A new working paper from the University of Chicago’s Energy Policy Institute is creating a bit of a stir. It concludes that one of the most popular, successful, and common state-level renewable energy policies – Renewable Portfolio Standards, or RPS – is far more expensive than previously believed. But there are reasons – multiple, in fact – to take the authors’ provocative conclusion with a grain of salt. So, too, their implied alternative, which would be to take a policy that seems to be working in 29 states and replace it with one that has nowhere been implemented on a scale sufficient to change the electric power mix as much as the RPS targets now being adopted by states around the country.

Although RPS programs differ considerably in their details, an RPS basically imposes a binding requirement that retail electricity providers sell electricity with a specific percentage of renewable energy, typically increasing gradually, year by year, or moving toward a fixed target by a certain date. The working paper claims, by analyzing retail electricity rates and other state-level data, that RPS programs, on average, raise retail rates by 11% (or about 1.3 cents/kWh), seven years after introduction, and by 17% (2 cents/kWh), 12 years after introduction. The price impacts at year seven correspond, according to the authors, to a net increase of 2% in the renewable energy content of electricity. A related finding, not surprisingly given this primary conclusion, is that RPS is an expensive policy for achieving carbon reductions (the only benefit the working paper considered).

The results of the working paper have been met with both praise and skepticism, including some excellent commentary in a meaty Twitter thread by Harvard University’s Jesse Jenkins (yes, Twitter – welcome to peer review in the 21st century!). After reading the working paper and much of the commentary on it, we urge you to use the paper with extreme caution, or better yet, don’t use it at all until it can be properly peer reviewed. Here are just some of the reasons:

• We reiterate that this is a working paper, meaning it has not been peer reviewed or subject to the normal scrutiny of a published manuscript. As the authors note, it also appears to be the first of its kind in terms of approach, which further argues for taking a wait-and-see stance before citing it as a definitive analysis. Simply put, we need time for the appropriate experts to “look under the hood” and independently confirm the methodology and analysis.
• A back of the envelope calculation on the primary results suggests that the reader should approach the central conclusions with healthy skepticism. If just 2% net additional renewable energy content raises retail rates by 11%, and if bulk supply costs (generation plus transmission) are roughly 3/4 of the retail rate, this means that the addition of 2% renewable energy to the mix raises bulk supply costs by about 15% (assuming distribution costs are largely unaffected, which seems reasonable at these low levels of incremental renewable energy generation). This means that the all-in cost of renewable energy would be about eight times that of conventional energy. This does not seem likely. If it were, then the price of electricity in states with high levels of wind and solar penetration (regardless of RPS policies) like Texas (wind power 15% of total generation in 2018) and Iowa (35% wind power), would have exploded. Instead, Texas and Iowa are among the 10 states with the lowest average retail electricity prices, according to the Energy Information Administration.
• The analytical approach sidesteps a “bottom up” analysis of the electricity system to understand cost drivers for renewable energy deployment, and instead uses retail rates as the key measure, attempting to correct for factors affecting the retail rate unrelated to the RPS policies they are studying. The study authors take this approach because the electricity system is incredibly complex, making a bottom-up analysis very difficult (“nearly impossible” according to the authors). But because retail rates smooth out this complexity over time, they can mask multiple trends that may either counteract or reinforce each other. Rates go up and down based on fuel costs, ongoing investments in generation, transmission and distribution infrastructure, investments in environmental controls at existing power plants, compliance with other policies (like RPS and energy efficiency programs), and changes in total demand and consumption patterns, among other things. Correcting for all of these and other effects so that the RPS “signal” in retail rates can be separated from the noise is not easy, and it is difficult to tell how close the authors have come to doing it.
• To compound this challenge, many RPS programs were implemented as part of much bigger changes in energy policies and market designs, namely the unbundling of vertically integrated utilities and the creation of competitive wholesale and retail electricity markets.
• By only looking at costs in direct dollars-and-cents terms, the paper draws focus away from the host of benefits that RPS provides. As the authors write in the Policy Implications section of the four-page Research Summary; "RPS policies have goals beyond reducing carbon emissions, such as spurring improvements in renewable technologies and improving the growth of these sectors of the economy. The study did not analyze the extent to which RPS policies have succeeded in meeting these goals. The authors note, however, that if RPS policies reduce renewable costs industrywide, then this would alter any cost-benefit analysis." Since that has undeniably happened, the authors’ bottom-line conclusion of how costly RPS is would seem due for alteration.

Given the body of literature on RPS costs and benefits (here’s one example from two national labs) it would be hard to argue that the policy has not added something to retail electricity prices. At the same time, the authors make arguments (in the Research Summary, not the working paper itself) that other policies, and specifically a carbon tax, would be more economically efficient. This is also likely true. But passing those types of policies at the levels of carbon pricing needed to drive change in the power mix has proved extremely difficult. And, in the absence of a policy as simple and elegant (to an economist, anyway) as a carbon tax or similar policy, an RPS does encourage competition (at least among eligible technologies) to help drive down costs.

The authors of this working paper have raised some very interesting and important questions, but clearly much more work is needed. In the meantime, RPS marches on, with several states recently ratcheting up their RPS targets or passing ambitious Clean Energy Standards, including several targeting 100% clean grids by 2050 or sooner.