In this chapter we examine the direction of the bias produced by externalities. Is it true, as often asserted, that when an activity generates external benefits, its competitive equilibrium level will always be below its optimum, and that where an activity imposes external costs, its equilibrium level must be excessive?

This is a proposition that underlies much of the policy advice given by economists on externalities issues: allocate more resources to goods that yield beneficial externalities and reduce their allocation to those that generate detrimental externalities. But suppose that advice is not always correct – what do such exceptions imply about the economist's advisory role?

As a matter of pure theory this problem is not as serious as the one discussed in the next chapter, for the difficulty we are considering here does not undermine the Pigouvian tax-subsidy solution. So long as the appropriate convexity conditions hold and the economy is competitive, one need merely impose the appropriate tax rates and the market equilibrium must occur at a Pareto optimum, wherever it may lie in relation to the equilibrium that would hold in the absence of Pigouvian taxes.

In practice, however, as will be emphasized in Part II, we do not know how to find or perhaps even to approximate optimal tax rates and so considerably coarser policy measures must be utilized. Usually these rely heavily on the conventional wisdom that constitutes the subject of this chapter: the acceptance of the view that one should expand outputs that yield beneficial externalities, and conversely.