In Depth: Government-Owned Enterprises
Direct government ownership of enterprises are among the most complex subsidies to evaluate. Nonetheless, they are important to examine closely. Around the world, government ownership of core production, refining, and distribution infrastructure is common in the energy industry. Included would be all construction, maintenance, and operating costs of energy-related facilities or service organizations owned directly by the federal government. Within the United States, federally-owned Power Marketing Administrations are a good example of this category of intervention. Uranium enrichment has historically been publicly-owned, though there have been recent efforts to privatize this capability in some countries. This function was privatized in the US, though required the taxpayer to retain large legacy liabilities. Also included are unrecovered investments into such items as transportation infrastructure which benefits energy as a major user of the waterways, although other groups may also benefit. These subsidies are prorated based on the energy sector's share of total use.
Government-ownership provides direct subsidies in a number of ways. The activity may not charge enough money even to cover its annual running costs (operating losses). Even if it covers annual operating costs, it may not be returning an appropriate amount of revenues to pay for the original capital (the dams, harbors, plants, etc.) For activities that have substantial post-closure liabilities (either through environmental cleanup or pension/health care obligations of workers), current earnings may be too low to properly accrue for these known future costs. In addition, the fact that government enterprises absorb significant operating risks, yet do not have any required rate of return for investors (the taxpayer) reduces the costs borne by the enterprise still further, increasing the barriers to entry for substitute energy services.1 Finally, government enterprises may cross-subsidize within its organization, offsetting losses in one area by increasing prices in another. Since the government has monopoly power in many of its areas of operation, cross-subsidies are much easier to maintain than for competitive private companies. These mechanisms of subsidization are discussed below:
Not all spending on plant and equipment proves to be productive, or productive for as long as expected. In such cases, the now worthless capital must be written off as a capital loss.2 In a private industry this loss is generally written off as soon as possible. Government assets in some agencies are also written off quickly once losses are recognized. However, until recently few federal enterprises had comprehensive-enough accrual accounting for these types of problems to show up.
Valuing subsidies through capital write-offs is a somewhat different challenge than that which would be used in preparing corporate accounts. Rather than develop accurate accounts for a particular enterprise, we are interested primarily in constructing a "steady-state" capital flow to approximate the annual costs of a given government activity. Therefore, capital losses are amortized back in time over the estimated period of loss. For example, billions of dollars of defunct investment into the Uranium Enrichment Enterprise (now operating as the privatized U.S. Enrichment Corporation) since 1969 were not recovered through sales of enriched uranium. Rather than recognizing the total loss in 1989, we converted them into an annual "premium" for these 20 years. This premium has the same net present value as writing off the losses entirely in 1989, but better reflects the fact that the UEE pricing of services or recognition of capital write-offs throughout the period was too low.3 This premium is a measure of how much more UEE would have put aside to protect against losses if it had known in 1969 how things were going to turn out.
As a practical matter, write-offs prior to the year of analysis will not affect that year's subsidy estimates. In some cases, the period of loss may go forward in time rather than backwards. For example, a modification of a loan or sales contract may reduce the future cash flow stream. In such cases, losses are amortized forwards rather than backwards. Using annualized subsidies yields a better approximation the annual value of subsidies to a particular energy sector.
Estimates of the magnitude of unrecovered federal government capital are often quite controversial. For example, independent estimates for the Uranium Enrichment Enterprise range from $3-10 billion. Much of this disparity is due to differing assumptions on the period of loss. In addition, part of the controversy stems from what price constitutes a reasonable charge for the sale of government assets to the private sector. This price, in turn, is determined by which operations are included in the amount the government is trying to recover through privatization. Part of the issue may also be political since agencies with large capital losses may be viewed as inefficient or mismanaged.
Losses in government operations may originate from current operations as well as from poor investment decisions regarding capital or R&D. Like capital losses, operating losses are usually covered through additional taxpayer money. Over the long-term, however, operating subsidies and capital losses may show up as "unrecovered federal investment," or some similar classification in entity accounting reports.
Absence of Accruals for Future Costs
Government entities, like private entities, must over time bear costs to close certain facilities or clean up certain wastes. While the timing and exact amounts of these outlays may be imprecise, it is clear to all participants that these are real costs which must be paid. As with issues such as pension accounting in the private sector, government endeavors that do not provide for these costs over the life of the facility confer a subsidy to the current users of the product or service created, making it appear cheaper than it really is. The clean-up costs (or similar expense) will then have to be borne either by future users or by future taxpayers, neither of whom benefited from the initial facility. Sometimes these costs can be imputed using external studies or engineering estimates. When there is no information on the expected costs of future cleanups, current cleanup expenditures can serve as a second-best proxy when available.4
No Rate of Return
Private entities will not stay in business unless they earn a high enough return to satisfy the various claims on funds - primarily from the owners and lenders. While the lack of a rate of return for government enterprises is not an out-of-pocket taxpayer subsidy, it does reduce the operating costs of one enterprise, creating a cost advantage over substitute energy forms. This advantage is not due to the underlying economics of the energy type, but rather to the type of owner.
Without a rate of return, the government is essentially absorbing operating and investment risks for free. Most activities involve some risk, both financial and otherwise. Areas in which national governments are traditionally involved, such as uranium enrichment, nuclear waste disposal, or constructing ports and harbors, often involve tremendous risk. Were a private entity completing these tasks, their prices would be driven up and their earnings would be driven down through a process of sharing risk with other parties -- through insurance, and sharing returns (to compensate for the higher risk) via higher interest rates from debt holders or a higher required rate of return from equity holders. Government entities do few of these things. Risks are mostly borne through government guarantees that whatever future damage or losses occur will be paid by taxpayers.5
One argument suggests that this form of risk internalization (de facto self-insurance) does not constitute a subsidy at all, since damages are paid for by the same entity that would have purchased insurance in the first place (the federal government). However, true self-insurance requires periodic risk assessments and the creation of provisions for expected losses. In the absence of such provisions, since damages are always uncertain and often irreversible, paying them after they occur does not generate the price signals (such as high insurance premiums and borrowing rates) that allow an investor to tell whether one future investment is more or less risky than another.
Furthermore, the party bearing the risk (the amorphous taxpayer) is, in fact, very different from the party gaining the benefit of lower current operating costs (e.g., the consumer of publicly-owned power). This differs from most self-insurance plans where the managers responsible for safety-related choices bear at least some financial risk (such as through stock holdings) for making the wrong choices. As a result of poor price signals and divergent interests of the actors, government self-insurance does little to ensure that relative risks are incorporated into investment choices.
Below-Market Provision of Services and Cross-Subsidies
Even where revenues earned by government activities adequately cover the costs of production, distortionary subsidies may remain if goods and services are marketed at below the market price. This commonly occurs with the sale of publicly-owned natural resource assets (e.g., timber, oil), where the acquisition cost of the resource to the government is at or near zero. Below-market sales accelerate resource depletion, rob the taxpayer of an appropriate return on the resource endowment, and sometimes exacerbate production-related environmental problems. However, below-market sales occur in other areas as well -- often with government-owned power resources, for example.
Cross-subsidization, where the prices for one product (e.g., irrigation for farmers) are subsidized with increased prices on another, more dependent or less-powerful customer (e.g., Bonneville Power Administration electricity sales to consumers), can introduce market distortions as well. Where identifiable, large cross-subsidies can be appropriately tracked in a subsidy analysis. In some cases, corrections for cross-subsidies can actually result in negative subsidy estimates for some fuels in some categories (for example where irrigation is cross-subsidized by hydroelectric power sales from the same dam).
Many federal enterprises provide energy goods and services which compete with private suppliers, but which are exempt from paying federal taxes. Because public firms do not pay federal taxes, do not require a rate of return, and often have the ability to sustain very large long-term losses, they can price their services below market, and net earnings will be non-existent. Again, the historical public ownership of uranium enrichment within the United States provides a salient example. At first glance, it would appear that the tax-exemption for the Uranium Enrichment Enterprise (the name prior to privatization), for example, is worthless since it always loses money, and no net income implies no tax liability.
This conclusion would be very wrong. UEE had a great deal of control over its product pricing for much of its operating life and could have earned market rates of return. It is because capital still flowed to this use regardless of the after-tax return on investment, that UEE -- and all of the government owned product and service organizations -- were able to price at the level they have. It was this depressed pricing that, in turn, created barriers to entry for alternative supply and demand technologies that had to earn a return on investment sufficient to pay investors and the federal tax coffers.
Government-owned enterprises compete in a number of ways. Some, such as the Army Corps waterways repair work, may compete directly with private firms that are capable of providing the same service. Other enterprises may not currently have direct private competitors. In some cases, this may be due only to the market power (through size or regulation) that the federal enterprise exerts, keeping available private substitutes out. For example, the private sector could build power plants to compete with the Power Marketing Administrations, or the PMAs themselves could be privately owned and operated. In other cases, economies of scale may preclude cost-effective direct competition, as with uranium enrichment or the Strategic Petroleum Reserve. Even in these cases, the tax-exempt status reduces the cost structure of the federally-owned enterprise, placing competing energy resources which must pay taxes and earn a rate of return, at a disadvantage.6
- 1. While some areas of government involvement, such as uranium enrichment, have historically been forbidden for private enterprise, the lack of an adequate return on capital still affects energy markets by making nuclear power more attractive relative to energy services provided primarily by private entities who don't have the luxury to operate it as a non-profit enterprise.
- 2. Losses in joint government-private ventures would be allocated to the owners based on share of ownership. Capital losses reflect a write-down of undepreciated capital.
- 3. Government capital write-offs sometimes result from improper depreciation of capital purchases. Depreciating capital purchases generate very different behavior in private sector tax accounting versus the public sector. The private sector will choose to minimize the write-off period since this maximizes tax benefits. The public sector agencies that depreciate their assets (not all do) may sometimes use the estimated service life, but will often choose to maximize the write-off period since this may help to improve the apparent cost-effectiveness of current activities.
- 4. The use of such proxies creates errors in the matching of cleanup costs and historic and current activity. For example, cleanup at DOE sites is just beginning and costs will likely escalate considerably. Thus, using current expenditures as a proxy for the accruals for future cleanup costs is likely to be too low, although technological developments may help decrease some costs over time.
- 5. Some agencies do purchase some insurance. For example, BPA does pay for property and liability insurance as required by the Price-Anderson act for its operation of nuclear plants. These costs are passed through to BPA by the parties with whom BPA contracts. However, even in cases such as this, a substantial portion of the operating risk may well be borne by the government and taxpayers, rather than the beneficiaries of the enterprise.
- 6. For a more detailed description of the divergent interests ("agency theory") see Michael Jensen and William Meckling, "Specific and General Knowledge, and Organizational Structure," presented at Nobel Symposium No. 77, Contracts: Determinants, Properties and Implications, Stockholm, August 18-20, 1990; and Eugene Fama and Michael Jensen, "Separation of Ownership and Control," Journal of Law and Economics, Vol. XXVI, June 1983, pp. 301-325.