by Jurgen Brauer, April 2007
Copyright: J. Brauer. No reproduction without permission.

There is no “government activity that has never been done or is not now being done by the private firm.”

The privatization of government functions is still the rage, much over the world. Privatizing railroads, privatizing prisons, privatizing streets (toll roads), privatizing waste hauling – there is no “government activity that has never been done or is not now being done by the private firm,” writes prominent professor James Q. Wilson – now at Pepperdine University – in his 1989 book Bureaucracy.

The privatization of government functions has many roots but one of them surely lies in the often correct claim that the latter are inefficient. But if government operations are inefficient, why then do they persist? Why do we not see full-scale privatization of every government function? Is it simply that politicians and bureaucrats wish to maintain a degree of power or are there more deep-seated reasons?

In an important article published in 1999, another professor, Oliver E. Williamson, who teaches business, economics, and law at the University of California, Berkeley, provides one answer. Taking his cues from an area of economics called transaction cost economics, he argues that the cost of doing business depends not only on technical factors such as the cost of labor, capital, materials, and the degree of competition among firms but also on so-called agency costs. In addition to technical efficiency, firms must also try to be agency efficient, he says.

“Agency” refers to a firm’s shareholders, managers, employees, suppliers, customers and others, namely all those with whom the firm has some sort of relation that might put its operations at a risk that cannot reasonably be resolved by later recourse to contracts and the courts. For example, a construction firm building a slaughterhouse for a chicken-processor in rural Arkansas or Georgia is making itself dependent on contract fulfillment by the chicken-processor as the slaughterhouse, once built, cannot readily be moved to another location. We say that the building is “asset-specific,” an asset with a specific, restricted, non-malleable use.

The more uncertain we are about cooperating with people in Uganda, Ukraine, Uruguay, or, for that matter, Utah, the higher the implicit risk we take.

Presumably, the construction firm will write a contract that considerably reduces the uncertainty of being paid in full and on time by the chicken-processor. But few contracts can foresee every eventuality – which is why so many contract disputes end up in court. Just like asset-specificity, limits to rationality (“foreseeability”) therefore add to agency costs. Contracts are uncertain. The more uncertain we are about cooperating with people in Uganda, Ukraine, Uruguay, or, for that matter, Utah, the higher the implicit risk we take. Uncertainty can sometimes be reduced when one deals repeatedly with the same people. In this case, one relies not just on the current transaction but also on the expectation of future transaction, so that the prospect of the future holds some sway over one’s behavior in the present. Thus, regularity or frequency of interaction, along with uncertainty and asset specificity need to be taken into account in decision making, and they need to be made as least costly – as efficient – as possible. This is agency efficiency. Think of it, loosely, as risk-management writ large.

If risks such as these are too great or too difficult to handle via contracts, it may be smart to move the transaction in-house instead of using the market. Transaction cost economics thus explains just why some transactions are contracted out whereas many others remain in-house. It at least partially explains why firms exist and why they are organized the way they are. This is a great accomplishment of the theory. Many firms contract out cleaning-services but if the CIA did so, the security risks would be unbearable and it therefore will just have to have its own cleaning staff, even if it is more expensive than outsourcing the task.

"Probity" refers to the "loyalty and rectitude" with which government's sovereign transactions are carried out.

Now, consider government (other than the CIA), specifically government’s foreign transactions, for example, its diplomatic relations to other sovereign states. In addition to the agency risks entailed by asset specificity, uncertainty, and frequency, Professor Williamson adds “probity,” by which he means the “loyalty and rectitude” with which government’s sovereign transactions are carried out.

In contrast to private firms, the quality of the U.S. diplomatic service depends on low-powered incentives. For example, we do not want diplomats to engage in undue zeal to earn the next promotion or a year-end bonus the way we might in a private market setup where high-powered incentives tend to prevail. We do not want foreign service officers to advance their own proposals directly to foreign powers and only check in with the boss later on (the President of the United States!) as we might in a private business. Instead we want them to say, and only to say, whatever the directives from the President and the Secretary of State dictate. Rules, regulations, and procedures must be followed, even if to a seemingly stifling degree, or a degree that might at any rate not be deemed proper in private business. Whereas private business has some but not much use anymore for careerists who spend the entirety of their work lives at a single firm, in government careerists are still valued. Fairly low but steady pay, combined with for-cause rather than at-will termination of employment are the rule. The reason, once more, is that we want to employ and keep people who in exchange for a near-guarantee of long-term employment and good health care and retirement benefits commit to follow the directives of their political masters. Creativity is discouraged. All we want is to curry the “loyalty and rectitude” of these employees.

The seeming inefficiency of many a government or public bureau, in contrast to that of private organizations, can thus be explained by the high agency cost of “probity,” or rather the risk entailed when probity might be violated. Imagine what could happen if the U.S. armed forces were fully privatized. It is one thing to have active-duty and reserve forces organized under the auspices of the U.S. Department of Defense; it is quite another to contract-out combat services to private companies. There are, however, very many military-related functions that have been contracted out, items such as meal services, laundry, and latrine cleaning. They are important services, to be sure, but when a contractor fails, this does not put the security and safety of the country at undue risk. Just how far along the risk-spectrum one may venture is a question that will only be answered by experiment and experience. Risks associated with uncertainty, asset-specificity, frequency, and probity need to be carefully weighed. If they appear to outweigh traditional cost savings to be had from privatizing government services, then it is in fact more efficient to retain the service within a government bureau – the cost of failure is too large to countenance.

Economists and other scientists love to study extremes as they often tell us more readily than otherwise about some underlying factors or principles that govern observed behavior.

The examples of a country’s diplomatic and military services are chosen precisely because it is “obvious” that there limits to contracting them out. Economists and other scientists love to study extremes as they often tell us more readily than otherwise about some underlying factors or principles that govern observed behavior. Uncertainty, asset-specificity, frequency, and probity – in addition to the usual cost factors of personnel, materials, and equipment – are important determinants which help decide just where along the spectrum from the purely private to the purely public tasks will be organized. Most of the time, the private sector has the advantage; but sometimes it is preferable – more efficient – to keep transactions within the public bureau.