In 2015, Mark Carney, then Governor of the Bank of England and now Prime Minister of Canada, named a problem that financial markets had begun to confront but had difficulty articulating, “the tragedy of the horizon.” As we and a colleague wrote in 2023, climate risks can unfold over decades, while corporate decisions are often made much sooner. If the results of today’s investments emerge long after today’s decision-makers have moved on, firms may underinvest in projects whose benefits are realized only in the distant future.

The typical response is to ask how corporate governance can give managers longer-term incentives. A better question is, how can a firm credibly promise today to reward a manager for an outcome far in the future?

From Full Commitment to a Puzzle

If the problem were simply that managers act in the short term, the solution would seem straightforward: give them longer-term incentives. And when a firm can bind itself today to a payment far in the future, it is. Edmans et al. (2012) provide an important benchmark. Under such full commitment, a difference in management and investment time horizons does not by itself generate underinvestment. A properly designed contract can align managers’ incentives with long-term shareholder value.

Yet the real world does not appear to fit this benchmark neatly. Cohen et al. (2023) document that the share of firms with ESG-linked executive performance indicators grew from roughly 3% in 2010 to over 30% by 2021. At the same time, companies often fall well short of their climate commitments. In a study of 25 major global companies, Day et al. (2022) find that announced net-zero pledges imply emissions reductions averaging roughly 40% rather than the 100% the term suggests.

These observations do not by themselves establish a contractual explanation, but they raise a question that contract theory is well placed to answer: Why does the proliferation of long-term climate commitments not translate into incentives capable of sustaining long time-horizon investment? In a recent paper, we show that incomplete contracting produces a gap of this kind, even when every other friction is absent. Compensation design still works, just not completely.

The Commitment Gap

Consider a manager who makes a climate-related investment whose ultimate effects will not emerge for a couple of decades. A board may want to promise compensation based on the eventual outcome. But by the time the outcome is observed, the CEO may have left, the board may have changed, ownership may have shifted, and the interests of the people responsible for administering the compensation may be different.

Our model captures this problem. The firm can contract on an earlier verifiable signal but cannot credibly commit to compensation based on the ultimate long-term outcome. Managerial effort under this constraint is positive and responds to incentives. It is simply lower than what full commitment to the offered compensation would deliver. We refer to this remaining difference as the commitment gap.

The commitment gap is not a problem of information. The parties may understand perfectly well that the long-term outcome is relevant. The problem is that it lies beyond the effective reach of the contract, so the firm must rely on early evidence that reveals less about the ultimate result. Different horizons are not the fundamental problem. The inability to commit across these horizons is.

Better Measurement Is Not Enough

This distinction produces a counterintuitive implication. Suppose this future outcome becomes measurable with substantial precision. One might expect this to make the contracting problem more manageable. Better information should make it easier to design effective compensation.

But this is true only if the contract can actually use the information. When the outcome cannot be the subject of a contract, making it more informative can make the problem more costly. The better the distant outcome is at revealing the impact of managerial effort, the more valuable it would be to include that outcome in compensation. And the higher the cost of being unable to do so. Only extending the contract to reach the future outcome can fully close the gap. No available mechanism reaches that far. The tragedy of the horizon is therefore not a problem of time alone. It is a problem of contractual reach.

Legal Duration and Effective Commitment

A supply chain agreement does not become void because the CEO who signed it has departed. The corporation remains bound. But climate-contingent executive compensation creates a different problem. The obligation may require measuring, decades later, what outcomes resulted from a specific managerial decision, distinguishing those outcomes from everything that happened in the intervening years, and applying metrics that may themselves have changed. Future boards administering the arrangement may not share the priorities of those that made the original promise. The contract may be formally binding. Yet its implementation may be much weaker than its legal duration suggests.

This suggests a practical question for lawyers advising boards. What is the effective period over which a contingent payment to a departed executive can be credibly enforced? The answer will depend on the jurisdiction, the structure of the contract, the nature of the underlying metric, and the institutions responsible for administering it. Making this horizon explicit would discipline how boards evaluate long-term compensation. A contract that nominally lasts 20 years may not provide a 20-year incentive if the parties cannot credibly preserve and enforce its terms for that period.

This distinction between legal duration and effective commitment points toward a different role for contract design. The objective is not to make a 20-year compensation agreement more detailed. It is to reduce the extent to which the agreement depends on the discretion of future decision makers. Can a contract be designed so that its essential terms survive a change in management, ownership, or the board itself?

More Commitment Is Not Always Better

Institutional mechanisms might move in this direction. Independent trusts, escrow structures with third-party administrators, and other arrangements that separate the administration of a long-term obligation from the future management of the company could reduce the ability of successor decision makers to alter the original commitment. Similar institutional separation is familiar in other areas of corporate and financial life, including pension arrangements.

But such arrangements raise questions of their own. Who controls the trust? Who determines whether the metric has been achieved? What happens if the metric becomes obsolete, or if the underlying objective ceases to make economic sense? What survives a merger? Institutional insulation creates commitment, but commitment reduces flexibility. A rigid contract can destroy value when circumstances change in ways nobody anticipated.

Renegotiation, at times, is efficient. The question for corporate lawyers is therefore not how to eliminate future discretion, but which elements of a long-term commitment should be insulated from renegotiation, and which should remain adaptable. That question brings governance transitions, administration, attribution, and measurement into the design of the contract itself.

Beyond Climate

Climate is an unusually visible case of a more general problem of investments whose consequences extend beyond the period over which managerial commitments can be credibly sustained. Long-cycle R&D, infrastructure, and organizational transformation may generate value only after the manager responsible for initiating them has left the firm. What makes climate a useful laboratory is the combination of decade-long horizons, difficult attribution, and evolving measurement standards.

The broader implication is that corporate governance should distinguish between long-term objectives and long-term commitments. A board can announce an objective that extends 20 years or more. It is harder to create an incentive that credibly connects a manager’s decisions today to a payment 20 years from now. A long-term objective is a statement about the future. A long-term incentive is a promise about the future. The commitment gap lies in the difference between the two. That is where the tragedy of the horizon becomes a problem of corporate design.

Christos Cabolis is an adjunct professor of economics and competitiveness, and Karl Schmedders is a professor of finance, both at IMD Lausanne. This post is based on their recent paper, “The tragedy of the horizon: A contracting account,” available here.