R&D investment decision and optimal subsidy

This article assumes that a firm facing technological uncertainty must decide whether to purchase R&D capital at each instant. R&D capital exhibits both irreversibility and externality through the learning‐by‐doing effect. The combination of irreversibility and uncertainty drives agents to be more prudent; the maxim 'better safe than sorry' applies. This maxim is more important if uncertainty is greater, technology progresses at a lower pace, the externality is stronger, or a catastrophic event is less likely to occur. A firm ignoring the externality will both invest later and disinvest earlier than a social planner who internalizes the externality. An equal rate of investment tax credits should be given to both costlessly reversible investments and irreversible ones, and the same rate of taxation should be imposed on disinvestment.