Do rich people steal children’s sweets?

This article,  “Interesting: being rich doesn’t make you good! Now, there’s a surprise!“, is a good example of how stories propagate around the internet. Its author, Richard Murphy, described as “the UK’s #1 economics blogger,” [edit: see one of the comments for some info on this] reproduces an infographic he’s found and states

“…it also looks properly sourced, so I thought it worth sharing.”

Here’s part of it:

You can see the full infographic reproduced in the article, with many more similar claims.

Let’s take two of the claims, paraphrased in the context they’re presented:

  1. Rich people steal more sweets from children
  2. Rich people are more likely to cheat others out of their money at games.


I checked the link given at the bottom of the infographic: This article doesn’t name the  paper, nor does it even give the sizes of the studies.

So the source of this “properly sourced” item is itself not even sourced (I tracked down the paper, it’s Higher social class predicts increased unethical behavior, PNAS (2012).)


For claim #1, 129 students were given a picture of a ladder with 10 rungs and told, “Think of the ladder as representing where people stand in the United States … Where would you place yourself on this ladder relative to these people at the very top? Imagine yourself in a getting acquainted interaction with one of the people you just thought about from the top/bottom of the ladder and write a brief description about how you think this interaction would go.”

Shortly thereafter they were asked to take a jar of 40 sweets to a child-specific laboratory, and were invited to help themselves to a few if they wanted. On average the 129 students took 1 sweet, with a S.D. of 1.

No more numbers than this are given, but I would say the evidence for claim #1 is something like: 60 students who thought about talking to poor people mostly accepted a couple of sweets when offered them, and 60 students who were thinking about rich people mostly declined. Right.


Claim #2 is based on 195 people responding to a Craigslist advert to take part in an online study after which they would be entered into a draw to win a $50 voucher for an online retailer.

They were divided into rich or poor based on answers to a survey about their background (sorry, but how rich exactly were the richest people in this group to need to take part in a study hoping to win $50?) Participants were presented with a computer program which randomly rolled five dice, the entrant with the highest score winning the $50 gift certificate. Test-takers had to report the scores themselves, but didn’t know the program was rigged to always generate a score of 12. A full 85% of people answered honestly, but out of the 31 participants reporting getting more than 12, there were a few more “richer” people than poorer people.

I’d restate claim #2 to suggest that poor people doing online surveys are probably poor because i) they’re not being productive with their time, and ii) they might help themselves by doing things such as typing the biggest number possible into a box in order to maximise their chances of making themselves $50 better off.


Perhaps the conclusion here is that people are inclined to believe what they want to believe, and are less questioning of claims consistent with that. The guy writing that this looks “properly sourced” was voted the seventh most influential left wing thinker in 2010/11, and as such likely has a world view on rich-v-poor with which the infographic resonates. Indeed, many web and print news outlets ran with stories based on this study, using headlines such as, e.g., Rich people more likely to take lollies from children, presumably because it’s a headline that would catch the attention of many of us and appeal to our biases.


Other dodgy stats:

BBC and selection bias

Subscribe to my blog: