Reading the fine print, episode 3: What’s in the bottle?

Caught out by the FDA.

I happened to be in the bug spray section of the store when I spotted a bottle of mosquito repellant that proudly proclaimed "100% 


But the FDA-mandated labelling tells a different story:

Active ingredients
N, N diethyl-m-toluamide      
Other isomers  

Similarly, foods labeled "zero fat" are actually allowed to contain up to a half gram of fat. (Well, up to but not including.) This is a definition of "zero" with which I had previously been unfamiliar.

(Episode 1, Episode 2.)

Comments (44)
  1. pcooper says:

    I noticed once that Tic-Tac mints had a serving size of 0.49 grams, which meant that had 0 grams of sugar and 0 grams of everything else, presumably since they could round those all down to 0.

  2. Moi says:

    I was just going to ask that very question. If something, say butter, was intended to be served in 0.49 gram portions, it would be "zero fat butter"? Say wha’?

  3. Jules says:

    "Other isomers" are probably DEET also, isomers being different structures that have the same components.  

  4. JS says:

    This is like a bad Slashdot joke. "100% DEET, for certain values of 100."

  5. Gabe says:

    Epsilon has to be some value, and 0.5 makes sense because it avoids adding a significant figure. I’m surprised that Raymond wasn’t aware of that definition, because I thought most computer programmers know that zero is actually any number that falls between negative epsilon and positive epsilon.

  6. Jeff Lewis says:

    I’m certain that you know you should always use an epsilon when comparing floating point numbers.  That’s all they’ve done here:

    #define EPSILON .5

    if (DBL_EQ(0, fatContent))


     cout << "Zero fat!" << endl;



  7. Miles Archer says:

    It’s impossible to have a jar of something that’s 100% any substance. To all intents and purposes it’s 100% – it may well be that it’s not possible to distill it any more pure that 95% due to azeotropes.

    However, it is kind of amusing.

  8. Raul says:

    The other five percent are probably still DEET, and just the "ortho" and "para" (as opposed to "meta") substitution isomers.

  9. mph says:

    Moi:  I think so.  I’ve seen cooking sprays, which are mostly or entirely fat (e.g. vegetable oil), claim to be fat-free because the serving size is so small.

    See, for example:

  10. AndyB says:

    I find it amusing that Raymond, an intelligent, articulate and sensible fellow, thinks you can get 100% of anything. Surely "close enough as not to matter" is as good as you’re going to get in the real world where nothing is absolutely exact.

  11. J says:

    The moral that I got from this post?  Never try comedy in front of nerds.

  12. Dan says:

    Raul – I agree.

    AndyB – to a certain point, yes, but 5%?  Try paying 95% of your mortgage and convince the bank that it’s close enough :)

  13. :: Wendy :: says:

    But why was Raymond in the Bug section of a store?  Surely he didn’t think that any percentage of DEET would resolve software bugs?  Is he working on some innovative approach to code de-bugging.  That’s WILD!

    Numbers are a descritipnve tool that can be misapplied.  Raymonds example is a wonderful example of how advertising,  law and common conceptions of what percentages mean have aligned to make for some bizarre labelling experiences.  Love it!  

  14. Johnny says:

    Bit of a tangent, but still related to food labeling.

    The Nesquik (Chocolate Syrup) bottle in our fridge has a label on the back which reads to the following effect:

    "Can be mixed with milk for a nutrious snack … (a sentence or two down) … contains all the nutritional value and goodness of a glass of milk."

    Presumably when added to milk, the milk continues to retain all of its nutritional qualities.

  15. RyanC says:

    Serving size is used everywhere to portray information in a manner that the manufacturer would prefer you see it.

    When you buy a 20oz bottle of Mountain Dew, do you drink the whole thing, or do you drink it 8oz at a time?  They bold things such as "Calories", "Total Fat", etc to draw your eyes away from the "Servings Per Container".

    110 calories doesn’t sound so bad.. 31g of sugar, again, doesn’t sound so bad.  Until you do the math and realize that 20oz bottle is 275 calories and 77.5g of sugar.  Yikes.

  16. Peter says:

    As the old computer joke goes:

    "Two plus two is five for large values of two, or small values of five"

  17. Moi says:

    The point is not whether you can have 100% of something or not, it is whether you should be able to claim that you do.

  18. Chris Moorhouse says:

    Well, that’s Raymond’s blog in a nutshell. Always interesting (for certain values of always).

  19. Karl says:

    As for claiming 100% of anything, think about it, are we better served with long warning labels discussing atomic weights of foreign matter or tiny print about statistic confidence intervals, all to hassle over the simple roundoff in everyday macroscopic life?  100% is a realistic, best effort, cost practical 100%.

    If you do, then I say just take the Microsoft approach to solving this: slap on a EULA to every product that says "by opening or using product, you agree to our definition of 100%".

  20. As annoying as it is that companies will abuse this threshold, there isn’t anything new with (zero < 0.5). In fact, in many CS applications, (zero < 1.0)- at least they aren’t just casting the amount of fat to an int.

  21. But who rounds 95 to 100?

  22. Claw says:

    But who rounds 95 to 100?

    Raymond: I think the point is that the other 5% is DEET as well; they’re just the other isomers of DEET, as pointed out by some of the previous posters.  The ortho- and para- isomers are also repellants, though they’re not as effective as the usual meta- isomer that makes up the 95%.

  23. SR says:

    So, if anything less than 0.5 can be rounded down to 0, then heavy cream (about 40% fat) can be advertised as "zero fat".

    0.4 < 0.4 after all!

  24. Claw: Hm, but the DEET fact sheet identifies the compound as "N, N-diethyl-m-toluamide", so they are explicitly specifying the meta isomer ("m"). Whatever those other isomers are, they aren’t DEET.

  25. njkayaker says:

    "DEET, N, N diethyl-m-toluamide,…". (Ie, a specific isomer.)

    I suspect that they would make it truely 100% DEET if they could. It’s funny but not a deception.

    "Serving size is used everywhere to portray information in a manner that the manufacturer would prefer you see it." I don’t think this isn’t exactly true. The serving size of drinks (at least) are a standard 8oz (usually).

    Anyway, anybody who thinks that a soda of any quantity is a "single serving" deserves to get fat. The reason that soda is sold as 20 oz is that it increases the profits of the soda manufacturers without having to increase the price. To maintain income when the components increase in price, you can either increase the price of the standard product size (12 oz), decrease the product size while maintaining price, or increase the product size with an corresponding increase in price.

  26. David Totzke says:

    Grabel’s Law:

    2 is not equal to 3, not even for large values of 2.

  27. njkayaker says:

    The other 5% is the same chemical. The label implies that the other 5% is active (but probably much less so).

  28. I’m not sure what fact sheet Raymond was using, but I went and looked up the EPA toxicology registration.  It defines DEET as a blend of diethyl toulamide isomers containing a minimum of 95% of the meta isomer as the technical active ingredient.  So, this looks like 100% Technical DEET.

  29. Jerry Pisk says:

    RyanC: Until you realize that the vast majority of drinks use 8 oz as the serving size. It makes it actually easier to compare various drinks to each other regardless of their serving size. Coke for example has something around 36 grams of suger per 8 oz, so Mountain Dew is slightly less sugary, assuming you drink a specific volume of liquids a day, not a specific number of various sized containers (ending up with a galon of liquids one day and half a quart the next).

    As for the FDA food labels – rounding down is pretty sneaky but then again it’s always better to check the list of ingredients, not the shortcut table.

  30. George Bailey says:

    Around here I can’t find an eight ounce bottle, but because of the serving size I keep having to throw away half of it.

  31. Jerry Pisk says:

    Well which is better, drinking your daily two quarts of something that comes in 20 oz bottles that have 78 g of sugar in them or something that comes in 24 oz bottles that have 90 g of sugar per bottle?

  32. Moz says:

    you’re comparing reals using the equality operator, Raymond. Bad user!

  33. Nicholas Allen: Well that settles it, then. DEET is "95% meta + other isomers", in which case the labelling is correct.

  34. Mike says:

    Could N, N diethyl-m-toluamide be a compound that reacts spontaneously to form other isomers of itself, with an eqilibrium constant Keq such that there will always be 5% of other isomers?  That’s at least what I’d say with an AP Chemistry-level knowledge.

    Any chemists out there who could confirm this (or refute my explanation as the blabberings of an idiot)?

  35. Ben says:

    The "zero fat" thing really bugs me as that also applies to trans fats.  If you care about what trans fats are, you probably want to consume none of them, and the 0g thing means you still have to go hunting on the ingrediants label for hydrogenated oils.  Especially since that aside from margarine, pretty much everything else with them gets to use 0g.

  36. Brooks Moses says:

    There’s another tricky thing going on in food labels that hasn’t been brought up yet, which I just noticed last week.

    I was comparing a couple of cans of soup, trying to figure out the relative calorie contents.  To a first guess from the labels, one can had slightly more than the other — but it was a notably smaller can.  So I investigated further, and discovered the following:

    Can 1: Net Wt. 14oz, calories per serving 200, serving size 1 cup, servings per container "about 2".

    Can 2: Net Wt. 19oz, calories per serving 180, serving size 1 cup, servings per container "about 2".

    Silly me, expecting that I could multiply the number of calories per serving and the number of servings per container and actually get something more than a rough guess!

  37. Gabe says:

    I think the proper thing to do here is apply some threshold whereby anything less than 0.5g and more the threshold would show up as "<1g" of fat. That way you could have something with 0.02g of fat reflect the fact that is has some fat without having to add confusing decimal places.

  38. Moi says:

    "trans fats"

    Fat chicks with d*cks?

  39. 8 says:

    It seems DEET is getting way too much over-used now. At this rate, it’s definately gonna stop working this century.

  40. Martin says:

    Alcohol-free beer is allowed to contain up to .5%vol of alcohol.

  41. Miles Archer says:

    So is alcohol free orange juice.

    Ever hear of fermentation?

  42. kbiel says:

    Brooks: Silly me, expecting that I could multiply the number of calories per serving and the number of servings per container and actually get something more than a rough guess!

    It’s usually not a good idea to try to compare to items using their weight and volume without knowing the density.  The denser item is going to weigh more (19 oz in this case) per volume than the lighter item (14 oz).

  43. David Walker says:

    Tangentially related to portion sizes:  

    I was at the store the other day and was buying a box of Equal (sugar substitute) packets.  The store brand had two sizes, 100 count and 200 count boxes.  The shelf tags that have the usually-helpful unit pricing were badly done.  

    The unit pricing is supposed to help you decide which size or which brand of any product is really a better value, by doing the arithmetic for you, so you can compare across sizes or brands.

    But shelf tag for the 100 count box showed the unit price "per each" (each packet), while the 200 count box showed the price "per ounce".  Not helpful.

    I did use my powers of arithmetic to realize that the 100-count box at $1.39 was a better deal than the 200-count box at $4.29.  If the counts had been different than they were, I might not have noticed.

  44. Eddy Boston says:

    "It seems DEET is getting way too much over-used now. At this rate, it’s definately gonna stop working this century."

    You’re confusing a repellant with a pesticide.  In the case of pesticides (or antibiotics), there is an evolutionary advantage to being resistant to it, since those that are less resistant will die.  In the case of a repellant like DEET, it doesn’t kill the less-resistant mosquitoes, just drives them away to more aromatic hunting grounds.  While this may make their evening meal a bit more of a challenge to find, I doubt it would contribute significantly to a development of DEET-resistant mosquitoes.  

Comments are closed.

Skip to main content