Internal Governance of Firms

In The Internal Governance of Firms, which was co-written with Viral Acharya and Stewart Myers, and which I recently presented at the Finance Seminar at Harvard Business School, we argue that there are important stakeholders in the firm, particularly its junior managers, who care about its future even if the CEO acts in his or her short-term self interest and shareholders are dispersed and powerless. These stakeholders, because of their power to withdraw their contributions to the firm, can force the CEO to act in a more public-spirited and far-sighted way. We call this process internal governance.

The basic intuition behind the model is as follows. Think of a partnership run by an old CEO who is about to retire. The CEO has a young manager working under him who will be the future CEO. Three ingredients go into producing the firm’s cash flow: the firm’s capital stock; the CEO’s ability to manage the firm, based on his skill and firm specific knowledge, and the young manager’s effort, which allows her to learn and prepare for promotion. We assume the CEO can commit to a pre-determined amount of investment. The CEO will leave the investment behind as the firm’s capital stock. The CEO can appropriate everything else: he can tunnel cash out of the firm, consume perks, or convert cash to leisure by shirking. Because the CEO has a short horizon, he could simply decide to take all of the cash flow, investing nothing for the future. But he needs the young manager’s effort in order to generate the cash flow. If the manager sees that the CEO will leave nothing behind, she has scant incentive to exert effort, and cash flow falls significantly. To forestall this, the CEO commits to investing some fraction of current cash flow, building or enhancing the firm’s capital stock in order to create a future for his young employee, thereby motivating her. This allows the firm to build substantial value, despite being led by a sequence of myopic and rapacious CEOs.

We show that internal governance is most effective when both the CEO and the manager contribute to the firm’s cash flows. We extend the basic model by allowing the CEO to commit to sell the firm to the manager when he retires. We show that this extends the horizon of the CEO so much that the first best level of capital investment is reached, given managerial effort. We call this the rolling partnership and it reduces agency problems at the firm to the problems of incentivizing managerial effort. However, suppose the firm is a public corporation. In equilibrium, the shareholders do not intervene. This suggests that external governance, even if crude and uninformed, can complement internal governance and improve efficiency. This leads to a theory of investment and dividend policy, where dividends are paid by self-interested CEOs to maintain a balance between internal and external control.

Overall, our models show that the traditional description of the firm falls short on three counts. First, control need not be exerted just top down, or from outside; it can also be asserted bottom-up. The CEO has to give his subordinates a reason to follow, and that, implicitly, is how they control him. Second, the view that there is one residual claimant in the firm, the shareholder, is too narrow. Anyone who shares in the quasi-rents generated by the firm has some residual claims, and thus there is no easy equivalence between maximizing shareholder value and maximizing efficiency. Third, the fact that CEOs and managers get rents at different horizons means that each one has to pay attention to others’ residual claims in order to elicit co-operation. The checks that parties inside the firm impose on each other ensure the firm functions well, even if outside governance is weak.

The full paper is available for download here.