Energy & Environment

Missing the forest for the trees on solar net metering

In a recent essay on the solar photovoltaic (PV or “rooftop”) power market, Mark Muro and Devashree Saha of the Brookings Institute applaud the net metering system of subsidizing such rooftop installations, concluding that the system is a “net benefit” for both the recipients and the non-recipients of the subsidy; that is, for all power consumers.

{mosads}They arrive at that conclusion largely through a literature review both limited and obviously cherry-picked, and then utterly eviscerated in two subsequent review essays. But the analytic problems inherent in the Muro/Saha paper are far deeper, in that their conclusions about the benefits of solar PV electricity are driven by deeply dubious assumptions on the costs of competing conventional power that solar PV allows the market to avoid, and on the purported environmental advantages of solar PV over conventional electricity.

First, a brief description of net metering: Power consumers who install solar PV panels — large explicit subsidies are paid for such installations — receive a credit from their electric utility for the power that they produce but do not consume. The excess electricity is transferred into the power grid for use by other consumers, and the owners of the solar systems pay only for their “net” electricity consumption.

So: What problems are created by net metering? First, in most jurisdictions, the credit price paid for the excess solar power sold back into the grid is higher than the cost of competing electricity production, usually from utilities or from the spot power market. Accordingly, higher-cost solar PV power is substituted in place of less costly electricity. The price paid often is the retail price, which includes transmission and distribution and other such costs that rooftop solar generation not only does not incur, but that solar PV customers do not pay for fully, even though they maintain access to conventional power as insurance against times that the sun is not shining (see below).

Consumers without such solar installations have to finance the payments for excessively expensive rooftop electricity, and for the components of the system for which solar PV customers do not pay fully. Accordingly, overall power prices are forced above the level that would prevail in the absence of the net metering system. This problem is exacerbated by the array of tax and other incentives to install solar systems: The combination of the installation subsidies and the excessive prices paid for the power fed into the grid means that more solar capacity is installed than otherwise would be the case, more and more expensive power is fed into the grid, and prices are forced up, in principle in a sort of upward spiral process.

There is the further matter that reliability is a hugely valuable attribute of power systems; no one likes blackouts. Electricity bills reflect the cost of that reliability in the form of “capacity” charges — that is, the part of the bill covering the cost of the physical system and its spare capacity, before fuel expenses and other such generation costs. People who install solar systems benefit from the reliability provided by the grid — they consume conventional power at night and at other times that the sun fails to shine — but because they pay only for their “net” power consumption, they get a free ride on the cost of the generation equipment and other capital that yield the reliability upon which they depend. Except the free ride is not free: Other consumers have to pay for it.

Over the long run — not necessarily a long period of time — the higher costs and prices mean that investment in maintenance and new generating capital will fall, and with it reliability and the economic benefits of inexpensive power. Only costs will rise, not a salutary outcome.

Net metering receives strong political support in substantial part because it is useful politically. All subsidies — direct, indirect, explicit or hidden — must be financed by someone, be it taxpayers, ratepayers or the beneficiaries of other government programs allocated fewer resources. Political incentives to hide the costs of such policies are powerful: It is better for bureaucrats and politicians that the losers not know that they are losing, and net metering serves that end beautifully.

Muro and Saha argue that solar PV systems yield net benefits for all power consumers — even those bearing the costs summarized above — because rooftop systems allow the grid writ large to avoid the costs of conventional generation, including those for transmission and distribution, and because of the purported environmental advantages of solar PV.

Both of those arguments are incorrect. With respect to the avoided cost argument: To say that substitution of a more expensive form of electricity in place of a less expensive one yields consumer benefits is, bluntly, preposterous. The true resource cost of rooftop solar power — that is, ignoring the subsidies, which hide, or shift, but do not reduce actual costs — is higher than that of conventional generation even given the need for transmission and distribution attendant upon the latter. That is why solar PV cannot compete without subsidies both direct and indirect; examples of such subsidies are the investment tax credit, the net metering subsidy already discussed, highly favorable depreciation allowances, guaranteed market shares (“renewable portfolio standards”) and others. This is acknowledged implicitly by Muro and Saha:

[T]he local utility in Nevada successfully wielded the cost-shift theory last winter to get the Nevada Public Utilities Commission to drastically curtail the state’s net-metering payments, prompting Solar City, Sunrun, and Vivint Solar — the state’s three largest providers of rooftop panels — to leave the Nevada market entirely. The result: New residential solar installation permits plunged 92 percent in Nevada in the first quarter of 2016.

The Energy Information Administration estimates that the “levelized” (loosely, spread evenly over the life of the generation facilities) cost of solar PV electricity is about $125 per megawatt-hour (mWh). For coal generation, it is about $95, for natural gas it is about $75, for nuclear power it is about $95 and for hydroelectric power it is about $83. Lazard has published similar computations (using a somewhat different methodological approach): Levelized costs per mWh are approximately $185 (solar PV), $107 (coal), $65 (natural gas) and about $124 (nuclear). (Lazard did not publish an estimate for hydroelectric generation.)

That EIA and Lazard use differing methodological approaches for the comparative cost analysis is irrelevant: In the context of the discussion here, the issue is whether the Muro/Saha argument is correct about the benefit of solar PV defined as the avoided cost of conventional power. If we use the most recent EIA data (2013) on delivered electricity costs (Table A8), including transmission and distribution, the U.S. average was about $101 per mWh, about 19 percent less than the $125 levelized cost estimate for solar PV. And even that figure is biased in favor of solar PV, in that the $101 figure includes the delivered costs of uneconomic renewable generation that has been expanded by subsidies of various kinds. The Muro/Saha argument that the net metering system allows the power market to avoid the costs of conventional electricity simply is a non sequitur, in that net metering leads to a substitution of expensive power in place of cheaper power.

And even that obvious reality understates matters, as the unreliability of solar energy creates a need for backup conventional capacity so as to avoid shortages, as discussed above. There is no single estimate of those backup costs applicable to a given power source, in substantial part because of the need to make assumptions about the percent of the time that the backup units would be producing power (the “capacity factor”). That parameter will vary with local conditions. My own estimate (pp. 26-31), based on EIA cost data and highly conservative assumptions, is $368 per mWh. That is for all renewables as a class; the figure for solar PV might be lower. But any assumption about that backup cost reduces the competitiveness of rooftop solar generation in the context of levelized costs, and weakens ever further the Muro/Saha avoided-cost argument for the purported net consumer benefits of the net metering system.

At a more fundamental level, the net metering system fundamentally harms consumer well-being because the excess consumption of resources used to produce solar PV electricity means by definition that the electricity sector is more costly than necessary, and therefore that the aggregate consumption basket must be smaller than otherwise would be the case. That is the reality in any world in which resources are limited. More narrowly, the net metering payments paid to power consumers installing solar PV systems have to be financed by other power consumers; no other outcome is possible in a world in which there are no free lunches. Therefore, the net metering system cannot yield net benefits for all power consumers even in principle; it must subsidize some consumers at the expense of others.

With respect to the second Muro/Saha argument, the purported environmental advantages of solar PV over power generated with conventional technologies: not so fast. Put aside the toxic-metal pollution problems attendant upon the production of the PV panels; those can be addressed with improvements in handling protocols, substitution of less dangerous materials in place of more toxic ones, and the like, although the extent to which improved environmental standards are likely actually to be adopted in China is less than obvious. To the extent that production of the panels is shifted to countries with more stringent standards, the much-ballyhooed “cost-competitiveness” of solar PV power is likely explicitly to be proven an illusion; in any event, it is obvious that solar PV generation substitutes one set of environmental effects in place of another.

Consider instead the implications of the unreliability of “renewable” electricity generally, and of solar PV power in particular. The need to avoid blackouts means that backup generation capacity — provided by conventional units usually powered by coal or natural gas — must be cycled up and down depending on sunlight (and wind) conditions in the various regional markets. But such cycling of conventional units means that they cannot be operated efficiently, the result of which is an increase in the output of conventional effluents and greenhouse gases per mWh generated; and recent analysis of Colorado and Texas shows that the cycling problem is sufficiently severe that it yields an increase in the absolute amount (not merely the amount per mWh) of conventional effluents and greenhouse gases emitted as the market share of renewables rises.

The Muro/Saha argument is the natural result of a quasi-religious belief in the virtues of unconventional electricity; all evidence must be interpreted in the context of that belief system, as distinct from the traditional scientific approach of allowing evidence to influence hypotheses (or beliefs). In the strange new world of “carbon” obsessions, expensive power is a benefit because it crowds out the costs of cheaper power, and environmental effects are assumed rather than analyzed in the light of actual evidence. Such are the fruits of intellectual flabbiness.

Zycher is the John G. Searle scholar at the American Enterprise Institute.

The views expressed by contributors are their own and not the views of The Hill.


The Hill has removed its comment section, as there are many other forums for readers to participate in the conversation. We invite you to join the discussion on Facebook and Twitter.

See all Hill.TV See all Video

Most Popular

Load more


See all Video