Western Australia's Domestic Gas Reservation Policy: The Elemental Economics

Authors


  • The author would like to acknowledge Rodney Tyers, Peter Hartley, David Prentice and anonymous peer reviewers for valuable comments and assistance. Thanks are also due to participants at the University of Western Australia's economics seminar. All errors are the authors’ own.

Abstract

Western Australia's domestic gas reservation policy nominally requires new gas export projects to supply the equivalent of 15 per cent of their exports to the domestic gas market. This export restriction suppresses the domestic price below the export netback price. A theoretical model shows that a binding reservation policy always causes a deadweight loss, and clarifies the source of this loss. Western Australia's reservation policy is estimated to cause losses with a present value of between $6.9 and $22.9 billion, depending on export netback prices.

1 Introduction

Under Western Australia's (WA) domestic gas reservation policy, the State Government nominally1 requires new gas developments to supply the equivalent of 15 per cent of their Liquefied Natural Gas (LNG) exports to the WA domestic gas market (Department of State Development, 2016). The policy is primarily enforced through the approvals process for LNG plants and infrastructure.2 It formalises domestic supply provisions contained in previous “State Agreements” between producers and the State Government.

Specifically, LNG producers are required to reserve a certain proportion of gas for sale to the WA market. They must ensure that “domestic gas is made available to coincide with the start of LNG production” and “show diligence and good faith in marketing gas to the domestic market” (Department of Finance – Public Utilities Office, 2012, p. 14). The price and terms of sale of this domestic gas are determined on the market.

Several studies, including by the Bureau of Resources and Energy Economics (BREE, 2014) and Economic Regulation Authority (ERA) (2014), have discussed the economic impacts of domestic gas reservation policies. These, generally qualitative, studies find that gas reservation likely has a negative economic impact. They equate the policy to a simultaneous tax on gas producers and subsidy to domestic gas consumers. This paper instead analyses the policy as an explicit restriction on the quantity of exports. It formally investigates the policy's effect on total gas production and domestic supply, and clarifies the source of the deadweight loss. The analysis also shows that the implicit production tax is equal to the domestic price discount multiplied by the proportion of total gas reserved for the domestic market. This paper is also the first to estimate the size of the loss caused by the WA policy, although it has formally been in place since 2006.

2 The WA Gas Reservation Policy

This section gives an overview of how WA's domestic gas reservation policy has been applied and considers whether it is binding.

2.1 WA's Gas Market

Until recently, eight projects produced gas in WA. Two of these supply LNG exports, while seven supply the domestic market.

The North West Shelf project (NWS), which supplies most of the state's gas, is the only project currently supplying both the domestic and export markets. Long-term contracts between the NWS and the WA Government supported expansion of the state's gas industry from its infancy by underwriting the project's development in the 1980s. However, now that the industry is mature, the effects of the domestic gas reservation policy should be assessed independently.

Domestic gas consumption in WA in 2014 was 366 PJ (Department of Mines and Petroleum, 2014), compared to exports of 1156 PJ. Therefore, domestic consumption was around 32 per cent of LNG exports, as shown in Figure 1.

Figure 1.

Domestic Consumption Relative to LNG Exports, PJ pa, Western Australia

Sources: (a) Department of Mines and Petroleum (2014), 2015 LNG volumes based on Woodside quarterly reports (b) IMO (2015).

  • Note: Data excludes gas used in LNG processing.

The Independent Market Operator (IMO) (2015) forecasts very slow growth in domestic demand from 2015, at an annual average between 0.5 and 1.5 per cent. On the other hand, four new LNG projects currently under construction will significantly expand exports over coming years. Thus, domestic gas consumption is expected to fall to around 15 per cent of LNG exports by 2018, and continue to slowly decline thereafter.

Two gas projects face a full reservation requirement of 15 per cent of LNG volumes. The Wheatstone domestic plant, which is scheduled to commence operations in 2018, will have capacity to supply seventy-three PJ per year (Chevron 2016). The NWS has also agreed to supply domestic gas equivalent to 15 per cent of its LNG exports from new expansions at Persephone and Greater Western Flank Two (Barnett, 2014).

Some projects have partial commitments. The Gorgon project agreed to supply 2000 PJ to the domestic market under a State Agreement. Its domestic gas plant will have capacity to supply 55 PJ per year for six years from 2016, expanding to 110 PJ per year thereafter (IMO, 2013). This amounts to 6 and then 12 per cent of the initial annual LNG capacity of the project. The deferred Browse FLNG project has gas reserves that are only partly in State waters, reducing its reservation requirement to around 10 per cent of LNG export volumes (Barnett, 2015). Finally, although Pluto has agreed to supply domestic gas equivalent to 15 per cent of LNG exports, “it remains highly contingent on the commercial viability of a domestic gas plant.” (Department of Mines and Petroleum, 2015, p. 40)

Two other export projects currently under construction have avoided the reservation policy, because their production facilities and gas reserves are located outside WA. Prelude will utilise Floating LNG technology, so that the gas is liquefied outside state borders. Ichthys will transport gas for liquefaction to Darwin via “one of the longest subsea pipelines ever built” (Inpex 2015).

2.2 The Reservation Policy Has Been Binding

There are a number of indicators that the domestic gas reservation policy has compelled gas producers to supply greater volumes of domestic gas than they otherwise would have, and is therefore likely to have been binding.

Firstly, although the industry-wide data presented above shows that domestic gas production will be around 15 per cent of LNG volumes after 2018, this does not imply that the reservation policy is not binding, because the forecasts include the additional domestic gas supply required by the policy. Furthermore, reservation requirements apply to new projects, rather than the industry as a whole. The policy's impact therefore depends on how it affects decisions made by each project.

Secondly, there are signs that the domestic gas market is currently facing overcapacity. According to IMO (2015) forecasts presented in Figure 2, domestic gas production capacity significantly exceeds domestic demand. The construction of the Gorgon and Wheatstone facilities under the reservation policy will raise domestic production capacity to around double consumption volumes by 2020.3

Figure 2.

Domestic Gas Capacity and Demand in Western Australia (TJ per day)

Source: IMO (2015).

Compounding overcapacity on the supply side of the market, growth in domestic gas demand is anticipated to be slow, as discussed in the previous section.

Finally, constructing a new gas project involves investing in a large, long-lived asset. When making investment decisions, producers must therefore form expectations about the anticipated trajectory of export and domestic prices many years into the future. Thus, one cannot judge whether the policy is likely to bind solely by looking at export netback versus domestic market prices in any one year. LNG prices in particular can fluctuate significantly, depending on global market conditions. Indeed, the decision to invest in the Gorgon and Wheatstone plants was made in an environment of significantly higher LNG prices than today. It is therefore likely that these projects would have preferred to make all of their reserves available to the LNG export market, rather than the domestic market.

In summary, the domestic gas reservation requirements applied to date are likely to have altered producer decisions, compelling the construction of domestic gas plants and diverting gas from the export to the domestic market. Three factors indicate that new LNG project developers in WA may not have allocated a significant proportion of reserves to the domestic market in the absence of the reservation policy. These are: overcapacity on the supply side of the domestic gas market; static or slow-growing domestic demand; and long-term expectations for LNG netback prices that exceed domestic prices.

3 Economics of Gas Reservation

Several recent studies from WA and the eastern states have discussed the economic impact of domestic gas reservation policies. Most studies have opposed reservation policies, claiming that the costs incurred in the upstream gas sector outweigh any benefits to domestic gas users. Some contrary studies have supported gas reservation, focussing on the benefits to gas-using industries. However, according to critics such as ACIL Allen (2014), these studies have significant flaws including disputed assumptions.

Most studies are qualitative in nature, pointing out that reservation policies are equivalent to an “implicit tax on gas producers that, rather than going to the government, provides domestic gas users with a price subsidy.” (BREE 2014, p. 112). Examples include studies by ACIL Allen (2014); Department of Industry and Bureau of Resources and Energy Economics (2014); BREE (2014); the ERA (2014); Simshauser and Nelson (2015a,b) and Wood et al. (2013). These studies find that domestic gas reservation “is likely to see a reduction in economic welfare if Australia foregoes export earnings (and tax revenues) in favour of (presumably lower value) domestic production, and lower future exploration and gas development activity.” (Department of Industry and Bureau of Resources and Energy Economics 2014, p. 107) The ERA goes further and recommends that the domestic gas reservation policy be rescinded as soon as possible (ERA, 2014, p. 383).4

The qualitative argument that reservation policies cause an economic loss can be summarised as follows.

Firstly, domestic gas reservation policies act as an implicit tax on Australian gas production. Additional domestic supply reduces the domestic price compared to the price that would prevail in the absence of the policy. This reduces firms’ ability to cover current costs or generate a return on investments in new capacity. These reduced incentives for production, investment and exploration impose economic losses.

Secondly, the subsidy for domestic consumers implicit in the domestic gas reservation policy also imposes economic losses. Artificially low gas prices reduce production costs in gas-using industries, enabling them to expand production. Thus, proponents of domestic gas reservation argue that diverting gas from the export market to the domestic economy generates additional value-added. However, while these domestic industries contribute to value added on average, this is not true at the margin. The value that additional gas contributes to the output of domestic industries is equal to the price that they are willing to pay for it (as embodied in the demand curve for domestic gas). In the absence of the reservation policy, producers would supply domestic gas whenever domestic users are willing to pay a price equal to or higher than the export netback price – that is, whenever the marginal value of gas is at least as high in the domestic economy as on the export market. However, under a reservation policy, the domestic price is below the export netback price. Thus, at the margin, the value of gas to the domestic economy is lower than its value on the export market. In other words, the reservation policy diverts gas to lower value uses, and results in an opportunity cost for the Australian economy.

Importantly, an expansion in domestic gas-using industries does not necessarily imply a long-run increase in overall employment or capital income for Australians. The labour and capital diverted to gas-using industries could otherwise have been employed in industries where no subsidy is necessary, that is, where they have a higher marginal value.5

Few papers explicitly estimate the loss from reservation policies, and those that do focus on the eastern states.

Deloitte Access Economics (2013) models the introduction of a reservation policy in the eastern states as a combined tax and subsidy regime. They find that the policy makes Australians worse off. Specifically, “every one per cent of future gas exports which is artificially re-directed towards the domestic market reduces GDP by an estimated $150 million at 2025” (Deloitte Access Economics, 2013, p. iii).

The Productivity Commission develops a detailed project-by-project model of the eastern gas market and also finds that the “cost to the community of diverting the gas from the export market to the eastern Australian gas market would outweigh any gains to domestic users, which are of themselves far from guaranteed” (Productivity Commission, 2015, p. 21). The Commission estimates that a 25 per cent reservation policy in the eastern Australian gas market could have a total welfare cost between $2.2b and $23.9b (net present value) depending on LNG prices. While the study discusses the sources of this loss, it does not provide a formal analysis of the policy.

The eastern gas market also faces different issues to the western market. For example, there is a risk that a reservation policy could be applied retrospectively, carrying additional economic costs, as discussed in Simshauser and Nelson (2015a). In addition, supply has been restricted by moratoriums on coal seam gas in NSW and Victoria. The political economy surrounding coal seam gas developments may partly explain why past research on reservation policies has focused on the eastern states.

4 Modelling the Domestic Gas Reservation Policy

This section develops a partial equilibrium model of the gas sector and uses it to identify the effects of a domestic gas reservation policy, including clarifying the source of the deadweight loss.

To identify the ongoing economic impact of the policy, this paper focuses on the long-run market equilibrium, considering a time frame over which gas producers can expand or contract capacity. As discussed in Section 'Reservation “Flexibility”', short-term dynamics can introduce additional considerations, but these are unlikely to affect conclusions based on the long-run modelling undertaken here.

4.1 The Undistorted Market

The market for LNG is increasingly global, and LNG competes with pipeline gas in many importing countries. By 2020, WA will have around 15 per cent of global LNG export capacity, based on International Energy Agency (2015) forecasts. Thus, WA producers are unlikely to significantly influence the international LNG price, and the export price is assumed to be exogenous. The relevant export price, PW, is the LNG netback price – the international LNG price less the additional liquefaction, shipping and regasification costs associated with exports.

The price that domestic consumers are willing to pay, PD, is lower the greater the quantity consumed, D. Thus the domestic demand curve is downward-sloping. Using the simplest functional form available (linear), the demand curve is assumed to be:

display math(1)

Here, b represents the “choke” price at which domestic demand is zero, or the price of the most expensive substitute for gas. The sensitivity of domestic demand to price is measured by m, with demand more elastic the larger is m (in absolute value).

Turning to supply, Figure 3 shows the estimated long-run marginal cost of WA gas production.6 This includes the fixed and variable costs of gas extraction and (domestic) processing, but excludes the cost of liquefaction, shipping and regasification. The cost of production is estimated separately for each existing and potential project, and ordered from lowest cost to highest cost.

Figure 3.

Long-Run Marginal Cost of Gas Production, Western Australia

Source: Author's calculations based on IMO (2013), BREE (2015), Department of Mines and Petroleum (2015) and company websites.

This long-run marginal cost curve is the industry's supply curve. To increase total supply, S, producers must move into gas fields that are increasingly expensive to develop, and thus require a higher producer price, PP, in order to do so.

The shape of the marginal cost curve in Figure 3 can be reasonably approximated using a linear function, which also has the advantage of being straightforward to work with. The supply curve is therefore:

display math(2)

Here, c represents the long-run marginal cost of the least-costly reserves in the state. The parameter n measures how rapidly costs escalate as output increases. In this paper, it is assumed that the domestic market is competitive and producers are price-takers. In a subsequent paper, non-competitive models of the domestic gas market are considered.

The total cost of production for the industry is the area under the supply curve between zero and the level of production, S:

display math(3)

Producers choose total supply, S, and the proportion of this supply, α, to deliver to the domestic market in order to maximise profits. Explicitly modelling the choice of α enables modelling of the reservation policy. Formally, the producer's profit maximisation problem is:

display math(4)

Solving the first-order conditions gives the familiar result that the representative producer chooses α to equate the domestic price and the export netback price: PD = PW. The resulting profit-maximising total supply and domestic market share are:

display math(5)
display math(6)

4.2 Impact of Gas Reservation

A binding reservation policy sets a minimum proportion of production7 to be supplied on the domestic market, math formula, where math formula. For any total supply, S, domestic supply, D, must be at least math formula, implying the following domestic supply curve.

display math(7)
Figure 4.

Domestic Supply Curve under the Reservation Policy

The undistorted equilibrium, D*, S* and PW = PD, discussed in section 4.1, is depicted in Figure 4. The figure also shows the domestic supply curve imposed by the reservation policy, from equation 7.

Under a binding reservation policy, the representative producer can no longer choose the proportion supplied to the domestic market, and can only choose total supply. Thus, the profit maximisation problem becomes:

display math(8)

The profit-maximising rule for total supply under the domestic gas reservation policy is therefore:

display math(9)

Producers choose total supply to equate the marginal cost of production with the weighted average of the world and domestic prices. Under a binding reservation policy, a proportion math formula of any additional production must always be sold on the domestic market so the marginal price becomes this weighted average price.

Figure 5 represents the market outcome under the domestic gas reservation policy, including domestic supply, math formula, and total supply, math formula. It can be shown that, for any math formula, the domestic price will be lower than the international price.8 Figure 5 also shows that a domestic gas reservation policy reduces both total production and exports, but increases domestic supply compared to the undistorted equilibrium.

Figure 5.

Gas Market under the Domestic Reservation Policy

This model of a domestic gas reservation policy is consistent with interpreting it as a simultaneous tax on production and subsidy for domestic consumption. The production tax rate, t, is the per cent difference between the world price and the producer price (the weighted average of domestic and export prices). The consumption subsidy rate, s, is the per cent difference between the world and domestic consumer prices. Figure 5 shows these rates.

It can also be shown9 that the tax and subsidy rates are proportional, as follows:

display math(10)

Thus, whenever the reservation policy imposes a domestic price discount compared to the export price, the effective tax imposed on producers is the proportion math formula of this discount. Thus, whenever there is a loss from the tax aspect of the policy, there will also be a loss from the subsidy aspect, and vice versa. This is a general result that relies only on the fact that domestic supply must be proportional to total supply, and is independent of the shape of the demand and supply functions.

The deadweight loss resulting from the domestic gas reservation policy is the difference between the loss in producer surplus and gain in consumer surplus10 :

display math(11)

A binding reservation policy, that is math formula, always produces a deadweight loss. The loss is the shaded areas in Figure 5. The triangle on the right is the loss from reduced total gas supply. The triangle on the left is the loss associated with the opportunity cost of diverting gas from the export to the domestic market.

4.3 Numerical Results for the WA Natural Gas Market

The IMO (2015) expects domestic gas consumption in 2016 to be 393 PJ and the price for new domestic contracts to be $5.68 per GJ in real (2015) terms. The price of diesel, a substitute for gas, was $15.89 per GJ at the end of 2015 (IMO, 2015). Using this “choke” price together with the 2016 data, the demand curve from Equation (1) can be characterised as:

display math(12)

The consumer price, PD, is measured in AUD per GJ, and WA demand, D, is measured in PJ per year. This function implies that the price elasticity of demand at 2016 consumption levels is approximately −0.56.11

The supply curve can be estimated using the long-run marginal cost data presented in Figure 3. Fitting a linear approximation gives the following supply curve:

display math(13)

Again, the producer price, PP, is measured in AUD per GJ, and total supply, S, is measured in PJ per year.12

Gas producers’ expectations for long-run LNG prices drive investment decisions, and therefore affect the size of the deadweight loss from the reservation policy.

Figure 6 shows World Bank data on the Japanese LNG import price, from January 1977 to September 2016. This includes cargos under both short and long-term contracts, and has been converted into Australian dollars per GJ, in real 2016 terms. Of course, prices have fluctuated, averaging AUD 15.29 per GJ between 2010 and 2014, when many recent Australian projects were under construction, but by September 2016 the average Japanese import price had fallen AUD 9.13 per GJ. Based on these data, the average real Japanese import price over the whole period of AUD 13.02 per GJ is probably a reasonable estimate of long-term anticipated average prices.

Figure 6.

Japanese LNG Import Prices

Sources: Nominal USD/MmBTU: World Bank (2016).Real AUD/GJ: Author's calculations based on World Bank (2016), RBA (2016) ABS (2016). Real prices are in 2016 terms.

From a given export price, the implied LNG netback can be derived by subtracting transport and liquefaction costs. Platts (November 2015) reports transport costs of AUD 0.77 per GJ (USD 0.55 per MmBTU). Liquefaction costs, including both capital and operating costs, are around AUD 4.31.13 Thus, the 1977–2016 average Japanese LNG price implies an export netback of AUD 7.94 per GJ over the same period.

This LNG netback price together with the supply and demand curves, imply long-run equilibrium production levels of 4460 PJ per year in the absence of the domestic gas reservation policy, which is similar to the total gas supply in the IMO's most optimistic forecasts for 2025 (IMO, 2015). Without reservation, 7 per cent of total production would be supplied to the domestic market, or 306 PJ per year. In terms of the model developed in the previous section, in an undistorted market, the netback price PW = PD is $7.94, total supply, S*, is estimated at 4460 PJ per year and domestic supply, D*, is 306 PJ per year.

Although the reservation policy states that 15 per cent of LNG production must be reserved for the domestic market, a number of projects have lower requirements, or have been able to completely circumvent the policy, as discussed in Section 'WA's Gas Market'. Based on IMO projections shown in Figure 1, around 13 per cent of LNG volumes, or 11.5 per cent of total supply, are likely to be supplied to the domestic market under the reservation policy.

With a netback price of $7.94, the reservation policy reduces estimated equilibrium production levels, math formula, to 3972 PJ per year. Domestic supply is greater, with math formula at 459 PJ per year, and is sold at a lower domestic price, PD, of $3.98 per GJ – an implicit subsidy of 50 per cent compared to the $7.94 netback. The deadweight loss from lower production levels and foregone export earnings is therefore estimated to be around $410 million each year, as shown in Figure 7. This loss can be broken into its two components. The loss caused by the implicit tax on gas production is estimated to be $110 million each year, while the loss related to the implicit consumption subsidy is $300 million.

Figure 7.

Deadweight Loss for Reservation Policy (13 per cent of LNG)

Source: Author's calculations.

A higher LNG netback price would raise the cost of restricting exports. An LNG netback price of $9 per GJ raises the estimated annual deadweight loss to $1.37 billion.

Importantly, there is never a positive economic outcome from domestic gas reservation. Even at very low LNG prices, the policy has zero impact, rather than a positive one. More specifically, if the long-run expected LNG netback price were to be $6.6 per GJ or lower, producers are estimated to be willing to supply greater than the required proportion of gas to the domestic market. This is because, at lower prices, total production and exports are also lower than otherwise. Thus, at sufficiently low LNG prices, the reservation policy would not affect domestic supply volumes, and so the deadweight loss from the policy would be zero.

In present value terms,14 the loss to the WA economy from the reservation policy is estimated to be $6.9 billion if long-run netback prices are the same as the average for 1977–2016. The loss rises to $22.9 billion if long-run netback prices are $9 per GJ. This is comparable to the loss estimated by the Productivity Commission (2015) for the eastern states, of between $2.2 billion and $23.9 billion.

5 Further Considerations

This section considers some further issues, pointing to areas for future research.

5.1 Reservation at the Project Level

Applying the reservation policy to each project, rather than the industry as a whole, imposes additional costs. For example, if all export projects are required to construct additional domestic gas facilities, the likely result will be a domestic supply chain with investment that is inefficiently sized or in the wrong location.

One way to lessen these inefficiencies is to allow exporting firms to contract with third parties to supply their domestic gas obligation. For example, this could be implemented as a property rights trading scheme, perhaps similar to the Mandatory Renewable Energy Target. Such a scheme may enable existing smaller onshore producers to supply gas to contribute to the domestic supply targets imposed on LNG exporters. There has been some movement towards this. While the Gorgon and Wheatstone projects are constructing new domestic gas facilities to supply domestic gas from their own reserves, the (now shelved) Browse FLNG project was permitted to contract with third parties, provided that it “results in a genuine net addition to energy supplied to the State” (Western Australian Parliament, 2015, p. 3).

5.2 Foreign Ownership

A number of studies raise the issue of foreign ownership of energy companies, including ACIL Allen (2014). In part, the reservation policy transfers income from gas-producing firms to gas-consuming firms.15 If gas-producing firms have higher foreign ownership than gas-consuming firms, then the reservation policy may act as an implicit income transfer from foreigners to Australians.

McGrath and Neill (2016) investigated foreign ownership in the WA gas industry. They find that around 45 per cent of income on each side of the WA gas market remains in Australia after taking tax payments into account. Therefore, it is arguable that the domestic gas reservation policy should not be considered as a mechanism to transfer income from foreign to local investors.

5.3 Reservation “Flexibility”

A WA Parliamentary Inquiry noted that, “The Reservation Policy allows for ‘case-by-case flexibility’, allowing potential producers to negotiate with government as to the amount to be reserved and the manner in which it is to be supplied.” (Parliament of Western Australia, Economics and Industry Standing Committee, 2011, p. 79)

However, based on the domestic supply requirements imposed to date, this paper has argued that the reservation policy has been binding, and therefore this “flexibility” is not sufficient to ameliorate the adverse effects on the gas market.

In addition, the possibility of 15 per cent domestic reservation will negatively affect the market by discouraging investment in exploration for new gas resources.

This “flexibility” also creates uncertainty which, in itself, can discourage investment. For example, an uncertain reservation percentage ranging from 11 to 15 per cent with a mean of 13 per cent would discourage investment to a greater extent than a commitment to a 13 per cent rate for all projects.16

5.4 Domestic Supply Dynamics

This paper has focused on the long-run loss caused by a domestic gas reservation policy, which is the most important policy consideration since it represents a permanent, ongoing loss in economic surplus. Nevertheless, some short-run dynamic considerations may mute the effects of the policy.

Firstly, the reservation policy requires that a certain volume be reserved for the domestic market and sold over the project's life. The producer can determine the level of supply in each year, but the policy still imposes an average level of domestic supply over the long term. Therefore, at best, this flexibility could partly reduce the adverse impact of the policy.

Secondly, a project may be willing to sell more to the lower priced domestic market if its LNG plant is operating at capacity. However, such additional domestic sales are likely to come at the expense of foregone future export earnings. In addition, the firm could add export capacity over the longer term.

Finally, given the currently low international LNG prices, it may be tempting to assert that the reservation policy has assisted LNG producers by requiring them to build infrastructure to access the domestic market. However, gas-producing firms have a greater awareness of the risks associated with fluctuating gas prices than the State Government, and are therefore better suited to making ex-ante investment decisions that maximise their dynamic efficiency.

6 Conclusions

This paper finds that domestic gas reservation policies inflict economic losses that can be divided into two parts.

Firstly, reservation policies impede overall gas production. Under the reservation policy, firms produce where the weighted average of the domestic and export prices is equal to the marginal cost of production. The policy suppresses the domestic price and thus discourages production.

Secondly, diverting gas from the export market results in foregone income for the Australian economy. At the margin, the value of gas to domestic users is lower than the value to the export market, so there is an opportunity cost from this diversion.

Whenever the reservation policy reduces the domestic price below the export netback price, the tax on gas production equals this discount multiplied by the proportion of supply reserved for domestic use (regardless of the shape of the supply and demand functions). Thus, a reservation policy is worse than an equivalent production tax because the reservation policy also generates losses from the embedded subsidy (and vice versa).

Using various data sources, the size of the deadweight loss caused by WA's domestic gas reservation policy is estimated to be between $0.41 and $1.37 billion each year, depending on export prices. The present value of these losses is between $6.9 and $22.9 billion.

Notes

  1. 1

    Some projects have a smaller reservation requirement.

  2. 2

    For gas reserves within State waters, the WA government could withhold the right to extract the gas. However, most reserves are outside State waters.

  3. 3

    While there may be value in having excess domestic gas capacity, a capacity of double consumption volumes is very high. For example, if supply capacity were 1.7 times actual annual gas consumption, this would mean that the gas supply infrastructure would bear much of the cost of fluctuations in gas demand from electricity producers. In 2014, annual electricity generation capacity for gas-fired generators was around 3.7 times actual generation. According to the Gas Bulletin Board data, these electricity generators account for approximately 26 per cent of total WA gas use. Thus, domestic gas suppliers need capacity around 1.7 times total gas consumption. However, this is likely to be an overestimate of the required excess capacity, because the electricity industry itself has significant over capacity (Department of Finance, 2014). In addition, the Mondarra Gas Storage Facility, as well as storage facilities at electricity generators, would assist with managing peak demand and reduce required excess capacity.

  4. 4

    Although many of these studies were completed before the recent falls in the international LNG price, the economic interpretation of reservation policies remains unchanged. See Section 'Domestic Supply Dynamics' for further discussion.

  5. 5

    Ceteris Paribus. This will be the case assuming that there are no industry-specific distortions in the alternative industries.

  6. 6

    Long-run marginal costs for most facilities use Energy Quest estimates reported in an IMO (2013) report. Costs for Prelude, Ichthys and Browse are from BREE (2014, p. 43). Costs for Pluto are the average of other projects in the Carnarvon Basin. Costs are adjusted to 2015 terms using the GDP price deflator. The cost of gas consumed during processing is 4 per cent of total gas, based on Core Energy Group (2013). These costs may be underestimates due to cost escalation issues experienced in recently constructed Australian LNG projects.

  7. 7

    Here, α is the proportion of total production supplied to the domestic market, whereas the reservation policy specifies the proportion of LNG exports that must be supplied to the domestic market.

  8. 8

    See Appendix S1.

  9. 9

    See Appendix S1.

  10. 10

    See Appendix S2.

  11. 11

    Two data points allow computation of the intercept and slope for the linear demand curve. The first is IMO estimates of the 2016 gas price and consumption volumes. The second is the price of the diesel substitute, at $15.89 per GJ, where demand for gas in WA is assumed to be eliminated.

  12. 12

    An OLS regression has been run on the data presented in Figure 3 to determine the intercept and slope of the linear supply curve in this model.

  13. 13

    The average annualised capital cost for liquefaction is estimated at AUD3.62 per GJ, based on total capital expenditure at NWS, Pluto, Gorgon and Wheatstone, assuming a plant life of 40 years and a required rate of return of 10 per cent. The key operating cost is gas consumed during the liquefaction process. Following the IMO (2015) assumption that 8 per cent of gas is consumed during liquefaction, operating costs are estimated at AUD0.69 per GJ for the 1977–2016 average LNG price.

  14. 14

    The discount rate is 6 per cent, following Productivity Commission (2015).

  15. 15

    However, as this paper has shown, the gain to gas consumers is smaller than the loss to producers.

  16. 16

    The negative impact of policy uncertainty on investment is analysed in Pastor and Veronesi (2012), Gulen and Ion (2013) and Nelson et al. (2010, 2012).

Ancillary