Boundary Conditions of Loss Aversion — When Past Performance Counteracts Evidence From a Field Experiment

How to cite

Hoffmann, Christin; Thommes, Kirsten. 2018. "Boundary Conditions of Loss Aversion — When Past Performance Counteracts Evidence From a Field Experiment." AEA RCT Registry. AEARCTR-0003411. https://doi.org/10.1257/rct.3411-1.0

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

Christin Hoffmann* , Kirsten Thommes

* Primary investigator

Registry ID
AEARCTR-0003411
DOI
https://doi.org/10.1257/rct.3411-1.0
First registered
Last updated
Status
On going
Trial dates
2018-01-01 – 2018-12-31
Countries
Private
JEL codes
C93; M12; M52
Registry record
http://www.socialscienceregistry.org/trials/3411

Abstract

We analyzed the effect of incentive-framing on performance in employment relations. If loss-aversion holds true, framing an incentive for employees as loss should lead to higher effort than framing the incentive as gain. However, current research sheds doubts on the universality of loss aversion. In most field settings, loss-framings do not increase performance as expected. Our results contribute to the understanding of the boundary conditions of loss-framings. Although loss-framings lead to increased performance among less-productive employees, they simultaneously lead to decreased performance among high-productive employees. Contrarily, gain-framings do not increase performance among less-productive employees but increase productivity among high-productive employees

Intervention

In order to analyze whether loss- or game framings in repeated settings work better and whether there are any time effects, we use a truck fleet of a medium-sized company in Germany. Here fuel consumption is one of the main productivity measures. Truck drivers can effectively save or waste fuel by their driving behavior without generating any side effects on speed or punctuality. As fuel is one of the main cost components in transportation, many truck companies apply telematics systems in their trucks, which, among other features, measure the energy efficiency of the driving style and provide feedback how to improve driving behavior. We conducted our experiment in one business site with 34 trucks and 41 truck drivers as all of these trucks were equipped with the same telematics system. We changed the incentive system to target every driver and pay everybody for their performance. Moreover, we tested whether a loss-framing or a gain-framing would be more effective.

Experimental design

We conducted our experiment at one company site of the large company with 34 trucks and 41 truck drivers, as all of these trucks were equipped with the same telematics system. The telematics system ranks driving behavior between 0 and 10 points with one decimal place, with 10 as the best result. We used this fuel-efficiency score as the dependent variable because it is largely unaffected by external circumstances. The telematics system is very sophisticated in terms of the data it provides. Using the data for travel distance, cargo weight, and driving behavior allows for fuel consumption to be estimated with a coefficient of determination of more than 95%. Any unobserved heterogeneity was purely due to air pressure and attrition of engine parts.

Design details

Randomization method
coin flip
Randomization unit
individual
Clusters
1
Observations
panel with 41 drivers, 1400 observations
Arms
20 drivers in each treatment, control phase before and after treatment
Primary outcomes
dependent variable: drivers performance explaining variables: incentive system, technological truck variables, variables measuring truck weight, topografie etc.

Cited by

No citation data available for this registration yet.

Keywords