One of the most remarkable developments of the past 100 years has been the growth of a substantial literature that documents and analyzes the negative characteristics of the state, such as its inefficiency, ineptness, and immorality. At the same time, defenders of the state, fighting a rearguard action, tout its necessity and egalitarianism. The American Founding Fathers recognized in their debates that the state consists of agents who act on behalf of the citizens as principals even if they did not use these terms. This recognition is the reason for many of the provisions of the U.S. Constitution such as the checks and balances. The development of principal/agent theory in the last few years now makes possible a much deeper and more consistent application of this idea to the state. The central idea in this article and the one that follows is that agents never act fully on behalf of principals and that this discrepancy causes losses that are known as agency costs. This occurs in all organizations in which agents act for principals, including corporations held by shareholders in which decision-making is delegated to managers. The analysis shows that agency costs are far higher in the state than in free market organizations like corporations, and this provides another substantial analysis of the essentially negative nature of the state. The third article in this series focuses on the state as an organization and how its very existence acts negatively vis–vis society.
Many people do not know what the state is. Others mistakenly think that states and major corporations are indistinguishable, both being in their eyes powerful and big. Some of us wish to make the state’s growth and sources of power intelligible, and we wish better to understand its nature. The agency cost analysis clarifies issues like these by outlining some theory of organizations and applying it to the state. The overall result is to de-mystify the state so that we can see it for what it is: nothing more than an organized group of men and women possessing extraordinary power over society. By looking at the state as an organization, we see how it has extended its control over society. We see more clearly the impediments to controlling the state. We glimpse the virtues of replacing it with free markets to accomplish such tasks as defense, justice, roads, and money.
The state is an organization
Franz Oppenheimer called the state the "organization of the political means," and Murray Rothbard defined it as "that organization in society which attempts to maintain a monopoly of the use of force and violence in a given territorial area." Both are right. The state is an organization, which means we can analyze it as such.
Americans have historically distrusted the accumulation of power in any organization, be it a business, a trust, a bank, a labor union, Wall Street, mob bosses, City Hall, or a branch of government. Yet by the usual standards of other private organizations that operate within society, we will see that Americans over a period of 200 or so years have systematically loosened their controls over the American state and allowed it to become a centralized power center far stronger than any private organization in society that was ever regarded with distrust.
We will see that there is a spectrum of political organization ranging from market anarchism to totalitarianism. In market anarchism, the members of society have the maximum control over the institutions that provide the traditional functions of the state. Under market anarchism, there is no single state with extraordinary power. The checks and balances sought by the Founding Fathers in a Constitution are dispersed even further into society where they operate naturally in truly free markets, not the regulated and crippled markets of welfare and regulatory states. With real free markets, taking away natural checks on power, removing them, or overriding them is far more difficult than under a Constitution. At the other end of the spectrum, under totalitarianism, society has the least control over the state. A small body of officials own and operate the state. They rule society. They have power over many details of the lives of society’s members, including the power of life and death.
By looking at the state as an organization, it becomes easier to raise pertinent questions that shed light on how a state works. Who are its members? What are its goals? How does it accomplish its goals? How does it attain power? How does it keep power? Why does it choose particular goals and not others? Even if I were able to, I could not answer all these questions in the space of a few articles. What I provide is a framework for productive thought with enough applications to show that it is a useful and coherent framework that helps us gain insight into the somewhat mysterious entity — organization — that we call the state.
Basics of organizations
There are those who think that because only individuals act, there are no such things as corporations, states, clubs, churches, armies, gangs, etc. They think organizations are somehow convenient or legal fictions, mere words that are being used to disguise or cover up the actions of individuals. After all, only individuals can be responsible for what they do. While it is true that only individuals act and that only individuals are truly responsible for their acts, these do not imply that organizations do not exist and have real meaning. Organizations consist of individuals who coordinate their acts. This does not preclude assigning moral or other responsibilities to those engaged in the acts.
I define an organization as a group of individuals who order their acts through a set of understandings, agreements, or contracts, implicit or explicit. The individuals come into the organization with personal aims, but they constrain their acts when they interact in the organization. They agree to act jointly.
The private and voluntary organizations we are most familiar with, such as charity, company, sports team, gang, club, association, and church, vary a great deal along a number of dimensions that help us differentiate them. The following are some important ways.
(1) Aims. All organizations have goals. The individuals in the organization coordinate for common purposes, and they help define the nature of the organization. The aims vary in clarity. They may be stated or unstated, written or unwritten. They may be well-understood, dimly known, or all shades in between.
(2) Boundaries. Some people are in an organization and others are not. It may not always be easy to tell, but this is a necessary idea. People belong to one church and not another, or to none at all. The Denver Broncos has an exclusive membership as do the U.S. Marines, the Hong Kong Bird Watching Society, and the U.S. Senate. Some organizations are more exclusive than others.
(3) Ownership. Organizations have capital or property (land, buildings, money, etc.) and some people own it. Call them principals. They bear the usual risks of ownership, which may be large or small. They may contract with agents to employ the capital to carry out various tasks.
(4) Contracting. Contracts, agreements or simply understandings, explicit or implicit, are set up to order the behavior of those in the organization. I use the term contracts to cover all this ground as well as the terms principals and agents. Agents act on behalf of principals. In a hierarchy, a person can be both a principal and an agent. For example, workers assemble cars on behalf of executives (as principals) who in turn manage (as agents) on behalf of a Board of Directors (as principals) who in turn (as agents) are responsible to shareholders and lenders, the ultimate principals.
(5) Incentives. The contracts provide incentives that influence how the agents behave.
(6) Agency costs. Michael C. Jensen and William Meckling pioneered the theory of agency costs in 1976. Agency costs are contracting costs, costs that inhere in contracts. They are (1) the costs of getting together to make the contracts, (2) the costs to monitor, bond and enforce the stipulated behavior, and (3) most important of all, the costs of the agents not living up to their obligations.
When agents do not fully live up to their obligations, which is usual, they don’t do fully what the principals want them to do. The principals lose, or there is a loss in value. The shortfall between the values the principals prefer to achieve and what they end up achieving when the agents act for them is part of the agency cost. The principals still employ the agents because they are better off than if they hadn’t, but they do not achieve as high a value as they might have if there were not agency costs. Agency costs are a necessary friction and part of organized human activity. Think of the old saying "While the cat’s away, the mice will play." If these costs get big enough, then the principals will stop employing the agents. There is good reason for entrepreneurs to devise new ways to keep these costs low in order to increase the size of the pie.
The state’s uniqueness
We can use the same six concepts to analyze the state. While this is very useful, at the same time we find that some of them either do not apply or apply in unusual ways. This tells us we are dealing with a very unusual organization that differs radically from other organizations. In fact, as we know, the state is not simply another organization, because it has power and it uses this power, when it wishes, to prevent any other group in society from being its rival. The state’s position as the maker of laws and imposer of taxes makes it different in kind. The state can coerce members of society without fear of punishment. No other group in society can use aggressive force and make others do things against their will without facing sanctions. The state can.
Thinking of taxpayers as principals and state officials as their agents leads us to understand why states tend not to work well for citizens and why free markets work better. But that is all the principal/agent model can do for us, help us analyze the state. When we analyze foods in terms of vitamins, we find that some are rich in nutrients and some have empty calories. By the same token, when we analyze the state in contractual terms, we shouldn’t fall into the trap of thinking that the members of society have come together and unanimously consented to create another kind of special corporation that will do particular tasks for them, or that there is some sort of social contract that has mysteriously arisen between voter and state. The state is relatively empty, some would say completely empty, of contractual content, like a food with empty calories.
Compare a corporation with many shareholders whose stock trades on an exchange. A shareholder in such a company can exit ownership at any time by selling the stock; and therefore all the owners hold the stock willingly and unanimously at a given time. A citizen of a country has no comparable low-cost means of terminating whatever arrangement he has with the state. Shareholders are not stuck with their company, but citizens are stuck with their state. Furthermore, voters disagree all the time about what they want the state to do. At any given time a large part of the electorate disagrees with what the state is doing. There is no voter unanimity, and if a voter disagrees he can’t do much about it. Shareholders for the most part are in agreement, virtually unanimous agreement, about one thing. They want their agents to select investments that raise the price of the stock. If they do not like what managers are doing, they can do something about it. They can part company by selling the stock.
Whereas private organizations are based on agreements that can be terminated and altered under terms of the agreements, and whereas the parties to these agreements continually participate in them, the state acts in mysterious ways according to laws it has promulgated years before citizens were born and it continues to impose new laws that a citizen cannot escape. Paradoxically, it is useful to analyze the state in terms of contract if only to see that the concept does not apply or applies only with grave and uncorrectable defects to the state.
Who bears agency costs?
The situation that arises again and again in many transactions is that the people who supply capital are not always the same people who demand capital. Savers transfer savings to business firms or those who have better investment opportunities. The savers then get higher returns. This process is the basis of an enormous increase in a society’s prosperity, but it always entails agency costs. Analogously, taxpayers are made to transfer funds to the state. Again, agency costs are involved. The analysis will show that the size, nature, and resolution of these costs is far different in the two cases.
It is impossible to align the interests of agents fully with those of principals. For example, in a firm owned by stockholders and managed by managers, the managers are agents and the stockholders are principals. Because shareholders can’t perfectly monitor what managers do on their behalf, managers waste some resources. If they dissipate a dollar of assets, which is a loss to the shareholders, the managers personally lose less than a dollar. Hence the managers (agents) make some investments that lower shareholder wealth, consume perquisites such as fancy surroundings and expensive cuisine that lower shareholder wealth, and fail to make some investments that would raise shareholder wealth. These losses to the shareholders (principals) are agency costs.
Who bears agency costs? It depends on the organizational setup. For corporations, the shareholders expect the managers to misbehave to a certain extent. They know that their behavior can’t be perfectly observed and controlled. They price in these prospective losses when supplying equity funds, that is, when they buy and sell stocks in the open market. They require a higher rate of return on securities. These expected costs then fall upon the demanders of capital (issuers of securities), not on the shareholders who supply capital (buyers of securities.)
If we provisionally think of a state as something like a corporation, with politicians as agents and voters as principals, again there will be agency costs. The voters exercise imperfect control over the politicians. The agency costs are the losses in value that occur when the state wastes taxpayer funds on projects that bring taxpayers no benefits and fails to use those funds for projects that would bring taxpayers benefits. For example, the citizens may want security from foreign attacks and instead pay an exorbitant price for insecurity due to unwise foreign wars. The citizens may want effective justice and instead get an expensive system that penalizes victimless crimes. But in fact the situation is even worse because the citizens do not have monolithic tastes for particular goods. Having a state is like having stockholders supply funds to a corporation that has no clear idea what goods to produce and produces all sorts of products haphazardly. Making matters still worse than this is that the funds are not provided voluntarily. The state extracts funds from its "shareholders" whether they like it or not.
The state’s poor choices and its waste create agency costs, and then, adding insult to injury, the state shifts the agency costs onto the group that finances it! The taxpayers and citizens bear the agency costs when their agents in the state misuse capital and/or act in ways the citizens dislike. This double-barreled whammy occurs because the state both imposes taxes and rules. The taxpayers have no choice but to pay; the citizens must obey. There is no market mechanism by which taxpayers can price in the expected agency costs. There is no easy escape hatch for citizens. It’s true that voters can staunch the agency costs in elections, or attempt to limit them in other ways. But it’s also apparent that voters pay all the bills and members of society suffer the vast majority of all the consequences of what their agents do in government, including the agency costs.
Corporations and states borrow money, and bond financing also causes agency costs. If a company finances by debt, agency costs crop up as (1) substitution of more risky for less risky projects, (2) underinvestment in wealth-increasing projects, and (3) potential bankruptcy costs. How so? The returns to a risky project mainly go to the shareholders, not the lenders. This gives managers an incentive to substitute more risky for less risky projects, and this adds risk to the bonds that bondholders dislike. Underinvestment occurs because the returns to any new investment partly benefit bondholders. Who bears these costs? Not bondholders. When stockholders borrow, the agency costs of debt fall upon them, not the lenders, because the lenders price in their expected agency costs via higher interest rates. It’s easy to see that higher expected bankruptcy costs cause higher lending rates.
Again, the state’s situation is different. The state’s agency costs of debt fall on taxpayers, not on the state’s members. Congress, for example, borrows and undertakes many projects that the body of taxpayers may not like. Members of Congress gain in various ways via these investments, but if they don’t pay off the taxpayers bear the losses. Congress overinvests in low and negative return projects like wars, space programs, and pork. At the same time, it underinvests in projects that raise taxpayer utility and do not give big gains to the Congressmen like levees or it forces underinvestment in oil refineries and power grids. Additionally, since the state can raise taxes to prevent bankruptcy, it unduly risks bankruptcy by issuing too many debt and other obligations. It overborrows.
Internal agency costs
Organizations have agency relationships within them, and these generate internal agency costs. "While the cat’s away, the mice will play," but for economic reasons, higher-ups delegate work despite the fact of agency costs. Top managers delegate decisions to lower managers and other employees because, for one thing, employees know more about certain specific matters than their bosses. There is division of labor. This process is not without its agency costs though. In this situation, the top managers act as principals and the lower managers are the agents, and the principals cannot fully control or even observe all the behavior of their agents.
The delegation involves agency costs inside the company. Internal agency costs include losses when the agents do not act in the interests of the principals, such as by shirking or missing out on good opportunities. They do not act fully in the boss’s interests because (1) they do not fully receive the benefits of doing so, (2) they have their own interests, and (3) bosses can’t costlessly monitor them. There are methods to reduce agency costs, and since agency costs in firms fall upon the agents, they accept measures that reduce these costs. These include internal budgeting and auditing, following orders and procedures, and monitoring of their work. Without these the managers would have to pay a lower wage because there would be more losses caused by employees. This is why employees accept these procedures, although it may sound far-fetched. (Things are not always what they seem when they are analyzed using economic analysis.)
The same idea applies to the workings within a state. There are frictions between Congress and the Executive, between Congress and its agencies, and between the top executives of these agencies and the employees in the bureaucracies. There are frictions and agency costs between the Executive and its agents, such as the CIA and the military. There are agency costs among the different armed services. This is not a complete list.
For example, the voters elect Congress. There is a first layer of agency costs right there. Congress acts as principal and it sets up a bureau like the Food and Drug Administration. Its members are the agents of Congress, a second layer. Next, the heads of the FDA delegate to those below, a third layer.
To begin to understand why the state is far less efficient than a business, notice that the FDA is not a profit-seeking organization. This implies that it lacks proper incentives to act on behalf of its customers, which, by the way, are not clear even to them. The FDA can’t obtain proper price signals to deliver services because it doesn’t operate in a market. Furthermore, the FDA’s civil service employees are entrenched, which makes controlling their behavior more difficult. Businesses that have to adhere to arbitrary work rules or union practices enforced by the state’s laws face similar issues.
On numerous occasions Congress considers and sometimes passes bills that delegate unconstitutional power to international bodies such as the U.N. or the WTO. Instead of a bureau within the U.S. controlling U.S. citizens, bureaus and authorities outside the U.S. gain such power. Congress has at least some control over internal agencies through budgets, reports, hearings, and appointments, but its control over international bodies is an order of magnitude lower. Accordingly, the agency costs to the typical American citizens rise steeply. Their control over these international bodies is so indirect and so diluted as to be nonexistent. Indeed, Americans end up being controlled. The international case is more dramatic but not different in kind from the domestic case. In the latter, citizens also end up on the receiving end of control exercised by the state.
We see that where the state is involved, numerous hierarchies incur internal agency costs. Legislators delegate funds to the executive branch. The top officials there delegate to the FEMAs and HUDs. They in turn delegate to lower-level employees. There are layers and layers of agency costs. The same thing occurs in companies, but with two significant differences. Companies have incentives (a) to minimize the agency costs and (b) adopt new and efficient organizational structures. If they don’t, competitors can put them out of business. In states, the entire structure of monitoring and incentives is weak because the taxpayers do not control their agents and because the state shovels the agency costs back upon the taxpayers. Periodic revelations of waste, scandals and frauds lead to no lasting reform. This shows that voting is a rather poor control device to control agency costs. The voters often do not know how big the waste is or who’s responsible. They do not know what their single vote means when there are dozens of programs and millions of other voters. It is quite clear that the taxpayer/state institutional structure fails to control agency costs as well as they are controlled in free markets. In a market, lenders can withhold money. Stockholders can sell their stocks. What can taxpayers do? They can’t sell their stock in the state. At best they can replace their current blue agents in Congress and the Executive with red ones. The new guys will misbehave in their own peculiar ways to be sure, but they will impose losses on society too because the institutional structure and incentives of the state remain exactly the same.
The next installment in this three-part series compares the free market control over agency costs with voter control in a state in greater detail. This buttresses the theoretical case that the state is less efficient than the free market. The article looks at some important events in American history and points out how, in terms of control over the state, the voters came to be the group being controlled and not the group doing the controlling.
Michael S. Rozeff [send him mail] is the Louis M. Jacobs Professor of Finance at University at Buffalo.