has increased faster. See Young and Bielinska-Kwapisz
(2002).
16
Data from NHTSA, Digest of Highway Safety Legislation, various years.
17
Data on the 0.08 law is from NHTSA's Digest of Highway Safety Legislation. Server training data provided by Alexander C.
School of Public Health.
Wagenaar, Alcohol Epidemiology
Program, University of Minnesota,
National
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
694
J. Young and Agnieszka Bielinska-Kwapisz
Douglas
Statistics
la. Descriptive
Table
Mean
Definition
Variable
RHS endogenous variables
Ethanol consumption
ACON
Ethanol price ($/gal)
APRICE
Minimum
Standard Deviation
Fatality rates (per 1000 population)
All ages, all times
TOTAL
TOTALWN All ages, weekend nights
TOTALOT All ages, other times
TEEN
Ages 16-20, all times
TEENWN
Ages 16-20, weekend nights
TEENOT
Ages 16-20, other times
0.172
0.042
0.087
0.346
0.115
0.138
0.066
0.051
0.015
0.027
0.106
0.045
0.048
0.011
0.029
0.132
0.022
0.020
0.430
25.0
1.24
212.6
2.37
269.3
(gal/capita)
Maximum
0.423
0.138
0.240
0.937
0.342
0.511
5.26
337.8
RHS exogenous variables
=
0.0
1.0
1
0.427
0.760
Seat belt law
SBELT
1.07
1.0 6.0
3.27
LEGALENV Legal environment index
0.384 1.0 0.0
0.08
0.180
ILLPER08
Illegal per se
0.422 1.0
0.0
0.237
SERVTRAN Server training law
3.12
36.715.3
25.3
Income per capita ($ 000s)
INCOME
7.75
2.09
12.9
Vehicle miles traveled (OOOs/driver)
VMTLIC
areas
46.4
0.0
4.28
8.49
in
DRY
(%)
dry
Population living
13.82
2.30
18.72
POP1829
Population ages 18-29 (%)
18.0
0.749
20.7
DRINKAGE Legal drinking age
1.0
0.0
0.312
0.110
KEGREG
Keg registration
1.00.0
0.485
0.420
YOUTHBAC Lower BAC for youth = 1
12.3 1.74
21A
63.5
CATHOLIC Catholic (%)
6.51
0.1
1.51
Mormon (%)
MORMON
10.1
0.0
37.4
7.96
Southern Baptist (%)
SBAPTIST
7.6 18.6
2.04
12.4
POP65
Population ages 65+ (%)
1.03
0.27
Share of GSP from hotels/restaurants (%) 0.76
TOURISM
22.4
24.0
21.0
76.2
15.1
regressions
include several laws directed specifically at this age group: the legal drinking age, keg
registration,
and
Southern
Baptist;
on
effect
and
the
included
for
the percentage
of
state
of gross
percentage
fatalities
via
drink less, but are more
Table
lb displays
consumption
(DISCUS
addition,
Instrumental
product
alcohol
states
variable
BAC
(lower)
the percentages
of
the population
over
and
the hotel
from
these
consumption,
sometimes
may
relationships
control
a separate
of
the existence
are also
Controls
be
level
65;
age
teens.18
that are Catholic,
and
restaurant
factors
a tourism
industry.19
also
may
For
conflicting.
for
the population
example,
variable
and
measuring
the
to an
In addition
a direct
have
Mormon,
on
effect
over
the population
65
indirect
fatalities,
to
tends
likely to be involved in fatal accidents.
descriptive
All
2000).
levy
taxes
estimation
on
statistics
states
and/or
employ
markups
is performed
the
per
tax measures
unit
based
excise
on
in LIMDEP
used
as
instruments
taxes
on
beer,
wine,
prices
of
the wholesale
8.0.20
First
stage
for prices
and
spirits
squared
and
spirits.
In
and wine.
correlations
18
Ibid.
Keg registration and youth BAC from Wagenaar,
19
based on Quinn et al. (1980) and Bradley et al. (1990). Age data from U.S.
Religion variables are interpolations/extrapolations
Bureau of the Census. Tourism data from the U.S. Bureau of Economic Analysis (http://www.bea.doc.gov/bea/regional/gsp/).
20
See Greene (2002), section E8.5.12 for details.
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
695
Alcohol Prices and Traffic Fatalities
Table lb.
Instrumental Variables
Beer excise tax ($/gal)
Wine excise tax ($/gal)
Spirits excise tax ($/gal)
Spirits excise tax (%)
BTAX
WTAX
STAX
STAXPERC
SMARKUP
WTAXPER
WMARKUP
Means
the fitted
between
lower
and
and
for price
values
is consistent
for price
R2
actual
a larger
with
Maximum
8.57
1.55
22.6
0.0
113.0
0.0
0.0
52.95
31.78
66.4
56.0
35.0
84.2
= 869.
by state population; N
are weighted
and standard deviations
6.64
6.91
7.46
9.52
22.4 0.0
4.10
10.0
18.86
11.69
40.58
3.72
12.7
1.06
2.61
Spirits markup (%)
Wine excise tax (%)
Wine markup (%)
Minimum
Standard Deviation
Mean
Definition
Variable
are,
consumption
error
and
0.90
respectively,
of measurement
amount
The
0.98.
in that variable.21
4. Results
tests
Hausman
of
the
six
rates.
fatality
and/or
of
endogeneity
are displayed
prices
1989, 1993). The null hypothesis of exogeneity
(Davidson andMacKinnon
for five
error
for measurement
As
the
last
two
columns
indicate,
in Table
2a
is rejected at the 1% level
for measurement
correcting
error/
endogeneity has a profound impact on the estimated price effects. OLS estimates are positive for five
of
the six fatality
estimates
rates,
that
imply
the
However,
and
three
of
in
increases
IV estimates
imply
are
the estimates
alcohol
quite
statistically
are
prices
All
the opposite:
associated
positively
with
are
the estimates
six of
at face
Taken
significant.
value,
traffic
fatalities.
five
and
negative,
these
of
the six are significant at the 5% level.
The estimated magnitudes suggest substantial effects of prices on fatalities. A 10% increase in
alcohol prices is predicted to reduce total fatalities by 5.8%. The estimated effect is somewhat larger
for weekend night fatalities (6.9%), and smaller for other times (3.9%). The estimated impact on all
youth fatalities (9%) is larger than for the total population. Less plausibly, the estimated impact on
weekend night fatalities among youth (3.5%) is smaller than the impact on youth at other times
is not significant at the 0.05 level.
(9.3%), although the difference
The
results
alcohol
using
in place
consumption
of
price
are
similar.
broadly
is
Exogeneity
rejected at the 10% significance level or less for five of the six fatality rates, and instrumental variable
estimates
per
is
11.3%.
estimated
than do OLS
effects
larger
alcohol
capita
estimate
the
indicate
is associated
consumption
The
effects
other
are
estimated
smaller
on
effects
weekend
particularly for youth, although the difference
The
regressions
also
provide
evidence
For
estimates.
on
with
a 9.9%
range
from
night
increase
10.2%
fatalities
a 10%
OLS,
using
example,
in fatalities,
to
14.1%.
than
on
Somewhat
fatalities
in
increase
whereas
the
IV
implausibly,
at
other
times,
is again not statistically significant.
a number
of other
determinants
of
fatalities.
Among
the total population (Table 3), income, vehicle miles traveled, and the population 18-29 years are
positively and significantly related to fatalities inmost specifications. Seat belt laws reduce fatalities
4-6%. The evidence is less strong for the legal environment index, which is negative for five of six
fatality rates, but not always statistically significant.22 The percentage of the population
21
22
living in dry
from the first-stage regressions are displayed in Table Al.
Entering the components of the index as separate variables results in about half negative and half positive coefficients, many
of which are not significant. It appears that the data are not sufficiently strong to distinguish among the effects of the large
number of legal initiatives that states have implemented.
Tax coefficients
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
696
J. Young
Douglas
and
Agnieszka
Table 2a. Tests for Endogeneity
Bielinska-Kwapisz
F-value
Rate
Fatality
Weekend
, Price Coefficient
r
_I_I_
All ages
28.4
0.16
All times (0.00) 2.2 3.4
All ages
0.11
17.3
1.1
3.0
(0.00)
All ages
12.6
0.29
3.5
2.1
(0.00)
16-20
0.08
15.3
All times(0.00)
0.7
3.1
Other times
Ages
0.4
16-20
Ages
16-20
Ages
-0.58
-0.69
-0.39
-0.90
-0.10
0.5
(0.52)
12.7
nights
|?-ratio I
IV
OLS
Level)
(Significance
nights
Weekend
Error in Prices
and/or Measurement
-0.35
0.8
0.44
-0.93
Other
times_(0.00)_^6_22_
no
has
counties
dry
on
effect
on
conditional
fatalities,
is positively
counties
to fatalities,
related
the price
of alcohol.
on
conditional
alcohol
However,
percentage
consumption.
This
in
living
is consistent
with the hypothesis that dry counties have a mixed relationship with fatalities: (i)more people living
in dry counties is associated with a lower demand for alcohol, and on that account fewer fatalities; (ii)
at the same time, people who do drink may be more
account
in more
result
There
little
is
(Baughman
that
evidence
0.08
The
or
laws
Mormon
percentage
likely to drive to obtain alcohol, and on that
et al. 2001).
server
is
not
Southern
and
above,
Baptists
preach
because
apparently
abstinence.
fatalities.
are
fatalities
and
positively
The latter relationship is unexpected, because
are positively
Fatalities
are higher
rates
fatality
in reducing
effective
whereas
significant,
significantly related to Southern Baptist membership.
many
are
training
related to fatalities, significantly so in two of three
in the Catholic church is negatively
Membership
equations.
fatalities
for
this
to the population
related
is not
Tourism
group.
65
aged
significantly
to fatalities.
related
Results for teenagers (age 16-20) are broadly similar (Table 4). Increasing the drinking age by
one
fatalities.
night
but
to reduce
is estimated
year
Table
5
is some
There
on weekend
not
A
nights.
assesses
the
teen
fatalities
evidence
youth
1-3%,
by
that keg
BAC
robustness
of
with
the
is associated
registration
law
is not
the
results
the
varying
on weekend
impact
with
teen
lower
fatalities,
to fatalities.
related
significantly
by
estimated
largest
estimation
and/or
technique
specification. These results use price as the right-hand side variable and total fatalities (all times) as the
variable.
dependent
Columns
least
two-stage
concern
One
is that
logit
are
specifications
sometimes
sensitive
to
weighting.23
1 and 4 of Table 5 display results for all ages and for teens with no weights; that is, unweighted
squares
errors
Standard
estimates.
are
using
computed
correction
White's
for group
wise heteroskedasticity, assuming that variances differ among states (Greene 2002). In comparison with
the weighted estimates in Tables 3 and 4, the unweighted price coefficient is slightly larger for all ages
and slightly smaller for teens. Both remain statistically significant. Thus, there is little evidence that
has
weighting
With
estimates,
23
Most
2000
an
state
important
and
are
which
dummies
sometimes
on
the results.
included,
our
misleading
in weights is attributable
inWyoming
in 1990.
of the variation
to 450,000
year
influence
are equivalent
estimates
because
to differences
of
serial
in population,
to "difference-in-difference"
correlation
which
(Bertrand,
ranges from 34,000,000
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
Duflo,
and
in California
in
Alcohol Prices and Traffic Fatalities
Table
for Endogeneity
Tests
2b.
Error
Measurement
and/or
in Consumption
|?-ratio |
IV
OLS
Level)
(Significance
All ages
All times
All ages
Weekend nights
All ages
Other times
Ages 16-20
All times
Ages 16-20
Weekend nights
Ages 16-20
Other times
Coefficient
Consumption
F-value
Fatality Rate
697
10.7
1.13
0.99
14.4
16.8
(0.00)
3.1
1.08
0.99
12.4
(0.08)
9.5
1.11
0.91
12.1
11.7
12.9
(0.00)
12.6
0.97
1.29
(0.00)
9.2
8.9
1.4
1.02
0.86
(0.23)
5.0
4.3
13.8
0.85
(0.00)
5.3
1.41
6.3
2004). Columns 2 and 5 of Table 5 provide estimates corrected for first-order serial
correlation (AR1).24 There is little effect on themagnitudes of the estimated coefficients. T-ratios are
Mullainathan
but
smaller,
mostly
are
There
whose
in dry
are
of
=
were
on
based
unaffected.
are
variables
control
server
0.08,
trends
time
the
coefficients
BAC
counties,
which
variables,
religion
some
estimated
is largely
coefficients
insignificant
when
results
variables
(population
the three
and
significant
different
dropping
insignificant
dropped
of
substantially
with
experimented
and/or
the pattern
and
training,
two
omitted.
We
of "wrong"
consistently
observations
from
sign
We
also
each
state,
tourism).
and were also often insignificant or of unexpected sign. As columns 3 and 6 of Table 5 indicate, omitting
these
variables
with
fatalities
respect
the magnitude
reduces
markedly
to the price
of alcohol
the
of
to 0.3
falls
estimated
for adults
The
response.
price
and 0.2
for teens,
and
of
elasticity
the teen estimate
is not significantly different from zero. Most of the rest of the coefficients display smaller changes in
magnitude and statistical significance. An exception is keg registration, which becomes highly
significant
and
should
treated
be
to reduce
is estimated
with
fatalities
as they
caution,
on
to depend
appear
results
these
13%. Overall,
by
the exact
suggest
specification
that point
of
estimates
the regression.
5. Discussion
There
is strong
fatalities.
state
Using
and
and/or
in the ACCRA
endogeneity
price
data
for
(OLS) estimators toward a finding of little or no effect of prices on
alcohol, which biases conventional
traffic
error
of measurement
evidence
federal
tax rates
as
instrumental
there
variables,
is strong
in
evidence
most specifications that fatalities are in fact negatively and significantly related to the price of alcohol,
ceteris
Our
paribus.
that an
indicate
estimates
point
increase
in the beer
tax of 50 cents
per
six pack
of
reduce traffic fatalities by about 4.5%; this amounted to 1900 lives in the year 2000.
Qualitatively similar results hold for the relationship between alcohol consumption and fatalities:
beer would
Estimates
Are
estimates
based
on
the partial
of
instrumental
reduced-form
the effects
of
price
variable
estimates
on
are
techniques
of
the price-fatality
consumption
and
than
larger
those
consumption
OLS.
consistent
relationship
on
from
fatalities?
By
with
structural
definition,
the
elasticity of fatalities with respect to the price of alcohol, Efp say, is equal to the product of the price
consumption
24
Estimated
elasticity,
Ecp,
and
the consumption-fatality
elasticity,
Efc.
Our
basic
rho is 0.46 for all ages and 0.12 for teens.
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
point
estimates
thus
698
J. Young and Agnieszka Bielinska-Kwapisz
Douglas
Table 3. Fatality Regressions
for Total Population
Alcohol
All
Right-hand
Side Variable
In (alcohol price)
(All Ages)
Price
Alcohol
Weekend
Other
All
Times
Times
Nights
Times
-0.58
-0.69
-0.39
3.4
3.0
2.1
Ln (alcohol consumption)
1.08
1.13
0.045
0.040
0.022
VMTLIC
18-29 (%)
Population
0.014
0.5
2.3
0.023
0.030
3.7
-0.045
-0.004
Catholic
(%)
-0.009
Mormon
(%)
-0.010
7.0
5.3
-0.041
-0.043
2.8
3.1
3.1
0.4
0.005
0.6
3.9
4.0
0.010
0.008
0.008
0.7
0.9
0.6
-0.009
-0.012
1.9
0.7
0.7
4.1
0.8
0.5
0.045
0.055
7.9
7.0
6.0
Population 65+
0.024
0.026
0.024
Adjusted R2
1.4
-0.018
0.5
1.6
-0.021
Southern Baptist (%)
(%)
0.3
0.006
5.1
0.9
2.1
1.8
-0.003
-0.013
0.010
1.5
0.038
1.7
1.9
-0.028
-0.021
-0.002
0.013
1.5
-0.003
3.8
-0.017
2.1
0.2
0.034
0.4
3.8
-0.008
-0.002
0.011
4.8
-0.046
0.009
1.8
0.005
2.8
-0.055
4.3
0.001
-0.011
2.2
training
Tourism
0.010
2.6
0.027
(%)
0.003
0.029
0.5
Server
5.9
0.009
5.0
0.001
0.08
0.022
0.055
0.8
BAC -
2.2
2.5
-0.004
in dry counties (%)
Population
0.010
5.6
0.037
3.1
Legal
0.017
5.4
-0.038
environment
1.11
11.7
9.5
0.051
7.2
Seat belt law
0.045
10.6
0.022
7.7
11.7
Other
Times
Nights
14.4
Income
Consumption
Weekend
-0.015
1.0
1.0
0.7
0.94
0.91
0.92
0.93
0.94
0.93
State and year dummies are included in every equation. Estimation is by 2SLS using combined state plus federal excise
taxes on beer, spirits, and wine, and for control states, the percentage excise taxes and/or markups on spirits and wine, as
=
applicable. Price or consumption is treated as endogenous. iV 869. Dependent variable: logit of fatality rate (absolute values of
f-statistics are below parameter estimates).
imply
the price-consumption
is
elasticity
?
Ecp
=
EfJEfp
-0.58/1.13
This
=-0.51.
is within
value
the
range of price elasticities for aggregate alcohol consumption estimated in previous work (Leung and
2003).
Phelps 1993; Young and Bielinska-Kwapisz
What do these estimates imply about the impact of alcohol taxes on traffic fatalities? The answer
depends on the degree towhich alcohol taxes are shifted forward to retail prices and on how important
taxes are as a share of retail prices. Young and Bielinska-Kwapisz
(2002) found that spirits, beer, and
wine
However,
taxes
are
excise
overshifted:
taxes
are only
Retail
prices
11-18%
rise
of retail
more
than
prices.
Thus,
one-for-one
the
1991
with
change
an
increase
in federal
in
excise
taxes.
taxes,
which doubled the beer tax from 16 cents per six pack of beer to 32 cents and increased the wine tax
by 500%, increased retail prices by only about 6%. Based on a price-fatality elasticity of 0.58, the
predicted decline in total fatalities is about 3.5%.
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
Alcohol Prices and Traffic Fatalities
Table 4. Fatality Regressions
for Teens
(Ages 16-20)
Alcohol
Side Variable
Right-hand
All Times Weekend
In (alcohol price)
Price
Nights
Alcohol
Other Times
-0.89
-0.35
-0.93
3.1
0.8
2.2
1.29
VMTLIC
0.065
0.023
7.5
4.3
0.031
0.015
Seat belt law
-0.042
2.3
2.3
environment
Legal
0.3
Population
1.2
BAC = 0.08
Server
training
-0.006
Catholic
(%)
-0.008
Mormon
(%)
-0.023
1.3
6.9
4.9
-0.009
Adjusted R2
-0.011
-0.023
-0.028
-0.014
0.022
1.7
0.8
3.0
2.3
1.2
-0.057
-0.046
-0.055
-0.071
-0.057
1.7
3.2
-0.012
-0.008
1.0
1.2
-0.000
-0.013
2.6
1.8
0.8
0.4
0.027
0.005
-0.006
0.8
0.6
0.4
-0.003
-0.014
2.7
1.6
0.9
-0.040
-0.015
-0.046
0.3
0.005
2.4
0.5
-0.016
0.004
1.4
-0.037
2.1
0.003
1.2
0.041
-0.003
1.0
0.2
0.023
-0.019
0.9
0.8
1.9
0.058
-0.127
Youth BAC
2.3
-0.022
1.1
0.059
registration
0.8
0.021
2.3
Keg
0.008
2.5
1.7
0.4
Southern Baptist (%)
0.015
0.0
4.0
4.0
0.041
0.052
0.078
0.033
1.8
4.5
0.010
-0.005
6.3
1.6
2.1
-0.002
in dry counties (%)
0.016
0.041
4.5
-0.019
1.41
1.02
4.3
9.4
Drinking age
Other Times
Nights
8.9
0.055
5.0
Consumption
All Times Weekend
Ln (alcohol consumption)
Income
699
0.056
4.4
4.5
-0.034
-0.182
-0.039
0.8
4.8
1.6
-0.019
0.019
-0.091
0.5
0.013
0.006
2.5
0.026
-0.005
0.5
0.7
0.5
0.4
0.2
1.1
0.85
0.79
0.76
0.87
0.79
0.77
State and year dummies are included in every equation. Estimation is by 2SLS using combined state plus federal excise
taxes on beer, spirits, and wine, and for control states, the percentage excise taxes and/or markups on spirits and wine, as
= %69.
applicable. Price or consumption is treated as endogenous. N
Dependent variable: logit of fatality rate (absolute values of
?-statistics are below parameter estimates).
These
can
results
also
be
expressed
as a
tax-fatality
Ignoring
elasticity.
the
in the wine
increase
tax for simplicity, the implied elasticity of fatalities with respect to the beer tax is 0.06. This figure is
about
one
than Evans,
lower
quarter
Neville,
and Graham's
(1991)
estimate
of 0.08,
and
about
one
half of Ruhm's
(1996, table 2) estimate of 0.11. Chaloupka, Saffer, and Grossman (1993, p. 181)
estimated that doubling the federal beer tax would have reduced fatalities by 3.9%, similar to the
finding in this paper.
The estimated tax elasticity for teen fatalities is about half again as large (0.09), because the teen
price elasticity of fatalities is estimated to be that much larger (Tables 2 and 4). This value is
substantially
smaller
than
some
previous
estimates.
For
example,
Ruhm's
(1996,
table
4)
tax elasticity
for 18- to 20-year-olds is twice as high (0.17-0.21), and Chaloupka, Saffer, and Grossman (1993, p.
181) report that doubling the federal beer tax would reduce fatalities among 18- to 20-year-olds by
11.8%,
implying
an even
larger
tax elasticity
of 0.21.
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
700
J. Young and Agnieszka Bielinska-Kwapisz
Douglas
Table 5. Robustness Checks
All Ages
Teens
AR1
Side Variable
Right-hand
In (alcohol price)
-0.70
Income
VMTLIC
-0.77
environment
Legal
-0.84
-0.74
2.2
2.5
1.8
0.036
0.026
0.041
0.048
0.041
8.2
4.2
6.5
4.6
8.7
0.020
0.007
0.030
0.028
0.021
0.038
7.9
4.0
2.7
6.8
-0.029
-0.026
1.4
0.053
5.2
9.8
-0.031
-0.023
2.5
2.3
2.0
-0.008
-0.008
-0.004
1.3
0.001
BAC = 0.08
Catholic
Mormon
(%)
Drinking
age
1.7
1.0
1.0
-0.004
0.7
0.027
1.1
1.6
0.9
0.008
0.9
0.0
0.3
0.005
0.010
0.006
2.0
0.0
2.4
-0.008
-0.013
2.0
1.0
-0.010
0.7
0.7
0.7
0.058
0.056
0.064
6.9
5.4
6.4
5.1
0.027
0.029
0.034
3.2
1.5
0.018
-0.018
0.7
-0.002
0.82
0.93
1.6
-0.034
-0.130
0.9
5.5
-0.003
0.9
0.8
0.2
0.78
0.75
0.86
-0.018
0.92
-0.013
-0.020
1.9
Youth BAC
-0.001
0.1
-0.059
registration
Adjusted R2
-0.013
0.041
0.2
Dependent
1.3
-0.005
0.000
(%)
variable:
-0.008
0.8
1.4
Keg
1.5
-0.010
0.038
2.4
(%)
-0.013
0.2
-0.017
0.044
1.3
0.022
1.0
Tourism
0.8
0.013
-0.023
Southern Baptist (%)
Population 65+
0.9
0.3
-0.005
(%)
-0.017
0.001
1.0
training
-0.19
0.062
8.8
1.8
Server
12.7
-0.030
in dry counties (%)
Population
-0.31
Omit Some
Xs
Corrected
2.8
5.2
Seat belt law
AR1
Unweighted
3.8
0.054
18-29 (%)
Population
Omit
Some Xs
Corrected
Unweighted
logit of fatality rate (absolute values of ?-statistics below parameter estimates).
these more modest
in this study should be regarded with caution,
however. One reason is that they still seem too large. Only a minority of fatalities involve alcohol;
currently the proportion is 30-40%. Thus, a 1% increase in per capita alcohol consumption could
fatalities increase by more than twice that
increase fatalities by 1.13% only if alcohol-involved
Even
amount. Second,
measures
is
responsive
nights
25
Dee
are
the pattern of estimated price and consumption effects across the different fatality
Teen
counterintuitive.
sometimes
to the price
of
apparently
less
(1999) and Dee
estimates obtained
and Evans
alcohol
than
responsive
are
to
fatalities
fatalities
alcohol
on
at other
consumption
weekend
times,
nights
and
than
total
fatalities
(2001) obtain similar results.
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
are
less
apparently
fatalities
at
on weekend
other
times.25
701
Alcohol Prices and Traffic Fatalities
In part,
Moore
these
are
variables
are
or be
a number
certain
ing
citizen
rent
effects.
is
of
on
point
infrequent
in state
changes
of price
modest
are
effects
obtained
effects
for DUI,
excise
taxes'"
to what
sensitive
some
when
of
approach
as
the resulting
taxes,
alcohol
and
(Cook
other
the
control
insignificant
effects
here,
estimators
by
enforcement
of
is not
confined
to this
estimates
a "reduced
are biased
and
inconsistent
study. Whether
form"
if taxes
took
by
are
stricter
legislating
and mobiliz
efforts,
to overstate
likely
and
fatalities,
simultaneously
taxes,
educational
instead
or
states
increasing
and
are
taxes
that
attitudes
underlying
to curb
intended
to the extent
abuse?say
up
stepping
policy"
is done
measures
policy
reflect
may
policies
In particular,
estimated
"endogenous
other
and
unmeasured
with
to reduce
the
of
taxes
alcohol
variables.
fronts
penalties
problem
a "structural"
that
correlated
groups?then
The
and
estimates
more
much
instrumental
improper
and more
"small
excluded.
concern
on
directly
with
included,
related
be
action
In addition,
alcohol,
A
toward
reflect
may
p. 421).
variables
thus
results
2001,
their
deter
actual
a researcher
takes
regressing
fatalities
endogenous.
Indeed,
all of the most frequently cited estimates of the impact of alcohol taxes rely on the assumption that
taxes
are exogenous.
all of
Thus,
these
studies
may
be biased,
and
a more
accurate
assessment
will
not
be possible until the determinants of policy are more fully understood and estimation procedures
are modified accordingly. This study has substantially resolved the discrepancy between estimates
on
based
propriate
tax
and
price
as exogenous
Tax Coefficients
Table Al.
but
data,
it remains
to be
seen
whether
the
tax
data
Dependent
Real
Real
Side Variable
beer
Variable
wine
Consumption
tax
excise
tax
excise
-0.01
spirits
excise
tax
Spirits excise tax (%)
(%)
2.98
0.0027
excise tax (%)
0.0053
-5.6
4.5
0.001
-0.0004
-0.0007
0.7
Wine markup (%)
Adjusted R2
1.75
-0.005
1.6
Wine
0.0041
0.0014
0.6
Spirits markup
Log of Ethanol
Price
3.4
1.1
Real
ap
from First-stage Regressions
Log of Ethanol
Right-hand
are
themselves
variables.
-0.0008
0.0016
1.2
0.00064
1.9
0.0034
2.7
-0.0002
2.3
0.45
0.98
0.89
Absolute values of ?-statistics below parameter estimates. Each equation also includes all of the right-hand
variables from the structural fatality equations (see Table 3).
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
side exogenous
702
Douglas
J. Young and Agnieszka Bielinska-Kwapisz
References
American
Chamber of Commerce Researchers Association
1982-2000. ACCRA cost of living index, quarterly
(ACCRA).
reports. Louisville, KY: ACCRA.
and John Pepper. 2001. Slippery when wet: The effects of local
Baughman, Reagan, Michael Conlin, Stacy Dickert-Conlin,
alcohol access laws on highway safety. Journal of Health Economics 20:1089-196.
Beard, T. Randolph,
application
Paula A. Gant, and Richard P. Saba. 1997. Border-crossing
to alcohol. Southern Economic Journal 641:293-306.
Bertrand, Marianne, Esther Duflo,
estimates? Quarterly Journal
and Sendhil Mullainathan.
2004. How
sales, tax avoidance,
much
should we
and state tax policies: An
trust difference-in-differences
119(1):249?75.
of Economics
Besley, Timothy J., and Harvey S. Rosen. 1999. Sales taxes and prices: An empirical analysis. National Tax Journal June: 157-78.
1990. Churches and church membership
Bradley, Martin B., Norman M. Green, Jr., Dale E. Jones, Mac Lynn, and Lou McNeil.
in the United States 1990. Atlanta: Glemnary Research Center.
Brown, Robert, R. Todd Jewell, and Jerell Richer.
Journal
1996. Endogenous
alcohol prohibition
and drunk driving. Southern Economic
62:1043-53.
1993. Alcohol control policies and motor-vehicle
fatalities. Journal of Legal
Chaloupka, F. J., H. Saffer, and M. Grossman.
Studies 22(1): 161-86.
Cook, Phillip J. 1981. The effect of liquor taxes on drinking, cirrhosis, and auto fatalities. InAlcohol and public policy: Beyond the
shadow of prohibition, edited by M. H. Moore and D. R. Gerstein. Washington, DC: National Academy Press, pp. 255-85.
J. Moore. 2001. Environment and persistence
in youthful drinking patterns. In Risky behavior
Cook, Philip J., and Michael
among youths, edited by Jonathon Gruber. Chicago: University of Chicago Press, pp. 375-437.
1989. Testing for consistency using artificial regressions.
Russell, and James G. MacKinnon.
5:363-84.
Davidson,
and James G. MacKinnon.
Davidson, Russell,
Press.
1993. Estimation
and inference
in econometrics.
Econometric
Theory
New York: Oxford University
S. 1999. State alcohol policies, teen drinking, and traffic fatalities. Journal of Public Economics
72:289-315.
S., and William N. Evans. 2001. Teens and traffic safety. In Risky behavior among youths, edited by Jonathon
Gruber. Chicago: University of Chicago Press, pp. 121-65.
Dee, Thomas
Dee, Thomas
DC: DISCUS.
Spirits Council of the U.S. (DISCUS). 2000. History of beverage alcohol tax changes. Washington,
Daniel. 2003. Evaluating
the effectiveness
of policies related to drunk driving. Journal of Policy Analysis and
22:249-74.
Management
Distilled
Eisenberg,
Evans, W. N., D. Neville, and J. Graham.
Risk Analysis
11:279-89.
Greene, William
H.
Software,
Greene, William
Inc.
LIMDEP
version
deterrence
8.0 econometric
of drunk driving: Evaluation
modeling
guide. Volume
of recent American
2. Plainview,
NY:
policies.
Econometric
H. 2003. Econometric
Grossman, Michael,
Monitoring
2002.
1991. General
analysis. 5th edition. Upper Saddle River, NJ: Prentice-Hall.
Frank J. Chaloupka, and Ismail Sirtalan. 1998. An empirical analysis of alcohol addiction: Results
the Future panels. Economic Inquiry 36:39-48.
from the
2002. Get high and get stupid: The effect of alcohol and marijuana
Grossman, Michael, Robert Kaestner, and Sara Markowitz.
use on teen sexual behavior. National Bureau of Economic Research Working Paper No. 9216.
of
Gruber, Jonathon. 2001. Introduction. In Risky behavior among youths, edited by Jonathon Gruber. Chicago: University
Chicago Press, pp. 1-28.
Gruenewald, P. J.,W. R. Ponicki, and H. D. Holder. 1993. The relationship of outlet densities to alcohol consumption:
series cross-sectional
17:38-47.
analysis. Alcoholism: Clinical and Experimental Research
Kenkel, Donald S. 1993. Drinking, driving and deterrence: The effectiveness
Law and Economics
36:861-76.
and social costs of alternative policies.
A time
Journal
of
Jeffrey D., and John R. Moran. 2003. Can policy changes be treated as natural experiments? Evidence from state excise
taxes. Working Paper, Department of Economics,
Syracuse University.
Leung, S. F., and C. E. Phelps. 1993. "My kingdom for a drink ... ?": A review of estimates of the price sensitivity of demand
for alcoholic beverages. In Economics and the prevention of alcohol-related
problems: Proceedings
of a workshop on
Kubik,
economic and socioeconomic
issues in the prevention of alcohol-related problems, October 10-11,1991,
edited by M. E.
and G. Bloss. Bethesda, MD: NIAAA Research Monograph No. 25:1-31.
1995. The demand for alcohol: The differential response to
Manning, Willard G., Linda Blumberg, and Lawrence H. Moulton.
14:123-48.
price. Journal of Health Economics
Hilton
Sara. 2000. The price of alcohol, wife abuse, and husband abuse. Southern Economic
S., P. Chatterji, R. Kaestner, and D. Dave. 2002. Substance use and suicidal behaviors
Working Paper No. 8810.
Markowitz,
Journal
Markowitz,
among young adults. NBER
Mast,
Brent D., Bruce L. Benson,
Economic Journal 66:214-49.
and David W.
Rasmussen.
1999. Beer
taxation and alcohol-related
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
67:279-303.
fatalities.
Southern
Alcohol Prices and Traffic Fatalities
J. P.
Nelson,
1997. Economic
Economics
and demographic
in U.S.
alcohol
demand: A growth
accounting
analysis.
Empirical
22:83-102.
J. P. 2003. Advertising bans, monopoly,
22:1-25.
of Industrial Organization
Nelson,
factors
703
and alcohol demand: Testing
for substitution
effects using
state panel data.
Review
T. M., G. D. Williams, H. Yi, A. K. Hoy, F. S. Stinson, and M. C. Dufour. 2002. NHTSA surveillance report #59:
state, and regional trends, 1970-99. Rockville, MD: National
Apparent per capita alcohol consumption: National,
Institute on Alcohol Abuse and Alcoholism, Division of Biometry and Epidemiology.
Quinn, Bernard, Herman Anderson, Martin Bradley, Paul Goetting, and Peggy Shriver. 1980. Churches and church membership
in the United States 1980. Atlanta: Glenmary Research Center.
Nephew,
15:435-54.
Ruhm, C. 1996. Alcohol policies and highway vehicle fatalities. Journal of Health Economics
1994. Effects of prices, civil and criminal sanctions, and law enforcement
Sloan, F. A., B. A. Reilly, and C. M. Schenzler.
alcohol-related mortality. Journal of Studies on Alcohol July: 454-65.
on
of the Census. Tourism data from US Bureau of Economic Analysis. Available http://www.bea.doc.gov/bea/
regional/gsp/. Accessed 24 October 2005.
School of Public Health.
Wagenaar, Alexander C. Alcohol Epidemiology
Program, University of Minnesota,
2002. Alcohol taxes and beverage prices. National Tax Journal LV:57-73.
Young, D. J., and A. Bielinska-Kwapisz.
U.S.
Bureau
Young,
D. J., and A. Bielinska-Kwapisz.
Studies on Alcohol 64:235-238.
Young,
D. J., and T. W.
20:107-26.
Likens.
2003. Alcohol
2000. Alcohol
consumption,
regulation
beverage
and auto fatalities.
prices,
and measurement
International
Review
This content downloaded from 153.90.170.56 on Tue, 10 Jun 2014 13:52:42 PM
All use subject to JSTOR Terms and Conditions
error. Journal
of Law and Economics
of