Higher social class predicts increased unethical behavior

[Pythagorean]

Seven studies using experimental and naturalistic methods reveal that upper-class individuals behave more unethically than lower-class individuals. In studies 1 and 2, upper-class individuals were more likely to break the law while driving, relative to lower-class individuals. In follow-up laboratory studies, upper-class individuals were more likely to exhibit unethical decision-making tendencies (study 3), take valued goods from others (study 4), lie in a negotiation (study 5), cheat to increase their chances of winning a prize (study 6), and endorse unethical behavior at work (study 7) than were lower-class individuals. Mediator and moderator data demonstrated that upper-class individuals’ unethical tendencies are accounted for, in part, by their more favorable attitudes toward greed.

http://www.pnas.org/content/early/2012/02/21/1118373109.full.pdf

 

[russ_watters]

I'll need to read this in more detail, but some early observations:

1. Who defined the ethics? Who decided greed was unethical?
2. How was the sample defined? Did it include people who were in jail? Rhetorical: people in jail aren't driving so they can't be observed while driving.
3. The introduction reflects common anti-success prejudices; namely that it isn't easy to succeed without cheating which therefore means that a significant fraction of winners must have cheated.
4. The study is rife with unnecessary rhetoric, indicating a strong bias in the authors.

One thing pointed out in the intro is that the lower classes give a higher fraction of their income to charity than the upper classes. But consider this: when you are in a plane and the oxygen masks deploy, you are instructed to put your mask on first and then the mask of your child. Why? Obviously it is because if you pass out you won't be able to help your child. Thus, being selfish is what enables you to help others; a poor person may give a higher fraction to charity, but a rich person gives vastly more actual money which means the rich person is doing more actual good...as a result of his selfishness.

 

[Ryan_m_b]

One thing pointed out in the intro is that the lower classes give a higher fraction of their income to charity than the upper classes. But consider this: when you are in a plane and the oxygen masks deploy, you are instructed to put your mask on first and then the mask of your child. Why? Obviously it is because if you pass out you won't be able to help your child. Thus, being selfish is what enables you to help others; a poor person may give a higher fraction to charity, but a rich person gives vastly more actual money which means the rich person is doing more actual good...as a result of his selfishness.

That analogy isn't appropriate here. Putting on your mask first ensures that you are in a position to help others after and therefore isn't selfish. This paper purports to show that people who "put on the mask first" then stay sat not really helping those around them.

Furthermore thanks to marginal utility giving away a set percentage of a larger wealth has less of an impact on the owner of that wealth than someone with a comparatively smaller amount so the idea of taking care of oneself first doesn't hold out in that case.

 

[Pythagorean]

I'll need to read this in more detail, but some early observations:

1. Who defined the ethics?

Society, really. I think most of us agree that lying, cheating, and stealing are unethical.

Who decided greed was unethical?

They didn't assume greed was unethical. They screened for people that saw greed as good and found them to be higher class and more likely to participate in unethical practices. They also cite several similar studies in the introduction that finds correlations between unethical behavior and greed.

2. How was the sample defined? Did it include people who were in jail? Rhetorical: people in jail aren't driving so they can't be observed while driving.

"We conducted seven studies using university, community, and
nationwide samples to test this general prediction."

3. The introduction reflects common anti-success prejudices; namely that it isn't easy to succeed without cheating which therefore means that a significant fraction of winners must have cheated.

That's not really a criticism about the quality of the research though. Certainly, we should all be aware of the possibility for a political agenda because of the topic, but we can criticize based on merit rather than an icky filling in your tummy.

4. The study is rife with unnecessary rhetoric, indicating a strong bias in the authors.

Again, please give concrete examples and criticize based on merit of the claims rather than making vague value judgments as you have in 3. and 4.

 

[atyy]

What is the appropriate definition of unethical behaviour in a lab study? For example, is anyone actually harmed by the unethical behaviour in the lab study? If no one is harmed, isn't it just fictional unethical behaviour, ie. no more unethical than an actor playing the role of a villain?

 

[Dotini]

Since I began taking fencing lessons, I have learned that fencing was originally developed as a means to safely practice for dueling with the sword and its variants. I also learned that dueling over questions of honor was a prerogative exclusive to the upper classes, nobility and more lately, military officers. I suppose dueling, since it was a matter of killing or being killed, was eventually outlawed by church and state because it is unethical, as well as an unconscionable waste of life.

So I submit that dueling is an (anachronistic) example of unethical behavior among upper classes, albeit motivated more by chivalry than greed.

 

[Pythagorean]

What is the appropriate definition of unethical behaviour in a lab study? For example, is anyone actually harmed by the unethical behaviour in the lab study? If no one is harmed, isn't it just fictional unethical behaviour, ie. no more unethical than an actor playing the role of a villain?

I think this is equivocating experimental ethics with social ethics. Just because the experimental ethics board doesn't want you harming vertebrates, doesn't mean they're saying it's ok to cheat, lie, or steal when it comes to reporting the research and there are legal and social repercussions for doing so... because as a society, we don't condone of lying, stealing, or cheating.

 

[marcus]

http://www.pnas.org/content/early/2012/02/21/1118373109.full.pdf

Thanks for the link. There is a nine-minute YouTube about this research, that Greg linked us to.
It shows some of their actual experimental set up, videos of experimenters and subjects interacting, videos of outdoors observations they made on the street. I thought it was entertaining actually. It was ingenious how they set up situations in which the subjects could cheat on board games, and swipe some candy that was supposedly intended for some children later and so on. A lot of the stuff, as I recall, was kind of harmless and funny when you think back on it. Sometimes they gave out different amounts of play money (monopoly money) to make some subjects *feel* richer than other subjects, even though in real life they weren't. In other cases there was a real world difference in wealth.

Here's the YouTube:


I wouldn't actually say they were correlating with social CLASS. that is probably somewhat different from wealth

Here's Greg's earlier post about this.
https://www.physicsforums.com/showthread.php?t=701749

 

[atyy]

Regarding Experiment 1 and 2: they had no evidence the driver was of higher social class. If the driver was a chauffeur, as is conceivable for a higher social class vehicle, he may have been of lower social class.

Regarding Experiment 3: they did not exclude that lower social class participants were less truthful in self reporting.

Regarding Experiment 4: I don't see anything unethical about taking candy from kids. It causes tooth decay!

Regarding Experiment 5: Are the p-values adjusted for multiple comparisons? If they have not been, an argument must be made why all the tests conducted are not independent. Also they were asked "What is the percentage chance that you will tell the job candidate that the position is certain to be eliminated in 6 months if she/he specifically asks about job security?" There is no lie if the person does not answer the question. I'm not convinced this is unethical behaviour.

Regarding Experiment 6: The experimenters clearly lied. Is that ethical? Also, no information is given as to whether the $50 voucher was given out.

Regarding Experiment 7: Same flaw as Experiment 3.

 

[Pythagorean]

Regarding Experiment 1 and 2: they had no evidence the driver was of higher social class. If the driver was a chauffeur, as is conceivable for a higher social class vehicle, he may have been of lower social class.

It's hard to say how representative chauffuer's would be. Probably not too representative of sports cars though. And anyway it still supports the general thesis from the wealth simulations that a large contributing factor is situation, not intrinsic character.

But there's really deeper problems with this study. How do we ensure the observers aren't subject to selection bias? Do they really know how to value cars or just obvious ones like Jaguars and Corvettes? Are they really able to track all the cars and associated events at once? It's a naturalistic field study, so it has all of those caveats that come with it.

Regarding Experiment 3: they did not exclude that lower social class participants were less truthful in self reporting.

Regarding Experiment 7: Same flaw as Experiment 3.

Social scientists know that self-reporting is a flawed method of data collection; that's why the authors qualify with "exhibit". They're essentially making the claim about self-reporting, not behavior. The author's interpretation I sensed from the video marcus posted is that higher class individuals are less ashamed of unethical behavior, that they even feel justified. So your criticism is essentially a straw man.

Regarding Experiment 5: Are the p-values adjusted for multiple comparisons? If they have not been, an argument must be made why all the tests conducted are not independent.

I don't know. Did you read the methods section?

you can see more details in their supplemental file here:
http://www.pnas.org/content/suppl/2...pplemental/pnas.201118373SI.pdf#nameddest=ST6

There is no lie if the person does not answer the question. I'm not convinced this is unethical behaviour.

Lying is a simple behavior. Dishonesty is a more complex set of behaviors, which include selective omission. It's not damning to be neglectful of disclosure, you could just be inconsiderate... but it would count negatively towards a scoring in a bigger set of questions. That seems reasonable to me. More ethical people will be more concerned about disclosure and will make sure you understand you only have 6 months of job security.

The following are kind of silly criticisms, so I don't think they need to be addressed:

Regarding Experiment 4: I don't see anything unethical about taking candy from kids. It causes tooth decay!

Regarding Experiment 6: The experimenters clearly lied. Is that ethical? Also, no information is given as to whether the $50 voucher was given out.

 

[russ_watters]

...but we can criticize based on merit rather than an icky filling in your tummy.

Of course. The icky feeling is an indicator that something might be wrong, but it doesn't actually tell us if/what is. I've figured it out though:

Society, really. I think most of us agree that lying, cheating, and stealing are unethical.

But those aren't all that ethics is, right? That's the basic flaw in the study. They sought to prove the upper class is more unethical by focusing on unethical tendencies the upper class is more likely to have. If they had investigated unethical tendencies the poor might be more likely to have - or ones that were neutral to financial state, they might have come to the opposite conclusion. How about:
-Sexual harassment?
-Cheating on your significant other?
-Mooching?
-Treatment of others?
-Treatment of property?

It amounts to a begging the question fallacy.

Related, only the first two or three of the 7 studies actually attempt to as "if". The rest are designed to explain why. In other words, they already had there conclusion and were exploring that narrowly focused conclusion, not actually testing it to see if it was true.

The reason I pointed out that they didn't survey people in jail is because there are class-based differences between who is in jail and who isn't. Thus the people in jail have excluded themselves from the study and may have skewed the results.

 

[russ_watters]

That analogy isn't appropriate here. Putting on your mask first ensures that you are in a position to help others after and therefore isn't selfish. This paper purports to show that people who "put on the mask first" then stay sat not really helping those around them.

You didn't respond to the example give. Do you recognize that it is fact that the rich give more actual dollars to charity than the poor?

Furthermore thanks to marginal utility giving away a set percentage of a larger wealth has less of an impact on the owner of that wealth than someone with a comparatively smaller amount so the idea of taking care of oneself first doesn't hold out in that case.

You're missing the point. Marginal utility (of the rich person's cash) has nothing to do with it; it is a matter of behaviors that help you become rich. Those behaviors are not easily turned-off...although they eventually tend to be. For example, the so-called "robber baron" super rich people of the turn of the century; many of them ended up giving the lions share of their wealth to charity later in life - after they decided they no longer needed to be bringing money in.

 

[russ_watters]

Regarding Experiment 1 and 2: they had no evidence the driver was of higher social class. If the driver was a chauffeur, as is conceivable for a higher social class vehicle, he may have been of lower social class.

I have more, but starting with this one, I had a similar complaint. While it is certainly true that a Mercedes (in the US) is more likely driven by a rich person than a poor, what they didn't try to filter out was rich people who choose to drive Civics. They exist and their choice would directly impact the study. Not only because it messes up the sample, but also because such people would seem likely to be less showy-money oriented.

 

[Pythagorean]

But those aren't all that ethics is, right? That's the basic flaw in the study. They sought to prove the upper class is more unethical by focusing on unethical tendencies the upper class is more likely to have. If they had investigated unethical tendencies the poor might be more likely to have - or ones that were neutral to financial state, they might have come to the opposite conclusion. How about:
-Sexual harassment?
-Cheating on your significant other?
-Mooching?
-Treatment of others?
-Treatment of property?

It amounts to a begging the question fallacy.

Good point

Related, only the first two or three of the 7 studies actually attempt to as "if". The rest are designed to explain why. In other words, they already had there conclusion and were exploring that narrowly focused conclusion, not actually testing it to see if it was true.

I'm not so sure, they seem to use the appropriate language to separate the issue. For example, in chapter 7; they're not saying rich people are rich by being scandalous, they're saying even if give a poor person power and money and they become more demanding and entitled. They are considering character vs. situational attribution and here they're saying it's situational.

The reason I pointed out that they didn't survey people in jail is because there are class-based differences between who is in jail and who isn't. Thus the people in jail have excluded themselves from the study and may have skewed the results.

That would be very difficult to pull apart the interdependencies. Class and prison is discussed here a bit. One of the difficulties is that a man on the street can go to prison for robbing someone for $15, but Enron executive (which equated to magnitude more than $15) were able to settle with the government for $300-million and avoid jail time.

http://www.sagepub.com/hanserintro/study/materials/reference/ref11.1.pdf

I myself have been given the option to pay a fine or do jail time by a judge for driving with a revoked license. This sets a much different set of punishments for rich people than poor people. Rich people suffer virtually no punishment, while poor people would be forced to take the jail time, causing them to miss more days of work...

what they didn't try to filter out was rich people who choose to drive Civics.

Another good point I hadn't thought of.

 

[Evo]

I myself have been given the option to pay a fine or do jail time by a judge for driving with a revoked license. This sets a much different set of punishments for rich people than poor people. Rich people suffer virtually no punishment, while poor people would be forced to take the jail time, causing them to miss more days of work...

Please post proof that the judge you went before doesn't offer "poor" people the same option. I guess he has a dosier on each person stating their finances so he knows which people to send to jail?

Let's keep this factual please.

 

[Pythagorean]

I think you're missing the point? The same offer is on the table for everyone. I was the poor person; I took the jail time because I couldn't afford the fine and I didn't want to make my poor parents pay for it either. Someone well off would just be able to pay the fine.

 

[Evo]

I think you're missing the point? The same offer is on the table for everyone. I was the poor person; I took the jail time because I couldn't afford the fine and I didn't want to make my poor parents pay for it either. Someone well off would just be able to pay the fine.

Ah, I see what you're saying.

 

[Pythagorean]

If they had investigated unethical tendencies the poor might be more likely to have - or ones that were neutral to financial state, they might have come to the opposite conclusion.

I see the authors actually address this in their paper on page 4, bottom half of left column. Not surprising though; it's rather common knowledge in the field of sociology that there is a correlation between poverty and crime.

There is a "critical criminology" paradigm in criminology that describe how poverty can cause crime, and I think what the authors are really showing in this paper is that wealth causes unethical behavior. They do so by intervention, even, simulating wealth in poor people.

In fact, I disagree with you that they're showing that people got rich by being unethical. They're showing that self-perception social status causes unethical behavior, much like poverty causes crime. So it's not "wealthy people are unethical", it's "wealth causes unethical behavior".

Of course, this doesn't justify any prejudices. These are statistical claims. Individual particles are unpredictable, and we're probably all surrounded with anecdotal exceptions.

 

[atyy]

Social scientists know that self-reporting is a flawed method of data collection; that's why the authors qualify with "exhibit". They're essentially making the claim about self-reporting, not behavior. The author's interpretation I sensed from the video marcus posted is that higher class individuals are less ashamed of unethical behavior, that they even feel justified. So your criticism is essentially a straw man.

Their title is "Higher social class predicts increased unethical behavior".

The following are kind of silly criticisms, so I don't think they need to be addressed:

I don't see how their lying is a silly criticism in a study which purports to study ethical behaviour.

It also impacts their claims. If they make a claim which is obviously false (it makes no sense that you roll dice online, but those who set up the game don't know what is rolled), then those playing may know they are in an experiment in which they have been "set up", and may consequently lie because they just think they are playing the same game. That is not unethical (like in poker).

 

[Pythagorean]

Their title is "Higher social class predicts increased unethical behavior".

We were talking about the conclusion of two of seven studies in which self-reporting was involved. The title is a synthesis of all seven studies.

I don't see how their lying is a silly criticism in a study which purports to study ethical behaviour.

Because it doesn't invalidate the merit of the claims. Likewise, police lie to suspects to get them to confess... and similarly it doesn't invalidate the conviction. It may not be fair or right, but it doesn't invalidate the conclusion.

 

[atyy]

We were talking about the conclusion of two of seven studies in which self-reporting was involved. The title is a synthesis of all seven studies.

But 1&2 based on the vehicle status are flawed (with additional criticisms from you and Russ Watters), 2 studies are based on self-reporting, one is frivolous (the candy one), leaving only 2 studies. One of which is not clearly unethical if they simply don't answer the question about job security (eg. if they replied, "Sorry, I am not able to give you infornation about job security at the moment), leaving only the one study in which the experimenters lied. As you correctly point out, their lie alone does not invalidate the claim, so I added in post #19 a criticism of the methodology based on the lie.

Here's a specific scenario. Let's say the rules allow me to play the game multiple times, or allow me and a group of friends to play. By comparing my rolls with those of my friends, I will be able to tell that the game is rigged (all our outcomes are identical). Once I know the game is rigged, ie. those who set up the game consider that lying is fair, then why not play the game according to their rules. One may say, for example, that since the outcomes are known, neither lying nor telling the truth makes a difference, so why not just lie to mess with those who set up the game? We know that lying does not affect the fairness of the game for other gamers, so it is not unethical.

 

[Pythagorean]

We are not in complete disagreement about 1&2. I think 1&2 do demonstrate something about self-perceptions of social image, but not about wealth in general. Similarly, the self-reporting demonstrates something about image (wealthier people tend to be less ashamed of unethical choices). The candy one is not frivolous, it demonstrates entitlement (it doesn't really matter that it was for children). Entitlement is not unethical alone, but it would contribute negatively towards an overall ethical score.

As for your criticism in #19 (I apologize, I have limited slots of free time that pop up so I can't completely reply to every post every time, especially more nuanced posts).

It also impacts their claims. If they make a claim which is obviously false (it makes no sense that you roll dice online, but those who set up the game don't know what is rolled), then those playing may know they are in an experiment in which they have been "set up", and may consequently lie because they just think they are playing the same game. That is not unethical (like in poker).

Again, the unethical conclusion comes form the synthesis. All that's being showed in this one study is the propensity to lie. It's not automatically unethical from there.

The synthesis is that given some low level of competition, a certain situations is more likely to lead people to lie, feel entitled, are less ashamed of unethical behavior, and neglect to disclose information (like job security to an employee) in experiments, then I think it's ok to predict that they, in general, have a lower threshold for unethical behavior in nature, where competition level is higher (more to lose, more to gain).

Each of the studies independently does not demonstrate unethical behavior.

 

[marcus]

Hi Pyth, I am glad to see you mentioned the feeling of *entitlement*. That was the key word that Greg Bernhardt used in his original post the started the other thread we had about this research.

AFAICS the research was not especially biased or badly designed and AFAICS it uncovered statistically real effects. People can differ as to how they want to TALK about the correlations in behavior that the experimenters seem to have found.

I personally would not use the word "unethical" at this point because that gets into the topic of ethics. What strikes me is that there could be something analogous to the resolution of cognitive dissonance. If you find that you are better off or higher status then maybe there is a natural tendency to want to feel that you *deserve* this and that you have extra prerogatives privileges, are entitled to more, and perhaps even that ordinary limitations don't apply as strictly.

Not sure about the last, different people would presumably respond differently to perception of higher status.
Some might become more honorable and scrupulous, be inspired to generosity and self-restraint. Doubtless the effects are merely statistical.

Anyway I certainly don't dismiss their research--I found it interesting and thought parts of the video were actually entertaining. I'm glad you started this thread about it and appreciate your clear patient discussion of it.

 

[Pythagorean]

Thank you marcus. I think that there's a tendency for people to jump to "second-option bias" with complexity topics like sociology. We are given the first default option, but once we hear an ounce of valid criticism, we over-correct and want to dismiss the study.

I think the study deserves some criticism, but I also think people go overboard and start trying to throw every noodle at the wall to see what sticks in an attempt to be critical. I also think people are misinterpreting the purpose of the study and assuming the authors are committing fundamental attribution error.

 

[atyy]

The synthesis is that given some low level of competition, a certain situations is more likely to lead people to lie, feel entitled, are less ashamed of unethical behavior, and neglect to disclose information (like job security to an employee) in experiments, then I think it's ok to predict that they, in general, have a lower threshold for unethical behavior in nature, where competition level is higher (more to lose, more to gain).

Each of the studies independently does not demonstrate unethical behavior.

Thank you marcus. I think that there's a tendency for people to jump to "second-option bias" with complexity topics like sociology. We are given the first default option, but once we hear an ounce of valid criticism, we over-correct and want to dismiss the study.

I think the study deserves some criticism, but I also think people go overboard and start trying to throw every noodle at the wall to see what sticks in an attempt to be critical. I also think people are misinterpreting the purpose of the study and assuming the authors are committing fundamental attribution error.

So the question is the strength of the evidence in favour of their synthesis. Have they established their synthesis to the point where no additional studies are needed to confirm it? Or should we still regard their synthesis as a proposal (as the Higgs boson was until recently).

If it is still regarded as a proposal, isn't it important to consider alternative explanations of their data?

Also, should the synthesis should be weakened to "entitlement" rather than "ethical", as marcus suggests in post #23?

 

[Chronos]

I view this as a pure survival behavior pattern. If you have access to superior resources, will you refrain from accessing them to spare yourself from hardship? Similarly, are you inclined to refrain from using those same resources to avoid 'bad' situations? Is lying 'evil'? In what context? Is outsmarting, or duping a business adversary into a 'bad' deal 'evil'? Whom is to say what is 'bad'? Survival is an unending struggle.

 

[atyy]

http://www.sciencedirect.com/science/article/pii/S0304405X13001992
http://www.economist.com/whichmba/luxury-and-lies