Exploring the association between financial and nonfinancial carbon-related incentives and carbon performance

Abstract

Firms increasingly respond to pressures to reduce their carbon emissions by providing financial and nonfinancial carbon-related incentives that should align and extrinsically motivate individuals’ behaviour towards improved carbon performance. We explore whether and how the provision of carbon-related incentives is associated with carbon performance. We employ data on carbon-related incentives and carbon emissions that S&P 500 firms voluntarily disclose to the CDP. Correcting for sample-induced endogeneity and time-series dependencies, we find that financial carbon-related incentives are associated with superior carbon performance, while nonfinancial carbon-related incentives are not associated with carbon performance. Financial carbon-related incentives appear to extrinsically motivate managers and employees and channel their efforts towards improving carbon performance. However, nonfinancial carbon-related incentives do not appear to be effective. These differences may be explained by the fact that financial carbon-related incentives trigger different cognitive and motivational mechanisms (e.g. utility, expectancies) in individuals than nonfinancial carbon-related incentives.

Keywords:

• carbon performance
• financial incentives
• nonfinancial incentives
• management control
• motivation

Acknowledgements

We would like to thank the editors Mark Clatworthy, Juan Manuel García Lara and Edward Lee, the associate editor Frank Hartmann and two anonymous reviewers for their valuable feedback.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The CDP is an international not-for-profit organisation founded in 2000. On behalf of 827 investors with US$100 trillion in assets under management, it works together with firms and cities to measure, disclose, and manage information regarding a broad set of environmental indicators (CDP 2016b).

2 While agency theory ignores intrinsic motivation or implicitly assumes it to be constant (Frey and Jegen 2001), the behavioural-based literature considers motivation as a continuum from intrinsic to extrinsic motivation (Groen et al. 2017, Ryan and Deci 2000). While incentives, both financial and nonfinancial ones, are deemed to increase extrinsic motivation, their effect on intrinsic motivation is not as clear a priori. The behavioural-based literature highlights the risk that incentives may weaken intrinsic motivation if they are perceived as control mechanisms or if they convey the impression that a formerly intrinsically pursued activity is seen as unattractive. However, to observe a negative ‘net effect’ on motivation, the decrease in intrinsic motivation would have to be larger than the increase in extrinsic motivation.

3 The CDP published its first report on the basis of the responses of 235 firms in 2003. Since 2007, the CDP collects information from S&P 500 firms. In 2016, about 5,600 firms world-wide responded to the CDP questionnaire.

4 Before the CDP 2010 questionnaire, the CDP also asks a number of open questions about the availability of carbon-related incentive systems. However, the evaluation of such questions would not only be more subjective, but also less comparable to the questions that we evaluate in this study.

5 We exclude financial services firms because their financial statements differ from those of other industries due to specific regulations and accounting rules, which means that the explanatory variables used are not always meaningful for them.

6 We exclude firms with negative equity and thus in financial distress, which decisively reduces their management’s room for maneuver, as in this case it must primarily focus on securing its existence.

7 The CDP sends out its questionnaire in the second quarter and typically receives responses in the third quarter. With the consideration of a time lag of one year in our basic analyses, we ensure that the incentive system is in place when we study its impact on carbon emissions. An (unreported) additional analysis indicates that the results are also stable when we omit the time lag.

8 The data from the CDP questionnaire allows to differentiate between different nonfinancial incentives (i.e., recognition, prizes/awards, other non-monetary rewards). However, an (unreported) additional analysis shows no differences for the different nonfinancial incentives with respect to their impact on carbon performance.

9 Alternatively, we use the proportion of a firm’s research and development expenses to a firm’s total assets (R&D_INT). If a firm is more innovative, it is more likely to implement changes in production processes or product design that help to improve its carbon performance (Anton et al. 2004). The results for this alternative proxy coincide with the results for IA2A.

10 The results are robust to employing carbon emissions data from the Thomson Reuters’ Asset4 database.

11 Alternatively, table 6 presents the results for the Heckman model using the Limited Information Maximum Likelihood (LIML) estimator that does not cluster standard errors at firm level. The LIML results for our variables of interest INC_FIN and INC_NON are directionally consistent with the FIML results.

12 In an unreported analysis, we examine the additional impact of all employees as beneficiaries of carbon-related incentives. The first (second) measure equals one if the all employees benefit from financial (nonfinancial) carbon-related incentives, and zero otherwise. The results are consistent with the results for the board and/or top management as beneficiaries.