I live in California. That gives me a front seat to virtually every new initiative and trend related to saving the planet, whether it is about turtles and plastic straws, banning single-use plastic bags, electric vehicles, or green energy. Although not the first state to adopt a Renewable Portfolio Standard (RPS), California has been one of the most aggressive in its timetable for replacing fossil fuel based electricity with carbon-free. In 2018, California updated its RPS to the requirement to achieve 60% of electricity sales from renewable sources by 2030 and 100% by 2045. Of course, California’s aggressive push toward renewables has triggered a wide range of reactions. For example, Michael Shellenberger of Environmental Progress has been pushing the idea that California’s electricity rates are significantly higher than the rest of the US (see Figure 1) and rising significantly faster because of its dependence on renewables. His culprit is renewable energy and his solution is to keep nuclear plants open. In contrast, Roger Sowell, who blogs about renewable energy issues, argues that California’s unique climate, geography, and large population make such differences to be expected.

The data in Figure 1 appears to show that California rates are higher than US averages by between 2 cents and 4 cents per kWh, but this has been the case for at least 25 years. It’s impossible to identify any causation from raw prices, which is why I was very interested when the Energy Policy Institute at the University of Chicago released its “Working Paper: Do Renewable Portfolio Standards Deliver?” on April 22nd. The paper summarizes the impact of 12 years of experience with RPSs on renewables adoption, electricity price, and carbon emissions reductions. Based on a quick read of the summary, the report is a pretty damning characterization of such a renewable energy policy, particularly one which is being implemented by a wide range of US states. In particular, it appears to imply that the push toward renewables has dramatically increased electricity prices for average Americans and failed in its effort to cost effectively reduce carbon emissions. As a supporter of renewable energy, this troubled me, and I decided to dig more deeply into the paper to make sure I understood the issues. What I found troubled me even more than the initial summary – either the University of Chicago (of which I have high regard) erred in their conclusions, were selective in their data choices and analysis, or poorly presented their results. Or possibly a combination of all three.

To start, let me quote the online summary: “…Using the most comprehensive panel data set ever compiled on program characteristics and key outcomes, we compare states that did and did not adopt RPS policies, exploiting the substantial differences in timing of adoption. The estimates indicate that 7 years after passage of an RPS program…average retail electricity prices are 1.3 cents per kWh, or 11% higher…the comparable figures for 12 years after adoption are…a price increase of 2.0 cents per kWh or 17%. These cost estimates significantly exceed the marginal operational costs of renewables and likely reflect costs that renewables impose on the generation system, including those associated with their intermittency, higher transmission costs, and any stranded asset costs assigned to ratepayers.” (Note: I’ve removed some items for the sake of brevity to focus on the issue of cost to ratepayers.) In an attempt to understand how the researchers arrived at these conclusions, I downloaded the working paper, then read through and analyzed the full document.

Figure 2 is taken from the working report itself (labeled therein as Figure 2). From the report: “Figure 2 plots the number of RPS programs passed into law in each year…Figure 2 also plots real national average retail electricity prices (right y-axis) which declined from about 12 cents per kWh to 10 cents per kWh from 1990 through 2002 but by the end of the sample in 2015 returned to 12 cents per kWh.” Let me point out here that although the figure doesn’t state it, the report specifies these rates as “real”, i.e., after removing the impact of inflation. This is the only place in the entire 51-page document where any distinction is made between real and nominal electricity rates. The importance of this distinction can be seen in Figure 3, which plots EIA electricity rate data over a time period that encompasses the range of dates examined in the study. Here we see a similar pattern, but with somewhat less variability from a low of about 11 cents per kWh in the early 2000s to 12 cents in 2016. The EIA quotes “The average retail price of electricity in the United States has risen about 1.5% per year between 2006 and 2016, about the same as the 1.6% per year general rate of inflation over those years.” In any single year, the inflation rate has been as low as about -2% and as high as about +10% per year. Nowhere in the report does it indicate how this highly variable and critical factor was accounted for in their evaluation.

Now let’s look at Figure 4 (which is taken from the working report Figure A.2b) and forms the basis for the claim regarding electricity cost increase due to RPSs.

According to the notes from the figure “Blue lines show the point estimates and gray lines contain the 95% confidence interval.” I have four fundamental issues with this figure: 1) again, it is silent on the question of real vs nominal values; 2) in year 12 the point estimate appears to be 1 cent per kWh increase from time of passage, and there is no way to link the results to the summary which quoted an increase of 2 cents per kWh; 3) the 95% confidence bands range from a decrease of about 0.3 cents per KWh to an increase of about 2.3 cents, i.e., some scenarios in their model show a decrease in cost; and 4) the report makes no effort to explain the significant drop at the high bound that occurs from year 10 to year 12 and is not mirrored in the low bound.

Based on the information supplied, I would argue that the small increases supposedly identified, less than 1% per year, are lost within the noise of inflation. Back to California, Figure 5 (from Sowell) shows EIA rate data, both real and nominal (adjusted by the consumer price index (CPI) to 2014 values.

California’s rates increased nominally from about 13 cents per kWh in 2002 (first year of its RPS) to about 16 cents, whereas the real cost decreased from about 17 cents to 16 cents. Then throw in the complexity of the electrical generation and distribution system in California, the impacts of drought on hydro power generation, the cost of repairing infrastructure damaged by forest fires, the disruption of the traditional power utility business model from the adoption of rooftop solar, etc., and it seems hard to get excited about a possible 1 cent per kWh variation with no clear causation.

My conclusion is that the University of Chicago has presented an incomplete analysis based upon a limited set of data. However, let me give them the benefit of the doubt. Let’s say that their conclusions are spot-on, and that the adoption of RPSs has resulted in an increase in electricity costs to residents of those states. Let’s look at the two key drivers of that increase:

• Intermittency of power generated by wind and solar, which requires back-up generation. This may be true in the short term. However, as more renewables get added with their generation being geographically and temporally distributed across a region, in the aggregate, intermittency will decrease. And, even more important, building expensive gas peaker-plants that only run a few hours per day is a legacy of the 100-year-old utility business model. As the cost of batteries and other storage technologies drop, renewables plus storage will become the norm rather than the exception to new generation, and the need for peaker-plants will decrease significantly.
• The cost of new transmission infrastructure to connect remotely located solar and wind generation to population centers. From a simple perspective this appears to be valid. The best wind and solar resources in the US tend to be distant from population centers that use the power. However, the reality is that old plants, both coal and single-cycle gas plants, need to be replaced, and no one wants them in their neighborhood any more. As an example, I can see the stacks from the AES Redondo Beach Power Plant (1310 MW capacity) from my house. The first of the existing units were built in the 1950s, and, today, it operates at less than 5% capacity factor. AES has been trying to come up with a plan to replace the old units with newer, smaller combined-cycle gas units for years and has been repeatedly blocked by resident activists. They are now in negotiations to sell the property to a developer. Power for the local residents that was once generated locally has slowly but steadily moved outside the metropolitan Los Angeles area. The bottom line is that old plants need to be replaced, and that will likely require investment in new transmission infrastructure regardless of generation source.

And, finally, if I step way back from the numbers and look at the big picture of global warming and climate change, I find it very hard to get concerned over a 2 cent per kWh increase in the average cost of electricity. Another venerable institution, the Massachusetts Institute of Technology (MIT), which has been on the forefront of investigation into climate change, has recently become much more negative on the chances that the world can dramatically reduce carbon emissions in time and cap the increase in global temperature to Paris Agreement levels. In fact, MIT researchers see strong evidence that the process of global warming has accelerated, and they are now promoting adapting to the new world that is coming. More extreme views are even using the phrase “climate apocalypse” to describe the impact on world civilization.

So, while I will continue to disagree with the University of Chicago’s conclusions until I see more details of their work, more importantly, I believe those results are meaningless in the larger context of global impact.