This piece is part of a compilation bringing together Regional Voices on the Challenges of Nuclear Deterrence Stability in Southern Asia.


Asim Bashir Khan
The author is a PhD scholar in the Department of Economics and Finance, Institute of Business Administration (IBA), Karachi. The views expressed are those of the author and should not be construed as necessarily represent the IBA. The usual disclaimer applies.

This study is not an oft-repeated conventional narrative on Pakistan’s nuclear security doctrine; nor does it critique, evaluate, or comment on the design, operations, and expansion of the program. It does not assess the strategic calculus or security doctrine. It is abundantly clear that investing in nonconventional security assets is vital for the modern security paradigm and serves as a peace-preserving investment given the multifaceted threats Pakistan is facing. I do agree with Crockett that “without nuclear weapons, Pakistan looses military parity,” and would thus have an incentive to keep a nuclear-capable military.1 This view is seconded by Krasner and Evans, who say that “nuclear weapons work for Pakistan: They are a deterrent against a much larger and stronger India.”2

This study is just an economist’s attempt to establish the opportunity cost of Pakistan’s nuclear program by estimating the cost of its nonconventional security assets and then computing an estimated opportunity cost in terms of conventional military expenditures and human security. All facts and figures referred are from secondary sources, with complete citation.

It is important to distinguish between accounting and economic costs. Accounting cost is the explicit cost paid, whereas economic cost includes implicit, nonmonetary costs as well. Economic cost includes a valuation of the trade-offs between nonconventional nuclear security and spending that affects conventional security and the state of human development. Economic cost is based on data that are very hard to find and information that is hard to estimate, due to nondisclosure. This groping in the dark, however, can still offer a better understanding than no investigation whatsoever.

The study’s scope is limited to conceptualizing the factors that contribute to the cost of Pakistan’s nuclear program, devising a workable methodology to estimate the running cost of the program, and assessing the program’s opportunity cost expressed in terms of conventional military assets and human security.

Conceptualization of Direct Costs

In the absence of annual budgetary data for Pakistan’s nuclear program and its components, determining its cost is a monumental task. It requires comprehensive knowledge of inputs produced domestically and imported from international markets. Determining the cost requires knowledge of the nature of items procured, the intensity and frequency of their usage, the availability of consumable supplies, cost centers, and the main drivers of the cost. Beyond these inputs, a number of other factors must also be included: supply chain management, procurement, security, repair and maintenance, stockpile maintenance and evaluation, inventory management, the cost of technical obsolescence, the decontamination cost, metallurgical and chemical processes, and project shutdown costs. Further, separating the fixed or capital costs from the operational costs is also necessary for a fair assessment of the running cost of the program. The total cost broadly sorts into delivery system costs and general program costs.3 The major cost components of any nuclear program are the development of new weaponry, the modernization of existing capacity, infrastructure, training, research and development, nuclear conversion, enrichment, exploration and mining, fuel fabrication, weapons testing, disarmament, decommissioning, and delivery systems, along with a number of ancillary activities.

The supply chain of inputs—most important, highly enriched uranium and weapons-grade plutonium—extracted or produced domestically are major components of program cost. For internationally sourced inputs, relevant data are compiled from the International Trade Centre (ITC) database and are coded as nuclear-related imports. It is difficult to establish the cost of capital-intensive projects in per-year terms. One solution is to construct time-series data for capital projects using annual Public Sector Development Program (PSDP) statistics. Another important determinant of unusual cost increase in such capital-intensive programs is the monopoly of original equipment manufacturers, which accounts for abnormal cost and time overruns in refits and overhauling operations.

Estimating the Program’s Direct Costs

Before presenting computations of the direct running cost as calculated using available data, it is very important to survey the literature available on the subject in order to draw fair comparisons and to evaluate or asses the data and methodologies. Thakur and John estimate a total cost of $2.2 billion for Pakistan’s nuclear program for 2011. Out of this number, $1.8 billion is quoted as the core cost of the program.4 The authors note that Pakistan and India do not provide any information on the costs of their nuclear programs. Others who have estimated the cost of Pakistan’s nuclear program note the same paucity of data.5

Mian and Ramana estimated the 2012 cost of the military as $6 billion.6 In fiscal year 2013–2014, this spending was 28 percent of the country’s budget outlay. There is still ambiguity as to whether this includes the nuclear weapons program, which they estimate at $2.5 billion per year. A similar figure has also been reported by Ghosh, who cites a report by the Women’s International League for Peace and Freedom.7 This estimate puts the cost of the nuclear program at 32 percent higher than the $1.9 billion cost of getting 5.1 million more children enrolled in school for three years.8

There is another important study by Mian, which says that “a more recent estimate suggests Pakistan’s nuclear spending could be about $800 million per year as of 2011, and possibly as much as $2 billion per year if health and environmental costs are included—and projected to rise significantly because of Pakistan’s expanding nuclear programme.”9 Mian values the negative externalities of the program at $1.2 billion per year, 67 percent higher than the actual running cost of the program. Without access to better data, however, Mian’s argument is hard to authenticate. Abbasi criticized the apparent cut of budgetary allocations for the nuclear program, arguing that “the strategic organisations of the country should be the top most priority of the government but still they got less than 0.5 percent of the [gross domestic product].”10

The 2011 Pakistan Nuclear Choronology says that “a Pakistani official reveals that Pakistan spent about $3.2 billion on its nuclear weapons and missile programs since the early 1970s. The official says that the country spent 2% of its national budget and 7.8% of its defense budget on developing nuclear weapons and missiles for past 32 years.”11 These statistics are not supported by background data, and are basically rough estimates. It is pertinent to note that the cost of the program is quoted for a period of four decades—from 1970 to 2011. Therefore, it is hard to compare this estimate with the annual running costs estimated by the studies mentioned above.

Blair and Brown call Pakistan’s nuclear program the world fastest-growing arsenal, with 120,000 to 130,000 people directly involved in the production of nuclear warheads and missiles. For 2009, they mention a credible cost estimate of $781 million—$300 million for weapons, and $481 million for missile delivery systems. They point out that the hidden cost of health and environmental degradation from the recent expansion of plutonium factories is also significant. In the absence of a practicable methodology and data on environmental degredation, assigning a numerical value to these externalities is a major hindrance to measuring indirect costs.12 Sud has estimated the cost of the program over a period of twelve years, from 1978 to 1989, at $12 billion, including both capital and revenue expenditures. A breakdown of the costs by component is given in table 1.

Table 1. The Cost of Pakistan’s Nuclear Program, 1978–1989
Classification of Costs Years Amount (billions of dollars)
Training and development 1978–82 1.5
Project management and related costs
(About 12% of installed cost)
1978–89 1.0
Hardware procurement, installation, and commissioning 1982–89 5.0
Direct labor (12,000 scientists and engineers) 1981–89 2.0
Utilities and infrastructure 1982–89 2.0
Other costs, including those absorbed by the army 0.5
Total 1978–89 12.0
Source: Hari Sud, “Cost Pakistan Incurred to Build the Nuclear Bomb,”, April 17, 2004,

The estimates provided by Thakur and Page, Mian and Ramana, Ghosh, Abbasi, Sud, Blair and Brown, and the Pakistan Nuclear Choronology all provide useful information about the program and its cost, but none of these studies ennumerates a clear methodology, components of cost, or the data used in the analysis. Further, the tendency to include negative externalities without a rigorous framework for doing so may lead to overestimates of economic cost. Validation of these estimates is difficult in the absence of appropriate data and knowledge of the authors’ criteria.

Given these data constraints, it is difficult to arrive at an educated estimate of the nuclear program’s running cost. Most of the program’s costs can be disentangled from the budgets of the Pakistan Atomic Energy Commission (PAEC), the Pakistan Nuclear Regulatory Authority (PNRA), and the Space and Upper Atmosphere Research Commission (SUPARCO), as reported by the PSDP.13 The cost of development schemes attributable to these three institutions during fiscal year 2014–2015 was $600.2 million, or 11.13 percent of the total PSDP. This certainly includes some of the “dual use” cost of nuclear infrastructure for civil and military purposes. Alger and Findlay specifically singled out the segregation of such data as prerequisite for cost determination, but the division of data into multiple sources is a problem as well.14 The civil nuclear program’s contribution to national electricity production was 3.5 percent in 2014–2015. If the windfall of this power production is subtracted from the total operational cost of civil nuclear institutions, we may arrive at the net cost without the civilian component of the program.

Alternately, the cost of producing the fissile material produced in an average year may be treated as a fair proxy for the annual running cost of the program. For instance, the annual production of fissile material is 150 kilograms of highly enriched uranium and 6 to 12 kilograms of weapons-grade plutonium.15 The cost of this fissile material and the cost of building and maintaining an average of ten new weapons per year can be set as a parameter for the annual running cost.

Nuclear operations in Pakistan are controlled by the Strategic Plans Divisions (SPD). The SPD directly commands the National Engineering and Scientific Commission (NESCOM), SUPARCO, the PAEC, and the uranium enrichment facility at the Kahuta Research Laboratories (KRL).16 Abdul Qadeer Khan reportedly claimed that in 1990, the KRL’s annual budget was about $20 million to $25 million—which, in 2015 dollars, would be between $240 million and $300 million. Further, he added that during the second term of Benazir Bhutto as prime minister of Pakistan (1993–1996), a deal to purchase long-range missiles to ensure deterrence against far-flung areas of India was finalized with North Korea at a cost of less than $50 million ($371 million in 2015 dollars).17 Contrarily, the 2011 Pakistan Nuclear Choronology claimed a payment of $200 million to North Korea for this technology used in the Ghauri ballistic missile. According to Chakma, during the period 1960–1968, the actual spending on the program was $67.5 million ($6.5 billion in 2015 dollars), in addition to a budgetary allocation of $83.3 million ($8 billion in 2015 dollars) for the Karachi Nuclear Power Plant.18 Chaudhri writes that an initial budget approval was sought from Prime Minister Zulfiqar Ali Bhutto for nuclear weapons on February 15, 1975, for uranium refining and conversions, building a production complex at Dera Ghazi Khan, building a centrifuge plant at Kahuta, and starting a weapons design program in PAEC. The total cost of the request was $450 million ($17.2 billion in 2015 dollars).19 The simple comparison of these statistics presented by Chakma and Chaudhri suggests that Khan’s estimates of the KRL budget deserve some skepticism.

An educated and fair estimate of the running cost should be based on the sum of traceable costs for a given year. An important qualification is to bear in mind that these calculations may include dual-use technology, the cost of which should be split between the nuclear program and civil nuclear energy. The cost of imports retrieved from the ITC database is believed to represent the KRL’s expenditures for such items as nuclear boilers and centrifuge-related imports. All the same, it is very hard to confirm.

The actual expenditures by key institutions for 2014–2015 totaled $600.2 million, plus the nuclear-related imports from the ITC database valued at $171.7 million, and human resource expenditures for the 120,000 personnel directly linked to the program, valued at $380.8 million.20 This sums to $1.1 billion, including the cost of the civil use of nuclear energy.

The estimated running cost of $1.1 billion (inclusive of cost of civil use) for 2014–2015 is close to Mian’s estimate of $800 million for 2011, excluding the environmental costs.21 The twelve-year cost estimates by Sud put the average per-year cost at $1 billion, but that is not strictly comparable, on two accounts. First, it involves both capital and current expenditures; second, it measures expenditures from 1978 to 1989, which are difficult to compare with more recent figures.22

Using a crude methodology, we have computed the running cost of the program, but this computation does not include the cost of negative externalities and hidden costs, which are difficult to value in monetary terms. Any rough guess could overstate the cost. The exact cost with the least possible margin of error is hard to estimate, given the secrecy that surrounds the program. Without comprehensive information about how expenditures correspond to the actual cost when technical and material support and the role of foreign aid are all accounted for, the reliability of estimates will remain low.

The Opportunity Cost of Nonconventional Security

Making a fair opportunity cost assessment is a challenge in the absence of reliable estimates of the direct running cost of the program. The cost estimate of $1.1 billion, based on the methodology presented above, closely reflects the running costs of the key institutions involved in the program. The cost of Pakistan’s conventional military is at least six times higher than the above-estimated cost of the nuclear program. In terms of conventional military equipment, the nuclear program’s cost is roughly equal to the cost of 73 JF-17 fighter aircraft,23 or to the cost of 3 or 4 Type-039/041 submarines.24 However, the nuclear program’s cost can be assessed in terms of opportunity cost in a number of other ways.

Pakistan’s poor human development indicators are a result of the state’s neglect of social and economic imperatives. Successive governments have ignored the necessity of human security, even after witnessing the collapse of the Soviet Union, which had no shortage of nuclear weapons.

The causes of Pakistan’s underperformance on education and health are multidimensional, but one of the major determinants is fiscal stress imposed by defense expenditures and debt servicing. Pakistan’s development expenditures as a proportion of the state’s budget have gradually declined, from 40.4 percent in 1975 to 20.3 percent in 2014. In the same year, 2011, that military spending was increased by $555.6 million, spending on socioeconomic development was reduced by $1.1 billion.25

For 2014–2015, defense and debt servicing cumulatively account for 42.5 percent of the federal government’s total expenditures. Nuclear spending, estimated at $1.1 billion, amounts to 2.6 percent of total expenditures. Pakistan spends 6.2 percent of its gross domestic product (GDP) on all its defense requirements, including about 1 percent of GDP on its nuclear program. This is twice as much as Abassi estimates. His prediction about budget cuts is correct, however. For the year 2015–2016, the PAEC’s budget has been cut by 48.7 percent, compared with the previous fiscal year.26

Statistics paint a bleak picture with regard to Pakistan’s investment in human development. Pakistan’s Human Development Index rank is 146 out of 185 countries. Page and Thakur, in a comparative analysis of nuclear-armed states, mention that Pakistan’s public spending on education and health is 2.4 percent and 0.8 percent of GDP, respectively, which is the lowest of all nuclear-armed states.27 According to the 2014–2015 Pakistan Economic Survey, education and health expenditures as a proportion of GDP have decreased further, to 2.1 and 0.4 percent of GDP, respectively.

The estimated cost of the nuclear program is half the size of the budget allocation for the federal national plan to enroll 5.5 million out-of-school children in government schools.28 This estimated cost is slightly more per year than the total annual budget of the Benazir Income Support Program, which supports 5 million beneficiaries.29


A huge cost is associated with investment in nonconventional security assets. For Pakistan, however, its nuclear program is an investment in a public good, insofar as it prevents war with India. This study has simply tried to estimate the cost of operating the program that constitutes this public good. In the absence of data on the program, we have devised a methodology for cost assessment. However, due to a number of data constraints, it is still hard to estimate the program’s running cost with any certainty. Many scholars have attempted to come up with estimates, but none of them can be validated because they have not adequately enumerated the assumptions they have made in the course of doing their estimates.

We have first tabulated the actual expenditures incurred during fiscal year 2014–2015 by various development schemes run by the key institutions of Pakistan’s nuclear program: the Pakistan Atomic Energy Commission, the Pakistan Nuclear Regulatory Authority, and the Space and Upper Atmosphere Research Commission. We then added the annual nuclear-related imports from the International Trade Centre database, as well as the human resource cost associated with the personnel working on the program. These components, which include the cost of the civil use of nuclear energy, sum to $1.1 billion, which is broadly in line with and validates Blair and Brown’s estimate of $781 million for 2009 and Mian’s estimate of $800 million for 2012. Other cost estimates are exaggerated; those done by Page and Thakur in 2013 are exactly double the estimates of this study, as are a 2014 estimate by Mian and Ramana and a 2012 estimate by Ghosh.30 Overall, the cost estimates for Pakistan’s nuclear program reported by these authors are not only exaggerated but also lack supporting data. The actual cost is far less than they say.

World Development Indicators data suggested that, as of 2010, 8.3 percent of Pakistanis lived below the poverty line of $1.90 a day. The 2007 Fourth National Report for the Convention on Nuclear Safety, which was published by two Pakistani nuclear agencies, clearly emphasized that Pakistan’s future challenges are food; clothing; education; environmental degradation and its effects on health, irrigation, and drinking water; and creating means of employment for the country’s very large and growing population.31 Although military security is no doubt essential for Pakistan, it is high time for the state to assign a high priority to investing in human capital, lest the country’s miserable state of human development continue indefinitely.

The author is a PhD scholar in the Department of Economics and Finance, Institute of Business Administration (IBA), Karachi. The views expressed are those of the author and should not be construed as necessarily represent the IBA. The usual disclaimer applies.


1 Robert Crockett, “National Security Implications of Eliminating Nuclear Weapons,” in Project on Nuclear Issues: A Collection of Papers from the 2013 Conference Series,ed. Sarah Minot (Washington, DC: Center for Strategic and International Studies, 2014).

2 Stephen D. Krasner and Alexander Evans, “Tough Talk Is Cheap: Washington’s Real Options in Islamabad,” Foreign Affairs 91, no. 3 (2012): 166–70.

3 According to Kristensen and Norris, Pakistan’s delivery system is based on 36 aircraft (F-16 A/B, Mirage III/V), 86 land-based ballistic missiles—Abdali (Hatf-2), Ghaznavi (Hatf-3), Shaheen-1 (Hatf-4), Shaheen-IA (Hatf-4), Shaheen-2 (Hatf-6), Shaheeh-3 (Hatf-?), Ghauri (Hatf-5), and NASR (Hatf-9)—and eight cruise missiles, Babur (Hatf-7), and Raa’ad (Hatf-8).
Hans M. Kristensen and Robert S. Norris, “Pakistani Nuclear Forces, 2015,” Bulletin of the Atomic Scientists 71 (2015): 77–83.

4 Ramesh Thakur and John Page, Nuclear Weapons: The Opportunity Cost (Zurich: International Relations and Security Network, 2014),

5 Justin Alger and Trevor Findlay, The Costs of Nuclear Disarmament (Canberra: International Commission on Nuclear Non-Proliferation and Disarmament, 2009).

6 Zia Mian and M. V. Ramana, “Asian War Machines,” Critical Asian Studies 46,no. 2 (2014): 345–60.

7 Palash Ghosh, “Pakistan Rapidly Expanding Nuclear Weapons Arsenal,” International Business Times, November 5, 2012,

8 Ministry of Finance, Pakistan Economic Survey 2014–15 (Islamabad: Ministry of Finance of the Government of Pakistan, 2015),

9 Zia Mian, “Pakistan,” in Assuring Destruction Forever: Nuclear Weapon Modernization Around the World, ed. Ray Acheson (New York: Reaching Critical Will, 2012).

10 Ansar Abbasi, “Pak Nuclear Programme Faces 35 pc Cut,” News, May 1, 2009,

11 Center for Non-Proliferation Studies, Pakistan Nuclear Choronology (Monterey, CA: Center for Non-Proliferation Studies at the Monetary Institute of International Studies, 2011).

12 Bruce G. Blair and Matthew A. Brown, Nuclear Weapons Cost Study 2011 (Washington, DC: Global Zero, 2011),

13 PAEC’s budgetary allocation for 2015–2016 has been reduced by $288.7 million, a cut of 48.7 percent compared with the actual expenditures for 2014–2015.

14 Alger and Findlay, Costs of Nuclear Disarmament.

15 Mian, “Pakistan.”

16 NTI Nuclear Security Index, Pakistan Country Profile (Washington, DC: NTI Nuclear Security Index, 2015),

17 Abdul Qadeer Khan, interview by Nadim Malik, Aaj News Television, August 31, 2009.

18 Bhumitra Chakma, “Road to Chagai: Pakistan’s Nuclear Programme, Its Sources and Motivations,” Modern Asian Studies 36, no. 4 (2002): 871–912.

19 M. A. Chaudhri, “Pakistan’s Nuclear History: Separating Myth from Reality,” Defence Journal 9, no. 10 (2006): 14–50.

20 The breakdown of these imports in 2014 includes $4,000 for HS Code 2845-Isotopes and their compounds; $1.8 million for HS Code 2844-Radioactive chemical elements, compounds, mixtures, and their residue; $44.7 million for HS Code 8401-Nuclear reactor and fuel element for reactor; and $125.1 million for HS Code 8421-Centrifuge, including centrifuge dyers, filtering and purification machinery. Employee-related expenditures are calculated by multiplying the $3.01 billion personnel total for the armed forces by the fraction of personnel estimated as part of the nuclear program, as given by Blair and Brown, Nuclear Weapons Cost Study 2011.

21 Mian, “Pakistan.”

22 Hari Sud, “Cost Pakistan Incurred to Build the Nuclear Bomb,”, April 17, 2004,

23 Each JF 17 costs $ 15 million per unit. “Pakistan Air Force (PAF) Inducts First Squadron of JF-17 Thunder Jet Aircraft,”Scholar Warrior 1, no. 2 (Spring 2011): 104–5.

24 Usman Ansari, “Pakistan to Buy 8 Submarines from China,” Defence News, April 3, 2015, china/25233481/.

25 Thakur and Page, Nuclear Weapons.

26 Abbasi, “Pak Nuclear Programme.”

27 Thakur and Page, Nuclear Weapons.

28 Ministry of Education, Trainings, and Standards in Higher Education, National Plan of Action 2013–16: Achieving Universal Primary Education in Pakistan (Islamabad: Government of Pakistan, 2013),

29 The Benazir Income Support Program is a social protection program launched in July 2008 to enhance the financial capacity of poor people and their dependent family members through the provision of cash transfers of Rs. 1,500 a month.

30 Thakur and Page, Nuclear Weapons; Mian and Ramana, “Asian War Machines”; Ghosh, “Pakistan Rapidly Expanding Nuclear Weapons Arsenal.”

31 Pakistan Nuclear Regulatory Authority and Pakistan Atomic Energy Commission, Fourth National Report for the Convention on Nuclear Safety (Islamabad: Pakistan Nuclear Regulatory Authority and Pakistan Atomic Energy Commission, 2007).