OIL AND NATIONAL SECURITY: AN INTEGRATED PROGRAM FOR SURVIVING AN OIL CRISIS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP83B00140R000100010031-0
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 29
• Document Creation Date: December 20, 2016
• Document Release Date: April 11, 2007
• Sequence Number: 31
• Publication Date: December 1, 1981
• Content Type: REPORT

This article summarizes much of the
work that I have done in the last
few years on oil interruptions. It
may be of interest to you in connec-
tion with your work on that c,Ihicr+
Henry S. Rowen, C/NIC
IROOM NO. BUILDING
7E62 CIA Hqs.
FORM NO 241 REPLACES FORM 36-8
WHICH MAY BE USED.
 ?
?
Ann. Rev. Energy. 1981. 6.171-98
Copyright ? 1981 by Annual Reviews Ina All rights reserved
OIL AND NATIONAL
SECURITY: An Integrated Program
for Surviving an Oil Crisis
Henry S. Rowen
Graduate School of Business, Stanford University, Stanford, California 94305
John P. Weyant
Department of Operations Research, Stanford University, Stanford,
California 94305
INTRODUCTION
The dependence of the United States and its principal allies on oil from the
Persian Gulf has brought with it a number of increasingly obvious and
serious problems. Assessments of the extent of these problems and the
merits of potential solutions have become critical national security con-
cerns. For the past two years a group of researchers from across the nation,
organized by Pan Heuristics, has studied the economic, political, and mili-
tary dimensions of problems accompanying dependence on oil from the
Persian Gulf.
Pan Heuristics issued its first comprehensive Report on Persian Gulf Oil
and Western Security to the Department of Energy on November 4, 1980
(1). In that report political trends in the Persian Gulf area are reviewed by
Khalilzad & Samore (2, 3); Wohlstetter and Brody describe the resulting
threat to the West and potential political-military responses (4, 5); Henry
Rowen & John Weyant attempt to quantify the economic effects of oil
supply interruptions (6); and a number of energy policy responses to the
economic problems caused by dependence on oil imports and by the threat
of oil supply interruptions are explored by Rowen, Weyant, Missner, Mc-
Donald, Pittinger, Kline, Hogan, Nye, Deese, Beverly Rowen, and Gregory
 Approved For Release 2007/04/11: CIA-RDP83B00140R00010001
? ?
Jones (7-18). Henry Rowen, as principal investigator, was responsible for
developing a succinct overview of the results of the project (19). The eco-
nomic dependence of the West on Persian Gulf oil also underlies most of
the military and political problems stemming from Persian Gulf depen-
dence. Thus, this review focuses primarily on the economic problems and
appropriate energy policy responses.
Two kinds of energy policies were considered as potential remedies for
the economic ills resulting from Persian Gulf oil dependence. First, policies
designed to directly decrease our dependence on imported oil (14, 16),
which can reduce the price of world oil under normal conditions and the
cost of oil supply interruptions when they occur, but will take some time
(probably 3-5 years) to have significant effects. Second, policies designed to
reduce the cost of oil supply interruptions once they occur. These emer-
gency preparedness measures (e.g. stockpiling oil) involve some costs and
generate no benefits under normal conditions, but could be valuable should
an oil supply interruption actually occur. Given the current state of the
world oil market and the long lead times required to increase the military
protection of the Gulf or to reduce the level of oil imports significantly, this
second type of energy policy, which could provide significant insurance
against oil supply interruptions within 1-3 years, has received far too little
attention. Thus, the present paper reviews various proposals for increasing
the supply of-or decreasing the demand for-oil during a crisis within the
framework established in the Pan Heuristics report (1).
This review cites and attempts to tie together insights from several related
studies of individual vulnerability-reducing policies. However, the simpli-
fied economic framework adopted here focuses on the domestic and interna-
tional interactions of the proposed emergency policies, and is, therefore, less
complete in its representation of the effects of any one policy on any particu-
lar country. Consequently, there is a natural complementarity between the
strategic interactions and policies identified here and the more operational
recommendations that flow from the more detailed studies that are cited.
One recent study of particular significance in this regard is the November
1.0, 1980 report on Reducing U.S. Oil Vulnerability from the Assistant
Secretary for Policy and Evaluation to the Secretary of Energy (20). The
interactions between the Department of Energy staff that prepared that
reports and our own were frequent and fruitful. In addition, our study team
overlapped with ones organized by Harvard University and the Electric
Power Research Institute (EPRI) to study similar issues. The Harvard
study (21) paralleled ours, but focused more on political factors and less on
'In particular, Lucian Pugliaresi, Thomas Neville, Roger Naill, John Stanley-Miller, Jerry
Blankenship, Michael Baron, Glenn Sweetnam, and Joe Eschbach of Assistant Secretary
William Lewis' staff.
 Approved For Release 2007/04/11: CIA-RDP83B00140R000100010031-0
SURVIVING AN OIL CRISIS 173
upwards of a year or more, would have a major depressing effect on the
world economy given the present state of preparation of the oil importing
countries. To take one case, a reduction of 9 X 106 bbl/d (millions of barrels
per day) for one year, about half the oil now exported from the Persian Gulf,
would have several economically destabilizing effects: the price of oil would
shoot up to over $100 per barrel, oil would be released from government
stockpiles, and efforts to conserve oil and to substitute other fuels and
capital and labor would be undertaken. Neglecting losses from unemploy-
ment and additional losses from government administrative controls, the
result would be an estimated loss on the order of 5% of GNP for the United
States, 7% for Western Europe, 8% for Japan, and 8% for the oil import-
ing, less developed countries (LDCs).
Recently, a number of actions that might greatly reduce the extent of
these economic losses have been proposed. Two types of measures have been
considered: those that increase supply and those that could limit demand
in a crisis. On the demand side, emergency tariffs have served as a proto-
type; on the supply side, the strategic petroleum reserve has commanded
almost all of the attention. In this paper, a number of additional supply-side
measures are identified and evaluated in concert with both oil stockpiles and
emergency tariffs.
The combined potential of the incremental supply possibilities for the
United States in a crisis comes to the equivalent of around 2 X 106 bbl/d
average for one year. The major components of this addition are: stored
natural gas, fuel switching to coal by electric utilities and industry, in-
creased oil and gas production, and more intensive operation of nuclear
power plants.
Outside of the United States, the principal extra non-Persian Gulf sup-
plies for an emergency are: increasing oil production to capacity levels
(assumed to add a 106 bbl/d), storage of natural gas, and more intensive use
of coal and nuclear plants (which is likely to require substantial coal stock-
piling). This potential adds up to a total of 2.5 X 106 bbl/d for one year.
It is assumed additional supplies beyond these levels would have to come
from prebuilt building oil stocks.
It appears that there is sufficient flexibility within the US energy system
to permit fuel substitutions on this scale; the uncertainties on this score
outside of the United States are a good deal larger.
economics, while the EPRI study (22) started when ours was nearing
completion, but promised to strengthen and expand upon the analytic
foundations of much of the previous work in this important area.
SUMMARY
A major interruption of Persian Gulf oil supplies for an extended period,
 Approved For Release 2007/04/11 : CIA-RDP83B00
174 ROWEN & WEYANT
The benefits of oil or oil-equivalent stockpiles, given an interruption say
of 9 X 106 bbl/d for one year, are very large. A billion barrel stockpile would
have an estimated value of $46 billion during a 9 X 106 bbl/d interuption;
with an equivalent stock held by all of the OECD countries, the saving to
the United States would be $92 billion. However, with the added supply
measures, the annual cost of gaining these benefits could be much reduced;
for instance, to around $1 billion annually for the equivalent of a 500 X
106 bbl oil stock versus $2.3 billion for an all-oil stock.
These benefits and costs can be expressed in terms of the breakeven
probability of a Persian Gulf oil disruption of given size in order to justify
holding given stockpile levels. An estimated annual probablility of a 9 X
106 bbl/d disruption of only 0.06 is sufficient to warrant a 1,000 X 106 bbl
combined gas plus oil stock plus coal switching strategy; for an equivalent
OECD-wide policy (as viewed from the perspective of the US economic
saving) the annual breakeven probability is 0.05. (By comparison with a
strategy of buying only oil stocks, these breakeven probabilities are reduced
by over one half at the 500 X 106 bbl-equivalent stockpile level and by
20-25% at the 1,500 X 106 bbl-equivalent stockpile level).
The combination of these supply-side measures with demand ones further
reduces the estimated economic loss. For the 9 X 106 bbl/d interruption,
an OECD-wide emergency tariff plus these supply measures only in the
United States (at the 1,000 X 106 bbl level) would cut US economic losses
in half. If such a tariff is combined with an OECD mixed fuel stockpile,
equivalent to 3 billion barrels of oil, economic losses in the noncommunist
world for this scenario would be practically eliminated. Even for a full
Persian Gulf closure for one year, on these supply and demand assumptions,
losses would be cut in half. The annual outlays by the United States in that
case would be around $4.6 billion.
BASELINE PROJECTIONS FOR 1990
Several baseline projections (6) set the stage for the evaluation of alternative
measures to reduce the vulnerability of the West to interruptions of its crude
oil supply. Projections of the uninterrupted world market and output of the
world's economies underly the calculation of the economic losses attribut-
able to several baseline oil supply interruption scenarios.
The rapidly escalating prices that have characterized the world oil maket
in recent months will reduce the world's demand for crude oil and increase
crude oil supplies from outside of the Persian Gulf during the 1980s relative
to the levels that would have resulted from previously prevailing price
trends. But the growth of the world's economies will exert an upward pull
on world oil demand that could counterbalance the downward push of the
higher prices.
 Approved For Release 2007/04/11: CIA-RDP83B00140R000100010031-0
Kline & Weyant (23), using a simple model of OECD oil demand incor-
porating both price and income effects, concluded that in the absence of new
policy initiatives, if crude oil prices grow at one percent per year in real
terms during the 1980s (starting from $32.00 per barrel in 1980), OECD
crude oil demand in 1990 will equal or be only slightly less than the current
level. That projection is consistent with a number of recent projections
made by other analysts. Furthermore, depletion effects are likely to counter-
balance much of the stimulating effect high prices will have on crude oil
supplies from outside of the Persian Gulf during the 1980s; 2 the least cost
(not price) oil in the world will continue to be that from the Persian Gulf.
Thus, the baseline projection of world crude flows in 1990 (Table 1) is not
much different from today's. LDC oil demand grows a bit, but is matched
by a modest increase in non-Persian Gulf supply, primarily from Mexico.
This result is consistent with results from ten models of the world oil market
included in the Energy Modeling Forum's study on world oil (25). So, in
the absence of new policy initiatives the OECD countries are about as
dependent on Persian Gulf oil in 1990 as they are today.
To utilize the simple analytical framework to calculate the economic
impacts of cutbacks of the supply of oil from the Persian Gulf, several
benchmarking assumptions are required for a future year of interest. The
year 1990 was chosen as a representative year for the economic loss calcula-
tions, and Table 2 shows the benchmarking assumptions.
Economic growth is assumed to be 5% per year in Japan and the oil
importing LDCs during the 1980s, and 3% per year in the rest of the
OECD. This results in a 3.4% economic growth rate for the OECD and
3.6% for importing nations outside communist areas (WOCA). Economic
impacts are generally reported for three key regions of the OECD: 1. the
United States, 2. Japan, and 3. other (primarily Western Europe); and the
oil importing LDCs. GNP in 1990 is assumed to be $3,696 billion 1980 US
dollars in the United States, $1,710 billion in Japan, $3,293 billion in
Western Europe and $1,873 in the oil importing LDCs.
The market price of world crude is assumed to grow at one percent per
year from $32 per barrel in 1980 US dollars in 1980, to $35 per barrel in
1980 dollars by 1990. (This, of course, assumes no further major, sustained
production declines.) Emergency oil stockpiles in 1990 are assumed to be
335 million barrels in the United States, 285 million barrels in Japan, and
432 million barrels in Europe. These stockpile figures are not predictions
of levels for the future; they simply enable testing the consequences of
today's posture continuing into the future.
'See (24) and its extensive list of references for studies of US oil production supporting this
projection.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
is ?
Supply (106 bbl/d)
Oil
Other oil
OECD
importing
LDCs
Persian
Gulf
exporting
LDCs
Total
consumption
OECD
14.0
0
15.5
8.5
38
United States
(9)
0
(2)
(6)
(17.0)
Japan
(0)
0
(4.5)
(1)
(5.5)
Western Europeb
(5.0)
0
(9)
(1.5)
(15.5)
Oil importing LDCs
0
3
2.5
4
9.5
Persian Gulf
0
0
3
0
3
Other exporting LDCsc
0
0
0
3
3
awe assume by 1990 no net imports or exports of oil from the Centrally Planned Econ-
omies.
bShorthand for OECD less the United States and Japan; i.e. Western Europe, Canada,
etc.
c E.g. other OPEC countries, Mexico, Egypt, Trinidad, Tobago, etc.
Additionally, it is assumed that the one-year price elasticity for crude oil
demand is 0.08 in the United States and the oil importing LDCs, 0.07 in
Western Europe, and 0.06 in Japan. These estimates were based on experi-
ments with the Kline/Weyant OECD oil demand model (23). These values
seem roughly consistent with the empirical evidence on the short- and
Oil
Emergency
GNP (billions
1980 dollars)
consumption
(106 bbl/d)
Oil imports
(106 bbl/d)
stocks
(106 bbl)
OECD
$6,250
$ 8,700
38.0
38.0
24.5
24.0
1,325
1,072
United States
$2,750
$ 3,696
17.0
17.0
8.0
8.0
488
355a
Japan
$1,050
$ 1,710
5.5
5.5
5.5
5.5
285
385b
Western Europe
$2,450
$ 3,293
15.5
15.5
11.0
10.5
552
432c
Oil importing LDCs
$1,150
$ 1,872
7.0
9.5
4.0
6.5
0
0
Total WOCA
$7,400
$10,572
45.0
47.5
28.5
30.5
1,325
1,072
a 100 X 106 bbl of public stocks plus 15 days of consumption in private stocks, in excess
of 40 days of consumption (680 X 106 bbl) assumed necessary for working inventories.
bStocks in excess of 30 days of consumption (164 X 106) assumed necessary for work-
ing inventories.
c200 X 106 bbl of government mandated stocks + 15 days of consumption in private
stocks in excess of 35 days of consumption (543 X 106 bbl) assumed necessary for work-
ing inventories.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
long-run demand elasticities for oil and aggregate energy.3 The elasticity is
smaller in Japan and Europe because the availability of alternative fuels
(e.g. natural gas and coal) is less and because of the larger divergence
between the price of crude and the price of oil products to consumers there
than in the United States. This latter condition can be attributed to the very
large oil product taxes in Japan and Europe and implies a lower crude oil
elasticity than in the United States if the elasticity of demand for oil pro-
` ducers is approximately equal in the two countries.
Higher prices would, of course, stimulate some additional production by
non-Gulf producers, some of which might come from small amounts of
excess capacity likely to exist at any given time. Unfortunately, most of the
world's spare production capacity is in the Persian Gulf area and, depend-
ing on the scenario, may be unavailable. The lead times to bring on new
production capacity are long, so that only modest additional amounts of oil
(or its close substitute, natural gas) can be expected in a crisis lasting one
or two years. An additional possible constraint on added production is the
willingness of oil exporters to increase output despite a much higher oil
price. Some, for domestic or international political reasons, might be unwil-
ling to do so.
The following assumptions about the potential for increased oil supplies
from outside the Persian Gulf during a crisis follow from analysis described
in (12). It is assumed that the non-Persian Gulf exporters would increase
production by 0.5 X 106 bbl/d during a 9-million barrel per day interruption
and by 1 X 106 bbl/d during an 18-million barrel per day interruption. The
increased production would come mainly from small increases in capacity
utilization in Libya and Nigeria and some increases in Mexican capacity
and output. In the United States, the incremental supply is responsive to
price.4 For example, 0.3 X 106 bbl/d of additional production of oil or close
substitutes is assumed to be available at a price of $100 per barrel, and 0.5
X 106 bbl/d at $200 per barrel. Of course, the 70% windfall profits tax in
the United States would retard this response considerably; a world price
increase of $215 per barrel would be required for producers to actually see
an effective price of $100 per barrel, and price controls on natural gas would
limit incentives to add to its supply. There is assumed to be no supply
response in Japan or in Western Europe; there is no oil production to speak
of in Japan, and the major portion of the Western European production is
from the North Sea, where difficult conditions probably result in lead times
of greater than one year for incremental production.
3See, for example, Energy Modeling Forum (26) and its extensive list of references.
`The short-run percentage increase in oil production in response to a one-percent increase
in oil price is assumed to be .03. This parameter is known as the short-run oil supply elasticity.
 Approved For Release 2007/04/11: CIA-RDP83B00140R00010001003-f--C
178 ROWEN & WEYANT
IDEALIZED INTERRUPTION ECONOMICS
Given the baseline projections, several simplifying assumptions (6) allow
relatively straightforward calculation of the economic costs of oil supply
interruptions. Most importantly we assume that, during the interruption,
compensating monetary and fiscal policies designed to keep the nation's
capital stock and labor force fully employed would be found and imple-
mented and that there would be no inflationary costs of the interruption
(e.g. 6-8). The exact specification of these policies is left to the macroecono-
mists (e.g 27-29). Reductions in payroll taxes and more liberal investment
tax credits would help a great deal, but other policies might work as well
or better.
It is also assumed that emergency energy policies fail to reduce the
demand for imported oil before the world oil price increases to clear the
market. Once the market clears, however, it is assumed that there is no
attempt-either internationally or within the major oil importing nations
-to reallocate supplies at the market clearing price.
The assumptions of full employment of capital and labor and no after-
the-fact oil allocation rules are extremely optimistic. There is no guarantee
that the required set of compensating macroeconomic policies exists, and
if it does, that it will be found or implemented. Additionally, it is almost
certain that equity considerations will lead to some non-price induced oil
supply allocations during an interruption, especially at very high prices.
Several recent studies (27-40) have attempted to consider macroeco-
nomic mechanisms (e.g. downwardly inflexible real wages, downwardly
inflexible money wages, price rigidities, etc) that can produce these unem-
ployment, financial market, and regulatory effects. However, there remains
considerable disagreement over exactly how important each of these mecha-
nisms will be during an oil shortage. Furthermore, the results of many of
these studies are sensitive to the macroeconomic policy responses assumed:
For example, the effects of small changes in the money supply can some-
times outweigh the unemployment, financial market, and regulatory effects
of an oil price increase. Finally, detailed macroeconomic studies of the
United States cannot be easily integrated with comparable representations
of the rest of the world's economies within the context of the world oil
market. Thus unemployment, financial market, and regulatory effects are
not considered here. However, one common result from the detailed macro
studies-that the unemployment, financial market, and regulatory effects
are always positive and could be larger than the full employment effects-
is factored into the interpretations of the results from the idealized interrup-
tion economics framework.
Although the optimistic assumptions adopted here lead to an understate-
ment of the value of measures designed to cope with the adverse economic
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
effects of oil supply interruptions, they allow for a much simplified descrip-
tion of the economics of an interruption.
A Simple Analytical Framework
This idealized view of the economics of an oil supply disruption abstracts
from the complexities of the real-world response, but sets the stage for the
use of a simple analytical framework to evaluate the relationships between
oil availability, oil price, and economic output. But to utilize that frame-
work, several additional simplifying aggregation assumptions must be
made.
AGGREGATION ASSUMPTIONS These estimates use a simple concep-
tual framework for studying energy-economic interactions.5 Although that
framework is specialized to oil, disaggregated by world region and linked
to various other simple analytic models in the present paper, the description
of the effects of an energy tax or an energy cost increase on the economic
output of an economy is solely due to Sweeney. In addition to the idealized
policy response assumptions described previously, several additional as-
sumptions underlie the implementation of this model.
First, it is assumed that actors in the economy under consideration seek
to maximize economic output within the technical limitations of an aggre-
gate production function. That production function represents aggregate
economic output as depending on the economy's aggregate inputs of crude
oil, capital, and labor. Additionally, it is assumed that the supplies of capital
and labor are fixed and that the relationship between the supply of domestic
oil and its cost does not change, assumptions that are more plausible in
short-run analyses than in long-run studies.
Finally, it is assumed that the economy's demand for crude oil depends
solely on its price and aggregate economic output. Higher economic output
results in higher oil demand if the price of oil remains constant. But if the
oil price increases, oil demand decreases, with a constant percentage de-
crease in oil demand (known as the "price elasticity" of oil demand) result-
ing from each percentage increase in price if economic output remains
constant. Of course, during an oil interruption the higher oil prices that
result decrease oil demand and economic output, with the decrease in
economic output leading to further decreases in oil demand.
'The authors thank Professor Sweeney for his model (41), his insights into its operation, and
for making his original computer code available to them. This model is closely related to the
seminal energy-economy model of Hogan-Manne (42) but focuses more directly on the differ-
ence between the effects of domestic energy tax increases and increases in the price of imported
oil.
 Approved For Release 2007/04/11 : CIA-R -
i S
180 ROWEN & WEYANT
MODEL DESCRIPTION Given these assumptions, the key relationship of
the Sweeney model is that, for small changes in the price of imported oil,
the change in the Gross National Product of a particular country is equal
to that price change times the level of oil imports. Mathematically, the total
differential for the change in Gross National Product (G) is dG =
-EIdP1 where El is the level of oil imports. The total decrease in GNP for
large price changes is obtained by accumulating the GNP losses for each
small price increment. Of course, as the price of oil increases, the demand
for oil imports-the difference between oil consumption and domestic oil
production-decreases; oil demand decreases in accordance with the aggre-
gate demand relationship in the model, and domestic oil production is
assumed to increase modestly in response to the higher prices prevailing
during the interruption. A constant elasticity of supply -percentage in-
crease in supply in response to a one percent increase in price-is assumed.
Thus, to calculate the resulting decrease in Gross National Product, the
level of oil imports must be updated for each small change in the price of
oil imports.
This model is the essence of simplicity. Given preinterruption levels of
total economic output, oil consumption, and domestic oil production for
any economy, and the precut price of world oil, the decrease in economic
output attributable to any increase in the world price of oil is calculated.
In this paper, the oil importing nations are grouped into four categories
to permit consistent calculation of the aggregate economic effects of an oil
supply interruption on the world and to isolate the effects on key oil import-
ers. The United States and Japan are considered individually, with all other
industrialized western economies included in an aggregate called "Western
Europe" and the oil importing less developed countries referred to simply
as "Importing LDCs."
A simple oil-economy model is calibrated to each of the four oil import-
ing regions. The decrease in economic output in each region is calculated
by increasing the world oil price by small increments and calculating the
economic loss for each increment. This process is continued and the eco-
nomic losses accumulated for each region until the sum of the imports of
the four regions is decreased to the level of available supplies (6).
ECONOMIC LOSSES RESULTING FROM OIL
SUPPLY INTERRUPTIONS
Three levels of Persian Gulf oil export interruption are examined: 3, 9, and
18 X 106 bbl/d. Each interruption is assumed to last for one year. Addition-
ally, it is assumed that half of the available emergency stocks are released
during the interruption and that the release rate during that time is con-
stant. Thus, the release rate in Japan is 285 - 2 _ 365 = 0.39 X 106 bbl/d.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
Given the benchmarking assumptions, the simple analytical framework
can be used to project the economic losses resulting from the baseline oil
supply interruptions. Figure 1 shows the economic loss projections by level
of cut and importing region (6). The comparison of projected losses by
importing region and level of cut identifies important relationships underly-
ing the computations.
The effects of a 3-million barrel per day interruption for a year are, as
expected, modest: they range from one to two percent of GNP across the
four regions. This corresponds approximately to the cut experienced re-
cently in Iran and is somewhat greater than that experienced in 1973-74.
However, the policy response in the earlier period, at least in the United
States, seems to have contributed to a greater GNP loss than this estimate.
For a 9-million barrel per day interruption, the damage is much greater;
about a 5% GNP loss for the United States, 7% for Western Europe, 8%
for Japan, and 8% in the Importing LDC's. Recall that these estimates
assume, optimistically, a smooth economic adjustment to the sharp increase
in oil prices.
For a full Persian Gulf interruption, 18 X 106 bbl/d for a year, the losses
are about 13% for the United States ($465 billion), 22% for Europe, and
25% for Japan and the LDCs.
Nonlinearity with Respect to the Depth of Cut
Perhaps the most striking feature of Figure 1 is the increasing rate of loss
with respect to level of the interruption in each region. The losses for the
18-million barrel per day interruption are much greater than twice those for
the 9-million barrel per day interruption, and the losses for the 9-million
0 3x106BD INTERRUPTION
EJ 9x106BD INTERRUPTION
m 18x106BD INTERRUPTION $725
WESTERN
EUROPE
IMPORTING
LDC's
Figure 1 Percent GNP losses resulting from Persian Gulf oil supply interruptions during
1990. Numbers on top of bars are absolute GNP losses in billions of 1980 dollars.
 Approved For Release 2007/04/11: CIA-RDP83B00140R000100010031-0
barrel per day cut are much greater than three times those for the 3-million
barrel per day cut. This nonlinearity is a result of the substitution assump-
tions embedded in the models. These assumptions are drawn from the
observation that in any economy that seeks to maximize output as the level
of oil input is reduced, the least cost substitution opportunities will be
exploited first; subsequent reductions are more costly. The increasing diffi-
culty of substitution also implies that a greater than proportional increase
in world oil price is required to clear the market in response to each
successive increment to the size of the interruption. A 20% reduction in oil
availability (9-million barrel per day Persian Gulf cut) leads to a threefold
increase in the price of world oil and a 7% decrease in WOCA GNP,
whereas a 40% reduction in oil availability (18-million barrel per day
Persian Gulf cutoff) leads to a ninefold increase in the price of world oil and
a 20% decrease in GNP.
Intercountry Comparisons
The economic losses vary significantly from one importing region to an-
other, with the losses the least in the United States and the greatest in the
oil importing LDCs. For a 9-million barrel per day interruption in Persian
Gulf oil supplies for one year, the loss of economic output is 5% in the
United States, 7% in Western Europe, 8% in Japan, and 8% in the oil
importing LDCs. The smaller losses in the United States are attributable
to its lower precut level of dependence and its greater potential for oil
import substitution. The precut value of imported oil is 2.8% of the US
economy, 4.1% of the Western European economy, 4.1% of the Japanese
economy, and 4.4% of the LDC economies. Additionally, the relatively
higher short-run crude oil demand elasticity in the United States and a
much larger proportion of oil produced domestically results in a much
higher short-run import elasticity than in Japan. Although the short-run oil
demand elasticity assumption in the United States is 25% higher than in
Japan, its import elasticity is a factor of three larger. This shows up quite
clearly in a comparison of the losses for the 9-million barrel per day inter-
ruption case.
In response to the $113 per barrel oil price that results in that case, the
United States is able to reduce oil imports by 36%, while only a 21%
decrease can be managed in Japan. And, in fact, this leads to such a large
income loss in Japan that oil consumption is actually reduced by a larger
percentage there than in the United States.
The lower preinterruption dependence on oil imports and greater poten-
tial for oil import substitution in the United States might have been ex-
pected to lead to even smaller losses relative to those in the other regions.
Two factors explain why the US losses are, in fact, not much smaller than
those in other nations. First, the market for oil is worldwide; lack of
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
substitution potential in any region leads to higher world oil prices for all.
In the case of a 9-million barrel per day interruption, US oil imports drop
by 2.25 X 106 bbl/d to 5.75 X 106 bbl/d. But an increase in the price of world
oil from $35 per barrel to $113 per barrel accompanies that decrease in US
imports. Thus the interruption increases the US import bill from $102
billion to $237 billion, or by nearly 4% of its precut level of GNP.
Differences in precut stock levels provide the other explanation for the
smaller than expected interregional loss differences. During a 9-million
barrel per day cut, Japan is assumed to utilize preexisting oil stocks at a rate
of 0.38 X 106 bbl/d, or 7% of its preinterruption oil consumption/import
level. This oil, which was purchased for $30 per barrel, is now worth over
$110 per barrel. Thus, Japan saves $11 billion, or nearly one percent of its
precut level of GNP by virtue of its oil stocks. The relative size and, hence,
value of the US stockpile is, given current policies, considerably less.
What Can be Done?
The view is widely held that little can be done, aside from stockpiling oil,
to increase available energy supplies during an oil supply crisis. Unlike the
1950s and much of the 1960s, there is little shut-in production capacity in
the United States, so that the country has been unable to buffer cuts abroad
by increasing its oil production. The present level of government and gov-
ernment-mandated oil stocks, around 600 million barrels in the OECD
countries, over minimum working inventories, although very valuable, is
inadequate for a deep, long interruption.
Existing energy assets, and others that could be created, could make a
very large contribution to reducing the economic impact of a major supply
disruption. In considering these possibilities it is important to, keep in mind
that the price of oil might climb in a deep crisis to a very high price, well
over $100 per barrel. Supply and demand responses (small in the crises
experienced so far) in which the price of oil moved from $2 to $10 per barrel
and then from $14 to $30 per barrel (in current dollars), are likely to be
much more vigorous at the very much higher prices that could occur.
EMERGENCY SUPPLY MEASURES: THE
POTENTIAL FOR IMPORT SUBSTITUTION
The first column of Table 3 lists the additional supply possibilities in terms
of their separate potential contributions in replacing imported oil (in the
early 1980s) for the United States only (9-12, 15, 43). It shows the assumed
average availability of these replacement fuels during the first year of a
major oil crisis.
One set of potential supply increase options outside of the United States
is listed the second column of Table 3 (9-12, 15). The total additional supply
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
identified comes to around 4-5 X 106 bbl/d during the first year of an
emergency in the world outside communist areas. This quantity is equiva-
lent to about 25% of the 1979 level of exports from the Persian Gulf. (In
the base case presented earlier, the availability of I X 106 bbl/d of this
enhanced supply was included in the economic loss estimates for the large
interruption cases.)
Several caveats apply to this aggregate estimate. First, there is a substan-
tial range of uncertainty around each of these values (see 9-12, 15); the
values shown in Table 3 assume an all-out effort and may be too optimistic.
Second, there is the possibility of double counting the potential saving of
oil. The combined substitution possibilities for replacing oil are likely to be
less than the sum of the independently estimated savings. Third, there is a
question about the net use of residual oil if these measures are capable of
saving more heavy oil than is "needed" in a crisis of a given size. This
"excess" saved oil would be highly useful in those end uses that have little
flexibility in the choice of fuels, e.g. transportation. The question is, to what
extent does the refining system have the flexibility to alter its mix of prod-
ucts to bring about this shift?6 Fourth, some supply options have been
omitted from this analysis; for instance, the possibility of replacing oil by
enhanced natural gas production, or stored natural gas, outside of the
United States.
In addition to the supply contribution from these sources, additional
amounts of oil can be stockpiled in either public or private hands. (For the
purposes of the analysis in this paper, no distinction is made between these
two types of ownership.) The analysis here assumes that the United States
and other countries have the choice of progressively adopting damage-
limiting measures beginning with the least costly ones and proceeding to the
United States
(106 bbl/d)
Outside
United States
(106 bbl/d)
Increased oil production
0.10
1.00
Increased gas production
0.30
0
Additional stored gas
1.00
0.30
Nuclear electric
0.30
0.20
Coal
0.70
1.00
Utilities (0.6)
Industry (0.1)
6Examination of the characteristics of the refinery sector suggests that it probably has
enough flexibility (44).
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
SURVIVING AN OIL CRISIS 185
most costly. Up to various thresholds, it is less costly to plan on fuel
switching to coal, storing natural gas, and increasing production of natural
gas and oil than it is to store additional oil. Then, as additional protection
is sought, more oil is bought and stored, up to the level where the marginal
cost of storing additional oil equals the expected marginal reduction in
economic loss at some breakeven, probability of oil supply disruption.
There is no attempt to determine the optimum combination of fuel
switching, gas storage, oil storage, etc. Instead, this paper tries to show the
value of a larger array of damage-limiting measures than has been consid-
ered so far and that at least one set is potentially superior to building a large
Strategic Petroleum Reserve.
After examining the possibilities for "oversaving" or double counting of
oil replacement measures, we use an estimate of an aggregate increased
supply potential in the United States of 2 X 106 bbl/d equivalent through
these measures and 2.5 X 106 bbl/d equivalent in the non-US OECD.
The Value of Stockpiles and a US Emergency Fuel
Switching Program
During an emergency, it is estimated that about 2.0 X 106 bbl/d of US oil
use in the utility, industrial, commercial, and residential sectors could be
switched to coal, natural gas, or nuclear sources by exploiting dual boiler
capabilities, increasing power plant capacity factors, and using gas appli-
ances more intensively. Of this total, about 1 X 106 bbl/d would come from
switching to fuels whose supply, it is assumed, could be sustained through-
out the crisis: coal, nuclear, and increased oil and gas production.? Some
of this response would occur through market forces, but utility and indus-
trial oil use switching is highly constrained by energy and environmental
regulations. Unless a program to relax these regulations in an emergency
is worked out in advance of a crisis, regulatory delays during the crisis could
prevent technically easy and low cost opportunities (certainly relative to
$100 per barrel of oil) from being implemented. Even with such a program
in place it is assumed that the amount of switching would depend on the
price of world oil, with the full 1-million barrel per day substitution
achieved only when the world oil price reaches $100 per barrel.
? The rest of the emergency supply would have to come from stocks.
Adding the assumed natural gas stockpile to the current Strategic Pe-
troleum Reserve level of about 100 X 106 bbl gives an oil-equivalent stock-
''On these assumptions, about 70 million tons of additional coal would be consumed in the
first year, a level that could be met through increased production and depletion of existing
stocks. This assumes that a strike has not depleted coal stocks in the period before the crisis
and that there is no coal strike during it. For purposes of estimating economic impacts for coal
stocks, the authors do not adopt the assumption used consistently for stocks of oil and gas that
only one half would be consumed in a year.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
pile level of 500 X 106 bbl/d, the lowest level examined; other levels
examined are 1000 X 106 bbl/d and 1500 X 106 bbl/d. It is assumed that
additional US stocks beyond 500 X 106 bbl/d of oil equivalents would be
composed entirely of oil.
There have been a number of excellent studies of the optimal size, fill, and
drawdown rates for the US Strategic Petroleum Reserve (45-55). The focus
here is primarily on the size for the reserve. In order to maintain an
interpolicy and international perspective, simplifying assumptions are made
about the fill and drawdown rates. It is assumed that the Strategic Pe-
troleum Reserve is filled at a constant rate over a five-year period, and that
precisely half of it will be used during a one-year oil supply interruption.
Of course, the amount that is actually released will depend upon what is
expected to follow; the exact depth and duration of an oil supply interrup-
tion will not be known in advance. Several recent innovative stockpile
studies have explicitly incorporated these expectations within a dynamic
modeling framework (45-47). In addition, some stockpile studies are for-
mulated as games between the oil exporting and importing countries (48),
including the potential sizable oil stockpiles to deter supply cutoffs. Such
studies are interesting, but pertain primarily to the 1973-74 embargo type
of interruption, not to those resulting from internal revolutions (e.g. Iran
in 1979), interregion conflicts (e.g. the Iran/Iraq war of 1980), and Soviet
control (5). Again, the complementarity between those studies and the
framework adopted here is obvious; they require inputs on international
interactions and the effects of other emergency policies and we require
inputs on the effects of expectations.
Two distinct benefits accrue from releasing stockpiled oil in a crisis (7).
One is a reduction in the world oil price; this benefits all oil importing
nations. The other is the capital gain from oil sold at a much higher price
than its purchase price; this benefit accrues only to the owner of the stock-
pile. Similarly, emergency fuel switching programs will reduce the world oil
price during an oil supply interruption, and could also produce capital gains
if the cost of fuel switching is less than the world oil price that would result
without it.
Figure 2 shows that in the absence of any other emergency policies a fuel
switching program without adding to the present Strategic Petroleum Re-
serve level would be worth $4 billion during a 3-million barrel per day world
oil supply interruption, $40 billion during a 9-million barrel per day world
oil supply interruption, and $99 billion during an 18-million barrel per day
world oil supply interruption (in this last case a $465 billion loss would be
reduced to $366 billion). The existence of this switching capability reduces
the value of having an oil, or oil-equivalent, stockpile, although for large
interruptions this value remains great. Figure 2 shows that this fuel switch-
ing capacity reduces the value of a one-billion barrel oil stock by $4 billion,
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
?
BASELINE
(100x106B U.S. ONLY STOCKPILE)
EMERGENCY FUEL SWITCH PROGRAM
(100x106B U.S. STOCKPILE)
1000x1068 U.S. ONLY STOCKPILE
3x109B OECD STOCKPILES
10004066 U.S. STOCKPILE &
0 10 20 30 40
CUT IN PERSIAN GULF OIL SUPPLY (1068D)
Figure 2 Reduction in US GNP losses attributable to US and OECD emergency supply
policies in billions of 1980 dollars.
$13 billion, and $15 billion for the 3, 9, and 18-million barrel per day cuts
respectively. For the deeper cuts, even with the assumed level of fuel switch-
ing, the value of a one-billion barrel oil stock remains large, with a GNP
saving of $115 billion for the biggest cut.
It is also true that releases from a one-billion barrel US oil stockpile,
which would limit the increase in the world oil price during a crisis, would
reduce the value of the fuel switching program (by $3 billion, $13 billion,
and $15 billion for the 3, 9, and 18-million barrel per day cuts respectively).
Again, the remaining value of fuel switching is large, especially for the
deepest crisis, $84 billion. Figure 3 shows the equally impressive benefits to
the other OECD nations resulting from the OECD-wide emergency oil
substitution program. Additionally, the interruption costs calculated here
30 I
TO SUPPLY PROGRAM
^ 3x106BD INTERRUPTION
9x106BD INTERRUPTION
I 18x106BD INTERRUPTION
U.S. WESTERN JAPAN IMPORTING
EUROPE LDC's
Figure 3 Economic benefits of 3 X 109 bbl OECD stockpiles and US emergency fuel switch
program. Reductions in GNP losses are in billions of 1980 dollars. Absolute baseline losses
are given in Figure 1.
 Approved For Release 2007/04/11: CIA-RDP83B00140R -
188 ROWEN & WEYANT
ignore unemployment and inflationary costs that would undoubtedly make
these supply-side oil emergency measures even more valuable.
Multifuel Stockpiles and Their Costs
Four components contribute to the cost of oil stockpiles: 1. physical storage
cost, 2. holding cost, 3. the cost of higher world oil prices during stockpile
acquisitions, and 4. a credit for the capital gain on the stockpiled oil (7).
If it i
s necessary to buy additional coal to ensure its availability in a crisis,
there would probably be little increase in its price, and if gas can be acquired
that would otherwise not be marketed in the next few years (say from
Canada), there would be only a small increase in its price as well. In any
case, the authors assume no price increase effect for these two fuels. Addi-
tionally, there is already a large amount of gas in storage for exceptionally
cold winters (11); some of this could be in excess supply in an average year
and be usable in an emergency.
It is assumed that in an emergency the United States can substitute gas
for one million barrels per day of oil use, that one half of this gas is
purchased at $5.00 per million cubic feet (Mcf), and that the annual storage
cost per barrel of oil equivalent would be $3.80.8 Thus, the United States
could achieve a 365-million barrel oil-equivalent stockpile at an annual cost
of only $3.80 X 365 million X 0.5 (i.e. $700) million annually. This is nearly
$1.5 billion per year cheaper than an equivalent oil-only stockpile (15). This
result is attributable solely to a smaller increase in the world oil price effects
in the former case. In fact, the assumed billion barrels of oil equivalent
abroad that is required to match a 500-million barrel US stockpile is made
up largely of coal. In this case, there would be no upward pressure on the
world oil price at all, with the stockpile cost to the US being independent
of the stockpile buildup elsewhere. Because of the limit on the increase of
the amount of gas that could be accommodated in the existing gas transpor-
tation and distribution system and substituted for oil uses, it is assumed that
additional US petroleum stocks would have to be in the form of oil.
The Breakeven Analysis
How do these lower stockpile costs affect their desirability? Table 4 shows
the stockpile breakeven probability analysis on the payoffs, given a 9-million
'In an average year about one trillion cubic feet of natural gas could be extracted from
existing stocks beyond seasonal demands in an emergency; therefore, this gas does not have
to be purchased. The annual cost of storing added gas bought at $5.00 per Mcf would come
to about $3.80 per barrel of oil equivalent. This assumes a 6% real discount rate, new storage
sites whose cost is $1 per Mcf of gas stored, annual operating costs of $.06 per Mcf, and an
additional 25% bought as cushion gas needed to replenish depleted gas fields assumed to be
used as added storage sites.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
is
0
0
O
b
O a
O
O% ON h N
l~
l-
O\
I- .O
p
'n 'A N
V
O
C
00
'.O 00
0
'd
O
O O
O a
00 00 N
O p
O
~O
'd
- - 0,
00
4
V
O
O O
O p
O
M
O
O
00
O
00
O O
O
O
Z
%0
- -
l-
O N
h
fl-
%0
IO
O
O O
O
'd
O p
O
O~
V
M
00 -"-
p
O
tiSO
O
M
M
W
O
N n
O O
M O\ ON N N
N O O O O O
z z. z z z z
c 0 0 G O O C O O
m w ti
00 0 O03OuOu
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
barrel per day interruption, from oil stockpiles, and for multifuel stockpiles.
In the absence of the emergency fuel switching program, US gas plus oil
stockpiles always have a lower breakeven probability than the oil-only
stockpiles; the benefits are the same and the costs lower. In fact, the gas
increment to the existing Strategic Petroleum Reserve stockpile is justified
by a one in thirty (.032) annual probability of a 9-million barrel per day
interruption. The presence of the fuel switching program decreases the
benefits of stocks and increases the breakeven probability by a little, with
the effect increasing with growing stockpile size. This is because fuel switch-
ing (largely to coal) costs very little, while stockpiling oil and gas costs more
for equivalent benefits.
The existence of equivalent stockpiles also in the other OECD countries
approximately doubles the benefits of a 500 X 106 bbl stock in the United
States; this reduces the breakeven probability associated with a gas plus oil
stockpile of that size for the United States to about .014 (i.e. a 1 in 70 chance
per year of an interruption of 9 X 106 bbl/d). For larger stocks, both benefits
and costs grow, but because the costs climb more rapidly, the breakeven
probability justifying these larger stocks is higher than when the United
States builds the stocks on its own. But it takes no more than a .07 annual
probability to justify an OECD oil-equivalent stockpile of 4-5 billion bar-
rels.
Measures designed to provide substitutes for oil imports during a crisis
appear to be extremely valuable. Furthermore, unlike complete reliance on
a strategic petroleum reserve, the use of other complementary options-e.g.
gas stockpiles, increased coal and nuclear electricity generation, emergency
fuel switching in industry-would require little investment before the crisis
emerges, which minimizes the burden on the federal budget.
In addition to these import-substitution measures, there are several op-
tions for reducing oil import demand during a crisis. Many emergency
allocation, rationing, and tax schemes have been proposed, but because of
their superior economic efficiency emergency oil import tariffs are consid-
ered as a prototype.
DEMAND-SIDE OPTIONS: EMERGENCY TARIFFS
AS THE PROTOTYPE
A sudden supply interruption is equivalent to a rapid leftward shifting of
the supply of oil production as a function of price. Because, in the short run,
consumers cannot easily reduce their demand for oil, the result is a large
increase in the price of oil and a large increase in the transfer of wealth to
the remaining producers of oil. If importers act to reduce their demand
beyond that induced by the change in the world price, they could save much
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
SURVIVING AN OIL CRISIS 191
of the wealth that would otherwise be transferred to the remaining pro-
ducers.
This demand reduction can result from either the imposition of a tariff
on imported oil, taxes on oil production, or, almost equivalently, a quota.
Recent fashion, as expressed in international Western summit meetings, has
been to announce (so far nonbinding) import quotas. As students of elemen-
tary economics know, setting physical quotas is nearly equivalent to setting
prices.9 The issue is whether the importing nation can retain a substantial
amount of huge wealth that will otherwise be transferred to the remaining
oil producers if they do not act to limit demand?
The idea of imposing an oil import tariff during an emergency has only
recently been proposed. Figure 4 illustrates the logic of an emergency tariff.
In Figure 4a, the world oil market clears and a tremendous amount of
additional payment is made to the remaining oil producers. In addition,
there is a "deadweight" loss in the domestic economy, as many of those who
once purchased oil at a price much less than its value to them can no longer
get it. Figure 4b shows that an emergency tariff raises the domestic oil price
considerably during the crisis; this increases the deadweight loss but reduces
the world oil price and the magnitude of the additional wealth transfer
considerably. And the total cost of the interruption (the shaded area in the
two diagrams) is reduced.
Costs and Benefits of Tariffs and Quotas
For example, in the absence of demand-reducing action by importers (and
leaving aside the effect of stocks of oil and other fuels), a 9-million barrel
per day interruption would result in the world oil price increasing to about
$113 per barrel. If the OECD, acting in concert, were to impose a tariff of
$100 per barrel, the result would be a lower world oil price of $83 per barrel,
Me difference, which may not be a trivial one, especially in the long run, is that the tariff
presents foreign suppliers with a demand schedule, whereas a quota presents foreign suppliers
with a specific target value, i.e. the stipulated amount allowed into the importing country. A
wealth maximizing oil exporter facing the quota knows exactly where to set his production,
at the permitted level. He has no incentive to set any lower price. On the other hand, when
faced with a tariff schedule, an increase in price further reduces demand and, because of the
effect of those demand reductions on his oil revenues it will not, in general, be in his best
interest to increase the price as much as he would when the quota sets a limit on the level of
demand. The assumption made here is that during a crisis the increase in the wealth transfer
to the remaining oil producers is so enormous and the political pressures so great they they
would not have sufficient motivation or time to fine-tune their pricing policy away from that
prescribed by the very strong market forces that would exist. The fact that the interrupted
suppliers would be those most influential in price setting in the past, those in the Persian Gulf,
lends credibility to this assumption.
Tariffs and quotas are analytically identical if the quota is continuously adjusted. However,
institutional practices prescribe intermittent decisions.
Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
0
0
OIL IMPORTS
Figure 4 Effect of emergency tariffs on world oil market during a supply interruption.
a $57 billion net saving to the US economy, and a $188 billion saving to
the OECD as a whole. The rest of the world would also benefit from this
action."
The saving would be much smaller if the United States were to act alone
in reducing demand in this way. If a tariff of $104 per barrel were to be
imposed by the United States only, the result would be a net saving to the
country of $15 billion and a saving to the OECD as a whole of $44 billion.
A US tariff of $104 per barrel in this case would be optimal.
Comparison of Emergency Quotas and Tariffs
A tariff is a tax that would be paid on each barrel of imported oil, and its
revenues could be returned to the domestic economy instead of being sent
to the oil producers. The increase in the price of oil paid by domestic
consumers eliminates some uses of oil that were worth more than the
preinterruption price, but not as much as the price during the interruption.
The higher energy price also decreases the value of the other factors of
production. The imposition of the tariff does not avoid this inevitable part
of the economic loss resulting from the interruption.
It is impossible to reduce imports in a more efficient way than by impos-
"In this case, a uniform tariff of $258 per barrel would result in the maximum saving in
OECD GNP ($277 billion) during the 9-million barrel per day interruption. However, a tariff
of that size would produce a smaller reduction in US GNP ($53 billion) than the optimal
OECD tariff, primarily because of its greater import substitution potential; the increase in the
deadweight loss in the United States starts to exceed the decrease in the wealth transfer to oil
producers at a lower tariff level than it does in Japan or Western Europe.
 Approved For Release 2007/04/11: CIA-RDP83B00140R000100010031-0
ing a tariff. As the tariff is increased, oil is progressively removed from its
least valuable remaining uses. Price controls and allocation rules will always
cost more unless the oil is allocated in exactly the same way as the tariff
would allocate it; if any use that is cut off has a higher value than one that
continues to be satisfied during the interruption, the economy pays a higher
price than it otherwise would. In practice the information requirement of
such an allocation system renders it less efficient. (However, an allocation
system may be perceived as more equitable, in which case a further loss of
GNP is accepted as the price for such fairness.)
The efficiency of the tariff as a means of reducing imports commends it
as an emergency measure. Any other demand-reducing measure will, at
best, have the same value via reductions in the world oil price, but will cost
at least as much to implement.
How the Cost of Imposing Emergency Tariffs is Calculated
A tariff on oil imports creates a gap between the world oil price and the price
domestic consumers pay for oil, which leads to a reduction in consumption
and in import demand, which can lead to a reduction in the world oil price.
But this reduction in demand is not free.
Another basic result from the use of Sweeney's model (41) is employed
to calculate the cost of imposing the tariff. As the tariff is increased, the
decrease in the value of the oil exceeds the decrease in the cost of importing
oil by precisely the size of the tariff. Thus the cost of increasing the tariff
by a small amount will be the product of the resulting decrease in imports
and the size of the tariff at that point. The total cost of imposing a tariff
of a given size will be the sum of the costs of the requisite number of
successive small tariff increases. For example, the calculations show that an
emergency tariff of $258 per barrel applied by all of the OECD countries
during a 9-million barrel per day oil supply interruption would cost the
United States $103 billion in deadweight losses and reduce US oil demand
by 3.6 million barrels. A $258 per barrel tariff during a 9-million barrel per
day world oil supply interruption is certainly possible. Such a tariff would
reduce the world oil price from $113 per barrel to $46 per barrel, which
would reduce the wealth transfer from the United States during the inter-
ruption by almost $70 billion, and the total cost to the United States
(increase in wealth transfer plus deadweight loss) attributable to higher
world oil prices during the interruption by $156 billion. Thus, the United
States would save $156 billion minus $103 billion, or $53 billion, and the
OECD as a whole would save $277 billion, or about half of the total cost
of the interruption of 9 X 106 bbl/d. This conclusion, and these numerical
estimates, obviously depend on the response of the remaining oil producers
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
to the importers' actions. This estimate assumes that the remaining pro-
ducers hold production constant; if they have other goals, a lower tariff is
likely to be optimal. Additionally, the higher domestic prices resulting from
the imposition of the tariff could lead to additional inflationary costs. Conse-
quently, the a fortiori argument concerning inflationary and unemployment
costs used in the evaluation of the emergency supply options does not hold
here. However, compensating monetary and fiscal policies in concert with
a well-designed program to rebate the tariff revenues to consumers could
enable the benefits calculated here to be achieved.
Emergency Tariffs in all OECD Countries
If all the OECD nations impose emergency tariffs, the effect on the world
oil price during the interruption will be much greater than that which
results from unilateral action by the United States.
Uniform OECD emergency tariffs are calculated in much the same way
as US emergency tariffs. A uniform emergency tariff is assumed to be
imposed in the United States, Japan, and Western Europe. This tariff re-
duces demand during the interruption and results in lower world oil prices
during the baseline interruptions. The imposition of the tariff and the subse-
quent world oil price increase both cost the oil importing economies. The
total reduction in the wealth transfer to the oil exporting nations leads to
a lower total cost than in the baseline interruption where no tariffs are
assumed. An optimal OECD emergency tariff is defined as the tariff that
minimizes the total cost of the interruption to the OECD as a whole.
O REDUCTION IN LOSS ATTRIBUTABLE
TO TARIFF
3x106BD INTERRUPTION
D 9x106BD INTERRUPTION
018*108BD INTERRUPTION
$25
El
WESTERN JAPAN IMPORTING
EUROPE LDC's
Figure 5 Economic benefits of optimal OECD emergency tariffs. Reduction in GNP losses
are in billions of 1980 dollars. Absolute baseline GNP losses are given in Figure 1.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
Figure 5 shows the benefits of optimal uniform OECD emergency tariffs
relative to the costs of the baseline interruption. These benefits are much
greater than those attributable to a tariff imposed by the United States
acting alone.
For the smallest interruption case, 3 X 106 bbl/d economic losses are
almost eliminated. It shows not only the large saving for the OECD coun-
tries but also that the free ride for the oil importing LDCs is significant.
They gain all the benefits of the lower world oil price resulting from the
OECD tariff but do not pay the cost of imposing a tariff on their own oil
imports. This leads to the question of whether mandatory restrictions on
LDC imports in a major world energy crisis would significantly increase the
value of demand-reducing measures elsewhere.
COMBINED VALUE OF SUPPLY AND DEMAND
MEASURES
Does the existence of an emergency oil/gas/coal supply program greatly
reduce the value of an emergency oil demand reduction program and vice
versa, or are the values of the two types of contingency plans roughly
additive? It would be surprising if the value of the simultaneous application
of the two kinds of programs equalled the sum of their individual benefits.
Each is focused on directly reducing the size of the economic loss to the
United States and indirectly reducing the size of the world oil price and
economic losses during the interruption to everyone.
A 1,000-million barrel US gas and oil stockpile" and emergency fuel
switch program is worth $87 billion during a 9 X 106 bbl/d interruption.
If an equivalent number of days of oil import coverage is provided by coal,
gas, and oil stocks outside the United States, the value of the US emergency
supply program to the United States grows from $87 billion to $128 billion.
Figure 6 shows that OECD-wide emergency tariffs reduce that value to
$102 billion. On the other hand, the existence of the OECD-wide emergency
supply program reduces the value of the OECD emergency tariff to the
United States from $53 billion to $27 billion. A smaller tariff is required in
this case than with the US-only emergency supply program. The United
States would derive a greater fraction of the benefits of that OECD tariff.
Thus, the OECD supply program increases the value of an optimal OECD
"The existing Strategic Petroleum Reserve level of about 100 X 106 bbl was assumed. This stock
plus the target level for the gas stockpile comes to around 500 X 106 bbl of oil equivalent. The
remaining 500 X 106 bbl would consist of new oil stocks.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
emergency tariff to the United States relative to a US-only emergency
supply program.
The most important message derived from Figure 6 is that the combina-
tion of the supply measures the authors propose in all of the OECD coun-
tries, together with an OECD-wide emergency tariff, can almost eliminate
economic losses in the United States, Western Europe, and Japan from a
9-million barrel per day interruption.
These measures would also, of course, greatly reduce losses from a full
Persian Gulf interruption. The combined effect of OECD emergency tariffs
and a three billion barrel oil-equivalent stockpile would be to reduce eco-
nomic losses in the United States by 60% and in Japan, Western Europe,
and the oil importing LDCs by even more. The resulting losses (aside from
possible macroeconomic effects) would be held to less than 10% of GNP
for all regions and to around 5% for the United States. In any case, these
losses are less than one half of the loss without the protection. And the cost
of such a program to the United States would be around five billion dollars
annually.
In summary, the value of emergency supply and demand measures accu-
mulates. The existence of emergency demand-limiting programs reduces the
value of emergency supply programs by less than emergency supply pro-
grams reduce the value of emergency demand programs, but even then
significant benefits remain. This suggests the efficacy of a balanced and
integrated emergency oil supply and demand program.
D REDUCTION IN LOSS
? LOSS WITH 3x1098 OECD STOCKPILES AND
U.S. EMERGENCY FUEL SWITCH PROGRAM
IS LOSS WITH OPTIMAL OECD EMERGENCY TARIFFS
- LOSS WITH BOTH SUPPLY PROGRAM AND TARIFF
Im
$182
U.S. WESTERN JAPAN IMPORTING
EUROPE LDC's
Figure 6 Comparative benefits of OECD emergency supply program and OECD emergency
tariffs during a 9 X 106 bbl/d oil supply interruption. Reduction in GNP losses are in billions
of 1980 dollars. Absolute baseline GNP losses are given in Figure 1.
 Approved For Release 2007/04/11: CIA-RDP83BOO14OR000100010031-0
?
Literature Cited
1. Pan Heuristics. 1980. Report on Persian
Gulf Oil and Western Security. Report
to the US Dept. Energy. Marina Del
Rey, Calif: Pan Heuristics. (18 Chs.)
2. Khalilzad, Z. 1980. Turmoil in the Per-
sian Gulf. See Ref. 1, Ch. 2. 29 pp.
3. Samore, G. S. 1980. The Persian Gulf
and energy security in the 1980s. See
Ref. 1. Ch. 3. 127 pp.
4. Wohlstetter, A. 1980. Protecting Per-
sian Gulf oil: US. and Alliance Military
Policy. See Ref. 1, Ch. 4. 89 pp.
5. Brody, R. 1980. The implications of
Soviet control of Persian Gulf Oil See
Ref. 1, Ch. 5. 29 pp.
6. Rowen, H. S., Weyant, J. P. 1980. The
economic impacts of oil supply interrup-
tions. See Ref. 1, Ch. 6. 37 pp.
7. Rowen, H. S., Weyant, J. P. 1980. The
value of oil stockpiles See Ref. 1, Ch. 7.
32 pp.
8. Missner, S. L. 1980. Private sector pe-
troleum stockpiling. See Ref. 1, Ch. 8.
31 pp.
9. Rowen, B. 1980. Coal in an oil crisis.
See Ref. 1, Ch. 9. 134 pp.
10. Jones, G. S. 1980. Increased nuclear
power in an oil crisis. See Ref. 1, Ch. 10.
19 pp.
11. McDonald, R. 1980. The potential for
storing natural gas See Ref. 1, Ch. 11.
25 pp.
12. Pittinger, L. 1980. Emergency pe-
troleum production capacity outside the
Persian Gulf. See Ref. 1, Ch. 12. 91 pp.
13. Rowen, H. S., Weyant, J. P. 1980. Re-
ducing oil demand in an emergency. See
Ref. 1, Ch. 13. 31 pp.
14. Hogan, W. W. 1980. Import Manage-
ment and Oil Emergencies. See Ref. 1,
Ch. 14. 65 pp.
15. Rowen, H. S., Weyant, J. P. 1980. An
integrative program for surviving an oil
crisis See Ref 1, Ch. 15. 29 pp.
16. Kline, D., Weyant, J. P. 1980. Reducing
Persian Gulf dependence over time. See
Ref. 1, Ch. 16. 92 pp.
17. Deese, D. A. 1980. The oil-importing
developing countries. See Ref. 1, Ch. 17.
47 pp.
18. Nye, J. S., Deese, D. A. 1980. Interna-
tional cooperation and energy security
policies in oil-importing nations. See
Ref. 1, Ch. 18. 87 pp.
19. Rowen, H. S. 1980. Overview. See Ref.
1, Ch. 1. 62 pp.
20. Assistant Secretary for Policy and Eval-
uation, US Dept. Energy. 1980. Reduc-
ing Oil Vulnerability: Energy Policy for
the 1980s, Rep. No. DOE/PE-0021. An
?
analytical report to the Secretary of En-
ergy. Wash. DC: DOE
21. Deese, D. A., Nye, J. S., eds. 1981. En-
ergy and Security Cambridge, Mass:
Ballinger. 424 pp.
22. Plummer, J. L. 1981. Methods for mea-
suring the oil import premium and the
oil stockpile premium Energy J 2(1):
1-18
23. Kline, D., Weyant, J. P. 1979. OECD
Oil Demand Projections. Stanford Int.
Energy Program, Stanford Univ., Calif.
24. Energy Modeling Forum. 1981. US Oil
and Gas Supply. Vols. 1, 2. Stanford
Univ., Calif: EMF
25. Energy Modeling Forum. 1981. World
Oil. Vols. 1, 2. Stanford Univ., Calif:
EMF
26. Energy Modeling Forum. 1980. Aggre-
gate Energy Demand Elasticity. Vols. 1,
2. Stanford Univ., Calif: EMF
27. Mork, K. A., Hall, R. E. 1980. Energy
prices and the U.S. economy in 19881. Energy J. 1(2):41-53
28. Mork, K. A., Hall, R. E. 1980. Energy
Prices, Inflation, and Recession. Energy
J. 1(3)31-63
29. Pindyck, R. S. 1980. Energy price in-
creases and macroeconomic policy. En-
ergy J. 1(4):1-20
30. Curtis, W. P. 1979. Macroeconomic
Effects of Petroleum Supply Interrup-
tions. US Dept. Energy, Energy Inf.
Adm., Wash. DC: DOE
31. Eckstein, 0. 1979. Macroeconomic
Analysis of Price Shocks. Presented at a
Con, f.Y on Energy Prices, Inflation, and
Econ. Activity, MIT Energy Lab. Work-
ing Pap. No. MIT-EL79-065WP. Cam-
bridge, Mass: MIT Energy Lab. 18 pp.
32. Eckstein, 0. 1979. Macroeconomic pol-
icy responses to energy price shocks See
Ref. 31. 8 pp
33. Gramlich, E. M. 1979. Macro policy
responses to price shocks. In Brookings
Papers on Economic Activity: pp. 125-
89. Wash. DC: Brookings Inst.
34. Nordhaus, W. D. 1980. The macroeco-
nomic cost of microeconomic shocks. In
Brookings Papers on Economic Activity,
pp. 341-88. Wash. DC: Brookings Inst.
35. Organization of Economic Cooperation
and Development. 1980. The impact of
oil in the world economy. In OECD
Economic Outlook; No. 27: pp. 114-30.
Paris: OECD
36. Probyn, D., Empey, W., Wyss, D. 1980.
OPEC III and the world economy. In
Data Resour. Rev. 4(1)):1.16-1.19
37. Smith, J. R. 1979. Recllation: Recession
and Inflation in the United States, its
 Approved For Release 2007/04/11 : CIA-RDP83BOO14UR009100010031-0
?
0
Causes and Influences from the Foreign
Sector. In Economic Notes by Monte Dei
Pasch,, 8(3):88-94
38. Solow, R. M. 1978. What to do macro-
economicall when OPEC comes.
NBER Conference on National Expecta-
tions and Economic Policy, pp. 249-67.
Chicago: Univ. Chicago Press
39. Thurmar, S., Berner, R. 1979. Analysis
of Oil Price Shocks in the MPS Model.
See Ref. 31. 25 pp.
40. US Dept. Energy, Economic Regula-
tory Administration. 1979. Standby
Gasoline Rationing Plan, Rep. No.
DOE/ERA-0046. Wash. DC: DOE
41. Sweeney, J. L. 1979. Energy and eco-
nomic growth: A conceptual frame-
work. In Directions in Energy Policy.A
Comprehensive Approach to Energy Re-
source Decision Making ed. B. Kursuo-
glic, A. Permutter, New York: Aca-
demic
42. Hogan, W. W., Manne, A. S. 1977. En-
ergy-economy interactions: The fable of
the elephant and the rabbit? In Energy
and the Economy. Energy Modeling
Forum Forum Rep. No. 1, Vol. 2. Stan-
ford Univ., Calif: EMF
43. Schlesinger, B. 1980. Potential of In-
creased Gas Supply Capability to Reduce
Impacts to the U.S. Economy of a Major
Oil Supply Disruption. Arlington, Va:
Am. Gas Assoc.
44. Purvin & Gertz, Inc. 1980. An Analysis
of Potential for Upgrading Domestic
Refining Capacity. Arlington, Va: Am.
Gas Assoc.
45. Teisberg, T. J. 1980. A Dynamic Pro-
gramming Model of the U.S. Strategic
Petroleum Reserve. Energy Lab. Mass.
Inst. Technol., Cambridge, Mass.
46. Pugliaresi, L., Sweetnam, G. 1980. Stra-
tegic Petroleum Reserve Analysis US
Dept. Energy. Wash. DC: DOE
47. Chao, H., Manne, A. S. 1980. Oil Stock-
piles and Import Reductions: A Dynamic
Programming Approach Palo Alto,
Calif. Electric Power Res. Inst. (Draft)
48. Balas, E. 1979. The Strategic Petroleum
Reserve: How Large Should it be? Grad-
uate School Ind. Admin., Carnegie-
Mellon Univ., Pittsburgh, Penn.
49. Holt, B. J., Berkman, M. 1980. An Eval-
uation of the Strategic Petroleum Re-
serve. Congressional Budget Office, US
Congress. Wash. DC: GPO
50. Blankenship, J., Barron, M. J., Lane,
W. 0., Eschbach, J., Bower, J. L. 1980.
The Energy Problem: Costs and Policy
Options US Dept. Energy, Assistant
Secretary for Policy & Evaluation,
Office of Oil Policy. Wash. DC: DOE
51. Coplon, G. V. 1979. DOE Analysis of
the Appropriate Size of the Strategic Pe-
troleum Reserve. US Dept. Energy, As-
sistant Secretary for Policy & Evalu-
ation, Office of Emergency Prepared-
ness. Wash. DC: DOE
52. Borison, A. 1980. Strategic Petroleum
Reserve Program: the Policy Questions.
Palo Alto, Calif: Applied Decision
Analysis
53. Butler, G. D., Kilgore, W. C. 1979. Pe-
troleum Supply Vulnerability, 1985. US
Dept. Energy, Energy Information Ad-
ministration, Office of Integrative Ana-
lysis. Report No. DOE/EIA-0102/44.
Wash. DC: DOE
54. Controller General of the United States.
1979. Factors Influencing the Size of the
U.S. Strategic Petroleum Reserve, Re-
port to Congress No. ID-79-8. Wash.
DC: General Accounting Office
55. Novicky, E. R. 1979. An Assessment of
the Macroeconomic Benefits of the Stra-
tegic Petroleum Reserve. Scientific Time
Sharing Corp. Management Technol.
Div. Bethesda, Md: STSC