Is real also ideal? A Field Experiment on Inflation Adjustment in Employee Performance Measures

We will collect data from a retail organization operating supermarkets in Germany. The experiment involves eleven supermarkets and focuses on department managers and their subordinates in the bakery, beverages & wine and spirits, dairy, fresh food counter (butchery, cheese, fish), food, as well as fruits and vegetables departments of each of the eleven stores. These department managers (and thus also their respective subordinates) will be randomly assigned to a control group and a treatment group. We will distribute two different sales reports (nominal and real sales) to them and observe their responses and measures to the inflation-adjusted reports.

To date, department managers receive a weekly sales report of their respective department from the store manager that shows department-level information on gross sales for the week, the percentage sales change compared to the previous year, and the absolute average receipt per customer, in each case compared to the ten other stores and without adjustment for inflation. Moreover, they receive, at the beginning of the report, the department’s own percentage sales change compared to the identical week of the previous year once again referred to separately without a market comparison and without inflation adjustment. Typically, the department managers display these sales reports in their department’s respective break rooms so that their subordinates can also access them.

2024-02-12

2024-04-14

Sales performance

Employee sick days and overtime

Control Group: No changes

Treatment Group: For the two months of the experiment the treatment group receives the identical information in the report as the control group, with the exception that instead of the separate percentage nominal sales change compared with the identical week of the previous year, the inflation-adjusted sales change (including the inflation adjustment rate and a comparison to the inflation-unadjusted value as a footnote) compared with the previous year is shown here at the beginning of the report.
The department managers also receive the sales reports from their store managers during the experiment to emphasize the significance of these reports. However, it is important to note that the reports, and thus also the inflation-adjusted sales numbers of the treatment group, are not further discussed in this context. For the inflation adjustment rates of the different departments of the retail industry, we rely on the retail chain’s in-house inflation indicators at the product category level.

Not available

Stratified Clustered Randomization: We stratify our randomization both across stores and within departments. On the one hand, this ensures that three departments of each store are randomly assigned to the treatment group and three departments to the control group. On the other hand, this ensures that 5 respectively 6 of the identical departments across all eleven stores are randomly assigned to the treatment group and the control group. Since the employees of the three fresh food counters (butcher, cheese, fish) have the same break room where the sales reports are laid out, they were combined into one department. The experiment is clustered on the department level, meaning that all employees within the same department receive the same treatment.

Store departments

Yes

Eleven stores and six departments each (66 clusters)

Eleven stores and six departments each (66 clusters)

Control group: 33 departments
Treatment group: 33 departments