Marshall School of Business

University of Southern California

Controls at the Bellagio Casino Resort
Teaching Note
Purpose of Case
The Controls at the Bellagio Casino Resort case was designed with several purposes in mind.
First, it illustrates a control system that is dominated by action and personnel controls, rather than results controls. The analysis of this system leads to insights about some of the factors that limit the feasibility of results controls.
Second, the case can lead into a discussion of what is meant by the term “tight control.” The case presents an excellent example of the application of tight action controls in the table games areas of the casino.
Finally, the case can lead into a discussion of the meaning of what auditors refer to as “internal control,” which is a subset of the broader area of management control. The case describes an excellent system of controls over cash and cash-equivalent stocks and movements thereof.
These controls, which fall in the category of internal controls, are necessary but not sufficient to guarantee good management control.

Suggested Assignment Questions
Since the case can be used for several purposes, the assignment can be slanted in several ways.
The following is one list of questions that might be used
1. Focus on three key roles at mainly three levels of authority in the casino – blackjack dealers, pit bosses, and the vice president of table games. How would you characterize the “control strategy” (e.g., tight vs. loose) used over each of these roles?

This note was prepared by Professors Kenneth A. Merchant (University of Southern California) and Wim A. Van der Stede (London
School of Economics) for the sole purpose of aiding classroom instructors in the use of the Controls at the Bellagio Casino Resort case. It provides analysis and questions that are intended to present alternative approaches to deepening students' comprehension of business issues and energizing classroom discussion.
Copyright
2007 by Kenneth A. Merchant and Wim A. Van der Stede. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without permission.

143
© Pearson Education Limited 2012

Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructor’s Manual

2. Prepare a list of the controls described in the case. What control problems are they designed to address? Are the managers interviewed for the case justified in being proud of their company’s control system? Why or why not?
3. Are any of the control systems in place at the Bellagio suitable for firms in other industries?

Case Analysis
Key Roles and Controls
We suggest starting the class by analyzing each of the organizational roles and the controls over them. (This addresses Suggested Assignment Question 1.) Start with the lowest level role described in the case—blackjack dealer—and build the chart shown in Attachment TN-1 on the blackboard.7 (Some students are not familiar with the blackjack game, so some discussion of how it works and what the dealer is expected to do may be necessary.) One point that comes out of this chart is that results controls are not important at the dealer and pit boss levels.
One major categorization of control systems is tight or loose. Asking students whether control over the dealers is tight or loose will invariably lead to an answer of tight. Labeling any control system as tight or loose will lead to a discussion of what it means to have tight control. Our definition is that tight controls provide a high degree of certainty that the person(s) being controlled will act as the organization wishes. This is a judgment call, but we believe that the
Bellagio certainly has tight control over dealers and pit bosses. But because they do not use action controls and do not have good performance measures, they do not have tight controls over the vice president of table games.
A minor point that is interesting here is whether control was enhanced by the high salaries the casino paid to line personnel, particularly dealers. The argument would be that well-paid employees would be less inclined to get involved in illegal activities. The managers we spoke with, however, said no! They believed that they could not pay anyone enough money to remove the temptation provided by the large sums of cash present on the casino floor.

Evaluation of Controls
The evaluation of the sets of controls in use has to be based on some consideration of what can go wrong. (This addresses Suggested Assignment Question 2.) A provisional, but not exhaustive, list of things that can go wrong at the blackjack tables is presented in Attachment
TN-2. Ask the students to build such a list (supplemented by items on the list shown where necessary) and think about whether the controls in place were sufficient to reduce the incidence of each undesirable activity to a reasonably low level.
After some discussion, instructors can summarize by stating that casino managers are aware of the risk of these problems (although new schemes are constantly being invented), and they have

7
As mentioned in the case, the levels of casino management above the pit bosses, such as shift managers and the vice president of table games, did little direct supervision of the gaming activity. They were mainly involved in trouble shooting (e.g., resolving cases of malfeasance), keeping good customers happy, resolving special problems that arose (e.g., staffing issues), and improving the casino operations. Below the pit bosses, on the other hand, there is another level of supervision by so-called floor supervisors.
However, to avoid over-structuring of the discussion, we recommend to essentially focus on three key roles or levels: dealers (front personnel), pit bosses (supervision), and vice president of table games (management).

144
© Pearson Education Limited 2012

Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructor’s Manual

taken steps to reduce their incidence. As a consequence, the evaluation of the controls in place should be very favorable as long as cost constraints are kept in mind.
Another problem that casinos face is “skimming.” Skimming means that money is taken from the casino, such as cash from the cage or counting room often requiring cover through forged documentation and changed audit trails. This problem can be used to lead into a discussion of internal control and the ways in which cash is controlled in the casino. The two major purposes of internal controls are (1) to safeguard assets and (2) to ensure reliability of the financial records. In the area of cash, the casino maintains imprest cash balances to safeguard the cash and chips; this means assigning individual responsibility for a certain amount of money. And they have elaborate procedures to ensure that the financial records are reliable (e.g., multiple approvals of transactions, checking of audit trails, multiple reconciliations, and trend analyses), some of which are described in great detail in the case.
Do these procedures prevent skimming? The answer is a tentative yes. The controls are effective against skimming unless collusion is involved, which is a frequent limitation to internal controls. Weakness of Results Controls
As part of the discussion of Suggested Assignment Question 2, instructors should point out that results controls are not very important in the casino, and it is important to spend some time to understand why this is so. Results controls are simply not feasible for control over the dealers because results are not measured until the 24-hour drop procedure is complete, and during that period more than one dealer has worked at a table. It would be possible to measure dealer productivity by keeping track of deals per hour, but this is not done. Modern technology, such as RFID, actually would make this quite feasible, but the Bellagio management has so far decided against implementing such a system.
Instructors may wish to spend some time discussing the advantages and disadvantages of RFID technology that could be used to track deals. Beyond relatively easily quantifiable costs and benefits (e.g., the cost of replacing the chips vs. the benefit of being able to design better customer comp programs); however, the managers felt very strong about perhaps the most significant “hidden cost” of RFID, best expressed by the following quote from the case:
“Having too much information—knowing to the penny who wins or loses how much—isn’t always the best for the customer. At some point, the customers might lose some entertainment value if we monitor every little thing that happens. And we’re not sure that we understand all the privacy implications of this just yet. What if the IRS comes to us and asks for this information? We are in the entertainment business, and so we shouldn’t do anything that diminishes customer enjoyment.”
Results can be measured for individual pits and, of course, the casino as a whole, and these measures could be used for control over pit bosses and casino managers. But even those results controls are not used because they are not considered good indicators of the desirability of actions taken. One key indicator is the hold percentage (win/drop), but as is discussed in the case, this is a flawed measure. Even if the handle could be measured instead of the drop, the hold ratio would not be a reliable indicator of the desirability of the actions of the person being controlled because of statistical variations. On any given day, week, or even month, poor results might be blamed on a “high roller on a hot streak.”

145
© Pearson Education Limited 2012

Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructor’s Manual

As an illustration of the limitations of the results measures, instructors can refer to Exhibit 12, which indicates rather large drop variances across the 24 blackjack tables listed (from 28.52% to +37.71%). Drop is a measure of the gambling activity at the tables, but negative drop variances cannot be taken to mean that the dealers were not productive. The drop is dependent on the table’s minimum and the number and type of players who chose to play at that table on that shift. To illustrate, ask if the drop results for blackjack Table 19 in Exhibit 12 (variance percentage of 28.52) suggest that the dealers on that table that day were doing something poorly (e.g., slow dealing) or dishonestly (stealing money before it was put into the drop box)?
If managers wanted to check the details of what happened at that table that day, they could examine a videotape of the action.
The results measures are more reliable, the larger the increment of the casino being measured
(e.g., an entire shift vs. a single table) and the longer the periods of time involved; hence, the
GCB’s requirement to investigate, on a monthly basis, all statistical fluctuations by game type in excess of ±5% resulting from the comparison of the previous calendar year to that of the current month. Reasons for the deviations could include the activity of customers whose play materially affected the results of the month (the so-called high-roller-on-a-hot-streak explanation); the effects of any changes to the rules, types of wagers, or game play procedures; the effect of any errors or mistakes made during the operation of the game during the month; the effect of any thefts or other improper acts by employees or patrons; or any other unusual occurrences during the month being reviewed.
And there is one other problem with the results measures that is mentioned in the case. Pit bosses and casino managers can resolve disputes by deciding in the casino’s favor, thereby increasing the daily hold percentages. But sometimes, up to a reasonable limit, it is more profitable to decide in favor of the player because if they are kept happy, they will continue to play in the house. For the same reason, no bonuses were provided to managers on the basis of win, as Rudloff explains in the case: “We obviously don’t want games where the win is too high, as that might jeopardize the enjoyment guests derive from gambling and coming into our casino in the first place.” These are examples of a form of short-term vs. long-term trade-offs.

Applicability to Other Industries
Are the controls observed at the Bellagio Casino applicable to other industries? (This addresses
Suggested Assignment Question 3.) The casino industry is in some ways unique, but some of the principles included in the Bellagio control systems are of use in other firms. Tight security and good controls over movements of assets are necessary where valuable, liquid assets are involved (e.g., gold in semiconductor firms, money in banks). And personnel and action controls, such as those used in the casino, will dominate where good results measures are not available. Managers in each business and company have to go through the process described here of thinking what is wanted and what can go wrong and then designing a set of controls.

Pedagogy
Our experience with this case suggests that the discussion should be relatively structured. A few students have never been in a casino and do not know how the game of blackjack works, and they may be reluctant to admit their lack of knowledge. As a consequence, the instructor should probably start class by asking a knowledgeable student to explain the basics of the games of blackjack and the layout of the tables and to act as an industry expert.

146
© Pearson Education Limited 2012

Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructor’s Manual

After the background information has been clarified, the discussion should proceed smoothly.
The casino business intrigues most students, and the energy level of the class is generally high.

Attachment TN-1: Analysis of roles and controls
Role

Key actions

Controls

Blackjack dealer - Accurate, honest, deals
- Productive
- Cheerful

Good, tight control;
- Licensing (P) some risks
- Hiring (P)
- Training (P)
- Frequent rest breaks (P)
- Supervision (A)
- Surveillance (A)
- Documentation (A)
- Mystery shoppers (A)

Pit boss

- Good supervision
- Licensing (P)
- Marketing (give comps to - Hiring (P) good customers)
- Promotion (P)
- Training (P)
- Documentation (A)
- Surveillance (A)

Good, tight control; have to (and probably can) trust these experienced people, given sufficient action controls, such as through creating transaction trails

Vice president of table games

Decisions:

Loose control

- Licensing (P)
- Hiring (P)
Hiring
Labor scheduling (when - MBO/bonus system (R) to open a pit)
When to change minimums Casino layout

Code:
R = results control
A = action control
P = personnel control

147
© Pearson Education Limited 2012

Evaluation (in light of things that can go wrong)

Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructor’s Manual

Attachment TN-2: Things that can go wrong at a blackjack table

1. Dealer overpayment, when bets are paid or change made.
2. Dealer payoff of a losing bet.
3. Player peeking at dealer’s card or dealer signaling (showing the card).
4. Player “past posting” (i.e., adding a bet to hand that looks promising) or “copping” (i.e., decreasing a bet on a hand that looks lost).
5. Dealers stealing chips. Dealers have discovered many devious ways to remove chips from the tables. Common schemes have been to “cup” chips in their hands or slip them into
“subs,” such as elaborate hairdo, false bandages, or hidden pockets in their apron.
6. Players counting cards. Skilled blackjack players can keep track of the aces and face cards that have been played and gain and advantage in their betting.
7. Players introducing counterfeit money or chips.
8. Pit bosses granting bad credit. (The casinos protect themselves by having this decision made by personnel in the cashier’s cage. Communication to the tables is done through computer terminals.) 9. Players stealing chips from each other (commonly by having an accomplice to create a diversion). 10. Players marking or switching cards.

148
© Pearson Education Limited 2012