In an NBER working paper, Geoffrey Heal discusses some aspects of the energy transition to come.
On infrastructure investments:
the likely net investment required to go carbon-free is now as little as $0.179 trillion
renewable power from wind and solar PV plants is now less expensive than power from gas, coal or nuclear plants … If it were not for the intermittency of renewables, we would save money by converting to clean power.
the social benefits from stopping the CO2 emissions from coal and gas in power generation in the U.S. amount to $200bn annually, roughly an order of magnitude greater than the costs. Furthermore, these benefits will continue for ever, whereas the costs are fully paid by 2050. … As greenhouse gases are a global public bad, many of these benefits will accrue to countries other than the U.S.
Carbon taxes only delay the extraction of fossil fuels except for those fuels whose marginal extraction cost is sufficiently high such that extraction cost plus tax exceeds the cost of alternative energy sources:
the Pigouvian and Hotelling frameworks lead to rather different conclusions when it comes to thinking about the effectiveness of a carbon tax. Pigou emphasizes the impact of a tax on substitution between commodities, in this case between energy sources. Hotelling on the other hand emphasizes the impact of a tax on an exhaustible resource on the time-path of consumption of that resource.
[in the Hotelling setting] the tax either has no effect at all on the cumulative consumption of the fossil fuel, or it drives it out of the market completely.
If we want to reduce cumulative oil consumption by for example 30%, then we need a tax of about $500 per ton of CO2: if we wanted to reduce oil consumption by two thirds we would need a tax of over $600 per ton CO2.
The marginal social cost of power from renewable sources is close to zero, as wind, solar and hydro all have essentially zero operating costs. So we would need much lower power prices to provide the correct incentives to use clean power rather than fossil fuels.
The classic response to this conundrum has been to recommend two-part tariffs, with a fixed charge or connection or membership charge recovering the fixed costs and a usage tariff covering the variable costs.