Employee theft in restaurants August 27, 2013Posted by Cameron Shelley in : STV202 , trackback
The New York Times has an interesting piece about the effects of installing theft detection systems in restaurants in the U.S. Employee theft is a problem for any business but it is especially difficult in the food service industry where turnover is so high. Since servers do not expect to have a long-term relationship with the business, the temptation to defect from its interests is perhaps more acute than usual.
Researchers studied the impact of theft-detection software that monitors transactions for patterns that suggest employee misdeeds. When a suspicious pattern is detected, its is flagged for the attention of management.
As with any system that depends on pattern recognition, the system will make at least two kinds of mistakes: (i) flagging transactions that are not dishonest and (ii) not flagging transactions that are. The system in question, National Cash Register’s (NCR) “Restaurant Guard”, tends to err on the side of the latter error so that managers will not be swamped with false alarms:
The software is intentionally set so that a restaurant manager gets only an electronic theft alert in cases that seem to clearly be misconduct. Otherwise, a manager might be mired in time-consuming detective work instead of running the restaurant.
According to the research, the direct effects of this effort were minimal: Restaurants using the system saved about $108 per week in theft reduction.
However, the presence of the system also altered the behavior of the servers. Instead of gaming their employers, servers turned their attention into cajoling more money out of customers:
Knowing they were being monitored, the servers not only pulled back on any unethical practices, but also channeled their efforts into, say, prompting customers to have that dessert or a second beer, raising revenue for the restaurant and tips for themselves.
In business terms, the servers adapted by becoming more productive. The study suggests that sales per restaurant using Restaurant Guard increased by $2,982 per week.
This outcome is an intriguing one in a number of ways. For starters, it illustrates a frequent theme in this blog that technology is more than just a tool. That is, the Restaurant Guard system not only reduced employee theft, it also increased employee productivity.
It also illustrates the trade-off between false positives and false negatives (that is, fairness). On the face of it, false negatives are more problematic for managers since each one represents a successful employee theft. False positives are more problematic for employees since each one constitutes a false accusation of crime, causing stress and decreasing future employment prospects. Restaurant Guard is designed to generate relatively few false positives, because managers are willing to trade off a certain level of theft for time to perform other work.
The increase in productivity is also a kind of externality. After all, the point of Restaurant Guard to was change the relationship between employers and employees. It did that but it also changed the relationship between employees and customers, the latter spending more money on average than before. Is that good for the customers or not? Portions sizes in many restaurants are often large, so consuming even more might not be a great outcome for consumers. Or, do they just take the extra home for later? Or, just throw it out? The study does not explore this point.
Of course, as the article makes note, part of the incentive for employee theft in restaurants is that servers are paid so little. Perhaps, if the money spent on Restaurant Guard were spend instead on higher wages, the theft problem would be reduced without the intervention of big data. It seems that the study is silent on this point also.