of imaginary tax payments from the rate base.

The interaction of full normalization with the tax benefits creates the

pattern in Figure 1. Annual profit is at its highest levels during the construction period and the first years of the operating period. Profit is actually negative for the remaining 20 years of operation and for the decommissioning period. The exception in this last 27 years is the first year after operations cease. It is assumed that all fuel costs, including spent fuel disposal, are charged to expenses during plant operations. The first year after the plant ceases operations, there is a sizable spent fuel disposal cost. This was previously charged to customers, and to expenses. But for tax purposes, this spent fuel cost is a deduction only in the time period in which it actually occurs, creating a tax deduction applied to income from other facilities. the anomaly of a profit.

Hence

Current tax expense is negative during the construction period, modest in the first 10 years of operation, and is largest in the last 20 years of

operation.

This time pattern provides a financial incentive for premature construction of nuclear power plants. Because of the tax benefits and the allowance for funds used during construction, the plant earns a profit valued at an accumulated value of $700 million during its construction period.

It should be pointed out that this timing pattern, as peculiar as it may seem, does not give shareholders an excessive return over the full 47-year period. The model indicates that, if shareholders had invested their construction money in other funds, the amount accumulating over 47 years would be equivalent in value to the accumulated value of the peculiar profit stream in Figure 1. However, the amortized present value of the current tax expenses is actually slightly negative for the Figure 1 values. In other words, with

present deductions, credits, and exclusions, the multi-billion dollar revenue from a new power plant is--over the full economic period of the facility--exempt from corporate income tax liability..

The significance of this tax subsidy is evident in Table 2. The cost of the hypothetical nuclear plant discussed above with the 1981 Tax Act provisions

is 15.6¢/kWh over the operating period 1990-2019. It should be remembered

that, through normalization, shareholders are earning a fair return and, eventually,

customers share in the tax benefits.

In the next column, the after-tax cost with the previous provisions indicate 16.1¢/kWh. The effect of the 1981 Tax Act is seen to be an increase in the value of tax subsidies to nuclear power.

The third column shows after-tax cost with no tax subsidies.

There is no

accelerated cost recovery system, no investment tax credit, and no interest deductions. AFUDC income, however, remains exempt when earned. The result: 22.5¢/kWh. Nuclear power, then, is attaining a tax subsidy of 6.9¢/kWh, equal to 44% of its cost to the utility. The typical 1,000 megawatt plant studied here produces 5.3 billion kWh in a typical year. The annual subsidy, then, is $365 million per year for this plant. The construction of an additional 100,000 megawatts of new capacity as currently planned creates a national annual equivalent subsidy for these plants which is approximately $35 to $40 billion each year. As noted above, the net amortized tax liability is negative for each new plant.

3. COAL AND NUCLEAR POWER ECONOMICS

The second row in Table 2 shows similar coal plant data from analyses prepared by Kathleen Cole. The economic assumptions are also given in the Appendix. The tax provisions are similar, except that the tax life for a coal

plant is now 15 years for a plant expected to operate 35 years. This is the only significant qualitative difference between coal and nuclear power taxation. The nuclear plant is enabled to claim full tax depreciation in 10 years, the

coal plant in 15.

The after-tax cost of the coal plant is 16.5¢/kWh, above the nuclear cost. The unsubsidized cost for coal would be 19.84/kWh. The corporate income tax system itself reversed the ordering. The subsidy accruing to nuclear power is more than twice that received by coal plants.

The coal subsidy is only increased by three-tenths of 14/kWh by the 1981 Tax Act. The significance of the new tax provisions is that they add an additional emphasis to an existing major tax advantage held by nuclear power.

Having made the point about the significance of tax provisions to nuclear power, it should be recognized that Table 2 does not indicate the full economic picture to an alert utility management. The coal capital cost of $836/kWh in 1980 dollars was selected from the actual reported experience of a New York utility in the process of constructing a new coal plant with full sulfur removal equipment. Hence the coal estimate might have been viewed as realistic.

The nuclear estimate, however, was a planning figure given by an architect engineer. That value of $966/kWh is considerably less than the costs currently being experienced. In addition, Table 2 and Figure 1 assume conventional

values for the future cost of decommissioning.

In other words, decommissioning

is assumed to cost a modest $50 million in 1980 dollars, and spent fuel disposal a modest $8 million annually, in 1980 dollars. With Three Mile Island decommissioning cost estimated to be an unknown amount in excess of $1 billion, and no operable programs or facilities for spent fuel, these conventional assumptions may become a serious legal question. In some ways, the litigation arising out of the uranium cartel, Three Mile Island, and West Valley may be relevant to

future controversy about responsibility for erroneous planning values in

decommissioning and spent fuel disposal.

4.

TAX INCENTIVES AND THE THREE MILE ISLAND ACCIDENT

Given the magnitude and timing of tax subsidies as they apply to nuclear power facilities, it is possible that the accident at Three Mile Island Unit 2

could have been influenced by these incentives10. The Rogovin study addresses

the problem in considerable detail.

It appears that General Public Utilities Corporation (GPU) received a $31 million tax benefit for 1978 by placing the plant in operation on December 30, 197811/. Current Pennsylvania and Federal income tax payments in the absence

of TMI-2 (Three Mile Island Unit 2) were estimated to be $15 million. Placing TMI-2 in operation on December 30 apparently allowed GPU to receive a net $16 million in 1978 rather than pay $15 million.

However, a $31 million tax benefit must be seen as rather insignificant for a corporation with GPU's annual revenues of $1.5 billion. Further, if operation was delayed from December 1978 to 1979, the tax benefit would not be lost, but simply transferred to 1979.

The Corporation also reported reserve capacity of 40% with TMI-2 and

25% without TMI-212/. Capacity availability, then, could not have been a

major motivation for operation of TMI-2.

In my judgment, it would not be logical for GPU management to encourage improper operation of TMI-2 in order to attain the tax benefit cited here. Rather, it seems to me that GPU operates in a normal tax and regulatory environment, and would not logically risk safe operations to attain very

There is not sufficient information publicly available to Judge whether