Using random assignment of compensation models to experiment participants, we find that compensation packages drive higher R&D investment as well as higher resource allocation towards clean production. Compensation models, then, play a crucial role in incentivising executives to create long-term financial and non-financial value. Shifting financial incentives towards long-term and/or clean production objectives leads to allocative decisions that increase R&D investments and clean production. 

 In order to assess the response of experiment participants to compensation mechanisms in relation to their (given) preferences for time, risk and pro-social outcomes, as well as to external pressures like shareholder focus, we survey respondent preferences and randomly vary (fictional) shareholder focus statements for long-term and/or clean production. When asked to decide on the allocation between current production and R&D, participants receive a ‘shareholder statement’ that either favours short-term profits or long-term profits. The other variant includes shareholder statements that either favour profit-maximisation or clean production. The analysis reveals that whilst compensation models continue to have a significant effect on allocative decisions, participant preferences for time, risk, and ESG and outside (‘shareholder’) preferences for long-term or clean production matter in their own right as well.  

 Additionally, the shareholder statement can influence the strategic decisions, even when there is no financial impact of that statement, namely when the compensation model is held constant. If participants receive a shareholder statement that urges them to focus on long-term profits, they invest more in R&D on average, compared to the situation with a shareholder statement that urges them to focus on short-term profits. Regarding investment in clean production, participants with a shareholder statement that focuses on clean productions allocate more resources to clean production, compared to participants that receive a pro-profit statement from their shareholders. 

 Finally, preferences can change the effect of a compensation model or shareholder statement on investment in R&D or clean production. Long-term oriented participants react stronger to long-term financial incentives than short-term oriented participants. This only holds when the shareholder statement is focused on the short term. However, shareholder statements do not have a different effect depending on the time preferences of the participants, it is positive in all cases. Additionally, risk-loving participants react stronger to a long-term shareholder focus, but the effect of the compensation model is the same for both groups. The effect of a compensation model or shareholder statement on investment in clean production is independent of the social preferences of the participants. Furthermore, preferences can also influence allocation decisions, irrespective of compensation models. For example, participants who believe businesses should focus more on ESG-related topics also invest more in clean production than other participants, regardless of the compensation model or shareholder statement. Additionally, risk-loving participants are more prone to investment in R&D than risk-averse participants.  

This report
The analysis consists of two experiments in which participants make strategic decisions on firm resource allocation. We measure the treatment effect of compensation models by randomly varying the temporal orientation of compensation models (i.e. short-term versus long-term objectives) as well as the pay-out for clean production, and by paying experiment participants concordant with long-term and clean production performance. In the first experiment, participants allocate resources towards either short-term production or R&D investments, in which the latter increases long-term production. In a second experiment, participants allocate resources towards either highly profitable dirty production with negative externalities or towards less profitable clean production without negative externalities.  

The analysis consists of an online laboratory experiment with over 3000 participants.