One of my favorite types of audit was on bars, especially the smaller ones operated as sole proprietors or partnerships. These owners were usually really interesting folks who knew their business inside and out. The key to auditing them was knowing the industry well―you needed to understand the lingo, know about internal controls and be able to relate to the owners on their own terms.
The basic bar audit is all about accounting for units of product―bottles of beer, glasses of wine or draft beer and ounces of hard alcohol. Bottles are easy. Glasses of wine or beer are fairly simple―x number of bottles or kegs divided by number of ounces per glass. But accounting for the hard liquor becomes much more subjective.
Most small local bars use a “free pour” where the bartender manually pours from a bottle into the glass, perhaps with the aid of a shot glass. Patrons won’t return if they see the bartender carefully pour the drink into the shot glass and then dump it into the serving glass. They expect an over pour where the shot glass is only a “suggested” measurement and the bartender continues to pour past the point where it is filled.
Measuring in this manner makes establishing the cost of the drink very challenging. One of the questions asked of the owner or bar manager at the start of the audit is what they expect that average pour to be. Most times the response is between 1 .0 and 1.5 ounces, which is reasonable, though 1.5 ounces is getting heavy. But some owners insist they are pouring more than that and stand by their answer even when the rest of the audit analysis shows this isn’t possible. I’m sure they wouldn’t want to lead an auditor astray but as B.B. King once said, “Only my mother loves me and she could be jiving too!”
So when a taxpayer insists on claiming an unusually high pour and the audit calculations don’t support it, it’s time for an undercover pour test.*
The Undercover Operation
I was fortunate enough to have participated in 6 or 7 of these tests during my years as an auditor and they all followed the same general pattern. Three auditors would go out to the bar. Two would sit at a table or booth and the third at the bar. Each would order a drink that did not contain any mixer other than water (mixes would taint the measurements described later). While they enjoyed the first round, they would be observing exactly how business was conducted, number and type of patrons, etc. These observations would later be compared to answers provided by the taxpayer during the detailed interview at the start of the audit.
After a reasonable period of time, a second round would be ordered. This time when the drinks were delivered, the auditors would show their badges and identify themselves and ask to speak with the manager or person in charge. (In one of my undercover tests, once we showed our badges, the staff in the kitchen area slipped out the back door. They thought we were immigration officials.) Once the manager arrived to witness, the drinks were poured into containers, sealed and labeled and a receipt given to the manager. The type and proof of alcohol in each drink was also recorded.
Most tests were conducted in the evenings, which is generally the busiest time for the bars and thus most representative of the average operation. The following morning, the drinks would be tested at the audit department. The temperature of each drink was taken and recorded as was the total volume of the drink. Then some of the drink was poured into a graduated cylinder and a hydrometer was floated in the liquid. A hydrometer is an instrument used to measure the relative density of a liquid; we compared the density of the liquid compared to the density of water.
If you know the temperature of the liquid, and the proof of the alcohol, it is possible to determine the amount of booze in the drink using a series of published conversion tables. This measurement would be conducted at least twice, usually three times on each drink to establish the average reading. Never did the results come close to the pour size claimed by the taxpayer.
These average readings were then used to re-work the audit to arrive at the expected sales, which in turn would be used to establish the understatement in reported sales. This understatement would be assessed against the taxpayer. All of the taxpayers involved in these undercover tests either agreed with the findings or at least reduced the size of their claimed pour. These reduced claims were usually accepted due to the overall uncertainty still inherent in these types of audits.
NOTE: One taxpayer was so adamant that our measurements could not possibly be correct that it was agreed to have the drinks analyzed by a local lab. Our results were within 1/8 of an ounce of their findings; the taxpayer paid the additional tax and related interest.
* A brief description of the undercover pour test can be found at CA BOE Audit Manual for Undercover Pour Tests.
Other recent “Audits and Sales Tax” posts by Lloyd Geggatt:
- Software Audits for Buyers & Sellers – and the 3 Key Questions
- Luxury Audits: Empty Boxes Full of Champagne Wishes & Caviar Dreams
- Sales Tax Audits: Actual Basis Approach Can Result in "Win-Win".
- What Auditors Should Understand About Outsourced Sales Tax Returns
- Auditing Sales Tax Credits: An Auditor's Top 2 Considerations