Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Appendix J A Method of Addressing Economic Irreversibility We discuss here an example of calculating the economic cost of long-term or irreversible environmental changes. The method is broadly applicable to the North Slope and beyond; the example of the Arctic National Wildlife Refuge was cho- sen because it is the only part of the region for which suit- able data are available. Chapters 7 and 9 describe physical and biotic effects of seismic exploration and their human effects. Here, the long-term economic costs are considered. From an economic perspective, damage done to the tundra by seismic explorations and road and pad building is basically economically irreversible. This means that roads, pads, and seismic tracks laid down today are visu- ally very evident for many decades later, as are other long- lived human interventions. Even if an effect does not last forever in a physical sense, it is an economic irreversibility in practice if it lasts for as long as, say, 50 years. The present value of a dollar every year forever differs by $.002 from the present value of a dollar every year for 50 years at the U.S. Geological Survey's (USGS) use of a 12% rate of interest when analyzing the value of oil in the Arctic Na- tional Wildlife Refuge. Irreversibility is an important ingredient for the accu- mulation of an effect because the decision to invest in explo- ration produces visual effects year after year even if no de- velopment follows. Most economic analysis assumes that one readily can undo what one has done. Thus the prospect of irreversibil- ity raises a major concern and requires analysis of the sig- nificance of the amount of development and whether de- velopment should be prohibited in areas of special value. The following example, calculated using the Arctic Na- tional Wildlife Refuge as the example, illustrates an ap- proach for thinking about irreversible accumulated effects. Data on environmental costs would have to be collected to reach a confident conclusion. The analysis does not include any noneconomic factors that usually also influence such . · . Decisions. 242 ALL OR NONE? One piece of relevant data for policy making is how much oil is economically, not technically, recoverable in an area. As an example, Figure J-1 illustrates the amount of oil recoverable from the refuge's 1002 Area as a function of pnce. A price less than $15 per barrel ($0.36 per gallon) makes oil development in the Arctic National Wildlife Ref- uge unprofitable. If the belief is that $25 per barrel ($0.60 per gallon) will prevail, then the gross value of oil in the refuge would be more than $130 billion if it were extracted and sold today. The cost per barrel in Figure J-1 represents discovery, development, production, and transport costs. The 40 - RESOURCE COSTSANWR ~ 002 AREA 6 BILLIONS OF BARRELS OF OIL in FIGURE J-1 Summary of the USGS estimates of economically recoverable oil that may occur beneath the federal 1002 Area of the Arctic National Wildlife Refuge. The three curves are based on estimates of economically recoverable oil volumes at the mean ex- pected value, and at the 95% (Fg5) and 5% (Fo5) probabilities. Each curve relates market price/cost to the volume of oil estimated to be profitably recoverable. Included are the costs of finding, develop- ing, producing, and transporting oil to market based on a 12% after- tax return on investment all calculated in constant 1996 dollars. SOURCE: Attanasi 1999.

APPENDIX J curve should be shifted up for other costs left out, particu- larly the expected costs due to oil spills, decommissioning, and other environmental damages that can be expressed on a per barrel basis. The relation of prices and quantities of oil in Figure I- 1 provides an estimate of net benefits from extracting oil in the Arctic National Wildlife Refuge as of today and depends on the prevailing pnce of oil. The figure portrays the incremen- tal or marginal costs of oil. From this, total cost can be calcu- lated. It is the area under the (mean) curve up to a particular quantity of oil. Total revenue is just the product of pnce and quantity, from which total cost is subtracted to yield a net benefit exclusive of environmental costs for any given ex- pected pnce. The expected future pnce of oil plays a vital role in the evaluation of a "go" or "no go" decision. Figure J-2 illus- trates the pnce of oil per barrel histoncally. The pnce of oil is not determined in the market because the oil market is not competitive now. Economists characterize the oil market as a monopoly with a competitive fnnge. The Organization of Petroleum Exporting Countnes (OPEC) is a collection of nations who seek to control pnce through collusion. When collusion is successful, the OPEC has driven the pnce of oil above $35 a barrel ($0.83 per gallon) in the past. However, it has not been possible to successfully collude for very long. The free market, competitive pnce is believed to be $10 a barrel ($0.24 per gallon), according to the oil minister for Kuwait (NYT, 11/16/2001, p. c2) The ability of the oil-producing countries to collude plays a key role in the potential development of an area. If the chance of persistent collusion is low, a resulting low market pnce of oil will preclude development. Growth of Venezuelan production Pennsylvania oil Boom 1 00 I ~ 901 80 7n 1 an 50 4n an 20 O 243 The second key factor in evaluating development is en- vironmental costs. Among potential environmental costs we focus here on the irreversible nature of the visual impact created by seismic trails, roads, and pads. These costs can be treated as the fixed social costs of oil development, which occur at the time of development. It will be assumed that people in general are not pleased with these seismic trails and the imprint of roads and pads, and that they would in principle be willing to pay to keep them from occurnng. One should think of this as an "as if' proposition. It is not that people actually would be asked to pay but rather people are asked to express displeasure in money terms. The thought experiment resembles what we do when we visit a restaurant and express expected gustatory pleasure in money terms by the choices made or not made. TWO SCENARIOS The crux question is whether the expected private net revenue from oil development, for which there are esti- mates, exceeds the social cost. Since no such costs have been estimated, a threshold analysis has to be made. Put another way, what is the least amount of money a represen- tative family in the U.S. must be willing to pay annually to make the losses from development exceed the benefits of development? Two scenarios are developed, one where the interest or discount rate is 12%, following the assumptions of the USGS in estimating the incremental costs of development. The other is a discount rate of 4%, which is more in line with what the economics profession would advocate for long- lived investment projects (Weitzman 2001~. Fears of - ? shortage in US product/on ~ ~ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 R i neg/ns Di f oil exports ~ Spindletop, begin ~ Texas 1 1 1 S money of the day ~ $2000 1861-1944 US average 1945-1985 Arabian Light posted at Has Tanura 1986-2000 Brent spot. ~~ ~~ mu. Post-war-' ~ Netback pricing reconstruction ~ ~ Yom Kippur introduced East Texas war Iraq invaded field - Suez ~ Iranian Kuwait discovered ~ ~ ~ crisis ~ , ~ ~ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1861-69 1870-79 1880-89 1890-99 1900-09 1910-19 1920-291930-39 1940-491950-591960-691970-791980-891990-2000 FIGURE J-2 Historical price of oil, per barrel. SOURCE: BP Exploration 2001b. Reprinted with permission, copyright 2001, BP.

244 Extraction is not an instantaneous decision. The land must be opened for leasing. It takes time to consummate actual discovery. Facilities have to be built. This process will take 7 to 12 or more years based on past experience. Here we assume that it takes 10 years before the first field actually produces.) Further we assume that it will take 40 years for the oil resources to be depleted.2 New fields will where come on line roughly in decreasing profitability. The extrac- tion period cannot be shortened very much because this re- duces the total amount of oil that can be extracted and re- duces overall profit to the owner (the public). Prudhoe Bay has been extracting for 24 years, and oil specialists expect there are more than 10 years of future productivity. The U.S. Department of Energy expects a total productive life of 65 years with a peak at 18 or 22 years depending on rates of development (DOE). This assumption plays an important role in the economic analysis below. To simplify the analy- sis, assume that all the oil is extracted at the midpoint of the extraction period 20 years and that value and costs are discounted back to the present, taken to be 10 years from now. A shorter extraction period entails less discounting and a higher profit from oil development. Discounting back to the actual present reduces the values by one-third. In the calculations that follow, an exponential function represents the incremental cost curve of the USGS (See Fig- ure J-1~. The estimated incremental or marginal private cost func- tions (MC) is3 MC = 0.61lQi725 + 15. Table J-1 is an auxiliary table illustrating the lump sum value of the environment necessary to equal the value of oil devel- opment in the Arctic National Wildlife Refuge for three al- ternative assumptions about the possible prices, conditioned on collusion occurring and two interest rates. TABLE J-1 Net Value (2011) of Oil Production in the Arctic National Wildlife Refuge, Present Value 2011 dollars (billions) Interest Rate = .12 Price/Barrel Probability of Collusion $20 1/4 ll2 314 Interest Rate = .04 Price/Barrel $25 $40 .24 .73 .49 1.4 .7 2.18 2.1 6.2 9.3 $20 $25 $40 1.20 3.59 15.28 2.40 7.9 30.56 3.61 10.78 45.84 ~ The basis for the calculation assumes no lags due to litigation by envi- ronmental interests and perhaps others. A lag increases the likelihood that oil development is not economically feasible. 2 President Bush recently spoke in terms of 47 years (NYT 5/18/01). The longer the depletion penod, the more distant the revenues, so the less valu- able the field for a given total volume. 3A C++ program was written to fit Figure J-1, and close to 600 data points were drawn. A software package, "Kaleida Graph," transformed the data points into the function. APPENDIX J The formula for calculating this table is the weighted prob- ability of expected net revenues in present value (2011) terms. ~ REP) ~ EV(P) = }:PQ(p) - | 0.611Q dQ15Qle (1FJ`) + P Fee A _ EV(P) expected net present value of oil production in the Arctic National Wildlife Refuge excluding envi- ronmental costs; 1- [I = the probability of successful collusion; _ = 0, because competition drives price below $15 per barrel (see text); QfP)= USGS estimate of oil volume when P = (P); and r = the interest rate. Explaining the equation in greater detail, when the price is high, P. with probability (l-n), the revenue is PEEP), where QfP) is computed from the equation for MC. When the price is low, P. with probability n there are no revenues because P is the competitive price and is too low to warrant extraction. So the expected revenue is 1) Exp Rev = P UP )~1 - H). The private cost = 0, when no oil is extracted, with prob- ability n The private cost, when there is extraction is the area under the marginal cost curve in Figure J-1, Q c-' o which occurs with probability 1-n. The expected net rev- enue is the difference between expected revenue and cost equals the difference between equations (1) and (2). The term, em, discounts the expected net revenue back to 2011 from 2031, where the latter date is the midpoint of the ex- traction period. The net present value (at a 12% interest rate) at the time extraction is assumed to begin, when the chance of collusion iS ~/2 and it achieves a price of $25 per barrel ($0.60 per gallon), is $1.45 billion. To match this development value requires a lump sum preservation value per family of about $14.50 since there are about 100 million families in the United States. This is less than one-half the value people were willing to pay to avoid another Exxon Valdez oil spill for 10 years, according to a study done for the state of Alaska. (See Carson et al. 1992.) To capture the idea that some environmental costs are economically irreversible, together with any other credible irreversible environmental cost, it is appropriate to express the threshold environmental values on an annual basis as in Table J-2. If there is a 50% chance of colluding and this market structure could achieve a price of $20 or $25 per

APPENDIX barrel ($0.48 or $0.60 per gallon), then a willingness-to-pay of about $0.96 or $2.87 annually per family in the U.S. would be necessary to match the value of oil development in the Arctic National Wildlife Refuge at a 4% rate of interest or $0.58 to $1.74 annually at the 12% rate of interest used by USGS. Expressed in present day values rather than 2011 values for the 50% chance of $25 per barrel ($0.60 per gal- lon), the threshold values per family are about $1.89 or $1.15 annually respectively for interest rates of 4% or 12% (Table J-3~. TABLE J-2 Per Family Net Value of Oil Production Arctic National Wildlife Refuge (2011 dollars) r = 0.12 and t = 20 245 TABLE J-3 Annual Family Willingness to Pay (2011 dollars) r = 0.12 and t = 20 Probability of Collusion Price $20 $25 $40 1/4 ll2 314 r = 0.04 and t = 20 0.29 0.58 0.87 0.87 1.74 2.61 3.70 7.40 11.11 Probability of Collusion Price $20 $25 $40 1/4 0.48 1,44 6.11 Probability Price ll2 0.96 2.88 12.22 of Collusion $20 $25 $40 1.44 4.31 18.34 1/4 ll2 314 r = 0.04 and t = 20 2.43 4.86 7.28 7.26 14.51 21.77 30.85 61.70 92.55 Probability of Collusion Price $20 $40 1/4 ll2 314 12.02 24.05 36.07 35.94 71.87 107.81 152.79 305.59 458.38 If the extraction period were shorter, the threshold will- ingness to pay would increase, as would be the case if more oil was economically available. A lower threshold willing- ness-to-pay would follow if seismic trails were laid down before extraction began and to the extent that there are other expected environmental costs associated with oil extraction not included in these assumed costs.