Arguments Against Interracial Marriage Identical to Arguments Today »« Jackson: No Sexual Abuse With Mother and Father

Atheist Discrimination at the Post Office?

A German company that makes shoes for atheists — yeah, I don’t really get the connection either — noticed that a lot of the packages they sent to customers in the United States were getting lost in the mail, so they decided to do a little experiment and found that packages marked with tape that said “atheist” on it were far more likely to get delayed or lost.

Before all of this, Atheist Shoes closed their shipping boxes with tape that said “atheist” on it. So they sent double packages to 89 customers in the US, 178 packages in all, with one box having the “atheist” tape and the other having regular tape. The result: nine of the packages marked “atheist” got lost and only one of the packages without those markings got lost. And the “atheist” packages took an average of three days longer. One “atheist” box sent here to Michigan took 37 days.

Now this is a fairly small sample set, but it’s still significant. I don’t know what the rate of lost packages is on items sent from Germany to the US, but over 5% seems awfully high. And the fact that it was skewed so heavily against the packages marked “atheist” is highly suspicious. It would not be difficult for a post office employee at several points in the handling of packages to delay or “accidentally” lose the box and it certainly wouldn’t be surprising if that’s exactly what happens.

Comments

  1. slc1 says

    Of course, it could have happened in Germany. Before blaming it on US Postal workers, German postal workers would have to be exonerated. By the way, Prof. Singham commented on this yesterday.

  2. Michael Heath says

    Very weird they’d not only market shoes to people because they’re atheists, but also identify their customers as atheists with packing tape. The latter also seems especially stupid.

  3. Pierce R. Butler says

    And Prof. Myers commented on this Wednesday.

    Something tells me that a German company was probably using the same tape to wrap packages in Germany before it shipped any products to the US.

    Did this experiment involve sending American customers twice as many shoes as they’d ordered? Did they even put shoes in the test packages (atheists are very very sneaky, y’know!)?

  4. tomh says

    @ #3

    That was the whole point. “Each person was sent two packages: One sealed with ATHEIST-branded packing tape, the other with neutral tape.” ATHEIST-branded packages were 10 times more likely to disappear.

  5. says

    @slc1

    The article stated that “having run a series of control tests in Germany and Europe, which demonstrate no such bias, the problem appears to lie in the USA”. So it sounds like they did other tests to ensure that the European postal system is working properly.

  6. tomh says

    @ #1

    Their site also says, “having run a series of control tests in Germany and Europe, which demonstrate no such bias, the problem appears to lie in the USA and is likely explained by the differential handling of packages by the employees of the US Postal Service.

  7. Michael Heath says

    lt’d be interesting to compare the delivery failure rates for “atheist” along with, “Tea Partier”, “Pro-Life”, “Pro-Choice”, “Anti-Union” or “Christian Nation Patriot”.

  8. says

    Do “atheist shoes” have a Darwin fish logo on the bottom or something? Or are only made with leather from cows who weren’t made kosher or halal before being butchered?

  9. nooneinparticular says

    We all should be skeptical of this claim. Simple power calculations suggest to me that this sample is far to small to draw any conclusions. The Postal service handles a shit ton of mail. I don’t have time today to do anything more than rough calculations (and I lack some key data, like what is the average volume of packages the USPS handles from Germany). The success of on time (+3 days) delivery of mail sent via First-Class Mail International (the USPS’s international mail service) ranges from a low of 86% to as high as 98.5% (meaning where they sent the packages matters) but sending 178 packages will just not be enough to detect any differences. By my calcs (and I admit I do not have sufficient data to do this accurately – I did some quesstimations and averaged the on time rates) they would need to send between 2,000 and 10,000 packages (depending on SD you want) to hope to detect a real difference even with the power at 80%.

    *this is another problem – the only data I could lay my hands on was for letters. But the Atheist shoe folks sent packages. I have no idea if the delivery success of packages and letters from International sources are the same, though they are very close for domestic.

  10. Moggie says

    Michael Heath:

    Very weird they’d not only market shoes to people because they’re atheists, but also identify their customers as atheists with packing tape. The latter also seems especially stupid.

    Well, it’s affinity marketing. For a lot of people, shoes are more than just functional: they say something about you. Think of people who obsess about buying exactly the right kind of trainers, for example. In this case, you’ve got a group of friends who are atheist (not exactly unusual in Berlin), and who probably figured that they’d sell more if they had a distinctive “hook”.

    Personally, I buy my footwear from a company called “Vegetarian Shoes”, who I suppose might have similar shipping problems to parts of the US. If someone were to start an outfit called “Vegetarian Atheist Shoes”, I’d probably give them a try.

  11. Moggie says

    fifthdentist:

    Do “atheist shoes” have a Darwin fish logo on the bottom or something?

    The site is here. They appear to have two “message” soles: “ICH BIN ATHEIST” and “DARWIN LOVES” (loves what? shoes, perhaps).

  12. Sassafras says

    A German company that makes shoes for atheists — yeah, I don’t really get the connection either

    The reasoning appears to be “We are shoemakers and designers who happen to be atheists and wanted to do something fun to promote atheism so we used the skills we already have”. Not really more odd than atheist jewelry or bumper stickers, I guess.

  13. Moggie says

    nooneinparticular:

    We all should be skeptical of this claim.

    Yep. As a piece of science, it’s not great. But as marketing, it’s genius. They’ve now got a lot of people talking about their products, and they’ll probably have a big uptick in sales.

  14. Reginald Selkirk says

    fifthdentist #10: Do “atheist shoes” have a Darwin fish logo on the bottom or something?

    What’s the matter, is your Google busted?

    Modusoperandi #9: … Do you really expect the LORD to protect shipments that deny Him?

    Clearly this is a miracle, but a sad one for believers. At one time, Gold Almighty created the entire universe. Eventually he was reduced to killing thousands of newborns; after a few thousand more years, all he could muster was walking on water and casting out demons. Now the most he can manage is to lose some shoes.

  15. steve oberski says

    I don’t now about this, wouldn’t fundie posties assume that un labeled plain brown wrapped packages contained pornography and steal them anyway ?

  16. psweet says

    nooneinparticular – I think you’re looking at this the wrong way. They sent paired samples (one of each to each customer), and looked at the difference between those samples. A simple t-test with paired samples gave me a highly significant result with 89 pairs of random values and a difference of only 1.5 days between means.

  17. nooneinparticular says

    Thanks psweet. I did the same and got similar results. But I don’t think this is a fair test. The path either packages goes through in the USPS service are not the same. I am not really comparing the success of two packages relative to each other, but relative to ANY package. If you ask the question; what is the likelihood of any package getting delayed relative to any other package (this is the null) there is simply no way to do it with only 178 parcels.

  18. psweet says

    I don’t see why it isn’t a fair test. I can understand why you might be interested in knowing how those packages relate to packages as a whole, but it seems to me that if you send two packages from the same place and time to the same place, then any consistent differences in the paths they travel is part of the effect you’re looking for. The difference between the two samples is statistically significant — what are you attributing that effect to?

  19. says

    psweet, nooneinparticular has decided that he wants to reject the test, and is trying to find reasons to do so. You’ve presented a detailed statistical argument in response to his vague statistical reference to a “power test” and now he’s making more demands…that we need to know exactly the route taken.

    Basically, he’s not doing a statistical analysis. He’s doing advocacy.

  20. nooneinparticular says

    psweet – It isn’t the right test unless you can show that you can detect a difference in successful on-time rates for 89 packages (0.5 * 178) sent in a system that sends tens of millions. The null hypothesis is that there is no discrimination going on and that the difference they observed was due to chance. My (admittedly flawed by lack of data) power calculations suggest that they are off by one or two magnitudes in being able to reject the null hypothesis. Therefore, although the paired T-test does show a difference, the sample is powered enough to distinguish it from the null.

  21. Lynn Dewees says

    Has the USPS been contacted? What did they say? I don’t know how many packages this shoe company ships, but if they are losing revenue due to USPS actions, it seems there might be some monetary damages.

  22. Sastra says

    Thanks for the link, Moggie #13. The shoes look stylish and are supposed to be “kitten soft” — but I hesitate to buy shoes I haven’t actually tried on. My feet are fussy.

    I must say that I don’t care much for what looks like their atheist symbol (logo?) A great big black giant hole of nothingness.

    THE atheist symbol is, of course, the one I appropriated for my gravitar. It’s getting so prevalent I predict it will take over the Darwinfish and Dawkins’ A in the next 5 years.

    As for the boxes being deliberately “lost,” it wouldn’t surprise me. My guess would be that it happens more at the local level than in the mass distribution centers. Brave freedom fighters daring to stand up for God against Satan the enemy.

  23. caseloweraz says

    Sassafras: “The reasoning appears to be ‘We are shoemakers and designers who happen to be atheists and wanted to do something fun to promote atheism so we used the skills we already have.’ Not really more odd than atheist jewelry or bumper stickers, I guess.”

    Well, I’m glad to hear that. My first thought on seeing this post — not a great thought, I admit — was, “Shoes for atheists? I bet they have no soles.”

    And when I read the customer’s remark that it’s like wearing kittens on your feet, I thought, “Those shoes must bite and scratch.”

    I guess I’m just cynical today.

  24. tbp1 says

    I’m sure I’m not the only one here who is a member of the Freedom from Religion Foundation. I’ve noticed that their mailings use only the organization’s initials, FFRF, and no indication at all on the envelopes or wrappers as to the contents. I’ve always assumed that this is to prevent having their materials “disappeared.”

  25. psweet says

    I still don’t understand your point, nooneinparticular. The essence of sampling theory, I thought, was that the size of the population wasn’t important — the size of the sample is what determines the efficacy of the test. (Actually, in a very small population the size is important, since you start getting into independence issues, but that clearly isn’t a concern here.)

  26. says

    tbp1 “I’m sure I’m not the only one here who is a member of the Freedom from Religion Foundation. I’ve noticed that their mailings use only the organization’s initials, FFRF, and no indication at all on the envelopes or wrappers as to the contents. I’ve always assumed that this is to prevent having their materials ‘disappeared.’”
    Plus, it sounds like a fart. More of a “trying to sneak one off” than a “got up too fast to stop it” variant. That one belongs to the Friends of Onkaparinga Park or the Marine’s Forward Arming and Refueling Point acronyms.

  27. Darkling says

    In terms of analysing that data to see if the differences are “significant” I think using bootstrap sampling is probably the best approach. From the overall study the USPS loss rate is 10 from 178 or 5.6% so how likely are we to see results that were reported (9 lost for the labeled ad 1 lost for the unlabeled) if the probability that the USPS loses them is the same?

    I set up a quick simulation to generate data using the 5.6% loss rate for both the unlabeled and labeled groups and then replicated this a large number of times. Obtaining results as extreme as the observed data happened very rarely (<0.2%).

    There is an obvious discrepancy between the rates of successful delivery for the two packages. Given this does it really matter whether or not there's enough data to detect a difference in time to delivery? The fact that the atheist labeled packages are lost at a higher rate tells us something is happening in the USPS that shouldn't.

  28. BobApril says

    A guy on Google+ is attempting to replicate the experiment and provide a little extra data, using postcards instead of packages. I don’t know if he’s looking for more volunteers or not, but I’m hoping to see whatever results he comes up with.

    As an atheist postal carrier, this story pretty much pisses me off. Especially since I’m careful to follow the postal regs rigorously when delivering crap from Peter Popoff and his ilk.

  29. says

    “Now the most he can manage is to lose some shoes.”

    And he’s cuttin’ in on St. Anthony of Padua’s territory. It’s not a good idea to fuck with mafia saints, just sayin’.

Leave a Reply