An agency cost is an economic concept concerning the cost to a "principal" (an organization, person or group of persons), when the principal chooses or hires an "agent" to act on its behalf. Because the two parties have different interests and the agent has more information, the principal cannot directly ensure that its agent is always acting in its (the principals') best interests.
Common examples of this cost include that borne by shareholders (the principal), when corporate management (the agent) buys other companies to expand its power, or spends money on wasteful pet projects, instead of maximizing the value of the corporation's worth; or by the voters of a politician's district (the principal) when the politician (the agent) passes legislation helpful to large contributors to their campaign rather than the voters. Though effects of agency cost are present in any agency relationship, the term is most used in business contexts.
Sources of the costs 
The costs consist of two main sources:
- The costs inherently associated with using an agent (e.g., the risk that agents will use organizational resource for their own benefit) and
- The costs of techniques used to mitigate the problems associated with using an agent—gathering more information on what the agent is doing (e.g., the costs of producing financial statements) or employing mechanisms to align the interests of the agent with those of the principal (e.g. compensating executives with equity payment such as stock options).
Agency costs in corporate governance 
The information asymmetry that exists between shareholders and the Chief Executive Officer is generally considered to be a classic example of a principal–agent problem. The agent (the manager) is working on behalf of the principal (the shareholders), who does not observe the actions, or many of the actions, or is not aware of the repercussions of many of the actions of the agent. Most importantly, even if there was no asymmetric information, the design of the manager's contract would be crucial in order to maintain the relationship between their actions and the interests of shareholders.
Agency costs mainly arise due to contracting costs and the divergence of control, separation of ownership and control and the different objectives (rather than shareholder maximization) the managers.
Professor Michael Jensen of the Harvard Business School and the late Professor William Meckling of the Simon School of Business, University of Rochester wrote an influential paper in 1976 titled "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure". Professor Jensen also wrote an important paper with Eugene Fama of University of Chicago titled "Agency Problems and Residual Claims".
There are various actors in the field and various objectives that can incur costly correctional behaviour. The various actors are mentioned and their objectives are given below.
The classic case of corporate agency cost is the professional manager—specifically the CEO—with only a small stake in ownership, having interests differing from those of firm's owners.
Instead of making the company more efficient and profitable, the CEO may be tempted into:
- empire-building (i.e. increasing the size of the corporation, rather than the size of its profits, "which usually increases the executives' prestige, perks, compensation", etc, but at the expense of the efficiency and thus value of the firm);
- not firing subordinates whose mediocrity or even incompetence may be outweighed by their value as yes-men or golf partners;
- retaining large amounts of cash, that while wasteful, gives independence from capital markets; and of course
- a maximum of compensation with a minimum of "strings"—in the form of pressure to perform—attached.
Venturing onto fraud, management may even manipulate financial figures to optimize bonuses and stock-price-related options.
Bondholders typically value a risk-averse strategy since they do not benefit from higher profits. Stockholders on the other hand have an interest in taking on more risk. If a risky projects succeeds shareholders will get all of the profits themselves, whereas if the projects fail the risk maybe shared with the bondholder (although the bondholder has a higher priority for repayment in case of issuer bankruptcy than a shareholder).
Because bondholders know this, they often have costly and large ex-ante contracts in place prohibiting the management from taking on very risky projects that might arise, or they will simply raise the interest rate demanded, increasing the cost of capital for the company.
Board of directors 
In the literature, the board of directors is typically viewed as aligned with either management or the shareholders.
Labour is sometimes aligned with stockholders and sometimes with management. They too share the same risk-averse strategy, since they cannot diversify their labour whereas the stockholders can diversify their stake in the equity. Risk averse projects reduce the risk of bankruptcy and in turn reduce the chances of job-loss. On the other hand, if the CEO is clearly underperforming then the company is in threat of a hostile takeover which is sometimes associated with job-loss. They are therefore likely to give the CEO considerable leeway in taking risk averse projects, but if the manager is clearly underperforming, they will likely signal that to the stockholders.
Other stakeholders 
Other stakeholders such as the government, suppliers and customers all have their specific interests to look after and that might incur additional costs. Agency costs in the government may include the likes of government wasting taxpayers money to suit their own interest, which may conflict with the general tax-paying public who may want it used elsewhere on things such as health care and education. The literature however mainly focuses on the above categories of agency costs.
Agency costs in agricultural contracts 
While complete contract theory is useful for explaining the terms of agricultural contracts, such as the sharing percentages in tenancy contracts (Steven N. S. Cheung, 1969), agency costs are typically needed to explain their forms. For example, piece rates are preferred for labor tasks where quality is readily observable, e.g. sharpened sugar cane stalks ready for planting. Where effort quality is difficult to observe, e.g. the uniformity of broadcast seeds or fertilizer, wage rates tend to be used. Allen and Lueck (2004) have found that farm organization is strongly influenced by diversity in the form of moral hazard such that crop and household characteristics explain the nature of the farm, even the lack of risk aversion. Roumasset (1995) finds that warranted intensification (e.g. due to land quality) jointly determines optimal specialization on the farm, along with the agency costs of alternative agricultural firms. Where warranted specialization is low, peasant farmers relying on household labor predominate. In high value-per-hectare agriculture, however, there is extensive horizontal specialization by task and vertical specialization between owner, supervisory personnel and workers. These agency theories of farm organization and agricultural allow for multiple shirking possibilities, in contrast to the principal-agency version of sharecropping and agricultural contracts (Stiglitz, 1974, 1988, 1988) which trades-off labor shirking vs. risk-bearing.
See also 
- Pay Without Performance by Lucian Bebchuk and Jesse Fried, Harvard University Press 2004 (preface and introduction)
- Investopedia explains 'Agency Costs'
- Jensen, Michael C.; Meckling, William H. (1976). "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure". Journal of Financial Economics 3 (4): 305–360. doi:10.2139/ssrn.94043. SSRN 94043.
- Fama, Eugene F.; Jenson, Michael C. (1983). "Agency Problems and Residual Claims". Journal of Law & Economics 26: 327–349. doi:10.2139/ssrn.94032. SSRN 94032.
- The Art of Capital Restructuring: Creating Shareholder Value Through Mergers ... By H. Kent Bake, p.60-1
- Bebchuk and Fried, Pay Without Performance (2004), p.16
- What Drives Corporate Excess Cash? Evidence from a Structural Estimation? Hamed Mahmudiy, Michael Pavlin, 2010
- Pay Without Performance - the Unfulfilled Promise of Executive Compensation by Lucian Bebchuk and Jesse Fried, Harvard University Press 2004, 15-17
- Stock Basics: An Investor's Guide Tradeking.com
- Cheung, Steven N S (1969). "Transaction Costs, Risk Aversion, and the Choice of Contractual Arrangements". Journal of Law & Economics. 12 (1): 23–42. doi:10.1086/466658. Retrieved 2009-06-14.
- Allen, Douglas W.; Lueck, Dean (2004). The Nature of the Farm: Contracts, Risk, and Organization in Agriculture. MIT Press. p. 258. ISBN 978-0-262-51185-8.
- Roumasset, James (March, 1995). "The nature of the agricultural firm". Journal of Economic Behavior & Organization 26 (2): 161–177. doi:10.1016/0167-2681(94)00007-2. Retrieved 2009-06-14.
- Stiglitz, Joseph (1974). "Incentives and Risk Sharing in Sharecropping". The Review of Economic Studies (The Review of Economic Studies Ltd.) 41 (2): 219–255. doi:10.2307/2296714. JSTOR 2296714.
- Stiglitz, Joseph (1988). "Principal And Agent". Princeton, Woodrow Wilson School - Discussion Paper (12). Retrieved 2009-06-14.
- Stiglitz, Joseph (1988). "Sharecropping". Princeton, Woodrow Wilson School - Discussion Paper (11). Retrieved 2009-06-14.