Overexploitation Risk in “Green Mountains and Clear Water”

Abstract

This study is motivated by a real project in China in the context of “Green Mountains and Clear Water”. As pointed out by Xi (2016), “to improve the environment is to boost productivity”. The purpose of this paper is to devise optimal capital allocation and compensation mechanisms when there exist social and environmental responsibility risks. A firm with access to a single investment project in an ecoregion may engage in environmental violations, and the ecoregion's proprietor (say municipal government) with preferences for ecological protection privately observes the process audit effort level. I extend the contract theory to a more general setting and find that given linear compensation schemes, the optimal solution is to set the non-negative share of project cash flows larger than 100% while the fixed component negative. The incentive-compatible mechanism gives rise to the overinvestment and overexploitation problems relative to the first-best level, and both problems become smaller as process audit effort level increases. Finally, I investigate how the model may vary with relevant characteristics such as multi-resource requirement, risk-averse government, and verifiable process audit information. Given unverifiable process audit information, municipal government faces a dilemma of gaining a higher compensation but losing a larger ecological area. In such a case, municipal government has to seek a compromise and may pin hopes on higher audit effort levels from superior administration units. The analysis also highlights the power of stricter process audit to reduce market surplus loss caused by information asymmetry.

Introduction

Internal capital budgeting is typically explained as a response to decentralized information and incentive problems (e.g., Harris and Raviv, 1996, Harris and Raviv, 1998). As claimed by Harris and Raviv, 1996, Harris and Raviv, 1998 and Bernardo et al., 2001, Bernardo et al., 2004, corporate headquarters should allocate more capital to high quality projects. The existing studies upon capital budgeting usually assume that the enterprises are fully ethical and thus comply with environmental and labor standards. However, the risks of corporate social responsibility problems in developing and emerging economies still give no reason for optimism.

Although the Chinese economy has grown rapidly during the last four decades, the environmental efficiency in the central regions is at a relative low level (Song et al., 2018). According to a recent survey (Lo et al., 2018), in the Chinese context the enterprise environmental violation events are not unusual and there is a significantly negative stock market reaction to these incidents. Furthermore, the empirical evidence in Lyon et al. (2013) and Lam et al. (2016) shows that Chinese investors react negatively to corporate environmental initiatives and green awards.

Indeed, the social and environmental responsibility problem is an important and challenging issue over the world. Recent tragic building collapse and garment factory fire incidents in Bangladesh have impelled global companies to concern about the potential supplier responsibility problems (Chen and Lee, 2017). Coincidently, Daigneault et al. (2018) investigate the impact of environmental policies on farm income in New Zealand, and find that net farm revenue decreases by between 0 and 11%, while not all environmental performance indexes improve. Wang et al. (2019) argue that there is no guaranteed link between voluntary disclosure of positive environmental information and firm environmental performance.

This paper is motivated by a real project in China in the context of “Green Mountains and Clear Water”, which implies that “to improve the environment is to boost productivity” (Xi, 2016). City X is a coastal city and owns hundreds of islands, all of which belong to the national ecological zone. The city plans to build a five-star hotel in an island with the largest economic potential. To balance the environmental performance and economic income, city X not only shifts the land use right with 40 years to an investment company at the beginning of the project, but also shares the cash flows of the project in the future. National environment agency, as a third-party organization, hires a professional audit team to deal with the potential environmental violations. The items about investment capital, land use right, and the compensation scheme are required to be specified in a contract.

In this paper, I consider a risk-neutral firm which has an investment project in an ecoregion. This project has to use some ecological resources like land and sea. In addition to the inescapable harm to the ecoregion, the firm may cut corner to lower project cost. That is, the firm is not fully ethical and may engage in social and environmental violations. Unlike quality violations that can be discovered by project inspection, responsibility problems such as environmental damage and poor working conditions can only be detected by process audit, public disclosure and actual accidents. For simplicity, I assume that there are no quality risks in the project.

Process audit is independently conducted by a third party like superior government or national environment agency. First, the audit effort level is unknown to the firm but known to municipal government. Second, the audit effort level cannot be verified ex post and thus cannot be specified in a contract. Third, municipal government may benefit from reporting a larger (false) audit effort level, reflecting high social and environmental standards. Hence, the firm should construct a compensation mechanism to induce truthful telling from municipal government.

It is not clear that how the insights from existing studies can be extended to a setting where an investment firm may engage in unethical practices. Such unethical practices are the activities violating social and environmental standards, for example, insecure working conditions, improper sewage discharge, gender discrimination, and so on. This paper aims to find the optimal capital allocation and compensation scheme under information asymmetry and responsibility risks. In particular, this paper will answer the following questions: (1) What is an effective compensation contract when there may exist unethical practices? (2) What are the impacts of information asymmetry on capital allocation and the area for project exploitation? (3) What are the impacts of process audit on capital allocation, development area, and market surplus loss?

To address these questions, I incorporate the capital allocation framework of Bernardo et al. (2001) into a setting with responsibility risks and analyze the optimal mechanism design problem. As in Bernardo et al. (2001), I employ a linear compensation scheme. In fact, since Holmstrom and Milgrom (1987) have shown that optimal compensation schemes are linear in the aggregated outcome under fairly general conditions, such schemes have become pervasive in the literature (e.g., Aggarwal and Samwick, 1999; Bernardo et al., 2001; Siemsen et al., 2007; Schlapp et al., 2015). Although the restrictions on linear schemes place some limits on the results, the basic model is descriptively reasonable and allows me to investigate a complex problem. In my model the compensation consists of a fixed component and a non-negative share of project cash flows. Unlike Bernardo et al. (2001), the optimal solution is to set the non-negative share larger than 100% while the fixed component negative. This is due to the fact that the firm's benefits brought by responsibility violations decreases in audit effort level.

This paper is most closely related to the work of Bernardo et al. (2001), which examines capital allocation and managerial compensation in a single-division firm, and Bernardo et al. (2004), which extends their earlier analysis to a multi-division firm. These papers assume that all parties are risk-neutral and division managers have private information about project quality and preferences for capital. As in my model, the optimal compensation contract is linear in project cash flows. Unlike my model, they do not consider the firm's responsibility risks or the effects of exogenous process audit that mitigates such risks. Interestingly, they find the underinvestment problem, whereas I find the overinvestment and overexploitation problems. In addition to capital budgeting, how to improve environmental and social consequences of ecoregional investment with potential violations is at the heart of my analysis.

Harris and Raviv (1996) also study the capital budgeting process in a single-division firm. In their model, a probabilistic auditing technology is adopted to guarantee incentive compatibility and makes the headquarters know the true project quality at some fixed costs. Harris and Raviv (1998) extend their earlier work to a multi-division firm and show that the same solution holds, with the difference that the initial spending limit is allocated for all the projects. Unlike my model, they assume division managers' private information is verifiable and the compensation contract is exogenously specified. Given the assumption of exogenous compensation contracts, they find regions of under- and over-investment.

Malenko (2019) examines optimal design of a continuous-time capital budgeting process based on Harris and Raviv (1996). In his model, division managers are privately informed about the arrival and properties of future projects, and headquarters can audit projects at a cost. The optimal mechanism is characterized as a combination of a dynamic spending account and a threshold for separate financing. As in Harris and Raviv, 1996, Harris and Raviv, 1998, his paper finds both under- and over-investment but does not endogenously determine monetary compensation.

It is worth noting that this paper uses the underlying idea in Chen and Lee (2017) for modelling process audit. To ease the exposition, I employ a simplified process audit model. That is, unlike Chen and Lee (2017), I do not consider the case that the party with responsibility risks can pretend to be the more ethical one. Nonetheless, the basic techniques of process audit introduce greater richness into my model and provide clearer understanding of real-world responsibility problems.

This paper demonstrates that when the project is worthy of investment, municipal government's expected utility increases in audit effort level. This result is consistent with the trend of sustainability that has prompted ecoregions to awaken to the threat of environmental violations and ask for stricter process audit. When the capital is allocated, the firm always allocates more than the first-best level, and perhaps surprisingly a larger area of ecoregion is exploited relative to the first-best solution (i.e., overexploitation problem). The basic intuition for the results is straightforward. First, to induce truthful reporting from municipal government, the firm needs to allocate more capital and offer higher compensation relative to the benchmark under symmetric information. Second, the area for project development responds positively to the capital in the optimal mechanism. This result is partially consistent with the motivating project in which the contracted (and final) ecological development area is significantly larger than the one municipal government initially planned.

In my model, capital allocation and development area are complementary in project cash flows. Such complementarity further explains why they change similarly. In particular, I show that both capital allocation and development area decrease in audit effort level. More importantly, both the overinvestment and overexploitation problems are smaller for the project with higher audit effort levels. This is because market surplus loss caused by information asymmetry decreases as audit effort level rises. Given unverifiable process audit information, municipal government faces a dilemma of gaining a higher compensation but losing a larger ecological area. In such a case, municipal government has to seek a compromise and may pin hopes on higher audit effort levels from superior administration units.

Finally, the similar mechanisms with additional peculiarities are still optimal in several extensions. In the first extension, I consider the setting where the investment firm needs several kinds of resources like land and sea. In the second case, I assume the municipal government is risk-averse. In the third case, I discuss the possibility of buying process audit information.

The paper is organized as follows. Section 2 describes the basic model. Section 3 presents the benchmark under symmetric information. Section 4 solves for the optimal mechanism under information asymmetry and considers several extensions of the model. Finally, Section 5 concludes and gives direction for future work.