A thought-provoking blog post about alcohol and it’s real and constructed effects.
Viktor examines a recent scientific study that made the headlines and attracted media attention for its findings. But Viktor asks if the findings actually are valid? And he explores the conclusions about how alcohol affects empathy and moral judgement.
With alcohol, not everything is as it seems and Viktor’s latest blog post helps to sort this “what is what” puzzle out…

I have come across an interesting study looking at the effects of alcohol on the ability to be empathatic and to make moral decisions.  The study received quite some attention and generated a number of sensational headlines.

But as I read it and a few articles about it, I asked myself: Does this study actually contribute with anything valuable? And are the results valid enough to be applied to any real life scenarios?

Hopefully these questions will become clear at the end of this post.

“Alcohol really is no excuse for bad behaviour – research reveals you’re still the same person after a drink” by Kathryn Francis, Lecturer in Psychology, University of Bradford, in The Conversation

Understanding research about alcohol’s effects on moral judgement

So what have these researchers been up to anyway?

They took 48 participants and randomly assigned them to one of three conditions: 1) placebo, 2) ethanol, low dose 3) ethanol, high dose. A randomized, placebo controlled trial in other words!

How did the researchers mask the ethanol and present the condition? Ethanol with a concentration of 37,5% of its volume with pure ethanol was mixed with lemonade and lime juice. For the placebo condition, ethanol was sprayed around the brim of glass to resemble the odor of an ethanol containing beverage.

Since this type of setup is considered to be the ‘golden standard’ in the medical community, let’s see if this study lives up to the associated expectations!

So what did they test?

The short answer is empathy and morality. The longer answer is that empathy was tested using a series of pictures of facial expressions and pictures of “dangerous” or potentially painful situations. The facial expression pictures either conveyed neutrality, happiness or sadness and the participants were asked how they felt about these different types of expressions. In the other pictures of painful situations, the participants were asked to put themselves in a ‘first person perspective’ and rate the pain response. These exercises have been conducted previously and are used to measure the ability to show empathy or detect traits of psychopathy.

To test the subjects’ morality, the test was split into two parts: one judgement-based part and one action-based part. The textual judgement based part was presenting a moral decision, for example, “kill one person to save five” where the judgement part assessed two different responses: 1) Is it morally acceptable to do that? and 2) Would you do it?
For the action based part the researchers used a VR (virtual reality) module where subjects could walk around with a joystick and interact with their environment. In the VR-module, a moral decision would present itself where the subjects could choose to react on or not. In this way, the researchers tried to capture both the intellectual and “real life” aspects of moral decision making.

Assessing the validity of these experimental tasks and measurements are outside my field of expertise, so we’ll have to assume that they suffice to measure empathy and moral decision making with some accuracy.

So what did the researchers find?

Without going into too much detail about the results, as the news article from the web portal The Conversation describes it, alcohol seemed to affect empathy, but not morality.

  • Effect on empathy was not throughly consistent but it seemed like people in the high alcohol group were more positive towards sad faces (sic) and more negative towards happy faces.
  • With regards to morality, judgement- or action-based, no effect on alcohol was seen.

However, there is a very important caveat that should be noted.

Not presented in the article itself, but in its supplement, the researcher presents the data for the awareness check conducted during the experiment. The awareness check is supposed to declare if the participants were aware of which condition they were a part of. Reading from the supplement, the awareness of assigned condition were the following:

  1. 12,5% for the placebo group
  2. 31,25% for the low-dose group and
  3. 68,75% (!) for the high-dose group.

In plain language, this means that even though the majority of placebo group were unaware that they were not consuming ethanol, awareness increased with every condition, where in the last condition the majority of the participants were aware of their consumption of ethanol during the experiment.

To the researcher’s defense, they’ve stated that “reported awareness of condition assignment was not associated with either moral or moral judgments […] ” (word by word from the supplement) so this basically means that awareness did not affect morality tests (though I still wonder if it would have been different with a larger sample size since there were quite few participants in every group).

However, the thing that surprises me the most is the following: “[…] awareness checks were not included in further analyses”.

So, if I understand this correctly, the one thing that could massively affect how subjects interpreted their surroundings (awareness of ethanol) was not checked for in the only test that showed a result of alcohol use. This is very strange to me since I view it as a hugely important confounder to be reckoned with.

Alcohol placebo: learned expectations and excuses

Why do I think that awareness of alcohol use would affect the empathic response with the participants? Because we have learned to associate alcohol use as a “safe space” for conducting socially deviant behaviors. Or just plain and simple: we have learned to use alcohol as an excuse for bad behavior.

How does this relate to study in question? According to the results, participants with high alcohol use showed less positive valence (psychology lingo for feelings) when they looked at happy faces and more positive valence when they looked at sad faces. In layman terms this means that you would describe a person with these sort of tendencies as a ‘douchebag’ or an ‘asshole’. Important to note however is that this was only seen when compared within the group itself, that is, how the ‘high alcohol-group’ where before ingestion and after ingestion of alcohol, compared with themselves. So here we had a significant difference in emotional reactions before and after the consumption of alcohol on these face pictures, but it was only seen within the group of high alcohol consumption.

You know what else was only seen in the high alcohol consumption group? A staggering 70% awareness of assigned condition! And the researchers didn’t even test this interaction statistically to rule out this fact! So how can we rule out the placebo effect of alcohol? And how can we make any assumptions about how alcohol affects our reactions to sad and happy faces if we are aware of the presence of alcohol as a potential excuse of being a jerk?

Until the researchers publish additional tests to examine if awareness of assigned condition affected the empathy tests, I would not make any assumptions of ethanol use on empathy.

Fake Free: it’s the person not the ethanol…

With regards to the study’s findings regarding morality, using ethanol doesn’t seem to affect how you would act morally. You are still the same person with your fixed set of values, ideas and interpretation of the world.

But isn’t this what we have been saying in the Fake Free movement for years?

Maybe, but there is a slight nuance here. In the work with challenging the assumptions and expectations around alcohol and questioning which effects come from consuming alcohol and which are placebo effects actually (a.k.a. the Fake Free approach) there have not been any claims about how people’s acts and opinions are affected that normative and socially neutral. There is a plethora of acts and opinions where individuals don’t use alcohol as an excuse to justify their behavior, because it is still accepted as perfectly normal. For example, I have never heard of anyone using alcohol to justify a change of their favorite color or alteration of their favorite cinnamon buns recipe. Just because these things are considered socially neutral.

But what about atrocious acts such as rape and abuse and abhorrent sexist and racist comments? Even though they in some ways can be explained as ethical or moral dilemmas, it is very far from the moral dilemmas presented in this study, which specifically used the trolley or footbridge dilemma. In my opinion the footbridge and trolley dilemma is very abstract and far from reality in relation to how people interact with each other in their everyday life (a philosopher might disagree, but come at me bro).

The big difference between opinions and acts relating to these dilemmas and being sexist or abusing someone is that it is very clear that the latter acts are strictly confined within some normative boundaries. That is, it is very obvious what is socially acceptable or not. So, with this analysis in mind I think we can reframe the results of this study and put it an another perspective.

Reframing the understanding of the study results

For acts that are deemed as socially unacceptable (such as being happy when seeing sad faces), alcohol seems to have an impact on these behaviors, making individuals slightly more deviant from the norm (although not taking into consideration the fact that individuals were aware of their consumption of alcohol).

For acts that are not associated with a normative structure (or in other words – where there is no ‘right’ or ‘wrong’) like the preference of ‘kantian’ or utilitarian framework of ethics, there doesn’t seem to be an effect of alcohol use on these opinions or acts based on these opinions.

And this is what we have been saying in the Fake Free movement for years: alcohol is no excuse for bad or socially deviant behavior and alcohol is not the reason for the “awesome dance moves” and “flirting skills”.

For further reading