Taxation and Investment Decisions in Petroleum.

• Author: Davis, Graham A.
• Position: Report - Statistical table

1. INTRODUCTION

   During the last forty years, the field of investment decisions has developed rapidly, both theoretically and practically. The adoption in industry of new tools does not happen simultaneously everywhere. The heterogeneity of practices is well documented. This is also true for the petroleum industry.

   When governments want to collect parts of the resource rent through taxation, it is useful to be able to predict the investment and production responses of companies. Academic economists have developed theories of taxation of firms, and of the taxation of resource extraction in particular, to make these predictions. These theories rely on the assumption that the efficient firms either undertake complex value-maximizing calculations prior to making investment and production decisions, or behave "as if they do. A particular topic that has been disputed is how generous deductions need to be in a rent tax system in order for the rent tax not to create investment and production distortions for these firms.

   An article by Osmundsen et al. (2015) in a special issue of this journal devoted to Morris Adelman, who himself wrote often about the distortionary effects of taxation, discusses actual oil company treatment of deductions, with harsh criticism of petroleum taxation in Norway. This article presents a different view. (1) The academic view is that firms correctly price the incentives to invest given to them by governments. Particular attention has been paid to depreciation allowances that represent risk-free income for the firm (Summers, 1987; Fane, 1987). Norway has gone to great lengths to reduce the risk connected to future tax deductions in the petroleum tax system, including surety of deduction carry-forwards. On this background the authorities assume that companies value tax deductions as if they are risk free. A review of the literature shows that practices in firms are more heterogeneous than suggested by Osmundsen et al. (2015), and that some firms report the valuation of low risk depreciation streams in line with academic predictions. This contributes to explaining why numerous academic articles, as well as reports on tax reform, give taxation policy recommendations that are at odds with those of Osmundsen et al. (2015).

   Osmunden et al. (2015) seem confounded that standard academic theory can suggest that firms use different discount rates for different cash flow streams, and refer repeatedly to the idea that companies discount only net cash flows based on some overall project or firm discount rate. Standard theory suggests superposition, also known as value additivity: cash flows have the same value to the firm no matter how they are aggregated. The positive effect from a guaranteed depreciation allowance can therefore be valued by discounting it at the riskless rate and adding it to the present value of the other cash flows. Or, as Osmundsen et al. (2015) admit, firms may instead discount net after-tax cash flows at an opportunity cost of capital that "reflects the average risk of the resulting aggregate" (p. 196), including the risk effects of the tax deductions. In Section 4 we present a simple model of investment decisions under uncertainty and taxation that illustrates how firms would make decisions in line with standard theory applied in this way.

   The essential question is not, as suggested by Osmundsen et al. (2015), whether firms apply one discount rate to net after-tax expected aggregate cash flows or use different discount rates for different cash flow elements that differ in risk. Superposition implies that both approaches yield the same valuation and investment behavior. The question is the extent to which firms price fiscal policy risk into their investment decisions, and whether they get either approach to valuing such risk approximately correct. Osmundsen et al. (2015) assume that firms calculate a project's after-tax beta from an average of stock market returns, that beta having nothing to do with a presumption of a riskless or even slightly risky deduction scheme. Our model shows a reasonable approximation of how the firm may arrive at the appropriate endogenous after-tax cost of capital using this second approach, be the deduction scheme riskless or risky. We conclude that the appropriate concern is not, therefore, over firm practice, but over the factual evidence as to the riskiness of the deduction scheme in Norway.

   Section 2 describes the Norwegian petroleum tax system and the political risk connected to it. Section 3 reviews the literature. Section 4 presents the model. Section 5 gives our discussion and interpretation of the model. Section 6 concludes.

2. NORWEGIAN PETROLEUM TAX AND RISK

   Our main topics are which methods to use in evaluation of investment projects under taxation and risk and some results on effects of taxation. This section will nevertheless provide some details about petroleum taxes in Norway and the related political risk. This is useful since we comment on a previous article that is applied to Norwegian petroleum, but also as an example to make our points more concrete. However, the remaining sections can be read without a comprehensive understanding of the details. More background is found in Lund (2002a; 2014b; 2018).

   Since 1992 the main features of the tax system have been unchanged from the viewpoint of a company which is never out of tax position. The system consists of a corporate income tax (CIT) and a special petroleum tax (SPT). There is no gross taxation, royalty, or similar. Neither of the two taxes is deductible in the other, so the total marginal rate is the sum of the two rates, constant at 78 percent since 1992. Today the rates are 23 percent and 55 percent, while they were 28 percent and 50 percent during 1992-2012. The CIT has a depreciation allowance, linear for six years, which is generally regarded as a bit too generous compared with true economic depreciation in the average project. (2) Interest expenses are deductible. Exploration costs can be expensed, i.e., deducted when they are incurred.

   The SPT has the same tax base as the CIT with two exceptions. There is an extra deduction, the "frunntekt," translated into English as an "uplift." This is like an extra depreciation allowance, currently of 21.2 percent, i.e., 5.3 percent of investment deductible for four years. The intention is that the CIT should tax the normal and supranormal return to capital, while the SPT should only tax the supranormal return, the rent. The other exception is targeted at thin capitalization and was introduced in 2007. The interest deduction in the SPT is limited to interest expenses on debt equal to 50 percent of the remaining value of capital after tax depreciation, reduced linearly to zero in six years. Apart from this limitation, the features described above were unchanged from 1992-2012. As a response to a gradual lowering of the CIT for other sectors in Norway from 28 to 23 percent during 2013-2018, the SPT rate was increased to keep the sum constant. The uplift was adjusted down from 30 to 22 percent in 2013, then lowered somewhat in 2016 and again in 2017.

   There have been three other changes that only affect companies out of tax position. Since 2002 there has been loss carry-forward with interest accumulation, in order to preserve the present value of tax deductions for the companies. Since 2005 a company that has exploration costs and is out of tax position will get the tax value of the costs refunded, like a negative tax. Furthermore, a company which closes down its operations without being able to deduct losses carried forward will get the tax value of these losses refunded. In a world context these three provisions are quite unusual, although Alaska had a refund for eventual losses during some years (Bradner, 2016, 2017). In the jargon of tax economists, the three provisions will create perfect loss offset as long as they are not changed by the authorities. For this reason, the Ministry of Finance of Norway calculates the value of future tax deductions based on the presumption that firms value the deductions as a riskless revenue stream (Ministry of Finance, 2012; Jensen, 2017).

   It may be useful to compare the system with rent tax systems that are known from the literature. The Resource Rent Tax (RRT) of Garnaut and Clunies Ross (1975) is different in that it is cash flow based, i.e., investments are expensed like operating costs. The Norwegian SPT is more like the Allowance for Corporate Capital (ACC) system (Boadway and Bruce, 1984). In addition to depreciation allowances, there is an extra deduction, the uplift, to protect the "normal" rate of return from the SPT. The uplift is proportional to investment, not only to the equity financed part, which is the case in the better-known Allowance for Corporate Equity (ACE) system (IFS, 1991; OECD, 2007). But the SPT uplift is a fixed, constant percentage of investment for four years. This differs from the ACC allowance, which each year typically is based on a market interest rate, which may vary over time. Moreover the ACC allowance is proportional to the remaining value of investment after tax depreciation, falling over time, and is thus theoretically just what is needed to protect the normal return from the tax. The SPT uplift is a simplified approximation to this. Since the SPT does not conform exactly to any of the systems known from theory, and since depreciation and interest deductions also differ from theoretical prescriptions, one needs to calculate the present value of deductions in order to evaluate whether the CIT-cum-SPT system gives incentives or disincentives for investment (see Lund, 2018).

   When the SPT uplift rate was held fixed for 20 years, its present value at a (time varying) market interest rate was not constant. This may have been one reason for the decision to decrease the uplift rate in 2013, since market rates had been decreasing. However, that problem is...