Quantcast

«

»

Oct 22 2013

Who do you trust to tell you lies?

Here’s a simple experiment: separate two people. Give one person, the sender information about two packets of money, a small sum and a larger sum. Let them then tell the other person, the receiver, about the two packets, and give them the choice of which one they can have. Will the sender lie or not, in order to trick the receiver into picking the smaller packet?

The result in college students in Canada is that roughly 50% of the subjects lie. Before we go on, be sure to put this in context: this result is only valid in one particular culture. Related economic tests have found that there are many other cultures that value giving over receiving, and that would skew the interpretation of this one, so we’re not looking at broader human properties, but solely at the properties of products of one narrow culture. OK? OK.

The interesting result is a finer breakdown of the individuals who were more likely to lie. Mostly, no consistent associations were found, except that membership in any of these three groups were more likely to predict lying: a) Business majors, b) children of divorce, and c) people who say religion is important in their lives. I’m sorry, business majors, but (a) doesn’t surprise me at all. (b) I would not expect; I wondered if financial insecurity played a role, but they report no correlation with socioeconomic status or student debt. Hmmm.

Again, (c) does not surprise me in the slightest. I’ve known too many Christians. Sorry, believers, you’re not all bad, but man have you got a lot of hypocrites in your ranks. I would actually expect that because parasites are more likely to choose to blend in with the dominant group.

I would suggest a variation on the experiment, though. Pair the senders; I’d guess that those students professing a strong religious belief would show a strong reversal, being far less likely to lie, if a co-believer were there to witness. On the other hand, I’d bet that two business majors acting as senders would high-five each other with a successful lie. Put a couple of Libertarians in there, though, and they’d grab both packets and sneak out of the room.

Yeah, I have a hierarchy of well-earned cynicism.

48 comments

Skip to comment form

  1. 1
    Crimson Clupeidae

    I trust me to…wait, what was the question again?

  2. 2
    Becca Stareyes

    I also wonder if you tell a Christian that the receiver is a co-religionist if sie would be less likely to lie*. My hypothesis would be ‘yes, sie would be more likely to tell the truth’ because humans tend to trust those we view as ‘like us’, even when the groups have grown large enough that we have no way of relying on observed behavior or even gossip.

    OTOH, I suspect a lot of open scams rely on that human instinct: that con-artists know that if they come in waving a cross and decorating their words with God language, that believers will assume they are part of the tribe and can be trusted.

    * With controls of ‘different religious beliefs, which would include a lack of religious beliefs’ and ‘religion is unmentioned’. And controls of non-Christians running the same test of ‘shares my views on religion’/'does not share my views of religion’/'is of an unknown religion’. The last might be interesting, since there is a tendency for many USAians to assume Christian by default.

  3. 3
    Nathaniel Frein

    My wife’s response is that she would lie so that she could get the larger packet, then give the other person enough from the larger packet to made them both even.

  4. 4
    aarrgghh

    “never do business with a religious son-of-a-bitch. his word ain’t worth a shit — not with the good lord telling him how to fuck you on the deal.” (william s. burroughs)

  5. 5
    everbleed

    Nathaniel…

    Your wife just might be a control freak. How you doin’ bro’?

  6. 6
    chigau (違う)

    I’d skip the lying part and make a deal with the other person to evenly split the money.

  7. 7
    Dave, ex-Kwisatz Haderach

    I’m not a Libertarian, but I’m still leaning toward grab both packets and sprint for the door.

  8. 8
    Anthony K

    What a waste of money. Like we needed science to tell us that Canadians are damned liars.

    [Narrows eyes suspiciously at chigau (違う)]

  9. 9
    Sven

    The 50% thing doesn’t surprise me, but for different reasons: what if the guy with the packets anticipates that the second person assumes that he is lying, so he tells the truth to mislead him? You really just eventually get that scene from The Princess Bride and it all becomes 50/50 anyway, right?

    Vizzini: But it’s so simple. All I have to do is divine from what I know of you: are you the sort of man who would put the poison into his own goblet or his enemy’s? Now, a clever man would put the poison into his own goblet, because he would know that only a great fool would reach for what he was given. I am not a great fool, so I can clearly not choose the wine in front of you. But you must have known I was not a great fool, you would have counted on it, so I can clearly not choose the wine in front of me.
    Man in Black: You’ve made your decision then?
    Vizzini: Not remotely. Because iocane comes from Australia, as everyone knows, and Australia is entirely peopled with criminals, and criminals are used to having people not trust them, as you are not trusted by me, so I can clearly not choose the wine in front of you.
    Man in Black: Truly, you have a dizzying intellect.
    Vizzini: Wait till I get going!

  10. 10
    Menyambal

    “Religion is important in my life because I scam religious people out of their money.”

  11. 11
    keinsignal

    If you’re doing business with a religious son of a bitch, get it in writing. His word isn’t worth shit, not with the good Lord telling him how to fuck you on the deal.

    William S. Burroughs – “Words of Advice for Young People”

  12. 12
    chigau (違う)

    Anthony K
    Honest. I’ll split it with you. 50-50.

  13. 13
    Leo Buzalsky

    For (b), I’d be curious if exposure to an environment where lying was perhaps a common occurrence would be the major factor. Do parents who get divorced lie more than parents who don’t? I’d expect so, though that would need investigation of its own. I would like to think that such an environment would teach a child a lesson that it is important to be honest, but then that assumes children (and humans in general) to be rational. That leads me to fear that the opposite is true — such an environment may teach that lying is a part of life, perhaps even acceptable.

  14. 14
    irisvanderpluym

    I would love to see an analysis of this data that accounts for political orientation. Those who would pick “none of the above” and arrange a deal to split the money are my kind of (lefty!) people: looking for the WIN-WIN. Business majors, who tend to skew conservative and are therefore pathologically obsessed with hierarchy (WIN-LOSE)? Me no likee.

  15. 15
    Jackie, all dressed in black

    A friend of mine has a business degree and she often complained at the lack of focus on ethics in her major. Profit always comes first. Obeying laws is further down the list of priorities.

  16. 16
    robro

    (b) might be the result of emotional insecurity rather than financial insecurity. Even if the divorce is fairly open and mutually agreed to, even if there isn’t a lot of lying in the marriage, these events could have a big impact on children.

    It would be interesting to see how religious people in a culture that does value giving over receiving responded to this game.

  17. 17
    ftltachyon

    One obvious question –

    Do we trust the statistics?

    There were 400 subjects. If you take an n=400 sample and test it against dozens and dozens of different hypotheses, purely by chance a few of them will turn out to have some correlation.

    And we know they tested at least “sex, age, grade point average, student debt, size of return, socioeconomic status, and average time spent in religious observation”. They probably tested a bunch of other things – at the very least, choice of major, parents marital status, but who knows what else. No wonder they got some positive correlations!

    I’s easy to let confirmation bias get the best of you and not question the results if you can think of a just-so story for why This Particular Group is more likely to lie (and you basically always can, making up stories is fun and easy). Has anyone here looked at the original paper and checked that their statistical approach is actually sound and didn’t just throw up false positives?

  18. 18
    naturalcynic

    I’d guess that those students professing a strong religious belief would show a strong reversal, being far less likely to lie, if a co-believer were there to witness.

    Sounds a lot like the old joke about Mormons on a fishing trip:
    How do make sure that all the beer gets consumed?
    Bring one Mormon.
    How do you make sure that you get all the beer for yourself?
    Bring two Mormons

  19. 19
    joe321

    Pair the senders; I’d guess that those students professing a strong religious belief would show a strong reversal, being far less likely to lie, if a co-believer were there to witness.

    This reminds me of one of my favourite fishing jokes:

    Q: Why should you always take two Mormons with you when you go fishing?

    A: If you bring only one, he will drink all the beer!

  20. 20
    PZ Myers

    Yeah, I don’t trust the statistics all that much.

    Also, as Anthony K points out, the study is marred by the extensive inclusion of Canadians.

  21. 21
    janiceclanfield

    But, but lying is sinful!

  22. 22
    twas brillig (stevem)

    Re OP:
    I totally misunderstand. mea culpa, to me, (a) business majors are more likely to lie than to predict lying. I’d think the business majors would more likely Suspect lying, than predict it (accurately). Do I have it completely backwards? When you said these groups are more likely to predict lying, are you saying they are more likely to accurately predict lying or just that they always expect that they are being lied to; that they tend to expect the sender will want to keep the larger sum and try to give away the smaller one even if they have to lie about it? Reading comprehension: FAIL.

    Another question about the study: ( I know I should read the paper, but I haven’t yet) did they correlate the lying behavior to the difference between the two envelopes? I.E. were they more likely to lie if the difference was large (e.g. a=$1, b=$1000) vs. when the difference was small (e.g. a=$1, b=$2)? [If they didn't, that'll be my proposal ;-) ]

  23. 23
    grouperfish

    I’d like to point out that lying is important and often a necessary part of life, but being deceitful is bad. These are different things, and the experiment is testing deceit.
    In the case of divorce, I can image any number of instances in which two divorced people would be less than honest with each other in order to keep the peace. Why bring up whether you are dating someone else right now if you know it will just hurt the other person? Why bring up that you just inherited some money if it will just make the other person more bitter about the financial settlement? Similarly, when working in a job, I’m not required to be honest with my co-workers about certain aspects of my personal life (dating, marriage prospects, kid prospects, illnesses, etc). Call them white lies or whatever, but it’s not the same. I’ve never felt like I’m required to be honest with people that might mis-use important information.

    As for the experiment – well, that’s the hitch, you can’t split it 50/50. Its not an option. So its sort of an artificial scenario, right? And I bet if the interaction was repeated, there would be more truth telling. Single events don’t much promote cooperation.

  24. 24
    johnharshman

    Stevem: What PZ means by “predict lying” isn’t that somebody will predict that someone else will lie. He means that some characteristic of a person correlates with an increased probability that person will lie. Nobody is making predictions except, in a scientific sense, the data.

  25. 25
    jamessweet

    Hmmm, in this contrived setup I’m not even entirely sure I feel that it even is immoral to lie. It’s a game, and it’s not at all clear to me that the rules of this game forbid subterfuge.

    If we’re playing poker, and I lead you to believe I have a better hand than I actually have, is that the moral equivalent of lying? Obviously not. So deception as part of game-playing is not immoral per se. I’d have to read the actual paper to get more details I guess, but my sort of feeling here is they gave the senders the money and then sort of let them do what they wanted in terms of the receivers. I see no particular reason why the senders were morally or ethically compelled to be truthful in this case.

    It might be different if the researchers emphasized that the senders were expected to be honest, that it would spoil the experiment if they lied or whatever. But as it stands, I don’t really see this as a test of willingness to lie for financial gain.

  26. 26
    protomoralist

    Re Anthony K: “…Canadians are damned liars”

    As a Canadian, would you believe me if I said this is true?

    Mostly, I tell the truth. But then, the best lies are dressed in a garden of truth.

    Also, re b): maybe lying, deceitful children are the cause of their parents’ divorces. A good liar could get the parents to blame each other. But since it didn’t work for me, maybe I’m not a good liar. You believe me, right?

  27. 27
    cuervocuero

    Not only were the participants Canadian, but university students!!!11! Would a factor in sharing/keeping/gambling be the economic status of the students and whether they were functioning on tuition costs and a steady diet of ramen? Does the ‘loaf of bread when starving doesn’t count as theft’ come into play at all?

    Although, even if I was in a money-poor student situation, if given opportunity to communicate with the receiver to deceive them into taking the lesser funds, I’d still be tempted, as others have, to hope for ability to take the greater and then split it with the other participant to even it out.

    I wonder what that says about self trust vs trust of others to do the altruistic thing for you, or whether finding the 2nd person likable(kin/kind group?)/financially vulnerable or not would affect sharing evenly after taking ‘altruistic’ control of the pay out parameters.

    Sidenote: Part of me wonders if the test holders wondered if religious types would ‘let jesus take the wheel’ and gamble as to who got the larger packet. If the results are really significant, it seems religious people tilt more towards ‘trust in allah but lock up your camels’.

  28. 28
    Bronze Dog

    It occurred to me to try and reverse the situation: “I’ll take the one you say is bigger and split the total 50-50 if you’re telling the truth.” But that’s not much of an incentive, since if they lie and I pick the smaller amount, they can just run off with the bigger.

    It occurs to me that the results could change by expanding the scenario over multiple rounds where liars risk developing a bad reputation. I recall something about a similar experiment where one person proposed a split. If the other person agreed, they’d split the money as agreed. If the other person refused, they both got nothing. If it was a one-time thing, it’d be rational to accept any proposal, no matter how unfair since you still got something, rather than nothing. With several rounds, people will deny unfair proposals to punish the proposer, essentially deferring their own gratification for greater benefit in future rounds.

  29. 29
    aarrgghh

    “Never do business with a religious son-of-a-bitch. His word ain’t worth a shit — not with the Good Lord telling him how to fuck you on the deal.” (William S. Burroughs)

  30. 30
    Nathaniel Frein

    Nathaniel…

    Your wife just might be a control freak. How you doin’ bro’?

    Just fine, fuck you very much.

  31. 31
    chris61

    I wonder what proportion of receivers would chose the ‘smaller’ packet because they assumed the sender was lying.

  32. 32
    Jadehawk

    When you said these groups are more likely to predict lying,

    PZ didn’t say that. He said membership in these groups predicts whether they will lie. It means there’s a significant correlation between being in these groups and the likelihood of them lying.

  33. 33
    ck

    Children of divorce being more likely to lie doesn’t surprise me. I’ve seen more than a few cases where the divorced parents are highly adversarial and attempt to use the child to get back at each other. I can imagine the impact that could have on a person.

  34. 34
    palefury

    Since when is P=0.061 statistically significant?????

  35. 35
    anuran

    Did anyone bother to read the whole article?
    Didn’t think so.
    Here’s the part I find interesting

    “We find that sex, age, grade point average, student debt, size of return, socioeconomic status, and average time spent in religious observation are not related to the decision to lie,”

    So the practice doesn’t seem to make one more or less religious. It’s the public profession or protestation of piety.

  36. 36
    anuran

    Gaaah. Not “religious” but “honest”

  37. 37
  38. 38
    Hairhead, whose head is entirely filled with Too Much Stuff

    My father was a minister for thirty years. His favourite saying was, “When the devil takes over the church, he will enter from the pulpit . . . or the choir loft.” (He was referring to the absolutely deadly rivalries between the ladies of the choir.)

  39. 39
    F [i'm not here, i'm gone]

    So the practice doesn’t seem to make one more or less religious. It’s the public profession or protestation of piety.

    Yeah. That’s where most of the trouble with religion comes in. For the most privileged, legislating their piety is a particularly public and enduring performance.

  40. 40
    robinjohnson

    I’m an atheist, a physics graduate with no business sense, and the product of a broken home. (b) surprised me at first, but I’m not stumped. I was around lies from a very young age, had various parents and step-parents who would Cause A Scene at the mention of each other, and quickly learnt that there were situations where an omission or a lie was the most likely thing to avoid severe unpleasantness. I wouldn’t like to think that made me willing to lie for gain, but it probably did make lying less instinctively black-and-white wrong.

  41. 41
    Acolyte of Sagan

    I have never told a lie in my entire life – except for this one ;-)

  42. 42
    Bicarbonate is back

    Robinjohnson,

    Lying is not necessarily wrong, it is often the right thing to do. When the Nazis come to the door and ask if you are hiding Jews in the attic, you don’t say that you are.

    What is right or wrong about lying is the reasons for doing it, not the fact in itself.

    The same goes for telling the truth.

  43. 43
    Moggie

    Bicarbonate:

    Lying is not necessarily wrong, it is often the right thing to do. When the Nazis come to the door and ask if you are hiding Jews in the attic, you don’t say that you are.

    Unless you’re Kant (see On a Supposed Right to Tell Lies from Benevolent Motives).

  44. 44
    Giliell, professional cynic -Ilk-

    Moggie

    Unless you’re Kant (see On a Supposed Right to Tell Lies from Benevolent Motives).

    Der kan(n)t mich mal
    (he can kiss my ass)
    German Students (Austrians probably, too) get way too much Kant and Nietzsche quoted at them during their highschool years.
    Lying can be moral, but lying for your personal gains when somebody else has a damage is.

  45. 45
    Giliell, professional cynic -Ilk-

    not

  46. 46
    David Marjanović

    What a waste of money. Like we needed science to tell us that Canadians are damned liars.

    Epimenides the Cretan said all Cretans are liars.

    There were 400 subjects. If you take an n=400 sample and test it against dozens and dozens of different hypotheses, purely by chance a few of them will turn out to have some correlation.

    Well, surely they did a Bonferroni correction for multiple testing, or something like that. Everyone does, right?

    Oh crap. Here is the paper.

    No trace of a correction. The methods section is ridiculously short.

    And this is most of the results section:

    Those subjects identified as majoring in business were 18.1% (p-value 0.016) more likely to send false messages than those in other disciplines. It could be that these students are more prone to lying by nature or training. It could also be that individuals strongly motivated by financial returns, and therefore more likely to lie for a monetary payoff, are more likely to pursue an education in business. This is discussed in detail with respect to the high returns treatment in Childs (2012b).

    Subjects whose parents were divorced were 29.3% (p-value 0.012) more likely to lie. This matches the classic work of Turkat (1994) who finds when parents divorce unamicably a child’s likelihood of anti-social behavior increases. Subjects raised by a single parent are 39% (p-value 0.002) less likely to lie for a monetary gain. These results should be treated with some caution as there were only 25 subjects whose parents were divorced and just 14 raised by a single parent (and only 3 whose parents were not divorced). This does suggest a promising avenue for future research with specifically recruited subject population, understanding that university students raised by a single parent may not be representative of all children of single parents.

    The importance of a subject’s religion increases the likelihood of lying (3.7% per point of importance, p-value 0.061). This is surprising as the Abrahamic religions most common at the University of Regina [footnote: "We did not ask subjects to specify their religion."] promote honesty as a virtue. It may be that subjects for whom religion was important feel separate from other students at this largely secular university. The impact of group membership and religion’s impact on lying warrant further investigation.

    The level of incentive had a surprising impact on the likelihood of a subject sending a false signal. Subjects facing low returns to lying were 23.6% (p-value 0.089) more likely to lie than those facing high returns. This marginally significant result contrasts with the work of Gneezy (2005) in which subjects were became more likely to lie as the return increased.

    The surprising impact of incentives on the decision to lie in these experiments is easier to understand when the salience of the incentive is considered. Due to the experimental design and timing of the sessions, subjects could not be paid immediately following their decisions. Instead subjects had go to the Department of Economics office and present a receipt to receive their pay in a sealed envelope. Overall, 110 out of 191 subjects came to get paid. Those subjects who came to collect their payoffs were 30.9% (p-value 0.021) more likely to lie than those who did not collect their payments. Finding that subjects who found the incentive salient were more likely to lie is consistent with Gneezy’s (2005) finding that the likelihood of dishonesty rises as the value of the reward rises.

    The salience of incentives, as shown by whether or not the subject came to get collect payment, varied by the level of returns to dishonesty. In the high return treatment 77 of 94 retrieved their pay while only 33 of 97 in the low return treatment did. This is suggests that many of the subjects in the low return treatment did not find the incentives salient. Those subjects claiming their payments in the low incentive treatment were 46.5% (p-value 0.000) less likely to lie than others. These subjects clearly did not see the reward as salient and exhibited an aversion to lying.

    It looks like economists – the journal is called “Economics Letters” – put the arbitrary cutoff for p-values at 0.1 instead of 0.05, report “p-value 0.000″ instead of “p-value < 0.001″, and are pretty bad at peer review.

    German Students (Austrians probably, too) get way too much Kant and Nietzsche quoted at them during their highschool years.

    All I had to read of him was Zum ewigen Frieden, which is an awesome work that makes me quite surprised Kant was such an asshole on other things.

  47. 47
    David Marjanović

    Correction: it looks like we didn’t read the whole thing.

  48. 48
    John Horstman

    Re: b), it’s a HUGE problem that “divorce” is simply treated as a single categorization. My guess would be that in acrimonious divorces, children wind up with anti-social behaviors (including lying) modeled for them at a much higher rate than married families (even if the marriage sucks, the partners will presumably be trying to make it suck less, as they apparently resigned themselves to not getting divorced), hence higher rates of anti-social behavior exhibited by the children themselves. I doubt that happens in (relatively) amicable divorces. It’s the acrimony that leads to the divorce (or results from it) that is the problem, not the divorce per se. It’s just that acrimony is more likely in divorced families, as cause, effect, or both regarding the divorce.

Comments have been disabled.