in press, Personality and Social Psychology Bulletin 1
The Drunk Utilitarian Revisited:
Does Alcohol Really Increase Utilitarianism in Moral Judgment?
University of Silesia in Katowice, Poland
University of Silesia in Katowice, Poland
Jim A.C. Everett
University of Kent, United Kingdom
University of Wroclaw, Poland
University of Texas at Austin, USA
The “drunk utilitarian” phenomenon suggests that people are more likely to accept harm for the greater good when they are
under the influence of alcohol. This phenomenon conflicts with the ideas that (1) acceptance of pro-sacrificial harm requires
inhibitory control of automatic emotional responses to the idea of causing harm and (2) alcohol impairs inhibitory control.
The current preregistered experiment aimed to provide deeper insights into the effects of alcohol on moral judgments by
using a formal modeling approach to disentangle three factors in moral dilemma judgments and by distinguishing between
instrumental harm and impartial beneficence as two distinct dimensions of utilitarian psychology. Despite the use of a
substantially larger sample and higher doses of alcohol compared to the ones in prior studies, alcohol had no significant
effect on moral judgments. The results pose a challenge to the idea that alcohol increases utilitarianism in moral judgments.
Keywords: alcohol, CNI model, moral dilemmas, moral judgment, utilitarianism
Most people would probably agree that physically
harming people is wrong; but what if doing so (e.g., by
torturing someone) would save the lives of ten
innocent people? Such moral dilemmas are often used
to highlight the tension between deontology and
consequentialism. Utilitarianism, as a particular type
of consequentialism, posits that moral decisions
should be guided solely by what brings about the best
consequences and so, if torturing someone saves lives,
it can be acceptable (e.g., Mill, 1863). In contrast,
deontological ethical theories posit that morality is
about more than just consequences, but should be
guided by rights, duties, and obligations (e.g., Kant,
1916), and therefore often assume that harming
someone is morally wrong.
In the current research, we investigated how
alcohol influences moral judgments when adhering to
“deontological” moral norms conflicts with the
“utilitarian” maximization of outcomes for the greater
good. According to the dual-process model of moral
dilemma judgments, “utilitarian” acceptance of harm
requires inhibitory control of automatic emotional
responses to the idea of causing harm (Greene et al.,
2001, 2004). In line with these assumptions,
acceptance of harm for the greater good has been
found to decrease under conditions assumed to impair
inhibitory control, including time pressure and
cognitive load (Białek & De Neys, 2017; Greene et al.,
2008; Suter & Hertwig, 2011). However, one
intriguing finding that seems rather puzzling in light of
these results is the phenomenon of the “drunk
utilitarian” (Duke & Bègue, 2015). This phenomenon
suggests that people are more likely to accept harm for
the greater good when they are under the influence of
alcohol, which seems difficult to reconcile with the
ideas that (1) acceptance of harm for the greater good
requires inhibitory control of automatic emotional
responses to the idea of causing harm (Greene et al.,
2001, 2004) and (2) alcohol impairs inhibitory control
(Day et al., 2015; Noël et al., 2010; Weafer &
The main goal of the current research was to
provide deeper insights into the effect of alcohol on
utilitarian moral judgment by using a formal modeling
approach to disentangle sensitivity to consequences,
sensitivity to moral norms, and general action
tendencies in responses to moral dilemmas
(Gawronski et al., 2017) and by distinguishing
between two dimensions of utilitarian psychology
called impartial beneficence and instrumental harm
(Everett & Kahane, 2020; Kahane et al., 2018).
Alcohol and Moral Judgment
The “drunk utilitarian” phenomenon was
discovered by Duke and Bègue (2015) in two studies
that investigated associations between moral
judgments and blood alcohol concentration measured
with a breathalyzer in patrons of French bars. In the
first study, participants were asked to imagine two
scenarios in which a runaway trolley is approaching a
group of five workers. In a scenario called the switch
in press, Personality and Social Psychology Bulletin 2
dilemma, participants were asked if they would pull a
switch to redirect the trolley to a different track where
it would kill one person and save the lives of the five
workers (Foot, 1967). In another scenario called the
footbridge dilemma, participants were asked if they
would push a man from a bridge, which would kill the
man but save the five workers (Thomson, 1976).
Blood alcohol concentration showed a significant
positive correlation with pro-sacrificial judgments in
the footbridge dilemma (r = .29, p = .023) but not the
switch dilemma (r = .17, p = .17). The second study
replicated the correlation between blood alcohol and
pro-sacrificial judgments in the footbridge dilemma (r
= .32, p = .039), further showing that the obtained
relation was not driven by self-reported behavioral
disinhibition or elevated positive mood.
To rule out ambiguities in Duke and Bègue’s
(2015) correlational findings, Arutyunova and
colleagues (2017) experimentally manipulated blood
alcohol levels in a community sample of volunteer
participants in a two-session study. In both sessions,
participants were asked to respond to a longer battery
of moral dilemmas that included the switch and the
footbridge dilemma (Cushman et al., 2006). In one of
the two sessions, participants responded to the moral
dilemmas under the influence of alcohol (i.e., 42
minutes after drinking juice mixed with vodka; 1 g of
alcohol at 40% strength for each 1 kg of body weight).
In the other session, they completed the same
dilemmas while being sober (i.e., after drinking juice
mixed with water). Results did not reveal any
significant difference in moral judgments across the
Francis and colleagues (2019) investigated effects
of alcohol on moral dilemma judgments among
psychology students who were randomly assigned to
one of three experimental conditions: (1) placebo, (2)
low-intoxication level (0.4 g of alcohol at 37.5%
strength for each 1 kg of body weight), and (3) high
intoxication-level (0.8 g of alcohol at 37.5% strength
for each 1 kg of body weight). Twenty minutes after
drinking either plain juice or juice mixed with different
amounts of vodka, participants responded to a battery
of moral dilemmas including the footbridge dilemma
(Greene et al., 2001). Similar to Arutyunova et al.’s
(2017) findings, moral judgments did not significantly
differ across experimental conditions.
The Current Research
In the current preregistered experiment, we aimed
to address five limitations of prior research on the
effects of alcohol on moral judgment: (1) relatively
low (or inconsistent) levels of alcohol consumption;
(2) lack of a placebo condition in some studies; (3)
small sample sizes; (4) confounds in the measurement
of moral dilemma judgments; and (5) generalization to
utilitarian judgments writ large based on responses to
First, because Duke and Bègue’s (2015) study
used a correlational design, the obtained relations
between blood alcohol concentration and moral
judgment may not necessarily reflect a causal effect of
alcohol. Subsequent laboratory experiments
(Arutyunova et al., 2017; Francis et al., 2019) aimed
to address this concern, and these studies did not find
any significant effects of alcohol on moral judgments.
However, the consumed doses of alcohol were
relatively small in these experiments. To illustrate this
concern: with the experimentally induced levels of
blood alcohol in permille, participants would have still
been allowed to drive in many countries (< 0.5‰).
Thus, it is possible that the lack of a significant effect
of alcohol in experimental studies is due to the
relatively low levels of alcohol, not lack of a causal
effect. In the current study, we aimed to address this
issue by administering a comparatively higher dose of
alcohol than previous experiments.
Second, only one of the three studies (Francis et
al., 2019) included a placebo condition in which
participants believed they were consuming alcohol
without actually consuming alcohol. Placebo
conditions are essential in this line of research to
distinguish actual effects of alcohol from effects of
people’s naïve beliefs about the influence of alcohol
(Bodnár et al., 2020). In the context of moral
judgments, it is also possible that people believe that
being intoxicated gives them a license to make more
pro-sacrificial judgments, even when alcohol itself has
no causal effect on moral judgments. In the current
study, we aimed to address this issue by using three
experimental conditions: (1) alcohol, (2) no-alcohol
control, (3) placebo control.
Third, the sample sizes in prior studies were rather
small overall, with N = 61 and N = 42 in the two studies
by Duke and Bègue (2015), N = 40 in the study by
Arutyunova et al. (2017), and N = 48 in the study by
Francis et al. (2019). Because small sample sizes can
lead to both false negatives (Maxwell et al., 2015) and
false positives (Button et al., 2013), evidence from a
larger sample would be helpful to reconcile the
conflicting findings in previous studies. In the current
experiment, we aimed to address this issue by
recruiting a relatively large sample of 300 participants.
Fourth, all three studies relied on the traditional
approach of using moral dilemmas that pit
“characteristically utilitarian” against
“characteristically deontological” options (Conway et
al., 2018). A major disadvantage of this approach is
that it includes two confounds in the measurement of
moral dilemma judgments. First, endorsement of the
“utilitarian” option requires rejection of the
“deontological” option, and vice versa. This approach
in press, Personality and Social Psychology Bulletin 3
confounds the measurement of utilitarian and
deontological tendencies underlying moral judgments,
which conflicts with the idea that the processes
underlying the two kinds of tendencies are
independent (Conway & Gawronski, 2013). Second,
“utilitarian” judgments are conflated with action (i.e.,
pulling the lever, pushing the man) while
“deontological” judgments are conflated with inaction
(i.e., not pulling the lever, not pushing the man),
confounding the two moral tendencies with general
action tendencies (Crone & Laham, 2017). These
considerations suggest that differences in responses to
traditional moral dilemmas (e.g., switch dilemma,
footbridge dilemma) may reflect either (1) differences
in outcome maximization, or (2) differences in
adherence to moral norms, or (3) differences in general
action tendencies (or any combination of the three). In
the current research, we aimed to disentangle these
three factors by using a mathematical model called the
CNI model to quantify (1) sensitivity to consequences,
(2) sensitivity to moral norms, and (3) general
preference for inaction versus action in responses to
moral dilemmas (Gawronski et al., 2017). The CNI
model disentangles these three factors by comparing
responses to four kinds of moral dilemmas that differ
in terms of (1) whether the benefits of the described
action are greater or smaller than the costs and (2)
whether the described action is prohibited or
prescribed by a moral norm. Disentangling the three
factors underlying moral dilemma judgments may
prove helpful for understanding the “drunk utilitarian”
phenomenon, in that alcohol seems unlikely to
increase sensitivity to consequences in a utilitarian
sense. Instead, it seems more likely that alcohol either
(1) reduces sensitivity to moral norms in a
deontological sense or (2) increases people’s
willingness to perform a focal action regardless of its
consequences and its consistency with moral norms. In
fact, if either of the latter two effects is sufficiently
large, they may conceal a simultaneous decrease in
sensitivity to consequences, which would suggest that
alcohol might reduce rather than increase utilitarian
concerns about the greater good (for an example, see
Luke & Gawronski, 2021).
Fifth, while work on the “drunk utilitarian”
phenomenon—along with work on the dual-process
model—has been used to draw conclusions about
utilitarian judgment in general, pro-sacrificial
judgments are just one part of utilitarian psychology.
According to the two-dimensional model of utilitarian
psychology (Everett & Kahane, 2020; Kahane et al.,
2018), utilitarianism has two dimensions that are
conceptually and psychologically distinct.
Instrumental harm (IH) captures willingness to cause
harm to achieve positive consequences for the greater
good. Impartial beneficence (IB) taps the extent to
which people endorse the radically demanding and
impartial helping utilitarianism requires. Different
from previous work inferring utilitarianism from
responses to sacrificial dilemmas, research guided by
the two-dimensional model of utilitarian psychology
infers endorsement of IH and IB from participants’
agreement with broad ethical statements about key
ideas of the two dimensions (Kahane et al., 2018).
Previous work using this approach has shown that the
two dimensions of utilitarianism show different
patterns of correlations with individual-difference
measures (Kahane et al., 2018), are affected differently
by priming manipulations (Capraro et al., 2019), and
have distinct consequences for social perception
(Everett et al., 2018, 2021). Moreover, although
endorsement of pro-sacrificial harm in moral
dilemmas has been found to be positively correlated
with IH, moral dilemma judgments were found to be
unrelated to IB (Kahane et al., 2018). Thus, based on
the known effects of alcohol, it seems possible that
alcohol increases the endorsement of IH. However, it
seems rather implausible that alcohol would increase
endorsement of IB.
In sum, the current study aimed to address the five
identified limitations in a preregistered lab experiment
testing the effects of a comparatively higher dose of
alcohol (1.6 g of alcohol at 40% strength for each 1 kg
of body weight) on “utilitarian” preferences. The study
included three types of measures: (1) traditional
sacrificial moral dilemmas (Foot, 1967; Thomson,
1976), (2) a battery of moral dilemmas for research
using CNI model (Körner et al., 2020), and (3) the
Oxford Utilitarianism Scale (OUS) measuring IH and
IB (Kahane et al., 2018). Participants were randomly
assigned to one of three experimental conditions: (1)
alcohol, (2) no-alcohol, (3) placebo. To obtain greater
statistical power than prior studies, we aimed for a
sample of 300 participants (100 per condition).
For responses to the two variants of the trolley
problem, we expected to obtain effects that correspond
to Duke and Bègue’s (2015) correlational findings. For
responses to the footbridge dilemma, we predicted
greater pro-sacrificial responding in the alcohol
condition compared to the no-alcohol and placebo
conditions. For responses to the switch dilemma, pro-
sacrificial responding was not expected to differ across
For the three factors captured by the CNI model,
we predicted that alcohol, compared to no-alcohol and
placebo conditions, would (1) decrease sensitivity to
consequences, (2) decrease sensitivity to moral norms,
and (3) decrease general preference for inaction over
Finally, for the IH dimension of the OUS, we
predicted that alcohol would increase scores compared
in press, Personality and Social Psychology Bulletin 4
to no-alcohol and placebo conditions. For the IB
dimension, scores were not expected to differ across
The study was approved by the Ethics Committee
of the University of Silesia. The preregistration, data,
analysis codes, and study materials are available at
Our target sample size was 300 participants after
preregistered exclusions, 100 (~50% female) per
experimental condition. For the predicted interaction
between experimental condition and type of trolley
problem, a sample of 300 provides 80% power in
detecting a small effect of f = 0.107, assuming a
correlation of r = .30 between measures and
nonsphericity correction of ε = 1. The same is true for
detecting the predicted interaction between
experimental condition and type of OUS subscale. For
the three parameters of the CNI model, a sample of 300
provides 80% power in detecting a small effect of f =
0.097, assuming a correlation of r = .30 between
measures and nonsphericity correction of ε = 1.
Participants were recruited through
advertisements in various media (e.g., university
websites, Facebook, newspapers). Individuals with
health problems, who were pregnant, who reported
alcohol addiction, or were younger than 18 years
before the laboratory invitation were not eligible for
participation. To verify these criteria, all individuals
who responded to the advertisements completed an
online screening questionnaire before receiving an
invitation to the lab study. Of the 1079 volunteers who
completed the screening survey, 387 met the criteria
and were invited to the laboratory (198 women, 189
men; Mage = 25.7, SDage = 7.4; age range: 18 to 52
years). All of the invited volunteers accepted the
invitation and participated in the study voluntarily
without monetary compensation. Participants were
asked to refrain from drinking alcohol for 24 hours,
taking any medication (e.g., painkillers) for 10 hours,
and from eating for at least 3 hours before coming to
the laboratory. Following our preregistered exclusion
criteria, data from 58 participants were excluded from
analyses because they failed to pass one or more of our
attention checks. The final sample included 329
participants, whose age ranged from 18 to 52 years (M
= 25.1, SD = 6.2). Of the 329 participants in the final
sample, 106 participants were in alcohol condition (53
women, 53 men), 114 in placebo control condition (57
women, 57 men), and 109 in the no-alcohol control
condition (53 women, 56 men). Following our
preregistered stopping rule, we ended the data
collection on the day we reached our target sample of
300 participants, but included the data from all
participants who had an appointment on the same day.
This led to an excess of 29 participants beyond our
target sample of 300 participants. All future
appointments after the day of completion were
canceled in line with our preregistered stopping rule.
Trolley problems. Participants were presented
with the switch dilemma (Foot, 1967) and the
footbridge dilemma (Thomson, 1976) and asked to
indicate whether they would perform the described
action on 7-point rating scales. The end-points were
labeled “I would definitely do nothing” (1) and “I
would definitely pull the level” (7) for the switch
dilemma, and “I would definitely do nothing” (1) and
“I would definitely push the man onto the track” (7)
for the footbridge dilemma.
CNI dilemmas. Participants were asked to
respond to a validated battery of 48 moral dilemmas
for research using the CNI model (Körner et al., 2020).
The battery included four variants of 12 basic
dilemmas, varying as a function of (1) whether the
benefits of the described action are greater or smaller
than the costs and (2) whether the described action is
prohibited or prescribed by a moral norm. Participants
were asked if they would perform the described action.
Responses were measured with dichotomous yes vs.
no response options. Using the CNI model template
files provided by Körner et al. (2020), the total
numbers of yes vs. no responses on each type of
dilemma were used to estimate three scores for each
participant via multinomial modeling (Hütter &
Klauer, 2016): a score reflecting sensitivity to
consequences (C parameter); a score reflecting
sensitivity to moral norms (N parameter); and a score
reflecting general preference for inaction versus action
(I parameter). Toward this end, the CNI model was
fitted to the data for each participant following the
procedures by Körner et al. (2020). CNI parameter
estimations were conducted with the freeware
multiTree (Moshagen, 2010), using random start
values, two replications, and a maximum of 90,000
OUS. Dimensions of utilitarianism were
measured using the OUS (Kahane et al., 2018). The IB
subscale includes five items measuring the extent to
which people endorse the utilitarian demand for
impartial helping (e.g., “It is morally wrong to keep
money that one doesn’t really need if one can donate
it to causes that provide effective help to those who will
benefit a great deal”). The IH subscale includes four
items measuring willingness to cause harm to achieve
positive consequences for the greater good (e.g., “It is
morally right to harm an innocent person if harming
them is a necessary means to helping several other
innocent people”). Participants were asked to indicate
how much they agree with each statement, using 7-
in press, Personality and Social Psychology Bulletin 5
point rating scales ranging from 1 (strongly disagree)
to 7 (strongly agree).
CRT. For exploratory purposes, the study also
included Primi et al.’s (2016) modified version of the
Cognitive Reflection Test (Frederick, 2005). The CRT
was included to identify potential effects of alcohol on
cognitive reflection and to explore whether effects of
alcohol on moral judgments are mediated by
differences in cognitive reflection.
Three research assistants were responsible for
different tasks during a given session. The first
assistant (informally referred to as policeman) was
responsible for measuring participants’ weight and
taking breathalyzer measurements. The second
assistant (informally referred to as bartender) was
responsible for preparing the drinks and the
randomized assignment to the three experimental
conditions. The third assistant (informally referred to
as courier) was responsible for ensuring that all
documents are signed before the study and for serving
the drinks (being unaware of the experimental
Participants in the no-alcohol condition consumed
a drink that included only juice and no alcohol.
Participants in this condition were told that there was
no alcohol in their drink. Participants in the placebo
condition were told that there was alcohol in the drink
and consumed a drink that included only juice and no
alcohol, but was sprayed with alcohol to create the
impression of alcohol consumption. Participants in the
alcohol condition consumed an alcoholic drink that
was prepared to contain 1.6 grams of alcohol at 40%
strength for each 1 kg of the participant’s body weight.
The drink was mixed with the same juice as in other
conditions. After the study, participants in the alcohol
condition had to wait to become sober or return home
with a sober driver.
Figure 1 presents an overview of the procedure.
When participants arrived in the laboratory, they
provided informed consent, had their weight and blood
alcohol measured using a breathalyzer, and then
completed a demographic survey that included
questions about participants’ age, marital status,
employment status, religion, political views,
subjectively perceived social status, and COVID
diagnoses for themselves close family and friends.
Next, participants consumed their assigned drink (up
to 10 minutes), after which they watched two
emotionally neutral movie clips comprising a period of
Because conducting a high-powered laboratory study on the effects
of alcohol requires a considerable amount of resources, we aimed to
maximize the utility of the invested resources by including several
survey instruments for a different project at the end. These
instruments included measures of self-concept (Stake, 1994), moral
identity (Aquino & Reed, 2002), personality (Gosling et al., 2003),
51 minutes to allow for alcohol absorption: (1) The
World From Above, season 10, episode 6 titled Iceland
- From Vatnajokull National Park to Gullfoss
Waterfall and (2) The World From Above, season 4,
episode 7 titled Yellowstone National Park. After the
movie, blood alcohol levels were measured a second
time (using the same sound signal in the placebo and
experimental groups). Next, participants completed
the main dependent measures. Participants first
completed the CNI dilemma battery and the OUS, with
the order of the two instruments being
counterbalanced across participants. Both the CNI
dilemmas and the items of the OUS were presented in
a fixed randomized order that was held constant for all
participants. The two measures were followed by the
two versions of the trolley problem, with their order
being counterbalanced independent of the order of the
CNI battery and the OUS. Finally, participants
completed the CRT and several supplementary
measures that were unrelated to the primary purpose
of this study.
The study concluded with a debriefing
and third measurement of blood alcohol. After the
debriefing, participants were also asked to guess the
condition to which they had been assigned.
Descriptive statistics of all measured variables are
presented in Table 1. Correlations between all
measured variables are presented in Table 2.
Endorsement of pro-sacrificial harm in the switch
dilemma showed significant positive correlations with
endorsement of pro-sacrificial harm in the footbridge
dilemma, the CNI model’s C parameter, and the IH
subscale of the OUS, as well as significant negative
correlations with the CNI model’s N parameter and
CRT scores. These results suggest that stronger
endorsement of pro-sacrificial harm in the switch
dilemma was associated with (1) stronger endorsement
of pro-sacrificial harm in the footbridge dilemma, (2)
stronger sensitivity to consequences, (3) stronger
endorsement of IH, (4) weaker sensitivity to moral
norms, and (5) weaker cognitive reflection.
Endorsement of pro-sacrificial harm in the footbridge
dilemma showed significant positive correlations with
the CNI model’s C parameter and the IH subscale of
the OUS, as well as a significant negative correlation
with the CNI model’s N parameter. These results
suggest that stronger endorsement of pro-sacrificial
harm in the footbridge dilemma was associated with
(1) stronger sensitivity to consequences, (2) stronger
and moral foundations (Graham & Haidt, 2012). Because these
supplementary measures were not intended for the current study and
the measures were administered at the end, the preregistration did
not include any of these instruments. Separate preregistrations were
submitted for the analyses of data obtained with these
in press, Personality and Social Psychology Bulletin 6
endorsement of IH, and (3) weaker sensitivity to moral
norms. Beyond these correlations, the C parameter
showed a significant negative correlation with the N
parameter, and significant positive correlations with
the I parameter, the IH subscale of the OUS, and CRT
scores. These results suggest that stronger sensitivity
to consequences was associated with (1) weaker
sensitivity to moral norms, (2) stronger action
aversion, (3) stronger endorsement of IH, and (4)
stronger cognitive reflection. Moreover, the N
parameter showed a significant positive correlation
with the I parameter and a significant negative
correlation with the IH subscale of the OUS,
suggesting that stronger sensitivity to moral norms
was associated with (1) weaker action aversion and (2)
weaker endorsement of IH. Finally, the IH subscale of
the OUS showed a significant positive correlation with
the IB subscale and CRT scores, suggesting that
stronger endorsement of IH was associated with (1)
stronger endorsement of IB and (2) stronger cognitive
To test the effectiveness of our alcohol
manipulation, blood alcohol levels measured with
breathalyzer were submitted to a 3 (Alcohol Group:
alcohol vs. no alcohol vs. placebo, between-subjects)
× 3 (Time: baseline vs. before survey vs. after survey,
within-subjects) mixed ANOVA, which revealed a
significant two-way interaction between Group and
Time, F(4, 652) = 872.00, p < .001, ηp2 = .842 (see
Table 1). Further analyses revealed that blood alcohol
levels significantly differed across the three groups
before the survey, F(2, 326) = 1248.75, p < .001, ηp2 =
.885, and after the survey, F(2, 326) = 2275.00, p <
.001, ηp2 = .933, but not at baseline, F(2, 326) = 0.00,
p = 1.00, ηp2 = .000. Before the survey, blood alcohol
levels were significantly higher in the alcohol group
compared to the no-alcohol group, t(326) = 43.20, p <
.001, d = 5.90, and compared to the placebo group,
t(326) = 43.70, p < .001, d = 5.90. Similarly, after the
survey, blood alcohol levels were significantly higher
in the alcohol group compared to the no-alcohol group,
t(326) = 58.30, p < .001, d = 7.96, and compared to the
placebo group, t(326) = 59.00, p < .001, d = 7.96.
Together, these results confirm the effectiveness of our
manipulation of blood alcohol.
Trolley problems. Responses to the two variants
of the trolley problem were submitted to a 3 (Alcohol
Group: alcohol vs. no alcohol vs. placebo, between-
subjects) × 2 (Dilemma Type: switch vs. footbridge,
within-subjects) mixed ANOVA. The ANOVA
revealed a significant main effect of Dilemma Type,
Note that the main effect of Parameter is uninterpretable, because
the neutral reference point of the I parameter (0.5) differs from the
indicating that participants were more willing to
redirect the trolley to a different track than to push a
man off the bridge, F(1, 326) = 332.02, p < .001, ηp2 =
.505 (see Table 1). Critically, there was no significant
main effect of Alcohol Group, F(2, 326) = 2.76, p =
.065, ηp2 = .017, and no significant interaction between
Alcohol Group and Dilemma Type, F(2, 326) = 1.57,
p = .210, ηp2 = .010.
CNI dilemmas. The three parameters of the CNI
model were submitted to a 3 (Alcohol Group: alcohol
vs. no alcohol vs. placebo, between-subjects) × 3
(Parameter: C vs. N vs. I, within-subjects) mixed
ANOVA (see Table 1).
Neither the main effect of
Alcohol Group, F(2, 326) = 1.24, p = .291, ηp2 = .008,
nor the interaction between Parameter and Alcohol
Group, F(4, 652) = 1.30, p = .269, ηp2 = .008, were
OUS. Responses on the OUS were submitted to a
3 (Alcohol Group: alcohol vs. no alcohol vs. placebo,
between-subjects) × 2 (Dimension: IH vs. IB, within-
subjects) mixed ANOVA (see Table 1). The main
effect of Alcohol Group was not significant, F(2, 326)
= 0.005, p = .995, ηp2 = .000, but the interaction
between Dimension and Alcohol Group was
statistically significant, F(2, 326) = 3.04, p = .049, ηp2
= .018. Descriptively, the placebo group showed
higher IH scores and lower IB scores compared to the
other two groups, but none of the relevant post-hoc
tests reached statistical significance (all ts < 1.35, all
ps > .193).
Guessed condition. Participants in the no-alcohol
condition were highly accurate in identifying the
condition to which they had been assigned (99.1%).
The same was true for participants in the alcohol
condition (96.3%). Accuracy was considerably lower
for participants in the placebo condition (65.8%), with
23.7% falsely believing that they had consumed
alcohol. Accuracy levels significantly differed across
the three groups, χ2(2) = 65.72, p < .001. To investigate
whether participants’ naïve beliefs about alcohol
consumption are associated with moral judgments, we
repeated the main analyses using “guessed alcohol
group” instead of “actual alcohol group” in the
ANOVA. There were no significant main or
interaction effects involving Guessed Alcohol Group
for responses to the trolley problems (all Fs < 2.51, all
ps > .083), the three CNI model parameters (all Fs <
1.54, all ps > .217), and the two dimensions of the OUS
(all Fs < 2.85, all ps > .059).
CRT. We performed a univariate ANOVA with
three conditions (alcohol vs. no alcohol vs. placebo)
on CRT scores, which revealed a significant difference
neutral reference point of the C and the N parameter (0), and N
scores are estimated in a manner that is conditional on C.
in press, Personality and Social Psychology Bulletin 7
across groups, F(2, 326) = 3.06, p = .048, ηp2 = .018.
A planned simple-contrast analysis indicated that this
effect was driven by higher CRT scores in the placebo
group (ΔCRT = 0.35), t(326) = 2.39, p = .017, and the
alcohol group (ΔCRT = 0.26), t(326) = 1.74, p = .083,
compared to the no-alcohol group (MCRT = 1.45).
These results suggest that believing one has consumed
alcohol led to improved performance on the CRT.
The “drunk utilitarian” phenomenon poses an
intriguing challenge to the dual-process model of
moral judgment, which suggests that alcohol-related
impairments in inhibitory control should reduce rather
than increase utilitarian judgments. However, because
the initial demonstration of the phenomenon was
based on correlational data (Duke & Bègue, 2015) and
subsequent experimental studies failed to obtain
significant effects of alcohol on moral dilemma
judgments (Arutyunova et al., 2017; Francis et al.,
2019), the reliability of the phenomenon is still
unclear. This state of affairs is exacerbated by several
limitations of prior research on the effects of alcohol
on moral judgment, which include (1) relatively low
(or inconsistent) levels of alcohol consumption; (2)
lack of a placebo condition in some studies; (3) small
sample sizes; (4) confounds in the measurement of
moral dilemma judgments; and (5) generalization to
utilitarian judgments writ large based on responses to
sacrificial dilemmas. To address these concerns, the
current preregistered experiment included a
manipulation of blood alcohol levels with
comparatively higher doses of alcohol and a placebo
condition to disentangle actual effects of alcohol from
effects of naïve beliefs about effects of alcohol. To
overcome the known problems associated with small
samples, the current study tested effects of alcohol
with a sample that was substantially larger compared
to prior studies. Finally, to overcome conceptual
limitations in the interpretation of pro-sacrificial
judgments in trolley problems, the current study used
the CNI model to disentangle different aspects of
moral dilemma judgments and the OUS to measure
different dimensions of utilitarian psychology.
Despite these improvements, we failed to obtain
any significant effect of alcohol on moral judgments.
Although our manipulation was highly effective in
influencing blood alcohol levels (measured with a
breathalyzer), there was no significant effect of
alcohol on pro-sacrificial judgments in the trolley
problem, the three parameters of the CNI model, and
only a weak, placebo-driven effect on the two
dimensions of utilitarian psychology captured by the
OUS. Moreover, although performance on the CRT
tended to be higher in the alcohol condition compared
to the no-alcohol condition, participants in the placebo
condition showed a similar performance boost,
suggesting that participants who believed that they
consumed alcohol invested extra efforts when
completing the CRT. Together, these results pose a
challenge to the “drunk utilitarian” phenomenon and
raise important questions about how alcohol may
influence moral judgments, if it has any such effect at
One possible explanation for the obtained null
effects is that the influence of alcohol on inhibitory
control might be more complex than commonly
assumed, given that effects of alcohol on inhibitory
control seem to be highly variable across tasks and
situations. Consistent with this concern, some studies
support the hypothesis that alcohol impairs inhibitory
control, while other studies report null effects of
alcohol on inhibitory control (e.g., Bartholow et al.,
2018). These mixed findings seem to be be partly
rooted in different conceptualizations of inhibitory
control and different approaches to measuring
inhibitory control. Although inhibitory control is
generally understood as the ability to suppress
attention, behavior, thoughts and/or emotions
(Diamond, 2013), inhibitory control is a multifaceted
construct that subsumes diverse aspects such as the
inhibition of prepotent response tendencies,
suppression of thoughts and memories, and delayed
gratification. A more nuanced analysis suggests that
alcohol might differentially affect different aspects of
inhibitory control (Riedel et al., 2021). Moreover,
although alcohol has been found to impair response
inhibition in the stop-signal (de Wit et al. 2000; Loeber
& Duka 2009; Gan et al. 2014; Roberts et al. 2016) and
go/no-go tasks (Fillmore & Weafer 2004; Marczinski
et al. 2005; Field et al. 2010; Korucuoglu et al. 2017;
Stock et al. 2016), recent evidence suggests that the
impact of alcohol on response inhibition may depend
on the particular measure of response inhibition
(Bartholow et al., 2018; Riedel et al., 2021). Based on
these findings, the influence of alcohol on inhibitory
control seems much more complex than presumed in
research on the effects of alcohol on moral judgment,
including the current study (see Button et al., 2013).
Nevertheless, it is worth noting that the lack of
experimental effects of alcohol in the current study and
prior research (Arutyunova et al., 2017; Francis et al.,
2019) does not necessarily question Duke and Bègue’s
(2015) correlational findings. Yet, the lack of
experimental effects does suggest a somewhat
different interpretation of their original findings, in
that blood alcohol may not have been the cause of the
obtained correlations. Instead, these correlations may
have been driven by individual differences that are
systematically associated with both alcohol
consumption and moral dilemma judgments. For
example, it is possible that individuals who tend to
in press, Personality and Social Psychology Bulletin 8
engage in excessive alcohol consumption are less
concerned about causing harm to themselves and
others, which could promote a positive correlation
between blood alcohol levels after a night at the bar
and pro-sacrificial judgments in the trolley problem.
To the extent that this association is more pronounced
for harm that involves direct contact, it would also
explain why Duke and Bègue (2015) found a stronger
correlation between blood alcohol and pro-sacrificial
judgments in the footbridge dilemma than in the
Although our findings provide more compelling
evidence regarding the effect of alcohol on moral
judgment than previous studies, it seems appropriate
to acknowledge a few limitations. The first limitation
is the controlled lab setting of the current study.
Alcohol consumption often occurs in social settings
(e.g., with friends at a party) and effects of alcohol may
differ depending on whether it is consumed
individually or in a social setting. Similar contextual
influences have been found for placebo effects, which
can be different in individual and social settings
(Bodnár et al., 2020). Thus, our lab findings may not
be representative of the effects of alcohol and alcohol-
related beliefs in general if their influence on moral
judgments depends on the context. This idea is
consistent with findings suggesting that moral
dilemma judgments differ depending on whether they
are made in an individual or social setting (Rom &
Conway, 2018). Future research comparing effects of
alcohol in individual and social settings may help to
provide deeper insights into the interactive role of
alcohol and social contexts in shaping moral
A second limitation is that we did not control for
biphasic effects of alcohol, in which blood alcohol
concentration rises to a peak following consumption
(i.e., ascending limb) and then gradually declines to a
sober state (i.e., descending limb). With the design
employed in the current study, it is possible that
participants’ blood alcohol concentration peaked
before the movie ended and was already on the
descending limb by the time they completed the moral
judgment tasks. Because alcohol can have different
effects during times of ascending vs. descending blood
alcohol concentrations, future studies should either
ensure that participants are making moral judgments at
the peak time of blood alcohol concentration or
directly test differential effects of alcohol during the
ascending vs. descending limb.
Third, although the current study used a
comparatively higher dose of alcohol than previous
studies, it is possible that the administered dose was
still too low to produce a detectable effect of alcohol
on moral judgments. Although higher doses of alcohol
may raise ethical questions about potential harm that
might be caused to participants, it is possible that the
correlations in Duke and Bègue’s (2015) study were
driven by intoxicated participants with higher levels of
blood alcohol after a night bar, and that the “drunk
utilitarian” phenomenon would emerge in
experimental studies with higher levels of blood
alcohol. In this case, insufficiently high doses of
alcohol might explain the discrepancy between Duke
and Bègue’s findings and the results of experimental
studies, including the current one.
Fourth, when determining the amount of to-be-
consumed alcohol, we followed the procedures of past
studies (e.g., Francis et al., 2019) and did not
differentiate alcohol doses according to gender.
However, because men and women differ in terms of
their alcohol metabolism (e.g., Bates et al., 2011;
Cofresí et al., 2020), one could argue that women
should have been given smaller doses than men to
obtain the comparable effects of alcohol even when
their body weight was comparable (Thomasson,
2002). Yet, counter to this concern about potential
gender differences, a 2 (Time) × 2 (Gender) mixed
ANOVA did not reveal any significant effects of
Gender on blood alcohol concentration; there was
neither a significant main effect of Gender, F(1, 327)
= 0.48, p = .488, nor a significant two-way interaction
between Time and Gender, F(1, 327) = 0.66, p = .419.
Finally, following previous studies on moral
judgment under the influence of alcohol (e.g., Francis
et al., 2019), we used body weight to determine the
amount of to-be-consumed alcohol. However, an
alternative approach is to use participants’ total body
water (rather than weight) to determine the ideal dose
of alcohol in laboratory studies (Watson, 1989). In this
method, the dose of alcohol required to produce a
specific peak blood alcohol level is assumed to be a
function of the participant’s total body water, duration
of the drinking period, time to peak blood alcohol
level, and alcohol metabolism rate. Future studies
might use Watson’s published formulas for this
alternative approach to determine the ideal dose of
alcohol (Curtin & Fairchild, 2003).
There are reasons to believe that alcohol may
influence moral judgments. On the one hand, alcohol
may impair inhibitory control and extant theories
suggest that impaired inhibitory control should reduce
the endorsement of pro-sacrificial harm for the greater
good. On the other hand, “drinking is […] like taking
one’s foot off the brake” (Heath & Hardy-Vallée,
2015, p. 2), which is consistent with the greater
willingness to cause pro-sacrificial harm in the “drunk
utilitarian” phenomenon. However, counter to either
of these ideas, we did not find any effects of alcohol
in press, Personality and Social Psychology Bulletin 9
on moral judgments. Because the current study
addressed several limitations of prior research on this
question and nevertheless did not find any evidence for
a causal effect of alcohol on moral judgments, we
conclude that the “drunk utilitarian” phenomenon
needs to be revisited.
Aquino, K., & Reed II, A. (2002). The self-importance
of moral identity. Journal of Personality and
Social Psychology, 83, 1423–1440.
Arutyunova, K. R., Bakhchina, A. V., Krylov, A. K.,
& Alexandrov, Yu. I. (2017). The effects of
alcohol on heart rate and evaluation of actions in
moral dilemma. Experimental Psychology
(Russia), 1, 5–22.
Bartholow, B. D., Fleming, K. A., Wood, P. K.,
Cowan, N., Saults, J. S., Altamirano, L., Miyake,
A., Martins, J., & Sher, K. J. (2018). Alcohol
effects on response inhibition: Variability across
tasks and individuals. Experimental and Clinical
Psychopharmacology, 26, 251–267.
Bates, M. E., Buckman, J. F., Vaschillo, E. G.,
Fonoberov, V. A., Fonoberova, M., Vaschillo, B.,
Mun, E. Y., Mezić, A., & Mezić, I. (2011). The
redistribution of power: neurocardiac signaling,
alcohol and gender. PloS One, 6, e28281.
Białek, M., & De Neys, W. (2017). Dual processes and
moral conflict: Evidence for deontological
reasoners’ intuitive utilitarian sensitivity.
Judgment and Decision Making, 12, 148–167.
Bodnár, V., Nagy, K., Cziboly, Á., & Bárdos, G.
(2020). Alcohol and placebo: The role of
expectations and social influence. International
Journal of Mental Health and Addiction.
Button, K. S., Ioannidis, J. P. A., Mokrysz, C., Nosek,
B. A., Flint, J., Robinson, E. S. J., & Munafò, M.
R. (2013). Power failure: Why small sample size
undermines the reliability of neuroscience. Nature
Reviews Neuroscience, 14, 365–376.
Capraro, V., Everett, J. A. C., & Earp, B. D. (2019).
Priming intuition disfavors instrumental harm but
not impartial beneficence. Journal of
Experimental Social Psychology, 83, 142–149.
Cofresí, R. U., Bartholow, B. D., & Fromme, K.
(2020). Female drinkers are more sensitive than
male drinkers to alcohol-induced heart rate
increase. Experimental and Clinical
Psychopharmacology, 28, 540–552.
Conway, P., & Gawronski, B. (2013). Deontological
and utilitarian inclinations in moral decision
making: A process dissociation approach. Journal
of Personality and Social Psychology, 104, 216–
Conway, P., Goldstein-Greenwood, J., Polacek, D., &
Greene, J. D. (2018). Sacrificial utilitarian
judgments do reflect concern for the greater good:
Clarification via process dissociation and the
judgments of philosophers. Cognition, 179, 241–
Crone, D. L., & Laham, S. M. (2017). Utilitarian
preferences or action preferences? De-
confounding action and moral code in sacrificial
dilemmas. Personality and Individual
Differences, 104, 476–481.
Curtin, J. J., & Fairchild, B. A. (2003). Alcohol and
cognitive control: Implications for regulation of
behavior during response conflict. Journal of
Abnormal Psychology, 112, 424–436.
Cushman, F., Young, L., & Hauser, M. (2006). The
role of conscious reasoning and intuition in moral
judgment: Testing three principles of harm.
Psychological Science, 17, 1082–1089.
Day, A. M., Kahler, C. W., Ahern, D. C., & Clark, U.
S. (2015). Executive Functioning in Alcohol Use
Studies: A Brief Review of Findings and
Challenges in Assessment. Current Drug Abuse
Reviews, 8, 26–40.
de Wit, H., Crean, J., & Richards, J. B. (2000). Effects
of d-Amphetamine and ethanol on a measure of
behavioral inhibition in humans. Behavioral
Neuroscience, 114, 830–837.
Diamond, A. (2013). Executive functions. Annual
Review of Psychology, 64, 135–168.
Duke, A. A., & Bègue, L. (2015). The drunk
utilitarian: Blood alcohol concentration predicts
utilitarian responses in moral dilemmas.
Cognition, 134, 121–127.
Everett, J. A. C., Faber, N. S., Savulescu, J., &
Crockett, M. J. (2018). The costs of being
consequentialist: Social inference from
instrumental harm and impartial beneficence.
Journal of Experimental Social Psychology, 79,
Everett, J. A., Colombatto, C., Awad, E., Boggio, P.,
Bos, B., Brady, W. J., ... & Crockett, M. J. (2021).
Moral dilemmas and trust in leaders during a
global health crisis. Nature Human Behaviour, 5,
Everett, J. A. C., & Kahane, G. (2020). Switching
tracks? Towards a multidimensional model of
utilitarian psychology. Trends in Cognitive
Sciences, 24, 124–134.
Field, M., Wiers, R. W., Christiansen, P., Fillmore, M.
T., & Verster, J. C. (2010). Acute alcohol effects
on inhibitory control and implicit cognition:
implications for loss of control over drinking.
Alcoholism, Clinical and Experimental Research,
Fillmore, M. T., & Weafer, J. (2004). Alcohol
impairment of behavior in men and women.
Addiction, 99, 1237–1246.
in press, Personality and Social Psychology Bulletin 10
Foot, P. (1967). The problem of abortion and the
doctrine of the double effect. Oxford Review, 5,
Francis, K. B., Gummerum, M., Ganis, G., Howard, I.
S., & Terbeck, S. (2019). Alcohol, empathy, and
morality: Acute effects of alcohol consumption on
affective empathy and moral decision-making.
Psychopharmacology, 236, 3477–3496.
Frederick, S. (2005). Cognitive reflection and decision
making. Journal of Economic Perspectives, 19,
Gan, G., Guevara, A., Marxen, M., Neumann, M.,
Jünger, E., Kobiella, A., Mennigen, E., Pilhatsch,
M., Schwarz, D., Zimmermann, U. S., & Smolka,
M. N. (2014). Alcohol-induced impairment of
inhibitory control is linked to attenuated brain
responses in right fronto-temporal cortex.
Biological Psychiatry, 76, 698–707.
Gawronski, B., Armstrong, J., Conway, P., Friesdorf,
R., & Hütter, M. (2017). Consequences, norms,
and generalized inaction in moral dilemmas: The
CNI model of moral decision-making. Journal of
Personality and Social Psychology, 113, 343–
Gosling, S. D., Rentfrow, P. J., & Swann, W. B., Jr.
(2003). A very brief measure of the Big-Five
personality domains. Journal of Research in
Personality, 37, 504–528.
Graham, J., & Haidt, J. (2012). Sacred values and evil
adversaries: A moral foundations approach. In M.
Mikulincer & P. R. Shaver (Eds.), The social
psychology of morality: Exploring the causes of
good and evil (pp. 11–31). American
Greene, J. D., Morelli, S. A., Lowenberg, K., Nystrom,
L. E., & Cohen, J. D. (2008). Cognitive load
selectively interferes with utilitarian moral
judgment. Cognition, 107, 1144–1154.
Greene, J. D., Nystrom, L. E., Engell, A. D., Darley, J.
M., & Cohen, J. D. (2004). The neural bases of
cognitive conflict and control in moral judgment.
Neuron, 44, 389–400.
Greene, J. D., Sommerville, R. B., Nystrom, L. E.,
Darley, J. M., & Cohen, J. D. (2001). An fMRI
investigation of emotional engagement in moral
judgment. Science, 293, 2105–2108.
Heath, J., & Hardy-Vallée, B. (2015). Why do people
behave immorally when drunk? Philosophical
Explorations, 18, 310–329.
Hütter, M., & Klauer, K. C. (2016). Applying
processing trees in social psychology. European
Review of Social Psychology, 27, 116–159.
Kahane, G., Everett, J. A. C., Earp, B. D., Caviola, L.,
Faber, N. S., Crockett, M. J., & Savulescu, J.
(2018). Beyond sacrificial harm: A two-
dimensional model of utilitarian psychology.
Psychological Review, 125, 131–164.
Kant, I. (1916). Fundamental principles of the
metaphysics of ethics. Longmans, Green.
Korucuoglu, O., Sher, K. J., Wood, P. K., Saults, J. S.,
Altamirano, L., Miyake, A., & Bartholow, B. D.
(2017). Acute alcohol effects on set-shifting and
its moderation by baseline individual differences:
a latent variable analysis. Addiction, 112, 442–
Körner, A., Deutsch, R., & Gawronski, B. (2020).
Using the CNI Model to investigate individual
differences in moral dilemma judgments.
Personality and Social Psychology Bulletin, 46,
Luke, D. M., & Gawronski, B. (2021). Psychopathy
and moral dilemma judgments: A CNI model
analysis of personal and perceived societal
standards. Social Cognition, 39, 41–58.
Loeber, S., & Duka, T. (2009). Acute alcohol impairs
conditioning of a behavioural reward-seeking
response and inhibitory control processes-
implications for addictive disorders. Addiction,
Maxwell, S. E., Lau, M. Y., & Howard, G. S. (2015).
Is psychology suffering from a replication crisis?
What does “failure to replicate” really mean?
American Psychologist, 70, 487–498.
Marczinski, C. A., Abroms, B. D., Van Selst, M., &
Fillmore, M. T. (2005). Alcohol-induced
impairment of behavioral control: differential
effects on engaging vs. disengaging responses.
Psychopharmacology, 182, 452–459.
Mill, J. S. (1863). Utilitarianism. Parker, Son and
Moshagen, M. (2010). MultiTree: A computer
program for the analysis of multinomial
processing tree models. Behavior Research
Methods, 42, 42–54.
Noël, X., Tomberg, C., Verbanck, P., & Campanella,
S. (2010). The Influence of Alcohol Ingestion on
Cognitive Response Inhibition and Error
Processing. Journal of Psychophysiology, 24,
Primi, C., Morsanyi, K., Chiesi, F., Donati, M. A., &
Hamilton, J. (2016). The development and testing
of a new version of the Cognitive Reflection Test
applying item response theory (IRT). Journal of
Behavioral Decision Making, 29, 453–469.
Riedel, P., Wolff, M., Spreer, M., Petzold, J.,
Plawecki, M. H., Goschke, T., Zimmermann, U.
S., & Smolka, M. N. (2021). Acute alcohol does
not impair attentional inhibition as measured with
Stroop interference scores but impairs Stroop
performance. Psychopharmacology, 238, 1593–
in press, Personality and Social Psychology Bulletin 11
Roberts, W., Monem, R. G., & Fillmore, M. T. (2016).
Multisensory Stop Signals Can Reduce the
Disinhibiting Effects of Alcohol in Adults.
Alcoholism, Clinical and Experimental Research,
Rom, S. C., & Conway, P. (2018). The strategic moral
self: Self-presentation shapes moral dilemma
judgments. Journal of Experimental Social
Psychology, 74, 24–37.
Stake, J. E. (1994). Development and validation of the
Six-Factor Self-Concept Scale for adults.
Educational and Psychological Measurement, 54,
Stock, A. K., Schulz, T., Lenhardt, M., Blaszkewicz,
M., & Beste, C. (2016). High-dose alcohol
intoxication differentially modulates cognitive
subprocesses involved in response inhibition.
Addiction Biology, 21, 136–145.
Suter, R. S., & Hertwig, R. (2011). Time and moral
judgment. Cognition, 119, 454–458.
Thomasson, H. R. (2002). Gender differences in
alcohol metabolism. In M. Galanter et al. (Eds.),
Recent developments in alcoholism, Volume 12
(pp. 163–179). Springer.
Thomson, J. J. (1976). Killing, letting die, and the
trolley problem. The Monist, 59, 204–217.
Watson, P. E. (1989). Total body water and blood
alcohol levels: Updating the fundamentals. In K.
Crow & R. Batt (Eds.), Human metabolism of
alcohol: Pharmacokinetics, medicolegal aspects,
and general interest (pp. 41–58). CRC Press.
Weafer, J., & Fillmore, M. T. (2016). Low-dose
alcohol effects on measures of inhibitory control,
delay discounting, and risk-taking. Current
Addiction Reports, 3, 75– 84.
The preregistration, materials, raw data, and
analysis files for the current study are publicly
available at https://osf.io/9vn5z/.
in press, Personality and Social Psychology Bulletin 12
Overview of the study procedure
Note. The schematic individuals in the lower right corner represent three research assistants (informally
referred to as policeman, bartender, and courier) responsible for the different tasks described in the
in press, Personality and Social Psychology Bulletin 13
Descriptive statistics for all measured variables
Note. BAC = Blood Alcohol Concentration in Permille (‰) Measured with Breathalyzer. Switch =
Switch Dilemma, Footbridge = Footbridge Dilemma, C = Sensitivity to Consequences, N = Sensitivity to
Moral Norms, I = General Preference for Inaction versus Action, OUS = Oxford Utilitarianism Scale, IH
= Instrumental Harm, IB = Impartial Beneficence, CRT = Cognitive Reflection Test.
Pearson’s correlations between measured variables
3. C Parameter
4. N Parameter
5. I Parameter
Note. * p < .05; ** p < .01. Switch = Switch Dilemma, Footbridge = Footbridge Dilemma, C = Sensitivity
to Consequences, N = Sensitivity to Moral Norms, I = General Preference for Inaction versus Action,
OUS = Oxford Utilitarianism Scale, IH = Instrumental Harm, IB = Impartial Beneficence, CRT =
Cognitive Reflection Test. Breathalyzer scores are not included in the table, because two thirds of
participants in the sample (i.e., those in the no-alcohol and the placebo conditions) have a score of zero.