The original paper, called “A revealed preference approach to valuing non-market recreational fishing losses from the Deepwater Horizon oil spill,” estimates the 2010 explosion of the
BP-owned drilling rig cost the Gulf-Coast recreational saltwater angler fishing industry alone nearly $600 million. But Kenneth Train, an economics professor at the University of California, Berkeley has questioned the methods used — both publicly, in a comment that was published in the Journal of Environmental Management, and privately through personal calls with the authors. The first author says Train asked them to retract the paper; he denies ever making that request. While Train, in his comment, says he doubts the accuracy of the $600 million estimate, he does not provide an alternative number.
Train was hired to review the study by BP, which owned the well that spilled millions of oil barrels into the Gulf.
Calculating the cost of oil spills is controversial. Since the Exxon Valdez spill in 1989, scientists have clashed over the statistical methods that can be appropriately applied to evaluate natural resources. Indeed, that was Train’s main argument – that the authors used the wrong model to estimate the costs. The discussion around the 2014 paper is part of that larger debate, which is why the dialogue between the authors and Train grew heated – even though the paper was not part of the official damage assessment.
The official assessment, which was finalized on March 21, allocates $8.8 billion to projects designed to restore the Gulf of Mexico’s ecosystem. It didn’t even include a specific estimate for the costs to saltwater angler fishing, just presented a more general figure for “Lost Recreational Use” — $693.2 million.
To complete the assessment, federal agencies and states — including the state of Florida — have been working to calculate the financial impact of the spill on gulf-state economies. Collectively, these stakeholders are known as natural resource trustees.
All four authors of the Journal of Environmental Management paper worked on the state of Florida’s official assessment of the BP spill. The first author, Sergio Alvarez, was a graduate research assistant at the University of Florida in 2011 when he did the bulk of work on the assessment, and the other three authors were consultants for the state.
In October 2015, the trustees published a draft of the Plan for Deepwater Horizon Oil Spill Natural Resource Injury Restoration. In an overview of the plan, the trustees do not specifically calculate the impact of the spill on recreational fishing. They do, however, estimate the cost of “Lost Recreational Use.” According to the document, “The Deepwater Horizon spill caused the public to lose more than 16 million user days of boating, fishing, and beach-going experiences. Total recreational use damages due to the spill are estimated at $693.2 million.”
Before publishing his comment to the 2014 paper, Train contacted several of the paper’s co-authors personally to discuss potential errors in their calculations.
“We agreed with some of the points that were made by Kenneth Train,” said John Whitehead, paper co-author and department chair and professor at Appalachian State University’s economics department. “We tried to set the record straight with the corrigendum.” The authors published the corrigendum in March, 2015. In this “addendum,” as it’s called in the note, the authors explain the methods they used to achieve their estimate, such as how they estimated the cost to angler fishers for each trip they had to relocate due to the BP spill, and explore other estimates that rely on different assumptions. The corrigendum also explains how the authors calculated the aggregate loss to all disrupted angler fishers over the course of the spill.
But even after the authors issued the update to the paper, Train personally called the first author to say that, regardless of the methods used to estimate the cost per-trip and the aggregate loss, the underlying formula the authors used was inappropriate. Here, the two sides remember events differently. First author Sergio Alvarez said that Train personally called him, and over the course of that call, requested a retraction. Train told us he did, in fact, call Alvarez, but said he asked for a revised version of the addendum rather than a retraction of the paper.
Ultimately, Train wrote a comment, published online in December in the Journal of Environmental Management. In it, he writes:
None of the estimated losses from the spill, in the original article and Corrigendum, is correctly calculated given their model specification.
Train’s main argument is that the authors didn’t use the proper sample of data to address the question they intended to answer, and that their model “specified an impossible form of time reversal where response to an unanticipated event (the spill) changes behavior before the event.” Ultimately, Train does not say whether the estimate of $585 million is too low or too high, merely that it was calculated incorrectly.
Following Train’s comment, the original authors submitted a reply, also published in December, which notes that Train’s comment does not advance the scientific literature, since they had already addressed his concerns in the corrigendum. Their response also calls out Train’s consulting role with BP:
Furthermore, the Alvarez et al. (2014) study was conceived, designed, and carried out completely independent of the legal process surrounding the Deepwater Horizon spill. It did not rely on funding from any stakeholder group, and its objectives or content were never reviewed by attorneys involved in either side of the legal process.
Despite the controversy, Train told us he does not believe the paper should be retracted at this point. “I think everything’s out there now,” he said. “My purpose was to make sure that these methods weren’t adopted elsewhere.” He noted that this type of back-and-forth between researchers is part of the typical scientific process.
Indeed, this debate about the proper methodology to calculate oil spill damages has been boiling in the literature before this paper. For example, the authors of the Journal of Environmental Management paper used a technique called contingent valuation to estimate the economic worth of non-market resources. But the method is controversial in some circles. In 2012, an economics professor at MIT named Jerry Hausman decries it in a Journal of Economic Perspectives paper titled, “Contingent Valuation: From Dubious to Hopeless.” Hausman, like Train, previously worked as an oil company consultant. He was hired by Exxon after the 1989 Valdez spill.
Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.