,
16 tweets,
4 min read
Read on Twitter
I bet you all thought Energy Twitter was all out of takes. But I can't help feel that many economists who are less embedded in wholesale electricity are drawing the wrong conclusions, both from the paper and the debate it spurred. (cc'ing some: @samori8 @KnittelMIT @noahqk)
To me, the key issue is that the estimates derived from the top-down regression suggest a cost of RPS that is roughly 10x what would be suggested by a bottom-up analysis. Three mechanisms (intermittency, transmission, stranded assets) are suggested to explain this.
I think why the economics crowd is missing the point is that a lot of the critiques center on things that, while important, are not core to that analysis.(learning effects, political feasibility, not including co-benefits, etc.)
A big problem with the discussion and interpretation sections of the papers is that a general economics reader can reasonably conclude something like this ...
... while somebody more familiar with wholesale electricity markets will recognize this:
The key observation is that the LBNL numbers include the cost of intermittency. A separate study estimating the cost of transmission (from different authors at LBNL) is also cited.
Of course there are caveats with those studies, but a reasonable estimate is that the cost of the technology, intermittency, and transmission explains about 10% of the result found by the top-down regression. Hence my "hot take" from earlier in the week:
So now the three most likely explanations seem to be: 1) the regression is wrong, 2) the three proposed mechanisms are not exhaustive, and 3) the problem of stranded assets is *gigantic*.
Others have talked about (1) and (2), and I have my questions along those lines. But implicitly the paper believes those have been adequately addressed.
Given this, it seems to me that the interpretation sections of the paper should almost entirely focus on how the cost of stranded assets gets passed through. This could be a fascinating discussion. And I don't think it would invite nearly the same vitriol from Energy Twitter.
And most importantly, it would be a more accurate interpretation of the results and less likely to be misinterpreted by readers. And it has very different implications for what we should expect from future growth of renewables, especially as it becomes more market-driven.
The reaction has at times been unfortunate (and often inaccurate from an economics standpoint). But as a consequence, I fear that the economics crowd has interpreted this as mainly being due to the unpopularity of the paper's conclusion.
I think a lot of the reaction is instead because the discussion around the 10x gap is not remotely plausible. Recognizing this, the people most attuned to the methodology can only give a soft defense of the validity of the approach.
If this gap didn't exist, and if the discussion didn't fall so far short, I'm not convinced anybody would have been so keen to dig into the methodological details, and I'm not convinced there would have been any accusations of ulterior motives.
I learned a lot from reading the paper, and I have mixed feelings about the "wide open peer review" that has been going on this week. But I hope this helps clarify some things and moves the discussion forward.
(Trying to remember other economists I have seen commenting on the paper in hopes that they’ll correct me if I’m off base: @ivanjrudik @bcshaffer @PeteManiloff)
