A Er 04 Addiction
A Er 04 Addiction
A Er 04 Addiction
r
s
, 0, R). Intuitively, u
s
a
represents the baseline
payoff associated with successful abstention in
state s and activity a, and b
s
a
represents the
marginal instantaneous benet from use the in-
dividual receives in state s after taking activity
a. Thus, u
s
a
b
s
a
is the payoff for usage. Let
p
s
(p
s
E
, p
s
A
, p
s
R
), u
s
(u
s
E
, u
s
A
, u
s
R
), b
s
(b
s
E
,
b
s
A
),
s
(p
s
, u
s
, b
s
), and (
0
, ... ,
S
). The
vector species all pertinent derivative pa-
rameters. It reects the properties of the sub-
stance, the method of administration, the
characteristics of the individual user, and the
public policy environment. We make the fol-
lowing assumption (the latter part of which is in
keeping with our earlier discussion).
ASSUMPTION 2: The payoff function w
s
is
increasing, unbounded, strictly concave, and
twice differentiable with bounded second deriv-
ative in the variable e (consumption of the non-
addictive good). Moreover, u
s
E
u
s
A
u
s
R
, and
u
s
E
b
s
E
u
s
A
b
s
A
.
For each state s, the DM follows one of ve
contingent plans: engage in activity E and then
21
Though we assume that the DM can commit to reha-
bilitation only one period at a time, this is without loss of
generality since he starts each period in the cold mode. In
practice, rehabilitation programs may also teach self-
management skills and desensitize addicts to cues. One can
model these possibilities by assuming that p
s
a
(for a given
state or states) declines subsequent to rehabilitation or ther-
apy. Since the evidence suggests that these treatments are
not completely effective (Goldstein, 2001, p. 188), the forces
described here would still come into play after treatment.
22
When w
s
depends on and p
s
A
p
s
R
, rehabilitation can
serve as a strategy for avoiding cues that trigger reductions
in hedonic payoffs (through cravings).
1566 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
use the substance when in the cold mode [(a,
x) (E, 1)], engage in E and refrain from use
when in the cold mode [(a, x) (E, 0), hence-
forth half-hearted abstention], engage in A
and use when in the cold mode [(a, x) ( A, 1)],
engage in A and refrain from use when in the
cold mode [(a, x) ( A, 0), henceforth con-
certed abstention], or enter rehabilitation [(a,
x) (R, 0)]. From Assumption 2, it follows that
(E, 1) always dominates ( A, 1), so there are in
practice only four pertinent choices.
The cold-mode DM is sophisticated in the
sense that he correctly anticipates his future
choices in either decision mode, and he under-
stands the process triggering the hot mode. Ac-
cordingly, his choices in the cold mode
correspond to the solution of a simple dynamic
stochastic programming problem with a value
function V
s
() (evaluated as of the beginning of
a period) satisfying
(1) V
s
max
a,xE,1,E,0,A,0,R,0
u
s
a
s
a,x
b
s
a
1
s
a,x
V
max1,s 1
s
a,x
V
minS,s 1
for s 1,
23
where
s
a, x
represents the probabil-
ity of consuming the substance in state s with
contingent plan (a, x) (so
s
E,1
1,
s
E,0
p
s
E
,
s
A,0
p
s
A
, and
s
R,0
0). Existence, unique-
ness, and continuity of V
s
() in follow from
standard arguments.
We close this section with several remarks.
First, though simple and stylized, our model
adheres closely to the three key premises de-
scribed in Section II. Specically, use among
addicts is potentially a mistake; experience with
an addictive substance sensitizes the user to
environmental cues that subsequently trigger
mistaken use; and the awareness of this possi-
bility leads users to manage their susceptibilities.
Second, our model reduces to the standard
rational addiction framework when p
s
a
0 for
all s and a. Thus, the novelty of our approach
involves the introduction of stochastic shocks
(occurring with probability p
s
a
0) that poten-
tially cause decisions to diverge from prefer-
ences. This possibility is a central feature of our
model since, without it, the DM would never
choose to avoid cues or enter rehabilitation
(with p
s
E
0, (E, 0) dominates both ( A, 0) and
(R, 0)). For the same reason, a naive DM who
incorrectly believes he does not suffer from
a self-control problem (that is, who acts as if
p
s
a
0) will never choose cue-avoidance or
rehabilitation.
Third, even though our model allows for the
possibility that choices and preferences may
diverge, with careful use of appropriate data it
should still be possible to recover preferences
and other critical parameters (such as hot-mode
probabilities) empirically. Since we assume that
preferences and choices are sometimes aligned,
the most obvious approach involves the selec-
tive application of the revealed preference prin-
ciple. The empirical challenge is to identify
instances of alignment. One cannot make this
determination using only information on choices.
We contend, however, that other evidence, such
as the research results summarized in Section II,
justies treating the central assumptions of our
model as maintained hypotheses. This means
that we can use choice data involving precom-
mitments and cue-avoidance to infer hot-mode
probabilities and the utility costs of unintended
use (recall the discussion in the preceding para-
graph). Furthermore, measures of physiologi-
cal arousal and/or self-reported affective states
could be used to differentiate cold choices
from hot choices in experimental settings. For
a more general discussion of preference mea-
surement when choices and preferences system-
atically diverge, see Bernheim and Rangel
(2005).
Fourth, unlike other economic theories of ad-
diction, ours does not necessarily assume that
present use increases the marginal benet of
future use (b
s1
b
s
). We show that, contrary
to some claims in the literature, it is possible to
explain the central features of addiction without
invoking intertemporal preference complemen-
tarities (provided the probability of cue-
triggered mistakes increases with s). This is
important because intertemporal complementa-
rities do not appear to drive some distinctive
addictive behaviors,
24
and these behaviors are
23
The associated expression for s 0 is virtually iden-
tical, except that V
0
() replaces V
max{1,s1}
().
24
The phenomenon of withdrawal is often interpreted
as the key manifestation of intertemporal comple-
1567 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
observed in contexts where such complementa-
rities are probably not present (e.g., compulsive
shopping and kleptomania).
Fifth, though one could incorporate the re-
alistic possibility that some individuals are
partially myopic with respect to the likelihood
and effects of becoming addicted, we assume
that the DM is sophisticated in the cold mode.
If, as we argue, counterproductive addictive
behaviors can arise even with sophisticated
decision-makers, efforts to eradicate addic-
tion solely through education and information are
misguided.
IV. Positive Analysis
A. Comparative Dynamics
Our comparative dynamic results concern the
intensity with which the DM voluntarily uses
the addictive substance. We study two notions
of intensity. We say that the disposition to use is
greatest for (E, 1), followed in order by (E, 0),
( A, 0), and (R, 0). Thus, for example, the dis-
position to use increases when the DMs choice
shifts from ( A, 0) to (E, 1). We judge the
intensity of intentional use by asking whether
the DM plans to consume the substance. Thus,
intentional use is highest for (E, 1), and equiv-
alent for (E, 0), ( A, 0), and (R, 0).
25
An increase
in intentional use implies an increase in the
disposition to use, but not vice versa. Both
denitions permit us to compare the intensity of
voluntary use both within states and across
states.
We study comparative dynamics with re-
spect to the elements of the parameter vector
. Since some of these are simple functions of
prices and income (q, r
s
, and y
s
), comparative
dynamics with respect to the latter variables
follow immediately. We are particularly in-
terested in the effects of the parameters p
s
E
and p
s
A
, since these are directly tied to the
novel aspects of our framework (stochastic
events that create pathological discrepancies
between preferences and choice). We are also
interested in the effects of u
s
A
, b
s
A
, and u
s
R
,
since these parameters are relevant only if the
novel components of our model are opera-
tional ( p
s
E
0).
1. Changes in Individual Parameters.In
practice, we are rarely interested in phenomena
that affect only one state-specic parameter.
However, examining these effects in isolation
lays the groundwork for subsequent results in-
volving changes in groups of parameters.
PROPOSITION 1: (i) The disposition to use in
state j is:
(i-a) weakly increasing in b
k
a
and u
k
a
, and
weakly decreasing in p
k
a
, for k j;
(i-b) weakly decreasing in b
k
a
and u
k
a
, and
weakly increasing in p
k
a
, for k j;
(i-c) weakly decreasing in p
j
E
and u
j
R
and
weakly increasing in b
j
E
.
(ii) Intentional use in state j is invariant with
respect to p
j
E
, p
j
A
, u
j
A
, b
j
A
, and u
j
R
.
Parts (i-a) and (i-b) establish the intuitive
property that benecial changes in parameters
for more (less) advanced states of addiction
increase (decrease) the disposition to use in the
current state. Thus, an increase in the likelihood
or severity of a cue-triggered mistake in state s
induces the DM to make choices that reduce the
likelihood of reaching state s. Part (i-c) is also
intuitive: the disposition to use in the current
state rises with the benets to current use and
falls with both the desirability of rehabilitation
and the likelihood that the exposure activity
triggers the hot mode (since this increases the
attractiveness of concerted abstention and reha-
bilitation relative to half-hearted abstention).
Part (ii) is perhaps less transparent. A change in
parameters can affect intentional use only if it
mentarities. Notably, W. E. McAuliffe (1982) showed
that only 27.5 percent of heroin addicts experienced
cue-triggered withdrawal symptoms, and only 5 percent
of these felt these symptoms were responsible for recid-
ivism.
25
For some parameter values, the DM may be indif-
ferent between two (but never more than two) choices in
any particular addictive state. When this occurs, the set of
optimal choices is always {(E, 1), (E, 0)}, {(E, 0), ( A,
0)}, {(E, 0), (R, 0)}, or {( A, 0), (R, 0)}. We say that
a change in parameters from to weakly increases the
disposition to use (intentional use) if it leads to a weak
increase in both the minimum and maximum disposition
to use (intentional use) among optimal choices, and
strictly increases the disposition to use (intentional use) if
either the minimum or the maximum strictly increases
and neither declines.
1568 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
tips the balance between (E, 1) and (E, 0).
26
Clearly, this comparison does not implicate p
j
A
, u
j
A
,
b
j
A
, or u
j
R
; neither does it depend on p
j
E
.
27
The characterization of directional effects in
Proposition 1 is not quite complete. This is
because the effects of b
j
A
, u
j
A
, and u
j
E
on the
disposition to use in state j can be positive or
negative, depending on the parameter values.
2. Changes in Groups of Parameters.To
examine the effects of policy and environmental
changes, and to make comparisons between op-
timal decision rules for different substances, we
must typically consider the effects of varying
many parameters simultaneously. For example,
a general reduction in the cost of rehabilitation
(due perhaps to the development of a new thera-
peutic drug) raises u
s
R
for all s. Likewise, when
one substance is more addictive than another, it is
natural to assume that p
s
a
is higher at every state s.
For any state-indexed variable z
s
, we say that
a change from (z
s
)
s0
S
to ( z
s
)
s0
S
represents a
general increase (decrease) if z
s
z
s
( z
s
z
s
)
for all s, with strict inequality for some s. Prop-
osition 1 suggests that compound parameter
changes of this type often have ambiguous ef-
fects on use. For example, a general increase in
u
s
R
or a general decrease in p
s
a
can reduce the
disposition to use in state j by making lower
states (weakly) more attractive but can also
increase this disposition by making higher
states (weakly) more attractive.
It is nevertheless possible to reach a number
of conclusions without imposing additional
structure. Public policy discussions often em-
phasize initial use, choices among casual users
who are at risk of becoming addicted, and pat-
terns of behavior among hard-core addicts. To
shed light on initial use, we study behavior in
state 0. To shed light on the choices of casual
users, we examine behavior in state 1, the length
of the rst intentional use interval (dened as
{1, ... , s
1
1} where s
1
is the largest integer
such that (E, 1) is chosen for all s {1, ... , s
1
1, ... , S} where s
3
is the smallest integer such
that (E, 1) is chosen for all s {s
3
1, ... , S}
but not for s
3
). While these aspects of behavior
respond ambiguously to general changes in some
parameters, other effects are unambiguous.
29
PROPOSITION 2: (i) A general increase in p
s
E
or p
s
A
, or a general reduction in u
s
A
or b
s
A
, weakly
decreases the disposition to use in state 0 (and
state 1 for p
s
E
), weakly shortens the rst inten-
tional use interval, weakly lengthens the initial
resistance interval, and weakly lengthens the
nal resignation interval.
(ii) A general increase in u
s
R
weakly increases
the disposition to use in all states (including
state 0) up to (but not including) the rst state in
which rehabilitation is an optimal choice after
the increase. It also weakly lengthens the rst
intentional use interval, weakly reduces the dis-
position to use in state S, and weakly shortens
the nal resignation interval.
(iii) A general increase in u
s
E
or b
s
E
weakly
shortens the initial resistance interval. In addi-
tion, a general increase in b
s
E
weakly increases
the disposition to use in states 0 and 1.
How do patterns of use compare for two sub-
stances that are the same in all respects, except
26
If (E, 0) yields a higher expected discounted payoff
than (E, 1), then (E, 1) is obviously not the DMs best
choice. Conversely, if (E, 1) yields a weakly higher ex-
pected discounted payoff than (E, 0), then (E, 1) is neces-
sarily preferred to both ( A, 0) and (R, 0). To understand
why, note that (a) u
s
E
b
s
E
V
min{S,s1}
() u
s
E
V
max{1,s1}
() u
s
a
V
max{1,s1}
() for a A, R
[where the rst inequality follows because the DM weakly
prefers (E, 1) to (E, 0), and the second inequality follows
from Assumption 2], and (b) u
s
E
b
s
E
V
min{S,s1}
()
u
s
A
b
s
A
V
min{S,s1}
() (by Assumption 2). For (R, 0),
the desired conclusion follows from (a); for ( A, 0), it fol-
lows from (a) and (b).
27
The DM prefers (E, 1) to (E, 0) if and only if he
prefers it to E with the certainty of abstention. The proba-
bility p
j
E
does not enter this comparison.
28
At least one of these intervals is always empty. The
length of the initial resistance interval is relevant only if pa-
rameters change after the DM starts using the substance (oth-
erwise he would never advance beyond state 1). In that case, it
sheds light on the DMs ability to achieve permanent recovery.
29
To allow for multiple optima, we say that a parameter
change weakly shortens (lengthens) an interval if it weakly
reduces (increases) both the minimum and maximum length
of the interval.
1569 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
that one is more addictive than the other (higher
values of p
s
E
and p
s
A
for all s)? Part (i) of the
proposition provides a partial answer. Not surpris-
ingly, an increase in addictiveness discourages use
among new users (reducing the disposition to use
in state 0, shortening the rst intentional use in-
terval, and lengthening the initial resistance inter-
val). Strikingly, it always has the opposite effect
on hard-core addicts, producing longer resignation
intervals. One might think that an increase in
addictiveness might discourage a relatively ad-
vanced user from taking actions likely to place
him in an even more highly addicted state, but this
effect never materializes in the nal resignation
interval. Instead, the DM is inuenced by the
increased futility of resisting use at lower states.
He resigns himself to severe addiction because he
recognizes his powerlessness to control his subse-
quent behavior adequately at lower states, despite
intentions to abstain. According to part (i), general
changes in the parameters governing payoffs from
the avoidance activity (u
s
A
and b
s
A
) have similar
effects.
How does an improvement in rehabilitation
technology (higher values of u
s
R
for all s) affect
patterns of use? According to part (ii) of the
proposition, use among those with low cumula-
tive exposure increases in a strong sense (the
disposition to use rising in all states up to the
point where the DM enters rehabilitation). Since
rehabilitation cushions the negative effects of
addiction, this is not surprising. As in part (i),
this development has the opposite effect on
hard-core addicts, shortening the resignation in-
terval. Notably, increasing u
s
R
only for states in
the resignation interval would have no effect on
behavior. Thus, for a general increase in u
s
R
, the
DM turns away from intentional use in the
resignation interval because rehabilitation be-
comes a more attractive option in lower states.
Part (ii) of Proposition 2 also implies that an
improvement in rehabilitation technology can
have the perverse effect of shifting the entire
population distribution to more addicted states.
Provided that all members of the population
start out at s 0, this occurs when a general
increase in u
s
R
raises the lowest state at which
the DM selects rehabilitation.
30
Part (iii) of the proposition concerns u
s
E
and
b
s
E
. These parameters do not relate to the novel
features of our model, but we have included
their effects for completeness.
Proposition 2 underscores the fact that
changes in the environment have complex ef-
fects on use, often driving consumption among
new users and hard-core addicts in opposite
directions. It is natural to wonder whether there
are any general parameter changes that always
have the same directional effect on the disposi-
tion to use in every addictive state. Our next
result provides an example: if baseline well-
being deteriorates more rapidly as the addictive
state rises, then the disposition to use is lower in
every state. This property holds in the standard
rational addiction framework (p
s
a
0) and is
preserved in the presence of cue-triggered
mistakes.
PROPOSITION 3: Consider
and derived,
respectively, from w
s
(e, x, a) and w
s
(e, x, a)
w
s
(e, x, a) d
s
(with the same values of y
s
, r
s
,
q, and p
s
a
). If d
s
is weakly increasing in s, the
disposition to use is weakly higher with
than
with for all s.
Propositions 2 and 3 shed light on the rela-
tionship between income and the consumption
of addictive substances. While an increase in
income raises the inclination to experiment rec-
reationally, it can reduce the inclination to use
at higher addictive states; accordingly, the
model can generate higher rates of addiction
among the poor. To see why, suppose the utility
function has the following separable form: w
s
(e,
x, a) u(e) v
s
( x, a). What happens when we
add a xed increment, , to income in all states?
The parameters u
s
E
and u
s
A
all rise by u( y
s
) u( y
s
), which is weakly increasing in s
(assuming y
s
is weakly decreasing in s). The
parameters u
s
R
increase by u( y
s
r
s
)
u( y
s
) u( y
s
) u( y
s
), and the
parameters b
s
a
increase by [u( y
s
q )
u( y
s
)] [u( y
s
q) u( y
s
)] 0. Thus, we
can decompose the effect into three pieces: (a) a
xed increase in u
s
a
equal to u( y
s
) u( y
s
)
for each s and a E, A, R; (b) a general
30
From simulations, we know that a general increase in
u
s
R
increases the lowest state at which the DM selects
rehabilitation for some parameter values and decreases it for
others.
1570 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
increase in u
s
R
, and (c) a general increase in b
s
a
for a E, A. Proposition 3 tells us that the
effect of the rst piece is to weakly increase the
disposition to use in every state. Part (ii) of
Proposition 2 tells us that the second piece
increases the disposition to use in state 0, and
parts (i) and (iii) tell us the same thing for the
third piece. Thus, the addictive substance is
normal in state 0. Note, however, that the sec-
ond and third pieces have ambiguous effects on
the disposition to use in more advanced states of
addiction, which is why the income effect can
change signs.
B. Patterns of Consumption
According to our theory, the particular pat-
tern of consumption that emerges in any
instance depends systematically on the charac-
teristics of the individual (including aptitude for
cognitive control), the substance, and the envi-
ronment. For reasonable parameter values, the
model generates a wide variety of observed
consumption patterns.
31
Consider a highly addictive substance (p
s
a
large). If baseline well-being declines rapidly
with consumption, the DM may choose never to
use [(E, 0) at s 0]. For most people, crack
cocaine appears to be a good example of this
(see Goldstein, 2001). In contrast, if the decline
in baseline well-being is initially gradual but
accelerates from one state to the next, the model
can produce a pattern of progressive resistance.
That is, the DM may begin using the substance
intentionally, engage in half-hearted abstention
(and therefore use intermittently) after reaching
an intermediate addictive state, and shift to con-
certed abstention after a string of bad luck. If
bad luck continues, precommitment to absten-
tion through rehabilitation may follow with sub-
sequent probabilistic recidivism. If baseline
well-being attens out for sufciently advanced
addictive states (the DM hits bottom), the
model can also produce resignation. That is, a
DM may give up, opting for (E, 1) once he
reaches a highly addicted state after an unsuc-
cessful battle to abstain.
Now consider an enjoyable substance for
which baseline well-being declines slowly with
consumption. Irrespective of whether the prob-
ability of entering the hot mode is high or low,
constant use often emerges. Caffeine potentially
ts this description.
Finally, a sufciently sharp drop in the plea-
sure generated by the substance from one ad-
dictive state to the next can produce intentional
recidivism. That is, the DM may choose (E, 1)
in one state and (R, 0) in the next, in which case
he oscillates between the two. He enters reha-
bilitation in each instance without any desire to
stay clean; he knows that he will resume using
the substance upon release from rehabilitation,
and fully expects to enter rehabilitation once
again. This pattern is in fact observed among
serious heroin users when repeated use dilutes
the high (see Michael Massing, 2000). It is
evidence of fairly sophisticated, forward think-
ing among junkies whose objective is to renew
the high by temporarily getting clean, and who
know that rehabilitation accomplishes this more
reliably than abstention.
C. Explaining the Distinctive Features of
Addiction
In Section I we argued that addiction is as-
sociated with ve distinctive behavioral pat-
terns. Our theory generates each of these
patterns.
1. Unsuccessful Attempts to Quit.Suppose
life circumstances change over time, gradually
shifting the parameters of the DMs problem
from to . Suppose the DMs best choice for
state 0 is (E, 1) if prevails forever, but that
the optimal decision rule prescribes either (E,
0), ( A, 0), or (R, 0) for all s if prevails
forever. If the shift from to is either
unanticipated or anticipated and sufciently
slow, the DM starts using the substance but
subsequently decides to quit unconditionally.
With p
1
a
0, the attempt is unsuccessful when
either (E, 0) or ( A, 0) is chosen in state 1.
2. Cue-Triggered Recidivism.For the set-
ting described in the previous paragraph, unsuc-
cessful attempts to quit are associated with high
realizations of c(a, ) (that is, exposure to rel-
atively intense cues).
31
We have generated each of the patterns described in
this section through numerical simulations, which we omit
to conserve space.
1571 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
3. Self-Described Mistakes.In our model,
choices and preferences diverge whenever the
DM selects (E, 0) or ( A, 0) and then enters the hot
mode. This constitutes a recognizable mistake.
32
4. Self-Control through Precommitment.
The choice (R, 0) is a costly precommitment;
under our assumptions, its only purpose is to
remove the option of consuming the substance.
5. Self-Control through Behavioral and Cog-
nitive Therapy.The choice ( A, 0) involves
costly cue-avoidance. Its only purpose is to re-
duce the probability of encountering cues that
trigger mistaken usage. Though not modeled
explicitly, cognitive therapy would inuence
behavior in our setting by increasing M
T
(that is,
raising the threshold impulse required to defeat
cognitive control).
Our three central premises play critical roles
in accounting for each pattern. We can remove
cue-triggered mistakes by setting M
T
, so
that the DM always exercises cognitive control.
With this change, p
s
a
0 for all a and s, and the
DM always choose either (E, 1) or (E, 0). All
attempts to quit are successful, and there is no
recidivism. Preferences and choices never di-
verge, so there are no mistakes. The DM never
exercises self-control through precommitment
by choosing (R, 0), or through cue-management
by choosing ( A, 0). Some sophistication is also
essential; otherwise the DM would ignore his
susceptibility to cue-triggered errors and make
choices based on the mistaken assumption that
p
s
a
0 for all a and s.
We also observed in Section I that aggregate
consumption of addictive substances responds
to prices and information in the usual way. This
too is consistent with our theory, as users some-
times make decisions in the cold mode.
V. Demand-Side Policy Analysis
In this section we study the welfare effects of
various public policies concerning addictive
substances. In keeping with the focus of the
preceding sections, we restrict attention to de-
mand side welfare effects, ignoring supply
side consequences associated with the devel-
opment of black markets, the spread of corrup-
tion, and enforcement costs.
33
A. The Welfare Criterion
In formulating our model, we retain the
standard assumption that each individual has
a single coherent set of preferences. Our de-
parture, which is grounded in the evidence
from neuroscience presented in Section II, is
to assume that there are imperfections in the
process by which the brain makes choices,
and that these imperfections give rise to mis-
takes in identiable circumstances. Since the
individual has only one set of preferences,
discounted experiential utility,
t0
t
w
s
t
(e
t
,
x
t
, a
t
), accurately measures his well-being,
and is unambiguously the appropriate welfare
standard.
34
It may be tempting to reinterpret our model as
one with multiple selves, hot and cold,
where the preferences of the hot self can be
inferred from choices in the hot mode. Under
that interpretation, our use of cold preferences
as a welfare standard is arbitrary. In our view,
this interpretation commits a fallacy. By assum-
ing that choices are always consistent with un-
derlying preferences, it assumes away the
possibility that individuals make systematic
mistakes. This possibility is a central premise of
our analysis and is justied based on the state of
knowledge concerning the neuroscience of ad-
diction. One can certainly dispute the validity of
this premise. However, given the premise and
our adherence to the standard formulation of
preferences, the correct welfare criterion is
unambiguous.
32
When the HFM-generated forecast is sufciently posi-
tive, cognitive override may not occur even when higher cog-
nition forecasts undesirable consequences. Thus, an individual
may use a substance while simultaneously recognizing (in
terms of higher cognitive judgment) that this is a mistake.
33
Supply-side effects are discussed elsewhere (see,
e.g., MacCoun and Reuter, 2001; J. Miron and J.
Zwiebel, 1995).
34
This is in contrast with a number of the behavioral
theories discussed in Section VI, for which one must either
use a weak welfare standard such as the Pareto criterion
(applied to multiple selves or multiple perspectives), or
select a particular method of resolving conicting prefer-
ences, for example, by respecting the tastes of only one self
or perspective.
1572 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
B. Policy Objectives
What might society hope to accomplish
through public policies regarding addictive sub-
stances? Possible objectives include protecting
third parties from externalities (e.g., second-
hand smoke), combatting misinformation and
ignorance, moderating the consequences of un-
insurable risks, and helping consumers avoid
mistakes. Both externalities and informational
problems provide well-understood rationales
for government intervention, and neither is in-
trinsically linked to the novel aspects of our
framework. We therefore focus primarily on the
last two sets of objectives.
1. Amelioration of Uninsurable Risks.Risk
and uncertainty relating to the effects of envi-
ronmental cues on decision processes are cen-
tral to our model. The DMs lack of knowledge
concerning future states of nature,
t
, prevents
him from perfectly forecasting future decision
modes and choices in states for which he plans
to select either (E, 0) or ( A, 0) and thereby
creates uncertainty about subsequent addictive
states. This translates into monetary risk be-
cause his resources depend on his addictive
state, and because variation in expenditures on
the addictive substance and rehabilitation imply
variation in consumption of the nonaddictive
good. Since the nancial consequences of ad-
dictionits effects on job retention, productiv-
ity, out-of-pocket medical costs (including
rehabilitation), and, for some substances (e.g.,
cocaine, heroin), direct expendituresare often
substantial, this risk is quantitatively signicant.
35
From the perspective of risk (ignoring other
considerations), policies that create actuarially
fair redistributions over realizations of future
states of nature are benecial (harmful) when
they distribute resources toward (away from)
outcomes for which the marginal utility of non-
addictive consumption is relatively high. In sub-
sequent sections, we initially impose the
assumption that w
s
(e, x, a) u(e) v
s
( x, a),
which implies that the marginal utility derived
from nonaddictive consumption depends only
(and inversely) on the level of nonaddictive
consumption. We focus on this case because we
regard it as a natural benchmark, but we also
discuss the implications of relaxing separability.
We assume throughout that private insurance
markets fail completely. As is well known, the
welfare effects of public policies that redistrib-
ute resources across states of nature can depend
on the specic factors that cause markets to fail
(see, e.g., Mark V. Pauly, 1974). It is therefore
important to specify the source of the market
failure and to explain how it interacts with the
policies considered.
We assume that private insurance companies
are unable either to observe or to verify the state
of nature
t
, cues, the DMs decision mode, the
addictive state, lifestyle activities, or consump-
tion of the addictive substance.
36
The govern-
ment is similarly handicapped. However, unlike
private companies, it can observe transactions
involving legal addictive substances (typically
without identifying purchasers), and it can ma-
nipulate the prices of these commodities
through taxation and subsidization.
Private companies can observe aspects of
treatment (rehabilitation and medical costs), but
we assume that treatment insurance is unavail-
able because (i) practical considerations pre-
clude ex ante contracting at age zero when risks
are homogeneous (e.g., before teen or even pre-
teen exposure), and (ii) adverse selection arising
from ex post heterogeneity precludes ex post
contracting.
37
The government is similarly
handicapped by the second problem but can
avoid the rst by imposing a universal policy on
all consumers ex ante.
2. Mistake Avoidance.Public policy can
potentially improve welfare by creating con-
ditions that reduce the frequency with which
individuals experience decision process mal-
functions or by forcing them to make alternative
35
Among chronic users, average annual expenditures on
cocaine and heroin exceeded $10,000 in 1999 (Ofce of
National Drug Control Policy, 2001a).
36
With respect to the addictive state, clinical diagnosis
of addiction is both costly and imprecise. Consumption is
potentially observable when the substance is dispensed as a
prescription medicine, but in that case the same problems
arise as for treatment (discussed in the next paragraph).
37
In practice, private health insurance policies do pro-
vide some coverage for the treatment of addiction. Yet
many people are not insured, and coverage is typically
incomplete.
1573 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
choices when malfunctions occur. In our model,
these possibilities correspond, respectively, to
reducing the probability of entering the hot mode
and to ensuring abstention when appropriate.
In solving the DMs optimization problem,
we treated the probabilities of entering the hot
mode, p
s
a
, as xed parameters. Many of the
policies considered below potentially change
these parameters; for example, one can model a
ban on advertising as a reduction in the volume
of cues encountered. For some portions of our
analysis, we also allow for the possibility that p
s
a
depends on the price of the addictive good, q,
and/or the DMs income, y. We reason that the
frontal cortex is more likely to generate stronger
cognitive incentives and override cues when the
immediate consequences of use are more se-
vere.
38
Accordingly, we assume that M
T
weakly
decreases with y, weakly increases with q, and
weakly increases with an equal increase in y and
q.
39
It is useful to state these assumptions more
compactly in terms of p
s
a
(with the added tech-
nical requirement of differentiability).
ASSUMPTION 3: The probability of entering
the hot mode, p
s
a
, is differentiable in q and y
with
q
p
s
a
y
p
s
a
0.
C. When Is Government Intervention
Justied?
When do the objectives discussed in Sec-
tion V B potentially justify government inter-
vention? The following result provides an
initial answer. Here and elsewhere, we say
that use is continual if the DM selects (E, 1)
in every state.
PROPOSITION 4: (i) Continual use solves the
DMs choice problem if and only if it is rst best
(in the sense that it solves the maximization
problem when p
s
a
0 for all a and s). (ii)
Suppose there is some state s with p
s
E
0 such
that (E, 1) is not a best choice in s. Then the
DMs choices are not rst best (in the sense that
setting p
s
a
0 for all a and s and reoptimizing
strictly increases the value function for some
states).
Part (i) tells us that noncontinual use is
necessary for the existence of a benecial
policy intervention. Laissez-faire is therefore
the best policy for substance users who make
no serious attempt to abstain (e.g., contented
smokers or coffee drinkers). Notably, this
conclusion follows even when the substance
in question is highly addictive ( p
s
a
rises
sharply with s) and well-being declines sig-
nicantly with long-term use. The intuition is
the same as for the nal portion of Proposi-
tion 1.
Part (ii) tells us that noncontinual use is suf-
cient for the existence of a theoretical policy
intervention that benets the DM, provided the
departure from continual use occurs in a state
for which the DM is susceptible to cue-triggered
mistakes. Of course, this intervention may be
impractical given the governments information
constraints.
D. A Framework for Tax Policy Analysis
The formal results below concern the de-
sirability of various types of tax policies. Fol-
lowing standard practice, we evaluate these
policies by embedding our decision-maker in
a simple economy and studying effects on
equilibrium allocations. Here we outline the
structure of the economic environment. Nota-
tion and some additional formal details ap-
pear in Appendix A.
The economy consists of an innite sequence
of generations. In the absence of government
intervention, every member of every generation
is identical and confronts the decision problem
38
Since cognitive control often must be asserted quickly
if at all, and since extrapolation of future consequences is
time consuming, we implicitly assume that the deployment
of cognitive control responds to variation in immediate
circumstance-specic consequences, but not to variation in
future circumstance-specic consequences.
39
When q is higher or y is lower, the immediate negative
consequences of use are plainly greater, and potentially
more likely to occupy the DMs awareness. When q and y
rise by equal amounts, the immediate hedonic payoff from
abstention rises while the immediate hedonic payoff from
use is unchanged, so the immediate negative consequence of
use is again more severe. It would also be natural to assume
that cue exposure, c(a, ), weakly declines in q (since use
among social contacts declines), and this would reinforce
Assumption 3. In principle, c(a, ) could rise or fall with
income.
1574 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
described in Section III. We interpret the DMs
discount factor as the product of a pure rate of
time preference and a constant single-period
survival probability. We assume that the size of
each new generation is just sufcient to keep the
total population constant. The total population
is very large, and realization of hot and cold
states are independent across DMs, so there is
essentially no aggregate uncertainty. Income ar-
rives in the form of the nonaddictive good, and
the addictive good is produced under competi-
tive conditions with constant-returns-to-scale
technology, as are rehabilitation services. Thus,
q and r
s
are xed and equal to unit production
costs (where the costs of rehabilitation services
potentially vary with the addictive state).
In each period, the government can tax (or
subsidize) either the addictive substance or re-
habilitation (we consider these instruments one
at a time). There is no revenue requirement;
taxes are purely corrective. By assumption, the
government cannot condition the associated tax
rates on either the DMs age or his addictive
state. This could reect either the practical dif-
culties associated with tailoring these taxes
and subsidies to an individuals conditions (in-
cluding the need for clinical diagnosis) or pri-
vacy concerns. Imagine in particular that the
tax/subsidy for the addictive substance is either
applied to anonymous transactions (like a sales
tax) or imposed on producers (like a value-
added tax), while the rehabilitation tax/subsidy
nominally falls on service providers (again, like
a value-added tax). The government can also use
age-specic (equivalently, generation-specic)
lump-sum instruments. An intertemporal policy
species values for all available tax/subsidy in-
struments in every period.
Since there is no borrowing or lending in our
model, and since we do not wish to advantage
the government articially, we assume that pol-
icies cannot redistribute resources across peri-
ods. We say that an intertemporal policy is
feasible if there is, for each generation, an op-
timal decision rule such that the governments
budget is balanced in every period. Feasible
policies permit within-period transfers across
generations, which can mimic borrowing and
lending, thereby leaving the government in an
articially advantageous position. One could
therefore argue for a stronger restriction requir-
ing government budget balance for each gener-
ation within each period. While we impose the
weaker requirement, our results also hold for
the stronger requirement.
40
A steady-state policy prescribes a constant
tax rate and constant age-specic lump-sum
taxes. Notably, each individuals problem is
potentially nonstationary because steady-state
lump-sum tax/subsidies may change with age.
The set of feasible steady-state policies includes
the zero-tax alternative, henceforth denoted ,
for which all tax/subsidy instruments are set to
zero.
In the next two sections, we focus on the
steady-state welfare effects of steady-state
policies (often dropping the modier steady-
state for brevity). For any steady-state pol-
icy, we use the lifetime expected discounted
hedonic payoff for the representative individ-
ual as our welfare measure. An optimal
steady-state policy maximizes this payoff
among all feasible steady-state policies. This
objective function respects each individuals
time preference over his own lifetime but is
innitely patient with regard to intergenera-
tional comparisons, in effect placing equal
weight on all generations. Since the DMs
choice set is discrete, best choices are often
insensitive to small parameter changes, so the
optimal policy is typically not unique.
E. Taxation and Subsidization of Addictive
Substances
Addictive substances are often heavily taxed
(e.g. nicotine and alcohol) and occasionally
subsidized (see, e.g., the description of a Swiss
heroin prescription program in MacCoun and
Reuter, 2001). Some policy analysts argue for
taxation of addictive substances on the grounds
that this discourages excessive use (e.g., Gruber
and Koszegi, 2001). Others suggest that, in the
absence of externalities, use is voluntary so
laissez-faire is best (e.g., Becker and Murphy,
1988). Our theory of addiction suggests a more
nuanced view.
Proposition 5 below relates the sign of the
optimal tax rate on the addictive substance to
40
In fact, the proof of Proposition 5 requires only minor
adjustments when we impose the strong requirement; Prop-
osition 6 holds as stated under either restriction.
1575 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
observable patterns of consumption. Notably,
the consumption patterns that determine opti-
mal tax rates are endogenous, and the proposi-
tion requires us to assess them at the optimal tax
rates.
41
This feature is common to many well-
known optimal-tax results. For example, the
Ramsey rule relates optimal commodity tax
rates to compensated demand elasticities evalu-
ated at the optimal tax rates (though the rule is
frequently stated in a way that disguises this
dependency).
The proposition refers to the following two
possible patterns involving the likelihood of
use.
42
Condition A: For every age t, the likelihood of
use is weakly increasing in s over states reached
with positive probability at that age, and the
DM does not enter rehabilitation in the lowest
such state.
43
Moreover, at some age t, at least
two addictive states are reached with positive
probability.
Condition B: For every age t, the likelihood of
use is weakly decreasing in s over states reached
with positive probability at that age.
44
More-
over, at some age t, at least two addictive states
are reached with positive probability, with nei-
ther expected use nor y
s
constant over such
states.
For each condition, the requirement for all t
is less demanding than it might initially appear.
Remember that, in the absence of taxes and
subsidies, each DMs problem is stationary, and
the best choices at each state are independent of
age. In a steady state for the economy, age
matters only because it affects the lump-sum tax
(or subsidy). If the lump sums are relatively
small, the general pattern of use will tend to be
similar at different ages provided it is not too
sensitive to small changes in income.
PROPOSITION 5: Suppose that w
s
(e, x, a)
u(e) v
s
( x, a), that y
s
is weakly decreasing in
s, and that p
s
a
does not depend on prices or
income.
(i) Consider an optimal steady-state policy for
which all budget-balancing optimal deci-
sion rules satisfy Condition A. The tax
rate on the addictive substance is strictly
negative.
(ii) Consider an optimal steady-state policy for
which all budget-balancing optimal deci-
sion rules satisfy Condition B. If q is suf-
ciently small, the tax rate on the addictive
substance is strictly positive.
To develop intuition for this result, note that
taxation (or subsidization) potentially affects
welfare through three channels. First, it can
change decisions in the cold mode. Second, it
can redistribute resources across uncertain out-
comes. Third, it can alter the effects of environ-
mental cues on operational decision modes
(through the trigger mapping T). With p
s
a
inde-
pendent of prices and income, the third channel
vanishes (we discuss the implications of rein-
stating it below). Effects involving the second
channel dominate welfare calculations for small
taxes and subsidies because they are generally
rst order, while effects involving the rst chan-
nel are not.
45
Accordingly, starting from a sit-
uation with no taxes, one can determine whether
a small tax or subsidy improves welfare by
focusing on the correlation between the taxed
41
Alternatively, one can make statements about welfare-
improving changes, assessing usage patterns at arbitrary
starting points. For example, from the proof of Proposition
5, we also have the following results: eliminating a positive
tax is benecial if initially Condition A holds; eliminating a
positive subsidy is benecial if initially Condition B holds
and q is small; if Condition A holds with the no-tax policy
, a small subsidy is welfare improving; if Condition B
holds with and q is small, a small tax is welfare improving.
42
In Appendix A, we dene
s
t
() as the probability of
use in state s at age t given a decision rule , accounting for
the possibility of entering the hot mode. Here, the likeli-
hood of use refers to
s
t
(). Note that the likelihood of use
is necessarily weakly increasing in s when the disposition to
use is weakly increasing in s.
43
Since the DMs decision problem is potentially non-
stationary, it is possible for him to nd himself in a state
beyond the lowest one in which he would select rehabilita-
tion during the same period.
44
This occurs, for example, if best choices are unique,
the disposition to use is weakly decreasing in s, the rst
intentional use interval is nonempty, and p
s
a
is constant
outside of this interval (i.e., the DM is fully addicted by the
time he attempts to refrain from consuming the substance).
45
With continuous decision variables and interior solu-
tions, the rst channel would be second order for small taxes
and subsidies. With discrete decision variables (as in our
current model), it is literally zero for sufciently small taxes
and subsidies.
1576 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
activity and the marginal utility of nonaddictive
consumption.
46
This helps to build intuition
concerning the signs of optimal tax rates. One
should bear in mind, however, that the optimal-
tax problem is more complex because, at the
optimum, welfare effects involving the rst
channel (incentive effects) are also rst order.
A policy provides de facto insurance if it
redistributes age-t resources toward outcomes
in which the marginal utility of nonaddictive
consumption is relatively high, which in the
benchmark case [w
s
(e, x, a) u(e) v
s
( x, a)]
means that the level of nonaddictive consump-
tion is relatively low. This occurs when other
expenditures are high and when the state is itself
high (since y
s
weakly declines with s).
47
A
subsidy necessarily redistributes resources to-
ward outcomes with relatively high expendi-
tures on addictive substances. Moreover, if the
likelihood of use increases with the addictive
state (Condition A), it also redistributes re-
sources toward outcomes for which income is
relatively low. Since both effects are benecial,
a subsidy is desirable [part (i)]. Conversely, if
the likelihood of use declines with the addictive
state (Condition B), a positive tax (with a
budget-balancing lump-sum payout) redistrib-
utes resources toward outcomes for which in-
come is relatively low and rehabilitation
expenditures are relatively high. It also redis-
tributes resources away from outcomes with
relatively high expenditures on addictive sub-
stances, but this effect is secondary when q is
small, rendering the tax benecial [part (ii)].
Proposition 5 underscores the fact that differ-
ent policies are appropriate for different addic-
tive substances, and that the characteristics of
good policies are related to usage patterns. As
we have seen in Section IV, usage patterns are
in turn systematically related to aspects of the
substance, the user, and the environment.
Part (i) suggests that a subsidy may be wel-
fare improving in the case of a substance for
which initial use tends to be spur of the mo-
ment, but where an intention to use becomes
increasingly predominant as the individual be-
comes more addicted. The argument for subsi-
dization is stronger when the substance in
question is more expensive. The apparent im-
plication that the government might benecially
subsidize substances such as cocaine and heroin
is provocative to say the least, and it should be
tempered by several considerations, including
the likely existence of externalities, the poten-
tial effects of price and income on the trigger
mechanism (discussed below), and the fact that
Condition A apparently does not hold univer-
sally, as many addicts seek treatment. Still, our
analysis adds a potentially important cautionary
note to existing discussions of the benets of sin
taxes (e.g., Gruber and Koszegi, 2001; Ted
ODonoghue and Matthew Rabin, 2004), which
can violate social insurance principles by penal-
izing those who have experienced bad luck. It
also provides a framework for understanding
the potential benets of somewhat more rened
approaches, such as the Swiss policy of provid-
ing cheap heroin to users who cross some diag-
nostic threshold of addiction, and who are not
interested in rehabilitation.
Part (ii) suggests that a tax may be welfare
improving in the case of an inexpensive sub-
stance that people initially use regularly, for
which attempts to abstain begin only after cue
triggers are well established and stable (so that
they change little with further use). Coffee, cig-
arettes, and alcohol arguably fall into this
category.
How robust are these ndings? Complemen-
tarity between addictive and nonaddictive con-
sumption would raise the marginal utility of
nonaddictive consumption whenever the DM
uses the addictive good, strengthening the ad-
vantages of a subsidy, thereby reinforcing part
(i) but potentially reversing part (ii). Substitut-
ability would reduce the marginal utility of non-
addictive consumption whenever the DM uses
the addictive good, strengthening the advan-
tages of a tax, thereby reinforcing part (ii) but
potentially reversing part (i).
46
There are some subtleties here. A small tax or subsidy
that changes cold-mode decisions can alter the correlation
between a taxed activity and the marginal utility of nonad-
dictive consumption, thereby changing effects through the
second channel. With discrete choice sets, the pertinent
correlation can change dramatically even for tiny taxes and
subsidies.
47
In principle, the government could also redistribute
resources through an income tax. Implicitly, we take the
income-tax system as exogenously given. This is reasonable
as long as addiction is not one of the primary factors
inuencing income distribution and the equityefciency
trade-offs that an optimal income tax system is intended to
address.
1577 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
We can also relax the assumption that p
s
a
is
invariant with respect to taxation and subsidi-
zation. A tax of per unit increases price by
and, from individuals of age t, raises less than
in per capita revenues. Suppose the government
distributes the revenue raised from each age
group back to the same age group as a lump
sum. Since the amount received by each indi-
vidual is less than the price increase, Assump-
tion 3 implies that p
s
a
falls in every state. Any
policy that reduces p
s
a
weakly increases welfare
through the third channel [strictly if the state is
reached with positive probability and the DM
selects (a, 0)]. This strengthens the advantages
of a tax, reinforcing part (ii) of the proposition,
and potentially reversing part (i).
48
F. Harm-Reduction Policies
Subsidization of rehabilitation is relatively
common. Popular justications appeal to the
notion that treatment should be affordable and
universally available, though sometimes posi-
tive externalities are invoked.
In the context of our model, there are at least
two reasons to subsidize rehabilitation. First,
this provides de facto insurance for a large,
uncertain expense. Second, under Assumption
3, the DM is less prone to make cue-triggered
mistakes when he receives resources in kind
(through a rehabilitation subsidy) rather than in
cash.
There is, however, an additional consider-
ation arising from the correlation between reha-
bilitation and income. If rehabilitation is more
likely at advanced stages of addiction, then a
subsidy benecially redistributes resources to-
ward low-income states. Since this reinforces
the considerations discussed in the previous
paragraph, subsidized rehabilitation is unambig-
uously desirable. If, however, rehabilitation is
less likely at advanced stages of addiction, then
a subsidy detrimentally redistributes resources
toward high-income states, offsetting the con-
siderations discussed in the previous paragraph.
Formally, one can prove a result analogous to
Proposition 5, relating the optimal tax/subsidy
treatment of rehabilitation to rehabilitation
patterns.
Our next result deals instead with the welfare
effects of small rehabilitation taxes and subsi-
dies. It shows that a small rehabilitation subsidy
is benecial, and a small tax harmful, under
extremely general conditions: at the no-tax al-
ternative , rehabilitation must be chosen in
some state, and there must be some random-
ness.
49
Here, we allow from the outset for the
possibility that p
s
a
depends on y
s
and q.
PROPOSITION 6: Suppose that w
s
(e, x, a)
u(e) v
s
( x, a), that y
s
is weakly decreasing in
s, and that r
s
q for all s. Suppose also that, in
the absence of taxes and subsidies (that is, with
policy ), the following conditions hold: rst,
there is at least one state in which rehabilitation
is a best choice; second, rehabilitation is the
unique best choice in the earliest of these; third,
for some earlier state (other than 0), (E, 1) is
not a best choice. Then, within the class of
policies that do not create net inter-cohort
transfers, a small steady-state subsidy for reha-
bilitation is benecial, and a small steady-state
tax is harmful.
Since Proposition 6 holds even when the cost
of rehabilitation is very small, it is not primarily
about the desirability of insuring a large, uncer-
tain expense. For the correct intuition, note that
with the no-tax alternative , each DMs prob-
lem is stationary, so best choices for each state
are independent of age. This implies that the
DM can never advance beyond the rst state in
which (R, 0) is the best choice. Consequently,
the likelihood of rehabilitation is positively cor-
related with the addictive state and negatively
correlated with income, so the three effects dis-
cussed at the outset of this section work in the
same direction, in favor of subsidization. The
practical lesson is simple: if addiction is rela-
tively unlikely to advance beyond the point
where people start to seek rehabilitation, then
subsidies are unambiguously desirable.
An appropriately modied version of our
model could address the effects of other harm-
48
The proof of Proposition 6 formally demonstrates a
closely related point in the context of a subsidy for rehabil-
itation services.
49
The assumption that (E, 1) is not a best choice in every
state up to the rst in which rehabilitation is selected en-
sures some randomness.
1578 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
reduction policies such as needle exchanges.
We leave this for future work.
G. Criminalization
Historically, criminalization has been the cor-
nerstone of U.S. drug policy, with more than
600,000 citizens incarcerated for drug-related
offenses in 1999 (Ofce of National Drug Con-
trol Policy, 2001b). It affects users through two
distinct channels: a price effect and a rationing
effect. The price effect refers to changes in the
marginal cost of using the substance resulting
from penalties and other costs imposed on users
and suppliers. The rationing effect refers to in-
terference with the process of matching buyers
and sellers: since criminalization forces buyers
and sellers to carry out transactions secretively,
buyers sometimes have difculty locating
supply.
50
It is instructive to consider the price and
rationing effects separately. The price effect is
equivalent to a tax policy in which the revenue
raised by the tax is destroyed. If criminalization
only created a price effect, taxation would dom-
inate it.
Now consider the rationing effect. Disrupt-
ing access to supply is potentially benecial
when the DM chooses (E, 0) or ( A, 0), and
potentially detrimental when he chooses (E,
1). However, the impact of the rationing ef-
fect on consumption may be smaller when the
DM chooses (E, 1). An individual who in-
tends to consume an illegal substance can set
about locating supply deliberately and sys-
tematically and can maintain stocks in antic-
ipation of transitory difculties. In the
extreme case where the rationing effect has
no impact on consumption when the DM se-
lects (E, 1), it is unambiguously benecial.
This conclusion is obviously weakened, or
even reversed, if unsuccessful search activity
is costly (e.g., because it exposes the DM to
physical harm).
51
It follows that, in some circumstances, crimi-
nalization may be superior to taxation and to
laissez-faire. This result deserves emphasis,
since it is difcult to justify a policy of crimi-
nalization based on demand-side welfare con-
siderations without adopting the nonstandard
perspective that supply disruptions can avert
mistakes.
52
Since it is better not to disrupt
planned consumption, the case for criminaliza-
tion is, ironically, strongest when enforcement
is imperfect.
H. Selective Legalization with Controlled
Distribution
Some policies permit transactions involving
addictive substances in certain circumstances
but not in others. Examples include a 1998
Swiss law legalizing the prescription of heroin
for severe addicts and blue laws prohibiting
alcohol sales on Sundays.
Policies of selective legalization with con-
trolled distribution often make deliberate
planning a prerequisite for availability, selec-
tively disrupting impulsive use without dis-
turbing planned use (assuming the hot mode
only activates behaviors that target immediate
consumption). This effect is potentially ben-
ecial, if unintended. For example, with blue
laws, alcoholics can make themselves less
vulnerable to compulsive drinking on Sun-
days by choosing not to stock up in advance.
These laws appear to reduce impulsive use in
practice (Peter T. Kilborn, 2003; T. Norstrom
and O. J. Skog, 2003).
A prescription requirement can play a similar
role, provided prescriptions are lled with a lag.
To represent this possibility formally, we mod-
ify our model as follows. Imagine that, in each
period m, the DM must decide whether to call
in a prescription for the substance. Taking this
action makes the substance available in period
m 1; otherwise it is unavailable, and con-
sumption is impossible.
With this option, the DM can always
achieve the rst-best outcome. Solving the
50
Probabilistic consumption following the choice (E, 1)
changes the value function somewhat, but the results from
Section IV extend to this case. See Goldstein and Kalant
(1990) for evidence that drug usage declines as substances
become less available.
51
The costs of a successful search are part of the price
effect.
52
Though the mechanisms considered in this paper in-
volve stochastic mistakes, the same conclusion would fol-
low in a model with deterministic mistakes, for example,
one in which the DM always errs by placing too much
weight on the immediate hedonic reward.
1579 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
dynamic programming problem with p
s
0
for all s yields a deterministic consumption
path. The DM can mimic this outcome by
calling in his prescription in period m if and
only if he consumes the substance in period
m 1 on the rst-best path. In this way, he
precommits to the rst-best choice by opti-
mally rationing himself.
53
If the hot mode also activates behaviors that
target future consumption, the preceding policy
is ineffective. However, a small modication
restores the rst-best outcome: allow the DM to
cancel irrevocably, at any point in period m, his
prescription for m 1. It is then optimal for
him to cancel during period m while in the cold
mode if and only if he does not wish to consume
the substance in period m 1.
54
In more realistic settings, these policies
might not permit consumers to achieve rst-
best outcomes. If, for example, the desirabil-
ity of using a substance in period m depends
upon conditions (e.g., mood) that are not re-
solved until the period is underway, the indi-
vidual may sometimes regret failing to call in
a prescription. However, the policy still
weakly benets consumers because it pro-
vides them with a tool for self-regulation
without mandating its use.
Heterogeneity across individuals makes se-
lective legalization with controlled distribution
even more attractive relative to other policies. A
prescription program accommodates heteroge-
neity by providing consumers with discretion:
intentional users can continue to indulge with-
out impediment, while unintentional users nev-
ertheless benet from improved self-control. In
contrast, any feasible tax, subsidy, or criminal
statute may be inappropriateeven harmful
for large subsets of consumers.
The policies considered in this section would
be advantageous in any model where the DM
makes similar types of mistakes and where he
understands this proclivity. The particular sto-
chastic mechanism discussed in this paper is not
essential. Our conclusions do depend on the
assumption that the government can limit resale
of the substance (e.g., by requiring on-site ad-
ministration) and can suppress illicit supply.
Notably, selective legalization impairs black
markets by siphoning off demand.
I. Policies Affecting Cue-Triggered Decision
Processes
In our model, public policy can potentially
help consumers by attenuating either exposure
or sensitivity to cues (i.e., reducing c(a, ) or
M(c, s, a, ) or raising M
T
). Arguably, the
producers of addictive substances raise the like-
lihood of triggering hot modes by exposing
consumers to ubiquitous cues through billboards,
television advertisements, product placement in
stores, and so forth. Advertising and marketing
restrictions of the type imposed on tobacco and
alcohol may eliminate a cause of compulsive use.
Restrictions on public consumption may have
similar effects.
Other public policies may reduce cue-
sensitivity by creating counter-cues. Brazil
and Canada require every pack of cigarettes to
display a prominent viscerally charged image
depicting some deleterious consequence of
smoking, such as erectile dysfunction, lung dis-
ease, or neonatal morbidity.
55
These counter-
cues are designed to activate the cognitive
control process described in Section II.
In our model, policies that reduce the likeli-
hood of cue-triggered mistakes by removing
problematic cues or establishing counter-cues
unambiguously increase welfare. As with selec-
tive legalization, these policies are attractive
because they are noncoercive, because they
accommodate individual heterogeneity, and
because they have the potential to reduce unin-
tended use without distorting choice in the cold
decision mode. Though individuals may have
some ability to avoid problematic cues and cre-
ate their own counter-cues, the government is
arguably better positioned to do this.
53
In a related analysis, Loewenstein et al. (2000) em-
phasize the role of mandatory waiting periods in a model
where agents systematically overconsume durable goods.
54
Alternatively, if the hot mode has a greater tendency
to activate behaviors targeting future consumption when the
planning horizon is short, one could restore (or at least
enhance) the policys efcacy simply by lengthening the lag
between prescription requests and availability (e.g., calling
in a prescription in period m makes the substance available
in m k, with k 1).
55
See http://www.hc-sc.gc.ca/hecs-sesc/tobacco/
research/archive/ for a description and some preliminary
evidence on the effectiveness of the Canadian program.
1580 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
VI. Related Literature
Existing economic theories of addiction in-
clude (i) variations on the standard model of
intertemporal decision making (Becker and
Murphy, 1988; Orphanides and Zervos, 1995),
including generalizations that allow for random
shocks and state-contingent utility (Hung, 2000;
Laibson, 2001), (ii) models with projection
bias wherein agents mistakenly assume that
future tastes will resemble current tastes, but
which otherwise conform to the standard model
(Loewenstein, 1996, 1999; Loewenstein et al.,
2001), (iii) models with present-biased prefer-
ences and either naive or sophisticated expecta-
tions (ODonoghue and Rabin, 1999, 2000;
Gruber and Koszegi, 2001), and (iv) models of
temptation wherein well-being depends not
only upon the chosen action, but also on actions
not chosen (Gul and Pesendorfer, 2001a, b;
Laibson, 2001). While all of these theories con-
tribute to our understanding of addiction and
share some important features with our model,
none adheres to all of the central premises set
forth and justied in Section II. In particular,
none of these models depicts addiction as a
progressive susceptibility to stochastic environ-
mental cues that can trigger mistaken usage.
All models of rational addiction (beginning
with Becker and Murphy, 1988) presuppose
complete alignment of choices and time-
consistent preferences, thereby denying the pos-
sibility of mistakes. Precommitments are never
strictly benecial, and a user would never state
a sincere, unconditional intention to quit with-
out following through. Stochastic environmen-
tal cues play a role in Laibsons (2001)
extension, but the mechanism involves hedonic
effects (cues trigger a change in taste for the
substance) rather than mistakes. Laibsons
framework can account for voluntary admission
to rehabilitation clinics and related behaviors
provided that these activities reduce the likeli-
hood of experiencing cravings. However, it
cannot account for the observation that many
addicts seek in-patient treatment not because
they expect to avoid cravings, but rather pre-
cisely because they anticipate cravings and wish
to control their reactions. Furthermore, even in
instances where entering a rehabilitation facility
does reduce the likelihood of cravings (e.g., by
removing environmental cues), the standard
framework implies counterfactually that the ad-
dict would nd the facilitys program more at-
tractive if it made the substance available upon
demand (in case of cravings).
Adding projection bias to the standard model
introduces the possibility that users may regard
past actions as mistakes. For example, an addict
may blame his initial drug use on a failure to
anticipate the escalating difculty of abstention.
Coupled with state-contingent utility shocks (as
in Laibsons model), projection bias could ac-
count both for the high frequency of attempted
quitting (when not triggered, users underesti-
mate the future difculty of abstention), and the
high frequency of failure (once triggered, users
overestimate the future difculty of continued
abstention). However, even with projection
bias, an otherwise standard decision-maker
would never anticipate making mistakes in the
future and sees no need for precommitments.
In models with present-biased decision-
makers, choice is always aligned with the pref-
erences prevailing at the moment when the
choice is made. Even so, one can interpret
present-bias as shorthand for considerations that
lead to systematic mistakes in favor of imme-
diate gratication, contrary to true (long-run)
preferences (see, e.g., Gruber and Koszegi,
2001). As a model of addiction, this framework
suffers from two main shortcomings. First, the
decision-making bias is not domain-specic. A
present-biased decision-maker mistakenly con-
sumes all pleasurable commodities excessively;
in this respect, there is nothing special about
addictive substances. Second, the bias is always
operativeit is not cue-conditioned.
In principle, one could formulate a model
with a powerful, narrow-domain, cue-triggered
present-bias. The resulting model (which does
not appear in the literature) would conform to
our premises; indeed, it would be nearly equiv-
alent to our approach. Our model is somewhat
simpler and more tractable than this alternative
because we treat behavior in the hot mode as
mechanical, whereas this present-bias approach
would portray even triggered choices as optimal
given well-behaved preferences. Naturally, for
our model, one can say that the triggered decision-
maker acts as if he optimizes subject to well-
behaved preferences that attach enormous
importance to consuming the addictive sub-
stance, but we think this as if representation is
1581 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
unenlightening. Since the decision-maker is as-
sumed always to consume the substance in the
hot mode, and since we regard this as a mistake
whenever he would behave differently in the
cold mode, the representation illuminates nei-
ther choices nor welfare.
Finally, Gul and Pesendorfer (2001a, b) model
addictive behaviors by dening preferences over
both the chosen action and actions not chosen,
thereby providing a potential role for temptation
and a rationale for precommitment. Their axiom-
atic approach embraces the doctrine of revealed
preference and therefore presupposes an align-
ment of choices and preferences, ruling out the
possibility of mistakes. In addition, their model, as
formulated, does not examine the role of stochas-
tic cues in stimulating use.
VII. Final Remarks
This paper develops an economic model of
addiction based on three premises: (i) use
among addicts is frequently a mistake (a patho-
logical divergence between choice and pref-
erence); (ii) experience with an addictive sub-
stance sensitizes an individual to environmental
cues that trigger mistaken usage; and (iii) ad-
dicts understand their susceptibility to cue-trig-
gered mistakes and act with some degree of
sophistication. We argue that these premises
nd strong support in evidence from psychol-
ogy, neuroscience, and clinical practice. Re-
search indicates that addictive substances
systematically interfere with the proper opera-
tion of an important process which the brain
uses to forecast near-term hedonic rewards
(pleasure), and this leads to strong, misguided,
cue-triggered impulses that often defeat higher
cognitive control. As a matter of formal math-
ematics, our model is tractable and involves a
small departure from the standard framework. It
generates a plausible mapping from the charac-
teristics of the user, substance, and environment
to dynamic behavior. It accounts for a number
of important patterns associated with addiction,
gives rise to a clear welfare standard, and has
novel implications for public policy.
Our theory also has potentially important im-
plications for empirical studies of addiction. It
suggests that users of addictive substances may
respond very differently to changes in prices,
with dramatically different implications for
welfare, depending on whether decisions reect
hot impulses or cold deliberation. In con-
trast, existing studies treat data on consumption
as if it were generated by a single process.
The model could be extended in a variety of
ways to improve realism and predictive power.
Possibilities include: developing a more com-
plete model of cognitive control in which future
consequences may inuence the likelihood of
overriding HFM-generated impulses (through
the threshold M
T
); adding stochastic taste shocks
realized at the outset of each period (to produce
variation in the contingent plan chosen for each
state); allowing payoffs (w
s
) to depend directly
on (to reect the hedonic effects of cravings);
allowing for imperfect information concerning
an individuals susceptibility to cue-triggered
mistakes; introducing partial, rather than full,
self-understanding; modeling life-cycle changes
(either anticipated or unanticipated) in prefer-
ences and susceptibilities resulting from aging
and changes in circumstances; and modeling the
long-term effects of early-life experiences.
It is natural to wonder whether the model
applies not just to addictive substances, but also
to other problematic behaviors such as overeat-
ing or compulsive shopping. These questions
are currently the subject of study among neuro-
scientists and psychologists, and it is too early
to say whether similar brain processes are at
work.
56
Notably, people who suffer from patho-
logical gambling, overeating, compulsive shop-
ping, and kleptomania describe their experience
as involving strong and often overwhelming
cravings, they respond to cues such as stress and
advertisements, and they exhibit cycles of
binges and abstention.
APPENDIX A: THE DYNAMIC ECONOMY
This appendix contains additional technical
details concerning the economy described in
56
Some preliminary evidence suggests that there may be
some connection. For example, compulsive gamblers and
kleptomaniacs respond to drugs such as naltrexone which
block the brains ability to experience euphoric states; com-
pulsive gamblers and bulimics experience sudden relapse even
after many years of abstinence. See Holden (2001a) for a
discussion of recent research concerning the commonalities
between various behavioral pathologies and substance
addiction.
1582 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
Section V D and referenced in Propositions 5
and 6.
Let g denote generation, t denote age, and m
denote time. Members of generation g are born
in period m g, and reach age t in period m
g t. Let denote the pure rate of time pref-
erence, and let denote the constant single-
period survival probability, so that . The
size of each new generation at age t 0 is (1
) N, where N is the constant size of the total
population.
Let
m
denote the taxes/subsidies applied in
period m, including either a tax on the addictive
substance,
m
, or a tax on rehabilitation,
m
, as
well as age-specic lump-sum instruments, T
tm
.
The period-m policy determines tax-inclusive
prices and incomes, from which we can com-
pute (as described in Section III), for each gen-
eration g m, a parameter vector
g,t
(
1
g,t
, ... ,
S
g,t
) with t m g applicable in
period m. An intertemporal policy assigns a
policy
m
to each period m, and induces, for
each generation, an innite sequence of param-
eter vectors,
g
(
g0
,
g1
, ...). Since
gt
can
vary over t (in contrast to the case treated in
sections III and IV), we must allow choice to
vary with age as well as the addictive state. A
decision rule maps age t and state s into a
probability distribution over {(E, 1), (E, 0), ( A,
0), (R, 0)} (note that we allow for randomiza-
tions) and implies a probability
s
t
() of use in
state s at age t. We use
g
to denote the decision
rule of generation g. The optimized value func-
tion V
s
t
(
g
) depends on the particular sequence
of parameters confronted by generation g, and
varies with age t. Since decisions are discrete,
an optimal decision rule need not be unique and,
indeed, is denitely not unique when it involves
randomizations.
The optimized usage probabilities generate a
state-transition probability matrix
t
(
g
). For a
large population of DMs starting in state 0 at
age 0 and following decision rule
g
, the pop-
ulation distribution across addictive states at
age t is z
t
(
g
) [
k0
t 1
k
(
g
)] z
0
, where z
0
is
an S-dimensional vector with a 1 in the rst
position and zeros elsewhere.
We say that an intertemporal policy is
feasible if there is, for each generation g, some
decision rule
g
solving the DMs choice prob-
lem given
g
induced by , such that the
governments budget is balanced in every pe-
riod. A steady-state policy prescribes a
constant tax rate, either or , and constant
age-specic lump-sum taxes, T
t
. Each genera-
tion faces the same sequence of parameters,
(
0
,
1
, ...), and V
0
0
() is the lifetime
discounted expected hedonic payoff for the rep-
resentative individual.
APPENDIX B: PROOFS
Here we prove Propositions 3 and 4, and
sketch the proofs of Propositions 1, 2, 5, and 6
to conserve space. Complete proofs are avail-
able on the AERs web site.
SKETCH OF PROOF FOR PROPOSITION 1:
Sketch for parts (i-a) and (i-b).The proof
involves three steps.
Step 1: Consider and such that: (1)
k
k
, (2)
i
i
for i k, and (3) V
s
()
V
s
() for all s. Then (a) for all j k, V
j
()
V
j
() V
j 1
() V
j 1
(), and (b) for all
j k, V
j
() V
j
() V
j 1
() V
j 1
().
The argument, omitted, involves induction
starting with j 1 for part (a), and with j S
for part (b).
Step 2: Consider and such that: (1)
k
k
, (2)
i
i
for i k, and (3) V
s
()
V
s
() for all s. Then (a) for j k, the disposi-
tion to use in state j is weakly higher with
than with , and (b) for j k, the disposition to
use in state j is weakly lower with than with
. These conclusions follow from step 1, which
implies that, for j k (j k), the difference in
continuation values following abstention and
use, and hence the disincentive to use, is weakly
greater (smaller) with than with .
Step 3: It is easy to verify that V
s
() is
weakly increasing in u
k
a
and b
k
a
, and weakly
decreasing in p
k
a
. Combining this with step 2
completes the proof of parts (i-a) and (i-b).
Sketch for part (i-c).Consider two param-
eter vectors, and
, such that b
j
E
b
j
E
with all
other components equal, or p
j
E
p
j
E
with all
other components equal. We argue, in two
steps, that the disposition to use in state j is
weakly higher with
than with .
Step 1: (a) If (E, 1) is optimal in state j with
, then it is optimal in state j with
, and (b) if
(E, 1) is the unique optimal choice in state j with
, then it is the unique optimal choice in state j
1583 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
with
. When p
j
E
p
j
E
and all other components
of and
j
E
b
j
E
and
all other components of and
j
E
b
j
E
. If the DM weakly
prefers use to abstention with , he must there-
fore strictly prefer it with
.
Step 2: (a) If neither (E, 1) nor (E, 0) are
optimal choices in state j for
and
; (b) if either ( A, 0) or (R, 0) is optimal in state
j with
), continues to
satisfy the valuation equation (1) when the pa-
rameter vector is changed from to
.
From part (a) of step 1 and part (a) of step 2,
the maximum disposition to use in state j is
weakly greater with
than with .
Now consider two parameter vectors, and
, such that u
j
R
u
j
R
with all other components
equal. We claim that if something other than (R,
0) is optimal in state j with , then it is also
optimal in state j with
.
Sketch for part (ii).Suppose
coincides
with except for p
j
E
, p
j
A
, u
j
A
, u
j
R
, and/or b
j
A
(subject to the restrictions imposed by Assump-
tions 1 and 2). We claim that, if (E, 1) is optimal
in state j for , it is also optimal in state j for
;
moreover, if (E, 1) is the unique optimum in the
rst instance, it is also the unique optimum in
the second instance. Analogously to step 2 of
part (i-c), these conclusions follow from the
easily veried fact that the same value function,
V
s
( ), continues to satisfy the valuation equa-
tion (1) when the parameter vector is changed
from to
with
p
s
E
p
s
E
for all
s (and all other parameters xed). With , let s
3
denote the rst state (working backward from S)
in which (E, 1) is not an optimal choice; this
denes the longest possible resignation interval.
Let s
0
3
s
3
denote the rst state (working back-
ward from S) in which something other than (E,
1) is an optimal choice; this denes the shortest
possible resignation interval. Remember that s
3
and s
0
3
may differ because the optimal choice in
each state is not necessarily unique. Consider
moving from to
p
s
E
to p
s
E
for s s
3
. Since (E, 1) is initially
optimal for all such states, this leaves all opti-
mal choices unchanged [Proposition 1, part (ii),
coupled with the observation that, when (E, 1) is
optimal, neither ( A, 0) nor (R, 0) is ever opti-
mal]. (2) Change from
p
s
E
to p
s
E
for s s
3
. This
weakly increases the disposition to use in states
s
3
1 through S [Proposition 1, part (i-b)].
Thus, the disposition to use in all states s s
3
is weakly lower with than with
. It follows
that (E, 1) continues to be an optimal choice in
states s s
3
with
than
with . Since nothing other than (E, 1) is opti-
mal in states s s
0
3
with , nothing other than
(E, 1) can be optimal in states s s
0
3
with
, so
the minimum nal resignation interval is
weakly longer with
than with .
PROOF OF PROPOSITION 3:
Select any state s. We can decompose the
change from
to derived from w
s
(e, x, a)
w
s
(e, x, a) d
s
, and (ii) a change from to .
The rst change reduces u
s
a
by d
s
for all states
s and actions a. This is simply a renormaliza-
tion, and has no effect on choices. The second
change weakly increases u
s
a
by d
s
d
s
for all
1584 THE AMERICAN ECONOMIC REVIEW DECEMBER 2004
s s, which weakly reduces the disposition to
use in state s by Proposition 1 part (i-b), and
weakly decreases u
s
a
by d
s
d
s
for all s s,
which also weakly reduces the disposition to
use in state s by Proposition 1 part (i-a). Thus,
the disposition to use in state s weakly
decreases.
PROOF OF PROPOSITION 4:
Part (i).Consider some parameter vector
,
and let denote the parameter vector obtained by
setting
p
s
a
0 for all a and s, leaving all other
elements of
max1 p
s
E
u
s
E
V
max1,s 1
p
s
E
u
s
E
b
s
E
V
minS,s 1
, 1 p
s
A
u
s
A
V
max1,s 1
p
s
A
u
s
A
b
s
A
V
minS,s 1
, u
j
R
V
max1,s 1
}.
Since (E, 1) is not a best choice in s, the rst
term in braces is strictly less than u
s
E
V
max{1,s1}
(
) u
s
E
V
max{1,s1}
(
V
max{1,s1}
(
), so V
s
( ) V
s
(
).
SKETCH OF PROOF FOR PROPOSITION 5:
Without loss of generality, we can proceed as if,
for the optimal policy, the net transfer to each
cohort is zero in each period. If this is not the case,
simply redene income in state s at age t as y
st
y
s
L
t
, where L
t
is the net transfer received at age
t; the original policy remains optimal.
Sketch for part (i).We prove this in two steps.
Step 1: An optimal tax rate must be weakly
negative. To prove this, we assume that there is
a strictly positive optimal tax rate and establish
a contradiction by showing that this policy must
be strictly inferior to (the no-tax policy).
Consider a decision rule (where we drop
the generational superscript g because we are
examining steady states) that is optimal and
satises budget balance with the optimal policy,
and any age t at which neither use nor nonuse
is a certainty from the perspective of period 0
(under the stated assumptions, there is always at
least one such age). Now suppose that policy
prevails but that the DM nevertheless continues
to follow . Through a series of algebraic steps,
one can show that E
0
[u(e
t
)x
t
1]
E
0
[u(e
t
)x
t
0]. That is, the expectation, as of
age zero, of the marginal utility of nonaddictive
consumption in t is greater when conditioned
on use than when conditioned on nonuse. Intu-
itively, use tends to occur when income is
lower, and it also entails a cost.
Suppose we switch from the optimal policy
to . Assume for the moment the DM continues
to follow . From the perspective of age 0, the
result is an actuarially fair redistribution across
age-t realizations of (s, ), from realizations
in which the DM does not use the substance
to realizations in which he does. Since
E
0
[u(e
t
)x
t
1] E
0
[u(e
t
)x
t
0] for the last
dollar redistributed, and since u is strictly con-
cave, the transfer makes him strictly better off.
Thus, his discounted expected hedonic payoff
weakly increases for every age t and strictly
increases for some. Reoptimizing the decision
rule given reinforces this conclusion.
Step 2: The no-tax policy, , is not optimal.
Intuitively, for the same reasons as in step 1, a
small subsidy coupled with lump-sum transfers
that achieve budget balance within each cohort
and period should generate a rst-order welfare
improvement by creating an actuarially fair re-
distribution from realizations in which the DM
does not use the substance to realizations in
which he does. Formally, this reasoning en-
counters two technical issues. First, we must
establish that policies with small tax rates and
budget balance within each cohort and period
are feasible. Allowing for randomized choices,
this is accomplished through standard argu-
ments and a routine application of the Kakutani
Fixed Point Theorem. Second, any such redis-
1585 VOL. 94 NO. 5 BERNHEIM AND RANGEL: ADDICTION
tribution must be actuarially fair relative to
probabilities associated with a decision rule that
is optimal with the new policy, not with .
To deal with this second issue, we consider a
sequence of tax rates, associated age-specic
lump-sum taxes, and optimal decision rules
with budget balance within each cohort and
period, (
j
, T
j
,
j
), with
j
, T
j
3 0, and
j
converging to some limit
. By standard argu-
ments,
t
denote the
likelihood of use at age t with
. Fixing the
choice rule at
t
from a cohort of age .
Distributing this back to the same cohort as a
lump sum, and taking the derivative of the ex-
pected age-t payoff with respect to , we obtain
(1 b
t
)b
t
(E
0
[u(e
t
)x
t
0] E
0
[u(e
t
)x
t
t
is 0 or 1, and by the
same arguments as in step 1, is strictly positive
for intermediate values. For large j,
j
is arbi-
trarily close to
, a switch from to (
j
, T
j
) creates a
redistribution that is almost actuarially fair for
the probabilities implicit in
. We therefore
know that redistribution is almost neutral for t
such that b
t
{0, 1}, and strictly benecial for
t such that b
t
(0, 1). Accordingly, there exists
j sufciently large such that the expected
present value of the DMs payoff is higher with
(
j
, T
j
) than with , assuming he chooses
.
Reoptimizing for (
j
, T
j
) reinforces this
conclusion.
Sketch for part (ii).The argument parallels
that given for part (i), except we use the fact that
E
0
[u(e
t
)x
t
0] E
0
[u(e
t
)x
t
1] when q is
sufciently small.
SKETCH OF PROOF FOR PROPOSITION 6:
First consider small subsidies. The argument
generally parallels step 2 of the sketch for Prop-
osition 5, part (i). Take any sequence of reha-
bilitation tax rates, associated age-specic
lump-sum taxes, and optimal decision rules
with budget balance within each cohort and
period, (
j
, T
j
,
j
), with
j
0,
j
3 0, T
j
3
0, and
j
converging to some limit
. Let B
t
denote the likelihood of rehabilitation at age t
with
, a small tax
generates per capita revenue B
t
from
a cohort of age t. Distributing this back to
the same cohort as a lump sum, and taking
the derivative of the expected age-t payoff
with respect to , we obtain (1
B
t
) B
t
(E
0
[u(e
t
)a
t
R] E
0
[u(e
t
)a
t
R]).
This equals zero when B
t
{0, 1}, and it is
strictly negative for B
t
(0, 1) (under the
conditions stated in the proposition, the DM
chooses R only in the highest state reached with
positive probability in t; rehabilitation therefore
occurs when income is lower, and it entails a
cost greater than q, so the expected marginal
utility of nonaddictive consumption must be
greater when conditioned on rehabilitation than
when conditioned on no rehabilitation).
We evaluate the change from (,
) to (
j
,
T
j
,
j
) in three steps. First, change the hot-mode
probabilities to those prevailing under (
j
, T
j
),
leaving everything else constant. Second,
change the policy from to (
j
, T
j
), still hold-
ing the choice rule xed at
. Third, reopti-
mize, changing the choice rule to
j
. The third
change is obviously weakly benecial, as is the
rst (with
j
0, the lump-sum transfers are
negative, so, under Assumption 3, the hot-mode
probabilities fall). Now consider the second
step. For large j, (
j
, T
j
,
j
) is arbitrarily close to
(,
, a switch from to (
j
, T
j
) creates a
redistribution that is almost actuarially fair for
the probabilities implicit in (,
). Thus,
j
(1 B
t
) B
t
(E
0
[u(e
t
)a
t
0] E
0
[u(e
t
)a
t
.
We evaluate the elimination of a small tax in
three steps. First, eliminate the tax (and associ-
ated lump-sum transfers) without changing the
hot-mode probabilities, and keeping the choice
rule xed at