Taxes, Investment and Q

Abstract

This paper attempts to provide a unified analysis of the effects of taxation on the equilibrium value of marginal q under alternative financial policies. It is shown that the q approach does not avoid all the specification problems associated with analyses based on the cost of capital. The (theoretically and empirically) crucial relationship between average and marginal q is also examined, and it is shown that, under UK and US tax rules, the possibility of winding up precludes persistent undervaluation in equilibrium. There have now been a considerable number of empirical studies of investment behaviour based on some variant of Tobin's q, defined as the ratio of the firm's market value to the value of its capital stock: see, for instance, Jenkinson (1981), Oulton (1981), Poterba and Summers (1981), Summers (1981) and von Furstenberg (1977). The q approach is seen as having the empirical advantage-compared in particular with the more traditional approach based on the cost of capital-of allowing expectations to be brought directly into the study of investment behaviour by relying on asset prices to digest all relevant expectations and reflect them in observable data. An immediate difficulty, however, is that, whilst the incentive to invest is related to the effect on shareholder wealth at the margin, and hence to marginal q, only the value of average q is directly observable. The consequent difficulty of implementation is not, however, considered here (see Bailey and Scarth (1980) and Hayashi (1982) for discussion of conditions under which marginal q can be inferred from average q). Rather this paper is concerned with the prior question of how personal and corporate taxation affects the equilibrium value of marginal q. For there has, as yet, been no systematic analysis of the effects of taxation on the equilibrium value of (marginal) q comparable with the analysis of the relationship between taxation and the cost of capital in King (1974). Thus Auerbach (1979a, b), Bradford (1981) and Edwards and Keen (1984) have all considered how taxes affect q when the marginal source of funds is either retained profit or the sale of equity, and the distinction between these two cases has also been recognized in the empirical work of Poterba and Summers (1983). But the important case in which debt is the marginal source has been ignored.' The effects on marginal q of the constraints that typically need to be imposed in modelling financial policy have also been neglected. This paper attempts to fill these gaps in the literature by providing a general analysis of the relationships between taxation, financial policy, investment and q. The paper also examines the implications of the special tax provisions that may apply in the event in the firm's ceasing operations (winding up). This bears upon one of the unresolved issues associated with the new view of equity finance (Auerbach (1983)), namely the suggestion that if dividends are taxed more heavily than capital gains,