• Recent Frauds in Hotel / Tourism Operations
• Quantifying the Loss - A Study of Eight Properties Operating in the Hospitality Sector in new South Wales 
• Factors that May Inhibit the Recording and Reporting of Fraud 

Abstract 

In  1998,  losses  through  fraud  and embezzlement by associates and employees at all levels  property. corporate and ownership levels -contribute to cost pressures. but there are limited studies that quantify the degree of loss. This paper discusses issues of internal control, especially some cultural factors that may inhibit the recording and reporting of fraud, and attempts to quantify' the possible cost effects of fraud and embezzlement on hospitality operations by using aspects of Wanhill's model (1994). The paper puts forward examples of recent frauds. and describes a study of eight properties operating in the hospitality sector in Australia. 
 
 

Introduction - The Extent of the Problem 

The manager of a motel in Canberra. Australia, recently commented on the subject of fraud, - "We 
don't like to admit the extent of theft in our industry, because of the special relationship we have with our guests. Customers come and go in a retail establishment, but in a hospitality operation, guests stay for an extended period and they trust us to ensure their safety. If the public realised how much employee theft goes on in hotels etc., their trust would be shaken." 

It may not be easy to actually see people -employees  -  committing  theft,  or  fraud,  or embezzlement. But can we prove anything about it? For a start, the extent of fraud and embezzlement surprises many. 

The Australian Institute of Criminology puts the cost to the Australian economy of fraud and misappropriation of funds at $AUS 3.5 billion per year, almost equal to the projected national budget deficit for 1997/98. Adding the expense of the criminal justice system that deals with fraud, the cost grows to $AUS 18 billion, or 4% of GDP. 

The accounting group KPMG surveyed companies in Australia on the issue of fraud in 1997. Of the frauds noted by the respondents, 77% were committed internally. The highest number of frauds  were  carried  out  by  non-management employees,  although  where  managers  were involved, the amount of the fraud was much higher. 

The conclusion appears clear - fraud is a serious issue and there is a need for management to improve internal control measures, but there are questions for the hospitality industry - Are fraud and embezzlement as prevalent in hotel and tourism operations as in commerce generally? Do they constitute a serious problem, and if so. what is the degree of likely loss? 

The questions are urgent ones, for in 1998, hotels and tourism operations are under pressure to contain  costs  arid  maintain  rate  structures, particularly in Southeast Asia and the associated region at a time when currency exchange rates are volatile. 
 

Recent Frauds in Hotel / Tourism Operations 

It is not difficult to find some instances of recent fraud in hotels and tourism operations. 
 

This bears out the Australian experience. Research by the Major Fraud Group of the police of the state of Victoria revealed that one half of those charged with fraud in 1997 were accountants or office managers. 

However,  it may  be that the  more substantial losses to an organisation are not through such publicised scams like these, but through quite small, "routine", regular embezzlements of small amounts. These do not catch the public's attention, and their imputed existence is often tolerated or ignored by senior management. 

A further question arises - Can and should management take more concerted action against minor acts of embezzlement? 

The Degree of Loss 

What is the likely loss to a hotel of minor fraud regarding stock, or wage payments? Is it worthwhile taking action? 

Wanhill (1994) has drawn on the work of Johnson (1983) and Mars and Nicod (1984) and estimated that losses in food and beverage areas, through "fiddles" and "knock-offs" by staff who consider such practices a part of their entitlements, of  up  to  5%  of  inventory   would have disproportionate effects on break-even revenue and contribution margin and could "seriously affect the viability of the business". A Coopers and Lybrand study (1994) estimates that losses due to pilferage etc., could be up to 8 or 9% of cost of sales. A 1997 study of member companies of the New South Wales Employers' Federation, including hotel and tourism  operators,  showed  that  merchandising operations accepted "stock shrinkage" of 2%. 

A study of Victorian retailers found that for the majority of operations, effective security control could lead to a 20% to 25% increase in net profit. 

In the area of wages costs, the American Payroll Association has estimated that fraudulent practices. such as unauthorized overtime or errors in clocking on and off, inflate payroll costs of the typical hotel by between I % and 1.5%. 

Are these estimates of losses due to fraud and embezzlement of 5% of inventory and 1 to 1.5% of payroll costs reasonable, possible and likely? 

Quantifying the Loss - A Study of Eight Properties Operating in the Hospitality Sector in new South Wales 

Examining financial statistics such as cost of sales and payroll details is a delicate task. Clearly, this data is highly confidential because of its commercial applications. For that reason, it is not  feasible  to  survey  businesses  and  ask management to disclose their statistics if the information is to be published. However, the author has had involvement with a number of properties in the course of accounting/audit work, and can put forward a  number of statistics  provided  the institutions are not identified or identifiable. The study therefore remains an indicative study only, but  some  conclusions  are  put  forward  for discussion. 

The intent of the examination is not to draw conclusions, for obviously the study group is far too small and the methodology is inadequate for that purpose, but the intent is to find some indicative percentages that can be used in drawing up a model. The model may or may not be general, but at least it is a model which can be applied in any property if figures for inventory shrinkage or payroll inflation can be imputed. 

Background to the Study 

In the New South Wales State Budget of 1995/96, the State Treasurer announced a new tax on New South Wales community clubs. These clubs, some of which are very large and multi-site, provide  entertainment,  resort  accommodation, gambling, bar and restaurant facilities for members and guests, and have hitherto been exempt from taxation on the basis that they are non-profit and any surplus on activities was returned to the community in one way or another. 

The new tax was on poker machine turnover in  excess  of one  million  dollars,  and  the announcement of the tax made many clubs review the financial aspects of their operations. 

The author was  involved directly  and indirectly on a professional basis with eight clubs in New South Wales. These clubs, some of which were in financial difficulty, sought to review and revise operations within the new financial parameters. Some of the clubs were associated, and the new taxation regime meant that gambling revenues were amalgamated and they would incur a liability for tax, so they were considering breaking up the associations into smaller entities. 

As the clubs had similar food and beverage operations, supplying the same range of alcoholic beverages and liquors and conducting broadly similar food services, it seemed that financial data such as cost of sales and payroll should be comparable and could provide insights into the experience of fraud and embezzlement in this sector of the hospitality industry. The incidence of fraud and embezzlement is otherwise difficult to quantify for various reasons, some of which are discussed later. 

The financial data is obviously of a very sensitive nature, and hence the clubs are listed according to an informal code, and no facility is identified nor intended to be identifiable for reasons of commercial confidentiality. The data presented, which is generally amalgamated and disclosed annually to club members, is examined only to discuss  possible  losses  through  fraud  and embezzlement and is not intended for any other purpose. The aim of the examination is to seek indicators of the presence of fraud or theft, nothing more than that. 

The clubs are not necessarily typical of other hospitality operations such as major hotel chains because all of them rely upon revenue from gambling, such as poker machines (although NSW public hotels are also relying more upon gambling revenue  from  poker  machines  after  recent legislative and regulatory changes). Hence the price of food and beverage products can be subsidised, as it were, by gambling revenues. This means that the cost of sales will generally be higher than in an operation conducted  on a full-cost recovery plus profit basis. For all of the clubs, there is also a large amount of "Other Revenue". This includes club membership fees, revenue from entertainment, or revenue from sources such as green fees for a country' club. 

Methodology of the Study 

Extracts of the income statements of the eight clubs were recorded and these were then dissected in a common form similar to that of the Uniform System of Accounts. Although similar, the recording of payroll costs had to be different from that of the Uniform Accounts, for three of the eight clubs recorded payroll costs under one heading and did not dissect it under direct and administrative costs, so it was not possible to provide a common format without amalgamating payroll costs for the other five clubs. 

Cost  of Sales  was  recalculated  as  a percentage of food and beverage revenue, and Payroll costs were recalculated as a percentage of total revenue. 

The club percentages can be examined in two ways: 

(1) - To determine whether any of the clubs have significant variances from the average of cost of sales and payroll percentages for the eight clubs. Because the sample is so small, mean deviations should not be tied, but instead the absolute percentage difference will be considered. 

A significant variation in excess of the average for all the clubs may indicate that the higher cost of sales or payroll are affected by unauthorised usage, such as wastage or pilferage, i.e. there is fraud or embezzlement occuring. 

(2) -  The spread of the percentages for cost of sales and payroll, if not explained by other factors, may be an indicator of the degree of fraud or embezzlement. 

The following tables summarise various data. Table I lists income statistics from all eight clubs and Table 2 draws out relevant percentages of cost of sales and payroll. 
 
 
 
 

Club 8 appeared to be the best run club, and returned a net income of 13.8% (which was not distributed as a profit, but held in various reserves and used for club and community purposes). Club 8 had a cost of sales of around 45%, as did club I (which returned 10.1% in net income). Club 7. which returned 4.6%, had cost of sales of 43.5%. Hence, cost of sales of around  45% seems reasonable for the industry, allowing for some subsidisation of meal costs through other club activities, notably the gambling revenue. 

These three clubs, clubs 1, 7 and 8, also showed unexceptional payroll figures, which are calculated in terms of total revenue because most of the clubs did not show Food and Beverage payroll separately at this reporting level. 

Club 5 also returned a good result in terms of net income (13.4%) but the higher cost of sales figure of 48.2% has been partially masked by the lower than average payroll statistic of 34.1 %. 

It is difficult to explain why one figure should be higher than average and the other much lower than average  In this case it may be  that management accepts a higher cost of sales in providing higher quality product and may incur higher than usual transport costs, although these were not immediately evident on a walkthrough of the property. 

Club 6 also performs well, both in terms of cost of sales and in terms of payroll percentage, but there is a proviso on its reported performance. Clubs 3,4 and 6 are related, and there is a reason why club 6 returns such a good percentage net income (15.3%), besides the good cost of sales figure and the mid-range payroll figure. The reason is that club 4 may have been allocated too high a proportion of fixed costs and club 6 allocated too low a figure.  It appears that club 6 has good management, and club 4 indifferent           management, so it may have been that the management of club 6 were more successful in putting an argument to the joint board of the three clubs that the allocation of fixed costs should be as it is - to the advantage of club 6 over both club 4 and club 3. 

Club 3, the smallest of the group under examination, returns a net loss, but that is mainly because it has allocated to it high fixed costs, partly because it is a new facility in a small town and the buildings bear high finance charges. 

The clubs that have problems are club 2 and club 4. 

Club 2 has a cost of sales of 52.4%, a high percentage, although the payroll percentage at 30.38% is less than the average for the group. 

The club returns a surplus of 4.9%, and it may be  that the board and senior management have not been under severe pressure to contain costs and have allowed the high cost of sales to continue. It may be that the products are not priced correctly, or that the high proportion of gambling revenue has allowed management to maintain a high subsidy on food and beverage prices. However, despite the evidence of tight management of payroll, it must be asked whether the high cost of sales is due in part to pilferage and wastage of inventory. 

Club 4, which is in a difficult financial position, has almost as high an inventory cost at 49.3%, and in addition its payroll costs are higher than average, exceeding 40% of total revenue. Both percentages.   taken   together,   imply   poor management of costs, and there must be a concern that there is a lack of adequate internal control over both inventory and payroll. 
 
Indicators  

(1) A first point to make, and it is a tentative one, given the very small size of the study group, was that the three newer properties were among the poorer performers. Clubs 2, 3 and 4 were all less than three years old. Generally, they had higher fixed costs, partly reflecting the higher costs of building of recent times. The other clubs, presenting statements prepared on a historical basis, did not record as high fixed costs. 

It may be that the high fixed costs obscured higher than usual operating costs at board level. It may also be that the newer properties did not have longer term data do use as comparisons for their pricing and costing policies. 

(2) Of the eight properties examined, one was  certainly  poorly  run  and  there  were opportunities for fraud and embezzlement that would have explained both higher than average cost of sales and payroll costs. A second property may have had problems with control of inventory. In casual conversation with managers and directors of these two properties, no one would consider the possibility of theft or fraud by their own employees. Loyalty is, of course, a virtue, but it can also impede an objective examination of a situation implied by statistics. 

Of the remaining six properties, there was a consistency of cost of sales between four and the other two were explainable. This very consistency suggests that aberrations should be the subject of a close and careful examination from the internal control/audit point of view. 

One or possibly two cases of inadequate internal control out of a group of eight properties is an indicator that the incidence of fraud and embezzlement may be significant. 

(3) The range of variation between the highest and the lowest figure for both cost of sales and payroll (10.5 percentage points between high and low cost of sales, and 10.3 percentage points between high and  low payroll) bears further examination. 

Such a spread cannot wholly be explained by discounting or transport costs in relation to beverages, for the most common products in clubs (beer etc) are similar and suppliers' prices are generally very competitive. There are reasons for a spread in food costs, because obviously the quality may differ, but once again there does not appear to be a sufficiently wide variety of food services in clubs to explain the difference in inventory costs. 

The spread of payroll costs is also unusual in an industry where generally bar and food services staff have their wages and conditions determined by an  award  that  is  applicable  in  all  similar workplaces. The only substantial cause of variation must be in hours worked, rather than in the hourly rate paid. 

The indicators are that ordinary hours and overtime hours must be subject to variation, and the question of control of these hours, especially overtime hours, arises. 

All eight properties had manual systems for recording hours,  even  if the  calculation  and payment of wages were computerised. In each case, employees submitted wage claims by means of time sheets or timecards, and these time sheets were subject to checking and approval by supervisors, often to a pre-prepared roster. However, overtime was generally undertaken on short notice. 

In the three clubs where there were above average payroll percentages, the management of each  believed  that  there  was  "almost  no overstatement of overtime", yet all three cases relied on these manually compiled time sheets. 

At this stage of the discussion, when the questions of control of payroll and opportunities to defraud by excessive wage claim arise, it is appropriate to consider the recent experience of US hotels which installed fully automated employee attendance systems. 
 
Implications from the Installation of Automated Payroll Systems  

Two cases where manual systems were replaced by automated systems indicate that this estimate could be a conservative one. The first case involved the Los Angeles Biltmore Hotel. After a two-lock automated system was installed, hotel management found that payroll costs declined even though occupancy rose. Overtime claims were more tightly controlled, and the overtime budget fell by 76%. 

The second case involved the three hotel-chain of JC Resorts in California. The installation of a similar system reduced unscheduled time, and eliminated the possibility of "phantom" employees on the payroll. 

Both these cases clearly indicate that hotels can lose a substantial portion of payroll costs through employees "stretching the limits" as it were, and embezzling small payments that they are not entitled to. 

What if overtime is 25% too much, or 50% too much, or 75% too much (as in the Biltmore's case)? 

Applying A Model of Inventory / Payroll Loss 

Professor Stephen Wanhill has developed a model for evaluating the impact of restaurant fraud on income statements. 

Wanhill considers losses through "fiddles" (e.g. taking cash through not recording sales) and "knock-offs" (physical theft of goods). 

In the examination of club figures above, both "fiddles" and "knock-offs" will affect the cost of sales. 

The basic income model for a restaurant is: 

where A is net income, R is revenue, I is other income, Cv is variable costs (i.e. assuming one half controllable expenses such as payroll and also cost of sales), and Cf is the remainder. 

Wanhill develops the model to take account of fiddles as a percentage of receipts. as: 

where a is the amount of fiddles of revenue. 

However, using the model in another way, to consider the effect of "knock-offs", where S is stock, and Ev  is other variable expenses, and d% is the amount of stock taken, then: 

The effect of theft of stock is that every dollar's worth of stock stolen not only has an effect on revenue, but has a direct effect on net income N.  In other words, every dollar stolen is a dollar's less profit, or a dollar's greater loss. 

Could loss by embezzlement be as high as 5%, even 8% of cost of sales, as postulated by Wanhill in sample calculations? 

The evidence from the examination of eight New South Wales clubs indicates that it could be, for the  range of variation  in  cost of sales percentages  is  10.5%.  Perhaps  one  or  two percentage points could be accounted for by wastage,  or  inadequate  paperwork  to  record inventory movements, but that still leaves up to 8% difference between highest and lowest cost of sales in the eight clubs. 

The direct relationship between increasing fiddles and knockoffs demonstrated by Wanhill means that frauds and embezzlement, even though a small percentage of cost of sales, translates into a large percentage of net income. For that reason, this examination of theft indicates that the problem is a significant one. 

As the problem can have significant effects, why is it not more widely recorded and reported? The answer may be that certain "cultural" factors peculiar to the hospitality industry inhibit the recording and reporting of fraud. 
 
Factors that May Inhibit the Recording and Reporting of Fraud  

• A  Reluctance  to  Consider  Certain  Behaviour Fraudulent 

When is a gift a gift and when is it an act of attempted bribery to give one party an advantage in dealing with purchasing agents of the enterprise? 

• A Reluctance to Admit that Fraudulent Practices could be Committed 

A report by the Crime Prevention Council of Victoria estimated that 75 to 80% of inventory losses were due to employee theft,. but 95% of the frauds  detected  and  referred  to  police  were committed by customers or guests - an indicator that employers may be turning their backs on the possibility of employee fraud. 

A US study found that only  18% of fraudulent actions were uncovered by internal audit and control - in fact the most common way that frauds are discovered is through anonymous tip -offs. 

• A Reluctance to Enforce Internal Controls on Employees 

• A Reluctance to Prosecute 

Conclusion 

In Wanhill's words, the practices of fraud have often been "condoned by management", as long as the Chef or the assistant manager "delivers the right percentages". The problems arise because fraud, " if left unchecked can result in substantial losses and the liquidation of the business." The small-scale study of eight clubs in New South Wales indicates that fraudulent losses in cost of sales and payroll, even if seemingly minor, can have substantial affects upon the bottom line. 
 

Albrecht, W.. Howe, K., Romney, M., (1984) Deterring Fraud - the Internal Auditor's Perspective. Florida: Institute of Internal Auditors Research Foundation 
Challinger, D. (Ed.) (1987). Studies in Shoplifting, Victoria: Australian Crime Prevention Council, pp9 1-104 
Hotel/Motel Security and Safety Management Journal 1992 to 1996 
Mann, P., (1997) Survey of Company Fraud, Melbourne: KPMG Forensic Accounting Unit 
Wanhill, S.R.C. (1994). The Economics of Restaurant Fraud, International Journal of Hosnitality Management, Vol 13 No 3 pp265- 273