Limited liability is a defining feature of corporations. Corporate law generally affords limited liability protection not only to the individual shareholders of corporations, but also to corporate parents of subsidiaries. The use of subsidiaries is pervasive, in the US and globally. However, because parents are generally insulated from the liabilities of their subsidiaries, this organisational structure creates incentives to engage in riskier behaviour than would be socially optimal. To mitigate this moral hazard problem, courts can impose liability on parent corporations in some instances.
In our paper, The Limits of Limited Liability: Evidence from Industrial Pollution, we shed light on how parent liability protections affect firms’ environmental behaviours. The threat of liability incentivises firms to limit toxic emissions, which may impose significant costs on various stakeholders, such as employees or the local community. Policymakers in many countries have adopted a “polluter pays” approach to environmental regulation to encourage companies to bear the brunt of such costs. However, the effectiveness of this regulatory framework is, to an extent, undercut by limited liability. If liability truly is limited, a parent will not bear the costs of environmental remediation that exceed the value of its subsidiary’s assets.
We look at a Supreme Court case that clarified parent company liability under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA). CERCLA authorises the Environmental Protection Agency (EPA) to impose ex-post liability on parties responsible for toxic sites. Costs associated with cleaning up such sites are often large and can result in the insolvency of responsible parties. Indeed, we estimate that liabilities associated with environmental clean-ups account for roughly 5.5% of a firm’s total assets in our sample. In United States v. Bestfoods (1998), the Supreme Court narrowed the circumstances under which parents are responsible for the environmental clean-up costs of insolvent subsidiaries.
The Bestfoods case centred on the question of when a parent is liable as an operator of the toxic site of an insolvent subsidiary. Prior to this decision, some appellate courts held parents liable as operators under relatively broad circumstances, namely if they had “actual control” of or the “ability to control” subsidiaries. In the Bestfoods decision, the Supreme Court invalidated these tests, holding parents liable as operators only if they directly operated the facility of a subsidiary responsible for emissions (e.g. if a parent is in a joint-venture with its subsidiary). We use this decision to study the effect of parent liability on environmental behaviours; analysing and comparing the plants of subsidiaries located in federal appellate court circuits that had weaker liability protection for parents prior to Bestfoods (the treatment group), with plants located in areas where a relatively narrow standard for parent liability was already in place (the control group).
We use data on toxic emissions from the EPA’s Toxic Release Inventory (TRI) to study the effects of the Bestfoods decision. Our findings indicate that stronger liability protections for parents is associated with higher emissions by subsidiaries. Specifically, plants in our treatment group increase ground emissions – the focus of CERCLA enforcement actions – by approximately 5% - 9% relative to the control group in the five years following Bestfoods. This increase in emissions is driven, in part, by chemicals that cause cancer and other chronic diseases. Figure 1 plots the evolution of this increase. Year t corresponds to the Bestfoods decision. In the years leading up this, toxic emissions are similar for treatment and control plants, but following the decision we see that emissions begin to go up for the treated plants relative to the control plants.
Figure 1. Impact of the Bestfoods decision on toxic emissions over time
We consider two potential explanations for this emissions increase. First, the increase in emissions may stem from reduced investment in emissions reductions because parents do not fully internalise the costs of environmental clean-ups, thus weakening incentives to limit pollution. To study this, we use data on plant reduction activities from the EPA. Our findings indicate the Bestfoods decision decreases incentives to invest in some types of emissions reduction. Specifically, plants decrease production-related reductions (e.g. modifying equipment or improving chemical reaction conditions) by 15% -17% following the decision. We do not find evidence of reduction changes related to operating practices (e.g. improved record-keeping).
Second, we consider whether the environmental effects of Bestfoods stem from changes in economic behaviour. Specifically, we test whether Bestfoods is associated with changes in output, using plant chemical-level production data from the EPA. We do not, however, find evidence consistent with this. The estimated effect of the court decision on production is economically small – consistent with the notion that costs associated with environmental clean-ups are often fixed and do not affect marginal production costs.
Finally, we show that insolvency and executive compensation play an important role in driving our findings. First, consistent with the idea that parents are only liable for clean-ups if a subsidiary is unable to pay, the increase in pollution and decrease in emissions reductions is concentrated in less-solvent subsidiaries. Second, the effects of Bestfoods are driven by the plants of parents that have existing environmental liabilities or are close to financial distress, suggesting such firms may prioritise short-term financing needs over the avoidance of long-term liabilities. Third, we find that firms with executives whose compensation is strongly tied to the volatility of their firms’ equity returns respond the most to Bestfoods.
Overall, our findings highlight the moral hazard problem associated with limited liability. While our setting precludes a rigorous welfare analysis, the results suggest that strengthening limited liability for parents leads to an increase in costs borne by other stakeholders. Efforts by policymakers to strengthen liability protections should carefully weigh the interests of corporate owners with those of other stakeholders.
This blog is written by academic guest contributors. Our goal is to contribute to the broader debate around topical issues and to help showcase research in support of our signatories and the wider community.
Please note that although you can expect to find some posts here that broadly accord with the PRI’s official views, the blog authors write in their individual capacity and there is no “house view”. Nor do the views and opinions expressed on this blog constitute financial or other professional advice.
If you have any questions, please contact us at firstname.lastname@example.org.