This discussion paper resulted in a publication in the <A href="http://www.sciencedirect.com/science/article/pii/S0095069612000927">'Journal of Environmental Economics and Management'</A>, 2012, 64(3), 342-363.<P> Optimal climate policy is studied in a Ramsey growth model with exhaustible oil reserves, an infinitelyelastic supply of renewables, stock-dependent oil extraction costs and convex climate damages. Weconcentrate on economies with an initial capital stock below that of the steady state of the carbon-freeeconomy and the initial cost of oil (extraction cost plus scarcity rent and social cost of carbon) below thatof renewables. There are then two regimes. If the oil stock is small, the social optimum path consists of aninitial oil-only phase followed by a renewables-only phase. With a lower cost of renewables or a lowerdiscount rate, more oil is left in situ and renewables are phased in more quickly. The optimal carbon taxrises along its development path during the oil-only phase, but the rise flattens off as less accessiblereserves are explored and the social cost of carbon increases. Subsidizing renewables without an optimalcarbon tax induces more oil to be left in situ and a quicker phasing in of renewables, but oil is depletedmore rapidly initially. The net effect on global warming is ambiguous. The second regime occurs if theinitial oil stock is large. The social optimum then consists of an initial oil-only phase followed with a finaloil-renewables phase. This regime converges to the steady state of the carbon-free economy as well. Thepaper also gives a full characterization of two other regimes that occur if the initial capital stock is aboveits steady state or renewables have an initial cost advantage.
