Developing Iraq’s Oil Industry to Maximize Government Net Revenues 1 DEVELOPING IRAQ’S OIL INDUSTRY TO MAZIMIZE GOVERNMENT NET REVENUES June 2009 Mohammad Mazin Hamid Ali Al-Moumen Department of Economics Stanford University Stanford, CA 94305 [email protected]under the direction of Professor Geoffrey Rothwell ABSTRACT This paper seeks to determine whether (1) nationalizing Iraq’s oil industry or (2) developing it through production-sharing agreements with international oil companies (IOCs) will generate the highest amount of net revenues for the Iraqi government. I propose to use the development of the Majnoon oil field in southern Iraq as a case study to answer this question. I consider a timeframe of forty years and construct the estimated revenues and costs associated with developing Majnoon under nationalization, comparing them to those associated with developing Majnoon under a PSA similar to that employed in Oman. I conclude that if the Iraqis believe that the Iraq National Oil Company (INOC) can be at least 87.5% as efficient as an IOC, then Majnoon would be more profitable under nationalization. If the INOC cannot be at least 75% as efficient, then Majnoon would be more profitable under a PSA. If the INOC’s relative efficiency is between 75%-87.5% (deemed the “Indecision Interval”), then Iraqi decision-makers must engage in further analysis to determine the profit-maximizing option. The case of Majnoon provides insights important for the decision-makers to consider when deciding on whether to nationalize Iraq’s oil industry or not. Keywords: Iraq, oil, Majnoon, international oil company, Iraq National Oil Company *Acknowledgments: I would like to thank Professor Geoffrey Rothwell for his guidance, kindness, support, patience, and mentorship; Junko Pierry and Koren Bakkegard for their administrative support; Tzvetan Tchoukalov, Andrew Nigrinis, and Bilal Badawi for their time and insight on analytical methods; Mohammad Ali, Ulugbek Baymuradov, Myles Bradley, Kevin Danna, Ali Habib, Samy Hamdouche, and Ahlia Kattan for their constant support;; the Alis and Hilfis for being family; and Mazin Al-Moumen, Nidal Douba Al-Moumen, Manaf Al- Moumen, and Mishaal Al-Moumen for being my spirit.
103
Embed
Developing Iraq’s Oil Industry to Maximize … Iraq’s Oil Industry to Maximize Government Net Revenues 3 Concessionary agreements were more common in the early 20th century, as
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
1
DEVELOPING IRAQ’S OIL INDUSTRY TO MAZIMIZE GOVERNMENT NET REVENUES
June 2009
Mohammad Mazin Hamid Ali Al-Moumen
Department of Economics Stanford University Stanford, CA 94305
under the direction of Professor Geoffrey Rothwell
ABSTRACT
This paper seeks to determine whether (1) nationalizing Iraq’s oil industry or (2) developing it through production-sharing agreements with international oil companies (IOCs) will generate the highest amount of net revenues for the Iraqi government. I propose to use the development of the Majnoon oil field in southern Iraq as a case study to answer this question. I consider a timeframe of forty years and construct the estimated revenues and costs associated with developing Majnoon under nationalization, comparing them to those associated with developing Majnoon under a PSA similar to that employed in Oman. I conclude that if the Iraqis believe that the Iraq National Oil Company (INOC) can be at least 87.5% as efficient as an IOC, then Majnoon would be more profitable under nationalization. If the INOC cannot be at least 75% as efficient, then Majnoon would be more profitable under a PSA. If the INOC’s relative efficiency is between 75%-87.5% (deemed the “Indecision Interval”), then Iraqi decision-makers must engage in further analysis to determine the profit-maximizing option. The case of Majnoon provides insights important for the decision-makers to consider when deciding on whether to nationalize Iraq’s oil industry or not. Keywords: Iraq, oil, Majnoon, international oil company, Iraq National Oil Company *Acknowledgments: I would like to thank Professor Geoffrey Rothwell for his guidance, kindness, support, patience, and mentorship; Junko Pierry and Koren Bakkegard for their administrative support; Tzvetan Tchoukalov, Andrew Nigrinis, and Bilal Badawi for their time and insight on analytical methods; Mohammad Ali, Ulugbek Baymuradov, Myles Bradley, Kevin Danna, Ali Habib, Samy Hamdouche, and Ahlia Kattan for their constant support;; the Alis and Hilfis for being family; and Mazin Al-Moumen, Nidal Douba Al-Moumen, Manaf Al-Moumen, and Mishaal Al-Moumen for being my spirit.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
2
1. Introduction
1.1 Developing Iraqi Oil
After three wars, 12 years of sanctions, and nearly 30 years of dictatorship, Iraq is
desperately in need of reconstruction. With the world’s third largest proven reserves of oil, the
importance of Iraq’s hydrocarbons industry to the country’s redevelopment is unquestioned. The
debate, rather, pertains to the means by which to develop this industry.
While some believe that nationalizing Iraq’s petroleum industry best serves the country’s
interests, others argue that privatizing the industry – through production-sharing agreements
(PSAs) – would be more beneficial for Iraq. To contribute to the debate, this paper examines the
possible development strategies of the Majnoon oil field in southern Iraq. I compare the
government’s net revenues from nationalizing Majnoon to its net revenues when employing an
Omani-style PSA, and I determine the breakeven efficiency level that the government must
achieve before privatization becomes the more lucrative option. The analysis concludes that Iraqi
decision-makers face an Indecision Interval of 70%-87.5%: if the Iraq National Oil Company
(INOC) can achieve an efficiency level – relative to an international oil company (IOC) – of
87.5% or greater, then the Majnoon study suggests that nationalization maximizes government
net revenues. If the INOC cannot achieve at least a 70% relative efficiency score, then a PSA
would maximize government net revenues. At any efficiency score between 70% and 87.5%, it is
unclear which development strategy is more profitable to the government, so further studies must
be conducted.
Concessions, PSAs, and nationalization are the three major methods employed to develop
a country’s hydrocarbon industry (Muttitt 2005). Under the concessionary model, a government
will grant an IOC, or a consortium of IOCs, the right to extract oil. Once extracted, this oil
becomes property of the IOC. In exchange, the IOC pays taxes and royalties to the country.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
3
Concessionary agreements were more common in the early 20th century, as they are currently
viewed as a threat to a country’s sovereignty. Today, concessions also cover smaller areas and
last for a shorter period of time. In 1901, William Knox D’Arcy signed a concessionary
agreement with the Persian Shah Qajar, securing exclusive rights to potential oil discoveries in
most of Iran for 60 years – this was the first Middle Eastern concessionary agreement (Kinzer
2003).
Under a PSA, a country will grant an IOC the right to extract oil (Muttitt 2005). The IOC
provides the capital investment necessary for exploration, infrastructure, and drilling, but the oil
legally remains property of the country. Once production begins, the IOC retains cost oil, which
is a percentage of production that makes up for the company’s costs and capital investment.
After the company recuperates its costs, the country and the IOC divide the remaining
production – known as profit oil – according to contractual agreements, with the state taxing the
IOC’s share. The Energy Information Administration (EIA) states that PSAs administer only
12% of the world’s oil reserves. Smaller nations with minor oil fields are most likely to employ
PSAs: lacking the financial means and technical know-how, and facing high extraction costs,
these countries engage in PSAs so that IOCs provide capital as well as de facto insurance in case
oil is not discovered.
The third form of development is nationalization. Currently employed in Iraq, the
nationalization model stipulates that a country owns 100% of the oil and has complete decision-
making power. The state may employ IOCs only through a technical service contract, in which
the IOC provides technical and consulting services in exchange for a fixed fee (Muttitt 2005).
The Iraqi government will not entertain any discussions on concessionary agreements
(Draft Iraq Oil and Gas Law 2007), as they are politically infeasible in a country still under
foreign military occupation. Thus, Iraq has two basic options: it can either develop its
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
4
hydrocarbons industry nationally, with technical assistance provided by IOCs, or it can engage in
PSAs with IOCs.
According to the Iraq Study Group, a panel appointed by the US Congress and led by
James A. Baker, Iraq’s stabilization depends heavily on its economy (EIA 2007a). Hydrocarbons
are the staple of this economy: oil exports account for nearly 90% of the government’s revenue
and 60% of the country’s GDP (EIA 2007a). The U.S. Government Accountability Office
(GAO) states that Iraq generated a total of $90.2 billion in crude oil export sales between 2005
and 2007 (2008), while Iraqi Ministry of Oil (MoO) spokesman Asim Jihad announced that oil
revenues in 2008 alone amounted to $61 billion (AFP 2009).
Iraq, however, has failed to meet hydrocarbon production and export targets since 2004
(EIA 2007a). The Special Inspector General for Iraq Reconstruction (SIGIR) reported in January
of 2007 that the industry was plagued by procurement, transportation, and storage issues and
struggled with managing pricing controls and budget execution. The report also highlighted
corruption and smuggling as major factors hampering production.
Significant investment is thus required to modernize the hydrocarbons industry.
International organizations have estimated that total reconstruction costs for the oil, gas, and
electricity sectors will amount to over $30 billion (Sakmar 2008), and the World Bank estimates
that maintaining current oil production levels would require an additional $1 billion (EIA 2007a).
1.2 Oil in Iraq
Iraqi oil was discovered in the early 20th century, and it has been subject to periods of
nationalization as well as privatization (Muttitt 2006). In 1925, King Faisal granted a concession
to a consortium of IOCs known as the Iraqi Petroleum Company. Granting the IPC full control of
Iraqi oil for 75 years, the agreement drew widespread dissatisfaction among Iraqis, who also
objected to the revenue-sharing terms and the degree to which the IPC controlled the industry’s
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
5
development. Thus, Iraq began nationalizing its oil in 1961, and by 1972 the state reclaimed all
Iraqi oil (Behn 2007). Today, the INOC controls virtually all oil production.
Iraq possesses 115 billion barrels (bbl) of proven oil reserves, according to the Oil and
Gas Journal (EIA 2007a). Nearly 65% of proven reserves are located in the south – particularly
in Basra – but there are also significant deposits near the northern cities Kirkuk, Khanaqin, and
Mosul, which account for 20% of Iraq’s proven reserves. According to the Iraqi government, the
country possesses nine “supergiant” fields and 22 “giant” fields. The southern fields historically
accounted for nearly two-thirds of production. But only three fields account for most of the
current production: North Rumaila, South Rumaila, and Kirkuk.
Estimates of the amount of unproven reserves vary: while oil industry consultant IHS
estimates that Iraq possesses up to 100 billion bbl of unproven oil, the US Geological Survey
believes this figure to be closer to 45 billion bbl (Blanchard 2007). Iraq’s former Oil Minister,
Thamer al-Ghadban, however, believes that Iraq possesses up to 214 billion bbl of unproven oil.
Appendix 1 shows production figures from 1960-2008. In 1960, Iraq produced an
average of 0.97 million bbl/day (EIA 2007b), and throughout this decade production generally
increased. In 1973, Iraq attained the 2 million bbl/day mark, but its most impressive feat
occurred in 1979: that year, the country produced approximately 3.5 million bbl/day, a 36%
increase over the previous year’s production and, to this day, the highest level achieved by Iraq.
But the onset of the Iran-Iraq War in 1980 put a dent in Iraq’s progress. By 1981, Iraq
was producing only 1 million bbl/day, and while this figure rose back to 2.9 million bbl/day in
1989, the Gulf War precluded a full recovery. In 1991, Iraq’s oil production plummeted: the
country produced an average of approximately 0.3 million bbl/day that year, an 85% decrease
from the previous year. This was a direct consequence of the heavy bombardment Coalition
forces inflicted on Iraq’s infrastructure as well as the economic sanctions imposed by the United
Nations, which prevented Iraq from exporting oil. Production remained below 0.6 million
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
6
bbl/day until 1997, the year after the UN Oil-for-Food Programme was initiated. Under the
program’s stipulations, the government could export oil in exchange for what were deemed to be
essential products. This newfound demand for Iraq’s oil boosted production, and by 2000, Iraq
was producing nearly 2.6 million bbl/day. The 2003 invasion of Iraq reduced production to levels
below 1.4 million bbl/day; but since 2004, production has hovered around 2 million bbl/day.
While Iraq is currently producing nearly 2.4 million bbl/day, the MoO aims to increase
production to six million bbl/day by the end of 2010 (EIA 2007a). To achieve this goal, the Iraqi
government has prioritized the development of four southern oil fields: Halfaya, Nahr Umar,
West Qurna, and Majnoon. Majnoon ranks among the largest oil fields in the world, with an
estimated 21 billion bbl of oil (Muttitt 2005). Actual output has never topped 60,000 bbl/day, and
the field currently produces only 40,000 bbl/day. Yet Thamer Al-Ghadban, the former Minister
of Oil believes that Majnoon’s potential output could peak at around 600,000 bbl/day. According
to Deutsche Bank, this would entail development costs of approximately $4 billion. The Iraqi
government firmly believes that succeeding in modernizing the petroleum sector will depend
heavily on Majnoon’s development.
1.3 Legal Background
Developing Iraq’s hydrocarbon industry requires a hydrocarbon law that outlines the
terms dictating the development and management of Iraqi oil and natural gas. In 2007, the Iraqi
Cabinet approved the Draft Iraq Oil and Gas Law, which included stipulations for restructuring
Iraq’s Ministry of Oil, creating an Iraqi National Oil Company, and defining revenue-sharing
policy. Appendix 2 is a translated version of the Draft Law.
One of the central elements of the proposed legislation is the creation of the Federal Oil
and Gas Council (FOGC), which “would become the most powerful body in Iraq’s oil sector”
(Blanchard 2007, p. 4). Along with the power to approve or reject the transfer of exploration and
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
7
production rights, it would also possess the authority to review all petroleum contracts and
determine all oil and natural gas industry policies. The Prime Minister or his nominee would
serve as the President of the FOGC, which would also include: the federal government’s
Ministers of Finance, Oil, and Planning; the Director of the Iraqi Central Bank; a regional
government minister from each region; a representative from any producing governorate not
included in a region; the CEOs of major petroleum companies like the INOC and the Oil
Marketing Company; and up to three experts in the fields of economics, finance, and petroleum.
The MoO’s mandate is also laid out in the draft law. Along with the FOGC, the MoO is
responsible for drafting legislation and federal petroleum policies (Blanchard 2007). It also
monitors activities in the industry and enforces legislation. It oversees petroleum operations,
ensures that documented costs are accurate and properly recuperated, and keeps track of
government revenues. The MoO also represents the Iraqi government in regional and
international forums, negotiating multilateral and bilateral treaties with other countries and
organizations. Essentially, the MoO is involved in proposing and enforcing legislation,
monitoring petroleum operations, and representing the Iraqi federal government within and
beyond Iraq.
The INOC is fully owned by the central government, but it “is financially and
administratively independent and runs on commercial bases” (Draft Oil and Gas Law 2007, p.
13). Its main function is to participate in exploration and production operations. It must sell the
crude oil it produces to the Oil Marketing Company at a price that covers delivery costs as well
as a “reasonable” profit. But it is also involved in the downstream processes of transportation,
storage, marketing, and sales. The INOC has the right to participate in international projects
involving both upstream (exploration, development, and production) and downstream
components, to acquire assets in local and foreign entities, and to form fully-owned subsidiaries
throughout Iraq. The federal government, however, must approve all decisions.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
8
The proposed law, however, is in legal limbo. Parliament has yet to approve the draft law
due to various points of contention (Blanchard 2007). While issues related to provincial
sovereignty, revenue sharing, and management of petroleum reserves remained pertinent, the
draft legislation’s provisions for foreign participation were the major stumbling block. While the
draft does not mandate the use of PSAs, Part 5 of Article 9 states, “The Model Contracts may be
based upon Service Contract, Field Development and Production Contract . . .” (Draft Oil and
Gas Law 2007, p. 16), essentially legalizing PSAs. Many Iraqis have expressed concern about
this de facto “denationalization,” pointing to the fact that Iraq would be the only major
petroleum-producing country in the region that would permit IOCs to control upstream
operations (Muttitt 2005).
1.4 What Course of Action Should Iraq Take?
Muttitt (2005) argues that Iraq should nationalize its petroleum industry. He looks at
PSAs employed in Russia, Libya, and Oman: Libya and Oman both share similar physical
conditions with Iraq, while Russia is the only country employing PSAs that has reserve amounts
comparable to Iraq’s. Comparing revenues generated under these three PSAs to expected
revenues from a nationalization scheme, Muttitt calculates that at a price of $40 per bbl, Iraq
would lose between $74 billion and $194 billion over the lifetime of the proposed PSAs. This
represents around two to seven times the current budget of the Iraqi government. Muttitt also
finds that the PSAs under examination would grant the oil companies annual rates of return
ranging “from 42% to 62% for a small field, or 98% to 162% for a large field.” (2005, p. 23) Oil
companies generally consider projects with internal rates of return of 12% to be profitable. He
also points out that Iraq’s oil-rich neighbors constitutionally ban PSAs, and that Iraq has more
lucrative investment-generating options. These include financing development through
government revenues, using future oil production as collateral to borrow money, or employing
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
9
IOCs to provide technical and consulting services. He therefore concludes that Iraq should
nationalize its oil industry to maximize revenues.
The purpose of this paper is to determine whether Iraq should nationalize its
hydrocarbons industry or open it up for IOCs to develop. I compare (1) the net revenues the Iraqi
federal government would generate over a period of 40 years by nationalizing Majnoon to (2)
those that would be generated under a PSA similar to that employed in Oman. I initially assume
that the INOC is as efficient as any IOC, because Iraq can employ IOCs for consulting services
and technical expertise under a technical service agreement (this assumption is not founded, so I
correct for this by multiplying production figures under nationalization by values between zero
and one. This is to account for the national company’s inherent relative inefficiency and also to
price the value of the technical service agreement, which is treated as an inefficiency).
For now, I assume the INOC is as efficient as an IOC and that a service agreement is free,
and I find that nationalization generates approximately $18.8 billion more in net revenues than a
PSA. Once I account for the aforementioned inefficiencies, I calculate a threshold inefficiency of
approximately 81.5%: this means that nationalization remains the net revenue-maximizing
option so long as it is at least 81.5% as efficient as an IOC. The Iraqi decision-maker must
therefore decide whether Iraqi nationalization with the use of a service agreement would meet
this threshold.
I would propose that if Iraq believes that the INOC can achieve 86.5% efficiency, it
should proceed with nationalizing Majnoon. If, on the other hand, it does not believe the INOC
can achieve at least 76.5% efficiency, then nationalization should be abandoned. This leaves an
interval of 76.5%-86.5% efficiency, which I term the “Indecision Interval”. If the INOC can
achieve an efficiency level within the Indecision Interval, then further study would be required to
determine whether nationalization is the most profitable alternative. How I drove these initial
results will be discussed in subsequent sections.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
10
2. Literature Review
2.1 Productivity and Demand/Supply Functions
The comparison between a national oil company (NOC) – like the INOC – and major,
integrated IOCs like ExxonMobil, Shell, and BP is essential to analyzing how a country’s
hydrocarbons industry should be developed. But how exactly does one rate the performance of
different firms within the same industry?
The essential function of a firm is to convert inputs into outputs (Coelli et al. 2005). One
basic performance measurement is productivity, which is a measure of the amount of input
required to attain a certain level of output. The basic productivity equation divides firm output by
its inputs:
Productivity ≡ Outputs/Inputs (1)
Yet firms often utilize multiple inputs to produce single or multiple outputs. In such cases, inputs
would need to be aggregated into a single index of inputs. This facilitates the calculation of Total
Factor Productivity (TFP), which is a productivity measure that includes all factors of
production.
An industry is made up of firms with many different productivity levels. Yet the industry
as a whole has a theoretical frontier that represents the maximum production that can be attained
by any given amount of inputs. This is known as the production frontier, and it represents the
current state of technology in the industry. Any firm that produces at the production frontier is
considered technically efficient: for a given level of inputs, it has produced the greatest possible
amount of output. Figure 1 illustrates the production frontier:
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
11
Figure 1: Production Frontier
Coelli, Timothy J, D.S. Prasada Rao, Christopher J. O’Donnel, George E. Battese. 2005. An Introduction to
Efficiency and Productivity Analysis. New York; Springer Science + Business Media, Inc.
The y-axis denotes the output produced by the inputs, which are represented by the x-axis. The
industry’s production frontier is depicted by the curve OF’. As seen, Firms B and C are at the
production frontier: they produce the maximum amount of output given the level of inputs they
utilize. Firm A, on the other hand, does not: although it uses the same amount of inputs as Firm
B, it produces less. Thus, Firms B and C are considered technically efficient, while Firm A is
not.
It is important to touch on the issue of productivity versus efficiency. Although a firm
may be on the production frontier and therefore technically efficient, it may still be able to
increase its productivity. Efficiency only means that a firm has maximized output at a given level
of input, but a firm may be able to achieve greater productivity at another level of input. This is
an issue of scale: the optimal scale of inputs is that level which maximizes productivity. Figure 2
illustrates this concept:
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
12
Figure 2: Optimal Scale of Productivity
Coelli, Timothy J, D.S. Prasada Rao, Christopher J. O’Donnel, George E. Battese. 2005. An Introduction to
Efficiency and Productivity Analysis. New York; Springer Science + Business Media, Inc.
In this graph, productivity is measured as the slope of the line going through the origin and the
point of production (y/x ≡ output/input ≡ productivity). As Figure 2 illustrates, line OA has a
smaller slope than line OB, demonstrating that Firm B is more efficient than Firm A. However,
line OC has an even larger slope than line OB, illustrating that while Firm B is technically
efficient, Firm C is more productive. In fact, theoretically speaking, a firm that is not efficient
can technically boast a greater productivity level than another firm that is.
While the production frontier defines the maximum output level for any given level of
input, technical changes in the industry can shift the frontier out. This implies an increase in the
maximum output level for all input levels. Technical changes may be spurred by new technology
or improvements in production processes. Graphically, these changes shift the production
frontier outward.
Essentially, if a firm’s productivity increases, it is due to one of three effects: firm-level
efficiency may have improved and drawn closer to the production frontier, a firm may have
exploited scale economies, and/or technical changes may have occurred in the industry.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
13
In the case of a firm that utilizes N number of inputs to produce a single output, the
firm’s technical possibilities can be summarized in a production function:
Q = f(x) (2)
x ≡ (x1, x2,…, xN)’
Q denotes the quantity of output, while x is an N x 1 vector of inputs. There are four properties
that are generally assumed:
1. Non-Negativity: the value of f(x) is real, finite, and non-negative.
2. Weak Essentiality: at least one input is required to produce positive output.
3. Monotonicity: increasing input cannot reduce output. If x0 ≥ x1, then f(x0) ≥ f(x1). If the
function is continuously differentiable, this implies that the marginal product of all inputs
is non-negative.
4. Concave in inputs: any linear combination of vectors x0 and x1 will produce an output
whose amount is greater than or equal to the same linear combination of f(x0) and f(x1).
Mathematically, f[φx0 + (1-φ)x1] ≥ φf(x0) + (1-φ)f(x1). If the production function is
continuously differentiable, concavity implies that all marginal products – the output
produced by the last unit of input – are non-increasing. This is the phenomenon of
diminishing marginal productivity, which is calculated as:
MPn = δf(x) / δxn (3)
Figure 1 graphically illustrated the production function representing a single-input,
single-output firm, where output is plotted on the y-axis and input is plotted on the x-axis. In the
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
14
case of a two-input firm, however, this is not feasible; rather, the two inputs are plotted on either
axis, while the output is held constant. Figure 3 illustrates this:
Figure 3: Isoquants for Two-Input Firm
Coelli, Timothy J, D.S. Prasada Rao, Christopher J. O’Donnel, George E. Battese. 2005. An Introduction to
Efficiency and Productivity Analysis. New York; Springer Science + Business Media, Inc.
In the graph above, the inputs are variable, while the output is fixed. The curves are known as
isoquants, since they represent the different combinations of inputs that produce a fixed quantity
of output. Another important element of the graph is the relationship between inputs: the
marginal rate of technical substitution (MRTS) measures the rate at which the input on the x-axis
must be substituted for the input on the y-axis to keep output unchanged. MRTS is calculated by
dividing the marginal product of the input on the y-axis by the marginal product of the input on
the x-axis, and multiplying the result by -1:
MRTSnm = - MPm / MPn (4)
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
15
The relationship between the outputs and inputs is also important and is captured in the output
elasticity, which is a measure of the change in output quantity associated with changes in one of
the input’s quantity:
En ≡ %Δ in output quantity / %Δ in input quantity = (δf(x) / δxn) *(xn / f(x)) (5)
The relationship between firm output and individual inputs is captured by the inputs’
marginal productivities. Another important relationship is that between output quantity and
simultaneous input scaling. What happens to output, for example, when both inputs are doubled?
The answer to this question determines the scalability of a firm:
If f(kx) < k(f(x)), then the firm experiences decreasing returns to scale (DRS)
If f(kx) = k(f(x)), then the firm experiences constant returns to scale (CRS)
If f(kx) > k(f(x)), then the firm experiences increasing returns to scale (IRS)
In an environment of DRS, doubling inputs leads to a less than doubling of outputs. This may
suggest that a firm is too large, as increased inputs have a diminished effect on output production
(possibly due to overcrowding or less-centralized management). With CRS, doubling inputs
leads to a doubling of outputs, while IRS implies that doubling inputs leads to a greater than
doubling of outputs. A firm experiencing IRS should consider expanding, as growth could, for
example, facilitate the specialization of labor.
Transformation functions can be used to generalize the production function of a firm that
produces multiple outputs:
T(x, q) = 0 (6)
q ≡ (q1, q2, …, qM)’ ≡ M x 1 vector of inputs
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
16
However, economists often circumnavigate transformation functions by one of two ways: either
they aggregate outputs into a single index measure, or they use price information and represent
technology using cost, revenue, and profit functions.
Up to this point, only technical efficiency has been discussed. With price information,
however, and the behavioral assumptions that firms minimize costs or maximize profits, we can
determine allocative efficiency – a firm’s ability to select the mix of inputs that produces a given
amount of output at minimum cost. Combining allocative and technical efficiency generates an
overall measure of economic efficiency.
A major firm objective is to minimize costs. In deciding the mix of inputs it will utilize to
do so, a competitive firm takes input prices as given. The cost minimization problem can be
depicted mathematically:
c(w,q) = min w’x so that T(q,x) = 0 (7) x
w ≡ (w1, w2,…, wN)’ ≡ vector input prices
Thus, w’x ≡ w1x1 + w2x2 + w3x3, which is the total cost faced by the firm. A firm’s cost function
satisfies five properties:
1. Non-negativity: costs cannot be negative.
2. Non-decreasing in w: an increase in input prices cannot decrease costs. Mathematically,
if w0 ≤ w1, then c(w0,q) ≤ c(w1, q).
3. Non-decreasing in q: an increase in output means an increase in costs. Mathematically, if
q0 ≤ q1, then c(w,q0) ≤ c(w, q1).
4. Homogeneity: multiplying all input prices by an amount k > 0 will lead to a k-fold
increase in costs. Mathematically, k[c(w, q)] = c(kw, q), for k > 0.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
17
5. Concave in w: any linear combination of vectors w0 and w1 will produce a cost which is
greater than or equal to the same linear combination of c(w0, q) and c(w1, q).
Mathematically, c[φw0 + (1-φ)w1,q] ≥ φ c(wo, q) + (1-φ) c(w1, q). This implies that input
demand functions cannon slope upwards.
In the case when the cost function is twice continuously differentiable, one method of deriving
the conditional input demand of a firm producing multiple outputs using multiple inputs is
known as Shephard’s Lemma:
xn(w, q) = ∂c(w, q) / ∂wn (8)
Essentially, Shepard’s Lemma states that the conditional demand for an input is equal to the
change in total cost associated with a change in the price of that input. If the cost function is
twice-continuously differentiable and satisfies the five aforementioned properties, then Shepard’s
Lemma shows that a firm’s input demand satisfies the properties of non-negativity, homogeneity,
symmetry (∂xn(w, q) / ∂wm = ∂xm(w, q) / ∂wn), and is non-increasing in w and non-decreasing in
q. Either Shepard’s Lemma or the constrained minimum cost approach can be used to determine
the minimum cost of producing a given output vector q.
Another approach taken by firms is that of maximizing revenues for a given input vector
x. For a multiple-input, multiple-output firm, revenue maximization can be illustrates as:
r(p,x) = max p’q so that T(q, x) = 0 (9) q
p ≡ (p1, p2,…, pN)’ ≡ vector output prices
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
18
Thus, p’x ≡ p1q1 + p2q2 + p3q3, which is the total revenue generated by the firm. A firm’s
revenue function satisfies the properties of non-negativity, homogeneity, convexity in p, and it is
non-decreasing in p and x.
We have looked at how firms decide input demand to minimize costs and output demand
to maximize revenues. In reality, however, firms simultaneously decide on the levels of inputs
and outputs to maximize profits:
π(p,w) = max p’q – w’x so that T(q, x) = 0 (10) q, x
The profit function satisfies the properties of non-negativity, homogeneity, convexity in p and w,
and is non-decreasing in p and non-increasing in w. If the profit function is twice-continuously
differentiable, then Hotelling’s Lemma can be used to derive input demand and output supply:
xn(p, w) = - ∂π(p, w) / ∂wn (11)
qm(p, w) = - ∂π(p, w) / ∂pm (12)
2.2 Data Envelopment Analysis
The two most prominent methods of estimating the production frontier – and hence firm-
level inefficiency – are data envelopment analysis (DEA) and stochastic frontier analysis.
DEA involves the use of mathematical linear programming methods to estimate the
production frontier (Coelli et al. 2005). Graphically, the analysis builds a non-parametric
piecewise surface over firm data. After it has estimated the frontier, DEA determines the
efficiency of individual firms relative to this frontier. While previous authors had used similar
estimation techniques, DEA owes its prominence – and coining – to Charnes, Cooper, and
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
19
Rhodes (1978), who proposed an input-oriented, constant returns to scale (CRS) model. Later,
papers by Fare, Gosskopf, and Logan (1983), and Banker, Charnes, and Cooper (1984)
introduced models assuming variable returns to scale.
Linear programming involves the maximization or minimization of a function, subject to
constraints. These constraints are vital: not only do they limit the domain of this optimization
problem, but their vertices also hold the optimization’s solution (Tchoukalov 2009).
Take the example of an industry with I firms, N inputs, and M outputs. The input matrix
A is made up of column vectors xi, which consist of the various inputs a firm employs for
production. The output matrix B is made up of column vectors ci, which represent the outputs
produced by a firm. Further, I define k as a vector of input weights and w as a vector of output
weights. The linear programming model solves for the values of k and w that maximizes each
firm’s ratio form productivity (Coelli et al. 2005):
max w, k (w’ci / k’xi) (13)
st w’cj / k’xj ≤ 1, j = 1, 2,…,I.
w, k ≥ 0
However, there are infinite solutions to this problem: for any solution (w, k), there are other
solutions in the form of (ρw, ρk). To come up with a single solution, the additional constraint of
k’xi = 1 is added. This yields a new linear programming problem:
max η , γ (η’ci) (14)
st κ’xi = 1
η’cj - κ’xj ≤ 0, j = 1, 2,…,I.
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
20
η , κ ≥ 0
The change in notation – from w to η and from k to κ signifies that this is a different linear
programming problem.
The dual nature of linear programming can then be utilized to simplify the problem.
Through duality, a linear program can be converted into another, unique linear program
(Tchoukalov 2009). In this conversion, the number of constraints becomes the number of
variables, and vice versa. The dual form has fewer constraints (Coelli et al. 2005), thus making
the problem easier to solve:
min ϕ ,µ ϕ (15)
st -ci + Bµ ≥ 0
ϕxi - Aµ ≥ 0
µ≥ 0
where ϕ is a scalar whose value is a firm’s efficiency score and µ is a vector of constraints. A
score of one implies that the firm is on the production frontier.
It is important to note, however, that the above analysis assumes constant returns to scale,
implying that firms are operating at optimal scale. But issues like regulation and imperfect
competition often prevent firms for operating optimally. Thus, the model must be adjusted to
account for production that exhibits variable returns to scale. To do so, the convexity constraint
of I1’µ = 1 is added to the above minimization problem to yield:
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
21
min ϕ ,µ ϕ (16)
st -ci + Bµ ≥ 0
ϕxi - Aµ ≥ 0
I1’µ = 1
µ≥ 0
where I1 is a Ix1 vector of ones. This constraint ensures that firms are only compared to those of
similar size, which was not the case in the CRS model.
2.3 Stochastic Frontier Analysis
Stochastic Frontier Analysis (SFA) is a second method of measuring firm-level
efficiency, introduced simultaneously by Aigner, Lovell, and Schmidt (1977), and by Meeusen
and Van den Broeck (1977) (Coelli et al. 2005). One problem with the frontier estimated under
DEA is it assumes that any deviation from the production frontier is due to technical
inefficiency; it does not account for measurement errors and statistical noise. Adding a random
variable representing statistical noise is a solution to this problem, which produces a stochastic
production function. Without the random variable, the production frontier can be written as:
yi = f(xi, β) * TEi (17)
where yi is the scalar output of Firm I, xi is a vector of inputs, β is a vector of technology
parameters to be estimated, and f(xi, β) is the production function. TEi is Firm I’s technical
efficiency, which is defined as the ratio of observed output over maximum possible output. Its
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
22
value lies between zero and one. A stochastic component accounting for random shocks
independent of the firm or technology can be added, modifying the above equation:
yi = f(xi, β) * TEi * exp {νi} (18)
Although each firm faces different shocks, it is assumed that the shocks are independent and are
described by a common distribution. TEi is assumed to be a stochastic variable with a specific
distribution. It can be rewritten as TEi = exp{-ui}, where ui ≥ 0. Assuming that the production
function f(xi, β) takes the log-linear Cobb-Douglas form, then the above equation can be written
as:
ln yi = β0 + Σ βnlnxni + νi - ui (19) n
where n is the number of inputs, νi is the “noise” component, and ui is the technical efficiency
component.
2.4 Studies on Firm-Level Efficiency
Eller, Hartley, and Medlock (2007) compare the revenue-generating efficiency of
national oil companies (NOCs) and private IOCs. They find that, in most cases, national oil
companies tend to be less efficient due to differences in the structural and institutional features of
a private firm (Eller, Hartley, and Medlock 2007). These differences tend to arise from different
firm objectives. A private firm focuses solely on financial objectives, while a national company
accounts for non-commercial goals – such as maximizing employment and “shifting resource
extraction away from the future towards the present” (Eller, Hartley, Medlock 2007, p. 1) – when
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
23
making decisions. In the petroleum industry, an NOC may also be forced to sell some of its oil
domestically and at subsidized prices. Such non-economic considerations hamper an NOC’s
ability to maximize revenues and thus make it less efficient at generating revenues for a given
level of inputs (labor and reserves).
Eller, Hartley, and Medlock (2007) also discuss principle-agent motivations as reasons
why private firms are more efficient at generating revenues. While managers generally try to
maximize their own income rather than their owners’, some privatization mechanisms limit this
problem. Tradable ownership shares, for example, give owners an incentive to monitor managers
and reduce inefficiency. On the manager’s side, maximizing the firm’s return is important to
maintain a good reputation, increase job security, and increase the value of shares or stock
options the manager may own. From an organizational standpoint, private firms’ decision-
making processes tend to be more decentralized and transparent. These qualities allow private
firms to better focus on strictly financial goals.
The authors collected data on 80 oil companies between 2002 and 2004, and they apply
DEA and SFA. SFA generates the following equation:
ln yn,t= 4.3644
(0.6561)+ 0.4847
(0.0666)* ln L
k ,n,t+ 0.0463
(0.0415)* lnOilRsv
k ,n,t+ 0.1695
(0.0493)* ln NGRsv
k ,n,t
+ 0.3022(0.0307)
* t2003
+ 0.4767(0.0312)
* t2004
+ vn,t! u
n
(20)
where y ≡ revenue, L ≡ labor, OilRsv ≡ oil reserves, NGRsv ≡ natural gas reserves; t ≡ time
effects (prevailing market prices of oil and gas, which are not constant, affect revenue), n refers
to the firm under study, vn ≡ stochastic component assumed to be normally distributed, un ≡ time-
invariant technical efficiency component, and e^(-un) ≡ firm-specific efficiency. The authors
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
24
conclude that major IOCs – BP, Chevron, ConocoPhillips, Exxon Mobil, and Shell – are the
most efficient firms, while NOCs tend to be the least efficient. Results are found in Appendix 3.
Al-Obaidan and Scully (1991) examine the efficiency of state-owned or controlled
enterprises – which they define as organizations that are at least 30% government-owned – in the
petroleum industry relative to private firms. They note that private firms and state-owned
enterprises (SOEs) differ in the function they seek to maximize: owners of private firms allocate
resources to their highest-valued use within their firm, and thus their objective function has a
single maximand (Al-Obaidan and Scully 1991). SOEs, on the other hand, pursue multiple goals,
many of which are mutually exclusive. They differ from government entities in that they are
susceptible to market pressures, but government intervention softens these market forces.
Because SOE ownership cannot be easily transferred like shares of a private firm, it is difficult to
meter an SOE’s performance; this reduces the incentive for management to achieve optimal
results. This, along with government intervention, contributes to a misallocation of resources.
The authors “estimate an Aigner-Chu deterministic frontier function, a maximum
likelihood stochastic frontier function, and a maximum likelihood Gamma frontier function” (Al-
Obaidan and Scully 1991, pp. 237) to examine a firm’s ability to use assets and employees to
produce output. They find that NOCs are only 63% - 65% as technically efficient as private firms
(Eller, Hartley, and Medlock 2007), and they conclude that “state firms could satisfy the demand
for their output with something less than half of their current resource inputs simply by being
converted to private, for profit enterprises” (Al-Obaidan and Scully 1991, pp. 245-246).
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
25
3. Methodology and Empirical Evidence
3.1 Revenue Assumptions
The purpose of the paper is to determine whether nationalizing or privatizing Majnoon
will generate the most net revenues – discounted over time – for the Iraqi government. This
entails calculating the revenues and costs associated with each development strategy. As
previously mentioned, I begin with the assumption that because Iraq can employ IOCs to provide
technical and consulting services, production figures under nationalization are equivalent to
those under a PSA. I first estimate the revenues and costs associated with nationalizing Majnoon
over 40 years, from 2009 to 2049. This length of time is convenient for comparisons involving
PSAs, which often last between 25 and 40 years.
The project begins with a three-year exploration and feasibility period, during which no
production is achieved. Thus, Iraqis will only be able to extract Majnoon oil beginning in 2012. I
estimate that Majnoon will produce only 30% of peak production in 2012, 60% of peak
production 2013, and 90% of peak production in 2014. According to Muttitt, three years of
production are required before a field begins to produce at its peak levels (2005). At that point, a
large field like Majnoon will continue to produce at its peak for 20 years, after which its
production declines exponentially at a rate of three percent per year. Thus, Majnoon will produce
at its peak level of 600,000 bbl/day from 2015 until 2035. After that, its production will decrease
exponentially at rate of three percent until 2049.
To calculate revenue, I need to predict the real price of oil over the next forty years. If the
price of oil is expected to rise at a rate greater than the risk-free real interest rate, then an oil-
producing nation has the incentive to reduce current production: the country expects that
producing oil in the subsequent time period will generate more revenues than producing it now,
selling it, and generating interest from the funds (Book 2 – Economics 2008). If, on the other
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
26
hand, the price of oil is expected to rise at a rate less than the risk-free real interest rate, then a
country has an incentive to produce more oil now, sell it, and invest the funds to generate the
risk-free rate. Based on this Hotelling Valuation Priciple (HPV), the equilibrium real price of oil
is expected to grow at the risk-free real interest rate. It should be noted, however, that this only
takes financial aspects into account. Other strategic, political, and utilitarian factors also affect a
country’s rate of extraction and the price of oil.
I assume that the risk-free real interest rate is 3%, and I use the current price of crude oil
– $56/bbl – as the initial real equilibrium price of oil. Therefore, I predict that the price of oil in
2010 will average $56*(1.03), or $57.68; in 2011, the price will be $56*(1.032), or $59.41, and
so on. Table 1 shows these calculations up to 2049:
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
27
Table 1: HVP Forecast of Real Price of Oil, 2009-2049
Year Real Price of Oil Prediction (2009 $)
2009 56.00
2010 57.68
2011 59.41
2012 61.19
2013 63.03
2014 64.92
2015 66.87
2016 68.87
2017 70.94
2018 73.07
2019 75.26
2020 77.52
2021 79.84
2022 82.24
2023 84.71
2024 87.25
2025 89.86
2026 92.56
2027 95.34
2028 98.20
2029 101.14
2030 104.18
2031 107.30
2032 110.52
2033 113.84
2034 117.25
2035 120.77
2036 124.39
2037 128.12
2038 131.97
2039 135.93
2040 140.00
2041 144.20
2042 148.53
2043 152.99
2044 157.58
2045 162.30
2046 167.17
2047 172.19
2048 177.35
2049 182.67
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
28
3.2 Costs Under Nationalization
There are five major costs associated with oil production, and those are feasibility,
development, fixed, and variable costs, along with the cost of a technical service agreement
(Muttitt 2005). The onset of a project consists of a feasibility study to assess whether a field
contains economically-extractable oil. Muttitt estimates that feasibility costs amount to
approximately $10 million a year for the three years prior to production, which I will assume.
Development costs consist of expenditures associated with developing a production plan and
establishing facilities to optimize production. Deutsche Bank pegs Majnoon’s development costs
at approximately $4 billion. I assume that half of these costs – $2 billion – are incurred up front
(2009), while $1 billion is incurred in 2010. An oil production project typically incurs
development costs until two years after peak production, so I divide the remaining $1 billion
equally from 2011 until 2037. This amounts to an annual cost of approximately $37 million.
Fixed and variable costs are also accounted for in this analysis. Muttitt estimates that
fixed costs generally amount to 5% of development costs, which in Majnoon’s case would
amount to $200 million. Iraqi authorities estimate the variable costs per barrel of oil to be
approximately $1-$1.5. I assume a variable cost of $1/bbl until the field ceases to produce at
peak level, at which point variable costs rise to $1.5/bbl. Multiplying the per-barrel variable cost
by the production amount yields total variable costs (VC).
The final – and most difficult – cost that must be accounted for is the cost of the technical
service agreement. Terms of such agreements are often kept secret and rarely published, thus
making it difficult to estimate their cost. For now, I assume that the technical services provided
by the IOC are free; I will account for this cost later in the analysis.
Net revenues are calculated by subtracting yearly costs from yearly revenues. But the
Iraqi decision-maker is interested in the net present value (NPV) of future net revenues, because
the value of future payments are discounted due to the opportunity cost of delaying payment: the
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
29
revenues could have been invested in other revenue-generating projects. NPV is the discounted
difference between future cash inflows and future cash outflows. It compares the value of a
dollar in different periods, taking into account rates of return. Generally, if the NPV of a project
is positive, a decision-maker should engage in the project. In my analysis, however, I am
comparing the NPV of nationalization to the NPV under a PSA to determine which project
possesses a greater NPV. NPV is calculated as follows:
NPV = ∑ [(Rt-Ct) / (1+r)t (21)
where Rt ≡ revenue at time t, Ct ≡ cost at time t, and r ≡ discount rate. How should a nationalized
oil company discount future cash flows? One method is to simply use the country’s real interest
rate: the Iraqi Central Bank has currently set a nominal interest rate of 11% (Iraq Directory
2009), while inflation has hovered around 5% (Index Mundi 2008), producing a real interest rate
of 6%. However, another method is to calculate the social discount rate; this is the discount rate
applied to social investments, and it measures the rate at which a society is willing to trade
present for future consumption. The social discount rate can be estimated by summing a
country’s population growth rate with its depreciation rate (Rothwell 2009), as this summation
yields the rate at which a country’s economy must grow to sustain itself. Iraq’s population
growth rate in 2009 is 2.5% (CIA 2009). There is little information on the rate of depreciation of
Iraq’s hydrocarbon capital. The depreciation rate of India’s oil and gas pipelines in 2005 was
adjusted to 3.17% (Mukul 2005). After years of destruction and underinvestment, Iraq’s
hydrocarbon infrastructure is sure to be in worse condition than India’s. Thus, I estimate that
Iraq’s hydrocarbon capital depreciation rate is 8.5%. Adding this rate to the population growth
rate yields an estimated social discount rate of 11%. I discount future cash flows from the oil
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
30
field using the social discount rate, as it is tailored to social projects and, in this case, to the oil
industry as well (later I will test this assumption).
Appendix 4 presents all of the revenues and costs associated with nationalization,
discounting future cash of net revenues over the next 40 years from Majnoon. The socially-
discounted cash flows are worth $98.5 billion in today’s terms.
3.3 Development under an Omani-style PSA
Under the terms of the Omani PSA, the IOC pays neither royalties nor taxes to the
government (Muttitt 2005). Cost oil is limited to 40% of production, and profit oil is split 80:20
in favor of the government. The government rewards the IOC with a $3 million bonus for
discovering oil in a field and with a $1 million bonus for every 25,000 bbl/day increase in
production levels, until the IOC produces at a rate of 150,000 bbl/day. For its part, the IOC
compensates the government with a signing bonus of $250,000.
This analysis adopts all of these terms except for the signing and discovery bonuses.
Since the presence of oil has already been confirmed in Majnoon, the hypothetical PSA
employed for Majnoon will not include a discovery bonus. As for the signing bonus, it is not
reflective of the sheer potential of Majnoon’s production. In 2007, Reliance Industries, India’s
largest private-sector conglomerate, paid a signing bonus of between $15.5 - $17.5 million for
the rights to explore and develop two Iraqi Kurdish blocks (Earth Times 2009). I therefore
include a signing bonus of $25 million for the Iraqi government as part of the PSA analysis, to be
paid up front.
Appendix 5 presents the revenues and costs associated with developing Majnoon through
a PSA. The timeline and production levels of the Majnoon field under the PSA are assumed to be
identical to those under a nationalized scheme. This is based on Muttitt’s assumption that, with a
technical service agreement, a national oil company is as efficient as the major international
Developing Iraq’s Oil Industry to Maximize Government Net Revenues
31
companies. More formally, this assumption states that the INOC would be at the production
frontier of the oil industry, achieving the highest level of industry efficiency (an assumption that
will later be relaxed). Therefore, production under the PSA is identical in timing and value to
that under nationalization. With the exception of the additional $25 million in the first year of
project development (2009), government revenues under both scenarios are identical.
Under a PSA, all costs are absorbed by the IOC, which subsequently recuperates them
through cost oil. The feasibility, development, fixed, and variable costs incurred by the IOC are
identical to those incurred by the government under nationalization. The IOC begins to
recuperate cost oil once it achieves production (2012). In this year, it will begin recuperating the
capital costs it has injected into the project, which are illustrated in the table below: