On the Taxation of Exhaustible Resources under Monopolistic Competition

Article excerpt

I. Introduction

Since the publication of Hotelling's classic paper, a number of insights have been gained into the problem of the optimal pattern of production of exhaustible resources. In that paper, Hotelling [1931] also considered the impact of a severance tax (either ad valorem or per unit) on the extraction path, based on the assumption that the stock of the resource (say, ore) is exogenously given. His conclusion that the tax induces a reduction in initial output in favor of increasing output in later periods has been confirmed by Burness [19761 and Dasgupta and Heal [1979, ch. 121. However, this "time-tilting" [Heaps and Helliwell, 1985, p. 423] is also accompanied by a tax-induced "high-grading," i.e., some resources will not be extracted even though their value before taxes exceeds the cost of extraction. This means that the tax policy affects or even determines the highest grade of ore that is mined. Thus, the cut-off grade of ore is an economic decision, and the stock of ore cannot be treated as exogenous, 'except in the very longest of long runs' [Conrad and Hool, 1981, p. 18].

While a number of authors(1) have attempted a complete taxonomy of the effects of different taxes on resource production and extraction, some important insights are given in Dasgupta, Heal, and Stiglitz [1980], in which they argue that 'there exists a pattern of taxation that can generate essentially any desired path of resource usage; that is taxation provides a sufficiently wide set of policy tools to influence resource depletion in, more or less, any manner that one cares to choose' [p. 150].

This is certainly true in the competitive case. In the present paper, which is narrower in scope, an attempt is made to see what policy tools might be used to capture resource rents through taxation when resource extraction is carried out under conditions of limited, or monopolistic competition, but bearing in mind the two tax-induced problems of time-tilting and high-grading that have much occupied the literature referred to above.

Before proceeding, it might be useful to explain the "mixed" market structure called monopolistic competition, which Stiglitz and Dasgupta [1982] also call limited competition. Some recent work on market structures (e.g., Novshek [1985], Roberts [1980], Stiglitz [1976, 1986(a), 1986(b)]), makes it clear that the analysis of imperfect competition is very sensitive to the assumptions embodied in it; for instance, it is no longer always the case that imperfect competition results lie between" the polar cases of pure competition and monopoly.

Furthermore, the large body of recent work called the "new industrial organization" (surveyed by Tirole [1988] and Jacquemin [1987]) contains a number of plausible models which capture the spirit of Chamberlinian competition, in which strategic interactions may be legitimately ignored. But in any case, it is an empirical matter as to which industries are best described as monopolistically competitive, in which strategic interactions are not of first order importance.

Next, the nature and force of competition is now regarded as multifaceted: firms do not compete in price only, and the number of firms in the industry is not as significant as once thought. The nature of competition will depend on factors other than the number of firms: such factors include sunk costs, informational structure, the possibility of international trade, and so on. In fact, when there is imperfect information about prices, the equilibrium price can be the monopoly price even when there are many firms in the industry [Diamond, 1971].

The diversity and richness of the new industrial organization notwithstanding, there is at least one unifying feature that is almost a defining characteristic of monopolistic competition, and this is some degree of market power, no matter how circumscribed. Operationally, this market power translates into a downward-sloping demand curve(2) faced by each firm in the industry, although only a part of the demand curve, close to the equilibrium price, is known to the firm. …