Preventing theft in the workplace: Field-experimental evidence from a peer monitoring intervention

How to cite

Thon, Max; Heinz, Matthias; Gürtler, Oliver; Sliwka, Dirk. 2024. "Preventing theft in the workplace: Field-experimental evidence from a peer monitoring intervention." AEA RCT Registry. AEARCTR-0014515. https://doi.org/10.1257/rct.14515-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Max Thon* , Matthias Heinz , Oliver Gürtler , Dirk Sliwka

* Primary investigator

Registry ID
AEARCTR-0014515
DOI
https://doi.org/10.1257/rct.14515-2.0
First registered
Last updated
Status
Completed
Trial dates
2024-10-01 – 2024-10-31
Countries
Germany ()
Registry record
http://www.socialscienceregistry.org/trials/14515

Abstract

We are collaborating with a restaurant chain in Germany. This company offers a diverse range of beverage and food products across its numerous restaurants throughout the country. A significant challenge faced by the company is employee theft, specifically waiters accepting payment from customers without properly recording the transaction in the cashier system. In our field experiment, we will implement a randomized peer-monitoring measure within these restaurants. Specifically, in the treated restaurants, the group of individuals involved in processing a particular order is expanded. If these products are not properly recorded in the cashier system, it will now involve other colleagues, preventing the waiter from concealing the transaction. Additionally, we will conduct an employee survey to collect data aimed at identifying the underlying behavioral mechanisms.

Intervention

During breakfast time, the restaurant generates a significant amount of revenue through its breakfast buffet. This product is highly vulnerable to employee theft. Specifically, employees take the order and send guests to the buffet, which operates as a self-service. However, this creates an opportunity for employees to confirm the order, allow guests to eat from the buffet without recording the transaction in the cashier system, and then take the money during the payment process. Hot drinks and other beverages (excluding orange juice, which prior to the intervention is part of the buffet) need to be entered separately, as these items are prepared by the bar staff. The bar staff receive the order via a printed message at the front station once it is entered into the cashier system. For the intervention, the self-service orange juice is removed from the buffet and replaced with a free glass of orange juice, prepared by the bar. However, the printed order for the orange juice will only appear if the buffet product is properly entered into the cashier system. Therefore, if waiters attempt to take payment from guests without entering the product in the system, they would have to convince the bar staff to give them the orange juice without a printed order, which is not the standard procedure.

Experimental design

The orange juice is removed from the breakfast buffet and prepared by bar staff at a random subset of restaurants (see for details above in the section intervention). The intervention aimed to start at 01 October 2024. However, for the intervention the orange juice is served in special containers (glasses), which need to be delivered. It depends on delivery times and logistics when the treatment is exactly rolled out. It is intended around October 1, and at the latest the roll out should be completed around October 8. We do not know yet, which restaurants already started. We expect to receive this information around end next week.

Design details

Randomization method
Randomization done in office by a computer.
Randomization unit
Restaurant level
Clusters
35 Restaurants (We needed to remove 5 restaurants as they could not take part in the experiment. The reason is that we found out that they were part of a confounding change earlier that year. They already had implemented measures similar to the treatment in January, which we had not been informed of before.) We may evaluate the treatment effects on the waiter level as it allows us to account for individual fixed-effects. However, we do not know yet how many waiters will work on the specific days.
Observations
16 in the control group and 19 in the treatment group
Arms
35 restaurants.
Primary outcomes
Number of breakfast buffets recorded in the cashier system
Secondary outcomes
Ratio buffet revenue to revenue from hot beverages, total revenue from breakfast buffets, total revenue during breakfast time, survey measures (workplace climate, job satisfaction, engagement, turnover intention, manager satisfaction), heterogeneity analyses with respect to previous monitoring intensity, turnover intention and manager satisfaction.

Cited by

No citation data available for this registration yet.

Keywords