Regulating Carbon Emissions: Efficiency vs. Redistribution

Bloomberg reports that New York state’s plan to eliminate its few remaining coal power plants has caused power prices for delivery in 2020, 2021, and 2022 to increase. Eyeballing the chart, the impact of the proposed regulation is on the order of $7/MWh, or about 25 percent of the 2019 price.

Coal represents a dwindling fraction of New York’s generation. The EIA reports 0 electricity from coal in October, 2018. As of 2014, the last full year for which I could find data on the EIA website, coal accounted for 4.6 million MWh, out of a total of 137 MWh of generation.

The efficiency impact of this depends on (a) the estimated social cost of carbon, (b) the kind of generation that will replace the shuttered coal plants, and (c) the non-energy costs that this replacement generation creates.

If you believe that the cost of carbon is $40/ton, if coal is replaced by zero emissions generation, the move is efficiency enhancing. A coal plant with a heat rate of a little more than 10 implies a carbon cost per MWh of $40. This is well above the price increase of around $7.

If coal is replaced by natural gas, with a carbon cost of about $20/MWh, the call is closer, but still comfortably in favor of eliminating coal.

Lower social costs of carbon of course affect the math. The other thing to keep in mind, though, is that the price is for energy only. Changing the generation mix also affects the need for ancillary services to maintain grid stability. In particular, substituting diffuse and intermittent renewables for coal increases the non-energy costs of supplying electricity. These costs can be appreciable, though again it’s difficult to see them being so large as to overcome the approximate $160 million in carbon cost savings from eliminating coal, based on a $40/MWh CO2 cost, ~4 MWh of coal fired generation, and replacement of coal by zero carbon emissions generation sources.

What’s truly startling about the numbers, though, is the redistributive impact. Price is driven by marginal cost, and the price impact suggests that the cost of the marginal megawatt hour from coal replacement generation is about $7/MWh above that of the eliminated coal units. Note: that $7/MWh price increase benefits every single MWh generated by inframarginal units (e.g., combined cycle NG). Coal represents (as noted before) ~3 pct of NY generation, but the remaining 97 percent will see a big increase in margins.

This is a crude calculation, but roughly speaking the regulation will result in a transfer of about $1 billion/year from consumers to owners of generation (~140 million MWh x $7/MWh). The vast bulk of this $1 billion will be a quasi rent for inframarginal generating assets. (About $28 million–4 mm MWh/year x $7/MWh–will cover the cost of the more expensive generation that replaces coal plants.)

As is often the case with regulation, the wealth transfers swamp the efficiency effects (which total at most $130 million=~4 MM MWh x $33/MWh in social cost savings). (Since coal generation has probably dropped from the 4 million in 2014, and the price impact reflects the elimination of the remaining coal generation, the total efficiency effects now are probably substantially smaller than $130 million.)

Thus, although this regulation is sold as one benefitting the environment, I strongly suspect that the political coalition that has given it birth is strongly supported by incumbent generation operators selling into the New York market. That is, it smacks of the typical special interest regulation that benefits a small concentrated group at the expense of a large diffuse one (i.e., the consumers in New York), all dressed up in pretty green (environmental green camouflaging Benjamins green, as it were).

Yes, in this instance perhaps–depending on one’s assumptions about the cost of carbon and the incremental uplift costs created by the regulation–this bargain has produced an efficient outcome. But the redistributive nature of this regulation, and those like it, creates a great risk that such regulations will be introduced even when they are inefficient.

Those harmed include ordinary New Yorkers lighting their homes, and commercial and especially manufacturing firms (and their employees) who pay higher power costs. (Employees will pay in lost employment and lower wages, due to a decline in derived demand for labor driven by higher costs of other inputs.) In France, a seemingly small imposition on a similar group sparked widespread social unrest. It hasn’t happened in the US yet (or in places like Germany, where consumers and employers are paying steeply higher electricity costs due to anti-carbon regulations), but US states should be aware that such policies could trigger resistance here as well–especially if and when the hoi polloi realize that the biggest winner from these policies is not the environment, but companies that are pretty unpopular to begin with.