Wednesday, January 28, 2015

Don't Write Blog Posts Late at Night

The original version of this post had the math backwards, as an astute reader pointed out. Sorry about that. The real news is that, based on a sample of 450 products analyzed by the U. S. Treasury Department’s Alcohol and Tobacco Tax and Trade Bureau (TTB) for the 2014 Alcoholic Beverage Sample Program (ABSP), 31 percent of the beer, wine, and spirits products sold in the United States do not comply with TTB rules. The 2013 results were about the same.

The 2014 sample included 190 distilled spirits, 155 malt beverages, and 105 wines. One-hundred-thirty-nine products were non-compliant; 73 distilled spirits, 46 malt beverages, and 20 wines.

The most common offense? Alcohol content that did not match the label. In some cases, the actual alcohol content places the product in a different tax classification.

Another common violation is labels that do not match their approved Certificate of Label Approval (COLA) due to changes that are not allowable revisions. Approximately 25 percent of the non-compliant labels contained such changes in information. As least one product in the sample went on the market without ever submitting a COLA.

Is 69 percent compliance good enough? Is that even a fair way to read this report? Four-hundred-fifty is a small sample size when TTB reviews more than 100,000 labels, advertisements, and other material each year. But it’s the only measure we have of TTB’s effectiveness at "assuring that alcohol beverages sold in the United States are properly described on the container."

(My statistics consultant advises that a sample size of 450 should be 95 percent accurate, regardless of the size of the universe.)

I know what TTB’s mission is because I read it in the TTB labeling brochure, under the heading “How TTB Protects the Public.” It begins:

“American adults who enjoy an occasional alcohol beverage of their choice do so without fear that the product they are consuming might not be labeled properly. Why don't they need to worry? Because a small Government agency takes pride in assuring that the alcohol beverages sold in the United States are properly described on the container.

“TTB takes tremendous pride in its strategic mission to ‘Protect the Public,’ which is designed to assure the integrity of alcohol beverages in the marketplace, verify and substantiate industry member compliance with laws and regulations, and to provide information to the public as a means of preventing consumer deception.”

I apologize to everyone who read the original post. My problem with math is that I often miss obvious mistakes, especially late at night. I should also give you the link to TTB's original release, so you can figure it out for yourself.


Anonymous said...

Don't you have the title backwards? Shouldn't it be "Three-Fourths of Alcoholic Beverages DO Follow Government Rules". Basic math here seems to suggest that if 139 products were non-compliant out of 450 tested, that would be about 31% don't follow the rules.

Anonymous said...

Where are you getting 71% from?

450 products, 139 non-compliant.
That's about 31% non-compliance, not 71.

I'm not saying this isn't high.

Chuck Cowdery said...

I've fixed the post but am leaving up the comments that pointed out my mistake. Thank you.

Anonymous said...

You're telling me a link to a pdf you posted on your blog says some crap about some crap you can't prove off of your own site? Can't believe I have a hard time believing that. AND you had to correct yourself?! Noooooo(sarcasm)ooooooo!!!!!

Ben said...

This is interesting, because in the supplement industry the estimate of non FDA compliant labeling is even higher than 30%.

Anonymous said...

Aren't they just testing the very easiest things a college kid with a chemistry set could test?

Alcohol content? One piece of paper matching the other?

Does compliance really boil down to just these two things?

What about the hard stuff like is it whiskey, four years old, made where you think it is? Is it really natural cinnamon flavor in there? Does the approved formula match the label and the liquid?

Anonymous said...

"The most common offense? Alcohol content that did not match the label. In some cases, the actual alcohol content places the product in a different tax classification."

That aspect is very interesting considering (aside from basic inaccuracy) the only way it would hurt TTB is if the Mfg labeled the ALC/VOL lower than actual ALC/VOL of the bottle’s contents, thus the Mfg would be paying less tax than due. But in that case, why would a Mfg bottle and distribute something higher in ACL/VOL when such a thing could have yielded them a higher retail price at the higher ALC/VOL? On the other hand, a product with a lower ALC/VOL (in the bottle), when the actual product label shows a higher ALC/VOL would mean the Mfg paid more tax to the TTB than they owed, thus lost profit for the Mfg. Since neither of these make any sense on the Mfg side; as both impact them significantly, I can only conclude that "someone" at the Mfg has truly fallen asleep at their assigned task as it pertains to quality control. But then again, could we be seeing the bottled content’s ALC/VOL lower than the Label’s ALC/VOL, because after the product was Mfg and the company paid their tax per proof gallon, they decided to “water down” their product to cheat the consumer and increase quantity of product for overall profit? I really wish I knew the stats as far as how many bottles were found lower than the label, and how many were higher than the label. I will likely be using my Proofing Hydrometer to check every bottle when I open it, just to satisfy my consumer curiosity! 

Pittsburgh, PA