Testing Lean & Green Corporate Incentives: Experimental Evidence

In an online behavioral experiment, participants take on the role of production workers and make a binary decision between working business-as-usual (BAU) and investing in continuous improvement processes (CIP).

In all treatment conditions, a Lean and Green measure is introduced. In the baseline treatment (T1), no additional corporate incentive scheme is implemented. In treatments 2-5, we additionally introduce and test different types of corporate interventions:

T1: Baseline – Basic Lean & Green setting without corporate intervention
T2: Corporate intervention – Fixed monetary reward
T3: Corporate intervention – Fixed social reward
T4: Corporate intervention – Variable monetary reward
T5: Corporate intervention – Variable social reward

The reward is granted if participants invest in continuous improvement processes and achieve an efficiency improvement. In the monetary reward treatments (T2 and T4), the reward is implemented as an additional monetary payment. In the social reward treatments, the reward is implemented as non-monetary recognition. In the fixed reward treatments, the reward is of the same size regardless of the magnitude of the efficiency improvement. In the variable reward treatments, the reward increases with the magnitude of the efficiency improvement.

In T3, participants who invest in continuous improvement processes and achieve an efficiency improvement are listed on the Lean & Green Recognition Board. The recognition board displays the participant’s self-chosen display name, the change in the company’s efficiency score, and the associated reduction in CO₂ emissions. The recognition board is made available only after the experiment has been completed and is visible only to participants assigned to the respective treatment.

In addition, in T5, participants who achieve a substantial efficiency improvement receive a certificate of recognition. The certificate includes the participant’s self-chosen display name, the achieved change in the company’s efficiency score, and the associated reduction in CO₂ emissions. The certificate is issued privately after completion of the experiment via Prolific and is accessible only to the respective participant.

2026-07-12

2026-09-30

The study aims to evaluate the effectiveness of different corporate incentive strategies in shaping employees’ decisions between continuing business-as-usual (BAU) and investing in continuous improvement processes (CIP).

Secondary outcomes include corporate efficiency, environmental impact, and employee earnings. All three outcomes are determined by participants’ decisions and the realized decision outcomes. Corporate efficiency is measured by changes in the company’s efficiency score. Environmental impact is measured by the real reduction in CO₂ emissions. Employee earnings are measured by the individual payoff, which additionally depends on the treatment-specific incentive scheme.

After the main decision, we will administer questionnaires to collect additional information on participants’ personal values, economic preferences, work motivation, and socio-economic background. Specifically, we will measure personal values using the Portrait Values Questionnaire (PVQ; Schwartz, 2012), economic preferences following Falk et al. (2018), work motivation using the Multidimensional Work Motivation Scale (MWMS), and socio-economic background characteristics.

Participants will be randomly allocated to one of several treatments by the experimental software. Each participant will take part in only one treatment. To comply with ethical principles, participants will be informed about their rights, the voluntary nature of participation, the possibility to withdraw from the study, and the processing of their data. By continuing with the study, participants provide their consent to voluntary participation and data processing.

After providing consent, participants will read the experimental instructions. They will then answer four comprehension check questions to ensure that they understand the experiment and the decision situation. Participants can only proceed if they answer all comprehension check questions correctly and do not provide a wrong answer twice for the same question. Participants who fail the comprehension checks will not be able to continue with the experiment.

Participants will then make their main decision, choosing between business-as-usual (BAU) and investing in continuous improvement processes (CIP). The outcome of this decision is determined probabilistically: for BAU, either no efficiency change, a small efficiency deterioration, or a substantial efficiency deterioration may occur; for CIP, either no efficiency improvement, a small efficiency improvement, or a substantial efficiency improvement may occur. To avoid biasing subsequent responses, participants will only learn the specific outcome of their decision, their corresponding monetary and non-monetary rewards, and the amount of CO₂ reduction at the end of the experiment. After the main decision, participants will complete several questionnaires.

In contrast to the basic treatment, participants in treatments 2-5 make the main decision additionally under incentives induced by corporate intervention schemes.

The experiment will be conducted online.

Not available

The experiment will be conducted online. Treatment assignment is randomized at the individual level using a computer-based lottery procedure implemented in the experimental software.

We randomize at the level of the individual. Each participant is assigned to one of the five treatment conditions (Treatment 1– Treatment 5).

No

500 English-fluent individuals from European countries with non-managerial work experience in industry sectors or occupational areas relevant to Lean and Green practices.

500 English-fluent individuals from European countries with non-managerial work experience in industry sectors or occupational areas relevant to Lean and Green practices.

Each treatment will include 100 participants, i.e., we will collect 100 observations per treatment.

In total, we will collect 5 x 100 = 500 observations.

In our literature search, we did not find any comparable studies from which reference values could have been derived for the sample size calculation. Therefore, we based our estimation on standard statistical assumptions. Specifically, we conservatively assumed a small-to-medium effect size (Cohen’s d = 0.40), following the conventional classification by Cohen (1988). Using a significance level of α = 0.05 and aiming for a statistical power of 0.80, the calculation indicated that a sample size of approximately 100 participants per treatment group would be sufficient to detect effects of this magnitude. In the absence of prior empirical data, we consider this sample size appropriate to address our research objectives while ensuring adequate statistical sensitivity. We therefore targeted 100 participants per treatment group, resulting in a total planned sample size of 500 participants across the five treatment conditions.