Towards a "Spatialized"" Cost of Capital Concept

In the neoclassical tradition, the investment demand (the setting up of a new firm, for example) is explained by the cost of capital. This indicator expresses the minimum rate of return the investment project must yield before taxes in order to provide the saver with an attractive net of tax return. It captures in a summary statistic the financial cost, the effective economic depreciation, all tax devices, and investment incentives that directly affect the return on capital. This approach has widely proved its relevance and its usefulness for comparing various location choices. While commonly used, this approach is not exempt of criticisms. In particular the plausibility of the main assumption is questioned : undifferentiated space. This hypothesis renders the marginal productivity of an investment project being absolutely the same wherever it locates. The assumption mentioned above entails other undesirable consequences of diverse importance, too: - the undifferentiated space involves that (fixed) capital has not to be split up into plant and equipment on the one hand, land on the other hand. Pricing both components is not required. Any difference in land price, for example, across space is therefore ignored; - the framework of competitive analysis is no longer the most appropriate one. The fact that some input suppliers are located closer to any locator than others is notably responsible of that. The object of this paper is to extend the cost of capital framework by accounting for space.