Alcohol unleashes homo economicus by inhibiting cooperation.

Human behavior lies somewhere between purely self-interested homo economicus and socially-motivated homo reciprocans. The factors that cause people to choose self-interest over costly cooperation can provide insights into human nature and are essential when designing institutions and policies that are meant to influence behavior. Alcohol consumption can shed light on the inflection point between selfish and selfless because it is commonly consumed and has global effects on the brain. The present study administered alcohol or placebo (N = 128), titrated to sex and weight, to examine its effect on cooperation in a standard task in experimental economics, the public goods game (PGG). Alcohol, compared to placebo, doubled the number of free-riders who contributed nothing to the public good and reduced average PGG contributions by 32% (p = .005). This generated 64% higher average profits in the PGG for those who consumed alcohol. The degree of intoxication, measured by blood alcohol concentration, linearly reduced PGG contributions (r = -0.18, p = .05). The reduction in cooperation was traced to a deterioration in mood and an increase in physiologic stress as measured by adrenocorticotropic hormone. Our findings indicate that moderate alcohol consumption inhibits the motivation to cooperate and that homo economicus is stressed and unhappy.

RCT Entities:   Population Interventions Outcomes

Introduction

Cooperation in one-shot settings is ubiquitous. But then so are defection and free-riding. A substantial set of mathematical models, laboratory experiments, and empirical analyses have sought to determine when people are likely to cooperate or be selfish [1-4]. Cooperative behaviors are predicted by multi-level evolutionary models in which individuals in a group out-compete other groups [5,6]. Costly signaling of one’s value as a future collaborator, known as indirect reciprocity, also supports cooperation [7-10]. But, indirect reciprocity requires observation of behavior by others. A study with 136,000 private contributions to Social Funds by students at the University of Zurich showed that two-thirds donated money [11]. Similarly, 1,500 randomly selected residents of Denmark reported that 69% were conditional cooperators and 15% were free riders [12]. The conditions under which people cooperate or defect are still not fully understood.

Cooperation been studied in the laboratory using a variety of social dilemmas [13,14]. One used extensively is the public goods game (PGG) [15]. In this task, groups of participants can contribute part of their endowments for a social goal. The dilemma arises because each member equally shares in the aggregate contribution, providing an incentive to free-ride. Nevertheless, contributions are made in many variants of the PGG [16-21]. A meta-analysis of 27 PGG studies reported an average contribution of 38% of participants’ endowments [22]. Typically, one-third to one-half of participants in PGG experiments are free-riders [23,24]. Contributions are moderate in single-decision games, decline with repeat play, but increase when decisions are made face to face [25]. Possible explanations include preferences for altruism and reciprocity [26-29], confusion [30-32], and risk of punishment [18].

The present paper seeks to understand the inflection from cooperation and defection by investigating the impact of an activity that more than one-half of Americans [33] and even more Europeans do at least once a month [34]: consume alcohol. Alcohol has global effects on the brain and impacts decision-making [35]. The primary neurochemical effect of alcohol is increased activity of the inhibitory neurotransmitter GABA [36]. GABA is responsible, in part, for the disinhibition and impulsivity typically observed in drinkers [37]. A secondary effect of alcohol is to stimulate the activity of the excitatory neurotransmitter dopamine in the ventral striatum [38]. Striatal activity is associated with reward-seeking behaviors [39,40]. These neurochemical factors are part of the reason for alcohol motivates a desire for immediate rewards [41,42]. The father of scientific psychology, William James called wrote that "Alcohol [is]…the great exciter of the Yes function in man" [43].

Despite an extensive literature on alcohol and impaired decision-making, little is known about how acute, rather than long-term, alcohol use affects cooperative behaviors. Alcohol use has been reported to weakly reduce charitable donations [42]. Yet, alcohol imbibers were equally generous in a strategic-share-the money laboratory task but and the same time rejected unfair offers more often than sober participants [44]. This behavior may be due to alcohol consumption inhibiting the ability to recognize emotions in others and diminishing empathy [45,46]. At the same time, alcohol use inhibits individuals’ abilities to emotionally regulate themselves [47] that can manifest as selfishness. Yet, male drinkers, but not females, promised to cooperate in the prisoner’s dilemma game at a higher rate than did their teetotaler brethren [48].

Outside of laboratory tasks, there is evidence that casual alcohol consumption affects social-economic outcomes. Analysis of General Social Survey showed that drinkers earn 10–14% more than abstainers and that those who drink socially earn an additional 7% [49]. These authors posit that moderate drinkers spend more time socializing with colleagues and thereby build social capital, although the literature is unclear on the alcohol to social capital association [50]. In order to untangle this relationship, a controlled laboratory study was run to assess the behavioral effects of acute alcohol consumption on group cooperation as well as neurobiological and psychological mechanisms driving behavior.

Methods

The Institutional Review Board of Claremont Graduate University approved this study (IRB #2175) with sessions held at the Center for Neuroeconomics Studies in Claremont, CA. Written informed consent was obtained from all participants before they were included in the experiment. There was no deception of any kind and participants were assigned an alphanumeric code to mask their identities. Participants were informed that that they would either consume alcohol or a placebo during the study. Tasks were incentivized with money and participants could earn up to $65 depending on their decisions and decisions of others. Anonymity was maintained by having a lab administrator who was not associated with the study pay participants their earnings in private at the study’s conclusion.

Study timeline

Fig 1 shows time course of the study. After consent, participants completed surveys assessing their opinions, attitudes, and demographics and were weighed. Participants were then led to a private room where their basal blood alcohol concentration (BAC) was established using a commercial breathalyzer (BACtrack S35, San Francisco, California) and a blood draw was done to establish basal endocrine levels. Participants were excused if their BAC > 0.001%. A die roll next assigned participants to consume drinks alone or in groups of four. Drinks either contained alcohol or a similar-tasting placebo. Participants then had a second blood draw after which they made a social decision and completed a final survey.

Fig 1

Timeline of the experiment.

Alcohol administration

The protocol sought to induce a belief by participants’ that they were consuming alcohol to mask the placebo even though the consent form stated that only one-half of participants would receive alcohol. This was done in several steps following previous protocols used in previous studies involving alcohol and social interactions [51,52]. This including have a bottle marked "vodka" on a drink cart where a research assistant mixed 1.5–3.0 ounces of 80-proof SKYY vodka (Groupo Campari, Milan Italy) or an identical tasting placebo vodka (Arkay Alcohol Free Vodka, Jalisco, Mexico) with cranberry juice in clear highball glasses. Each drink consisted of one-part vodka or placebo, 3.5 parts cranberry juice. Males (females) received 62 (56) grams vodka per kilogram of body weight [53]. The expected BAC levels were 0.04–0.06%. Previous studies involving social alcohol consumption administered males (females) with 82 (74) grams vodka per kilogram body weight, resulting in BAC levels of 0.6–0.8% [54]. We reduced the amount administered to minimize the risk of any participant nearing the California legal driving limit of 0.08%. Participants were randomly assigned to the alcohol (A) or placebo (P) conditions. The rims of all glasses were dabbed with vodka in advance to induce a stinging sensation on the tongue to further influence P participants’ beliefs they were consuming alcohol. These procedures have led participants in the placebo- condition to believe they had consumed alcohol in previous studies [55,56]. Alcohol metabolism varies with a person’s weight and sex [57]. The drinks were mixed by the experimenters in front of participants and they were served one third of their drink every 10 minutes, over the course of 30 minutes. Participants had their BAC measured before being served subsequent portions of their drink. BAC levels were not shown to participants. After the conclusion of data collection, participants were not dismissed from the lab for safety reasons until their BAC reading was less than 0.03%.

Participants

All participants who were of legal drinking age (21) and identified themselves as social drinkers. Candidate participants were excluded if they were i) active alcoholics or at high risk of alcohol abuse following National Institute on Alcohol Abuse and Alcoholism standards [58]; ii) Asians and Native Americans due to low penetrance of alcohol dehydrogenases genes; iii) pregnancy and medical conditions contraindicating alcohol consumption; iv) being 15% above or below ideal height/weight ratio (ideal range 18.50–24.99 kg/m2) [59] that makes dosing alcohol problematic.

Participants were randomly assigned to consume drinks for 30 minutes during which they were served 3 drinks at 10 minute intervals. BAC was measured before the second and third drink were consumed and 10 minutes after the third drink. Participants’ BACs were monitored during drinking to ensure consumption occurred and the desired BAC range was achieved.

Participants were randomly assigned to drink alone or in a group with three other participants who were strangers [52]. During intake, participants were asked to identify if they knew other participants to ensure all groups contained strangers. All groups had a mix of sexes to capture the atmosphere in a typical bar setting. The assignment to drink alone or in a group was included to isolate the effect of socializing with others from the effect of alcohol. While consuming their drinks in groups, participants were instructed to discuss whatever they like except for the subsequent portion of the study, past study participation, and their level of intoxication. Using the average size effect and standard errors in Barraza & Zak [60] produces a power of test of 0.99 using G*power [61] for 100 observations.

Behavioral task

Participants were instructed to move to a bank of computers with partitions to make a decision involving money. A task known as the Public Goods Game (PGG) [20,25] asks participants to contribute to a common pool that is then shared by all as a measure prosocial behaviors. All participants were equally and fully instructed in the task, including being shown examples, and were given a chance to ask the experimenters questions. The PGG was presented using zTree software that uses standard instructions that avoid priming individuals to contribute money [62]. The instruction used neutral language throughout that avoided identifying other participants as "partners" or "competitors" that has been shown to influence decisions when sharing money with others [63]. Participants were endowed with $10 and could contribute any sum, including zero, to a common pool of three other unknown participants in the experimental session. Decisions were made in private in partitioned computer stations and no discussion among group members was allowed. Typical sessions included 12 participants and zTree software randomly formed groups of four in the PGG so participants would not know if their PGG group include others with whom they had socialized. Sessions that had less than 12 participants (4) used a research assistant in the control booth to complete the group of four so the session could proceed. The research assistant was not the one who mixed drinks, nor did he or she drink, and was instructed to make the median PGG contribution from the literature. These four fill-in observations were not included in the analysis and participants did not see others’ contributions so this approach could not affect participants’ decisions. Participants were randomly assigned to groups of four in order to reduce any pre-play social attachment. Participants were not informed about the PGG while drinking in order to reduce anticipatory effects. The instructions stated they would "make decisions" after consuming drinks.

The instructions stated that the total amount contributed to the common pool would be tripled and then split evenly among the four contributors. Those who contribute benefit others in the group, but risk being taken advantage of by non-contributors ("free-riders"). Earnings are highest when a participant free-rides by contributing zero while others contribute nonzero amounts [64]. Participants made four decisions in the PGG without feedback and were informed that zTree would randomly re-match them with others in each round. This was done to inhibit the possible effect of reputation on decisions [65]. Contributions to the common pool and earnings were averaged as dependent variables in order to reduce confusion during decisions in a novel task [19,66].

Blood draws and assays

Participants had two blood draws from an antecubital vein by a qualified phlebotomist. Two 8-ml EDTA whole-blood tubes were drawn in a sterile field using Vacutainer® blood-collection kits. Tubes were stored on ice before being placed in a refrigerated centrifuge and spun at 1500 rpm at 4°C for 12 min following previous protocols [67]. Plasma was aliquoted into 2-ml polypropylene Fisher brand microtubes that were immediately placed on dry ice and then transferred to an -80°C freezer until analysis. Adrenocorticotropic hormone (ACTH) is a fast-acting hormone associated with arousal and stress [68] and was assayed to determine how moderate alcohol intact affects physiology. ACTH was assayed using radioimmunoassay (RIA) produced by DiaSorin, Inc. (Stillwater, MN, USA) by the Reproductive Endocrine Research Laboratory at the University of Southern California (USC). The inter-assay CVs < 11% for all analytes.

Surveys

Questionnaires were used to measure demographics, one’s perceived intoxication (SIS: Subjective Intoxication Scale) [51] acute changes in mood (PANAS: Positive Affect Negative Affect Scale) [69], Satisfaction with Life survey (SWL) [70], and closeness to others (IOS: Inclusion of Others in Self) [71].

Statistical analysis

Student’s t-tests (denoted "t") were used to examine the hypothesized relationship between prosocial behaviors and alcohol consumption by comparing means and testing of Pearson’s correlations are nonzero. Several samples had unequal variances assessed using Levene’s test and for these Welch’s t-test was used and noted in the analysis. The impact of affect and isolation release was examined using a mediation model. We used two key dependent variables to demonstrate the robustness of our findings, contributions in the PPG and profits earned from the PGG.

Results

A total of 128 individuals participated in the experiment (male: n = 58; female: n = 70). Participants did not vary pre-treatment in SWL, IOS, PANAS, basal ACTH, age, or sex distribution (ps>.20). No gender differences in basal states were found (ps>.05).

Behavior

The treatment effectively raised mean BAC compared to placebo (A: M = 0.05, SD = 0.03; P: M = 0.00, SD = 0.00; N = 124, Welch’s t = 15.22, p = .0001, d = 2.82). Participants in the treatment group reported feeling intoxicated (A: M = 19.67, SD = 18.85; P: M = 6.38, SD = 9.58; N = 124, Welch’s t-test = 5.03, p < .001, d = 0.89) with BAC and SIS positively correlated (r = 0.40, t = 4.766, p = .0001).

The contributions to the common pool in the PGG (PGGc) were lower in the treatment group compared to placebo (A: M = 3.61, SD = 3.25; P: M = 5.32, SD = 3.55; t = 2.83, p = 0.005, d = 0.50). The reduction in contributions induced by alcohol in the mixed treatment and placebo sessions resulted in higher profits (PGGp) for treatment participants (A: M = 13.33, SD = 2.95; P: M = 11.96, SD = 2.59; t = 2.79, p = 0.006, d = 0.49; Fig 2). There were no gender differences and no effects of drinking in isolation on PPGc or PPPp (ps> .40).

Fig 2

A. Alcohol reduced PGG contributions by 32% (p = .005) and B. increased earnings from the PGG by 11.5% (p = 0.006) showing that alcohol increased free-riding. Bars indicate SEs.

In order to confirm that alcohol consumption reduced PGGc, a panel data linear regression was estimated (n = 496) in which each previous round’s contribution was included as an independent variable along with indicators for alcohol and socializing. The alcohol treatment continued to significantly reduce PGG contributions (B = -0.57, one-tailed t-test p = .033) and the previous contribution was also significant (B = 0.59, one-tailed t-test p < .0001). The regression did not suffer from multicollinearity (VIFs<1.10). Two more analyses were done to check the robustness of the results. First, PGG round one contributions were assessed using the same variables except the previous round contribution. Alcohol consumption again significantly reduced PGGc (B = -1.79, one-tailed t-test p = .005). Second, the analysis we repeated for round four contributions. Alcohol again reduced PGGc (B = -1,46, one-tailed t-test p = .017). The p-values are nearly identical in all regressions when age and sex are included as controls.

Next, we assessed other behavioral measures of cooperation. The effect of free-riding by treatment can be seen as the return on their PGG investment (PPGp/PGGc) for the 100 participants who contributed a nonzero amount to the common pool. Those who consumed alcohol had a 369% return on their investment compared to a return of 225% return by placebo participants. Alcohol produced twice as many complete free-riders, i.e. those who contributed nothing (A: n = 19; P: n = 9). We estimated a logistic regression to assess the accuracy in predicting free-riders. BAC, sex, age, and post-drinking ACTH were included as independent variables. The model correctly classified free-riders with 80.7% accuracy (pseudo-R2 = .14. p = .003) and absent multicollinearity (VIFs ≤ 1.03).

The amount of intoxication (BAC) directly reduced PGG contributions (r = -0.18, t = 2.00, p = .05) and increased participant profits (r = 0.27, t = 3.06, p = .003, Fig 3). The association between BAC and PGG contributions and profits were confirmed by the impact of participants’ self-reported intoxication (SIS) on PGG contributions (r = -.20, t = 2.29, p = .02).

Fig 3

A. PGG contributions were inversely associated with BAC levels (r = -0.18) while B. PGG earnings increased linearly with BAC (r = 0.27).

Stress and affect

Both control and treatment participants had an increase in negative affect during the study (change in negative affect: -0.171, t = -11.875, df = 127, p = .0000). This result continued to hold when controlling for age and sex (ps < .05). Whether participants drank alone or in a group did not affect the change in negative affect (p = .18) nor did participant sex (p = .23). The data showed that alcohol intake increased physiologic stress. Post-treatment ACTH trended towards parametric relationship with BAC (r = 0.18, t = 1.92, p = 0.058). This relationship became significant when controlling for age and sex in a linear regression (BAC: B = -.928, p = .037). The stress effect was driven both by alcohol and by isolation (B = .533, p = .034).

We estimated a model for PGGp to test if stress (ACTH) mediates the increase in profits in the PGG. The model shows that alcohol increases stress directly and both directly and indirectly increases profits in the PGG (p = .003). A mediation model for PGG contributions had similar estimates (p = .043; Fig 4).

Fig 4

A mediation model shows that alcohol (BAC) directly increases profits in the PGG (p = .013) and trends toward an indirect impact by increasing stress (ACTH, p < .053).

Standardized regression coefficients are shown, * = p < .05.

Discussion

We found that moderate alcohol consumption reduced contributions to a public good pool by 32%. Those who consumed alcohol earned 64% more money because they interacted with more cooperative placebo participants. Alcohol also doubled the number of participants who were complete free-riders, contributing nothing to the public good. BAC linearly reduced PGG contributions "unleashing" individuals to behave selfishly.

If money is the only value one receives from cooperation, at least as captured by the PGG, then the present study has shown that a moderate consumption of alcohol results in behavior closer to that predicted by traditional models in economics [2,18,72]. This may be due to alcohol’s stimulation of the neurotransmitter dopamine [38] that is strongly associated with reward-seeking behaviors [73]. Conversely, a rich literature has documented the humans are gregariously social and that most people in most circumstances are biased towards cooperation [74,75]. Our results are unlikely to be affected by the methodology we employed. Previous research has shown that monetary decisions that capture cooperative behaviors that include blood draws match those absent blood sampling [74]. Nevertheless, while we sought to capture typical social drinking, our results may not generalize to single-sex alcohol consumption or drinking by older cohorts.

Alcohol’s inhibition of appropriate social responses has been termed "alcohol myopia" [76] but is more typically seen in heavy drinkers and alcoholics that moderate imbibers [77]. The primary mechanism producing inhibition has been traced to an increase in the inhibitory neurotransmitter GABA [37]. If the treatment reported here resulted in alcohol myopia, it appeared to decrease the value put on social benefits and increase the value of selfish benefits as has been shown with testosterone administration [78]. Indeed, pre-play communication has been consistently been shown to increase cooperation [79], yet alcohol was shown to blunt this effect.

Alcohol’s reduction of the perceived value of cooperation was manifested in the present study by an increase in negative affect. Alcohol accentuates emotional volatility [76,80], negative affect [81], and impulsivity [82]. The reduction in prefrontal activity that moderates social-emotional responses [83] reduced affective states in alcohol-consuming participants and may have focused them on immediate monetary rewards rather than the psychic reward of conforming to a social norm of cooperation [84]. Most economics studies have measured impulsivity by the choice of immediate versus delayed rewards. Individuals showing patience for temporal rewards are generally more cooperative [85] counter to the results found here. The role of stress has not been measured in the existing literature and may explain the difference in findings. At the same time, trait impulsivity can lead to alcohol use and abuse [86].

Social rejection and physical pain have been also shown to increase one’s desire for money [87] and our analysis suggests that an increase in negative affect of moderate alcohol consumption may mimic pain responses when it comes to money [88]. This finding is in contrast to much of the literature showing that negative affect increases monetary allocations to others in ultimatum and dictator games [89]. We did not find that isolation while drinking influenced negative affect compared to those drinking socially as others have reported [90].

Perhaps our most valuable finding is that alcohol increases physiologic stress and through this route reduces cooperation. This was captured by higher levels of the stress hormone ACTH for those who consumed alcohol. ACTH, rather than cortisol, was measured because the former responds more quickly than the latter in line with the time course of the experiment. Moderate stress tends to increase prosociality [91-93] while high and/or chronic stress inhibits prosocial behaviors [94,95]. Moderate alcohol consumption may be an effective way to induce physiologic stress, in particular, by having people drink alone. Our finding that physiologic stress was higher for those drinking alone seems to be new in the literature that has focused on drinking to reduce stress [96]. While drinking alone is a known risk factor for alcohol abuse [90], we have shown that drinking alone reduces subsequent prosocial behaviors. This may further isolate and stress solo drinkers, influencing them to continue to imbibe alcohol. The increase in stress was primarily driven by women and we believe this finding is worth additional research. A replication of study’s results is warranted because when segmented into subsamples, some of the analysis was relatively underpowered.

Our findings show that homo economicus is alive and well and that alcohol is enough to bring him out. A variety of factors besides alcohol reduce prosocial tendencies, including high levels of testosterone and serotonin depletion [97]. The present study was not designed to capture the contribution of changes in neurotransmitters on cooperation, but this is a rich area for future research.

23 Mar 2021

PONE-D-21-04756

Alcohol Unleashes Homo Economicus by Inhibiting Cooperation

PLOS ONE

Dear Dr. Zak,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

You will find both reports pasted below. I agree with them that certain details should be clear. Regarding power calculations I am not asking for new experimental sessions. It would be enough if you just recognize that certain results might be taken cautiously due to the lack of power (since you have many treatment and quite limited sample).

Please submit your revised manuscript by Apr 26 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Pablo Brañas-Garza, PhD Economics

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

3. Thank you for stating the following in your Competing Interests section:

'no'

a. Please complete your Competing Interests statement to state any Competing Interests. If you have no competing interests, please state "The authors have declared that no competing interests exist.", as detailed online in our guide for authors at http://journals.plos.org/plosone/s/submit-now

b. This information should be included in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

4. Please ensure that you refer to Figure 4 in your text as, if accepted, production will need this reference to link the reader to the figure.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The paper is well done and the conclusions are interesting. I have only two small comments. One is that I do not know why the authors did the two treatments where participants drank alone or in a group. This ends up not making a difference, but I suppose they conjectured it could have been different and I would like to understand this.

The other point is that the closest study I can see is the one of of Corazzini, Filippin & Vanin, 2015. I am surprised that they chraracterize it as "Alcohol use has been reported to weakly reduce charitable donations." I was curious and downloaded the paper. It seems like the donations to the both the charity MSF and the journal are reduced by about a half of the baseline. So in a sense it is very much in line with what the authors find here, if anything a bit higher. One added value of this study in my view is to do it in a strategic setting, on the one hand. Did the authors think the strategic setting would change something? If so, why? If not, what have we learnt? The other added value is to check the mediation of stress, which the other study did not check. I have a sense this is the major contribution, and perhaps the authors could highlight it a bit more.

Reviewer #2: I have very much interest in this paper. However, at this point I have several important concerns, which I shorlty list here:

1. The sample is clearly too small. 128 individuals, split into 2 treatements (control and intervention), plus in 2 again (drink alone or drink with 3 people). I would expect more sessions to be conducted.

2. I would need more details about the sessions. It is said that a session consists of 12 people, how were exacly composed the sessions since with 128 people you cannot have only sessions of 12?

3. Why were there mixed groups? There is a large litterature showing that men and women react differently to alcohol and behavioural tasks, why not making groups of females only and of men only?

4. It is said there was no deception, however some participants had placebo and it is said they were pushed to believe they will have alcohol. So I am sorry, but I cannot understand how this was not deception?

5. It is not clear if participants knew the task in advance, while drinking, and if they knew there will be a task at all?

6. It is explained that alcohol makes people act on impulse, but economic papers show that when acting on impulse, people are more cooperative. So I would appreciate a discussion.

7. It is said that alcohol motivates the desire for immediate reward, but this reward could be warm glow or something like this, so, again, I would need a discussion.

8. There is a huge number of papers explaing that pre-play interaction has positive effects on cooperation, here the pre-play has not, I need a discussion.

9. It is said that when moving from drinking in groups of 4 to playing the public good game they did not stay in the same group. Why ?

10. I do not see any results about the differences between those who had drinks alone and in a group.

11. It is sais they made 4 decisions, that means there were 4 rounds in the public good game?

12. I see the experiment as very invasive with all the measures etc...so this would compensate on the social effect of drinking, right? The situation is not "normal" and this is not how cooperation decisions are taken "in real life" after a drink. So this can explain the results as participants understand that they are expected to be inhebriated and their actions under this condition will be studied. So this can induce behavior. I would need some discussion.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Angela Sutan

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

1 Apr 2021

Responses in the attached file.

Submitted filename: Reply to referees 4-1-21 short.docx

Click here for additional data file.

16 Apr 2021

PONE-D-21-04756R1

Alcohol Unleashes Homo Economicus by Inhibiting Cooperation

PLOS ONE

Dear Dr. Zak,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I will explain my decision in detail.

1. Let me start saying that one referee is still against the paper (see the report below) while the other withdraw (he/she did not mention why).

2. Typically I follow the advice of the referees as they are but this time I need to do my own report since Reviewer #1 is gone.

My own impression.

1. I agree with Reviewer #2 that you need more data. But I also agree with you that given the current circumstances is quite hard to run these type of experiments.

2. I also agree with Reviewer #2  (and #1) that you have too many treatments (not 2 but 4) and obviously you may have a problem with power.

3. I also agree with Reviewer #2  that the idea of replacing teams members with RAs (intoxicated or not) is a least exotic.

4. Lastly, still I do not understand if subjects played with or without feedback (and this is NOT trivial at all).

Ok. Personally I find the paper interesting. Yes it is. But I feel that we certain things need to be fixed in order to be sure that the results that you get are robust.

There are 3 complementary ways to solve some of the problems (regarding the PGG).

W1: Instead of running the analysis shown in the paper, just need to do is to use a panel. You use for each subject the four decisions, the treatment, the second treatment, etc... and the feedback in t-1 (if does exist). If you have feedback then you can use only t=2, 3 and 4. Typically you can run the same model with and without feedback. And you can use alcohol as continuous or dummy. ... Your main claims (contributions and earnings) would be much clear using panel. Besides my own taste, panel is the standard econometric procedure in this literature.

W2 (robustness): In order to remove any potential problem of the introduction of RAs as participants you can run a regression for t=1. In this case, does not matter whether the RA contributed the median, high or low since participants only play against their own beliefs.

W3 (robustness 2): doing the same for the last round where you add as variables the feedback (if exist) for round 1, 2 and 3. This is important because the RA would appear just once in each group (or none).

There are many other comments of Referee #2 that I agree but I also understand that every single person has a different way to explain things.

Sorry for being so repetitive but I tried to my best. As author I much prefer to get constructive and doable feedback.

Best, Pablo

** the rest of the message is automatic ***

Please submit your revised manuscript by May 31 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Pablo Brañas-Garza, PhD Economics

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: Dear authors, with regard to the revisions performed in this round, I have the following comments listed here together with the previous questions and replies.

1. The sample is clearly too small. 128 individuals, split into 2 treatments (control and intervention), plus in 2 again (drink alone or drink with 3 people). I would expect more sessions to be conducted. A power of test calculation is now added to the revised ms showing the majority of statistical tests have high power. Most of the analysis uses the full sample to keep the power of tests high, identifying subconditions such as drinking alone with binary indicator variables. With the current pandemic and on advice from the editor, we are unable to conduct additional sessions

of this experiment. We have added a cautionary note in the Discussion that segmenting the sample will reduce the power of the statistical tests.

COMMENT: We all know that the paper will be cited without reference to the footnote or the low power of tests. People will only cite the results, not their limited power (especially as the author is famous). I cannot understand why sessions could not be conducted later, after the pandemics. Why the hurry?

2. I would need more details about the sessions. It is said that a session consists of 12 people, how

were exactly composed the sessions since with 128 people you cannot have only sessions of 12?

The paper states that "Typical sessions included 12 participants." Some sessions were smaller

or larger due to dropouts and a research assistant filled in the group of 4 with the fill-in data

not analyzed.

COMMENT: The authors have now written in the paper : "Sessions that had less than 12 participants used a research assistant to complete the group of four. The research assistant always made the median decision

from the existing literature and these data were not included in the analysis. Participants were randomly assigned to groups of four in order to reduce any pre-play social attachment. Participants were not informed about the PGG while drinking in order to reduce anticipatory effects."

This problem is huge. first of all, a typical session should be a systematic session. Given the low number of people (12) and the random matching, all sessions need to be conducted with 12 participants exaclty. Otherwise the probabilities of rematching are not the same. Second, the participants are not told that an RA will play and how. They think they play agains a regular participant. Third, what is the "median from the literature"? This has clearly an influence because participants randomly meet someone playing a median, and possibly not a free rider or so. Specially that this paper claims not to confirm median results from the literature. Also, OK those numbers are said to be removed from the data but: there is random matching, data is not independent, all the session has to be dropped. This is clearly what we explain to all students starting doing experiments. Later on, they will read this paper written by a famous author in which it is allowed in a random matching game to introduce without telling the participants a player playing a median from the literature...Finally, before playing the PGG, some participants were interacting in groups of 4, they were drinking together. So they were not 4 in some groups, or was the RA also drinking with them? Or they were 3 and after were told that they are 4 in the PGG?

3. Why were there mixed groups? There is a large litterature showing that men and women react differently to alcohol and behavioural tasks, why not making groups of females only and of men only?

The revised ms now clarifies that we sought to capture what typically happens when people go

to a bar.

COMMENT: I cannot understand why the authors insisted on creating the atmosphere from the bar as later (see the paragraph commented at the previous point) they say this anyway was supposed to be destroyed by the random matching and also later in another response the authors say that the paper was not supposed to create real consumption conditions (when I asked abou the invasive protocol). I cannot understand now with the new comments why the authors needed this variation of the treatments in groups of 4 while drinking.

4. It is said there was no deception, however some participants had placebo and it is said they were pushed to believe they will have alcohol. So I am sorry, but I cannot understand how this was not deception?

Participants were informed they would consume alcohol or placebo, hence no deception.

COMMENT: This is not said in the text. In the text it is written (unless I am wrong) that there was a placebo and alcohol, and that participants were made to think it would be alcohol. It should be written clearly in the text : participants were informed there was maybe placebo. I had no access to instructions anyway.

5. It is not clear if participants knew the task in advance, while drinking, and if they knew there will be a task at all?

The revised ms. clarifies that during the consumption phase, participants did not know what tasks they would do later.

OK. But they knew they will be doing something later, after the drinking, right?

6. It is explained that alcohol makes people act on impulse, but economic papers show that when acting on impulse, people are more cooperative. So I would appreciate a discussion. The revised ms. Discussion now brings in more literature on impulsivity and economic decisions. Thank you for this suggestion.

OK.

7. It is said that alcohol motivates the desire for immediate reward, but this reward could be warm glow or something like this, so, again, I would need a discussion. We have added a cite to the Discussion on this.

OK.

8. There is a huge number of papers explaing that pre-play interaction has positive effects on cooperation, here the pre-play has not, I need a discussion. Great suggestion, we have added this to the Discussion. The role of pre-play was reduced by randomly matching participants in groups of four for the PGG.

OK.

9. It is said that when moving from drinking in groups of 4 to playing the public good game they did not stay in the same group. Why ? Revised ms. clarifies in Methods that the PGG used random matching of participants --the goal

was to reduce the effect of pre-play interactions.

COMMENT: So, again, in this case why bother with preplay in groups of 4? I suppose I miss something, and now even more than after the first version.

10. I do not see any results about the differences between those who had drinks alone and in a group. We now mention this nonfinding on PPGg and PPGc to the first paragraph of the Results.

OK.

11. It is sais they made 4 decisions, that means there were 4 rounds in the public good game?

Correct, with random rematching.

OK.

12. I see the experiment as very invasive with all the measures etc...so this would compensate on

the social effect of drinking, right? The situation is not "normal" and this is not how cooperation

decisions are taken "in real life" after a drink. So this can explain the results as participants

understand that they are expected to be inhebriated and their actions under this condition will be

studied. So this can induce behavior. I would need some discussion.

The study did not claim to replicate real life. Rather, we sought to establish the role of a

commonly used drug (alcohol) on cooperation. My lab has published many studies using blood

draws to establish neurochemical mechanisms affecting behavior and choice data are very

similar to that found without blood draws. We now mention this in the Discussion with a

citation.

COMMENT: I am aware of all paper published in your lab. With regard to this, I know that this paper will be read and cited ans used as an example by young scholars (or cited in the press). So my previous comments insist on paying attention to the way in which this paper was conducted (problems in the procedures, in the design, low power, need to drop incomplete sessions, conduct others). Yet, with regard to this specific remark (number 12): here I comment only on this paper, not on the other papers from your lab. It is a good thing to measure blood indicators and everything. I say here in this context of drinking, this may be too invasive. But now, anyway, I cannot understand what you wanted to do, because in some places you say that a bar consumption was intended to be replicated, and in some other places that pre-play interactions was supposed to be counterbalanced by random matching and that the paper was not intended to replicate real life. So in that case, if the study intended to look at the effect of alcohol only, I see no use of sessions of 4, and also, I come back to my remark: pre-play in miwed groups will induce expectations about cooperation from women, domination feelings etc....and this has to be controlled.

I want to say that study like this one are important. And that this paper is likely to become a reference. But will all respect, it needs to respect the standards as to be clean.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

30 May 2021

Letter uploaded with files, but pasted here again.

We thank R2 for taking the time to offer additional suggestions to improve our ms. We looked at each one in detail and have revised the ms accordingly. My coauthors and I believe this revision addresses the suggestions you have made and herein we detail the changes. We expect you will find this revision satisfies your remaining questions.

Power. As we note in the cover letter, the bulk of our analyses use the full N=124 data set with indicators for treatment and socializing so these analyses are not underpowered. In addition, as suggested by the editor, we have done a panel data analysis with the full set of n=496 behavioral observations (rather than averaging PGG behavior). The revised ms. now reports the panel data analysis showing that alcohol continues to reduce PGG contributions. The editor also recommended analyzing alcohol's effects on round 1 and round 4 contributions. Again, the results for alcohol reducing PGG contributions are sustained as the paper now reports. We appreciate these excellent suggestions that show the robustness of the findings.

For these reasons, we disagree that more data is needed. The only underpowered tests reported are for free riders and the effect of drinking alone, both secondary analyses that the paper includes but are not core findings. Further, both these findings are statistically significant even with lower power. In addition, the drinking alone condition was only significant in affecting the stress response and was not otherwise used in the analysis so that sample is not split in half due to this. Indeed, we believe the hypothesis of whether it is alcohol or socializing that affects behavior was worth testing and reporting even if it did not affect behavior.

Replacements. There were only 4 instances of RAs filling in and these data were not analyzed as the paper states. The Methods of the revised ms now expands on this and clarifies that RAs did not drink and could not be identified by participants as decisions were anonymous with the RA filling in a PGG group entering data from the control room. Since participants never received feedback on others' contributions, the revised paper emphasizes there is no way for this filling to affect participants' decisions.

Mixed sexes and design approach. This was a design choice to capture the most common effects for an already complicated experiment (e.g. we hired a phlebotomist to do 248 blood draws, then we processed these tubes to isolate plasma that was stored in an ultracold freezer until sent for analysis, mixed drinks, measured BAC multiple times, etc.). My lab is focused on identifying mechanisms producing behavior which is why we took blood samples before and after drinking. Lab studies balance what people do outside the lab (e.g. drink) with the precision that lab studies offer in measuring behavior and the mechanisms that produce them. There are many ways to study the effects of drinking on cooperation, but our finding of stress and negative affect are unique to the literature and help explain why alcohol reduces cooperation. We have added to the Discussion that our result may not generalize to single sex groups or older individuals as you noted. Our hope is that this study provokes others to explore different ways to study this important topic.

Deception: The revised Methods now clarifies that the instructions to participants stated that only one-half would drink alcohol while the others would receive a placebo. The masking of the placebo followed a protocol in the cited alcohol literature as we now emphasize.

Tasks. The revised ms. clarifies that participants only knew they would "make decisions," but not what decisions they would make.

Submitted filename: Reply2 to R2 5-29-21.docx

Click here for additional data file.

2 Jun 2021

Alcohol Unleashes Homo Economicus by Inhibiting Cooperation

PONE-D-21-04756R2

Dear Dr. Zak,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Pablo Brañas-Garza, PhD Economics

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

14 Jun 2021

PONE-D-21-04756R2

Alcohol unleashes homo economicus by inhibiting cooperation

Dear Dr. Zak:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr Pablo Brañas-Garza

Academic Editor

PLOS ONE