Pay-for-performance is often championed as a possible solution to agency problems, aligning managers’ incentives with shareholder interests. But what happens when solving one agency problem creates another – between the firm and society?

In a recent paper, we develop a model to study how a firm’s choice of governance framework—the array of mechanisms it employs to mitigate agency problems – shapes its production decisions and the resulting externalities. We apply this model to the U.S. coal mining industry and find that when firms face higher monitoring costs, they compensate by tying managers’ pay more closely to their performance. This shift boosts output, but at a steep price: Workplace safety deteriorates significantly. Our findings raise important questions about how governance reforms and investor actions – like divestment campaigns – shape not only firm value but also society and the environment.

Governance Trade-Offs and Socially Costly Production

The heart of our argument is straightforward. A firm’s governance framework comprises tools like incentive pay (e.g., production bonuses) and monitoring mechanisms (e.g., internal audits, board oversight, institutional-investor influence). These tools are meant to address a classic principal-agent problem: Managers may prefer to avoid hard work (live “the quiet life”), while shareholders want performance. But here’s the catch: Strong incentives to increase production can create negative externalities that firms only partially internalize through fines or litigation. For instance, aggressive sales targets may pressure employees to cut corners or even break the law. In the 2016 cross-selling scandal, Wells Fargo employees opened millions of unauthorized accounts to meet unrealistic quotas. Likewise, pharmaceutical companies have been held responsible for contributing to the opioid crisis by misbranding drugs – most notably by claiming that certain opioids were nonaddictive.

We formalize this idea in a principal-agent model where stronger production-based incentives drive managers to increase output but also raise social costs when managers push production beyond a certain point. Critically, a firm’s owners internalize only part of these social costs (e.g., through fines or reputational damage), and thus may prefer stronger production-based incentives even when they harm society.

Our model predicts that when it becomes more challenging to provide effective oversight of management, the company leans more heavily on performance-based incentives. This change increases both production and negative externalities. Importantly, the effect is not uniform. Firms with alternative governance tools – such as influential institutional shareholders, independent boards, or strong labor and regulatory oversight – are less likely to make this trade-off, dampening the increase in social costs.

Evidence from the Coal Industry

We test these predictions using rich asset-level data from the U.S. coal industry, a setting that offers three key advantages.

• Coal mining involves a well-documented trade-off between output and safety. Halting production is often necessary to maintain safety. The 2010 Upper Big Branch Mine disaster, which killed 29 miners, tragically illustrates the costs of prioritizing output over safety.
• The industry is highly standardized, with limited product differentiation, making it easier to isolate governance effects.
• Recent climate-driven divestment campaigns by major institutional investors (such as CalPERS and Norway’s NBIM) provide a natural experiment: Politically motivated divestments raised monitoring costs for coal firms by removing key activist owners.

Ownership Changes and Social Costs

We start by analyzing private-to-public transitions, which increase monitoring costs by diluting ownership and weakening oversight. Using detailed mine-level data over four decades, we find that these transitions lead to:

• 11 percent higher production
• 33 percent more safety violations

The effects are concentrated in firms that experienced the largest dilution or relied heavily on bonus pay, consistent with our model’s predictions. Notably, the increase in violations per unit of output suggests that this effect is not just about scale – it reflects a shift toward “socially costly production.”

The Role of Alternative Governance Mechanisms

Firms with stronger alternative governance – such as influential institutional blockholders, independent boards, or stock prices that more accurately represent the company’s true value  – see smaller increases in production and safety violations post-IPO. Regulatory mechanisms matter too: Tight enforcement and strong unions dampen the effects of increased monitoring costs.

A Natural Experiment: Climate Divestment Shocks

To establish causality, we exploit a policy shock. Between 2015 and 2017, CalPERS and NBIM were forced by their governments to divest from thermal coal producers (but not metallurgical coal producers) for environmental reasons. This divestment removed the monitoring provided by these large investors and was plausibly unrelated to thermal coal firms’ performance and safety records.

We compare thermal and metallurgical coal firms before and after the divestment and find:

• Significant increases in production and safety violations in thermal coal firms only
• Rising bonus thresholds for thermal coal managers, indicating stronger pay-for-performance contracts

These results strongly support our hypothesis that divestment increased monitoring costs, which in turn led firms to shift toward incentives that amplified social costs.

Policy and Regulatory Implications 

Our study offers four main contributions, each with policy relevance:

• A new trade-off in governance: We document how solving the agency problem between managers and shareholders can exacerbate the one between shareholders and society, particularly when output creates negative externalities.
• The ESG puzzle: Our findings help explain why some firms with poor environmental and social records also score poorly on governance. It’s not just a lack of values; it may be a byproduct of weakened oversight and misaligned incentives.
• Rethinking ownership structures: As asset ownership becomes more dispersed or passive through, say, index funds or after IPOs, traditional monitoring mechanisms erode. Our work suggests that absent compensating governance tools, this trend in ownership structures can have significant social consequences.
• Unintended consequences of sustainable investing: Divestment campaigns are often aimed at punishing polluters. Our evidence, however, shows that without strong substitutes for investor oversight, divestment may backfire, leading to higher production and greater externalities. Investors and policymakers should account for these effects when designing ESG strategies.

Conclusion

Corporate governance reforms and sustainable investing strategies can profoundly influence firm behavior. But these efforts must be carefully calibrated. Our findings reveal a subtle yet powerful mechanism through which governance decisions can amplify or mitigate social harm.

When oversight is costly or weakened – whether due to ownership dispersion, regulatory design, or divestment campaigns – firms may double down on performance pay. The result may be more output, but also more harm. The challenge is not just to align managers with shareholders – but to align both with the broader interests of society.

This post comes to us from professors Alvin Chen at Stockholm School of Economics and Michael D. Wittry at Ohio State University. It is based on their recent article, “Production and Externalities: How Corporate Governance Shapes Social Costs,” available here.