It’s a serious argument anchored in principal-agent theory that should be addressed on its merits.
Principal-agent problem
The government wants something done, such as providing telecom services or leisure services for citizens. It has a choice. It can do what needs to be done directly, through its own employees, or it can do it through a private company. In the latter case, there is a clear principal-agent relationship. Government is the principal; the private company is the agent.
Principal-agent relationships are bedeviled by two inter-related problems: the agent’s actions are driven by incentives different from those of the principal; and the agent has better knowledge about what is going on than the principal. The knowledge asymmetry makes the alignment of incentives difficult. The solution is a contract that aligns the incentives of the two parties and, to the extent possible, ensures the flow of accurate and relevant information to the principal. Think of an arrangement governing the working of paddy fields by non-owners. Who provides the inputs? Who bears the risks? How is the owner’s share determined?
Regulation, through detailed concession contracts and licenses as well as through technically proficient specialized regulatory agencies, is the modern solution to the principal-agent problem in government supply of services. But it is simplistic to assume that the principal-agent problem exists only when government services are outsourced to private entities.
Railway services, electricity supply, etc. were originally supplied by private companies. These activities were not subject to marketplace discipline and, to use language from the American courts, “affected with the public interest.†Therefore, very quickly, governments became involved in price setting, ensuring non-discrimination in the supply of services and so on. But they had great difficulty in aligning the incentives of the private entities to those of government. Even more frustrating was the inability of government to extract accurate information from the private entities.
So, as governments are wont, they decided to nationalize the private entities (in most countries, except the US, the Philippines and a few Latin American countries). It was thought that direct government supply was a solution to the principal-agent problem.
Internalizing the principal-agent problem
But alas, it reappeared. The incentives of the managers appointed to run these enterprises were different from those of the “government†represented by the Minister. Indeed the Minister’s incentives were different from those of the government as a whole; and the incentives of the governing party different from those of the people who elected them to power. The managers obviously had better information about the supply of electricity or whatever than the Minister; the Minister had better information than the government and so on. So the government could not be sure that the Minister was doing what it wanted done; the Minister could not be sure the managers were doing what she wanted done and so on.
There was no further nationalization to be done. So one had various crude solutions, such as the appointment of trusted “working directors†and even physical assaults on managers by Ministers. Various dysfunctional equilibriums emerged. The managers allowed the Minister to stuff the organization with political appointees and use public resources for elections; the Minister allowed managers to pay themselves multiple allowances disconnected from performance. Inefficiency reigned. Losses mounted. The taxpayer paid the bills.
Outsourcing based on detailed contracts, regulation by specialized agencies, etc. were proposed as solutions. The beneficiaries of the prior arrangements tried their best to stall the reforms, pointing out the difficulties of solving the “pure†principal-agent problem. Will the private entities not subvert the contracts, however carefully worded? Will the regulatory agencies be competent? Will they be able to extract accurate information from the regulated entities?
So we have dueling examples: SAGT is not only profitable for its owners JKH; it keeps afloat its “principal†the Sri Lanka Ports Authority. One for the proponents of private enterprise. See how the quality of service has not deteriorated at Water’s Edge and at Lanka Hospitals. Two for the proponents of government ownership. The debate is at a stalemate, at least in Sri Lanka.
Can government organizations take risks?
This column was triggered by my recent visit to Colombia, where the same debates are playing out. There, municipal governments are strong, supplying, among other things, fixed telephony and broadband services. In the interests of ensuring good governance and safeguarding the public patrimony, the central government and municipalities are subject to the supervision of “Controlarias,†powerful independent audit authorities.
One such Controlaria had hauled up a former Mayor for wasting the public patrimony by offering ISDN services. This was a primitive form of data communication (128 kbps) that was flogged by US telecom equipment suppliers in the 1990s (we used to call it “I Still Don’t [k]Nowâ€). It was deployed all over the world, including by SLT in 1999, and sank without a trace. Presumably, the Mayor was advised by his managers that ISDN was the way to go; he had said yes; the service bombed and the municipality lost some money. Years later the Controlaria was demanding that the losses be covered by the ex-Mayor out of his own pocket.
For those in the sector, this looks patently wrong. Lots of new technologies come along accompanied by great hype. ISDN, DECT, videotext, WiMAX, even CDMA. Some fail instantly. Others fail slowly. A few succeed. No one can tell what will win in the marketplace. There is no alternative to taking risks. It is easy to laugh at ISDN now (though I was skeptical even in the 1990s). But many serious people believed in ISDN then and many serious companies offered the service. Why pick on the poor Mayor?
But there is a difference between taking risks with private funds and taking risks with the public patrimony. In the former, one simply has a hard time at the AGM or is handed a golden parachute. In the latter, one has to deal with the Controlaria in one’s old age.
I was reminded of a question I was asked about hedging by the Ceylon Petroleum Corporation. I said that any normal enterprise would engage in hedging where one would win some times and lose on others; gave examples of Southwest Airlines and even a local company exporting cables.
But CPC was not a normal enterprise, I said. It could not afford to lose. If it could not afford to lose, it should not engage in hedging. If it could not engage in hedging, it could not be an efficient player in the market. This was the problem with government ownership.
So, I will add another element to the government supply of services debate: who is better equipped to take risks? There can be no innovation without risk. Do we want government-supplied services that lack innovation? Are we capable of distinguishing risk taking from wasting the public patrimony?
Rohan Samarajiva heads LirneAsia, a regional think tank. He was also a former telecoms regulator in Sri Lanka. To read previous columns go to LBOs main navigation panel and click on the 'Choices' category.