To the Editor:
Last month, I published an opinion piece in these pages on how social and career incentives surrounding climate researchers cause a good portion of the full story on the climate problem to be left out of the high-impact literature (“Does High-Profile Climate Science Tell the Full Story?” The Chronicle Review, February 27). My position was based on at least six broad lines of evidence.
1) Noting the vast majority of the climate papers that receive the most attention appeared in high-impact journals and that surveys/studies demonstrate that publishing in these journals is beneficial for career advancement.
2) Noting that the messaging espoused by the leadership of high-profile journals (e.g., officially endorsing policies and politicians) gives a strong indication to researchers that some results are more welcome than others.
3) Showing that, when it comes to climate impacts, the predominant political goals were established prior to most of the research that ostensibly supports those goals, indicating that there has been plenty of ‘policy-based science’ solicited over the years.
4) Documenting and cataloging the patterns that I am concerned about in Google Scholar’s top 50 climate change papers from Nature and Science journals — most notably highlighting that none of the papers focus on the risks of overly restrictive energy policies.
5) My first-hand experience as a researcher in the field which has shown me that research is treated differently depending on how it is framed.
6) My first-hand experience of how you are treated as a researcher when you are perceived as being on the good team (raising concern about climate impacts) compared to when they are perceived as being on the bad team (in my case, raising concern about biases in high-impact climate publishing).
Chris Haufe and Cyrus Taylor recently wrote a letter to the editor in response, “Essay’s Claim That Full Story of Climate Science Isn’t Being Told Doesn’t Withstand Scrutiny,” Letters to the Editors, April 5). Unfortunately, their letter did not even attempt to address most of the thesis summarized above. Instead, they erroneously claimed my assertions were based on only “three lines of evidence,” which they claimed to refute. However, even in the case of these three issues, their supposed refutations fail.
The first claim Haufe and Taylor make is that the paper I published last year in Nature only received considerable attention because I subsequently published an essay critiquing it. This supposedly undermines my claim that publishing in high-impact journals is desirable. Both assertions are false. In my piece, I was referring to the attention that my Nature paper received prior to my essay critiquing it, which was considerable. That attention included domestic coverage from NPR, The Los Angeles Times, The San Francisco Chronicle, and Yahoo News and international coverage in the Japan Times, and The Times of India. Hauf and Taylor make no attempt to refute my demonstration that climate articles with the most fanfare tend to come from high-impact journals, nor do they refute the evidence that published articles and surveys demonstrate the desirability of high-impact papers for career advancement.
Haufe and Taylor then make the claim that “Nature referees explicitly criticize Brown’s article for failing to quantify key aspects other than climate change.” As I thoroughly documented here, nowhere in the peer review process did reviewers challenge the usefulness of focusing solely on the impact of climate change when projecting long-term changes in wildfire behavior.
Concerning the two papers rejected by Nature, Haufe and Taylor make more misleading and false assertions.
First, they imply that a quote from the IPCC summary for policymakers somehow refutes my claim that a certain flawed methodology “has been incorporated into major IPCC reports.” Just one of the papers that uses the flawed methodology was cited five times in the IPCC report in question, so clearly, my claim that “this methodology has proliferated; it has been incorporated into major IPCC reports” is on solid footing.
Then, Haufe and Taylor misquote me, implying that I said that I had made a “major discovery” (their quotation). By moving the adjective “major” to the noun “discovery” rather than “flaw,” Haufe and Taylor seek to paint me as a delusional narcissist. The mischaracterization of someone’s position is one thing, but intentionally misquoting someone is, according to various university policies, “extremely poor scholarship and could constitute academic misconduct.”
On the second paper rejected by Nature, Haufe and Taylor claim that my work supports the predominant narrative on climate change. This, too, is very misleading. The topic of the paper in question is whether adding costs to a previously published analysis that considered only benefits supports or undermines the conclusion of the original benefit-only analysis. I found that adding costs reverses the conclusions of the benefit-only analysis. Here is how it was put in my paper:
“Under a discount rate of 3%/year and a time horizon through 2100 (the central values used in Burke at al. [39]), … we calculate …a ~40 trillion US$ loss from limiting global warming to 1.5 °C relative to 2.0 °C under Burke-DICE (Fig 10a) which is in contrast to the corresponding calculation in Burke at al. [39] which found a central estimate of a ~40 trillion US$ benefit from limiting global warming to 1.5 °C relative to 2.0 °C but did not consider mitigation costs.”
The observation that other research had done similar things (with the implication that the bench rejection was warranted) further makes my point: My research was published alongside other papers in lower-tier journals, where the results were disseminated far less widely than in the original paper.
The final claim by Haufe and Taylor is that there is no reason to believe that academics are self-censoring because a survey I cite in support of the claim had a low response rate. The survey in question was developed by a survey and market research firm, it was thoroughly documented, and it has been cited in the peer-reviewed literature. It had 1,491 respondents, and its response rate was partially a product of a very high number of academics included in the initial survey database. Haufe and Taylor offer no evidence that the sample was biased in one direction or another, and they offer no alternative survey that undermines the broader claim that self-censorship is a problem in academia.
Overall, Haufe and Taylor avoid addressing the main pillars of my thesis and instead choose to pull out three claims that they apparently thought would be the easiest for them to refute. Even using this avoidance strategy, they failed. All their letter does is demonstrate one of the main points of my essay. If you go against the predominant narrative on climate change, you will be seen as being on the bad team and will be attached personally as a nefarious character. This is unfortunate and counter to the substantive debates that we should be having, which will elucidate the full picture on the climate-change problem.
Patrick T. Brown
Lecturer, Energy Policy and Climate Program
Johns Hopkins University
Baltimore